Biohydrogen production from agricultural and livestock residues within an integrated bioenergy concept

Tra i possibili approcci biologici per sfruttare la biomassa per la produzione di idrogeno, questa tesi ha focalizzato il suo interesse sulla fermentazione anaerobica effettuata alk buio, che può garantire contemporaneamente la produzione di un elevato valore di H2 e il trattamento dei rifiuti, trasformando in questo modo così trasformato una fonte di emissioni inquinanti ambientali e gas serra in una risorsa preziosa.

Scarica il PDF Scarica il PDF
Aggiungi ai preferiti Aggiungi ai preferiti

Articoli tecnico scientifici o articoli contenenti case history
Tesi di Dottorato, Università degli Studi di Milano, Anno Accademico 2010-2011

da Alessia De Giosa
VerticaleSegui aziendaSegui


Parole chiave: 

Estratto del testo
Doctoral School in "Innovazione tecnologica per le scienze agro-alimentari e ambientali", XXIV cycle Faculty of Agriculture Department of Agricultural Engineering (AGR/09) Biohydrogen production from agricultural and livestock residues within an integrated bioenergy concept Ph.D. Thesis Alberto Tenca (Matr. Number R08109) Tutor: Prof. Luigi M. Bodria Doctoral School Coordinator: Prof. Roberto Pretolani Academic Year 2010/2011 i Abstract

Concerns about energy security, fossil fuel prices, and climate change
issues, are leading to increasing renewable energy demand. Hydrogen is
considered as one of the main possible energy carriers in future, due to its
environmental (it can be converted to energy with the solely emission of
water) and energetic (energy content of 120 MJ/kg, three times higher of the
gasoline content of 44 MJ/kg) unique properties. If hydrogen is currently being produced mainly by fossil sources, its
production from renewable sources answers to the demand of more
environment-friendly exploiting alternatives, possibly leading to a
renewable-based hydrogen economy. Biomasses are an important renewable
source ranging from energy-dedicated crops to livestock waste effluents,
agro-industrial wastewaters, food-processing industry residues and organic
fractions of the municipal solid waste (OFMSW). Thus the agricultural
sector may acquire a renewed importance in the mid-term as a producer of
energy sources for renewable biohydrogen production. Among the biological ways to exploit biomasses for hydrogen production,
this thesis focused its interest on anaerobic dark fermentation, which can
simultaneously guarantee the production of an high-value product (H2) at
high evolution rates and the treatment of wastes, thus transformed from an
environmental pollution and greenhouse gases emissions source into a
valuable resource. If on the one hand this process has lots in common with
anaerobic digestion, which already is a well-established technology for
treating different biomass types in real-scale plants, on the other hand it is a
relatively new approach, which needs to be further studied for improving its
performances and being concretely applicable. As a matter of fact, the main disadvantage of dark fermentation is its
relatively low yield, compared to other bio-hydrogen production methods,
which typically are between 2.4 and 3 mol H2/mol glucose. This represents
just the 20-25% of the 12 mol of H2 theoretically obtainable by glucose
fermentation. Therefore, generally two different (but not mutually
exclusive) options could be chosen for improving the process and making it
ready for full-scale applications: the optimization of the biological,
biochemical, chemical-physical operative parameters that regulate process;
or the coupling of this bioprocess with other technologies capable of
exploiting the organic matter not fully used by the dark fermentative
approach. For example, Microbial Electrolysis Cells (MECs) are able to
biologically oxidize the organic matter (from simple substrates like volatile ii fatty acids, lactic acid, glucose, cellulose, to actual wastewaters) releasing
electrons from an anode to a cathode where potentially pure hydrogen can
be formed from protons in the water. Papers I and II basically belong to the first strategy. In Paper I indeed, two
waste biomasses were co-digested: in consideration that in the Po Valley
area (Italy) swine manures (SM) are yearly produced at high waste density
levels and could be a cause of environmental problems, this waste was used
as a co-substrate for biohydrogen production by the thermophilic
fermentation of easily degradable and carbohydrate-rich materials, such as
fruit and vegetable market waste (FVMW). Biohydrogen production rates
and process stability were thus simultaneously maximized, thanks to the
endogenous buffer capacity of manure, through the combination of a
suitable composition (as FVMW/SM) of the feeding material and the
hydraulic retention time (HRT) of the process. Thus, livestock manure
represented not only a renewable source for supplying the production of
biohydrogen, but also a source of alkali to be used for avoiding the addition
of exogenous chemicals (alkali) to maintain the pH, and so the metabolic
pathways and bacterial communities, into an optimal domain for
biohydrogen production. To further study and optimize the bio-H2 production in laboratory-scale
processes, but also to find applicable tools for favoring dark fermentation
application in full-scale biogas plants, Paper II succeeded in obtaining
mixed microbial cultures from natural sources (soil-inocula and
anaerobically digested materials) which reached high hydrogen yields with
glucose and were used to explore the potential of bio-hydrogen production
from four organic substrates of possible interest for full-scale plants (market
bio-wastes, maize silage, swine manure, OFMSW). In direct prosecution of
the positive co-digestion results shown in Paper I and looking for future
transfer of this bioprocess technical solutions to full-scale systems, Paper II
used the enriched mixed microflora for evaluating the co-fermentation of a
mixture of OFMSW and swine manure in a lab-scale continuously-fed
CSTR (continuously stirred tank reactor) digester. Despite the good results
obtained, our study suggested that further efforts are needed for future
applications of effective biohydrogen fermentation in full-scale plants. Paper III and IV are more focused on the second scientific strategy. Paper
III joins the interest toward implementation of bio-H2 in full-scale plants
and the strategy of improving the overall recovery of the energy contained
in the biomass associating hydrogen production to other bioprocesses. Many
authors report that the two-stage anaerobic digestion (AD) process, if iii compared to traditional and extensively real-scale applied single-stage AD,
has also other advantages, such as differentiating the biofuel production
(bio-hydrogen and bio-methane), potentially reducing the plant dimensions
and costs, improving the overall biogas production yields and allowing
higher CH4 concentrations in the biogas produced in the second stage, thus
decreasing the biogas purification costs. Therefore, a two-stage laboratory-
scale CSTR anaerobic digester, fed with a mixture of agricultural and
livestock residues, was monitored for a long run (approximately 700 hours)
and compared to a similar one-stage reactor. This study obtained a good
hydrogen yield per kg of biomass treated and partially confirmed the
advantages previously illustrated, even if it reached almost the same overall
energy recovery of the single stage process. Aiming at other possible biological strategies to improve the energy and
hydrogen recovery efficiency with the use of effluents from a first dark
fermentative stage, a relatively new electrohydrogenesis device (MEC) was
studied. Paper IV explores the rate and the yield of biogas (a mixture of H2,
CH4 and CO2) produced by MEC exploiting an actual industrial wastewater
with high methanol content, a compound never before reported to be used in
a MEC device. The energetic recovery and treatment performance of the
process was evaluated and also compared with a simulation of anaerobic
digestion of the same wastewater, revealing the economical competitiveness
of the MEC technology with the AD process. This leads to future research
perspectives aiming to realize a laboratory-scale two-stage reactor with a
MEC using the volatile-rich effluent of a first dark fermentative stage.
Keywords: Bio-hydrogen, biomasses, MEC, biorefinery
Author's address: Alberto Tenca, Università degli Studi di Milano, Facoltà
di Agraria, Department of Agricultural Engineering. Via Celoria 2, 20133,
Milano (MI), Italy E-mail: iv Table of Contents List of International Refereed Papers and Congress
1. Hydrogen energy infrastructure and biological production of
hydrogen 1
1.1 Energy and environmental issues: the factors favoring the
hydrogen energy 1
1.2 Technology for hydrogen energy use 6 1.2.1 Internal combustion engines 7
1.2.2 Fuel cells 9 1.3 Biohydrogen 15 1.3.1 Biological processes for hydrogen production 15
1.3.2 Hydrogen production by Dark Fermentation 23 1.4 Key parameters regulating the dark fermentative hydrogen
production and their optimization 29 2. Biomasses for hydrogen production via Dark Fermentation 53
2.1 Agricultural and livestock biomasses 57
2.2 Organic residues and wastewaters 63 3. Biohydrogen production via bioelectrochemical systems: the
MEC application 77
3.1 General operating principles 77
3.2 Main process regulating factors and their optimization 90
3.3 MEC application and scalability 108 4. Biohydrogen and integrated energy production 113
4.1 Dark fermentation + Photo fermentation 115
4.2 Dark fermentation + Anaerobic Digestion 123
4.3 Dark fermentation + Bioelectrochemical systems 137
4.4 Other approaches 143 5. Conclusions and final remarks 147 6. References 151 Acknowledgements 179 Appendix A - International Refereed Papers Appendix B - Congress Communications v List of International Refereed Papers and
Congress Communications
International Refereed Papers This thesis is based on the work contained in the following papers,
referred to by their Roman numerals in the text: I. Tenca A., Schievano A., Perazzolo F., Adani F., Oberti R.
2011. Biohydrogen from thermophilic co-fermentation of
swine manure with fruit and vegetable waste: Maximizing
stable production without pH control. Bioresource
Technology 102(18): 8582-8588. II. Tenca A., Schievano A., Lonati S., Malagutti L., Oberti R.,
Adani F. 2011. Looking for practical tools to achieve next-
future applicability of dark fermentation to produce bio-
hydrogen from organic materials in Continuously Stirred
Tank Reactors. Bioresource Technology 102(17): 7910-
7916. III. Tenca A., Schievano A., Scaglia B., Oberti R., Adani F.
Two-stage vs. single-stage fermentation process: comparison
of energetic performances and chemical characterization. (To
be submitted to International Journal of Hydrogen Energy) IV. Tenca A., Cusick R., Logan B.E. MEC and Anaerobic
Digestion performance comparison with complex industrial
wastewater. (To be submitted to Biotechnology and
Bioengineering) Papers I-II are reproduced with the permission of the publishers. Congress Communications I. Schievano A., Consonni E., Tenca A., Oberti R., Adani F.
2009. Bio-hydrogen production from bio-waste: ready for
full-scale applications' (Ecomondo 2009 Proceedings) II. Schievano A., Tenca A., Oberti R., Adani F. 2009. An
operational strategy to produce biohydrogen: the use of
digestate for process control (Hypothesis VIII Proceedings)

1 Chapter 1 Hydrogen energy infrastructure and biological production of hydrogen

1.1 Energy and environmental issues: the factors
favoring the hydrogen energy
The increasing global demand for finite oil and natural gas reserves
together with the national energy security need, are driving the
scientific community into the search for alternatives to fossil fuels.
During the 20th century, the industrial and societal development was
driven by and depended on the fossil resources and related
technologies, but with the rapidly expanding world population and
the increase of prosperity in less developed countries, it is expected
that the world consumption of energy and resources will increase
with a minimum factor of 3 by the year 2050 (Boeriu et al., 2005).
Stating that fossil fuel reserves are a diminishing raw material and
that the fossil resources are non-uniformly distributed, Rifkin argues 2 that global oil production will fail to meet this increasing demand for
energy in the next 10-20 years (Rifkin, 2002). Even if oil and fossil
fuel production wouldn't reach an international crisis, many nations
of the world are already and increasingly adopting alternatives to
fossil fuels, taking increased carbon dioxide issues seriously and
implementing measures to reduce greenhouse gas emissions. Indeed,
fossil fuels possess undoubtedly very useful properties not shared by
non-conventional energy sources, but in addition to being not
renewable, they are not environmentally friendly. The pollutants
emitted by fossil energy systems (e.g. CO, CO2, CnHm, SOx, NOx,
radioactivity, heavy metals, ashes, etc.) are greater and more
damaging than those that might be produced by a renewable based
energy system (Momirlan and Veziroglu, 2002). Although there is
disagreement about the specific effects of greenhouses gases on
global temperature, it is a fact that the concentration of atmospheric
CO2 have increased about 30% in the past 150 years and that the
fossil resources are the major source of this additional CO2
(Chamberlain et al., 1982).
Therefore the world is striving for the search of a newer, cleaner and
renewable energy source that can be easily transported, used in
vehicles and that in principle will never run out. Hydraulic, solar,
wind, tidal, geothermal energy and also energy from renewable raw
materials such as biomass, are all expected to significantly contribute
in the mid- to long-term to the energy demand(Soetaert and
Vandamme, 2005). Within this scenario, hydrogen has demonstrated a big potential to
both reduce the dependence on oil and to lower the greenhouses gas
emissions. Hydrogen was reported to be the "fuel of future" since the
energy crisis of the '70s, and its popularity as a fuel source has
always followed the crisis coming from the extensive use of
nonrenewable fuels. As a matter of fact, after great efforts put into
hydrogen research during the early '70s, like other alternative energy 3 technologies hydrogen lost its importance with the drop of the oil
price at the end of the crisis. But again, with the emerging concerns
about the greenhouse effect in the '90s, a new crisis reignited
mainstream interest in post-fossil fuel hydrogen-based economy
(Benemann, 1996).
Although H2 is an energy carrier rather than a traditional fuel found
in and harvested from nature, it joins the advantages to be a mobile
source of energy and to be environmentally friendly (especially if
manufactured by capturing energy from renewable starting materials)
(Logan, 2004; Veziroglu and Barbir, 1992). Nowadays, the "Global
Hydrogen Vision" envisions hydrogen as a flexible, safe, affordable,
domestic energy resource to be used in all sectors of the economy, in
a wide variety of applications, including fuel for automobiles and in
all regions of the world. Since it can be used either as the fuel for
direct combustion in an internal combustion engine or as the fuel for
a fuel cell, hydrogen could join electricity as a primary energy carrier
and provide the foundation for a globally sustainable energy system
(Kotay and Das, 2008).
H2 has the highest energy content per unit weight among the known
gaseous fuels (143 GJ ton ''1, which is 2.75 times greater than hydrocarbon fuels) and at the same time it is a carbon-free fuel which
oxidizes to water as a combustion product (Armor, 2005). The only
pollutant that eventually can arise is the nitrogen oxide that comes
from the combination of the oxygen and nitrogen in the air. This
occurs only if the hydrogen is not recombined with pure oxygen, that
is using air as an oxidant, and with high reaction temperatures
(Momirlan and Veziroglu, 2002). Therefore hydrogen energy system
could be presented as a carbon-free, natural close cycle where the
water generated from its combustion becomes, together with
renewable primary energy for splitting it, a source of clean and
abundant energy (Fig. 1.1). 4 Fig. 1.1 Hydrogen energy system (Source: International Association of Hydrogen Energy,
USA; Courtesy of Nath and Das, 2003). However, independently of the source of hydrogen, there are many
logistical and market challenges that must be overcome before a
"hydrogen economy" can become a reality. Among them: - Production technologies
Although hydrogen is the most abundant element in the universe, it
must be produced from other hydrogen-containing compounds, such
as fossil fuels, biomass, or water, and each method of production
requires a source of energy, i.e. thermal (heat), electrolytic
(electricity), or photolytic (light) energy (Kotay and Das, 2008).
Currently, the hydrogen production depends mostly on natural gas
and therefore is highly energy intensive and not environmental-
friendly. Nearly 90% of hydrogen is obtained by steam reforming of
naphtha or natural gas and by gasification of coal or by electrolysis
of water (Nath and Das, 2003). For sure these strategies, especially in
the case of steam reforming of methane (SRM) or other
hydrocarbons (SRH) and of non-catalytic partial oxidation of fossil
fuels (POX), have reached maturity for commercial exploitation but
are really energy-intensive processes, requiring high temperatures 5 (>850 °C) (Momirlan and Veziroglu, 2002). In consideration of both
energy security and environmental issues related to fossil-fuel
reserves exploitation, production of hydrogen by renewable sources
(such as biomasses) seems to be imperative (in-depth examination in
chapter 1.3.1) - Hydrogen market and the required infrastructures
If nowadays technologies challenges are firstly addressed to lowering
the cost of hydrogen production, also hydrogen delivery, storage and
conversion are required to be elaborated. Indeed, contrarily to what is
commonly thought, demand of hydrogen is not limited to its
utilization as a source of energy. Hydrogen gas is a widely used
feedstock for the production of chemicals and electronic devices, the
hydrogenation of fats and oils in food industry, the steel processing
and the desulfurization and re-formulation of gasoline in refineries.
H2 can acts to saturate compounds, crack hydrocarbons or remove
sulphur, nitrogen compounds and traces of oxygen (thus preventing
the oxidative corrosion thanks to its oxygen scavenging property).
Such an amount of end-use applications makes a 50 million tons of
hydrogen annually trading worldwide and is pushing the contribution
of hydrogen to total energy market up to 8-10% by 2025 (United
States National Hydrogen Program esteem) (Kapdan and Kargi,
Among the technologies for hydrogen storage and transport, there are
four technologies available today to store hydrogen aboard vehicles
(Hynek et al., 1994):
' liquefied hydrogen (used by NASA and considered for airliners);
' metal hydrides (used for example by Mazda and by Daimler- Benz in passenger cars); ' compressed hydrogen gas (used on urban transport bus built by Ballard); ' carbon sorption (yet to be used on vehicles). 6 Moreover in recent years, new technologies have been developed as
reported by an interesting state of the art review by Momirlan and
Veziroglu (2002):
' new carbon variants: graphitic nanofibres and (carbon) nanotubes as hydrogen storage materials; ' new hydrogen storage alloys: La/Ce mixture, Mg or Mg 2Ni (nanocrystalline) system, nanocrystalline Zr-based AB2 alloys; ' liquid-film type catalytic decalin dehydrogeno-aromatization for mobile storage of hydrogen; ' innovative hydrogen densification in a two stage metal hydride system. - Upgrading requirements
Especially dealing with the biological processes, hydrogen produced
is mostly found together with other different gaseous impurities like
O2, CO, CO2, CH4 and some amount of moisture. The presence of
these gases lowers the heating value of hydrogen, in addition to
posing some problems in efficient burning of fuels (see section 1.2).
Therefore, as proposed by Nath and Das (2003) the following
upgrading procedures may be used:
- CO2 acts as a fire extinguisher and it is sparingly soluble in water: it
can be separated by scrubbers or absorbed by a fifty per cent (w/v)
KOH solution or monoethanolamine;
- O2 in the gas may cause a fire hazard: it can be absorbed by an
alkaline pyrogallol solution;
- Moisture in the gas mixture reduce the heating value of hydrogen:
passing the gas mixture either through a dryer or a chilling unit (by
condensing vapor in the form of water) is the more practical solution.
1.2 Technology for hydrogen energy use The widespread use of hydrogen as an energy carrier will depend
significantly on the availability of efficient, clean and economic 7 techniques for its utilization and conversion to electricity/heat.
Among the main technologies for hydrogen utilization as a fuel, it is
possible to distinguish internal combustion engines (ICE) from the
fuel cell (FC) technologies. If the first is widely known and used,
regardless of hydrogen use, the latter is now emerging as a leading
technology able to replace the more polluting ICEs both in vehicle
and stationary distributed energy applications and it will need deeper

1.2.1 Internal Combustion Engines An internal combustion engine is an engine in which the combustion
of a fuel (from fossil fuels - petroleum and carbon - to biofuels,
vegoils and hydrogen) occurs with an oxidizer (usually air) in a
combustion chamber. In an ICE, the expansion of the high-
temperature and high-pressure gases produced by combustion applies
direct force to some component of the engine (such as pistons,
turbine blades, nozzle), thus generating useful mechanical energy.
A large number of different designs for ICEs have been developed
and built, each one with different points of strength and weakness,
and even if there are many stationary applications, the real strength
of internal combustion engines is in mobile/automotive applications.
Gasoline, Diesel, Wankel engines and open gas turbines are all
examples of internal combustion engines.
Hydrogen could eventually replace conventional fossil fuels in
traditional ICE. Apart for its heating volume, other key properties of
hydrogen that are relevant to its employment as an engine fuel
(especially for transport processes) are its remarkably high values of
thermal conductivity and diffusion coefficient, in comparison to
those of gasoline fuel (hydrogen: thermal conductivity at 300 K:
182.0 mW/m K, diffusion coefficient into air at NTP: 0.61 cm 2/s; gasoline: 11.2 mW/m K and 0.05 cm 2/s, respectively) (Yamin, 2006). 8 So according to Yamin et al. (2000), hydrogen employment in ICEs
has the following advantages over gasoline:
1. Reduced deposits due to more homogeneous mixture formation.
2. Reduced engine oil dilution and increased oil life.
3. Reduced engine wear, hence increased engine life.
4. Higher compression ratios can be used which may solve the
problem of reduced power output due to reduction in volumetric
5. Elimination of emissions of CO and HC.
6. Increased fuel economy due to possible operation at leaner
Moreover, hydrogen-fueled internal combustion engines, which
already have typical efficiency of 30% (and maximum of 45%) could
help in quickly filling the gap between the existing hydrocarbon-
fueled ICE and future hydrogen fuel-cell technology. On the other side, hydrogen use in ICE has some problems.
The first is its introduction into the engine, because of hydrogen wide
flammability range, low energy density by volume and high flame
speed (U.S. D.o.E., 2001).
The second is that its use in ICE (as in catalytic burners) may
produce nitrogen oxides (NOx) emissions (Zurawski et al., 2005)
due to the high temperatures generated within the combustion
chamber. Indeed, H2 produces only water if combusted with oxygen,
following the reaction
H2 + 1/2 O2 '' H2O but also NOx are produced if it is combusted with air as follows
H2 + O2 + N2 '' H2O + N2 + NOx So, depending on the operating strategy used (i.e. the air/fuel ratio,
the engine speed and compression ratio, the ignition timing, etc.), an
hydrogen engine can produce from almost zero emissions to high
NOx and even carbon monoxide emissions. 9 Alternative strategies to the use of pure hydrogen in ICE are also
possible. For example H2 can be used as supplementary fuel to
enhance the combustion properties of natural gas (a commercially
available mixture known as Hythane - i.e. 20% hydrogen and 80%
natural gas -), without requiring modifications to a natural gas
engine: both gases can be stored in the same tank, and emissions
reduction by more than 20% are possible (U.S. D.o.E., 2001).
Indeed, the hydrogen presence allows lower combustion
temperatures, simultaneously leading to lower NOx emissions and
enhancing the combustion process with recovery of the power and
energy consumption penalties associated with natural gas.

1.2.2 Fuel cells A fuel cell (FC) is an electrochemical device that converts the
chemical energy of gaseous (e.g. hydrogen, natural gas, and biomass
derived gas) or solid (mainly coal) fuels directly into electrical
energy (and heat) via an electrochemical process with high
conversion efficiency (Edwards et al., 2007). Differently from a
battery, which is an energy storage device, a fuel cell is an energy
conversion device that can produce electricity as long as the fuel and
the oxidant are provided to the electrodes and that doesn't run out
(unless undesired events, such as component degradation/corrosion).
The concept of the FC was firstly developed by W.R. Grove in 1839,
who successfully produced an electric current and water by
combining hydrogen gas with oxygen.
The basic FC consists of an electrolyte layer in contact with a porous
anode and cathode on either side. A fuel, such as hydrogen, is fed to
the anode where negatively charged electrons are catalytically
separated from positively charged ions. From the anode and through
an electrolyte a ionic current flows toward the cathode where protons
combine with oxygen or air, resulting in water production. 10 Simultaneously, the excess electrons flows through an external
electric circuit, generating an electric current (Edwards et al., 2007).
Depending on the FC type and its conversion efficiency, different
amount of the energy developed by the hydrogen oxidation reaction
(enthalpy of 285.8 kJ/mol under standard conditions) are converted
into electricity, the remaining into heat. Because fuel cells are not
subject to the intrinsic limitations of the Carnot cycle, they convert
fuel into electricity at more than double the efficiency of internal
combustion engines. In transportation, hydrogen fuel cell engines
operate at an efficiency of up to 65%, compared to 25-30% for
current oil-fueled car engines. As the reaction at the basis of the FC
is an exothermic one, when heat generated in fuel cells is also used in
combined heat and power (CHP) systems, an extremely high overall
efficiency of 85% or more can be achieved (Dutton, 2002).
Moreover, unlike internal combustion engines, fuel cells demonstrate
high efficiency across most of their output power range. This
scalability makes them ideal for a variety of applications from
mobile phone batteries through vehicle applications to large-scale
centralized or decentralized stationary power generation.
Even from the environmental point of view the fuel cells seem to be
better than ICEs because, operating at lower temperature, only water
and virtually no pollutant (NOx) are released. If the hydrogen fuel
could be sourced from renewable routes and not be hydrocarbon-
based, real zero emission will be reached by hydrogen-powered fuel
cell vehicles (HFCV). Even if compared with other zero-emission
vehicles (as battery-driven electric cars) HFCV may relay on a
technology (FC) characterized by a much longer operational lifetime
and providing the same high specific energy as traditional
combustion engines (Winter and Brodd, 2004). Several types of fuel cells operating on a variety of fuels and suitable
for different energy applications have been developed, so they are 11 generally categorized by low (LT) or high (HT) operation
Alternatively, if classified according to the material used as the
electrolyte we can distinguish:
' alkaline fuel cells (AFC) (both LT and HT);
' phosphoric acid fuel cell (PAFC) (LT);
' proton exchange membrane fuel cell (PEMFC) (LT);
' molten carbonate fuel cell (MCFC) (HT);
' solid oxide fuel cell (SOFC) (HT). Fig. 1.2 Schematic of a PEM fuel cell (left) and a SOFC fuel cell (right). - Low-temperature fuel cells
The most interesting technology among the low-temperature FCs is
the PEMFC (Figure 1.2), which means Polymer Electrolyte
Membrane or Proton Exchange Membrane Fuel Cell.
Its particular architecture uses a thin solid polymer as an electrolyte
(typically Nafion kept moist using liquid water) and porous carbon
electrodes containing a platinum catalyst. Typically they are fueled
with high purity hydrogen and operate at relatively low temperatures, 12 around 80 °C, which positively affects the components used,
resulting in better durability.
In details in a PEMFC:
1. Hydrogen gas is pressurized and channeled through flow field
plates and throughout the anode gas diffusion electrode.
2. The catalyst layer (usually platinum powder mounted on very thin
carbon paper) between the anode and the PEM electrolyte
accelerates the separation of the hydrogen gas into two negatively
charged free electrons and two positively charged hydrogen proton
3. The positively charged hydrogen protons travel through the moist
PEM membrane to the cathode while the free electrons travel
through the external load to the cathode.
5. Oxygen gas (or normal air) is pressurized and channeled
throughout the cathode gas diffusion electrode.
6. The catalyst layer between the cathode and the PEM accelerates
the separation of the oxygen gas into two oxygen molecules.
7. Hydrogen protons, free electrons and oxygen molecules combine
inside the cathode to form a water molecule (H2O) while releasing
heat (then used outside of the FC or just exhausted). A typical PEM running at 0.6 to 0.8 volts has an efficiency
(conversion of hydrogen into electricity) of about 50 ± 10%, which is
much better than 20% efficiency of common gasoline engines.
PEMFC stacks (a set of connected cells) reach high-power density
and have low weight and volume, compared with other FC, therefore
they are particularly suitable for the scalability of this technology.
Indeed, PEM fuel cells are very close to mass production and are
primarily used for automotive applications (50 to 125 kW), because
of their fast warm-up/start-up time, low sensitivity to orientation, and
favorable power-to-weight ratio (Figure 1.3). However they can be
found also in some stationary applications, from homes (1-5 kW) to
electrical generation plants (up to 250 MW or more). 13 The system cost is one of the disadvantages of the PEMFC, due
mainly to the noble-metal used as catalyst (platinum). Secondly, they
show sensitivity to impurities in the fuel source, especially carbon
monoxide (CO): even tens of parts per million of CO can poison a
pure platinum catalyst, therefore the addition of a pre-reactor to
reduce CO in the fuel gas is necessary (if H2 is derived from an
alcohol or hydrocarbon fuel), together with the development of more
resistant alternatives, such as platinum/ruthenium combination.
Another obstacle, specific for their use in vehicles, is the pure-
hydrogen storage on-board in pressurized tanks. Due to the low-
energy density of hydrogen, it is difficult to store enough hydrogen
on-board to allow vehicles to travel the same distance as gasoline-
powered vehicles before refueling, typically 300-400 miles. A
solution could be the use of an on-board reformer, which convert
higher-density liquid fuels (such as methanol, ethanol, natural gas,
liquefied petroleum gas, and gasoline) to hydrogen. Nevertheless this
would increase costs and maintenance and release carbon dioxide. Fig. 1.3 A fuel cell car powered by hydrogen and sold by Hyundai®. - High-temperature fuel cells
Among the fuel cells working at high temperature, the solid oxide
fuel cells (SOFCs) (Figure 1.2) are a particularly promising 14 technology. Thanks to higher electric efficiency and high
temperature of co-generated heat, to the possibility of realizing
generators of good size (up to some MWe), to the capability of
operating with different fuels (hydrogen, methane, bio-fuels, etc.)
and to high tolerance to gas impurities, SOFCs are better than other
FCs (both LT and HT) to adapt to stationary co-generation
SOFCs use a hard, non-porous ceramic compound as electrolyte and
differently from PEMFC the electrolyte conduct negative oxygen
ions from the cathode to the anode, where they can be
electrochemically oxidized with hydrogen or other fuels. The solid
nature of the electrolyte allows SOFC to be constructed in different
configurations than the typical plate shape of other FCs (Figure 1.4).
SOFCs are expected to be around 50%-60% efficient at converting
fuel to electricity and with the exploitation of the co-generated waste
heat they could reach overall efficiencies of 80%-85%.
Solid oxide fuel cells operate at very high temperatures, around
1,000 °C, which allows to remove the need of precious-metal
catalyst, thereby reducing cost. It also allows SOFCs to reform fuels
internally, thus enabling the use of a variety of fuels and reducing the
cost associated with addition of a reformer to the system. However,
high-temperature operation has also disadvantages: it results in a
slow start-up and requires significant thermal shielding to retain heat
and protect personnel, which may be acceptable for utility
applications but not for transportation and small portable applications
(U.S. D.o.E., 2001). The high operating temperatures also deeply
affects the durability requirements on materials, and so low-cost
materials with good durability at high operating temperatures are
needed. Simultaneously, researchers are studying an alternative
strategy: SOFCs operating at lower temperatures would have fewer
durability problems and lower costs for materials, even if they would
probably produce less electrical power. 15 Another big advantage of the SOFCs is their resistance-to-
contaminants: they can tolerate several orders of magnitude more
sulfur than other cell types and, in addition, they are not reported to
be poisoned by carbon monoxide (CO), which can even be used as
fuel. This property makes SOFCs the preferred FC technology to use
gases made from coal or renewable sources like biomasses. Fig. 1.4 Bloom Boxes® by Bloom Energy Corporation of Sunnyvale, California, USA. On-site
"energy servers" using SOFCs for distributed power, each one producing 100 kW of electricity. 1.3 Bio-Hydrogen 1.3.1 Biological processes for hydrogen production Biological production of hydrogen is known since the late 1800,
when basic research established that algae and bacteria could
produce hydrogen (Jackson and Ellms, 1896). With the oil crisis of
the early '70s, the National Science Foundation (NSF, Washington,
DC, USA) sponsored several meetings on biological hydrogen
production, which firstly roused the interest about the matter,
especially focusing on photosynthetic processes (Benemann, 1996).
With the renewal of interest in the '90s about renewable energy
sources, biological hydrogen production became a focus of 16 additional governmental support, particularly in Germany, the US
and Japan, with minor efforts in other countries. This led to an
enlargement in the biohydrogen production scenario, including also
strategies different than photosynthetic/photolytic processes (Kotay
and Das, 2008).
In general biological hydrogen production processes have some
advantages over their chemical or electrochemical counterparts. They
are catalyzed by microorganisms in an aqueous environment at
ambient temperature and pressure and are well suited for
decentralized energy production in small-scale installations in
locations where biomass or wastes are available, thus avoiding
energy expenditure and costs for transport.
From a thermodynamic perspective, as the organic substrates
dissolved and diluted in wastewater are in a high entropy state, it is
somewhat difficult to obtain their combustion enthalpy by
mechanical means (Kotay and Das, 2008). On the contrary
microorganisms can naturally recover and concentrate the energy
from high water content organic resources, such as industrial
wastewater and sludge in a usable form. Thus, biohydrogen
production is an entropy reducing process, that could not be realized
by mechanical or chemical systems (Nandi and Sengupta, 1998). Broadly speaking, the variety of biological processes for hydrogen
production can be split in light-dependent or light-independent
processes. Light mediated processes include direct or indirect
biophotolysis and photosynthetic strategies (photofermentation and
microbial water shift), whereas dark fermentation is the major light
independent process. Figure 1.5 describes from a biochemical point
of view these different bio-hydrogen production approaches and
additional information about each of them will follow.
17 Fig. 1.5 Biological hydrogen production approaches. It must be said that it is too early to predict which of them will be
ultimately successful, or how they would appear in case of large-
scale production processes or small scale conversion devices,
because their practical development still requires scientific and H2O O2 Photosystems Ferrodoxin Hydrogenase H2 1) Direct biophotolysis 2) Heterocystous nitrogen-fixing cyanobacteria 3) Indirect biophotolysis: Non-heterocystous nitrogen-fixing cyanobacteria H2O O2 Photosystems Ferrodoxin Nitrogenase H2 CO2 CO2 Recycle [CH2O]2 [CH2O]2 Vegetative cell Heterocyst NADPH H2O O2 Photosystems Ferrodoxin Nitrogenase or
Hydrogenase H2 CO2 CO2 Recycle [CH2O]2 [CH2O]2 1° stage
(photosynthesis) NADPH 2° stage
(H2 production) Photosystem I ATP 12 H2O 12 H2 + 6 O2 Ex: 4) Photo fermentation: Photosynthetic bacteria 5) Microbial shift reaction: Photosynthetic bacteria 6) Dark fermentation CO + H2O H2 + CO2 [CH2O]2 Ferrodoxin Hydrogenase H2 [CH2O]2 Ferrodoxin Nitrogenase H2 Bacterial
photosystem ATP ATP NADPH C6H12O6 + 6 H2O 12 H2 + 6 CO2 Ex: 18 technological advances, and middle- to long-term applied R&D
(Benemann, 1996).
For example, the group of bio-technologies based on photosynthetic
systems refers to a theoretically perfect process, directly
transforming solar energy into hydrogen by photosynthetic bacteria.
For real, due to the low utilization efficiency of light and difficulties
in designing light reactor, this strategy is hard to be applied in
practice (Liu et al., 2008). Indeed, the reducing power generated by
photosynthesis must be produced as close as possible to the maximal
possible solar conversion efficiencies (about 10%) and then
efficiently transferred to hydrogenase enzyme. On the contrary,
photosynthetic organisms like higher plants currently capture only 3-
4% of sunlight's available energy at most (Benemann, 1996). 1 - Direct biophotolysis
The (direct) splitting of water to generate hydrogen by solar radiation
is a process achieved either in photochemical cells where, for
example, TiO2 is illuminated as the catalyst, or by applying
photovoltaics, which indirectly utilize solar radiation for the
electrolysis of water into H2 and O2 (Kotay and Das, 2008).
A biological alternative to this process is the direct biophotolysis,
which involves light-driven decomposition of water through micro-
algae or cyanobacteria (Benemann, 1996). Green algae are able to
evolve hydrogen by means of a reversible hydrogenase which
receives the reductants generated by photosynthesis from a reduced
ferrodoxin. On one hand, under laboratory condition at low light
intensities, it has been demonstrated that the green alga
Chlamydomonas converts up to 22% of light energy into hydrogen
energy, equivalent to a 10% solar energy conversion efficiency
(Greenbaum, 1988). On the other hand, this process is based on two
low compatible reactions: in the first step, water is split to produce
oxygen, and in the second, the reducing power of electrons is passed
to protons to make hydrogen through hydrogenase. Since oxygen is a 19 strong inhibitor of hydrogenase activity, a feedback inhibition
mechanism is inherent in the system. Aiming at real world
application, this is a great obstacle: since laboratory strategies to
overcome it by either consuming or sweeping out the oxygen as fast
as it is produced would not be practical for large scale operations, the
only applicable strategy could be the cultivation of algae under sulfur
deprivation for 2 or 3 days to provide anaerobic conditions in the
light (Winkler et al., 2002). 2 - Heterocystous nitrogen-fixing cyanobacteria
This strategy assumes to employ algae (such as Anabaena cylindrica,
a filamentous cyanobacteria) as systems where the two previously
described incompatible reactions can be separated by compartmentalization, with CO2 acting as an intermediate to shuttle
between the two compartments. In this type of organisms, oxygenic
photosynthesis is restricted to vegetative cells, then microoxic
heterocysts evolves hydrogen from reductants generated by
photosynthesis through the use of ferrodoxin and nitrogenase
enzymes, after N2 reduction process has been blocked. Nitrogenase is
the key enzyme that catalyzes hydrogen gas production and it must
be noticed that its activity is inhibited in the presence of oxygen,
ammonia or at high N/C ratios (Kapdan and Kargi, 2006) 3 - Indirect biophotolysis: non-heterocystous nitrogen-fixing
Differently from the previous strategy developed by heterocystous
cyanobacteria, non-heterocystous nitrogen-fixing cyanobacteria are
able to separate the H2 and O2 evolution steps temporally, such as a
day-night cycle, or spatially, through separate bioreactions rather
than through two cell types. However, as before, CO2, which is
fixated and released, is the intermediate between the two reactions.
Moreover, these microorganisms can use both nitrogenase and
reversible hydrogenase for hydrogen production: unfortunately 20 nitrogenase has a high ATP requirement, which lowers potential
solar-energy conversion efficiencies to unacceptable levels.
Considering the real-world applicability of this process (which has
been already tested in a two-stage power plant in Osaka, Japan, by
Akano et al., 1996), the advantage is that the CO2 fixation stage,
representing up to 90% of the total area required for the plant, would
take place in open ponds, which are much cheaper than the closed
photobioreactors typically required for the H2 photo-evolution stage. 4 - Photo fermentation: photosynthetic bacteria
In the light, some photosynthetic bacteria can convert organic
substrates, including many wastes, into hydrogen and CO2 (Figure
1.5, pathway 4). In principle relatively little light-energy input
(which means small photobioreactors) should be required to drive
this reaction, as most of the hydrogen energy comes from the organic
substrates themselves. However, the high-energy demands by the
nitrogenase catalyzing hydrogen evolution in these bacteria and the
relatively low light intensities at which these bacteria operate, make
their photosynthetic efficiencies disappointing.
Hydrogen production rates vary depending on the light intensity,
carbon source and the type of microbial culture. The organisms seem
to prefer organic acids as carbon source, such as acetic, butyric,
propionic, lactic and malic acid. On the basis of available literature,
the highest conversion efficiency, between 80 and 86%, was
obtained using lactic acid as the carbon source (Kapdan and Kargi,
2006). Carbohydrates and industrial effluents may also be used:
using three different substrates hydrogen was produced by four
strains of photofermentative Rhodopseudomonas sp. bacteria.
Among them, sugarcane juice supported the maximum level of
hydrogen production followed by potato starch and whey at the rates
of 45, 30 and 25 ml H2 mg -1 bacterial cell(dry weight) h -1, respectively (Ike et al., 1999). 21 5 - Microbial shift reaction: photosynthetic bacteria
Photosynthetic bacteria can also act as biological catalysts in the
conversion of carbon monoxide to hydrogen, a strategy which has
the potential for near term practical application. The microbial shift
reaction shown in Figure 1.5 (pathway 5) can accomplish this
conversion at room temperature and in just one step (in contrast to
chemical catalysts use). This pathway could be particularly useful for
small scale application, but in order to make this process
economically feasible it would be advantageous to use gas phase
bioreactors and overcome the mass transfer limitation (Markov et al.,
1996). 6 - Dark fermentation
Dark fermentative hydrogen production has a lower technical
complexity compared to the photo fermentation process and can
exploit a broad spectrum of applicable substrates with high hydrogen
evolution rate (Nath and Das, 2004). The word "dark" is used to
distinguish this technology from other fermentations that need light
to work. This technology will be discussed in details in the following
section (1.3.2). 22 P ro ce ss A d v a n ta g e s D is a d v a n ta g e s T y p e o f o rg a n is m - R e p re se n ta ti v e s tr a in M ax im u m r e p o rt e d r a te (m m o l H 2/ L h ) R e fe re n ce s D ir e ct B io p h o to ly si s C an p ro d u ce H 2 d ir e ct ly f ro m w a te r a n d s u n li g h t R e q u ir e s h ig h i n te n si ty o f li g h t G re e n a lg a e - C h la m y d o m o n a s R e in h a rd ii 0 .0 7 Le v in e t a l. 2 0 0 4 ; M a h y u d in e t a l. 1 9 9 7 S o la r co n v e rs io n e n e rg y i n cr e a se d b y t e n fo ld s a s co m p a re d t o t re e s a n d c ro p s O 2 ca n b e d a n g e ro u s fo r th e s y st e m Lo w e r p h o to ch e m ic a l e ff ic ie n cy In d ir e ct B io p h o to ly si s C a n p ro d u ce H 2 fr o m w a te r U p ta k e h y d ro g e n a se e n zy m e s a re t o b e r e m o v e d to s to p d e g ra d a ti o n o f H 2 C y a n o b a ct e ri a - A n a b a e n a V a ri a b ili s 0 .3 6 T a n is h o e t a l. 1 9 9 8 ; K a ta o k a e t a l. 1 9 9 7 H a s th e a b il it y t o f ix N 2 fr o m a tm o sp h e re A b o u t 3 0 % O 2 p re se n t in g a s m ix tu re O 2 h a s a n i n h ib it o ry e ff e ct o n n it ro g e n a se P h o to fe rm e n ta ti o n A w id e s p e ct ra l li g h t e n e rg y c a n b e u se d b y t h e se b a ct e ri a Li g h t co n v e rs io n e ff ic ie n cy i s v e ry l o w , o n ly 1 -5 % P h o to sy n th e ti c b a ct e ri a - R h o d o b a ct e r S p h e ro id e s 0 .1 6 M iz u n o e t a l. 2 0 0 0 C a n u se d if fe re n t w a st e m a te ri a ls l ik e d is ti ll e ry e ff lu e n ts , w a st e e tc O 2 is a s tr o n g i n h ib it o r o f h y d ro g e n a se D a rk f e rm e n ta ti o n It c a n p ro d u ce H 2 a ll d a y l o n g w it h o u t li g h t R e la ti v e ly l o w e r a ch ie v a b le y ie ld s o f H 2 F e rm e n ta ti v e b a ct e ri a - E n te ro b a ct e r cl o a ca e D M 1 1 7 5 .6 H u ss y e t a l. 2 0 0 3 A v a ri e ty o f ca rb o n s o u rc e s ca n b e u se d a s su b st ra te s A s y ie ld s in cr e a se H 2 fe rm e n ta ti o n b e co m e s th e rm o d y n a m ic a ll y u n fa v o ra b le C lo st ri d iu m s p . S tr a in N o 2 6 4 .5 La y 2 0 0 0 It p ro d u ce s v a lu a b le m e ta b o li te s su ch a s b u ty ri c, la ct ic a n d a ce ti c a ci d s a s b y p ro d u ct s P ro d u ct g a s m ix tu re c o n ta in s C O 2 w h ic h h a s to b e se p a ra te d It i s a n a e ro b ic p ro ce ss , so t h e re i s n o O 2 l im it a ti o n p ro b le m T a b . 1 .1 B io lo g ic al h y d ro g en p ro d u c ti o n p ro c es se s. 23 1.3.2 Hydrogen production by Dark Fermentation Dark fermentative hydrogen (Pathway 6, Figure 1.5) is produced by
the intermediate steps (acidogenesis and acetogenesis) of the well
known Anaerobic Digestion (AD) process, and therefore it shares
some of the advantages of this well-established technology, like the
capability of exploiting low-cost substrates/organic wastes at high
Before delving into the hydrogen production, it is essential to briefly
present the AD process, a complex microbiological process
involving a community of several populations and many different
metabolic pathways.
As known, microbial conversions of organic matter occur in
sequence from more to less energetically favorable reactions and if
several electron acceptors (oxygen, nitrate, sulfate, iron (III), and
CO2) are available, the most energetically favorable electron
acceptor will be utilized. So in the absence of the electron couple O2/H2O (aerobic respiration, '' = -78.3 kJ/mol) or other strong acceptors such as in AD, the organic matter will be reduced to the end products CH4 and CO2 via methanogenesis ('' = +23.5 kJ/mol).
The AD could be divided into four main steps, as follows:
' Hydrolysis: conversion of non-soluble biopolymers to soluble organic compounds; ' Acidogenesis: conversion of soluble organic compounds to volatile fatty acids (VFA) and CO2; ' Acetogenesis: conversion of volatile fatty acids to acetate and H 2; ' Methanogenesis: conversion of acetate and H 2 to methane gas. Figure 1.6 gives a schematic representation of anaerobic degradation
of organic matter. 24 Fig. 1.6 Simplified schematic representation of the anaerobic digestion process. A great advantage of fermentation is the fast degradation of solids
and other complex organics, such as those found in wastes and
agricultural products. But despite its speed, fermentation is not yet
very efficient for capturing the energy value of biomass into
hydrogen (Kotay and Das, 2008). Indeed, the main problem is that
the dark fermentative bacteria produce only relatively small amounts
of hydrogen and, as yield increase, hydrogen fermentation becomes
thermodynamically unfavorable, decreasing the feasibility of its
application to an industrial scale. Benemann (1996) estimated that
fermentative hydrogen production from wastewater has the greatest
potential for economical near-term production of bio-hydrogen, but
only if hydrogen conversion efficiency could reach 60-80%. 25 If biomass could be categorized as a carbohydrate such as glucose
(C6H12O6), the complete conversion of each mole of glucose would
produce 12 mol of hydrogen (Equation 1). C6H12O6 + 6 H2O '' 12 H2 + 6 CO2 (1) The maximum result about fermentative hydrogen yield was reported
by Woodward et al. (2000) who achieved a 96.7% conversion
efficiency (11.6 molH2/molglucose-6-phosphate) by combining the enzymes
of the oxidative pentose phosphate cycle with hydrogenases from
Pyrococcus furiosus under optimized in vitro conditions. However
this result was obtained by using exclusively enzymes, not bacteria.
Indeed, in bacterial fermentation the complete oxidation of glucose
into hydrogen and carbon dioxide is not possible, as the
corresponding reaction (1) is not feasible thermodynamically, having
a positive ''Go of +3.2 kJ. According to Logan (2004) fermentation of glucose by all known microbiological routes (primarily by
Clostridia) produces up to 4 molH2/molglucose. Thus, there are no
known naturally occurring biochemical routes for achieving the
required 60-80% conversion efficiency.
Moreover, while a conversion efficiency of 33% is theoretically
possible for laboratory hydrogen production from glucose
(corresponding to 4 mol H2/mol glucose), only half of this is usually
obtained under batch or continuous fermentation conditions (Logan
et al., 2002; Van Ginkel et al., 2001). Indeed, glucose gives 4 mol of
hydrogen if exclusively 2 mol of acetate are simultaneously
produced (Equation 2), while only 2 mol H2 are achieved when
butyrate is the main fermentation product (Equation 3). C6H12O6 + 2 H2O '' 4 H2 + 2 CO2 + 2 C2H4O2 (2) C6H12O6 '' 2 H2 + 2 CO2 + C4H8O2 (3) In particular, according to Liu et al. (2002), using mixed microbial
cultures for inoculation of the process, a combined production of
both acetic and butyric acid often occurs, with butyrate typically
amounting to 60-70% of the aqueous products (Equation 4): 26 4 C6H12O6 + 2 H2O '' 8 H2 + 8 CO2 + 3 C4H8O2 + 2 C2H4O2 (4) This results again in a maximum hydrogen conversion yield of 2
Moreover, other compounds may be present as end products of the
fermentation, further lowering the yields. If propionic acid is the end
product, the H2 conversion yield is just 1 molH2/molglucose : C6H12O6 '' H2 + CO2 + C3H6O2 (5) while the formation of lactic acid (CH3CHOHCOOH) by anaerobic
degradation of glucose is associated with no production of hydrogen
(Equation 6). C6H12O6 '' 2 C3H6O3 (6) In addition lactic acid bacteria (e.g. Lactobacillus paracasei or
Enterococcus durans) can form intermediate catabolic products (so-
called bacteriocrine), inhibiting hydrogen producing bacteria (Noike
et al., 2002).
Another reason suggested for lower yields is that glucose is usually
just partially biodegraded through this bioprocess (Kapdan and
Kargi, 2006). However, it is more probable that utilization of
substrate for bacterial growth is the main reason for obtaining yields
lower than theoretical estimations, because it was shown that even
when more than 95% glucose was degraded, the yield could be less
than 1.7 mol H2/mol glucose (Lin and Chang, 2004).
In Paper II we collected the best results in terms of yield
(molH2/molglucose) reported in the literature for glucose dark
fermentation in batch test with mixed microbial cultures or with
pure/selected/GM microorganisms (Figure 1.6; see also Table S.2 in
supplementary information section of Paper II). H2 yield obtained in
our research (Paper II) by mixed microbial cultures enriched from
natural sources (soil-inocula and anaerobically digested materials)
are reported in the same figure. With low substrate concentration (1 g 27 glucose/L) we achieved high H2 yields (2.8 ± 0.66 molH2/molglucose), comparable to pure microbial cultures achievements. Fig. 1.6 Effect of substrate concentrations and type of inoculum on hydrogen yield: our Paper
II results in comparison with literature results. ( ') acclimated inocula (Paper II), (') soil inocula (Paper II), ( '') literature results with naturally-sourced mixed microbial cultures, (') literature results with pure/selected wild type microbial cultures and (+) literature results with
genetically modified pure cultures. Tab. 1.2 Reported maximum fermentative hydrogen yield achieved by dark fermentation in
continuous system (with or without optimization efforts). Full references are reported in the
Reference section of this Ph.D. thesis. Process Substrate Max H2 yield (mol H2/mol hexose) References Fluidized bed reactor Sucrose 1.3 Wu et al. 2003 N2 sparging, CSTR Wheat starch co-product 1.9 Hussy et al. 2003 Upflow reactor Wastewater 2.1 Yu et al. 2002 Fermentor Sucrose 2.1 Fang et al. 2002 Fermentor Glucose 2.1 Fang and Liu 2002 CSTR Glucose, sucrose 2.2 Chen and Lin 2001 N2 sparging, CSTR, HRT 8.5 h Glucose 1.43 Mizuno et al. 2000 N2 sparging, continuous Glucose 1.4-2.3 Kataoka et al. 1997 Chemostate, HRT 17 h 0.75% soluble starch 2.14 Lay 2000 Continuous (immobilized bioreactor) Glucose 2.3 Kumar and Das 2000 Continuous (immobilized bioreactor) Glucose 3.8 Kumar et al. 2001 28 Similarly, Table 1.2 collects the recent maximum hydrogen yield
achieved in continuous experiments, also including optimization
efforts. Again, those yields typically reach 2.1 molH2/molhexose. However, Benemann (1996) stated also that the economics of
hydrogen fermentation could be favorable even at less then
stoichiometric yields: production costs of methane fermentations
range from $3 to $8 per MMBTU (million British thermal units),
whereas hydrogen produced by the same types of hardware could be
sold as much as $15 per MMTBU, depending on location, scale,
purity and other factors. For real, even if hydrogen fermentation
would embrace similar hardware to that currently used in industrial
anaerobic digesters, economic feasibility will not be sustainable until
these yields reach the 60-80% efficiency mark, even without taking
into account the substrate used for the process. It has been speculated
that a fermentation yield of 10 molH2/molglucose and a glucose cost of
5 cents per dry pound will be required for this process to approach
costs competitive with traditional fuels. Based on these considerations, it is possible to summarize and to
split in three different strategies the possible ways to increase the
biological energy recovery via hydrogen dark fermentation:
1. Increasing the biohydrogen yield to around 85% through chemical-physical process conditions optimization, efficient
bioreactor design, use of suitable microbial strains, genetic and
metabolic engineering of hydrogen-producing microorganisms,
redirection of the metabolic pathways, etc. (details in section 1.3) 2. Using essentially cost-free substrates like solid waste materials, such as those from farms, and dissolved organic matter from
various industrial and domestic wastewaters (details in chapter 2) 3. Finding methods to harness the remaining 85% of the energy, i.e. including hydrogen production into a modular energy production
concept (details in chapter 4). 29 1.4 Key parameters regulating the dark fermentative hydrogen production and their
Process engineering, such as bioreactor design and operating
parameters optimization, may enhance the performances of the
fermentative hydrogen-producing microorganisms, thus positively
affecting the final hydrogen yield. Also, the unstable hydrogen
production could be minimized, since it is possibly attributed to the
metabolic shift of bacteria or system imbalance.
Some regulating factors among those that can be manipulated to steer
a bioconversion process to the desirable product (in this case,
hydrogen) would be discussed in this section. - Substrates (pure)
Considering their higher yields of hydrogen per mole of substrate,
carbohydrates are the preferred pure substrate for laboratory-scale
dark fermentative hydrogen production. They can be monosaccharides, but may also be polymers such as starch, cellulose
or xylan.
Table 1.3 (Courtesy of Kapdan and Kargi, 2006) summarizes the
yields and the rates of hydrogen production obtained from batch and
continuous tests when hexoses (glucose and sucrose) and
polysaccharides (starch and cellulose) were used as the substrate.
Dealing with carbohydrates polymers and according to the reaction
stoichiometry, a maximum of 553 mL of hydrogen gas can be
produced from 1 gram of starch, with acetate as a by-product.
Considering that the yield may be lower than the theoretical value
because of utilization of substrate for cell synthesis, Zhang et al.
(2003) obtained a noticeable result, with high specific yield of 480
mLH2/gVSS d feeding a mixed sludge with 4.6 g starch/L solution. Very
high hydrogen yield (2.4 molH2/molglucose) was also obtained by 30 mixed culture from starch in the presence of 0.1% polypepton
(Yokoi et al., 2001). Tab. 1.3 Yields and rates of bio-hydrogen production from pure carbohydrates by batch or
continuous dark fermentations (Courtesy of Kapdan and Kargi, 2006). On the other side, proteins, peptides and amino acids seem to be less
suitable for dark hydrogen production, whereas biopolymers like
lipids are unsuited. Formate and some peptides have been studied
before as substrates for dark hydrogen production by De Vrije and Continuous experiments Batch experiments 31 Claassen (2005), who obtained amino acids oxidation to hydrogen by
specific strains of extreme thermophilic bacteria. However it is still
unclear whether specific amino acids, entering bacterial metabolism
at the level of pyruvate, are selected or whether this phenomenon is
more general.
The vast range of potential organic substrates for dark hydrogen
fermentation will be dealt also in chapter 2 of this Ph.D. thesis,
specifically referring to complex substrates such as agricultural
biomasses and actual organic waste streams (from solid wastes to
wastewaters) which have been successfully used for H2 production. - Biomass pre-treatment processes
Hydrolysis is the first step of the biologic conversion of biomass.
During the hydrolysis, both solubilization of insoluble particulate
matter and biologic decomposition of organic polymers to dimers
and monomers (e.g. simple sugars, amino acids, long-chain fatty
acids and aromatic compounds) take place. This step is especially
required for recalcitrant biomasses like the lignocellulosic ones and it
may be accomplished biologically or by means of chemical and/or
physical techniques. So, a number of pre-treatments can be chosen,
depending on the biomass solid structure, the process used for
biomass fermentation and the desired microbial products. It is
possible to include (Gavala et al., 2003):
' Heat treatment: in the temperature range of 40-275 °C. In case of lignocellulosic biomasses, among celluloses, hemicelluloses and
lignin, the hemicelluloses are the most sensitive to the thermal
treatment and thus are the first to be degraded (Ntaikou et al.,
2010a), with just a partial solubilization of lignin at 160 °C or
above. ' Chemical treatment: using ozone, acids, alkali, etc. Dealing with lignocellulosic biomass, both dilute or concentrated acids, such
as H2SO4 and HCl, can be used. Mainly hemicellulose is
hydrolyzed in this process, while lignin is hardly dissolved, 32 nevertheless being disrupted to a high degree it leads to
increased cellulose susceptibility to enzymes. In case of alkaline
pretreatment, diluted bases cause the decrease of polymerization
degree and crystallinity, the destruction of links between lignin
and other polymers, and the lignin breakdown. ' Mechanical treatment: using ultrasounds, mills or homogenizers. Mechanical pre-treatment is almost always applied before any
other kind of pre-treatment, and aims at the reduction of particle
size and crystallinity of biomass, leading to an increased specific
surface available for enzymatic attack. ' Biological hydrolysis (enzyme addition): specific enzymes may drastically improve the depolymerization of the organic matter.
Indeed, either enzymes -hydrolases - or extracellular enzyme-
producing microbial cells may be added to accomplish biomass
hydrolysis, as typically occurring in the food processing
industry, where enzymes are used in order to decrease food
wastes with simultaneous formation of higher value chemicals or
biofuels (in particular, production of ethanol from agricultural
and forestry residues) (Galbe and Zacchi, 2002). Biological pre-
treatment seem very effective especially with lignocellulosic
biomass, proceeding through the concerted action of specific
enzymes, i.e. lignin peroxidase, manganese peroxidase, H2O2-
generating enzymes and laccase, which produce strong oxidants
and combust the lignin framework. ' Combination of two or more of the aforementioned methods: for example wet oxidation (combination of oxygen pressure,
alkaline conditions and elevated temperature) of the lignocellulosic material has been used to enhance the
accessibility of the carbohydrates, as it dissolves the
hemicellulose fraction and makes the solid cellulose fraction
susceptible for enzymatic hydrolysis and fermentation (Lyberatos et al., 2005). 33 - Fermentation pathways and their terminal products
The dark fermentation of organic wastewaters is basically of three
types: butyric acid-type fermentation, metacetonic acid-type
fermentation, and ethanol-type fermentation. Each fermentation has
specific properties and terminal products that can affect hydrogen-
producing ability and metabolic pathways of fermentative
microflora. For example, ethanol fermentation achieves high ethanol
production with simultaneously high hydrogen production (specifically: higher biogas and H2 production rate and higher
hydrogen content than metacetonic acid-type fermentation) under
equal quantities of aqueous terminal products. As showed by Ren et
al. (2006) ethanol seems to have little inhibitory effect on
fermentative hydrogen production, while acetic acid has strong
inhibitory effect on hydrogen production.
However, the common major products in hydrogen production by
anaerobic dark fermentation of carbohydrates are acetic, butyric and
propionic acids, and even formation of lactic acid was observed
when lactose and molasses (sucrose) were used as substrates
(Kapdan and Kargi, 2006). High concentration of organic acids may
result in a collapse of the pH gradient across the membrane and
cause the total inhibition of all the metabolic functions in the cell
(Jones and Woods, 1986).
It has been claimed that both the total acetate or butyrate acid
concentration and the undissociated form of these acids can inhibit
the dark hydrogen fermentation process (Jones and Woods, 1986;
Van Ginkel and Logan, 2005b; Van Niel et al., 2002). In particular,
Van Ginkel and Logan (2005b) studied the inhibition of biohydrogen
production by using undissociated acetic and butyric acids: with
acetic acid addition to give total undissociated acid concentrations in
the reactor of 63 mM, which occurred at pH 5.5 and 165 mM of
added acetate, complete H2 production inhibition was reached. In Liu
et al. (2008) hydrogen yields were inhibited more by self-produced 34 acids than by similar concentrations of externally added acids and the
acetate concentration started to inhibit extreme-thermophiles
hydrogen production at more than 50mM. At acetate concentration of
200 mM, the hydrogen production (36 mL/gVSadded) was 7 time lower
than at 5-25 mM acetate (250 mL/gVSadded), and the lag phase was
more than 100 hours. - Microorganisms
Hydrogen production can be achieved either through selected
hydrogen producing bacteria (pure cultures of strict anaerobes,
facultative anaerobes and even some aerobes) or mixed microbial
cultures derived from natural environments (soil, wastewater sludge,
compost...) (Liang, 2002, Nandi and Sengupta, 1998).
The advantages of pure cultures are the selectivity of substrates, the
easy manipulation of the metabolism by altering growth conditions,
the common high hydrogen yields as an effect of the reduction of
undesired by-products and the repeatability of the process. On the
other hand, they are sensitive to contaminations, thus requiring
aseptic process conditions unfeasible for industrial production of H2
(Ntaikou et al., 2010a).
Among the obligate anaerobes spore-forming organisms (Clostridia,
methylotrophs, rumen bacteria, archaea), those belonging to the
genus Clostridium (C. buytricum, C. thermolacticum, C.
pasteurianum, C. paraputrificum M-21 and C. bifermentants) are
widely exploited. They produce hydrogen gas during the exponential
growth phase but when the stationary growth phase is reached, their
metabolism could shift from a hydrogen/acid production phase to an
undesirable solvent production phase (Kapdan and Kargi, 2006).
Interestingly, a culture with Clostridia dominance can be easily
obtained by heat treatment of biological sludge: the spores formed at
high temperatures can be activated when required environmental
conditions are provided for hydrogen gas production (Sung et al.,
2003). 35 Among the facultative anaerobes, the species of the genus
Enterobacteriaceae have the ability to metabolize glucose by mixed
acid or the 2-3 butanediol fermentation pathway. In both patterns,
CO2 and H2 are produced from formic acid in addition to ethanol and
the 2-3 butanediol (Kapdan and Kargi, 2006). Their H2 yield on
glucose is normally lower compared to that of Clostridia, however
hydrogen production of anaerobic facultative bacteria (Enterobacter
aerogenes and Enterobacter cloacae strain ITT-BY 08) have given
interesting results (2.2 mol H2/mol glucose; Kumar and Das, 2000).
Recently, hydrogen producing aerobic cultures such as Aeromonas
spp., Pseudomonas spp. and Vibrio spp. were also identified.
Hydrogen production performance of anaerobic thermophilic and
hyperthermophilic organisms has also been investigated. Shin et al.
(2004) reported Thermoanaerobacterium thermosaccharolyticum
and Desulfotomaculum geothermicum strains producing hydrogen
gas at high yield. Other bacteria, (Thermococcus kodakaraensis
KOD1 isolated from a geothermal spring in Japan; Clostridium
thermolacticum producing hydrogen from lactose at 58 °C;
Klebsiella oxytoca HP1 isolated from hot springs) have been used by
different researchers, as reported by Kapdan and Kargi (2006). Compared with that of selected hydrogen-producing bacteria (pure
culture), the hydrogen-producing ability of mixed cultured bacteria
fed with complex organic substance is usually higher and the control
and operation of the process easier (no medium sterilization is
required). Therefore, beyond being a cheaper process, the use of
mixed culture makes the microbial ecology more tolerant to stress
and system imbalance. It depends on complex microbial interactions:
the various species will grow in an interactive manner with
commensalistic, ammensalistic, competitive and more complicated
impacts of one specie to another. Thus the total resulting
fermentative pathway will not correspond to what could be
obtainable from individual species and each metabolic step will be 36 modulated by the overall interspecies available enzymatic activity
(Lyberatos et al., 2005).
Interestingly, Lin et al. (2003) studied the cooperation of hydrogen-
producing fermentation bacteria in mixed culture with a batch test.
Their results demonstrated a tight cooperation within the mixed
culture, but on the other side, also demonstrated that the cooperation
is conditional on the substrates: when fed with glucose (easily used
by H2-producing bacteria), the hydrogen-producing ability might be
restrained because of the competition for the substrate between
hydrogen-producing bacteria and other fermentation bacteria.
Indeed, the main disadvantage of the mixed culture is the possible
predominance of non-hydrogen producing species such as
methanogens, homoacetogens and lactic acid bacteria: to minimize
this risk an initial pretreatment of the seed, together with the
maintenance of environmental conditions unsuitable for the
hydrogen consuming species, is usually required. De Vrije and Claassen (2005) made an interesting overview of the
different pure strains and mixed culture used in dark fermentative
biohydrogen production tests, resumed in a table reported here
(Table 1.4, Courtesy of De Vrije and Claassen, 2005) together with
experimental conditions used, H2 yields and H2 production. 37 Tab. 1.4 Hydrogen yields and production rates by microorganisms as reported in the scientific
literature (Courtesy of De Vrije and Claassen, 2005). - Genetic engineering strategies
Different studies on hydrogen production by dark fermentation
focused on artificially regulating and controlling bacteria metabolic
pathways, aiming at enhancing hydrogen-producing efficiency at a
microbial molecular level. The development of the following areas
rouses the scientific community's interest:
' Identification and isolation/selection of high-efficient hydrogen- producing bacteria. Among the microbial H2-producing species
(Bacteroides, Zymomonas, Clostridium, Fusobacterium, etc.)
some bacteria with peculiar characteristics were reported. For
example, Ren et al. (2003) found a new genus of fermentative
H2-producing bacteria, including the strains Rennanqilyf 1 and 38 B49. The latter in particular had a good acid resistance (optimal
pH under stirring cultivation: 3.9-4.2) and high H2 production.
Specific hydrogen conversion rate and hydrogen production rate
of strain YUAN-3T, isolated by Xing et al. (2006), were 2.81
molH2/molglucose and 27.6 mmolH2/gdry cell h, respectively, and
additionally this strain is the only auto-aggregative bacterium
among the hydrogen-producers reported in literature up to date. ' Regulation and control of enzyme genes involved in fermentative hydrogen production or in correlated metabolic pathways.
Activity and time of expression of hydrogenase, the last rate-
limiting enzyme of hydrogen production pathway, directly affect
the metabolism of hydrogen-producing bacteria, and thereby
influence the production rate and yield of hydrogen (Liu et al.,
2008). Many researchers made studies about the expression level
of hydrogenase and showed that its overexpression typically
enhances the hydrogen-producing rate (Mishra et al., 2004;
Morimoto et al., 2005). Also other enzymes can be regulated: for
example Yoshida et al. (2005) performed the genetic
overexpression of the formate hydrogen lyase (FHL) in E. coli,
effectively resulting in a 2.8-fold increase in hydrogen
productivity of the mutant strain compared with the wild type
strain ' Increased application of microbial molecular breeding and development of new techniques for breeding hydrogen-producing
bacteria. Presently, gene chip, microarray, real-time quantitative
polymerase chain reaction, protein two-dimensional electrophoresis, multidimensional liquid chromatography, and
surface plasmon resonance are playing a promoting role in
microbial molecular biology and microbial molecular breeding
(Liu et al., 2008). However, some transformation and expression
systems of fermentative hydrogen-producing microbes are still
imperfect, and the study on molecular breeding aiming at 39 improving the hydrogen-producing fermentation pathways is still
in the initial stages. - Hydrogen producing mixed microflora from natural sources:
inoculum acclimation and methanogens control
Zuo et al. (2005) used (pre-heated) river sediments as seed sludge to
achieve anaerobic biohydrogen production. But this is just one of the
different natural sources exploitable in a dark fermentative process.
Previous authors demonstrated that H2-producing consortia can be
obtained from various environmental sources, such as soil, compost,
sewage sludge and various fermented organic materials (Kyazze et
al., 2006; Li and Fang, 2007) and in Paper II we report interesting
results from the preparation and use of both soil-inocula (a rice soil,
a green urban soil and a vegetables-cultured soil ) and anaerobically
digested materials. All our works (Paper I, II and III) employ similar
inoculum acclimation strategy.
Indeed, since dark fermentation has been shown to have great
potential as applicable process to produce biohydrogen from a
variety of organic materials, mixed cultures easily obtainable from
natural sources and able to operate on non-sterile feedstock are
required for future real implementation of this bioprocess. (Hawkes
et al., 2002). Lyberatos et al. (2005) state that a mixed culture with
numerous microorganisms capable of degrading different organic
compounds should be ensured in the initial inoculum, whose choice
should be based on both the biomass used as substrate and the
desirable products. Therefore the acclimation of a microbial culture
is a very important process which brings significant changes to the
microbial population and adapts the microbial to a specific substrate.
A well acclimated heterogeneous culture is characterized by better
performance, concerning the efficiency and the selectivity of the
fermentation process, compared to the initial inoculum (Lyberatos et
al., 2005). For example, Liu (2008) and other authors (Radmann et
al., 2007; Yokoi et al., 2002) report many operational advantages 40 using the "repeated batch cultivation technique", a well-known
method for acclimating bacteria and enhancing the productivity of
microbial cultures while simultaneously controlling the nutrients
feed rate.
Nevertheless, in order to scale-up processes to make industrial
production of biohydrogen economical, not only obtaining the
inoculum from natural sources, but also avoiding loss of hydrogen
through interspecies transfer - primarily to methanogens naturally
occurring in non selected mixed cultures - is required. Three
methanogens-inhibiting treatment methods have been reported so far,
which are heat shock, pH control, and 2-Bromoethanesulfonic (BES)
acid control. Most bio-hydrogen researchers use high temperature
(range 75 °C - 121 °C) to inactivate hydrogenotrophic bacteria and
harvest anaerobic spore-forming bacteria such as Clostridium: this
process is pretty fast and its duration varies between 15 min and 2 h
(Gavala et al., 2006; Ntaikou et al., 2010a; Oh et al., 2003; Wang et
al., 2003a). In our works (Paper I, II and III) we typically heat
shocked the inoculum at 100 °C for 1 or 2 h.
The pH control method is based on inactivating the methanogens
maintaining them in a low pH (5 - 5.5) environment, on the other
hand suitable for H2-producing bacteria (Chang et al., 2002; Oh et
al., 2003). Otherwise, BES (C2H4BrO3SNa) is introduced as a
specific methanogen chemical inhibitor but failures have also been
reported about the use of this chemical in biologic fermentative
systems, probably due to the high concentration used, far from the
requirements of real situations (Lee et al., 2009; Liu et al., 2006). - Reactor type and design optimization
Reactor design deeply influences the process performances through
different aspects (the reactor microenvironment, the established
hydrodynamic behavior, the contact between substrate and consortia,
etc.), achieving substantial increases in hydrogen yield if well
optimized. 41 Assuming that batch mode is more suitable for research purposes
about fermentative hydrogen production (Ntaikou et al., 2010a),
industrially feasible process would work in continuous or
semicontinuous mode and should exploit the well-established and
commercial available AD technologies, currently used for
wastewater treatment and biogas production. Among them, it is
possible to distinguish three categories, based on the feedstock type:
' wet fermentation system: it fits to waste streams of less than 15% total solids, therefore solid waste streams are often diluted with
recycled process water to form a slurry. The continuous stirred
tank reactor (CSTR) is the most used reactor type to digest low
solid waste streams (animal manure, sewage sludge, household
waste, agricultural wastes, feces, urine, kitchen waste or
mixtures of these substrates), offering simple construction, ease
of operation and effective homogenous mixing as well as
temperature and pH control. These digesters consist of a void
recipient stirred by biogas recirculation, liquid recirculation or
mechanical means (Figure 1.7). Mechanically stirred CSTRs
were used in our continuous tests about hydrogen and methane
production from anaerobic digestion of organic substrates/wastewaters (Paper II and III). Fig. 1.7 Schematic diagram of a CSTR mechanically stirred (top) and stirred by biogas
recirculation (bottom) (Courtesy of de Mes et al., 2005). 42 CSTRs have the advantage of maintaining constant, homogeneous and well characterized process conditions, usually
preventing stratification and formation of a surface crust and
ensuring that solids remain in suspension. On the other hand, if
process parameters are not correctly designed, bacterial washout
may occur, thus decreasing the reactor performance. ' dry fermentation system: it fits to waste streams with a total solid percentage of more than 20% and to energy crops. ' high rate systems: they fit to wastewaters in high rate continuous flow. Reactors that allow high hydraulic loading rates without
the washout of microorganisms could be divided into systems
with fixed bacterial films on solid surfaces or into systems with a
suspended bacterial mass, where retention is achieved through
external or internal settling (like the contact digester or the
upflow anaerobic sludge blanket - UASB - often adopted in
hydrogen production tests). Fixed-bed bioreactors containing a consortium of mesophilic bacteria
have been reported to enhance the rate of hydrogen production to a
greater extent than reported by other approaches (121 mmolH2/Ldigester
h; Chang et al., 2002). In this case activated carbon as a support
matrix allowed retention of the H2-producing bacteria within the
In another study (Van Groenestijn et al., 2002), hydrogen was
produced in a high-rate bioreactor in the presence of
hyperthermophilic bacteria, which formed a biofilm within an
anaerobic trickling filter containing packing materials with a very
high surface area. This resulted in the continuous flow of liquid-
suspended biomass substrate through the filter, so that the biomass
substrate, the H2-producing bacteria and the resulting gas phase were
in close proximity. The energy required to run this process was at
least four times lower than the combustion value of the H2 gas
produced in such reactors (Van Groenestijn et al., 2002). 43 Lastly, gas hold-up is one main problem for bioreactors: to mitigate
the problem and to improve performance in terms of both the rate of
hydrogen production and gas hold-up, tapered and rhomboidal
bioreactors have been proposed. Gas hold-up was found to be
reduced by 67% using a rhomboid bioreactor compared with a
tubular bioreactor and high hydrogen production rate was achieved
(75.6 mmolH2/L h; Kumar and Das, 2001). - Organic loading and hydraulic retention time (HRT)
In batch tests in the mesophilic temperature range, it has been
observed that the overall biohydrogen yield is a function of the
organic load (OL, e.g. the amount of organic material added to a
particular environment), with the highest yields obtained at the
lowest OL (Van Ginkel et al., 2001). Our work (Paper II) confirms
this result, as increased glucose concentrations (1-7 g/L) caused a
progressive decrease in H2 yield (from 2.8 to 1.78 molH2/molglucose).
The increase of substrate load to the system may cause higher levels
of inhibitory metabolites and change chemical equilibrium,
depressing further hydrogen production (Van Ginkel et al., 2001;
Zhang et al., 2003).
This seems to be confirmed also in continuous culture, since
comparing tests by different authors (all using glucose as substrate
and adopting similar hydraulic retention time (6-12 hours)) the yield
achieved at 30, 20, 7 and 3 gCOD/L were respectively 1.1, 1.7, 2.1 and
2.4 molH2/molglucose (Fang and Liu, 2002; Kim et al., 2006; Lin and
Chang, 1999; Taguchi et al., 1995). Again, Kataoka et al. (1997)
found that increasing the glucose concentration from 5 to 10 g/L
decreased H2 yields from 2.0-2.3 to 1.4-2.0 molH2/molglucose.
Therefore, Van Ginkel and Logan (2005a) assume that biohydrogen
yields would be optimized for more dilute feeds and lower organic
loading rates than those typically used in biohydrogen reactor
studies. 44 Related to the OL, the HRT (hydraulic retention time) and/or the
SRT (solid retention time) strongly affect the stability and the
performance of a fermentation system and consequently the yield
and selectivity of specific metabolic products. The HRT is defined as , where Vr is the active volume of the fermenter and Q is the influent flow rate. The SRT is defined as the mass of solids in the
fermenter divided by the solids effluent rate and in CSTR reactor is
equal to HRT. For experiments with pilot or laboratory scale
reactors, HRT needs to be optimally selected and adapted to specific
reactor design and metabolic products (Lyberatos et al., 2005). In
CSTR system, short HRTs (< 3 days) are generally used together
with easily fermentable substrates, in order to wash out the slow
growing methanogens (requiring more than approx. 3 days to grow)
and to select for the acid producing bacteria (Chen et al., 2001).
Indeed, Ntaikou et al. (2010a) state that in CSTR a HRT of 12-36 h,
depending on the substrate, provides complete conversion of
carbohydrates and highest hydrogen yields. However, it is also true
that too short HRT could lead to bad hydrolysis of organic wastes
(Han and Shin, 2004a) or to pH unbalance and thus to lower H2
production. Our considerations about HRT influence on maximizing
biohydrogen production and process stability are reported in Paper I,
where we investigated a wide range of HRT (1-4.5 d) associated with
fruit-vegetable waste and swine manure mixture fermentation. - Temperature
Temperature is an important factor for microbial activity, which
regulates the metabolic and reproduction rates of microorganisms.
There are mainly three temperature intervals considered valid for
hydrogen production by anaerobic fermentation: the psychrophilic
(Temperature range 0-20 °C, optimum 15 °C), the mesophilic
(Temperature range 15-45 °C, optimum 37 °C) and the thermophilic
(Temperature range 45-75 °C, optimum 55 °C). Moreover, extreme 45 thermophiles or hyperthermophiles H2-producing microorganisms
exist, whose growth optimal temperature is above 65 °C and 80 °C,
respectively (Levin et al., 2004).
Psychrophilic digestion is not commonly used, except for some
applications in the northern countries of the world, mesophilic
temperature is the most common range adopted, while thermophilic
digestion has been reported to have several advantages over the
others, such as higher reaction rates and pathogen-killing effect
(Meulepas et al., 2005). Therefore, we generally conducted our
experiments (Paper I, II and III) in thermophilic conditions (55 ± 2
°C). However, interesting studies reported that extreme-thermophilic
fermentation can further minimize the contamination by pathogens
and hydrogen consumers (Liu, 2008), achieve higher hydrogen total
production and production rate than mesophilic hydrogen
fermentation (van Groenestijin et al., 2002), and reach the theoretical
maximum yield of 4 molH2/molglucose (van Niel et al., 2002). - pH
The hydrogen-producing bacteria are quite sensitive to pH
fluctuation because pH change may result in the change of their
metabolic pathway: medium pH affects enzyme activity in
microorganisms (since each enzyme is active only in a specific pH
range) and thus hydrogen production yield, biogas content, type of
the organic acids produced and specific hydrogen production rate.
Under not optimal pH, the hydrogen fermentation process may
prolong the lag phase or shift to other pathways, such as solvent
production (Cheng et al., 2002; Liu, 2008; Temudo et al., 2007).
Very low initial pH of 4.0 - 4.5 causes long lag periods such as 20 h
(Khanal et al., 2004; Liu and Shen, 2004), while initial pH of 9.0
may decrease the lag time but gives lower biohydrogen yield and
higher risk of hydrogen consuming activities (Zhang et al., 2003).
The optimal pH for ethanol-type fermentative bacteria ranges from
4.0 to 4.5, while the pH range for the maximum hydrogen yield in 46 metacetonic acid-type fermentation is between pH 5.0 and 6.0
(Gomez et al., 2006; Kapdan and Kargi, 2006; Liu et al., 2006).
However, fermentative bacteria own balancing and regulating
abilities and can withstand pH also different from their optimum. For
example for the (extreme)thermophilic acid-type hydrogen
fermentation the optimum pH range was reported to be between 6.8
and 8.0 (Liu, 2008; Van Niel et al., 2002; Yokoyama et al., 2007) or
oppositely around pH 4.5 (Shin et al., 2004).
For real, without pH adjustment of the media, many studies report
that at the end of anaerobic hydrogen production via metacetonic
acid-type fermentation, the medium pH shifts away from its optimum
and it can reach values between 4.0 and 4.8, regardless of initial pH
(Liu et al., 2003; Liu and Shen, 2004; Morimoto et al., 2004; Yokoi
et al., 2001). Indeed, in an unbuffered system, the pH decreases due
to production and accumulation of organic acids which deplete the
buffering capacity of the medium. This, often caused by system
overloading, may inhibit hydrogen production by affecting the
activity of iron containing hydrogenase enzyme (Kapdan and Kargi,
2006; Khanal et al., 2004).
Therefore, medium pH adjustment through addition of chemicals
(acid or base) is often required to maintain the pH around its
optimum. On the other hand, this approach may not be optimal for
large-scale transfers, and when looking for full-scale applications
different strategies may be considered for maintaining acceptable
chemical equilibrium in fermentation broth.
So, in our studies (Paper I and III) enhancement of the buffer
capacity of the system to avoid the pH drop and maintain it around
its optimum, was reached by mixing the fermentable substrates with
alkalinity rich wastewater (such as swine slurry). Even effluents
recycled from AD process may be used in this strategy.

47 - H2 and CO2 partial pressure and gas sparging
The accumulation of hydrogen and carbon dioxide can lead to
repression of H2 production (due to end-product inhibition) and to
formation of more reduced products (Classen et al., 1999; Nath and Das, 2004).
In anaerobic digestion with mixed anaerobic cultures, the
accumulation of hydrogen is normally balanced by rapid hydrogen
consumption by methanogens, resulting in little net hydrogen
accumulation in the system (Mahyudin et al., 1997). On the contrary,
in Dark Fermentation hydrogen must be produced and accumulated
without being consumed. However, if hydrogen concentrations
increase over some fixed limits (pH2 of >50 kPa at 60 °C, >20 kPa at
70 °C, and >2 kPa at 98 °C), H2 synthesis decreases and metabolic
pathways shift towards production of more reduced substrates, such
as lactate, ethanol, acetone, butanol or alanine (Levin et al., 2004).
Therefore, the system used for Dark Fermentation must be designed
to both remove hydrogen before it leads to repression of its
production and to prevent interspecies hydrogen transfer leading to
methanogenesis (Mahyudin et al., 1997; Tanisho et al., 1998).
Therefore gas sparging has been found to be a useful technique to
reduce hydrogen partial pressure (pp) in the liquid phase and enhance
H2 yield.
In a study by Mizuno et al. (2000), it was observed that the specific
hydrogen production rate increased from 1.446 mLH2/min gbiomass to
3.131 mLH2/min gbiomass under nitrogen sparging conditions. With N2
sparging at a flow rate approximately 15 times the hydrogen
production rate, the hydrogen yield was 1.43 molH2/molglucose. This
meant an increase in hydrogen yield due to nitrogen sparging of
almost 50%. A report by Tanisho et al. (1998) revealed that sparging
with argon results in an increase of residual NADH, which might be
expected to increase hydrogen production. A hollow fiber/silicone
rubber membrane effectively reduced biogas partial pressure in a 48 dark fermentation system, resulting in a 10% improvement in the rate
of hydrogen production and a 15% increase in H2 yield (Liang et al.,
For real, the advantages of gas sparging are also connected to the
CO2 removal, because its accumulation may decrease the yield of
hydrogen due to electrons consumption for succinate and formate
synthesis via CO2, pyruvate and NADH (Das and Veziroglu, 2001).
Tanisho et al. (1998) stated also that CO2 partial pressure may have
higher inhibition effect to the dark fermentative hydrogen production
than H2 pp. Several attempts to remove CO2 have been made either
by inert gas sparging to drive out hydrogen and carbon-dioxide from
the reactor or by employing other membrane-based processes. Not
only inert gases like argon but also hydrogen itself was effective in
the removal of CO2. H2 sparging may also be economical because the
production plant won't need to separate the mixed gas if the produced
hydrogen will be used as the removing gas (Tanisho et al., 1998).
Thus, Tanisho et al. (1998) increased the hydrogen yield from 0.52
up to 1.58 molH2/molglucose by the combined effects of CO2 removal
and conditions of sufficient nitrogen source. Table 1.5 (Courtesy of
Nath and Das, 2004) provides an excellent overview of the state-of-
the-art technologies for hydrogen removal from a reaction system. Tab. 1.5 Different approaches of hydrogen removal. (Courtesy of Nath and Das, 2004). 49 - Nutrients and metal ions requirements
In general, hydrogen production in AD has lower requirements of
macro and micro nutrients than aerobic processes, due to the process
lower microbial biomass yield. Of course, carbon, nitrogen and
phosphorous are fundamental, with nitrogen that is especially
important not only for bacterial growth and multiplication but also
because the digestion of nitrogenous compounds could contribute to
the pH buffering of the bioprocess by releasing ammonium cations
(Liu and Shen, 2004). Even the balance between the nutrients is
fundamental, such as the C/N ratio which is reported to deeply affect
the hydrogen productivity and the process stability (Tanisho et al.,
1998). Lin and Lay (2004) demonstrated that increasing the C/N
ratio from 40 to 47, hydrogen production in mesophilic hydrogen
fermentation from sewage sludge was 5 times higher.
Among the micronutrients, the most influencing elements are sulfur,
vitamins and traces of minerals: many function of anaerobic bacteria
are strongly dependent on the availability of trace elements, since
they form part of the active sites of several key enzymes (Meulepas
et al., 2005). In particular, iron (Fe) shortage could influence the
growth, metabolism, and hydrogen-producing ability of B49 (an
anaerobic bacterium strain). Ren et al. (2003) suggest that adding
Fe 2+ could increase hydrogen enzyme activities (such as NADH-Fd reductase) and consequently enhance bacteria hydrogen-producing
ability. Indeed, ferrodoxin (Fd) is an iron-sulfur protein which
requires Fe and functions primarily as an electron carrier, being
involved in pyruvate oxidation to acetyl-CoA and CO2 and in proton
reduction to molecular H2. Fe is involved also in the control of lag
phase (increased by high iron concentrations, such as 100 mgFe/L
added in batch hydrogen production by Liu and Shen, 2004) and
metabolic pathways (butyric acid-type fermentation may turn into
ethanol-type by adding Fe; Wang et al., 2003b). 50 Another important influencing factor is magnesium (Mg 2+): its shortage may limit the growth anabolism of hydrogen-producing
fermentative bacteria (such as B49), and its hydrogen-producing
ability. Addition of Mg 2+ is reported to promote the growth of ethanol type hydrogen-producing fermentative bacteria and enhance
their hydrogen-producing ability (Liu et al., 2008). - Toxicants
Toxic compounds can be found in the feedstocks or can be produced
by microorganisms converting non inhibitory substances to
inhibitory ones. Among the toxicants, it is possible to find inorganic
substances (heavy metal cations, hydrogen sulphide, salts and
ammonia at relative high concentrations) and organic compounds
(polyphenols, furfural and hydroxyfurfural compounds). Their toxic
effect may be even increased by other process factors: for example
ammonia inhibiting effect is higher with higher pH, since this makes
easier the release of free ammonia into the medium (Meulepas et al.,
2005). As for the inhibitory organic compounds, they are commonly
generated during physicochemical or biological pretreatment of
lignocellulosic biomass. Also wastes originating from different
agricultural products typically contain polyphenolic compounds,
even frequently found in animal manure. These toxicants presence
could be a big issue also for a second stage fed with the effluent of a
first dark fermentative stage (see Chapter 4). Indeed, phenol, indole
and benzene, with and without substituents, were found by us in the
first stage effluent of a two-stage H2-CH4 producing reactor, deeply
studied and characterized in Paper III. Lastly, also oxygen could be
considered a toxicant for obligate anaerobic microorganisms (see the
following section) and in addition any highly oxidized material, as
nitrate or nitrite, can exhibit inhibition.

51 - Anaerobic conditions and reducing agents
Among the fermentative hydrogen-producing microorganisms,
obligate anaerobes such as Clostridium butyricum are extremely
sensitive to oxygen and their hydrogen-producing activities are
completely inhibited by the presence of a very slight amount of
oxygen in the feeding medium (Yokoi et al., 1995). Therefore
reducing agents such as argon, nitrogen, l-cysteine and the same
hydrogen gas may be used to remove trace amounts of oxygen
present in the medium and to decrease redox electric potential.
However, the use of such reducing agents is relatively expensive, and
therefore uneconomical for industrial biological production of H2.
Alternatively, Enterobacter aerogenes or other facultative anaerobes,
which have the ability to survive and work in the presence of slight
amount of oxygen within the bioreactor, may be exploited. Yokoi et
al. (1998) suggested the use of a mixed culture of strict and
facultative anaerobic bacteria where E. aerogenes rapidly consumes
oxygen, thus recovering immediately strict anaerobic condition
optimal for the more performing Clostridium bacteria.


53 Chapter 2 Biomasses for hydrogen production via Dark Fermentation

Biomass is considered an intrinsically safe and clean material, with
unlimited availability and high potential to be used as a renewable
source for the production of energy and alternative fuels, new
materials for technical applications and organic materials and
chemicals. In particular, biomass can be defined as ''the
biodegradable part of products, waste and residues from agriculture
(including vegetable and animal substances), forestry and related
industries, as well as the biodegradable fraction of industrial and
municipal waste' (Italian Legislative Decree 29/12/03, n. 387 -
Implementation of directive 2001/77/CE on the promotion of
electricity produced from renewable energy sources in the internal
electricity market).
On the basis of the increasing biomass exploitation by emerging
technologies, assumptions have been made about the capability of 54 biomasses to contribute for 15% of the total world energy demand in
2020. Calculations based on an optimistically estimated maximal
yield of 50 ton/ha for agricultural biomasses, indicate that an amount
of 50 Gton biomass/year (80% non-food biomass, 10% forestry and
10% waste streams) could be available for non food applications in
2040 (Okkerse and van Bekkum, 1999).
However, in order to really compete with fossil fuel-based energy
technology without any tax support, biomass-based energy systems
must achieve feedstock and processing advantages over fossil fuels.
Unfortunately, nowadays biomass energy production has costs that
are inherently high compared with the gas/oil/coal-fired electricity
generation processes, because:
- biomass fuels have low bulk density, they are expensive to gather,
process, transport and handle;
- biomass power generators are smaller than conventional generators,
which makes bioenergy generation economically disadvantageous;
- biomass energy technology is not as advanced and integrated with
the needs of the society as those oil/natural gas/coal -based (Akay et
al., 2005);
- whereas the oil-based chemical technology for converting fossil
feedstocks into a variety of useful products is very efficient and
mature, the technology for converting agricultural raw materials is
still in its infancy (Soetaert and Vandamme, 2005).
At the present, energy from biomass can be produced basically in
two ways: chemical decomposition through thermal processes and
biological conversion. Thermal processes (incineration in excess of
oxygen, pyrolysis and gasification) have the general disadvantage of
causing atmospheric pollution, unless costly purification of the
effluent gases is applied.
Thus, a valid alternative are the biological processes, which
essentially are anaerobic fermentative processes with production of
ethanol, methane and hydrogen (Lyberatos et al., 2005). As shown in 55 chapter 1, medium and long term strategies, from bioprocess
intensification and miniaturization to genetic engineering of
microorganisms and plants, have been developed for biologic
biomass conversion and new programs have been started in order to
harvest the benefits of the bio-based economy and to enhance the use
of biomass for energy and chemicals. Still, the cost and the retrieval
of substrates (biomasses) to be used in fermentative process is one of
the main expenditure items. A report of ITABIA (2008) (Table 2.1) summarized the quantity of
overall available biomass in Italy every year, considering different
organic residues, animal manure and dedicated energy crops,
amounting to 24-30 Mtoe (millions of equivalent tons of oil) per
year. This study stated also that the actual availability of biomass in
Italy, regardless of collection and supply problems, is about 80% of
potential availability, thus corresponding to 19-24 Mtoe/year.
According to Coldiretti association, this could guarantee a saving of
10-12 millions of tons per year in oil consumption, with a
simultaneous reduction of CO2 emissions of 30 million tons. Tab. 2.1 Million of equivalent tons of oil (Mtoe) yearly available from organic matter in Italy
(Courtesy of ITABIA, 2008). Making specifically reference to renewable hydrogen production,
this amount of available biomass would have the potential to make 56 the biohydrogen-based technologies cost-competitive to those natural
gas-based (by now cheaper).
The Department of Energy of the United States has recently set
ambitious goals about biomass conversion into hydrogen by dark
fermentation (Logan, 2004), e.g.
- to reach 50% conversion efficiency for hydrogen production from
- to reduce hydrogen costs from $6 to $1.50/kgbiomass;
- to extract high-purity hydrogen from biomass at relatively low cost
of $2.60/kgbiomass in short-term period.
However, waiting to fulfill these requirements economic
competitiveness of biohydrogen/bioenergy production could be
increased through biomass integration in relatively small scale (0.1-
20 MWe) applications, aiming at local needs satisfaction with
simultaneous environmental impact reduction.
Therefore, specific attention must be paid to biomass chemical
composition, because, as stated previously (see chapter 1), biomasses
rich in sugars and/or complex carbohydrates are the preferred ones
for fermentative hydrogen production, potentially achieving a 20
times higher production than fat-rich wastes (fat meat, chicken
skin...) and protein-rich wastes (Lay et al., 2003; see Table 2.2,
Courtesy of Show et al., 2011). Tab. 2.2 Comparison of maximum hydrogen yields of different substrates (Courtesy of Show
et al., 2011). Substrates Constituents Seed sludge Hydrogen yield conversion
(L H2/kg VS) Conversion efficiency
(%) References Carbohydrates Pure glucose Thermotoga maritima 497.8 100 Shroeder et al. 1994 Cellulose Sludge compost 298.7 a 60 Ueno et al. 1995 Starch Thermococcus kodakaraensis KOD1 414.4 b 83.25 Kanai et al. 2005 Proteins Peptone UASB sludge inhibited by chloroform 33.6 6.75 Liang et al. 2001 Egg Digested sludge 7.07 1.42 Okamoto et al. 2000 Lean meat Digested sludge 7.68 1.54 Okamoto et al. 2000 Lipids Chicken skin Digested sludge 10.1 2.05 Okamoto et al. 2000 Fat Digested sludge 11.2 2.23 Okamoto et al. 2000 aCalculated from the reported value of 2.40 mol H2/mol hexose
bCalculated from the reported value of 3.33 mol H2/mol hexose (starch) 57 In the following sections the types and the availability of the
exploitable actual biomasses will be explored in details: section 2.1
focuses on dedicated agro-zootechnical and lignocellulosic (from
agriculture, forests, energy crops) feedstocks; section 2.2 on
heterogeneous wastes and wastewaters, such as municipal solid
waste (MSW) and its organic fraction (OFMSW), sewage sludge and
industrial wastes.
2.1 Agricultural and livestock biomasses Yearly, a total of about 170 Gton biomass is worldwide produced
through photosynthesis in green plants and the energy value of this
annual production (2.74 x 10 19 Btus - British thermal units -) is eight times as much as the annual world energy consumption (3.4 x 10 18 Btus) (Lyberatos et al., 2005). However only 3.5% (6 Gton) of the
biomass produced is being cultivated, harvested and used. Of this,
62% (3.7 Gton) is consumed for the production of food, 33% as fuels
for energy and housing and just approximately 0.3 Gton (5%) are
used by non food industry (Eggersdorfer et al., 1992).
The word "energy crops" refers therefore to that 33% of biomass
cultivated for being further exploited (either whole or part of it) as
feedstock for energy production, i.e. energy gain through combustion
or biotransformation to biofuels. Ntaikou et al. (2010a) state that the
sustainability of such processes can only be assured if:
' the crops are produced at low cost, thus with minimum nutrient and water requirements; ' the crops are resistant to environmental stresses;
' the crops are highly biomass yielding;
' the crops have high sugar and/or carbohydrates'' content and low lignin content (specifically for hydrogen production via dark
fermentation). 58 If employed in wet processes, such as in anaerobic digestion, also the
water content is one of the key parameters for the choice of a specific
crop. Indeed, water is important for biological activity, since
nutrients must be dissolved in water before they can be assimilated,
and it enhances the mobility of microorganism, improving the mass
transport and the penetration/diffusion of microorganisms throughout
the substrate, thus facilitating the digestion process (Meulepas et al.,
2005). Moreover a water content usually higher than 35% (together
with a C/N ratio in the range of 20-30) make biomasses fit better to
bioprocessing wet technologies than to direct incineration, where
supplemental fuel would be required, proportionally to the water
content. Moisture, ash content and gross calorific values (CV) of
different solid biomass feedstock are given in Table 2.3 (Nath and
Das, 2003). Tab. 2.3 Moisture and ash content and gross calorific value of different biomass feedstock
(Courtesy of Nath and Das, 2003). However, the main criterion driving the adoption of a specific
biomass, especially for biohydrogen production, is its chemical
composition: energy crops can be divided in sugar based crops (e.g.
sweet sorghum, sugar cane and sugar beet), starch based crops (e.g. 59 corn and wheat), and lignocellulose based crops, including
herbaceous (e.g. switch grass and fodder grass) and woody crops
(e.g. Miscanthus and poplar). These crops can be processed by
means of so-called biorefineries into relatively pure carbohydrate
feedstocks, the primary raw material for most fermentation
processes, or directly employed in fermentative hydrogen production
processes. Therefore, the exploitation of lignocellulosic raw
materials, consisting of tightly bound lignin, cellulose and
hemicellulose, for hydrogen production via fermentation depends on
the capability of exploiting cellulose and hemicellulose and
simultaneously avoiding the lignin constituent. Indeed, the bonding
in lignocellulose resists mobilization, as lignin is not degraded under
anaerobic conditions and is often inhibitory to microbial growth (De
Vrije and Claassen, 2005). In view of producing cheap feedstocks for
dark hydrogen fermentation from lignocellulosic biomasses,
development of cost effective pretreatment methods with a low
energy demand must be studied.
Several energy crops (mainly starch- and sugar-based) have already
been used in laboratory-scale experiments, such as maize, rye,
Jerusalem artichoke, oat, sunflower, triticale, rape and wheat (Ahrens
and Weiland, 2004) and Table 2.4 (courtesy of Ntaikou et al.,
2010a), shows the maximum hydrogen production and yield
achieved via dark fermentation of different energy crops, with
noticeable results from pretreated Miscanthus.
In our Paper II, Maize silage showed a relatively interesting
biohydrogen production potential of 118 ± 2 NLH2/kgVS, almost the
half of that of the more promising market bio-wastes and organic
fraction of municipal solid wastes. 60 Tab. 2.4 Hydrogen production via dark fermentation from different energy crops (Courtesy of
Ntaikou et al., 2010a). As for animal (livestock) residues employment in biohydrogen
production via fermentation, due to their chemical characteristics
(alkaline pH, no carbohydrate content,...) they are commonly not
suitable for this process. Only a few studies have addressed the
potential use of swine manure as feedstock for biohydrogen
production and when used as a single substrate alone, very little
biohydrogen could be recovered from fermentation both at
mesophilic temperatures (Wagner et al., 2009), as well as at
hyperthermophilic temperatures, with production yields lower than 4
LH2/kgVS (Kotsopoulos et al., 2009). Our study (Paper II) confirmed
low biohydrogen production potential of swine manure, achieving
just 14 ± 1 NLH2/kgVS.
On the other hand, hydrogen yield as high as 200 LH2/kghexose (Wu et
al., 2009; Zhu et al., 2009) were obtained when swine manure was 61 added to glucose, reinforcing the hypothesis that livestock residues
are a suitable co-substrate to be fermented by a mixture culture in
addition to a carbohydrate-rich, promptly hydrolysable material.
Speculating on full scale hydrogen production via dark fermentation,
this co-digestion strategy seems particularly advantageous, as
demonstrated nowadays in Europe conventional AD process plants
which hardly use agricultural crops alone, while co-digest them with
manure or other organic wastes because of the positive influence of
manure as source of essential trace elements and buffering
substances for the biogas process (Holm-Nielsen and Al Seadi,
From an applicative point of view, over 65% of the biogas plants
now operating in Germany utilize energy crops and crop residues in
co-digestion with manure (Weiland et al., 2003). Eurobserv'ER
(2010) reported that the introduction of energy crops in co-digestion
strategies, made the European biogas production increase and, as a
consequence, made the electric energy production via biogas
employment in cogeneration jump forward of about 20.5% between
2006 and 2007. This is especially true for UK and Germany, among
the EU member countries, which have been playing the leading role
because of their renewable energy policies providing incentives for
farmers producing energy with small digestion units.
The same can be said for Italy, where, as indicated by ITABIA
(2008), nowadays crops dedicated to energy production (competing
with crops planted for food or industrial purposes) are limited to a
few thousand hectares of sunflower, soy and rapeseed for biodiesel
and other few thousand hectares of rapid growth poplars (Short
Rotation Forestry, SRF) located in Northern Italy. The Italian agro-
energetic system interest in biogas production by means of anaerobic
digestion of agricultural biomasses is focused, on one hand, on the
chance of giving integration to farmers' income and, on the other, on
the significant economic and social role that AD can play as it 62 effectively can contribute to mitigate and even solve the
environmental issues connected to the high concentration of
livestock production units that can be found in some Italian Regions
(e.g. Lombardy). The same remarks would be valid for biohydrogen
production via Dark Fermentation.
By an analysis of Italian territory, ITABIA estimated that
approximately 500,000 - 600,000 ha of arable land may be used to
grow energy crops for bioenergy production, and that about 100,000
ha of marginal land may be used to grow low input energy crops that
are ideal for both producing lignocellulosic biomass and biofuels.
Overall vegetable production (lignocellulosic, oleoplants, sugary
plants, etc.) can be estimated to be about 13-16 million tons (of fresh
matter)/year, corresponding to 4-5 Mtoe/year of primary energy.
In consideration of animal manures, in Italy the pig industry
alone consists of 9.5 millions animals and is mainly concentrated in
highly specialized districts. It can be estimated that in the Po Valley
area, more than 10 million tons of swine manure are produced yearly
on a territory of about 5000 km 2. As a whole, 330 million tons of liquid wastes are produced every year, but only a part of it is used for
anaerobic digestion, even if no other alternative competitive uses
exist (apart for the limited share of liquid waste that can be spread on
the fields) and if this is potentially a cause of environmental and
societal problems, when not properly managed. On the contrary, as
reported by the ITABIA report (2008), in Germany and The
Netherlands there is a liquid waste stock exchange where trading
rates range from 1.5 Euro/t within 5 km from the producing farm to
up to 5 Euro/t beyond 5 km. So, similarly to other Countries and considering the huge amount of
animal slurries, crops and their residues (223 million tons/year,
altogether; Piccinini - CRPA, 2008) yearly available in Italy, co-
digestion of these wastes for biogas or biohydrogen production via
fermentation is an advisable strategy for our Country. 63 However, despite the good biogas and biohydrogen productions
guaranteed by energy crops, the continuously rising food prices, the
sustainability doubts and the energy-equation challenges have
recently led to a backlash against their use as feedstocks for biofuels
generation. In particular a food-vs-fuel debate took place, since in
many countries huge agricultural areas have been turned into
feedstock "industries" for the production of chemicals, transport
biofuels and energy (Ntaikou et al., 2010a; 2010b).
Therefore, the second generation biofuels, produced by feedstocks
that are not competitive to edible crops, such as wastes and residues,
can be a valid solution to the "energy deviation" from the plants
primary function of supporting human dietary needs. Substrates used
for "second generation hydrogen" will be further discussed in the
following section.
2.2 Organic residues and wastewaters Wastes from agricultural and municipal processes should be the
preferred biomass for energy production. Indeed, if the biomass
employed in the bioprocesses would preferentially be derived from
wastes and discarded residues, this will not increase the pressure on
natural habitats or the conflicts with the global food availability and
the preservation of biodiversity.
Nowadays unused biomasses are mainly burnt, land-filled or
accumulated as excess biomass, potentially leading to leachates,
greenhouse gases emission, soil and water contamination and so on.
Moreover, the disposal of wastes is already an economic burden on
communities and industries.
So, creating a marketable product (biofuel) from wastes would
immediately make money by both reducing waste treatment costs
and recovering energy (in form of methane or hydrogen) and/or
valuable materials from their processing. An example about how
much energy could be theoretically available in wastewaters is 64 reported by Logan (2004), considering the United States wastewaters
- the organic content in wastewater produced annually by humans is
equivalent to 0.11 quadrillion of Btus (British thermal units) and
worth $2 billion (assuming 330 million people producing 230 L wastewaters/day, 300 mg biological oxygen demand (BOD)/L wastewaters, and 3.5 kcal/g BOD); - animal wastewaters have the potential for energy harvesting of an
additional 0.3 quad. Btus;
- food processing wastewater, the most readily available source
because of its high sugar content and low indigenous bacterial
concentration, could have an overall energy content of 0.1 quad. of
energy (assuming the exploitation of 5% of the total U.S. food
industries wastewaters, having an average organic matter content of
2 g chemical oxygen demand (COD)/L wastewaters).
Logan (2004) stated also that, in order to compete with current
electric power plant costs, a wastewater source annually containing
0.1 quad. would have to be harvested for less than $3.3 billion.
However, higher costs could be tolerated in the United States if they
are included in the $45 billion already needed over the next 20 years
for wastewater treatment infrastructure or if they are used to reduce
annual expenditures in the $25 billion wastewater industry.
To further maximize economic competitiveness of bioenergy
production processes toward common fossil energy strategies, wastes
should be digested on site, so that their exploitation by small-scale
power generation plants can compete with fossil fuel-based energy
production, advantaged for example by feedstock logistics (no
transportation from one place to another). An interesting report by ITABIA (2008) assessed the italian
availability of various (biomass) residues, mainly from the five most
relevant sectors: agriculture, forestry, agro-industry, wood industry
and urban waste. The estimated total quantity of organic residues and 65 by-products produced in Italy every year amounts to more than 25
million tons of dry matter (see Figure 2.1). Unfortunately only a part
of it can be used nowadays, due to:
1) Competition with the non-energy uses of the biogenic matter;
2) Problems with collection of materials and their subsequent supply
to the energy conversion plant. Fig. 2.1 Total annual quantity (k.ton of dry matter per year) of biomass residues in Italy. Table 2.5 presents detailed esteems for specific residues availability
in Italy. Specifically:
- Agricultural residues
Lignocellulosic wastes (straw, stalks, prunings, etc.) from agriculture
sector (herbaceous and woody plants) have an estimated available
quantity (excluding the share of existing but unusable residues) that
amounts to approximately 9.3Mt/year of dry matter;
- Forestry biomass
Forestry residues that can be used for energy are estimated by
analyzing actual firewood production: firewood production in Italy
today amounts to approximately 2.2. Mt/year of dry matter (4.5
Mt/year of wet matter) but a significant increase in resources (4.3
Mt/year in d.m.) for the wood-energy supply chain would be required
in order to reach the calculated 6.5 Mt/year from today''s production.

66 - Industrial residues
The overall availability of industrial residues, expressed as dry
matter, amounts to around 8.4 Mt/year, of which 3.9 Mt come from
agro-industry and 4.5 Mt/year from the wood industry;
- Urban wastes
Today, in Italy only 8% - 10% of the wastes of urban origin is used
as fuel, in compliance with Legislative decree 152/06. Indeed:
' Recycling produces almost 0.9 Mt/year of wet matter, which means about 100,000 t/year of dry matter. ' The maintenance of public greeneries yields over 9 Mt/year of wet matter, which means about 380,000 t/year of dry matter
(ITABIA, 2008). The organic fraction of urban solid waste (OFMSW) that can be
obtained from waste treatment plants (the existing ones and the
plants under construction) amounts to about 2 Mt/year, which
corresponds to about 400,000 t/year of dry matter. This value may
also quadruple if plants treating all national wastes were developed. 67 Tab 2.5 Italian biomass residues availability: type of wastes and their quantity. Nevertheless, availability of the waste materials and their cost are
just two of the main criteria for the selection of feedstocks to be used
in biohydrogen production.
The others are waste biodegradability, low concentration of
inhibitory to microbiological activity compounds, and especially
their carbohydrate content, as stated previously. For example,
wastewaters from food-processing industries and breweries, and SECTOR TYPE OF WASTE QUANTITY (k.ton of dry matter per year) AGRICOLTURE
SOFT WHEAT Straw 500 HARD WHEAT Straw 1,600 BARLEY Straw 380 OATS Straw 120 RICE Straw 550 MAIZE Stalks/Cops 3,100 TOBACCO Stalks 10 SUNFLOWER Stalks 350 GRAPEVINE Shoots 880 OLIVE TREE Wood, branches, fronds 800 APPLE TREE 90 PEAR TREE Branches 50 PEACH TREE Branches 150 CITRUS TREE Branches 480 ALMOND TREE Branches 95 HAZEL TREE Branches 85 ACTINIDIA Shoots 25 APRICOT, CHERRY, PLUM, TREE Branches 35 TOTAL Straw, stalks, stems, leaves, etc. 9,300 FOREST BIOMASS
HIGH FORESTS (broad-leaved trees, conifers) Branches, tops and small residues 1,800 COPPICE WOODLANDS (simple, compound) Whole plant 4,700 TOTAL 6,500 AGRO-INDUSTRY RESIDUES
SUGAR REFINERY Molasses, dry pulp, sludge 1,570 TOMATOES Peels and seeds 135 CITRUS FRUIT Pulp and peels 210 FRESH FRUIT Stones 35 DRIED FRUIT Peels 135 FLOUR MILLING Bran 185 PASTA INDUSTRY Part breaking off 60 RICE INDUSTRY Husk, chaff, starch, green grains, broken parts 520 OIL Virgin residues, exhausted residues 750 WINE Virgin pomace, exhausted pomace, grape stalks 300 TOTAL 3,900 WOOD INDUSTRY RESIDUES
PRIMARY WOOD PROCESSING Barks, wane, etc. 2,500 SECONDARY WOOD PROCESSING Sawdust, woodchips, etc. 1,700 PAPER INDUSTRY Pulp-paper, pulper 300 TOTAL 4,500 68 agricultural wastewaters from animal confinements are ideal
candidates for bioprocessing because they contain very high levels of
easily degradable organic material, which results in a net positive
energy or economic balance, even when heating of the liquid is
required (Angenent et al., 2004). In addition, they have typically a
high water content, which circumvents the necessity to add water for
bioprocessing them in wet technologies. Considering the plant-based raw waste materials, the following main
categories can be identified (Boeriu et al., 2005):
- Agricultural residues from primary agricultural production, such as
straw, bran, corn cobs, corn stover, foliage, and hulls;
- Agro-industrial wastes generated in the food production at various
links in the chain, as well as resources lost in inefficiency of food
uptake and animal production, vegetable, fruit garden-waste and
- forestry residues and grass.
The majority of them are mainly lignocellulosic materials (sugar
cane and sweet sorghum bagasse, corn stalks and stover, fodder
maize, wheat straw, etc.) that are either poorly valorized or left to
decay on the land and that are attracting increasing attention as an
abundantly available and cheap renewable feedstock utilized via
bioconversions (Soetaert and Vandamme, 2005). Ntaikou et al.
(2010a) state that around 2.9 x 10 3 million tons from cereal crops, 1.6 x 10 2 million tons from pulse crops, 1.4 x 10 million tons from oil seed crops and 5.4 x 10 2 million tons from plantation crops are produced annually worldwide. In Table 2.6 different types of
lignocellulosic residues used as feedstocks for hydrogen production
are reported, along with their biohydrogen yields and production
rates and the pretreatment adopted for their solubilization.

69 Tab. 2.6 Hydrogen production via dark fermentation with different types of lignocellulosic
residues (Courtesy of Ntaikou et al., 2010a). The second most abundant group of plant-based residues is that of
agricultural or industrial wastes containing starch and cellulose.
These feedstocks are easy to be used by AD processes, but have
different characteristics and properties influencing the process
parameter optimization. Starch containing solid wastes are generally
easier to process for hydrogen gas formation because starch can be
hydrolyzed to glucose and maltose by acid or enzymatic hydrolysis
followed by conversion of carbohydrates to organic acids and then to
H2. Many authors studied the suitability of starch-based residues for
hydrogen-production: lactate-containing wastewater, cow dung
slurry, vegetable starch, sugar-cane juice and whey, bean-product
wastewater, tofu wastewater have been extensively used for 70 biohydrogen production in lab-scale (Nath and Das, 2003). Claassen
et al. (2005) reported the use of potato steam peels, in form of a
highly viscose starch-rich slurry, obtained as a by-product in the
potato processing industry and commonly used in the fodder industry
(wet feed component).
One of the highest specific hydrogen production rate was 237
LH2/kgVSS d when edible corn starch was used as the substrate by C.
pasteurianum (Liu and Shen, 2004), while specific yield of 480
LH2/kgVSS with 4.6 g/L starch concentration at 37 °C using a mixed
sludge was obtained by Zhang et al., 2003. Yokoi et al. (2001) used
dried sweet potato starch residue (2.0% starch residue content) to
feed a mixed culture of C. butyricum and E. aerogenes: H2 yield
obtained in long term repeated batch operations was 2.4
Differently, cellulose containing wastes (paper wastes, agricultural
wastes - wheat straw, corn stover, rice straw, corn cobs...- and
others) require further pre-treatment and therefore are less favorable
to be used. Cellulose (and hemicellulose) content of wastes can be
hydrolyzed to carbohydrates and then further processed to hydrogen,
but very often these wastes must be preventively grounded and then
delignified by mechanical or chemical means before fermentation. Enlarging our interest to other wastes than those strictly plant-based,
heterogeneous and complex solid wastes or wastewaters have
already been tested as feedstocks for fermentative hydrogen
production even if, differently from the wastes previously described,
they own quite high content of proteins and fats together with
carbohydrates. The different origin and composition of these wastes
bring to general lower conversion (to hydrogen) efficiencies than
pure carbohydrates and Lay et al. (2003) stated that the hydrogen
production potential of carbohydrate-based wastes may be 20 times
higher than that of fat-based and protein-based waste. This is
possibly explained by the consumption of hydrogen towards 71 ammonium using nitrogen generated from protein biodegradation.
Table 2.7 shows sources of possible degradable waste streams which
are presently used for methane production and that, with an
appropriate setting of the operating parameters, can be exploited also
for hydrogen production. The waste streams are divided into solid
wastes, waste slurries and wastewaters. Tab. 2.7 Origin of organic waste streams that can be utilized for the production of biogas
(adapted from Weiland, 2000). These waste streams may require particular bio-processing
technologies to hydrolyze their carbohydrate fraction, to remove
undesirable components and for nutritional balancing. As an
example, hydrogen yields of 1.2 mgH2/gCOD were reported by Wang
et al. (2003a) when waste sludge (biosolids) generated in wastewater
treatment plants was used as the raw material, with higher yields (15
mgH2/gCOD) obtained from the sludge filtrate.
Moreover they may show different conditions for optimal
fermentation: for example, cheese whey, a rich in readily
fermentable sugars wastewater, reached its highest hydrogen yield Solid wastes Domestic Separately collected vegetable, fruti and yard waste The organic fraction of source-sorted household waste
Organic residual fraction after mechanical separation of integral collected household waste Agricultural Crop residues
Undiluted manure Waste slurries Domestic Primary and secondary sewage sludge Agricultural Liquid manure Industrial Slaughterhouse and meat-processing Fish-processing Wastewater Domestic Sewage, Black Water sewage Industrial Dairy, sugar, starch, coffe processing, breweries and
beverages, distilleries and fermentation, chemical,
pulp and paper, fruit and vegetable processing 72 and production rate both at pH between 6 and 7 (Davila-Vazquez et
al., 2008) and at pH in the range 4 - 5 (Yang et al., 2007).
However, among all wastes, the research activity on fermentative H2
production mainly focuses on food-related wastes, e.g. food-industry
wastes/wastewaters or municipal solid organic wastes. Even if these
show high variations in carbohydrate and protein types and
concentrations in the mixture, they are commonly high-strength
organic wastes, whose biotransformation to biohydrogen can be
considered particularly appealing from both the environmental and
the economic standpoint (Ntaikou et al., 2010a). Rice winery,
noodle, sugar, and molasses manufacturing, olive mill wastewater,
olive pulp, dairy industry, baker''s yeast, brewery wastewaters and
cheese whey were successfully tested for hydrogen production at
laboratory scale (Table 2.8; Courtesy of Ntaikou et al., 2010a). Tab. 2.8 Hydrogen production via dark fermentation from different types of waste and
wastewaters (Courtesy of Ntaikou et al., 2010a). The hydrogen yield obtained from food wastes/wastewaters range
from 0.7 and 2.7 molH2/molhexose, results quite comparable to those 73 from pure carbohydrates. An interesting study was made by Van
Ginkel et al. (2005a), who reported hydrogen production from four
different food-processing industries (two confectioners and apple and
potato processing industries). The H2 production rates obtained were
in the range of 0.1 and 2.8 LH2/Lwastewater, with potato wastewater as
the best performing waste.
Food wastes were also used in different studies aiming at optimizing
some of the factors regulating the process. Shin et al. (2004) reported
higher H2 production potential and production rates from food wastes
under thermophilic conditions than in mesophilic processes. The
effect of HRT on H2 production from food wastes was studied by
Han and Shin (2004a), who obtained 58% COD reduction and 70%
hydrogen formation efficiency at very fast retention time of 5.3 h.
Market bio-wastes were also used in our works individually (Paper
II), showing promising biohydrogen production potential of 176 ± 2
NLH2/kgVS, or co-digested with swine slurry (Paper I and III). In
particular market bio-wastes/swine manure ratio of 35/65 and HRT
of 2 d gave the highest production rate (among the tested conditions)
of 3.27 ± 0.51 LH2/Lreactor d, with a corresponding hydrogen yield of
126 ± 22 NLH2/kgVS added and H2 content in the biogas of 42 ± 5%. At
these operating conditions the process exhibited also one of the
highest measured stability, with daily productions deviating for less
than 14% from the average (Paper I). Municipal solid waste (MSW) is generally defined as "household
waste plus other waste of a similar composition collected by (or on
behalf of) the local authority". In practice, this means that if the
waste generated by a particular commercial business is collected
along the household waste the material is classed as MSW. The
combustion of MSW for energy production is an effective use of
wastes that significantly reduces the problems of waste disposal but
it generates greenhouse gas emissions. Alternatively, a fraction of
municipal waste can be composted, although its main part is 74 typically landfilled, which is being severely curtailed due to
unavailability of land and environmental concerns. So, AD processes
seem to be a valid and environmental-friendly alternative.
The biological hydrogen production from organic fraction of
municipal solid wastes (OFMSW) is quite promising, since this
waste can represent up to 70% of the total MSW produced,
consisting of paper (up to 40%), garden residues, food wastes and
wood. With specific regards to Italy, in 2008 it was reported that the
total municipal solid waste (MSW) production in Italy is about 2
Mt/year (ITABIA, 2008). The process feasibility was investigated by
many authors, such as Lay et al. (1999) who used different mixed
anaerobic microflora under mesophilic conditions for OFMSW
digestion. Varying the "food-to-microorganisms" ratio they reported
that at 0.4 g OFMSW/g bacterial biomass, high hydrogenic activity took place,
with 43 LH2/kgVSS h specific production rate and 125 LH2/kgVS h
production potential. Okamoto et al. (2000) reported a hydrogen
production of 19.3-96.0 LH2/kgVSadded by mesophilic batch
fermentation of MSW composed mainly by rice and carrots.
Similarly, Valdez-Vazquez et al. (2005) reported a yield of 95
LH2/kgVSadded with semi-continuous CSTR treating municipal organic
In our research (Paper II), the biohydrogen potential (BHP) test
applied to a representative sample of OFMSW collected in northern
Italy achieved very high hydrogen yield (202 ± 3 LH2/kgVSadded) in
thermophilic condition. Considering also the huge amount available
of this waste, OFMSW could represent a huge source of renewable
energy if AD and hydrogen economy would take place in our
Country. To gain an insight into possible production rates of real
scale (CSTR) plant, typically digesting highly concentrated organic
mixtures (50''150 gVS/L), a concentrated organic mixture of OFMSW
(126 gTS/L) was also tested in Paper II through a continuous lab-scale
fermenter. Hydrogen production (60 ± 4 NLH2/kgVS added), although 75 comparable with previously mentioned results, resulted in only 30%
of its BHP, showing that further improvements are still needed for
future full-scale applications of dark fermentation.


77 Chapter 3 Biohydrogen production via bioelectrochemical system: the MEC application

3.1 General operating principles Bioelectrochemical systems (BES) use the catabolism of living
microbial cells to convert reducible organic materials to:
- electricity, in microbial fuel cells (MFCs);
- hydrogen, in microbial electrolysis cells (MEC);
- other products (Rabaey and Rozendal, 2010).
These systems, based on the bacteria ability to generate electric
potential and firstly exploited in MFCs, result nowadays in a number
of different applications and could be overall named "MxCs" (Logan,
2010). For example, it is possible to include those systems that,
through the addition of external voltage to the potential generated by 78 the bacteria, generate other products than hydrogen, such as methane
and hydrogen peroxide (Cheng et al., 2009; Liu et al., 2005;
Rozendal et al., 2006a) and those MFCs that through a specific
technique use membranes to allow for simultaneous water
desalination and electrical power production (Cao et al., 2009; Kim
and Logan, 2011). The first connection between electricity and biology was made in
1791 by Luigi Galvani, who discovered that severed frog''s legs
contract every time the muscle and nerve endings are connected to a
static electricity generator (Wrana et al., 2010). More than a century
spent since, in 1911, M.C. Potter discovered the ability of certain
bacteria to produce electrical current (Potter, 1911): however no
practical applications of this ability was firstly assumed. Then
followed experiments about bacteria-electrode interaction (by
examining overall potentiometric intensities of chemical reactions
during bacterial growth; Cohen, 1931) and, during the '50s and '60s,
about converting organic material into electrical energy. For real, it
was just with the '70s, under the boost of US space program which
saw MFCs as potential waste disposal units that could generate
power during space missions (Shukla et al., 2004), that this new
biological technology began to be explored. During those years it
was already well known that when bacteria oxidize a chemical,
electrons are captured and transferred to a series of respiratory
enzymes used to store energy (in the form of ATP) within the cell.
These electrons are then released to an electron acceptor such as iron,
nitrate, sulfate, or oxygen. This biochemical property let the
scientific community to assume that the same bacteria that can
respire using iron could be able to transfer electrons to an insoluble
acceptor, such as an electrode, acting as a catalyst for oxidizing the
organic matter (Shukla et al., 2004). However, despite an improved
understanding of the biological mechanisms, the laboratory
experiments of the 70's showed several discouraging results: 79 maximum power densities reached were low, chemical mediators
used to ease the transfer of the electrons from the microorganisms to
the electrodes had to be often added into the medium and were toxic,
and rich media were needed to cultivate the bacteria (Allen, 1972).
Essentially, it was determined that current could not be produced at a
rate or quantities large enough to be a viable source of electrical
energy (Wrana et al., 2010) and the research on this field was
(temporarily) stopped.
Finally, several events happened at the same time that completely
changed the prospects for electrical current generation by microbial
fuel cells (MFCs). It was discovered that:
' mediators did not need to be added into solution as bacteria capable of direct extracellular electron transfer to fuel cell
anodes were found (Kim et al., 1999); ' wastewaters could be used as a source of fuel, while accomplishing wastewater treatment (Liu et al., 2004); ' much higher power densities were attained through cell architecture and parameters optimization (Rabaey et al., 2003). Hereinafter these features, together with the basis of the BES
technologies, will be discussed.
A MFC acts very similarly to a conventional fuel cell, where
hydrogen gas is injected into the anode chamber and usually is split
on a platinum-coated electrode into protons and electrons (see
chapter 1). In a MFC organic matter is used instead of hydrogen and
its breakdown into protons and electrons (and CO2) is accomplished
by electrochemically active microbes growing on the surface of the
anode. From here the electrons and protons travel through an
external circuit and electrolyte solution, respectively: the former
creates an electric current, while the latter uses oxygen as the
electron acceptor at the (usually Platinum-coated-) cathode
generating just water as reaction product (Figure 3.1). 80 Fig. 3.1 Single chamber air cathode microbial fuel cell showing the anode, where bacteria form
a biofilm on the surface, and the cathode, which is exposed to the air (Courtesy of Liu et al.,
2010). Specific bacteria, naturally present in waste organic matters, are
involved into the process and permit to avoid the expensive addition
of chemical mediators to the culture. These microorganisms capable
of exocellular electron transfer have been previously referred to as
electricigens or anode-respiring bacteria and are now widely known
as exoelectrogens (Liu et al., 2011). Even if generally there is a
predominance of the same type of microorganisms on the anode
surface, the microbial community appears to be highly dependent
upon the design and the operational parameters of the MFC system
and even upon the culture technique (batch or flow-through mode).
For example, in systems harvesting electricity under highly anoxic
conditions, Geobacter species were found to be the predominant
organisms involved in electricity production, while in other systems,
where the reactor design permits substantial leakage of oxygen into
the anode chamber, organisms more tolerant to oxygen exposure
could predominate (Lovley, 2006; Rabaey et al., 2004).
All these bacteria rely on three different strategies of extracellular
transport of electrons, well described by Wrana et al. (2010) (Figure
3.2): 81 a. Direct electron transfer using outer membrane c-type cytochromes; b. Long range electron transfer via biogenic soluble mediators (shuttle); c. Long range electron transfer via conductive bacterial appendages (conductive pili or microbial nanowires). a. Microbial strains with mutations or deletions in some genes that
encode for outer membrane cytochromes and for direct and indirect
mineral reduction (cyma and omcB for S. oneidensis; omcS and
omcE for G. sulfurreducens) gave evidence of the role of these c-
type cytochromes in direct electrical contacts between microbe and
electrode (Holmes et al., 2006; Lies et al., 2005). b. Certain microorganisms produce soluble exogenous mediators that
shuttle electrons from cells to insoluble compounds via diffusion
(Wrana et al., 2010). Specifically, riboflavin secretion for this
transfer has been observed both in Shewanella sp. and Geothrix
fermentans. The main disadvantage of this mechanism is that it is
energetically taxing and may not be the most desirable system for the
bacteria (Mahadevan et al., 2006). c. Unlike other pili that aid bacteria in cell motility or adhesion to
solid surfaces, nanowires are electrically conductive protein
filaments (composed of the repeated single unit PilA; Reguera et al.,
2005), that enable communications between microorganisms.
Therefore they are responsible for maximizing biofilm health by
coordinating a cooperative electronic community and by aggregating
and interconnecting cells into a network capable of effectively
distributing and dissipating electrons (Wrana et al., 2010; Figure
3.2). In MFCs they enable communications both between the
microorganisms composing the biofilm and between the biofilm and
the electrode. 82 Among the mechanisms described, one may be dominant in the
electron transfer, but it is known the possibility of the involvement of
all three mechanisms in electron transfer by a single bacterial specie. Fig. 3.2 Mechanisms for electron transfer from the exoelectrogenes to the anode: i) Long-
range electron transfer via electron shuttles (yellow hexagon), ii) direct electron transfer via
outer-surface c-type cytochromes (red circles), and iii) long-range electron transfer via
conductive pili or ''microbial nanowires' (orange rods). (Courtesy of Wrana et al., 2010). The substrate exploitable by MFC is another advantage of this
technology, as current can be produced by using various organic
matters, from the simpler ones, such as acetate, lactate, and glucose,
to more complex materials, such as domestic and industrial
wastewaters. In 2004, for the first time microbial fuel cells were
successfully applied to real (domestic) wastewaters treatment,
simultaneously generating electricity (26 mW/m 2) and removing up to 80% of the BOD win the wastewater (Liu et al., 2004). In the
following years many other wastewaters have been successfully
exploited through MFC technology (Table 3.1, Courtesy of Logan,
83 Tab. 3.1 Real complex wastewaters used in MFC and the achieved power density (Courtesy of
Logan, 2005). Lastly, just in recent years, microbial fuel cells with an enhanced
power output have been developed, providing possible opportunities
for practical applications (Rabaey and Verstraete, 2005). Since 1999,
power production has increased by five to six orders-of-magnitude
(based on projected surface area; Logan, 2010). Power densities of
MFCs using oxygen have reached from 2.7 W/m 2 (power normalized to the cathode; Xing et al., 2008) up to 6.9 W/m 2 (much larger cathodes than anodes, power normalized to the anode area; Fan et al.,
2008). However, in flow-through systems required for treating large
volumes of liquid waste, the highest power outputs are less than 2
W/m 2 anode surface, even when treating readily degradable pure substrates like glucose (Lovley, 2006). This power output is unlikely to be
sufficient to recover the power expended in pumping the fluid
through the system. Therefore, scaling these laboratory systems to
the size that would be required to handle large volumes of wastes is
still an issue.
These preliminary remarks are useful to better understand MEC
(microbial electrolysis cell) technology, which represents the next
generation of the MFCs, converted into an innovative way for
biological hydrogen production. They also share some operational 84 and functional advantages of MFCs, such as direct and high efficient
conversion of organic substrate energy into usable energy (H2) or
efficient operation at ambient, and even at low, temperatures
(differently from all the current bio-energy processes).
MECs were independently realized by two research groups (Liu and
Logan, 2005; Rozendal and Buisman, 2005) just a few years ago and
firstly named as BEAMR (bio-electrochemically assisted microbial
reactor) or BEC (biocatalyzed electrolysis cells) (Liu et al., 2010).
MECs share many attributes with MFCs because the design of the
anodes and the electrogenic reactions occurring are similar.
However, differently from a MFC, in a MEC the protons generated
on the anode surface by biomass breakdown by microbes combine
between themselves and the electrons released at the cathode to
generate hydrogen (Figure 3.3). Fig. 3.3 Single-chamber microbial electrolysis cell with a power supply as the driving force for
electron flow from anode to cathode (Courtesy of Liu et al., 2010). Another operational difference is that reaction at the cathode must
occur in an anaerobic environment. This simplifies cathode design,
but, since the product is a gas rather than electricity, the cell
architecture must be modified for collecting this gas (see details in
section 3.2) (Logan et al., 2008). Moreover, while in MFC oxygen
diffusion into the anode chamber substantially reduces recovery of 85 electrons from substrate as current (defined as Coulombic Efficiency,
CE), in MEC the lack of oxygen results, on average, in greater CE. A
fully anoxic MEC also enables a better growth of strict anaerobes,
but simultaneously the lack of exposure of the microorganisms to
oxygen enhances the likelihood for methanogenesis, which can lower
hydrogen recovery (Call and Logan, 2008; Rozendal et al., 2006a).
More important, if in MFC, due to the higher redox potential of
oxygen than that of the microbial anode, electrons flow
spontaneously from the anode to the cathode generating electricity,
in MEC the reduction reaction of H + ions to H 2 at the cathode has a lower redox potential than the anode, thus the electrons do not flow
spontaneously through the circuit and the hydrogen gas generation is
not spontaneous. Therefore MECs require the addition of energy to
overcome the thermodynamic limitations set by the chemical
reactions at the electrodes and by the potential losses occurring
within the system.
In details, under standard biological conditions (T = 25 °C, P = 1 bar, pH = 7) the Gibbs free energy of reaction ('' ') for acetate oxidation to hydrogen is positive and therefore acetate cannot be
fermented to hydrogen: CH3COO - + 4 H 2O '' 2 HCO3 - + H+ + 4 H 2 ( ''G ' '= +104.6 kJ'mol) Additional energy added to the system in order to favor hydrogen
evolution is supplied by applying a small voltage, named applied
voltage (Eap). In order to drive the H2 production, the applied voltage
needs to be at least larger than a value, usually referred as the
equilibrium voltage (Eeq), which could be calculated as: E = '' ''G ' ' nF where n is the amount of electrons involved in the reaction, and F =
96 485 C/mol e - is Faraday''s constant. 86 Alternatively it could also be calculated as the theoretical overall cell
potential expressed as an electromotive force (Eemf), by evaluating
the difference between the cathode and the anode theoretical
potentials (Eelectrode): Eemf = ECat - EAn Each Eelectrode can be calculated from tabulated values under standard
conditions by using the Nernst equation, which is: E = E '' RT nF ln [reduction] [oxidation] where R is the universal gas constant (8.31447 J/mol K) and T is the
absolute temperature.
Following the example of acetate oxidation to H2 and considering the
following two half-reactions occurring at the anode and at the
cathode (under standard biological conditions), ANODE: CH3COO - + 4 H 2O '' 2 HCO3 - + 9 H+ + 8 e- ; CATHODE: 2 H + + 2 e- '' H 2 , Eeq is very low: E #$%&#&% = '' 104.6 ' 10. 8 ' 96485 = (''0.414 V) '' (''0.279 V) = ''0.14 V The negative sign indicates that the reaction is not spontaneous and
that a voltage has to be applied in order for the reaction to proceed;
also, this value translates to a theoretical energy requirement of 0.29
kWh/m 3 H 2. For real, depending on the substrate consumed at the anode by the
bacteria and on the operating conditions, including hydrogen partial
pressure, pH, resistance in the system and polarization at the
electrodes, the Eeq could vary.
Also, under typical operating conditions, the applied voltage (Eap) is
always larger than Eeq because of internal losses in the system.
These losses (anodic and cathodic overpotential and ohmic losses) 87 are all a function of the current, are comparable to those observed for
MFCs and they will be explained in details in section 3.2.
Experiments have shown that the microbial electrolysis reactions
typically start to occur at Eap above 0.2 V, which corresponds to an
energy requirement of 0.43 kWh/m 3 H 2 (at 100% cathodic hydrogen recovery; Logan et al., 2008). For real, even applied voltages lower
than 0.3 V may result in a low hydrogen-production rate and erratic
system performance, while Eap > 1 V are not recommended because
the electrical energy input would be equal or larger than that of the
water electrolysis process (1.2 V; Liu et al., 2010).
Eap is typically provided by a power supply unit or a potentiostat, two
different external power source devices that are able to provide the
low voltages required by MEC (Logan et al., 2008). Evaluating the performance of this novel technology, recently the
overall MEC efficiencies and hydrogen production rates have
increased. Cheng and Logan (2007) reported that hydrogen
production yield can reach high value of 2.0 - 3.95 molH2/molacetate
(representing the 50-99% of the theoretical maximum) by improving
the materials and reactor architecture in a single chamber MEC.
Also, increasing the applied voltages from 0.2 to 0.8 V, the
production rate increased from 0.03 to 1.5 m 3 H2/m 3 reactor (total volume) d, simultaneously reducing the time needed for a complete batch cycle
(from 30 h to 3 h). At the same time the energy efficiency (η) of the
system ranged from ηW = 681-243% (when evaluated in terms of
only the voltage addition) to ηW+S = 62-86% (when evaluated on the
basis of both the voltage added and the heat of combustion of the
acetate added). The biogas produced was nearly pure hydrogen (''
99.5%) with only trace amounts of CO2 and CH4 in all experiments.
The same authors reported hydrogen production in MEC at high
yields from a variety of substrates (Table 3.2; Courtesy of Cheng and
Logan, 2007). Glucose was converted to hydrogen gas at a rate (1.23
m 3 H 2/ m 3 reactor (total volume) d) similar to that of acetate but at a lower 88 overall maximum recovery of 71% (8.55 molH2/molsubstrate). This
recovery is 4- to 5-times larger, however, than that typically achieved
through cellulose fermentation (Cheng and Logan, 2007). Even all
the predominant acids typically produced by glucose fermentation
(acetic, butyric, lactic, propionic, and valeric acid) were successfully
used to generate H2 in the electrohydrogenic process at energy
recoveries (ηW+S) of 66-82% (Table 3.2). Tab. 3.2 Hydrogen production using cellulose, glucose, or five different organic acids at an
applied voltage of 0.6V (Courtesy of Cheng and Logan, 2007). Improving the single-chamber membraneless MEC architecture, by
using graphite fiber brush anode and acetate as feeding, in 2008 Call
and Logan achieved a very high cathodic hydrogen recovery (78-
96%) and production rate (3.12 ± 0.02 m 3 H 2/ m 3 reactor (total volume) d). This production rate is more than double than that obtained in
previous MEC studies and demonstrates how fast this technology is
improving towards higher production rate (which has increased more
than 100-fold in less than 5 years) and higher recover efficiency. An
updated overview of the major MEC systems reported in literature
has been summarized together with their performance and the
adopted parameters (applied voltage, cathode type, membrane,
substrate...) by Liu et al. (2010) (Table 3.3).
As Table 3.3 shows, the current density generated by the single-
chamber MECs (4.2-12 A/m 2) is generally higher than that by the two-chambers MECs with membranes (0.4-3.3 A/m 2). The highest value of 12 A/m 2 has been achieved in a single-chamber MEC with 89 equal-sized anode and cathode and it is almost the half of the highest
microbial anodic current density (26 A/m 2) obtained in MFC with carbon cloth anode and a large air cathode (Fan et al., 2008). The
same Table shows also that MECs energy efficiencies based on
electricity input (ηW) range typically between 114% and 270%, and
they are expected to increase through further reduction of internal
Thus, big efforts are needed in improving the architectural MEC
designs and the materials used in the process (aiming at increasing
the electrode-area-to-volume ratio) and in avoiding those factors
potentially limiting H2 production (such as hydrogen consumption by
hydrogenotrophic methanogens or the direct acceptance of electrons
from the cathode by electromethanogenic bacteria). Indeed, although
exoelectrogens generally can outcompete methanogenic bacteria for
acetate on a MEC anode, a significant amount of methane can be
detected in single-chamber MECs after a few weeks operation,
especially with the use of wastewaters as substrate (Call and Logan,
2008; Hu et al., 2008). Also in our work (Paper IV), where single
chamber MECs with different cathodes (stainless steel, MoS2 and
Platinum) were fed with industrial wastewater with high-content of
methanol, persistent and high methanogenesis (always > 55%) was
detected. This persuaded us to compare MEC with (simulated)
anaerobic digestion process.
In the next section (3.2) some of the main MEC regulating factors
will be presented, together with recent studies aiming at their
optimization. 90 Tab. 3.3 Components and performances of microbial electrolysis cell system by microbial cell
design (Courtesy of Liu et al., 2010). 3.2 Main process regulating factors and their
- MEC architecture and its influence on performance
MEC could have several different architectures, but typically they
are a two-chamber system with the anode chamber separated from
the cathode chamber by a membrane. Figure 3.4 shows how many
different designs, used for batch or continuous tests, have been 91 developed in the past few years. The cells can be bottle-type, cube-
type, disc-type or rectangular-type, with both the electrodes soaked
in solution and collection of the hydrogen produced in the headspace
of the cathode chamber.
Model A and E (Figure 3.4) are the first systems designed and thus
are not optimized for performance: in particular the H-type reactor
(A) developed by Liu et al. (2005) had a high internal resistance
caused by the large anode-to-cathode distance (with the electrodes
placed in two distinct bottles) and by the small size of the membrane
between them. However the authors reached yield of 2.2
molH2/molacetate (CE = 60%; Eap = 0.25 V).
Model B represents a larger reactor that increased the anode surface
area using graphite granules and reduced the electrode spacing, but
that didn't increase performance.
With model D (Cheng and Logan, 2007) performances were really
increased, due to the adoption of a larger membrane surface, relative
to the electrode-projected surface areas, to the membrane type
(which allowed charge transfer via phosphate buffer anions), to the
higher anode surface area (graphite granules) and to the architecture
(a cube-shaped reactor with small electrode spacing). The advantages
of this design include the production of relatively pure hydrogen gas
in the cathode chamber, the reduction of biocontamination of cathode
metal catalysts and the control of different microbial species or
communities in the anode and the cathode chambers, which may be
especially required for MECs with microbial biocathode (Liu et al.,
Coming from model D alteration, a cubic-shape MEC (model C)
without a membrane, thus single-chamber type, was firstly proposed
by Call and Logan (2008), in order to further simplify the reactor
design and to reduce the capital costs. Removing the membrane
brings some advantages that will be dealt in the "Membranes"
section of this chapter. This cell also adopted a graphite fiber brush 92 anode to provide a high surface area for the exoelectrogenes, and a
cathode placed in close proximity to the anode. Improved current
densities were reported for this model, which also reached a
maximum hydrogen production rate of 3.12 m 3 H2/ m 3 reactor (total volume) d at an applied voltage of 0.8 V over a fed-batch cycle time of 12 h,
together with high CEs (92% average) and cathodic hydrogen
recovery between 78% and 96%. The main disadvantage of this
design is the likely hydrogen consumption by methanogens growing
on the cathode or in the solution, especially with actual wastewaters
used as feeding. Therefore further work is needed to investigate the
long-term stability of this system and to avoid methane generation.
Aiming at real scale scalability of MEC technology (no membrane,
no buffer or amendments addition and wastewaters exploitation),
model C was adopted in our study (Paper IV).
Other novel versions of the single-chamber MEC have been recently
presented (Figure 3.5; models H, I and J), which use different
materials for the main body (plastic tube, a glass bottle, a glass
tube...) and aim at miniaturizing the system (Hu et al., 2008; Hu et
al., 2009; Lee et al., 2009). Although no membranes are used in
single-chamber system, a separator, such as a J-cloth, is typically
used to avoid a short circuit when the electrode distance is reduced to
decrease the internal resistance (Liu et al., 2010). As for the models used for continuous flow tests, Rozendal et al.
(2007) examined a MEC design using a membrane electrode
assembly (MEA) architecture, i.e. a membrane integrated with the
cathode and a platinum catalyst (1 g/m 2) layer faced a gas collection chamber (Figure 3.4, model F). This eliminated the liquid
surrounding the cathode and reduced the reactor volume of the cell.
This study achieved a hydrogen production rate of 0.3 m 3 H2/ m 3 reactor (total volume) d with an Eap of 1.0 V. 93 Fig. 3.4 MECs used in different studies.
Types used in fed-batch experiments: (A) H-type construction using two bottles (320 mL each)
separated by a membrane; (B and D) two cube-type MECs (512 and 42 mL, respectively)
where anode and cathode are separated by a membrane; (C) cube-type single chamber (28
MEC lacking a membrane.
Types used in continuous flow tests: (E) Disc-shaped two-chamber MEC (each chamber 3.3
L); (F) disk-shaped membrane electrode assembly MEC (3.3 L) with gas diffusion electrode;
(G) rectangular-shaped MEC with serpentine-shaped flow channels through the reactor that
allow the gas to be released at the top of each flow path (each chamber 280 mL) (Courtesy of
Logan et al., 2008). Figure 3.5 Single-chamber microbial electrolysis cell (MEC) systems: (H) Bottle MEC with
carbon cloth anode and cathode; (I) Bottle-type MECs with graphite rod electrodes;
type MEC with carbon cloth anode and cathode separated by cloth (Courtesy of Liu
2010). type construction using two bottles (320 mL each) type MECs (512 and 42 mL, respectively) type single chamber (28 mL) chamber MEC (each chamber 3.3 shaped membrane electrode assembly MEC (3.3 L) with gas diffusion electrode; flow channels through the reactor that flow path (each chamber 280 mL) (Courtesy of ) Bottle MEC with Cs with graphite rod electrodes; (J) Tube- (Courtesy of Liu et al., 94 Another continuous vertically-orientated flow reactor architecture
was recently tested, together with an innovative biocathode (Figure
3.4, model G). The cell contained many parallel flow channels and
the produced hydrogen gas was continuously collected from the
headspace at the top of each flow channel, while development of
stagnant areas on the electrode, potentially inhibiting biocathode
performance due to pH increase, was prevented. - Electrode materials
Electrodes should be highly conductive (with low overpotentials) and
non-corrosive, should possess a high specific surface area, should be
non-fouling, inexpensive, easy to fabricate, scalable and versatile in
morphologies, and should assure a good adhesion to bacteria, in
order to achieve good electrical connection (Logan, 2008; Wrana et
al., 2010). Carbon and graphite meet many of these properties and
therefore are the most used materials. ANODE: Although carbon materials tend to corrode at high
potentials on the oxygen electrode side during water electrolysis,
they are chemically stable under the anaerobic anodic process in
MECs (Liu et al., 2010). Carbon materials with high surface area
have already been used in MEC as anodes able to increase the
number of bacteria attached to the surface and thereby increasing
anodic current output. They include carbon cloth, carbon paper,
graphite felt, graphite granules (with a graphite rod inserted into the
bed of granules as a current collector), and graphite brushes (Logan
et al., 2008). For the brush, the graphite fibers are usually connected
to a core made by twisted wires of a conductive and noncorrosive
metal (such as titanium or stainless steel) while for the other shapes
the electrode is pressed or glued to an insulated wire. Graphite brush
and granules are widely used, especially because they maximize the
surface/volume ratio and enhance the biofilm surface. However,
there is also a limit to this because small pores can become clogged 95 rapidly by bacteria, which could die off and hence decrease the
active surface of the electrode before lysis (Rabaey and Verstraete,
2005). Compared to other anode materials, these two anodes set two
main challenges:
- they can add electrode resistance to the system (due to loosely
packed graphite granules that may be disconnected and to the lack of
contact between the fibers in the brush)
- they can bring to large average distance between the anode and the
cathode: reducing this distance is the most direct way to reduce
internal resistance due to low proton concentration in MEC system
(Liu et al., 2010).
Cheng and Logan (2007) filled the anode chamber of a two-
chambers MEC with graphite granules, increasing total electrode
surface area to 528 cm 2 (assuming an average particle size of 4 mm). At an applied voltage of 0.6 V and coulombic efficiency of 88%, 3.5
molH2/molacetate were generated. The same authors (Cheng and Logan,
2007b) suggested also to pretreat the carbon-based anodes with a
high temperature ammonia gas process to chemically modify their
surface and thus increase their performance (e.g. faster start-up and
higher current densities in MFCs, probably due to a more favorable
adhesion of microorganisms to the positively charged anode). Other
researchers developed improved anode materials, by impregnating
them with chemical catalysts: for example Park and Zeikus (2003)
used manganese modified kaolin electrodes, yielding power outputs
up to 788 mW/m 2. CATHODE: Similar to the cathode of a water electrolyzer, a MEC
cathode normally consists of metal catalysts and catalyst-supporting
materials. A catalyst is a substance that increases the rate of
hydrogen production at the cathode surface by reducing the
activation energy barrier considerably, thereby lowering the
"cathodic activation overpotential" (see the next section within this
paragraph). The reaction involved leaves the catalyst unchanged. 96 Platinum (Pt) is well known as the best catalyst material for this
reaction and is commonly used in MEC systems: platinum catalyzed
electrodes are commercially available or prepared in the laboratory
by mixing platinum with a chemical binder (for example 5% Nafion
solution or 2% PTFE solution; Logan et al., 2008) and then applying
this mixture to one side of the cathode base (for example carbon
paper). The high cost of Pt and its being subject to be poisoned by
chemicals such as sulfide, easily to be found in common
wastewaters, are unfavorable factors to the scale-up of MECs and
rouse the search of better and cheaper alternatives to platinum-
catalyzed cathodes. Nevertheless, differently from water electrolyzers, MECs operate under complicated biological and
chemical conditions (presence of bacteria, bacterial metabolites, pH
buffers and medium nutrients) and therefore it's not easy to find cost-
effective alternatives, able to guarantee performances comparable
with Pt catalyst under realistic MEC operative conditions.
Recently, some good results were achieved, using non-precious
(costs about one to two orders of magnitude lower than Pt ) metal
catalysts such as nickel oxide (Selembo et al., 2009), nickel alloys
(Hu et al., 2009), tungsten carbide (Harnish et al., 2009) and
stainless steel. Compared to conventional platinum catalyst, stainless
steel made the cathodic hydrogen recovery improve from 47% to
61% and the overall energy recovery from 35% to 46% (Selembo et
al., 2009). As for the metal-alloy catalysts, they normally exhibit
better catalytic capability than that of their pure components, due to
synergistic electronic effects among alloys, but they are also easily
prone to reduction in mechanical stability (with related decreasing
Another recent alternative is the development of low cost
biocathodes (i.e., absence of an inorganic metal catalyst) that use
bacteria as cathode catalysts for hydrogen evolution in MEC. To
develop this biocathode, Rozendal et al. (2008) proposed to create an 97 electrochemically active culture by enriching a biofilm of hydrogen-
oxidizing bacteria on the anode and then to reverse the polarity of the
electrode to obtain an active biocathode.
In our study (Paper IV), graphite fiber brush non-ammonia-treated
anode and three different cathodes (carbon base layer plus platinum
or molibdenum disulphide or stainless steel) were used to compare
performances of MEC fed with high-COD industrial wastewater.
Despite better performances of Pt and substantial methane
production in all the MEC used, MoS2 proved to be a valid
alternative to Pt at the cathode, much more affordable for pilot- or
real-scale appliances. - Membranes
MECs may contain or not a membrane, distinct architectures called
two-chambers and single-chamber MEC, respectively. A membrane
is used to create a chamber where the microorganisms degrading the
organic matter are kept separated from the cathode (where the
hydrogen is evolved), and at the same time to regulate the ionic flux
between the chambers. This configuration minimizes hydrogen
losses to microbes on the anode and in the liquid and prevents
mixing of the hydrogen product gas with carbon dioxide from the
anode (Logan et al., 2008), thus maintaining the high purity of the H2
gas evolved. Moreover, membrane functions as a separator to avoid
any short circuit.
By now, four main membrane types are used: the cation exchange
membrane (CEM), the anion exchange membrane (AEM), the
bipolar membrane (BPM) and the charge mosaic membrane (CMM).
With respect to the transport numbers for protons and/or hydroxyl
ions and the ability to prevent pH increase in the cathode chamber,
the ion-exchange membranes are rated in the order: BPM, AEM,
CMM, CEM (Rozendal et al., 2008). However, the more widely used
membrane type is the CEM, or more specifically the PEM (proton
exchange membranes, such as Nafion'), which would selectively 98 permeate the positive charges from the anode to the cathode
chamber. During operation, however, cation species other than
protons are responsible for the positive charge transport through the
cation exchange membrane (Rozendal et al., 2006b) because
concentrations of Na +, K+, NH 4 +, and Ca2+ in wastewaters ('' pH 7) are typically present at concentrations 10 times higher than the
protons (Logan et al., 2008). As a result, protons consumed at the
cathode are not replenished by protons generated at the anode and
this leads to a membrane pH gradient (the pH increases at the
cathode and decreases in the anode chamber) which reduces MEC
performance. Indeed, every pH unit difference between the two
chambers will increase Eeq by 0.06 V, which corresponds to an
additional energy requirement of about 0.13 kWh/m 3 H2 per pH unit (Logan et al., 2008). Fan et al. (2008) reported that a Nafion'
membrane in a two-chamber MFC developed resistance up to 86% of
total internal resistance.
As a solution, the anion exchange membrane, which allows for the
transport across the membrane of negatively charged chemical
buffers, such as phosphate and bicarbonate alkalinity, is
progressively more widely adopted. It has been reported that AEM is
better than CEM, as it helps to buffer pH changes between the two
chambers and thus to substantially increase MEC performance
(Cheng and Logan, 2007; Rozendal et al., 2007). Alternatively membranes could also be avoided, because, contrarily
to water electrolyzers where they are required to prevent the
explosive mixtures of oxygen and hydrogen gases, there is no
oxygen evolution/presence in a MEC. Removing the membrane
reduces ohmic resistance and helps to reduce the bulk pH gradient in
the liquid, although it does not prevent localized pH gradients at the
electrodes (Logan et al., 2008). Using a single chamber MEC with
carbon cloth anode and carbon cloth loaded with platinum (0.5 mg 99 Pt/cm 2) for the cathode, Hu et al. (2008) produced 2.5 mol H2/molacetate at remarkable coulombic efficiency of 75% (with an Eap of 0.6 V). - Overpotentials causing energy losses
Compared to the maximal open circuit potentials (potential observed
when no current is running through the electrical circuit), the real
attainable voltage in a biologic cell is much lower, due to the so-
called overpotentials, which are potential losses owing to electron
transfer resistances and internal resistances (Figure 3.6). That's why,
as previously stated, the voltage requirements to allow the hydrogen
production at the cathode are considerably higher than the Eeq.
Therefore, in order to reduce the hydrogen costs, it is important to
minimize the irreversible energy losses as much as possible, while
maintaining an acceptable hydrogen production rate (Logan, 2008). Fig 3.6 Potential losses during electron transfer in a MFC. 1. Loss owing to bacterial electron
transfer; 2. Losses owing to electrolyte resistance; 3 Losses at the anode; 4. Losses at the MFC
resistance (useful potential difference) and membrane resistance losses; 5. Losses at the
cathode; 6: Losses owing to electron acceptor reduction (Courtesy of Rabaey and Verstraete,
2005). Two main kinds of overpotentials can be defined, the electrodes
overpotentials and the ohmic losses, and they both must be
considered in order to calculate the actual energy requirements, Ecell,
according to the equation: 100 E = E 78 '' (9η: +9η + IR') where '' >'and '' >@ are the sum of the overpotentials at the anode and at the cathode respectively and ABC is the sum of all ohmic losses within the system (Logan et al., 2008). Each electrode can be
treated individually (Wrana et al., 2010) and among the parameters
influencing the electrode overpotentials there are the electrode
surface, the electrochemical characteristics of the electrode, the
electrode potential, and the kinetics/mechanism of the electron
According to Logan et al. (2006), overpotentials in MECs are
classified as:
1. activation losses: losses related to the transfer of electrons to or from a substance reacting at the electrode surface (such as the
electron transfer mechanism between microbe and electrode at
the anode). Wrana et al. (2010) suggest to focus on improving
catalyst reaction kinetics by increasing operating temperatures
and reaction surface areas. 2. coulombic losses: Coulombic Efficiency (CE) is defined as the amount of electrons recovered as current versus the amount of
electrons available in the substrate. Therefore coulombic losses
(e.g. the amount of electrons not recovered as current) are mainly
due to microorganisms, that use a variable amount of energy for
their growth and metabolism. A good balance between bacterial
energy gain and electrode (terminal electron acceptor) may limit
these losses. 3. concentration losses: poor mass transfer kinetics may limit the supply or the elimination of each substance provided to the cell
and involved in the reactions. Substrate flux to the biofilm,
diffusional gradients resulting from improper mixing, or
unbalanced ratios of oxidized to reduced species at the electrode 101 surface are all contributing factors that will result in potential
losses (Wrana et al., 2010). The ohmic voltage losses (ABC) are independent of current and are
due to both resistance to electron flow through electrical conductors
(electrodes and external circuitry) and resistance to ion flow through
ionic conductors (electrolyte and membrane) (Logan et al., 2006).
The proposed solutions for decreasing ohmic losses are the reduction
of electrode spacing, the increase of electrolyte conductivity, the
selection of electrode and membrane materials with low resistivity
and also an adequate mixing of the medium within the cell. - Microorganisms
Compared to MFCs, little is known about the composition of the
microbial communities in MECs (Logan et al., 2008). However, the
same exoelectrogenic bacteria (Shewanella spp., Geobacter spp. and
Pseudomonas spp.) were found in thick biofilms grown on the
cathodes of both MFCs and MECs (Liu et al., 2005) and it is
assumed that the same exoelectrogens act in MFC and MEC.
The study of exoelectrogens has emerged as a new subfield of
microbiology and many research efforts have brought to discover
new strains and enrich their presence from various environments,
such as domestic wastewaters, ocean sediments and anaerobic
sewage sludges. Liu et al. (2010) have listed the isolated
exoelectrogens in diverse genetic groups, including four of the five
proteobacteria (a, b, g and d), firmicutes, acidobacteria and
actinobacteria (here presented in Table 3.4 together with the current
densities produced and the specific MFC configuration adopted).

102 Tab. 3.4 Taxa of bacteria, current density generated and reactor configuration used in MEC
experiments (Courtesy of Liu et al., 2010). Among them, some, such as Shewanella and Geothrix species, are
reported to produce their own electron shuttles and therefore have the
advantage to be positioned at a distance from the electrode and yet
still transfer electrons to the electrode surface. On the other hand, an
isolated strain of Pseudomonas aeruginosa capable of utilizing
glucose and producing an electron shuttle was inefficient in
converting glucose to electricity, probably due to the incomplete
oxidation of glucose to fatty acids (Rabaey et al., 2005).
Therefore, if pure culture are useful to evaluate the mechanisms of
electricity production and the strategies for optimizing this process,
mixed cultures could have the advantage of increasing MEC 103 versatility with respect to substrate utilization, of enhancing the
system robustness (due to biological diversity) and of being more
adaptable to wastewater treatment. It is also interesting to note that
while some isolates from mixed cultures demonstrate electrochemically active properties, most exhibit lower current
densities when grown as pure cultures than as mixed cultures (Liu et
al., 2010).
Mixed cultures are usually enriched from domestic wastewaters and
anaerobic sewage sludge (Liu et al., 2005; Logan, 2008) or even
from soils (Cheng and Logan, 2007) and have already been used in
MEC studies (Call and Logan, 2008; Ditzig et al., 2007; Rozendal et
al., 2006a; Tartakovsky et al., 2009; Wagner et al., 2009). The main
practice for enriching a bacterial community for a MEC (start-up
procedure) is to operate a MFC and then directly transfer the anode
into a MEC: this procedure ensures the formation on the anode of a
biofilm of suitable bacterial consortia and preselects an
exoelectrogenic community for MEC operation (Logan et al., 2008).
During the start-up, the MFC, operated at low external resistance,
will initially generate low current during biomass build-up, and
hence will achieve a high anode potential and preferably select
facultative anaerobes (Rabaey and Verstraete, 2005). Upon growth of
the culture, the metabolic turnover rate, and hence the current, will
increase. The moderate anode potential registered in this phase will
favor, on the contrary, lower redox facultative anaerobes.
This procedure was also followed in our study (Paper IV).
Alternative practices for MEC acclimation is the use of effluents
from working MFCs or MECs, full of active exoelectrogenic
bacteria, or the creation of a solution made by suspension of biofilm
scraped from the anode of an active cell (Cheng and Logan, 2007;
Rozendal et al., 2008). 104 Moreover, microflora enrichment directly in MEC is also possible
(even if more risky), that has the advantage of direct bacterial
adaptation to the operative reactor conditions. Lastly, the complete absence of oxygen in MEC promotes not only
the growth of obligate anaerobic exoelectrogenic bacteria, but also
that of non-exoelectrogenic fermentative or methanogenic microorganisms, which could decrease the process performances (H2
production and yield). Methanogens hydrogenotrophs are also
favored by the presence of high concentrations of hydrogen gas
within the MEC and could compete with exoelectrogenic bacteria for
the substrate or use the hydrogen produced by them for converting it
into methane. The suggested methods to inhibit the methanogens are:
' Avoiding buffer addition to the MEC (or at least no carbonate buffer). Rozendal et al. (2008) found that the use of a
bicarbonate buffer with a biocathode encouraged the growth of
hydrogenotrophic methanogens, able to use the buffer as a
carbon source. ' MEC exposure to oxygen: Call and Logan (2008) showed in a membrane-less MEC, at an applied voltage of 0.6 V, that the
exposure of the electrodes to air in between batch cycles reduced
methane concentrations to < 1% in the product gas and did not
impact current densities. In contrast, a lack of air exposure at the
same applied voltage resulted in methane concentrations of 3.4%
or more. The main problems of this strategy are that some
substrates (e.g., cellulose) require strictly anaerobic conditions
and that potential explosive mixtures (hydrogen + oxygen) could
be created. ' Alternative strategies (to be investigated): preventing the methanogens growth by varying culture conditions, such as
lowering pH, heat shocking of the inoculum, and short operation
retention times (Logan et al., 2008). 105 However, rather than to consider CH4 as an undesired by-product,
our study (Paper IV) tried to turn methane production in a MEC into
a valuable result. - Feeding substrates
Acetate is the typical carbon source for MEC, as it is the preferred
substrate for exoelectrogenes, particularly Geobacter sulfurreducens.
For example, MEC fed only with acetic acid achieved high H2
recovery yields (91%) and production rates (1.1 m 3 H2/m 3 reactor (total volume) d) (Cheng and Logan, 2007). However, in practice acetate generally does not exist alone and is typically found mixed with
other organic acids (in wastewaters, landfill lecheates, and
fermentation (acetate-rich) digestates). Thus, in the recent years,
some studies have been made with simple substrates similar to
acetate, such as butyric acid, lactic acid, propionic acid, and valeric
acid (Cheng and Logan, 2007), while other studies aimed at treating
acetate rich-wastes or diluted carbon sources found in wastewaters
(Ditzig et al., 2007; Wagner et al., 2009).
Compared with fermentative hydrogen production, MECs can
potentially utilize a wider variety of organic materials, because dark
fermentation basically requires a fermentable (carbohydrate-rich)
substrate, while MEC is subjected to theoretically less limitations.
Indeed, bacterial respiration can occur on a wide array of substrates,
ranging from pure compounds of sugars, carboxylic acids, alcohols
and proteins to complex mixtures such as biomass hydrolysates,
biowastes, agriculture residues and domestic, animal and food-
processing wastewaters (Liu et al., 2010).
However, using fermentable substrates such as glucose in MECs
entailed several problems, as it resulted in increased diversion of
electrons to nonelectricity sinks, such as biomass and organic
byproducts, and in the reduction of exoelectrogens (in proportion to
fermentative bacteria) in the community (Liu et al., 2010). As a
result, the kinetics of electron transport through the biofilm from 106 substrate to anode may be slowed, reducing the potential efficiency
and the overall energy conversion efficiency of the system.
Moreover, the use of a fermentable substrate may result in the
accumulation of organic acids from fermentation, which lowers the
pH of the reactor solution to as low as 4.5, a level that drastically
reduces hydrogen production from a MEC, perhaps due to permanent
damage to the exoelectrogens (Lu et al., 2009).
Alternatively, cellulose was used for direct current generation in
MECs using a mixed culture enriched from soil or a pure culture of
Enterobacter cloacae (Cheng and Logan, 2007; Rezaey et al., 2009).
However, the current density generated from cellulose was ten-times
lower than that from glucose and acetate. Therefore, pretreatment
and hydrolysis, similarly to the pretreatment processes in conversion
of cellulose to ethanol or fermentative hydrogen, may be needed to
break down complex cellulosic biomass structures and to release
soluble molecules, that can be more easily utilized by bacteria.
MEC can also use derivatives from sugar processing/fermentation
include polyacolcohols, which can be found in the water-soluble
materials of biomass, and uronic acids, which can be released from
the hydrolysis of hemicelluloses (Liu et al., 2011). Attention must be
paid to some phenolic compounds, such as acetophenone and 3-4-
dimethoxybenzyl alcohol, which have demonstrated strong inhibitory
effects on exoelectrogens and must be removed.
Apart from issues about process performances optimization (H2 yield
and production rate), the choice of the substrate is fundamental also
for making the process economically feasible. Indeed, taking into
account sugar as feeding, 1 kg of sugar contains 4.41 kWh of energy
and, considering that in the United Europe (EU) 1 kWh is worth up
to ' 0.16, even recovering all its energy content a maximum value of
' 0.70 could be obtained, that is less than its market value
(approximately ' 1 in the EU). This means that, in order to produce
enough hydrogen with MEC technology to supply the needs of 107 transportation and industry, it will be essential the use of an
abundant, renewable biomass, preferably available on the market for
low or negative (in case of waste) prices.
Therefore, new efforts have been recently dedicated to the study of
wastewaters or waste biomass sources as sustainable substrates for
the MEC process. In this way, a huge amount of the energy
contained in the wastewater (it may amount up to 9.3 times the
energy required to treat the waste with aerobic technologies
according to Logan, 2004) may be recovered in a high value product
(H2), providing very high margins in profit and energy gain for a
treatment process.
Easily digestible organic contaminants and some xenobiotics are
metabolizable under microbial fuel cell conditions (Jang et al., 2006)
while particulate substrates have not been extensively studied in
either MEC or MFC (Logan et al., 2008). Focusing on actual
wastewaters, different types have been tested as fuel in MFCs (Kim
et al., 2004; Liu et al., 2004; Min and Logan, 2004; Min et al., 2005;
Oh and Logan, 2005; Yokoyama et al., 2006; You et al., 2006),
while fewer studies with wastewaters have been conducted in MECs.
MECs efficiently exploited domestic wastewater (Ditzig et al., 2007)
and swine wastewater (Wagner et al., 2009) and hydrogen was also
successfully generated, although at low rates compared to pure
compounds in well buffered solutions, using winery and domestic
wastewaters (0.17 and 0.28 m 3 H2/ m 3 reactor (total volume) d, respectively; Cusick et al., 2010). However, slow degradation of complex
substrates and low conductivities of real wastewaters (typically 0.8 -
2 mS/cm) have been shown to decrease performance of MFCs and
MECs, compared to tests developed under optimal laboratory
conditions (Logan et al., 2008). Thus, additional studies are needed
about MEC using non-amended wastewaters (e.g. without nutrients
or buffers addition). Therefore in Paper IV an industrial wastewater,
with high-COD content and significant presence of methanol, was 108 used as feeding substrate without any amendment addition (buffer or
NaCl). This work shows satisfactory COD removal and methanol
exploitation in MEC, which has never been previously reported.
3.3 MEC application and scalability Compared to other biofuel production processes, MEC shows
interesting advantages from the applicative point of view:
' it directly generates high purity hydrogen, avoiding the pro- duction of undesired side products as in ethanol production from
biomass; ' it doesn't require particular detoxification processes for the removal of inhibitory compounds (such as required in the
ethanol-production process); ' it offers easy separation of high value product (H 2) from the liquid biomass solution, with reduction of separation/purification
costs. MECs are also advantageous for wastewater treatment because
' they generate less excess sludge in a more stable condition than the aerobic treatment process, where one-third to half of the
operating costs are associated with solids handling and treatment
(Kim et al., 2007; Logan et al., 2008); ' they can possibly limit the release of odors (Logan et al., 2008). All these aspects are promoting MEC application out of the lab-
scale. However, since electrical energy is consumed in MECs, first
of all they need to be more cost-effective than other existing
technologies in the same field. Logan et al. (2008) compared aerobic
(activated sludge, AS) and anaerobic processes (Anaerobic
Digestion, AD) for wastewater treatment to microbial electrolysis
(Table 3.5). These authors state that to be competitive, MEC must
recover sufficient hydrogen from the wastewater and simultaneously
minimize the required applied voltage. 109 Tab. 3.5 Energy requirements and production for wastewater treatment process. MEC may have an energy consumption similar to that required for
aerators in AS systems (0.5 - 2.4 kWh/kgCOD) and Rozendal et al.
(2007) stated that full-scale MEC systems are expected to require ca.
1 kWh/m 3 H2 and to produce 10 m 3 H2/m 3 d with a 100% overall hydrogen recovery efficiency. This means an energy consumption of
1.5 kWh/kgCOD, hence competitive with AS.
Moreover, compared to AS, they have the advantages of potentially
treating higher volumes of wastewater (higher loading rate registered
in continuous MEC system by Rozendal et al., 2007) and of
producing additional energy as hydrogen with lower excess sludge.
In addition, MEC process (similarly to AD), developing in an
environment not exposed to air as it is the AS, does not release
odors: removal of chemicals associated with odors has already been
demonstrated in MFC (Kim et al., 2008). All these advantages
potentially make MEC an environmental-friendly and healthy
technology, acceptable by the media and the people.
Compared to AD, the first advantage of MEC is that from the same
amount of COD, the gas produced (H2) is more valuable than
methane ($0.75/kgH2 COD vs $0.11/kgCH4 COD; Logan et al., 2008).
Capability of MEC for the nutrients (ammonia, nitrate,
phosphorus,...) removal has not already been examined, but it could
represent an additional advantage over AD process (Table 3.5). On
the other hand, AD is already a well established and applied to real-
scale technology which doesn't require any significant electrical
energy input to produce methane. Moreover, CH4 can be exploited
by existing cheap technologies for electricity and heat production Process Volumetric Loading rate (kgCOD/m 3 d) Sludge Production Nutrient removal Energy consumption (kWh/kgCOD) Energy production Activated Sludge 0.5 - 2 high Yes 0.7 - 2 No Anaerobic Digestion 8 - 20 low No low Yes, CH4 Microbial Electrolysis '' 6.5 low (expected) Possibly 0.5 - 2.4 Yes, H2 110 (then easily sold or used for energetic self maintenance of the plant).
However, it is also true that MEC is a new technology, whose
performance can be greatly improved and design could be simplified.
For sure, the commercialization of an efficient and scalable MEC
system will depend on the full exploitation of the high-value gas
produced (in order to compensate for the electrical energy costs of
the system), on the adoption of an efficient design, applicable to real-
scale, and on the cost effectiveness of a possible biomass
pretreatment. Also, research is needed on whether MEC systems will
be capable of stand-alone operation or if they will require aerobic
effluent polishing step (as commonly is the case for ADs) (Logan et
al., 2008). At present, while the high hydrogen yield by the MEC process has
been demonstrated in laboratory tests, the scalability of MEC tech-
nology remains a challenge. No published detailed reports on MECs
stand-alone pilot-scale plant are available. However, Logan (2010)
reports the building of the first MEC pilot-scale plant (Figure 3.7),
which has been conducted at the Napa Wine Company (Oakville,
CA, USA) by Penn State researchers and engineers by Brown and
Caldwell (Walnut Creek, CA, USA). The reactor design is based on
the approach of immersing brush anodes and flat cathodes made of
stainless steel into a tank (Call et al., 2009; Logan, 2008; Selembo et
al., 2009). The reactor contains 24 modules, each with six pairs of
electrodes, and is approximately 1 m 3 in total volume. No performance of this system are reported in the paper. 111 Fig. 3.7 Pilot-scale microbial electrolysis cell fed with winery wastewater at the Napa Wine
Company in California, USA (Courtesy of Logan, 2010). Another possibility for the scaling up of the MEC technology is
connected to its ability to use the biodegradable components of a
wide range of residual waste streams, even at very low COD
concentrations such as reported by Kim et al. (2010a; 2010b).
Therefore, the VFAs-rich effluents of dark fermentative or
methanogenic (AD) bioprocesses could be used as feeding substrate
for MEC, which may extract most of the chemical energy left in the
effluents (otherwise representing a disposal burden and a waste of
energy) and may also act as an effluent-polishing unit (Kim et al.,
2010a; 2010b). Therefore the association of a BES system to other
biological technologies represents a good driving force for its real
scale application. The combination of MEC technology with dark
fermentative biohydrogen production stage will be dealt in details in
section 4.3 of this PhD thesis.


113 Chapter 4 Biohydrogen and integrated energy production

By now, the main biological processing strategies able to treat
complex substrates, such as industrial and agricultural wastewaters,
and to simultaneously produce high-value products are the
methanogenic anaerobic digestion, the biological hydrogen
production, minor other fermentation processes and the bioelectrochemical systems. Biohydrogen production might be
considered as more environmental-friendly process if associated to
other technologies, which can exploit water, nutrients and by-
products (mainly the VFAs) of the biohydrogen step for the
production of fuels, energy or chemicals. In this way, the
disadvantages of the dark fermentation stage (first of all, the low
hydrogen yield theoretically achievable) may be overcome by the
overall performances of a multi-stage process. This means that the
H2 stage can be potentially a major part of a complex (bio)factory 114 which includes pre-treatments, separations and multiple steps of
biological or biochemical transformation of the organic matter. By
producing multiple products, it is also possible to take advantage of
the differences in biomass components and process intermediates and
to maximize the value derived from the biomass feedstock.
Agriculture sector with its well established knowledge of
bioprocesses, may play a leading role in this type of production
complex, and therefore countries with large agricultural economies
have potential for significant economic development through
incorporation of bioenergy into bioindustry.
Some studies presented hydrogen as a potential by-product of other
processes, such as gluconic acid production, an high-value
biotechnology product (Woodward et al., 1996). However this
process, due to the low market of gluconic acid (about only 50,000
tons/year in the US), was not so economically convincing. This
example helps to understand how one of the main problems of the
multi stage strategies is the need to produce, together with
biohydrogen, one or more high-value products for which there is also
a large need, such as a fuel or electricity or a chemical.
In this Ph.D. thesis we present hydrogen production by dark
fermentation of biomasses or wastewaters as the first stage of
different possible "biofactory" stages, needed to counter the
incomplete oxidation of organic matters achieved by the hydrogen
stage and to recover the remaining energy. In this way the 85% of the
energy still contained in the by-products of the dark fermentation
stage (Logan, 2004) could be further and efficiently exploited.
The multi- (often two-) stage system also includes the idea that the
overall biomass conversion is obtained through a sequence of
biochemical reactions which do not necessarily share the same
optimal environmental conditions (De Vrije and Claassen, 2005).
Therefore every different fermentation steps, even conducted in
physically separated reactors, must be optimized in order to generate 115 at the maximum reaction rate and yield the different valuable
products of the process.
Finally, the choice of what reactions could efficiently been coupled
must consider the overall process economy. For example,
considering that the maximum hydrogen yield from glucose
fermentation is 4 mol per mole of glucose (see section 1.3.2),
paradoxically a doubling of the hydrogen yield can be achieved by
fermenting 1 mol of glucose to 2 mol of ethanol and then reform the
2 mol of ethanol to 8 mol of H2 (Deluga et al., 2004). Thus, the
synergy of the processes involved must be maximized in order to
fully exploit the substrate, which otherwise fails to achieve complete
conversion due to thermodynamic limitation.
In the next paragraphs, different conceptual integrated system
configurations will be presented, even if this chapter doesn't pretend
to include all the possible and near to infinite combinations of
4.1 Dark fermentation + Photo fermentation Sequential dark and photo fermentation is a rather new approach in
biological hydrogen gas production. The conversion of the end-
products of the dark fermentation stage is a process thermodynamically unfavored (''Go = +75.2 kJ/mol), but
phototrophic purple, non-sulphur bacteria are able to overcome this
barrier by employing energy from light (hv) while using the organic
acids (mainly acetate) as the prime carbon source, according to the
following reaction (Claassen et al., 2005): CH3COOH + 2 H2O + ''''hv'''' '' 2 CO2 + 4 H2 Specifically, in the first thermophilic dark fermentation stage, the
biomass is fermented to acetate, carbon dioxide and hydrogen, while
a separate second photobioreactor under anaerobic conditions
converts the acetate formed in the first stage to additional hydrogen 116 (and CO2), therefore reaching theoretical maximum production of 12
molH2/molglucose equivalent : (i) Stage I - dark fermentation
C6H12O6 + 2 H2O '' 2 CH3COOH + 2 CO2 + 4 H2 (ii) Stage II - Photofermentation
2 CH3COOH + 4 H2O '' 8 H2 + 4 CO2 Hydrogen gas production capabilities of (purple) photosynthetic
bacteria such as Rhodobacter spheroides, Rhodobacter capsulatus,
Rhodovulum sulfidophilum W-1S, Rhodopseudomonas palustris,
Rhodospirillum rubrum and Rodopseudomonos palsutris P4 have
been previously investigated, together with the range of organic acids
(acetic, butyric, propionic, lactic and malic acid) which can be used
by them as carbon source. Table 4.1 (Courtesy of Kapdan and Kargi,
2006) summarizes the yields and the rates of hydrogen production by
photofermentation of different organic acids, as reported in literature.
Figure 4.1 shows instead how this two-stage bioprocess could be
physically realized, including also the possibility of gas recirculation
in the first reactor (aiming at improving the H2 recovery by gas
sparging technique) and a gas separator system (for the upgrading of
hydrogen, thus employable in end-use technologies like PEM-Fuel
Cells). The major types of photo-bioreactors developed for hydrogen
production are commonly tubular, flat panel and bubble column
reactors, which have different features, especially if we consider the
importance of photochemical efficiency parameter (theoretical
maximum 10%). For example, reactors highly illuminated have
lower light conversion yields but higher hydrogen production rates.
On the other hand, they are more exposed to photo-inhibition of
bacteria by excess light, potentially resulting in decreases in
hydrogen production rate (Kapdan and Kargi, 2006). 117 Tab. 4.1 Yields and rates of biohydrogen production from organic acids by photo fermentation
(Courtesy of Kapdan and Kargi, 2006). Fig. 4.1 Outline of a two stage fermentation process for hydrogen production from biomass.
Stage 1: heterotrophic dark fermentation; stage 2: photoheterotrophic fermentation (Courtesy
of De Vrije and Claassen, 2005). 118 Hereafter some of the most recent and relevant research experiences
about a two-stage dark- and photo-heterotrophic bioprocess will be
explored. Kim et al. (2001) combined dark fermentation (using Clostridium
butyricum NCIB 9576) with photo-fermentation (using immobilized
Rhodopseudomonas sphaeroides E15-1 in hollow fibers) in a two
stage process treating raw rice wine or Tofu (soybean curd)
wastewaters (a carbohydrate-rich substrate). The dark stage had low
HRT (1 day or less) and achieved 0.9 - 1 LH2/Lwastewater d, along with
organic acids and ethanol production with rice wine waste. The
effluent of this process was used for feeding the slower second stage
(HRT of 10 days), where 0.44 and 0.2 LH2/Leffluent d were obtained
with rice wine or Tofu wastewater, respectively. The authors
reported also preliminary tests about mixing the effluent of the
acidogenic stage with a pre-treated sewage sludge to increase the
performance of the second stage. Yokoi et al. (2002) reached high yield of 2.7 molH2/molglucose in the
first dark fermentative stage, by feeding the digester with mixed
starch manifacturing wastes (sweet potato starch residue as a carbon
source and corn steep liquor as a nitrogen source). Mixed culture of
Clostridium butyricum and Enterobacter aerogenes HO-39 was used
in this stage. The supernatant of the culture broth from the first stage
was then used to feed Rhodobacter sp. M-19 culture, attaining
hydrogen yield of 4.5 molH2/molglucose (with pH adjustment at 7.5 and
addition of EDTA and Na2MoO4·2H2O to the medium). Therefore a
very promising overall hydrogen yield (7.2 molH2/molglucose) was
reached. Nath et al. (2005) used glucose as substrate for a first dark
fermentation step in batch (with pure culture of Enterobacter cloacae
DM11), whose acid rich effluent underwent photo-fermentation by
Rhodobacter sphaeroides strain O.U.001 in a column photo- 119 bioreactor, which constituted the second stage of the process. The
dark fermentation stage reached yield of about 1.86 molH2/molglucose,
while the yield in the photo-fermentation stage was about 1.5 - 1.72
molH2/molacetic acid, thus achieving an overall recovery of hydrogen of
about 5 molH2/molglucose. The authors also observed that one of the
main limitation of the process was the low light conversion
efficiency of the second stage (which reached at maximum only
0.51%). Also Tao et al. (2007) simulated a dark-fermentation stage through a
batch test, where mixed microflora fermented sucrose and produced
hydrogen and a mixture of fatty acids (mainly butyrate and acetate).
The maximum hydrogen production rate was higher than 360
mLH2/L h and the maximum hydrogen yield was 3.67
molH2/molsucrose. Then photo-fermentation by Rhodobacter sphaeroides SH2C was employed to convert the fatty acids into more
hydrogen, achieving a total hydrogen yield of 6.63 molH2/molsucrose
for the whole two-step process. Complete consumption of fatty acids
was reported (no butyrate, acetate, propionate, or valerate were
detected at the end of the process). In 2008, Manish and Banerjee proposed a new process flow diagram
(Figure 4.2) which would further complicate the system, introducing
the idea of a three stage process where the effluent of dark-
fermentation is sent to a photo-fermentation stage and that of photo-
fermentation is sent to an anaerobic digester to produce biogas.
Hydrogen and carbon dioxide gas mixture produced during the
fermentation stages would be sent to a pressure swing adsorber for
being upgraded.
The authors reported that this system reached higher H2 yield than
that of the single dark-fermentation process and reduced the
requirement of sugarcane (in term of kg sugarcane input/kgH2)
approximately by 65% as compared to dark-fermentation. This lower 120 requirement also reduced the amount of electricity required during
the milling process by 30%.
The process had the highest energy efficiency, the highest net energy
ratio and lower greenhouse gases (GHG) emissions among different
biohydrogen processes considered (photo or dark single stage
fermentation systems or electrochemically assisted process).
Compared to Steam Methane Reforming process, for each kg of
hydrogen produced this system reduces GHG emissions by 7.31 -
9.37 kg CO2 ('' 57%) and non-renewable energy use by 123.2 - 148.7
MJ ('' 65.79%) (Manish and Banerjee, 2008). Fig. 4.2 Three stage fermentation process proposed by Manish and Banerjee (2008). Aiming at real scale applications, a group of researchers (De Vrije
and Claassen, 2005; Claassen et al., 2005) designed a system
consisting of a thermobioreactor (capacity of 95 m 3) for dark fermentation, followed by a photobioreactor (300 m 3) equipped with a sunlight collector (Figure 4.3). A preliminary system (extruder)
was also used to exploit complex feedstocks as process substrates.
Hydrogen was produced in the thermobioreactor by extreme
thermophilic microflora (Caldicellulosiruptor saccharolyticus) and
by Rhodobacter capsulatus pure culture used in the second stage.
The size of the plant was set at a production capacity of 425
Nm 3 H2/h, which is equivalent to an energy production of 5.4 GJ/h, based on the upper combustion value of 12.74 MJ/Nm 3 H2. 121 Using glucose (derived from potato steam peels) as initial substrate,
the process produced one third of the total hydrogen in the first stage
(where it was recovered through gas stripping technique and further
purified by pressure swing adsorption system) and the remaining two
thirds in the following step. As a whole, the process achieved a
conversion efficiency yield of 47%. This data is quite noticeable if
compared with the maximum achievable efficiency of 69%,
calculated assuming that the two separate fermentations may operate
each one at 80% conversion efficiency. Fig. 4.3 Simplified flow sheet of the two stage hydrogen production bioprocess by Claassen et
al. (2005). Details: 1: extruder; 2: tank; 3: heat exchanger; 4: thermoreactor; 5: photoreactor; 6:
titled plate settler. In conclusion, this two-stage production system has both advantages
and disadvantages over single stage dark- or photo- fermentation
Among the advantages, the limitation of the photo biological process
requiring organic acids as substrates would be eliminated,
considering the high acid content of the dark anaerobic process
effluent. Further exploitation of its organic acids means also better 122 effluent quality in terms of COD. Most of all, higher hydrogen
production yields can be obtained with the two systems combined
and optimized.
On the other hand, there are many factors limiting the practical
application of such a process.
A major challenge is the low photosynthetic efficiency, either with
solar radiation or tungsten lamps, since at even moderate light
intensities the main part (> 80%) of captured light is dissipated as
heat (Hallenbeck, 2009). This involves the demand of large surface
areas for anaerobic photobioreactors, which could be extremely
expensive for practical applications. For example, Claassen et al.
(2005) calculated that the required area of a photobioreactor
converting the effluent of a 450 m 3 dark fermentation trickle bed reactor would be 12 ha. Therefore, photofermentative hydrogen
production using current state of the art organisms and technology
has been considered economically unrealistic by many authors
(Westermann et al., 2007).
Secondly, the system requires a continuous control in order to
provide optimum media composition and environmental conditions
for the two different microflora involved in the process. For example,
dilution (also useful for reducing suspended solids, which decrease
light penetration) and neutralization of dark fermentation effluents
are required before feeding the photo-fermentation stage, in order to
adjust the organic acid concentration and the pH to an optimal level
for photosynthetic bacteria. Also the nitrogen content of the dark
fermentation effluent must be monitored, since nitrogen, especially
in the form of NH4 +, not only inhibits the nitrogenase enzymatic activity, but also represses the synthesis of this enzyme fundamental
for the photo-fermentation (Ntaikou et al., 2010a). Establishing and
maintaining nitrogen deficient environmental conditions or
selecting/genetically modifying nitrogenase enzyme and its
regulating system could be useful. 123 4.2 Dark fermentation + Anaerobic Digestion The integration of a dark fermentative hydrolytic/acidogenic process
with a subsequent, physically separated methanogenic process for
combined hydrogen and methane generation, has been proposed by
many researchers and shows many advantages, mainly an higher
waste stabilization efficiency and an higher net energy recovery for
the overall process (Angenent et al., 2004; Ntaikou et al., 2010a).
Basically this strategy splits the traditional AD process in two stages,
with the first step assigned to hydrogen and organic fatty acids
production and the second to methanogenic (prevalently acetoclastic)
fermentative reactions for methane production through acids
consumption. These steps must occur in two physically separated
reactors (as shown in Figure 4.4), providing the preferred
environments for acidogenic hydrogenesis and methanogenesis (Han
et al., 2005). To this end, in order to reduce the characteristic
differential between the two stages in hydraulic retention time
(which in the second stage must be usually an order of magnitude
higher than the first stage, due to the slower growth rate of the
methanogenic archaea), many different reactors designs have been
proposed, such as immobilized bioreactor (trickling filters and
upflow anaerobic sludge blankets - UASBs -) for the AD stage
(Guwy et al., 2011). Fig. 4.4 Flow diagram of the two-stage (hydrogen production by dark-fermentation and
methane production by anaerobic digestion) process. Hydrogen stage Methane stage Feeding Digestate H2 CH4 AGV Biogas upgrading 124 The strictly methanogenic anaerobic digestion (AD) as a second
stage is particularly advantageous because of its high organic
removal rates, low energy-input requirement, energy production (i.e.
methane) and low sludge production. Also, methane can be used as a
fuel source for on-site heating or electricity production, or converted
to other useful products, such as methanol for biodiesel production
(Angenent et al., 2004).
On my opinion, compared to the other strategies presented in this
chapter, converting the organic matter remaining from the dark
fermentative hydrogen step into methane gas also has the advantage
to be the most feasible process for bioenergy production in the near
term in real scale. Indeed, the technologies for methane production
with simultaneous treatment of solid wastes and wastewaters already
include single- and multistage processes and are well developed and
established. This is particular true for our country (Italy), which in
2009 was Europe''s number four biogas producer with 444.3 ktoe,
and where it is expected the production of at least 2000 MWe from
agro-energy plants within 2015 (Biogas barometer; Eurobserv'ER,
2010). For real, this type of two-stage process has been traditionally
suggested and used for merely improving the methane production
and the AD process efficiency (Liu, 2008). Indeed, the separation of
the process in two stages led in some studies to a larger overall
reaction rate, organic matter conversion rate and biogas yield
(Blonskaja et al., 2003; Mata-Alvarez et al., 1993). Also, the process
achieved a better pathogenic destruction combining a short
hydrolysis stage performed at thermophilic temperatures and a
methanogenic stage at thermophilic or mesophilic temperatures
(Bendixen, 1994).
However, the simultaneous production of two different gaseous
energy sources/vectors (i.e. hydrogen and methane) brings new
additional advantages to this process, from those intrinsically 125 connected to the H2 (its energy content and environmental-friendly
nature) to new ones deriving from the cooperation of the processes.
For example, stronger hydrolytic conditions typical of the first
acidogenic step make the range of exploitable feedstocks wider: this
strategy has been used so far for food wastes, cheese whey, olive mill
wastewaters, household solid wastes, shredded paper wastes and
wastewater sludges (Antonopoulou et al. 2008; Han and Shin,
2004b; Koutrouli et al., 2009; Ntaikou et al., 2010b).
Ueno et al. (2007b) reported that this kind of two-stage process
achieves enhanced biological stability for wastes and higher organic
loading rate capacities for the methanogenesis process, compared
with the traditional single-stage process, while Hawkes et al. (2007)
confirmed the general higher total efficiency on waste treatment and
energy recovery than the traditional one stage process.
Moreover, recovering the hydrogen before bacteria turn it into
methane, as it commonly happens in anaerobic digestion, makes the
process advantageous also from economic point of view. Hydrogen
gas ($6/kg) is more valuable than methane ($0.43/kg) and on the
mass basis it also contains 2.2 times more energy than CH4.
Logan (2004) assumed the possible monetary income deriving from
the exploitation of a wastewater (2 g/L of BOD) from a single large
food processing plant, generating 1.4 x 10 6 m3/year of wastewater. He demonstrated that compared to the $310,000/year gained if all the
BOD was just converted to methane, a two stage plant could produce
$350,000 worth of hydrogen annually (assuming that all the organics
in the wastewater were sugars and converted with a ratio of 2
molH2/molglucose), together with $260,000/year coming from the first
stage effluent, fully converted into methane (assuming 0.4
LCH4/gBOD). Thus, at current gas prices the simultaneous recovery of
hydrogen and methane would be economically more favorable than
that of methane from a traditional AD process. 126 Another advantage is that the hydrogen-methane mixture is reported
to significantly reduce air pollutants and to increase combustion
efficiencies if adopted as fuel for internal combustion engines instead
of pure methane (Bauer and Forest, 2001; Hallenbeck, 2009).
Figure 4.5 shows some of the inputs/outputs of the system, together
with some mentions to operative aspects and advantages of the
process. Fig. 4.5 Principle diagram of two-stage process for hydrogen and methane production
(Courtesy of Liu, 2008). Even the final outputs of the system (i.e. biogas and digestate) seem
to possibly be positively affected by the separation in two stages. The
biogas, a mixture of methane and carbon dioxide, is the gaseous end
product of the organic matter degradation by anaerobic
microorganisms and its quality depends on its composition, which
usually consists of CH4 (50-70%), CO2 (30-50%) and smaller
amounts of hydrogen sulphide and ammonia. Nevertheless, its
composition strictly depends on the feedstock used as substrate, the 127 process conditions and the type of digester used. Beyond the
advantages guaranteed by the two-stage anaerobic digestion, our
study (Paper III) reports that the second methanogenic stage produce
a biogas with a stably higher (> 70%) methane content than the
typical range for AD process.
The digestate, i.e. the organic residue of the fermentation which
shows a good content of essential nutrients (N,P and K), is the other
main product of Anaerobic Digestion. Similarly to the one-stage
process, also in the two stage system the digestate deriving from the
methanogenic step can be split into a solid and a liquid fraction, with
the first collecting the most of the phosphorous, and the second the
most of ammonia-nitrogen. Thus, those fractions could be employed
as fertilizer and as soil conditioner on farmland or gardens to
improve soil quality. But in addition, in the two-stage process the
digestate could be also used (instead of water) for achieving a better
control on the reaction pH of the first stage and for simultaneously
diluting the VFAs within the first bioreactor, thus warding off their
inhibitory effect on hydrogen producing bacteria. Preliminary
research was made by our group about this topic (Congress
Communication II) but further research shall be carried out to test
this strategy with different kinds of fermentable organic substrates
and operational conditions. Nevertheless, the two-stage hydrogen and methane strategy is not
always advantageous: for example concentrated slurries and wastes
with a high lipid concentration should preferably be treated in a one-
stage digester (lipids will not be hydrolysed in the absence of
methanogenic activity). Similarly, hydrolysis and acidification of
proteins is not fully promoted by acidogenic conditions (de Vrije and
Claassen, 2005).
Also, the two-stage strategy remains basically unproven in real scale
as it adds complexity to the system and, as a consequence, increases
investments and operational costs. Currently, just the 10% of full 128 scale biogas plants in Europe rely on two-stage (both methanogenic)
processes (Choi et al., 1997; De Baere, 2000).
Finally, the effect of increasing biogas production has not been
accepted broadly, as separation of the two processes negatively
affects syntrophic association and prevents interspecies hydrogen
transfer between acidogenens/acetogens and methanogens (Reith et
al., 2005). However, the scientific community made many researches in the last
years about two-stage H2-CH4 anaerobic digestion process and some
noticeable or new examples of them will be given in the next lines. Liu et al. (2006) studied a two-stage hydrogen-methane fermentation
process with household solid waste as substrate at mesophilic
temperature, setting the HRT at 2 and 15 days for hydrogen stage
and methane stage, respectively (Figure 4.6). The short HRT of the
first stage was effective for separating acidogenesis from
methanogenesis and no other control was used for preventing
methanogenesis in the H2 stage. This short HRT helped to maintain a
stable pH (a key factor affecting the hydrogenic fermentation
pathway) in the first bioreactor, where the optimum range for pH was
found to be between 5 to 5.5.
The authors reported an hydrogen production of 43 LH2/kgVSadded and
a methane production of approximately 500 LCH4/kgVSadded. The
overall methane production was 21% higher than one-stage process
which was simultaneously run as control and which produced just
413 LCH4/kgVSadded. The same authors stated that these results were
similar to other studies results, as those shown by Mata-Alverez et
al. (1993) who achieved 510 LCH4/kgVS in a two-stage process for
household solid waste fermentation and only 428 LCH4/kgVS in one-
stage process (i.e. 19% methane increase). However, Mata-Alverez
et al. (1993) didn't evaluate the first stage performance (e.g. H2
129 Fig 4.6 Schematic diagram of two-stage hydrogen-methane process by Liu et al. (2006). 1:
Hydrogen reactor; 2: Effluent bottle; 3: Methane reactor; 4. Gas meter and counter; 5: Mixer. Similarly, Paper III reports the comparison between a two-stage and
one-stage AD process made by our group. Briefly, the hydrogen-
stage (2 L working volume; 3 d HRT) of the two-stage thermophilic
reactor gave relatively partial contribution to the total energy yield
achieved by the system (13% of the total energy produced),
according to previous works (Van Ginkel and Logan, 2005a). In our
study the overall energy recovery doesn't show significant
differences between the two- and single-stage AD systems (13-14
kJ/g VS added), probably due to partial inefficiency of the methanogenic
reactor (14.7 L working volume; 22 d HRT) of the two-stage
process, but highlights the increase of the methane content for the
methanogenic reactor of the two-stage system ('' 70%), as compared
to the single-stage process (55%). Energy production shown in Paper III are also in agreement with
results by DiStefano and Palomar (2010), who focused on the effect
of two-stage process reactor configuration on the system energy
(hydrogen and methane) yield. Five different reactor systems, all fed
with a 40 ± 1 g/L complex synthetic substrate (Ensure media ® +
micronutrients) and maintained at 35 °C, were investigated and
compared: suspended growth, two-phase mixed, two-stage mixed,
UASB reactor, and two-phase UASB. All reactor configurations
achieved very high COD removals, on the order of 99%, but the
highest energy productions, although lower than the maximum 130 theoretical value (15.2 kJ/g COD removed), were reached by the mixed
two-phase and two-stage configurations, amounting to 13.3 and 13.4
kJ/g COD fed respectively (Figure 4.7). The (two)phase or stage reactor
configuration influenced microbial pathways in acidogenic reactors,
since butyrate was the predominant volatile acid in phased
configurations, whereas acetate was predominant in the staged
However, the idea suggested by these authors (DiStefano and
Palomar, 2010), which is contrary to other studies, is that the
complex substrates fermentation in a two-step AD process is
justifiable primarily for improved process stability, improved
methanogenesis performances and for the end-use zero GHG
emissions technologies associated with H2, whereas just secondarily
for the total energy recovery point of view. Indeed, regardless of the
reactor process configuration, hydrogen represented a minor
proportion of the recovered energy input and theoretical analysis
revealed that the maximum specific energy production from the two-
phase suspended-growth configuration is only 9% higher than that
from a single-stage mixed reactor (15.2 kJ/g COD versus 13.9 kJ/g COD). On the other hand, it is noticeable that the specific methane
production for the two-phase and two-stage reactors was 22%-26%
higher than that of the suspended growth reactor, which produced
0.30 LCH4/g COD fed. The authors justify this significant increment with
the acidification achieved in the first stage, which has enhanced the
biodegradability of the organic matter for the subsequent
methanogenic reactor. 131 Fig. 4.7 Average specific energy production by process configurations (Courtesy of DiStefano
and Palomar, 2010). Sus: suspended growth; 2PSus: two-phase mixed; 2SSus: two-stage
mixed; UASB: upflow anaerobic sludge blanket; 2PUASB: two-phase UASB. The laboratory experiment by Ueno et al. (2007b) aimed instead at
evaluating the combination of a methanogenic process (with
thermophilic down-flow packed-bed reactor) with a previous
thermophilic (60 °C), pH-controlled acidogenic stage, alternatively
used for hydrogenic or solubilizing operation. The first stage was fed
with artificial garbage slurry containing milled paper. Hydrogenic
operation resulted to be more suitable to combine with methanogenic
process than solubilizing operation, since the retention time of the
former was shorter (0.5 d) than that of the latter (4 d), nevertheless
obtaining almost the same levels of overall removal efficiency in
both COD and VSS. At 25 h of total retention time, with hydrogenic
and methanogenic processes combined, overall COD removal and
VSS decomposition were 82% and 96%, respectively. The process
produced 199 mmolH2/Lreactor day (which means approximately a
yield of 2.0 molH2/mol hexose consumed) and 442 mmolCH4/Lreactor day
(with an average content of 60% of CH4 in the biogas). Comparing
this two-stage system with single methanogenic process at the same
HRT, again two times higher methane gas was produced and also 132 higher allowable OLR in methanogenic process was sustained (Ueno
et al., 2007b). In the same year, Cooney et al. (2007) used a two-stage magnetic
stirred hydrogen-methane system of 2 and 15 L working volume,
respectively (Figure 4.8). This relative volume ratio (1:7) with a
consequent short retention time in the methanogenic reactor was
selected by the authors to test the assumption that separation of phase
can enhance metabolism in the second reactor. Temperature and pH
in both reactors were controlled and maintained at 35 ± 0.1 °C and at
pH 5.5 in the first reactor and 7.0 in the second reactor through
automated addition of chemicals. The reactor system was inoculated
with conventional anaerobic digester sludge without pre-treatment
(to simulate full scale operation) and fed with a glucose (10 g/L) -
yeast extract (2 g/L) - peptone medium (2 g/L).
The authors showed that the selection pressure of pH and dilution
rate is sufficient to select for acidogenic and methanogenic bacteria
in their respective stages. The percentage of hydrogen in the head
space of the first reactor was between 30% and 40%, regardless of
varying dilution rates. However relatively low H2 yields (between
0.05 and 0.35 molH2/molglucose) and rates (range: 11 - 85 mmolH2/L d)
were observed, probably due to an unfavorable predominant
production of lactate (from 6 to up 31 g/L found in the first stage
effluent). As for the second stage, after its preliminary independent
operation, the integration into the two-phase system improved its
CH4 yields and production rates at all dilution rates, achieving up to
0.57 molCH4/molglucose and 12.95 mmolCH4/L d. However, the two
stage integration did not show any accelerated metabolism in the
second reactor that would allow the use of shorter retention times for
it. Therefore, the functioning of a system with a small second stage
reactor and with overall processing time shorter than usual (both of
which would greatly decrease the cost of application in industry) was
not supported. 133 Fig. 4.8 The two-stage system design by Cooney et al. (2007). Differently, Kraemer and Bagley (2005) focused on the effects of
methanogenic effluent recycle in a two-phase anaerobic (Figure 4.9).
The hydrogen reactor (2 L of working volume), fed with a synthetic
glucose-rich medium at 37.3 kgCOD/m 3 d, was operated as a chemostat at 35 °C and pH 5.5 with a 10 h hydraulic retention time.
The second stage was an up-flow reactor (12.5 L of working
volume), operated at 28 °C and pH between 6.9 and 7.2. Without
methanogenic effluent recycle, the H2 productivity (0.115 g H2 COD/g feed COD, data incorporating both gaseous and dissolved H2) was higher than with recycle (0.015 H2 COD/g feed COD), where presence of methane
in the biogas up to 17% (v/v) was also detected. On the other hand,
effluent recycle reduced the required alkalinity for pH control by
approximately 40% and it could still be used without decreasing the
yield, maybe through membrane filtration or heat treatment of the
effluent, in order to exclude the return of hydrogen-consuming
organisms. The improvement of this strategy could be particularly
important for full-scale application of the two-stage AD process. 134 Fig. 4.9 Diagram of the experimental two-stage system of Kraemer and Bagley (2005). Other minor examples of two-stage H2-CH4 producing reactors are
reported in literature, and among them it could be interesting to
mention of the BIOCELL system (Han and Shin, 2004b), due to its
innovative design based on phase separation, reactor rotation mode,
and sequential batch technique. The BIOCELL process consisted of
four leaching-bed reactors for H2 recovery, fed with food waste and
operated in a rotation mode with a 2-day interval between
degradation stages, and of one UASB reactor for CH4 recovery
trough the post-treatment of hydrogenic stage effluent. At the high
volatile solids (VS) loading rate of 11.9 kg/m 3 reactor day, this system could remove 72.5% of VS and convert the VS removed to H2
(28.2%) and CH4 (69.9%) in 8 days. H2 gas production rate was 3.63
m 3/m3 reactor day, while CH4 gas production rate was 1.75 m 3/m3 reactor day. The yield reached for H2 and CH4 were 0.31 and 0.21 m 3/kg VS added, respectively. 135 However, although a good amount of laboratory two-stage
fermentative systems have been reported, at the present there isn't
any real scale application of this process and just few pilot scale
experiences are known. A pilot plant for combined hydrogen and
methane generation from a mixture of pulverized garbage and
shredded paper wastes has been evaluated by Ueno et al. (2007a)
(Figure 4.10). The first stage was a CSTR inoculated with
thermophilic microflora enriched from excess activated sludge
compost and kept at 60 °C, while the methanogenic stage was
operated at 55 °C using an internal recirculation packed-bed reactor.
The two stages had a HRT of 1.2 d (hydrogenesis) and 6.8 d
(methanogenesis) and produced 5.4 m 3/m3 reactor d of hydrogen and 6.1 m 3/m3 reactor d of methane, with high chemical oxygen demand and volatile suspended solid removal efficiencies of 79.3% and 87.8%,
respectively. Maximum hydrogen production yield was 2.4 mol/mol hexose and 56 L/kg COD added. This study also showed that the methane yields were two fold higher than a comparable single-stage process. Fig. 4.10 Schematic diagram of the two stage pilot scale process by Ueno et al. (2007a). The
arrows show the flow direction of raw material (solid lines) or biogas (dashed lines). The Energy Technology Research Institute of the National Institute
of Advanced Industrial Science and Technology in Japan operated a
semi-pilot scale two-stage hydrogen-methane plant using wastes 136 collected from a local cafeteria: kitchen waste (50 kg/d), paper waste
(3-5 kg/d) and food waste (10 kg/d). The plant configuration shows a
solubilization/hydrogen fermentation tank of 1 m 3 capacity and a methane fermentation tank of 0.4 m 3 capacity (Figure 4.12). Few information are available about this plant, but the authors aim at
reducing the overall process HRT from 25 to 15 days and at
increasing both the decomposition of organic wastes from 60-65% to
80% and the energy recovery from 40-46% to 55% (AIST, 2004). Fig. 4.11 Outer view of the semi-pilot scale two-stage plant by AIST. Lastly, Wang and Zhao (2009) used a 0.2 m 3 rotating drum biohydrogen reactor integrated with a methanogenic CSTR of 0.8
m 3, fed with restaurant waste. They reported that 88% of the substrate was converted to biogas with a hydrogen yield of 0.065 m 3 H2/kg and methane yield of 0.055 m 3 CH4/kg. Lee and Chung (2010) used a 0.5 m 3 mesophilic CSTR biohydrogen reactor fed with a supernatant produced from a ground, pressed and
dehydrated food waste. The hydrogenic stage effluent was used to
fed a 2.3 m 3 methanogenic UASB reactor operated at an HRT of 3.6 days and 36 °C for more than 40 days. The authors reported a
hydrogen yield of 1.82 molH2/molglucose and relate little difference in 137 the production costs between the single stage AD and the two-stage
strategy, whose potential electricity generation was 10-12% more
than that of single stage.
4.3 Dark fermentation + Bioelectrochemical
Due to the limitations of current technologies, acidogenic digesters
for biohydrogen production and microbial fuel/electrolytic cells
could hardly be used as sole system for wastewater treatment.
Indeed, although the former is a good energy resource, its COD
removal and energy efficiency remain low. Therefore, the new
bioelectrochemical systems under development may be used for
recovering the 85% of the energy that remains in wastewaters after
hydrogen production by dark fermentation, in order to significantly
enhance the overall hydrogen-production rate and yield (Hallenbeck
and Ghosh, 2009), Otherwise, BES could act as a polishing step for
the removal of the residual organics present in the effluent of the
digester, which could still contain 0.5 to a few grams of residual
volatile fatty acids (van Lier et al., 2001).
The joint of these two bioprocesses into a two-stage treatment system
may also represent a driving force for overcoming the practical
limitations existing for BES application in real scale (due to high
cost of the electrode materials, low or unstable materials
performance and longevity at larger scale, low maximum achievable
current densities, unfavorable wastewater composition...; Logan,
2010). For example, although BESs have low requirements for
substrate specificity (Thygesen et al., 2010), their process
performances may be enhanced if the biomass used as feeding would
be subjected to pretreatment or other treatment processes (which
however must be cost effective in order to commercialize an efficient
and scalable MEC system). Therefore, high-strength carbohydrate- 138 rich wastewater can be at first treated by dark fermentation to
generate hydrogen and other metabolites favorable to be used in
MECs, such as aliphatic acids and alcohols (Liu et al., 2010).
The bioelectrochemical systems adopted after the first dark
fermentative stage could be both microbial fuel cells (MFC) or
microbial electrolysis cells (MEC) (Mohanakrishna et al., 2010;
Sharma and Li, 2010) and this choice could be based on that product
(electric energy or hydrogen gas) thought to be more useful for a
specific on-site application. MFC or MEC connected with anaerobic fermentative hydrogen
production processes has already been suggested (Pham et al., 2006;
Rozendal et al., 2008), however, nowadays very few two-stage
processes for the integration of these technologies are known or have
been published.
Manish and Banerjee (2008) proposed a two/three stage process
treating sugarcane biomass, where effluent of dark-fermentation
stage is sent to a biocatalyzed electrolyzer to produce hydrogen and
carbon dioxide (Figure 4.12). They assumed to use a two-chamber
BES, in order to produce H2 and CO2 in different chambers and to
avoid the use of pressure swing adsorption systems (or analogous)
for the hydrogen upgrading. Compared to a single stage dark-
fermentation process for biohydrogen production, the higher
hydrogen yields in electrochemically assisted process led to lower
need for sugarcane biomass input, thus increasing the energy
efficiency of the process. On the other hand, this process had the
least value of net energy ratio among the biohydrogen processes
considered (photo or dark single stage fermentation and two-stage
photo-dark fermentation), mainly because of significant electricity
consumption by the electrolyzer. Improvements in the cell design
and process parameters are required, in order to achieve less
electricity consumption. 139 Fig. 4.12 Schematic diagram of a two/three stage process applying AD systems and a
biocatalyzed electrolyzer (Courtesy of Manish and Banerjee, 2008). The first studies that really exploited actual hydrogenic fermentation
effluents for feeding MECs were those by Lalaurette et al. (2009)
and Lu et al. (2009).
Lalaurette experimented a two-stage process for converting
recalcitrant lignocellulosic material into hydrogen through a first
thermophilic dark-fermentation stage followed by electrohydrogenesis, with optimization of the two steps in separate
reactors. The lignocellulosic biomass was converted by the first stage
into hydrogen, carbon dioxide, acetic, formic, succinic, and lactic
acids, plus ethanol; then the residual volatile fatty acids (VFAs) and
alcohols were transformed into hydrogen gas by Pt-catalyzed MEC.
For the fermentation process, the authors used a pure culture of
Clostridium thermocellum and achieved 1.67 molH2/mol glucose at a
rate of 0.25 LH2/Lreactor d with a corn stover lignocellulose feed, or
1.64 molH2/mol glucose and 1.65 LH2/Lreactor d with a cellobiose feed. In
the second stage, hydrogen yields and production rates from the
fermentation effluents were 900 ± 140 LH2/kg COD and 0.96 ± 0.16
LH2/Lreactor d with cellobiose, while 750 ± 180 LH2/kg COD and 1.00 ±
0.19 LH2/Lreactor d with lignocellulose. Remarkable removal efficiency
(70-85% based on VFA removal or 65% as for the COD removal)
were obtained by MEC fed with the cellobiose and lignocellulose
effluent. In particular, the overall hydrogen yield of the process was 140 9.95 molH2/mol glucose for the cellobiose, a noticeable result near to the
maximal theoretical yield of 12 molH2/mol glucose, thus favoring the
techno-economic feasibility of the process. Lu et al. (2009) used the effluent of an ethanol-type dark-
fermentation CSTR reactor fed with a molasses wastewater as
substrate for a single-chamber MEC. The first stage was fed at an
organic loading rate of 22.8 kg COD/m 3 d and produced hydrogen gas at a maximum rate of 0.7 LH2/L reactor d and yield of 0.017 gH2/g COD
(equal to 0.27 molH2/mol COD). Its effluent had a pH in the range of
4.5-4.6, and a COD of 6500 ± 120 mg/L, mainly constituted by
ethanol and acetate in solution. Due to this low pH, the effluent was
added to the MEC both buffered (pH 6.7-7.0) or not. At an applied
voltage of 0.6 V, the MEC fed with buffered effluent achieved an
overall hydrogen recovery of 83 ± 4%, with a hydrogen production
rate of 1.41 ± 0.08 LH2/L reactor d. So, considering also the
fermentation system, the overall hydrogen recovery was 96%, with a
production rate of 2.11 LH2/L reactor d, corresponding to an electrical
energy efficiency of 287%. Moreover, this two-stage process showed
an electrical energy demand of only 1.12 kWh/m 3 H2, which is much less than that needed for water electrolysis (5.6 kWh/m 3 H2). Lastly, the authors reported that the addition of a buffer to the first stage
fermentation effluent was critical to MEC performance, as there was
little hydrogen production using unbuffered effluent (i.e. 0.037 LH2/L reactor d at Eap = 0.6 V and pH 4.5). Another couple of studies were later published about two-stage Dark
fermentation-MFC process.
Sharma and Li (2010) used as first stage a 2 liters biofermenter
(HPB) at continuous flow (Figure 4.13), fed at different organic
loading rates (OLR) by changing the substrate COD and the
hydraulic retention time. The reactor had a mixed microflora
enriched by soil inoculum, it was fed with glucose and its
temperature and pH were kept constant at 30 °C and 5.5 using a 141 build-in control system. The authors demonstrated that the hydrogen
yield by this first stage increased with the decrease of OLR, and
reached the maximum value of 2.72 molH2/mol glucose at the lowest
OLR of 4 g/L d. The effluent from the hydrogen producing fermenter
was then fed to a single chamber 100 mL glass bottle MFC, with
graphite fiber brush as anode and a carbon cloth containing 0.35
mg/cm 2 Pt as the cathode. The MFC power density increased with the increase in influent COD concentration and the highest power
density and coulombic efficiency of 4200 mW/m 3 and 5.3%, respectively, were reached at 6.3 gCOD/L. In MFC the energy
conversion efficiency reached the peak value of 4.24% at the OLR of
2.35 g/L d and then decreased with higher OLR. The combination of
the two system increased the overall energy recovery and the COD
removal of the system, which reached a maximum of 29% and 71%,
respectively. Fig. 4.13 Experimental setup of the continuous flow hydrogen fermenter plus the single
chamber MFC by Sharma and Li (2010). Mohanakrishna et al. (2010) integrated an acidogenic sequential
batch biofilm reactor (AcSBBR), producing H2 by fermenting
vegetable waste, with a single chamber MFC which generated
bioelectricity from the acid-rich effluents produced by the first stage 142 (Figure 4.14). AcSBBR was operated at high OLR (70.4 gCOD/L d)
and a maximum yield of 2.93 molH2/kg COD removed was reached,
showing an average COD removal efficiency of 28%. The effluent
produced had an high average VFA content of 7713 mgacetate equiv./L
(with acetate as the major metabolite (42%), followed by butyric acid
(23.5%) and propionic acid (34%)). This effluent was chemically
buffered to reach pH 7 and was then fed to the MFC at three variable
organic loading rates. Higher power output (111.76 mW/m 2) was observed at lower substrate loading condition and MFC was able to
efficiently remove effluent COD (80%), volatile fatty acids (79%),
carbohydrates (78%) and turbidity (65%). Thus, the authors
demonstrated the feasibility of utilizing acid-rich effluents by MFC
for both additional energy generation and wastewater treatment. Fig. 4.14 Schematic experimental design of two-stage system by Mohanakrishna et al. (2010).
AcSBBR: acidogenic sequencing batch biofilm reactor; FPT: feed preparation tank; IFPT:
intermediate feed preparation tank/outlet of AcSBBR collection tank; ECT: effluent collection
tank; MFC: microbial fuel cell; HST: H2 storage tank; P: peristaltic pump; T: preprogrammed
timer; V: volt meter; A: ammeter; PEM: proton exchange membrane; VR: variable resistor. However, two-stage dark fermentation-BES systems technology
need further research efforts and improvements to be applied in real

143 4.4 Other approaches - Dark fermentation + biopolymers production stage
In the chemical industry biomass can nowadays be fully exploited
integrating in a single facility various processing options to produce
simultaneously chemicals, fuels and energy.
Indeed, hydrogen production via dark fermentation could act as a
pre-treatment step followed by a process of bioconversion of volatile
fatty acids to other high added-value products, such as biopolymers
like polyhydroxyalkanoates (PHAs). PHAs are biodegradable
polyesters that specific bacteria can produce and accumulate under
stress conditions as intracellular storage reserves (in the form of
inclusion bodies/granules) of carbon and energy (Kessler et al.,
2001). There are several types of biopolyesters but poly(3-
hydroxybutyrate) and poly(3-hydroxyvalerate) are the most well-
known, since they bear similar structural properties to conventional
plastics such as polypropylene, polyethylene and polyvinylochloride.
They are also less penetrable to oxygen than conventional plastics,
thus being ideal as packaging material in food industry or in many
other applications of the medical sector. Therefore, PHAs may
substitute the nowadays widely used petrochemical plastics, thus
decreasing the environmental pollution due to anthropic activities
(Ntaikou et al., 2010a).
The production of PHAs from acidified wastewaters had been
previously investigated in both lab (Dionisi et al., 2005) and pilot
(Kellerhals et al., 2000) scale, showing very promising results. But it
was just in 2010 that Ntaikou et al. (2010b) proposed a two-stage
continuous system, which degraded a three phase olive mill
wastewater and simultaneously produced hydrogen via dark
fermentation in a CSTR and PHAs in an aerobic SBR using the first
stage effluent. In the first stage they achieved hydrogen production
rates between 165 and 202 mL/d (with a maximum yield of 196
LH2/kg consumed solids), while the second aerobic stage consumed 144 preferably butyrate to produce polyhydroxybutyrate, reaching a
PHAs yield of 8.94% (w/w) of dry biomass weight. Following these
promising results, the same authors lately reported the scale up of the
process, conducting the two stage PHAs production at semi-pilot
scale (Ntaikou et al., 2010a).
The exploitation of low-cost substrates/wastewaters, such as the
olive mill wastes used by Ntaikou et al. (2010b), together with the
use of highly productive microorganisms (some of them can
accumulate bioplastic up to 80% of their cell dry mass; Kim and
Lenz, 2001) may be the right way to increase the overall economical
viability of the process. However, issues concerning the separation of
the soluble biopolymers from the fermentation broth and the stability
of pure- or co-culture fermentation processes remain to be addressed.
In particular, the downstream processing efficiency (separation and
purification of the products from the bulk liquid) represents the
highest percentage of the manufacturing cost, therefore one of the
key goals is to adopt selective, efficient, and short separation routes. - The Maxifuel Danish Concept
The Danish Bioenergy Concept is a combinatory system approach
which can produce hydrogen, methane and ethanol altogether (Figure
4.15) (Westermann et al., 2007). In this system biomass is pre-
treated by wet oxidation to convert lignocellulosic compounds and to
increase the availability of fermentable sugars, which are
subsequently fermented to ethanol by yeast. Pentoses, which are not
converted by yeasts, are then fermented again to ethanol and also to
hydrogen in a (preferably) thermophilic fermentation process. Then
the effluent of this stage is used to produce methane, and therefore
the authors suggest to add it with manure as a cheap source of water
and as a way to increase production of methane. The system also
schedules to partly re-circulate the process water (after purification)
into the methane step. 145 The advantage of this strategy is the creation of a close-cycle system
and the possibility to specifically optimize each step for the
production of the preferred energy carrier (hydrogen, methane, and
ethanol) by choice of optimal microorganisms and operation
conditions. Currently the Danish Bioenergy Concept is optimized
with respect to ethanol production. Fig. 4.15 The Danish Bioenergy Concept (Courtesy of Westermann et al., 2007).


147 Chapter 5 Conclusions and final remarks

Hydrogen economy will be possible only with developments in
existing technical and engineering challenges (H2 upgrading, storage,
transport and end-use technologies) and by lowering the costs and
the sustainability of the present production strategies. Contrarily to
the expensive, energy intensive and fossil fuel-demanding productive
methods, our research efforts were dedicated exclusively to
biological production of hydrogen via anaerobic (dark) fermentation
of biomasses. Indeed, this biotechnology typically achieves high bio-
hydrogen production rate and could be an economical strategy for
waste disposal.
As first step, we managed to obtain efficient hydrogen-producing
mixed microbial cultures directly from acclimation/enrichment of
natural sources (soils and digested materials) (Paper II). Such type of
inoculum, used in batch reactors with proper substrate and 148 metabolites concentrations, allowed H2 yields comparable to those
reached by more expensive pure/selected/GM microbial cultures.
Being the Dark Fermentation process influenced by the cost,
availability and type of the raw material (biomasses) used, a new
methodology (BHP test) was established and applied to test the
biohydrogen production potential (BHP) of different organic
substrates of possible interest for future real-scale applications.
Organic wastes/wastewaters from agriculture, livestock and food
industry (in particular market bio-wastes, yielding 176 ± 2
NLH2/kgVS; Paper II) were proved to be attractive biomasses.
However, a concentrated mixture of organic fraction of municipal
solid waste, although it showed very high and repeatable BHP in
batch (202 ± 3 NLH2/kgVS), resulted in just 30% of its BHP when
fermented in laboratory scale continuously stirred tank reactors
(CSTRs), suggesting that further efforts are needed for future
applications of dark fermentation in full-scale plants (Paper II).
Also co-digestion of wastes was proposed and applied to
fermentative hydrogen production. Indeed, this process may suffer of
inhibition or instability due to volatile fatty acids production and pH
deviations from optimality. Therefore, the co-fermentation of
promptly degradable feedstock with alkali-rich materials, such as
livestock wastes, may represent a feasible and easy-to-implement
approach to avoid external adjustments of pH. The natural buffer
capacity of swine slurry was able to avoid pH drops in semi-
continuous and continuous processes when fed together with fruit
and vegetables residues. Optimal environment for high biohydrogen
production rate and yield (3.27 LH2/Lreactor d, 126 NLH2/kgVS added)
highly stable (deviations from daily average less than 14%) was
maintained, without any chemical addition for process control (Paper
I). Simultaneously, multi stage system strategies were investigated, in
order to extend past the present metabolic limitation of dark 149 fermentation (maximum yield 4 molH2/molglucose) and come closer to
the 10 molH2/molglucose required to approach costs competitive with
traditional fuels.
Considering that the major part (85%) of the feeding substrate energy
remains unused within the soluble compounds in the effluent of the
H2-producing process, association of this process with an anaerobic
digestion stage for methane generation was tested. A two-stage
laboratory-scale CSTR digester, fed with a mixture of agricultural
and livestock residues, was monitored for a long run (approximately
700 hours) (Paper III). High hydrogen yields (140 Ndm 3 H 2 kg -1 VS- added) were reached, with subsequent methane production of 351 Ndm 3 CH 4 kg -1 VS-added. However, overall energy recovery similar to that of a single-stage reactor run as control was produced (13-14 kJ
kg -1 VS-added) and partial inhibition of the methanogenic reactor of the two-stage process was assumed. Nevertheless, biogas with high CH4
content (''70%) was produced, that is advantageous for lowering the
biogas upgrading cost.
These results push to further in-depth researches about the two-stage
process, that will be done both on laboratory scale continuous
digesters and on a pilot scale two-stage reactor, located in the
experimental farm of the University of Milan.
Lastly, initial investigations about bioelectrochemical systems for
pure hydrogen biological production (MEC) were made. The variety
of fuel sources and high yield of high-purity H2 with just a relatively
small electrical energy input make this process a promising approach
and a new core technology for economically viable biohydrogen
production, particularly from biomass with low or negative economic
value. Furthermore if dark fermentation hydrogen production
effluent may be used as feeding substrate for MEC.
Paper IV explored the rate and the yield of gas (for real a mixture of
H2, CH4 and CO2) produced by membraneless MEC exploiting an
actual industrial wastewater with high methanol content, a compound 150 never before reported to be used in a MEC device. MEC energy
recovery was positive (3.76 and 3.38 kWh/kg TCOD removed with
platinum and molybdenum disulfide cathode, respectively) and 14-
16% higher than an AD lab-simulated process. Also the TCOD
removal efficiency was high (85-87%) with complete degradation of
methanol. MEC emerged to be competitive with the AD process,
especially using cheaper alternative catalysts such as molybdenum
disulfide (MoS2).
Starting from these preliminary remarks, a two- or three-stage
reactor, which combines the technologies here introduced and
studied (dark fermentation, bioelectrochemical system - MFC or
MEC - and anaerobic digestion) will be realized and studied by our
group within this year.

151 Chapter 6 References

Ahrens T., Weiland P., 2004. Electricity production from agricoltural
wastes through valorization of biogas. In: Resource Recovery and
Reuse in organic solid waste Management. Edited by Lens P.,
Hamelers B., Hoitink H., Bidlingmaier W., pp 395-410. IWA
Publishing, London, UK.

AIST, 2004. World First Biogas Plant to Recover Hydrogen and
Methane Quickly from Kitchen Waste. Translation of the AIST press
released on July 14, 2004.

Allen M.J., 1972. Cellular electrophysiology. In: Methods microbial.
Edited by Norris J.R., Ribbon D.W., pp 247-283. Academic Press,
New York, USA. 152 Akano T., Miura Y., Fukatsu H., Miyasaka K., Ikuta Y., Matsumoto
H., Hamasaki A., Shioji N., Mizoguchi T., Yagi K., Maeda I., 1996.
Hydrogen production by photosynthetic microorganisms. Appl.
Biochem. Biotechnol. 57/58, 677-688.

Akay G., Dogru M., Calkan O.F., Calkan B., 2005. Biomass
processing in Biofuels applications. In: Biofuels for Fuel cells -
Renewable energy from biomass fermentation. Edited by Lens P.,
Westermann P., Haberbauer M., Moreno A., pp 51-76. IWA
Publishing, London, UK.

Angenent L.T., Karim K., Al-Dahhan M.H., Wrenn B.A., Domìguez-
Espinosa R., 2004. Production of bioenergy and biochemicals from
industrial and agricultural wastewater. Trends Biotechnol. 22(9),

Antonopoulou G., Stamatelatou K., Venetsaneas N., Kornaros M.,
Lyberatos G., 2008. Biohydrogen and methane production from
cheese whey in a two-stage anaerobic process. Ind. Eng. Chem. Res.
47, 5227-5233.

Armor J.N., 1999. The multiple roles for catalysis in the production
of H2. Appl. Catal. A 176, 159-176.

Bauer C.G., Forest T.W., 2001. Effect of hydrogen addition on the
performance of methane-fueled vehicles. Part I: effect on S.I. engine
performance. Int. J. Hydrog. Energy 26, 55-70.

Bendixen H.J., 1994. Safeguards against Pathogens in Danish Biogas
Plants. Water Sci. Technol. 30(12), 171-180.

Benemann J.R., 1996. Hydrogen biotechnology: Progress and
prospects. Nat. Biotechnol. 14, 1101-1103.

Blonskaja V., Menert A., Vilu R., 2003. Use of two-stage anaerobic
treatment for distillery waste. Adv. Environ. Res. 7(3), 671-678.
153 Boeriu C.G., van Dam J.E.G., Sanders J.P.M., 2005. Biomass
valorisation for sustainable development. In Biofuels for Fuel cells -
Renewable energy from biomass fermentation. Edited by Lens P.,
Westermann P., Haberbauer M., Moreno A., pp 17-34. IWA
Publishing, London, UK.

Call D., Logan B.E., 2008. Hydrogen production in a single chamber
microbial electrolysis cell (MEC) lacking a membrane. Environ. Sci.
Technol. 42, 3401-3406.

Call D., Merrill M.D., Logan B.E., 2009. High surface area stainless
steel brushes as cathodes in microbial electrolysis cells (MECs).
Environ. Sci. Technol. 43(6), 2179-2183.

Cao X., Huang X., Liang P., Xiao K., Zhou Y., Zhang X., Logan
B.E., 2009. A new method for water desalination using microbial
desalination cells. Environ. Sci. Technol. 43(18), 7148-7152.

Chamberlain J.C., Foley H.M., MacDonald G.J., Ruderman M.A.,
1982. Climate effects of minor atmospheric constituents. In: Carbon
dioxide Review. Edited by Clark W.C., pp.255. Oxford University
Press, New York, USA.

Chang J.S., Lee K.S., Lin P.J., 2002. Biohydrogen production with
fixed-bed bioreactors. Int. J. Hydrog. Energy 27(11-12), 1167-1174.

Chen C.C., Lin C.Y., 2001. Start-up of anaerobic hydrogen
producing reactors seeded with sewage sludge. Acta Biotechnol.
21(4), 371-379.

Chen C.C., Lin C.Y., Chang J.S., 2001. Kinetics of hydrogen
production with continuous anaerobic cultures utilizing sucrose as
the limiting substrate. Appl. Microbiol. Biotechnol. 57(1-2), 56-64.

Cheng S., Logan B.E., 2007. Sustainable and efficient biohydrogen
production via electrohydrogenesis. PNAS 104(47), 18871-18873.
154 Cheng S., Logan B.E., 2007b. Ammonia treatment of carbon cloth
anodes to enhance power generation of microbial fuel cells.
Electrochem. Commun. 9, 492-496.

Cheng S.S., Chang S.M., Chen S.T., 2002. Effects of volatile fatty
acids on a thermophilic anaerobic hydrogen fermentation process
degrading peptone. Water Sci. Technol. 46(4-5), 209-214.

Cheng S., Xing D., Call D., Logan B.E., 2009. Direct biological
conversion of electrical current into methane by electromethanogenesis. Environ. Sci. Technol. 43(10), 3953-3958.

Choi H.B., Hwang K.Y., Shin E.B., 1997. Effects on anaerobic
digestion of sewage sludge pretreatment. Water Sci. Technol. 35(10),

Claassen P.A.M., Budde M.A.W., van Niel E.W.J., de Vrije T.,
2005. Utilization of biomass for hydrogen fermentation. In: Biofuels
for Fuel Cells - Renewable energy from biomass fermentation.
Edited by Lens P., Westermann P., Haberbauer M., Moreno A., pp
221-230. IWA Publishing, London, UK.

Cohen, B., 1931. The bacterial culture as an electrical half-cell. J.
Bacteriol. 21(1), 18-19.

Cooney M., Maynard N., Cannizzaro C., Benemann J., 2007.
Twophase anaerobic digestion for production of hydrogen''methane
mixtures. Bioresour. Technol. 98(14), 2641-2651.

Cusick R.D., Kiely P.D., Logan B.E., 2010. A monetary comparison
of energy recovered from microbial fuel cells and microbial
electrolysis cells fed winery or domestic wastewaters. Int. J. Hydrog.
Energy 35, 8855-8861.

Das D., Veziroglu T.N., 2001. Hydrogen production by biological
processes: survey of literature. Int. J. Hydrog. Energy 26, 13-28.
155 De Baere L., 2000. Anaerobic digestion of solid waste: state-of-the-
art. Water Sci. Technol. 41(3), 283-290.

De Mes T.Z.D., Stams A.J.M., Reith J.H., Zeeman G., 2005.
Methane production by anaerobic digestion of wastewater and solid
wastes. In: Bio-methane & Bio-hydrogen. Status and perspectives of
biological methane and hydrogen production. Edited by Reith J.H.,
Wijffels R.H., Barten H., pp 58-102. Dutch Biological Hydrogen
Foundation Publishing, Petten, The Netherlands.

De Vrije T., Claassen P.A.M., 2005. Dark hydrogen fermentations.
In: Bio-methane & Bio-hydrogen. Status and perspectives of
biological methane and hydrogen production. Edited by Reith J.H.,
Wijffels R.H., Barten H., pp 103-123. Dutch Biological Hydrogen
Foundation Publishing, Petten, The Netherlands.

Deluga G.A., Salge J.R., Schmidt L.D., Verykios X.E., 2004.
Renewable Hydrogen from Ethanol by Autothermal Reforming. Sci.
13(303), 993-997.

Dionisi D., Carucci G., Papini M.P., Riccardi C., Majone M.,
Carrasco F., 2005. Olive oil mill effluents as a feedstock for
production of biodegradable polymers. Water Res. 39, 2076-2084.

DiStefano T.D., Palomar A., 2010. Effect of anaerobic reactor
process configuration on useful energy production. Water Res. 44,

Ditzig J., Liu H., Logan B.E., 2007. Production of hydrogen from
domestic wastewater using a bioelectrochemically assisted microbial
reactor (BEAMR). Int. J. Hydrog. Energy 32, 2296-2304.

Dutton A.G., 2002. Hydrogen energy technology. Tyndall Working
Paper TWP 17. Edited by Tyndall Centre for Climate Change,
Norwich, UK.
156 Edwards P.P., Kutznetsov V.L., David W.I.F., 2007. Hydrogen
Energy. Phil. Trans. R. Soc. A 365(1853), 1043-1056.

Eggersdorfer M., Meijer J., Eckes P., 1992. Use of renewable
resources for non-food materials. FEMS Microbiol. Rev. 103, 355.

EUROBSERV'ER, 2010. Biogas Barometer. Systèmes Solaires: le
journal des énergies renouvelables 200, 104-119.

Fan Y., Sharbrough E., Liu H., 2008. Quantification of the internal
resistance distribution of microbial fuel cells. Environ. Sci. Technol.
42(21), 8101-8107.

Fang H.H.P., Liu H., 2002. Effect of pH on hydrogen production
from glucose by mixed culture. Bioresour. Technol. 82, 87-93.

Fang H.H.P., Zhang T., Liu H., 2002. Microbial diversity of
mesophilic hydrogen producing sludge. Appl. Microbiol. Biotechnol.
58, 112-118.

Galbe M., Zacchi G., 2002. Review production ethanol from
softwood. Appl. Microbiol. Biotechnol. 59, 618-628.

Gavala H.N., Yenal U., Skiadas I.V., Westermann P., Ahring B.K.,
2003. Mesophilic and thermophilic anaerobic digestionof primary
and secondary sludge. Effect of pre-treatment at 70°C. Water Res.
37(19), 4561-4572.

Gavala H.N., Skiadas I.V., Ahring B.K., 2006. Biological hydrogen
production in suspended and attached growth anaerobic reactor
systems. Int. J. Hydrog. Energy 31, 1164-1175.

Gomez X., Moran A., Cuetos M.J., Sanchez M.E., 2006. The
production of hydrogen by dark fermentation of municipal solid
wastes and slaughterhouse waste: A two phase process. J. Power
Sour. 157(2), 727-732.
157 Greenbaum E., 1988. Energetic Efficiency of Hydrogen Photoevolution by Algal Water Splitting. Biophys. J. 54, 365-368.

Guwy A.J., Dinsdale R.M., Kim J.R., Massanet-Nicolau J., Premier
G., 2011. Fermentative biohydrogen production systems integration.
Bioresour. Technol. 102, 8534-8542.

Hallenbeck P.C., 2009. Fermentative hydrogen production:
Principles, progress, and prognosis. Int. J. Hydrog. Energy 34, 7379-

Hallenbeck P.C., Ghosh D., 2009. Advances in fermentative
biohydrogen production: the way forward' Trends Biotechnol. 27(5),

Han S.K., Shin H.S. 2004a. Biohydrogen production by anaerobic
fermentation of food waste. Int. J. Hydrog. Energy 29(6), 569-577.

Han S.K, Shin H.S., 2004b. Performance of an Innovative two-stage
process converting Food Waste to Hydrogen and Methane. J. Air
Waste Manag. Assoc. 54, 242-249.

Han S.K., Kim S.H., Kim H.W., Shin H.S., 2005. Pilot-scale two-
stage process: a combination of acidogenic hydrogenesis and
methanogenesis. Water Sci. Technol. 52(1-2), 131-138.

Harnisch F., Sievers G., Schroder U., 2009. Tungsten carbide as
electrocatalyst for the hydrogen evolution reaction in pH neutral
electrolyte solutions. Appl. Catal. B. Environ. 89, 455-458.

Hawkes F.R., Dinsdale R., Hawkes D.L., Hussy I., 2002. Sustainable
fermentative hydrogen production: challenges for process optimisation. Int. J. Hydrog. Energy 27(11-12), 1339-1347.

Hawkes F.R., Hussy I., Kyazze G., Dinsdale R., Hawkes D.L., 2007.
Continuous dark fermentative hydrogen production by mesophilic 158 microflora: Principles and progress. Int. J. Hydrog. Energy

Holm-Nielsen J.B., Al Seadi T., 2004. Manure based biogas systems.
In: Resource recovery and Reuse in organic solid waste
Management. Edited by Lens P., Hamelers B., Hoitink H.,
Bidlingmaier W., pp 377-394. IWA Publishing, London, UK.

Holmes D.E., Chaudhuri S.K., Nevin K.P., Mehta T., Methe B.A.,
Ward J.E., Woodward T.L., Webster J., Lovley D.R., 2006.
Microarray and genetic analysis of electron transfer to electrodes in
Geobacter sulfurreducens. Environ. Microbiol. 8(10),1805-1815.

Hu H., Fan Y., Liu H., 2008. Hydrogen production using single-
chamber membrane-free microbial electrolysis cells. Water Res. 42,

Hu H., Fan Y., Liu H., 2009. Hydrogen production in single chamber
tubular microbial electrolysis cells using non-precious metal
catalysts (NiMo, NiW). Int. J. Hydrog. Energy 34, 8535-8542.

Hussy I., Hawkes F.R., Dinsdale R., Hawkes D.L., 2003. Continuous
fermentative hydrogen production from a wheat starch co-product by
mixed microflora. Biotechnol. Bioeng. 84, 619-626.

Hynek S., Fuller W., Bentley J., McCullough J., 1994. Hydrogen
energy progress X. Proceedings of the 10th World Hydrogen Energy
Conference; Cocoa Beach, Florida, USA; 22-24 June 1994.

Ike A., Murakawa T., Kawaguchi H., Hirata K., Miyamoto K., 1999.
Photoproduction of hydrogen from raw starch using a halophilic
bacterial community. J. Biosci. Bioeng. 88, 72-77.

ITABIA - Italian Biomass association, 2008. Goals of Bioenergy in
Italy- Report 2008, key elements for 2020 objectives. Edited by
ITABIA and Ministero dell'Ambiente e della Tutela del Territorio e
del Mare, Roma, ITALY. 159
Jackson D.D., Ellms J.W. 1896. On odors and tastes of surface
waters with special reference to Anabaena, a microscopial organism
found in certain water supplies in Massachusetts. Rep. Mass State
Board Health, 410-420.

Jang J.K., Chang I.S., Moon H., Kang K.H., Kim B.H., 2006.
Nitrilotriacetic acid degradation under microbial fuel cell
environment. Biotechnol. Bioeng. 95, 772-774.

Jones D.T., Woods D.R., 1986. Acetone-Butanol Fermentation
Revisited. Microbiol. Rev. 50(4), 484-524.

Kanai T., Imanaka H., Nakajima A., Uwamori K., Omori Y., Fukui
T., Atomi H, Imanaka T., 2005. Continuous hydrogen production by
the hyperthermophilic archaeon, Thermococcus kodakaraensis
KOD1. J. Biotechnol. 116(3), 271-282.

Kapdan I.K., Kargi F. 2006. Bio-hydrogen production from waste
materials. Enzyme Microb. Technol. 38, 569-582.

Kataoka N., Miya A., Kiriyama K., 1997. Studies on hydrogen
production by continuous culture system of hydrogen-producing
anaerobic bacteria. Water Sci. Technol. 36(6-7), 41-47.

Kellerhals M.B., Kessler B., Witholt B., Tchouboukov A., Brandl H.,
2000. Renewable long-chain fatty acids for production of
biodegradable medioum-chain-length polyhydroxyalkanoates (mcl-
PHAs) at laboratory and pilot scales. Macromol. 33, 4690-4698.

Kessler B., Weusthuis R., Witholt B., Eggink G., 2001. Production
of microbial polyesters: fermentation and downstream processes.
Adv. Biochem. Eng. 71, 159-182.

Khanal S.K., Chen W.H., Li L., Sung S., 2004. Biological hydrogen
production: effects of pH and intermediate products. Int. J. Hydrog.
Energy 29, 1123-1131. 160
Kim B.H., Ikeda T., Park H.S., Kim H.J., Hyun M.S., Kano K.,
Takagi K., Tatsumi H., 1999. Electrochemical activity of an Fe(III)-
reducing bacterium, Shewanella putrefaciens IR-1, in the presence of
alternative electron acceptors. Biotechnol. Tech. 13, 475-478.

Kim B.H., Park H.S., Kim H.J., Kim G.T., Chang I.S., Lee J., Phung
N.T., 2004. Enrichment of microbial community generating
electricity using a fuel-cell-type electrochemical cell. Appl.
Biotechnol. 63, 672-681.

Kim B.H., Chang I.S., Gadd G.M., 2007. Challenges in microbial
fuel cell development and operation. Appl. Microbiol. Biotechnol.
76, 485-494.

Kim D.H., Kim S.H., Kim K.Y., Shin H.S., 2010a. Experience of a
pilot-scale hydrogen-producing anaerobic sequencing batch reactor
(ASBR) treating food waste. Int. J. Hydrog. Energy 35(4), 1590-

Kim J., Dec J., Bruns M. A., Logan B. E., 2008. Odor removal from
swine wastewater using microbial fuel cells. Appl. Environ.
Microbiol. 74, 2540-2543.

Kim J.R., Premier G.C., Hawkes F.R., Rodriguez J., Dinsdale R.M.,
Guwy A.J., 2010b. Modular tubular microbial fuel cells for energy
recovery during sucrose wastewater treatment at low organic loading
rate. Bioresour. Technol. 101(4), 1190-1198.

Kim M.S., Lee T.J., Yoon Y.S., Lee I.G., Moon K.W., 2001.
Hydrogen production from food processing wastewater and sewage
sludge by anaerobic dark fermentation combined with photo-
fermentation. In Biohydrogen II: an approach to environmentally
acceptable technology. Edited by Miyake J., Matsunaga T., San
Pietro A.G., pp. 263-272. Pergamon, Oxford, UK.
161 Kim S.H., Han S.K., Shin H.S., 2006. Effect of substrate
concentration on hydrogen production and 16S rDNA-based analysis
of the microbial community in a continuous fermenter. Process
biochem. 41(1), 199-207.

Kim Y., Logan B.E., 2011. Microbial reverse electrodialysis cells for
synergistically enhanced power production. Environ. Sci. Technol.
45(13), 5834-5839.

Kim Y.B., Lenz. R.W., 2001. Polyesters from microorganisms. Adv.
Biochem. Eng. Biotechnol. 71, 51-79.

Kotay S.M., Das D. 2008. Biohydrogen as a renewable energy
resource: Prospects and potentials. Int. J. Hydrog. Energy 33, 258-

Kotsopoulos T.A., Fotidis I.A., Tsolakis N., Martzopoulos G.G.,
2009. Biohydrogen production from pig slurry in a CSTR reactor
system with mixed cultures under hyperthermophilic temperature.
Biomass Bioenergy 33, 1168-1174.

Koutrouli H.C., Kalfas H., Gavala H.N., Skiadas I.V., Stamatelatou
K., Lyberatos G., 2009. Hydrogen and methane production through
two-stage mesophilic anaerobic digestion of olive pulp. Bioresour.
Technol. 100(15), 3718-3723.

Kraemer J.T, Bagley D.M, 2005. Continuous Fermentative Hydrogen
Production Using a Two-Phase Reactor System with Recycle.
Environ. Sci. Technol. 39, 3819-3825.

Kumar N., Das D., 2000. Enhancement of hydrogen production by
Enterobacter cloacae IIT-BT 08. Process Biochem. 35, 589-593.

Kumar N., Das D., 2001. Continuous hydrogen production by
immobilized Enterobacter cloacae IIT-BT 08 using lignocellulosic
materials as solid matrices. Enzyme Microbiol. Technol. 29, 280-
287. 162
Kumar N., Ghosh A., Das D., 2001. Redirection of biochemical
pathways for the enhancement of H2 production by Enterobacter
cloacae. Biotechnol. Lett. 23, 537-541.

Kyazze G., Martinez-Perez N., Dinsdale R., Premier G.C., Hawkes
F.R., Guwy A.J., Hawkes D.L., 2006. Influence of substrate
concentration on the stability and yield of continuous biohydrogen
production. Biotechnol. Bioeng. 93, 971-979.

Lalaurette E., Thammannagowda S., Mohagheghi A., Maness P.C.,
Logan B.E., 2009. Hydrogen production from cellulose in a two-
stage process combining fermentation and electrohydrogenesis. Int.
J. Hydrog. Energy 34, 6201-6210.

Lay J.J., 2000. Modeling and optimization of anaerobic digested
sludge converting starch to hydrogen. Biotechnol. Bioeng.68, 269-

Lay J.J., Lee Y.J., Noike T., 1999. Feasibility of biological hydrogen
production from organic fraction of municipal solid waste. Water
Res. 33, 2579-2586.

Lay J.J., Fan K.S., Chang J.L., Ku C.H., 2003. Influence of chemical
nature of organic wastes on their conversion to hydrogen by heat-
shock digested sludge. Int. J. Hydrog. Energy 28, 1361-1367.

Lee H.S., Torres C.I., Parameswaran P., Rittmann B.E., 2009. Fate of
H 2 in an upflow single-chamber microbial electrolysis cell using a metal-catalyst-free cathode. Environ. Sci. Technol. 43, 7971-7976.

Lee S.Y., Yang S. H., Lee W. S., Kim H. S., Shin D. E., Ha J.K.,
2009. Effect of 2-Bromoethanesulfonic Acid on In vitro
Fermentation Characteristics and Methanogen Population. Asian-
Aust. J. Anim. Sci. 22(1), 42-48.
163 Lee Y., Chung J., 2010. Bioproduction of hydrogen from food waste
by pilot-scale combined hydrogen/methane fermentation. Int. J.
Hydrog. Energy 35(21), 11746-11755.

Levin D.B., Pitt L., Love M., 2004. Biohydrogen production:
prospects and limitations to practical application. Int. J. Hydrog.
Energy 29(2), 173-185.

Li C., Fang H.H.P., 2007. Fermentative hydrogen production from
wastewater and solid wastes by mixed cultures. Crit. Rev. Environ.
Sci. Technol. 37, 1-39.

Liang T.M., Wu K.L., Cheng S.S., 2001. Hydrogen production of
chloroform inhibited granular sludge. Proceedings of the IWA
WaterQual. 2001, Asia-Pacific Regional Conference; Fukuoka,
Japan, 863-868.

Liang T.M., Cheng S.S., Wu K.L., 2002. Behavioral study on
hydrogen fermentation reactor installed with silicone rubber
membrane. Int. J. Hydrog. Energy 27, 1157-1165.

Lies D.P., Hernandez M.E., Kappler A., Mielke R.E., Gralnick J.A.,
Newman D.K., 2005. Shewanella oneidensis MR-1 uses overlapping
pathways for iron reduction at a distance and by direct contact under
conditions relevant for biofilms. Appl. Environ. Microbiol. 71(8),

Lin C.Y., Chang R.C., 1999. Hydrogen production during the
anaerobic acidogenic conversion of glucose. J. Chem. Technol.
Biotechnol. 74, 498-500.

Lin C.Y., Chang R.C., 2004. Fermentative hydrogen production at
ambient temperature. Int. J. Hydrog. Energy 29, 715-720.

Lin C.Y., Lay C.H., 2004. Carbon/nitrogen-ratio effect on
fermentative hydrogen production by mixed microflora. Int. J.
Hydrog. Energy 29(1), 41-45. 164
Lin M., Ren N., Wang A., Wang X., 2003. Cooperation of mixed
culturing bacteria in the hydrogen production by fermentation. Chin.
J. Environ. Sci. 24(2), 54-59.

Liu D., 2008. Bio-hydrogen Production by Dark Fermentation from
Organic Wastes and Residues. Phd Thesis, Department of
Environmental Engineering, DTU Environment - Technical
University of Denmark, June 2008.

Liu D.W., Liu D.P., Zeng R.J., Angelidaki I., 2006. Hydrogen and
methane production from household solid waste in the two-stage
fermentation process. Water Res. 40(11), 2230-2236.

Liu G., Shen J., 2004. Effects of culture medium and medium
conditions on hydrogen production from starch using anaerobic
bacteria. J. Biosci. Bioeng. 98, 251-256.

Liu H., Ramnarayanan R., Logan B.E., 2004. Production of
electricity during wastewater treatment using a single chamber
microbial fuel cell. Environ. Sci. Technol. 38, 2281-2285.

Liu H., Zhang T., Fang H.P.P., 2003. Thermophilic H2 production
from cellulose containing wastewater. Biotechnol. Lett. 25, 365-369.

Liu H., Grot S., Logan B.E., 2005. Electrochemically assisted
production of hydrogen from acetate. Environ. Sci. Technol. 39,

Liu H., Hu H., Chignell J., Fan Y., 2010. Microbial electrolysis
novel technology for hydrogen production from biomass. Biofuels
1(1), 129-142.

Liu W.T., Chan O.C., Fang H.H.P., 2002. Microbial community
dynamics during start-up of acidogenic anaerobic reactors. Water
Res. 36(13), 3203-3213.
165 Liu X., Ren N., Song F., Yang C., Wang A., 2008. Recent advances
in fermentative biohydrogen production. Prog. Nat. Sci. 18, 253-258.

Logan B.E., 2004. Extracting hydrogen and electricity from
renewable resources. Environ. Sci. Technol., 160-167.

Logan B.E., 2005. Simultaneous wastewater treatment and biological
electricity generation. Water Sci. Technol. 52(1-2), 31-37.

Logan B.E., 2008. Microbial Fuel Cells. Edited by John Wiley &
Sons, Hoboken, USA.

Logan B.E., 2010. Scaling up microbial fuel cells and other
bioelectrochemical systems. Appl. Microbiol. Biotechnol. 85(6),

Logan B.E., Oh S.E., Kim I.S., Van Ginkel S., 2002. Biological
Hydrogen Production Measured in Batch Anaerobic Respirometers.
Environ. Sci. Technol. 36, 2530-2535.

Logan B.E., Hamelers B., Rozendal R., Schroder U., Keller J.,
Freguia S., Aelterman P., Verstraete W., Rabaey K., 2006. Microbial
fuel cells: methodology and technology. Environ. Sci. Technol.
40(17), 5181-5192.

Logan B.E., Call D., Cheng S., Hamelers H. V. M., Sleutels
T.H.J.A., Jeremiasse A.W., Rozendal R.A., 2008. Microbial
Electrolysis Cells for High Yield Hydrogen Gas Production from
Organic Matter. Environ. Sci. Technol. 42(23), 8630-8640.

Lovley D.R., 2006. Microbial fuel cells: novel microbial
physiologies and engineering approaches. Curr. Opin. Biotechnol.
17, 327-332.

Lu L., Ren N., Xing D., Logan B.E., 2009. Hydrogen production
with effluent from an ethanol-H2-coproducing fermentation reactor 166 using a single-chamber microbial electrolysis cell. Biosens.
Bioelectron. 24, 3055-3060.

Lyberatos G., Gavala H.N., Skiadas I.V., 2005. Modeling of biomass
fermentation: control of product formation. In: Biofuels for Fuel cells
- Renewable energy from biomass fermentation. Edited by Lens P.,
Westermann P., Haberbauer M., Moreno A., pp 95-118. IWA
Publishing, London, UK.

Mahadevan R., Bond D.R., Butler J.E., Esteve-Nunez A., Coppi
M.V., Palsson B.O., Schilling C.H., Lovley D.R., 2006. Genome-
based modeling of metabolism reveals physiological properties of
Geobacter sulfurreducens. Appl. Environ. Microbiol. 72, 1558-1568.

Mahyudin A.R., Furutani Y., Nakashimada Y., Kakizono T., Nishio
N., 1997. Enhanced hydrogen production in altered mixed acid
fermentation of glucose by Enterobacter aerogenes. J. Ferm.
Bioeng. 83(4), 358-363.

Manish S., Banerjee R., 2008. Comparison of biohydrogen
production processes. Int. J. Hydrog. Energy 33, 279-286.

Markov S.A., Weaver R., Seibert M., 1996. Hydrogen production
using microorganisms in hollow-fiber bioreactors. Energy Progress
XI: Proceedings of the 11th World Hydrogen Energy Conference;
Stuttgart, Germany; 23-28 June 1996.

Mata-Alvarez J., Mtzviturtia A., Llabresluengo P., Cecchi F., 1993.
Kinetic and Performance Study of a Batch 2-Phase Anaerobic-
Digestion of Fruit and Vegetable Wastes. Biomass Bioenergy 5(6),

Meulepas R.J.W., Nordberg A., Mata-Alvarez J., Lens P.N.L. 2005.
Methane production from wastewater, solid waste and biomass. In:
Biofuels for Fuel cells - Renewable energy from biomass
fermentation. Edited by Lens P., Westermann P., Haberbauer M.,
Moreno A., pp 121-138. IWA Publishing, London, UK. 167
Min B., Logan B.E., 2004. Continuous electricity generation from
domestic wastewater and organic substrates in a flat plate microbial
fuel cell. Environ. Sci. Technol. 38, 5809-5814.

Min B., Kim J.R., Oh S.E., Regan J.M., Logan B.E., 2005.
Electricity generation from swine wastewater using microbial fuel
cells. Water Res. 39, 4961-4968.

Mishra J., Khurana S., Kumar N., Ghosh A.K., Das D. 2004.
Molecular cloning, characterization, and overexpression of a novel
[Fe]-hydrogenase isolated from a high rate of hydrogen producing
Enterobacter cloacae IIT-BT 08. Biochem. Biophys. Res. Commun.
324(2), 679-685.

Mizuno O., Dinsdale R., Hawkes F.R., Hawkes D.L., Noike T.,
2000. Enhancement of hydrogen production from glucose by
nitrogen gas. Bioresour. Technol. 73, 59-65.

Mohanakrishna G., Venkata Mohan S., Sarma P.N., 2010. Utilizing
acid-rich effluents of fermentative hydrogen production process as
substrate for harnessing. Int. J. Hydrog. Energy 35, 3440-3449.

Momirlan M., Veziroglu T.N. 2002. Current status of hydrogen
energy. Renew. Sustain. Energy Rev. 6, 141-179.

Morimoto K., Kimura T., Sakka K., Ohmiya K., 2005.
Overexpression of a hydrogenase gene in Clostridium
paraputrificum to enhance hydrogen gas production. FEMS
Microbiol. Lett. 246(2), 229-234.

Morimoto M., Atsuko M., Atif A.A.Y., Ngan M.A., Fakhru''l-Razi
A., Iyuke S.E., Bakir A.M., 2004. Biological production of hydrogen
from glucose by natural anaerobic microflora. Int. J. Hydrog. Energy
29, 709-713.
168 Nandi R., Sengupta S., 1998. Microbial production of hydrogen: an
overview. Crit. Rev. Microbiol. 24, 61-84.

Nath K., Das D., 2003. Hydrogen from Biomass. Curr. Sci. 85(3),

Nath K., Das D., 2004. Improvement of fermentative hydrogen
production various approaches. Appl. Microbiol. Biotechnol. 65,

Nath K., Kumar A., Das D., 2005. Hydrogen production by
Rhodobacter sphaeroides strain O.U.001 using spent media of
Enterobacter cloacae strain DM11. Appl. Microbiol. Biotechnol. 68,

Noike T., Takabatake H., Mizuno O., Ohba M., 2002. Inhibition of
hydrogen fermentation of organic wastes by lactic acid bacteria. Int.
J. Hydrog. Energy 27, 1367-1371.

Ntaikou I., Antonopoulou G., Lyberatos G., 2010a. Biohydrogen
Production from Biomass and Wastes via Dark Fermentation: A
Review. Waste Biomass Valor. 1, 21-39.

Ntaikou I., Peroni C.V., Kourmentza C., Stoller M., Iliena V.I.,
Chianese A., Chiellini E., Lyberatos G., 2010b. Production of poly-
hydroxy-alkanoates (PHAs) from 3-phase oline oil mill wastewater
at a two stage system of semi-pilot scale. Proceedings of the 3rd
International Conference on Engineering for Waste and Biomass
Valorisation; Beijing, China; 17-19 May 2010.

Oh S.E., Logan B.E., 2005. Hydrogen and electricity production
from a food processing wastewater using fermentation and microbial
fuel cell technologies. Water Res. 39, 4673-4682.

Oh S.E., Van Ginkel S., Logan B.E., 2003. The Relative
Effectiveness of pH Control and Heat Treatment for Enhancing
Biohydrogen Gas Production. Environ. Sci. Technol. 37, 5186-5190. 169
Okamoto M., Miyahara T., Mizuno O., Noike T., 2000. Biological
hydrogen potential of materials characteristic of the organic fraction
of municipal solid wastes. Water Sci. Technol. 41(3), 25-32.

Okkerse C., van Bekkum H., 1999. From fossil to green. Green
Chem. 1, 107-114.

Park D.H., Zeikus J.G., 2003. Improved fuel cell and electrode
designs for producing electricity from microbial degradation.
Biotechnol. Bioeng. 81, 348-355.

Pham T. H., Rabaey K., Aelterman P., Clauwaert P., De
Schamphelaire L., Boon N., Verstraete W., 2006. Microbial Fuel
Cells in Relation to Conventional Anaerobic Digestion Technology.
Eng. Life Sci. 6(3), 285-292.

Piccinini S., 2008. Biogas in Italia: lo stato dell'arte. Proceedings of
CRPA (centro ricerche produzioni animali) national meeting,
Digestione anaerobica: opportunità per l'agricoltura e per l'ambiente;
Milano, Italy; 25 gennaio 2008.

Potter M.C., 1911. Electrical effects accompanying the decomposition of organic compounds. Proc. R. Soc. London Ser. B
84(571), 260-276.

Rabaey K., Rozendal R.A., 2010. Microbial electrosynthesis:
revisiting the electrical route for microbial production. Nat. Rev.
Microbiol. 8, 706-716.

Rabaey K., Verstraete W., 2005. Microbial fuel cells: novel
biotechnology for energy generation. Trends Biotechnol. 23(6), 291-

Rabaey K., Lissens G., Siciliano S.D., Verstraete W., 2003. A
microbial fuel cell capable of converting glucose to electricity at high
rate and efficiency. Biotechnol. Lett. 25(18), 1531-1535. 170
Rabaey K., Boon N., Siciliano S.D., Verhaege M., Verstraete W.,
2004. Biofuel cells select for microbial consortia that self-mediate
electron transfer. Appl. Environ. Microbiol. 70, 5373-5382.

Rabaey K., Boon N., Hofte M., Verstraete W., 2005. Microbial
phenazine production enhances electron transfer in biofuel cells.
Environ. Sci. Technol. 39, 3401-3408.

Radmann E.M., Reinehr C.O., Costa J.A.V., 2007. Optimization of
the repeated batch cultivation of microalga Spirulina platensis in
open raceway ponds. Aquacult. 265(1-4), 118-126.

Reguera G., McCarthy K.D., Mehta T., Nicoll J., Tuominen M.T.,
Lovley D.R., 2005. Extracellular electron transfer via microbial
nanowires. Nat. 435, 1098-1101.

Reith J.H., Wijffels R.H., Barten H., 2005. Introduction: the
perspectives of biological methane and hydrogen production. In:
Bio-methane & Bio-hydrogen. Status and perspectives of biological
methane and hydrogen production. Edited by Reith J.H., Wijffels
R.H. and Barten H., pp 103-123. Dutch Biological Hydrogen
Foundation Publishing, Petten, The Netherlands.

Ren N.Q., Lin M., Ma X.P., Wang A.J., Li J., 2003. A strain of
anaerobic bacteria screened for high efficient hydrogen production
and its aciduric character. Acta Energ. Solaris Sin. 24(1), 80-84.

Ren N.Q., Song J.X., An D., Zhang R.J., 2006. Effects of terminal
products on hydrogen production by ethanol hydrogen-producing
microflora. Environ. Sci. 27(8), 1608-1612.

Rezaei F., Xing D., Wagner R., Regan J.M., Richard T.M., Logan
B.E., 2009. Simultaneous cellulose degradation and electricity
production by Enterobacter cloacae in a microbial fuel cell. Appl.
Microbiol. Biotechnol. 75, 3673-3678.
171 Rifkin J., 2002. The Hydrogen Economy. Edited by Tarcher/Putnam,
New York, USA.

Rozendal R.A., Buisman C.J.N., 2005. Bio-electrochemical process
for producing hydrogen. Patent WO 2005005981.

Rozendal R.A., Hamelers H.V.M., Euverink G.J.W., Metz S.J.,
Buisman C.J.N., 2006a. Principle and perspectives of hydrogen
production through biocatalyzed electrolysis. Int. J. Hydrog. Energy
31, 1632-1640.

Rozendal R.A., Hamelers H.V.M., Buisman C.J.N., 2006b. Effects of
membrane cation transport on pH and Microbial Fuel cell
Performance. Environ. Sci. Technol. 40, 5206-5211.

Rozendal R.A., Hamelers H.V.M., Molenkamp R.J., Buisman,
C.J.N., 2007. Performance of single chamber biocatalyzed
electrolysis with different types of ion exchange membranes. Water
Res. 41, 1984-1994.

Rozendal R.A., Jeremiasse A.W., Hamelers H.V.M., Buisman,
C.J.N., 2008. Hydrogen production with a microbial biocathode.
Environ. Sci. Technol. 42, 629-634.

Schroder C., Selig M., Schonheit P., 1994. Glucose Fermentation to
Acetate, CO2 and H2 in the Anaerobic Hyperthermophilic
Eubacterium Thermotoga Maritima: Involvement of the Embden-
Meyerhof Pathway. Arch. Microbiol. 161, 460-470.

Selembo P.A., Merrill M.D., Logan B.E., 2009. The use of stainless
steel and nickel alloys as low-cost cathodes in microbial electrolysis
cells. J. Power Sour. 190, 271-278.

Sharma Y., Li B., 2010. Optimizing energy harvest in wastewater
treatment by combining anaerobic hydrogen producing biofermentor
(HPB) and microbial fuel cell (MFC). Int. J. Hydrog. Energy 35,
3789-3797. 172
Shin H.S., Youn J.H., Kim S.H., 2004. Hydrogen production from
food waste in anaerobic mesophilic and thermophilic acidogenesis.
Int. J. Hydrog. Energy 29, 1355-1363.

Show K.Y., Lee D.J., Zhang Z.P., 2011. Production of Biohydrogen:
Current Perspectives and Future Prospects. In: Biofuels: Alternative
feedstocks and Conversion Processes. Edited by Pandey A., Larroche
C., Ricke S.C., Dussap C.G., Gnansounou E., pp 467-478. Academic
Press, Oxford, UK.

Shukla A.K., Suresh P., Berchmans S., Rajendran A., 2004.
Biological fuel cells and their applications. Curr. Sci. 87(4), 455-468.

Soetaert W., Vandamme E., 2005. Biofuels production from
agricultural crops. In Biofuels for Fuel cells - Renewable energy
from biomass fermentation. Edited by Lens P., Westermann P.,
Haberbauer M., Moreno A., pp 37-50. IWA Publishing, London,

Sung S., Bazylinski D.A., Raskin L., 2003. Biohydrogen Production
from Renewable Organic Wastes. FY 2003 Progress Report, 1-5.

Taguchi F., Mizukami N., Saito-Taki T., Hasegawa K., 1995.
Hydrogen production from continuous fermentation of xylose during
growth of Clostridium sp. strain No. 2. Can. J. Microbiol. 41, 536-

Tanisho S., Kuromoto M., Kadokura N., 1998. Effect of CO2
removal on hydrogen production by fermentation. Int. J. Hydrog.
Energy 23(7), 559-563.

Tao Y., Chen Y., Wu Y., He Y., Zhou Z., 2007. High hydrogen yield
from a two step process of dark- and photo-fermentation of sucrose.
Int. J. Hydrog. Energy 32, 200-206.
173 Tartakovsky B., Manuel M.F., Wang H., Guiot S.R., 2009. High rate
membrane-less microbial electrolysis cell for continuous hydrogen
production. Int. J. Hydrog. Energy 34, 672-677.

Temudo M.F., Kleerebezem R., van Loosdrecht M., 2007. Influence
of the pH on (open) mixed culture fermentation of glucose: A
chemostat study. Biotechnol. Bioeng. 98(1), 69-79.

Thygesen A., Thomsen A.B., Possemiers S., Verstraete W., 2010.
Integration of Microbial Electrolysis Cells (MECs) in the Biorefinery
for Production of Ethanol, H2 and Phenolics. Waste Biomass Valor.
1, 9-20.

U.S. Department of energy, 2001. Hydrogen use in internal
combustion engines (module 3). In: Hydrogen Fuel Cell Engines and
Related Technologies.

Ueno Y., Kawai T., Sato S., Otsuka S., Morimoto M., 1995.
Biological Production of Hydrogen from Cellulose by Natural
anaerobic microflora. J. Ferment. Bioeng. 79(4), 395-397.

Ueno Y., Fukui H., Goto M., 2007a. Operation of a Two-Stage
Fermentation Process Producing Hydrogen and Methane from
Organic Waste. Environ. Sci. Technol. 41, 1413-1419.

Ueno Y., Tatara M., Fukui H., Makiuchi T., Goto M., Sode K.,
2007b. Production of hydrogen and methane from organic solid
wastes by phase-separation of anaerobic process. Bioresour. Technol.
98, 1861-1865.

Valdez-Vazquez I., Rios-Leal E., Esparza-Garcia F., Cecchi F.,
Poggi-Varaldo H.M., 2005. Semi-continuous solid substrate
anaerobic reactors for H2 production from organic waste: Mesophilic
versus thermophilic regime. Int. J. Hydrog. Energy 30, 1383-1391.
174 Van Ginkel S., Logan B.E., 2005a. Increased biological hydrogen
production with reduced organic loading. Water Res. 39, 3819-3826.

Van Ginkel S., Logan B.E., 2005b. Inhibition of biohydrogen
production by undissociated acetic and butyric acids. Environ. Sci.
Technol. 39(23), 9351-9356.

Van Ginkel S., Sung S., Lay J.J., 2001. Biohydrogen Production as a
Function of pH and Substrate Concentration. Environ. Sci. Technol.
35(24), 4726-4730.

Van Ginkel S., Oh S.E., Logan B.E., 2005. Biohydrogen production
from food processing and domestic wastewaters. Int. J. Hydrog.
Energy 30, 1535-1542.

Van Groenestijn J.W., Hazewinkel J.H.O., Nienoord M., Bussmann
P.J.T., 2002. Energy aspects of biological hydrogen production in
high rate bioreactors operated in the thermophilic temperature range.
Int. J. Hydrog. Energy 27(11-12), 1141-1147.

Van Lier J.B., Grolle K.C.F., Frijters C.T.M.J., Stams A.J.M.,
Lettinga G., 1993. Effects of Acetate, Propionate, and Butyrate on
the Thermophilic Anaerobic Degradation of Propionate by
Methanogenic Sludge and Defined Cultures. Appl. Environ.
Microbiol. 59(4), 1003-1011.

Van Niel E.W.J., Budde M.A.W., de Haas G.G., van der Wal F.J.,
Claasen P.A.M., Stams A.J.M., 2002. Distinctive properties of high
hydrogen producing extreme thermophiles, Caldicellulosiruptor
saccharolyticus and Thermotoga elfii. Int. J. Hydrog. Energy 27(11-
12), 1391-1398.

Veziroglu T.N., Barbir F., 1992. Hydrogen: the wonderful fuel. Int.
J. Hydrog. Energy 17,391-398.
175 Wagner R.C., Regan J.M., Oh S.E., Zuo Y., Logan B.E., 2009.
Hydrogen and methane production from swine wastewater using
microbial electrolysis cells. Water Res. 43, 1480-1488.

Wang C.C., Chang C.W., Chu C.P., Lee D.J., Chang B.V., Liao C.S.,
Tay J.H., 2003a. Using filtrate of waste biosolids to effectively
produce bio-hydrogen by anaerobic fermentation. Water Res. 37,

Wang Y., Ren N.Q., Sun Y.J., 2003b. Analysis of the ferment
process and prohydrogen ability of pro-hydrogen and ferment
bacterium influenced by Fe. Acta Energ. Solaris Sin. 24(2), 222-226.

Wang X., Zhao Y., 2009. A bench scale study of fermentative
hydrogen and methane production from food waste in integrated
two-stage process. Int. J. Hydrog. Energy 34(1), 245-254.

Weiland P., 2000. Anaerobic waste digestion in Germany - Status
and recent developments. Biodegrad. 11, 415-421.

Weiland P., Rieger C., Ehrmann T. 2003. Evaluation of the newest
biogas plants in Germany with respect to renewable energy
production, greenhouse gas reduction and nutrient management. In:
Proceedings of the future of Biogas in Europe II, European Biogas
workshop; Esbjerg, Denmark; 2-4 October 2003.

Westermann P., Jorgensen B., Lange L., Ahring B.K., Christensen
C.H., 2007. Maximizing renewable hydrogen production from
in a bio/catalytic refinery. Int. J. Hydrog. Energy 32, 4135-4141.

Winkler M., Hemschemeier A., Gotor C., Melis A., Happer T., 2002.
[Fe]-hydrogenases in green algae: photo-fermentation and hydrogen
evolution under sulfur deprivation. Int. J. Hydrog. Energy 27, 1431-
176 Winter M., Brodd R.J., 2004. What are batteries, fuel cells, and
supercapacitors' Chem. Rev. 104, 4245-4269.

Woodward J., Mattingly S.M., Danson M., Hough D., Ward N.,
Adams M., 1996. In vitro hydrogen production by glucose
dehydrogenase and hydrogenase. Nat. Biotechnol. 14, 872-874.

Woodward J., Orr M., Cordray K., Greenbaum E., 2000. Enzymatic
production of biohydrogen. Nat. 405(29), 1014-1015.

Wrana N., Sparling R., Cicek N., Levin D.B., 2010. Hydrogen gas
production in a microbial electrolysis cell by electrohydrogenesis. J.
Clean. Prod. 18, S105-S111.

Wu S.Y., Lin C.N., Chang J.S., 2003. Hydrogen production with
immobilized sewage sludge in three-phase fluidized bed bioreactors.
Biotechnol. Prog. 19, 828-832.

Wu X., Zhu J., Dong C., Miller C., Li, Y., Wang L., Yao W., 2009.
Continuous biohydrogen production from liquid swine manure
supplemented with glucose using an anaerobic sequencing batch
reactor. Int. J. Hydrog. Energy 34, 6636-6645.

Xing D.F., Ren N.Q., Li Q., Lin M., Wang A., Zhao L., 2006.
Ethanoligenens harbinense gen nov., sp. Nov., isolated from
molasses wastewater. Int. J. Syst. Evol. Microbiol. 56(4), 755-760.

Xing D., Zuo Y., Cheng S., Regan J.M., Logan B.E., 2008.
Electricity generation by Rhodopseudomonas palustris DX-1.
Environ. Sci. Technol. 42(11), 4146-4151.

Yamin J.A.A., 2006. Comparative study using hydrogen and
gasoline as fuels: Combustion duration effect. Int. J. Energy Res.

Yamin J.A.A., Gupta H.N., Bansal B.B., Srivastava O.N., 2000.
Effect of combustion duration on the performance and emission 177 characteristics of a spark ignition engine using hydrogen as a fuel.
Int. J. Hydrog. Energy 25(6), 581-589.

Yokoi H., Ohkawara T., Hirose J., Hayashi S., Takasaki Y., 1995.
Characteristics of Hydrogen Production by Aciduric Enterobacter
aerogenes Strain HO-39. J Ferment. Bioeng. 80(6), 571-574.

Yokoi H., Tokushige T., Hirose J., Hayashi S., Takasaki Y., 1998. H2
production from starch by mixed culture of Clostridium buytricum
and Enterobacter aerogenes. Biotechnol. Lett. 20, 143-147.

Yokoi H., Saitsu A.S., Uchida H., Hirose J., Hayashi S., Takasaki Y.,
2001. Microbial hydrogen production from sweet potato starch
residue. J. Biosci. Bioeng. 91, 58-63.

Yokoi H., Maki R., Hirose J., Hayashi S., 2002. Microbial
production of hydrogen from starch-manufacturing wastes. Biomass
Bioenergy 22(5), 389-395.

Yokoyama H., Ohmori H., Ishida M., Waki M., Tanaka Y., 2006.
Treatment of cow-waste slurry by a microbial fuel cell and the
properties of the treated slurry as a liquid manure. Anim. Sci. J. 77,

Yokoyama H., Moriya N., Ohmori H., Waki M., Ogino A., Tanaka
Y., 2007. Community analysis of hydrogen-producing extreme
thermophilic anaerobic microflora enriched from cow manure with
five substrates. Appl. Microbiol. Biotechnol. 77(1), 213-222.

Yoshida A., Nishimura T., Kawaguchi H., Inui M., Yukawa H.,
2005. Enhanced hydrogen production from formic acid by formate
hydrogen lyase-overexpressing Escherichia coli strains. Appl.
Environ. Microbiol. 71(11), 6762-6768.

You S.J., Zhao Q.L., Jiang J.Q., Zhang J.N., Zhao S.Q., 2006.
Sustainable approach for leachate treatment: electricity generation in 178 microbial fuel cell. J. Environ. Sci. Health A - Environ. Sci. Eng.
Toxic Hazard Subst. Control 41, 2721-2734.

Yu H., Zhu Z., Hu W., Zhang H., 2002. Hydrogen production from
rice winery wastewater in an uplow anaerobic reactor by mixed
anaerobic cultures. Int. J. Hydrog. Energy 27, 1359-1365.

Zhang T., Liu H., Fang H.H.P., 2003. Biohydrogen production from
starch in wastewater under thermophilic conditions. J. Environ.
Manag. 69, 149-156.

Zhu J., Li Y., Wu X., Miller C., Chen P., Ruan R., 2009. Swine
manure fermentation for hydrogen production. Bioresour. Technol.
100, 5472-5477.

Zuo J., Zuo Y., Zhang W., Chen J., 2005. Anaerobic bio-hydrogen
production using pre-heated river sediments as seed sludge. Water
Sci. Technol. 52(10), 31-39.

Zurawski D., Meyer M., Stegmann R., 2005. Fermentative
production of biohydrogen from biowaste using digested sewage
sludge as inoculum. Proceedings of Sardinia 2005, 10th International
Waste Management and Landfill Symposium; S. Margherita di Pula,
Italy; 3-7 October 2005.

179 Acknowledgements

I would like to thank all those who have contributed to the
completion of this thesis work. Prof. Bodria and Dr. Oberti for having introduced me in this
interesting and innovative research topic and having supervised and
advised me patiently during these years. Prof. Adani for having supported this research with place,
opportunities and time and for having extended my scientific
knowledge. Prof. Logan, for having taught to me everything I know about BES
systems and having given me a warm welcome into his research
group. All my colleagues of Diprove (Ricicla Group) for supporting and
sharing expertise, successes&failures (and lunchbreaks) together. All my colleagues of the Logan Group..I felt at home, even at 6,000
miles away. All the students and the colleagues I collaborated with during these
years...glad to be improved together! My family, relatives and Daniela, for all their advices,
encouragements and love: without you I couldn't have made it! 180 APPENDIX A International Refereed Papers _ I_
Biohydrogen from thermophilic co-fermentation of swine manure with fruit
and vegetable waste: Maximizing stable production without pH control A. Tenca a, A. Schievano b, F. Perazzolo a, F. Adani b, R. Oberti a,'' a Department of Agricultural Engineering, Università degli Studi di Milano, Via Celoria 2, 20133 Milano, Italy
b Department of Crop Science, Università degli Studi di Milano, Via Celoria 2, 20133 Milano, Italy a r t i c l e i n f o Article history:
Received 1 December 2010
Received in revised form 28 March 2011
Accepted 30 March 2011
Available online 3 April 2011 Keywords:
Swine manure
Fruit and vegetables waste
Response surface analysis a b s t r a c t Hydrogen production by dark fermentation may suffer of inhibition or instability due to pH deviations
from optimality. The co-fermentation of promptly degradable feedstock with alkali-rich materials, such
as livestock wastes, may represent a feasible and easy to implement approach to avoid external adjust-
ments of pH. Experiments were designed to investigate the effect of the mixing ratio of fruit''vegetable waste with swine manure with the aim of maximizing biohydrogen production while obtaining process stability
through the endogenous alkalinity of manure. Fruit''vegetable/swine manure ratio of 35/65 and HRT of 2 d resulted to give the highest production rate of 3.27 ± 0.51 LH2 Lÿ 1 dÿ1, with a corresponding hydrogen yield of 126 ± 22 mL H2 g ÿ1 VS-added and H2 content in the biogas of 42 ± 5%. At these operating conditions the process exhibited also one of the high-
est measured stability, with daily productions deviating for less than 14% from the average. ' 2011 Elsevier Ltd. All rights reserved. 1. Introduction In the last decade, scientific and societal interests in possible applications of hydrogen as an energy carrier have grown tremen-
dously due to hydrogen''s unique environmental features. Indeed,
molecular hydrogen H2 stores the highest amount of chemical
energy per mass unit (HHV = 142 MJ/kg) and can release it by react-
ing with oxygen, i.e., air, virtually without any exhaust emissions in
the atmosphere except for water. Interest in hydrogen energy appli-
cations are also associated with very high conversion efficiency
(45''60%), obtained with fuel cells as well with new homogeneous
charge compression ignition (HCCI) engines (efficiency close to
45%) (Edwards et al., 2007; Gomes Antunes et al., 2008). Despite present technological limitations and challenges, hydrogen is considered to be a possible energy carrier of the future,
and developing sustainable methods to obtain hydrogen from
renewable sources, in substitution of current fossil-fuel based
technologies, is paramount in order to fully achieve all the poten-
tial benefits and environmental sustainability. A first significant contribution to this goal may come from bio- hydrogen, i.e., hydrogen produced via biological conversion of
organic matter by microbial consortia. In particular, mixed culture
fermentation of substrates such as organic waste or biomass
appears to be one of the most promising approaches able to bring in the next future the distributed production of renewable hydro-
gen, sustainable also on smaller scales. Biohydrogen production by acidogenic fermentation under mesophilic, thermophilic and hyperthermophilic conditions has
been demonstrated for a number of organic substrates especially
rich in carbohydrates, such as simple sugars (often used as a
model substrate), starch, sugar beets and potatoes processing
wastewaters, cheese whey, brewery waste, etc. Hydrogen yields
obtained with these substrates typically range from 50 to
150 mLH2 gÿ1 VS , depending on the biodegradability and complexity of the substrate, and operation mode (Li et al., 2008; Shin and
Youn, 2005; Venetsaneas et al., 2009; Yokoi et al., 2002) while
that range can easily exceed 200 mLH2 gÿ1 hexose for solutions of pure simple sugars (Aceves-Lara et al., 2008; Fang and Liu,
2002; Lin and Cheng, 2006; Van Ginkel and Logan, 2005; Zheng
and Yu, 2005). The equilibrium of the fermentation process can be disturbed by many biochemical and operating parameters. When promptly
biodegradable substrates are used as feedstock, one of the most
frequent factors of process inhibition or instability is the rapid
build-up of undissociated volatile fatty acids (VFA) accompanying
hydrogen production by acidogenic consortia. This leads to the
depletion of medial buffer capacity, directly affecting the pH,
which is known to play a crucial role in maintaining metabolic
pathways and bacterial communities in an optimal domain for bio-
hydrogen production (Lay, 2000; Lee et al., 2002; Mu et al., 2006;
Ye et al., 2007; Zheng and Yu, 2005). Indeed, even if significant
production has been reported for a wider range of conditions, 0960-8524/$ - see front matter ' 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biortech.2011.03.102 '' Corresponding author. Tel.: +39 0250316867; fax: +39 0250316845. E-mail address: (R. Oberti). Bioresource Technology 102 (2011) 8582''8588 Contents lists available at ScienceDirect Bioresource Technology j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / b i o r t e c h fermentation pHs from 5.0 to 6.0 are very often considered optimal
to avoid both methanogenesis and solventogenesis and to enhance
biohydrogen generation (Hawkes et al., 2002; Van Ginkel et al.,
2001; Wu et al., 2010). pH levels lower than 4.5 are known to neg-
atively affect the activity of hydrogenase enzyme (Dabrock et al.,
1992; Van Ginkel et al., 2001), driving the microbial community to-
ward other fermentation pathways; on the other hand, a neutral or
weakly alkaline pH favors methanogenic and homoacetogenic mi-
crobes development and their direct consumption of produced
hydrogen, which can therefore be avoided by guaranteeing weak
acidity in the environment. To maintain the conditions within an optimal range, pH adjust- ment with the addition of exogenous acids and bases is often pro-
posed as a process control solution in laboratory conditions. On the
other hand, this approach may not be optimal for large-scale trans-
fers, and when looking for full-scale applications different strate-
gies may be considered for maintaining acceptable chemical
equilibrium in fermentation broth. For example, livestock manure is an abundant source of alkali and nutrients for cell growth, thus representing an ideal co-
substrate for fermentation with carbohydrate rich and promptly
degradable materials. In intensive animal industry areas, livestock
waste could then represent a primary and renewable co-substrate
for biohydrogen production. In Italy, for example, the pig industry
consists of 9.5 millions animals and is mainly concentrated in
highly specialized districts. It can be estimated (data not shown)
that in the Po Valley area, more than 10 million tons of swine man-
ure are produced yearly on a territory of about 5000 km 2. Such waste density levels are a major cause of environmental and soci-
etal problems, when not properly managed. Demonstrating the sustainability of using pig slurries as a renewable source for supplying the production of biohydrogen
would disclose new opportunities in a high-grade energy vector
chain, while treating large amounts of agricultural waste. Only a few studies have addressed the potential use of swine manure as feedstock for biohydrogen production. When used as a
single substrate alone, very little biohydrogen could be recovered
from fermentation both at mesophilic temperatures (Wagner
et al., 2009), as well as at hyperthermophilic temperatures, with
production yields lower than 4 mLH2 gÿ1 VS (Kotsopoulos et al., 2009). On the other hand, hydrogen yield as high as 203 mLH2 gÿ1 hexose (Wu et al., 2009) and as 209 mLH2 gÿ1 hexose (Zhu et al., 2009) were ob- tained when swine manure was added to glucose, reinforcing the
hypothesis that it is a suitable co-substrate to be fermented by a
mixture culture in addition to a carbohydrate-rich, promptly
hydrolysable material. Fruits and vegetables waste could be massively available to appropriate waste management approaches, and they represent a
form of highly degradable feedstock to be used for biohydrogen
production. Indeed, batch experiments at mesophilic temperatures
conducted on cabbage and carrots pulp resulted in a maximal yield
of 62 and 71 mLH2 gÿ1 VS , respectively (Okamoto et al., 2000), while a composite mixture of vegetables reached 89 mLH2 gCODÿ 1 (Venkata Mohan et al., 2009b); lettuce and potato yielded 50 and
106 mLH2 gÿ1 VS , respectively (Dong et al., 2009). High biohydrogen yields were also obtained with fruit waste such as sweet lime peel-
ings extracts (Venkata Mohan et al., 2009a) and jackfruit peel
(Vijayaraghavan et al., 2006), obtaining biohydrogen yields of
76 mLH2 gÿ1 COD and 198 mLH2 g ÿ1 VS , respectively. The objective of this work was to study the use of swine manure as a co-substrate for biohydrogen production by the thermophilic
fermentation of easily degradable and carbohydrate-rich materials,
such as fruit and vegetable market waste. In particular, the study
aimed to maximize biohydrogen production rates while obtaining
process stability through the indigenous buffer capacity of manure,
and avoiding the use of any external alkali source. 2. Methods Experiments were conducted in semi-continuously operated reactors, without pH adjustment of the fermentative broth. A cen-
tral composite experimental matrix was designed to evaluate the
effect and interactions of the two co-substrates'' mixing ratio and
hydraulic retention time, and to determine the conditions allowing
for the highest biohydrogen production rates and self-stability at
thermophilic temperature. 2.1. Startup inoculum The seed microflora used in this study was collected from a 10 L laboratory-scale reactor, producing hydrogen by digesting glucose
and fruit wastes. The reactor had been continuously operating un-
der thermophilic conditions (55 °C) for months, prior to the begin-
ning of this study, showing a stable production of biohydrogen at
an average rate of 0.9 LH2 Lÿ 1 ferm d ÿ1, and with a H 2 concentration in the biogas of 45 ± 5% (v/v). No methane was detected during
the operation. Before being used in the experiments, the startup
inoculum was transferred into a stirred flask and kept in anaerobic,
thermophilic conditions for 3''4 d without adding any feeding sub-
strate. During this time, the gas production rate sharply decreased
to insignificant levels (less than 0.05 Lbiogas Lÿ 1 dÿ1) and the pre- pared inoculum was considered ready to seed the experimental
reactors. The total solids (TS) and volatile solid (VS) concentrations
and the pH of the inoculum resulted in 25.1 ± 4.3 g kgÿ 1, 18.9 ± 3.6 g kgÿ 1 and 5.45 ± 0.15, respectively. 2.2. Substrates The feeding substrate was a mixture of swine manure (SM) and fruit and vegetable market waste (FVMW). SM was collected from
four different pig farms near Milano (Italy) and filtered through a
stainless steel sieve (US Mesh No. 10, sieve opening of 2.0 mm).
FVMW consisting of raw fruits and vegetables residues (apples,
pears, potatoes, zucchini, etc.) was collected at different dates from
a municipal market in Milano (Italy), freshly shredded in a blender
and immediately frozen at ÿ20 °C to avoid feedstock acidification. TS and VS of SM used for the experiments were 10.0 ± 1.0 g kgÿ 1 and 8.1 ± 0.5 g kgÿ 1, respectively, while FVMW had a TS content of 133.0 ± 8.0 g kgÿ 1 and a VS content of 99.8 ± 4.0 g kgÿ1. SM and FVMW showed pH of 8.1 ± 0.20 and 4.60 ± 0.10, respectively, and to-
tal alkalinity (TA) of 10.7 ± 0.2 gCaCO3 kgÿ 1 and 3.5 ± 0.1 g CaCO3 kg ÿ1, respectively. Before feeding the reactors, shredded FVMW was acclimated to room temperature and mixed with filtered SM, according to mixing
ratios defined by the experimental design. No other ingredient was
added to the FVMW + SM mixtures, which were then used as the
feeding substrate. 2.3. Experimental design A Box''Wilson central composite design (CCD) (NIST/SEMA- TECH, 2010) was applied to study the effect of two operating
parameters (the controllable factors) on biohydrogen production
and process stability (the experimental responses), and therefore
to find the optimal region in which to operate the fermentation.
The two operating parameters considered in the study are: (a)
the ratio of FVMW/SM of the co-substrates composing the feeding
material; and (b) the hydraulic retention time (HRT) of the
fermentation. In a CCD, the experimental values of each controllable factor are defined to be uniformly distributed around a centerpoint, accord-
ing to factorial design levels coded from ÿ1 to +1. These levels A. Tenca et al. / Bioresource Technology 102 (2011) 8582''8588 8583 are then augmented with star points that, in a two-factor CCD, are
axially placed at a coded distance of ÿ p 2 and + p 2 from the center of the design. As a result, the mixing ratio of co-substrates and HRT were investigated at five levels, coded as (ÿ p 2, ÿ1, 0, +1, + p 2). The level code reflects the step change in the actual value chosen for the two
operating parameters. All the evaluated levels were arranged in nine different treat- ments, corresponding to nine combinations of mixing ratios with
HRT values. Each treatment consisted of three replicated assays,
except for the centerpoint treatment, which was replicated six
times. Two sets of experiments were designed applying the described methodology: a preliminary set A aimed toward an exploratory
screening of the experimental domain, and a subsequent, more
focused set B, aimed toward deeper insight into the process. In the preliminary set A, selected ranges for factors were 10''70% for FVMW content in the mixture with SM, and 1.5''4.5 d for HRT,
with a design centerpoint of [30%; 3 d]. The resulting investigated
range for organic loading rate is from 6 gVS Lÿ 1 dÿ1 to 32 g VS L ÿ1 dÿ1. According to results of experiments conducted on preliminary set A, more focused ranges for factors were chosen when defining
set B. In this second set, FVMW content in the mixture was inves-
tigated in the range of 15''55%, while HRT varied within 1''3 d, cen-
terpoint of the design being [35%; 2 d]. The corresponding range for
the organic loading rate is approximately from 13 gVS Lÿ 1 dÿ1 to 53 gVS Lÿ 1 dÿ1. All the coded levels and corresponding real values of operating variables considered in the experimental design are summarized in
Table 1. 2.4. Semi-continuous fermentation and operating procedure The fermentation assays were carried out using 500 mL Whea- ton batch serum bottles, with an operating volume of 300 mL. The
reactors were run on a semi-continuous basis, with broth with-
drawal and feeding addiction operated at least every 12 h. The bottles were firstly filled with the prepared startup inocu- lum. Even if known to influence the duration of the lag phase in
batch biohydrogen production, the initial pH of the inoculum was not adjusted with chemical ingredients, as its value (5.45 ±
0.15) was considered adequate to guarantee optimal conditions
for biohydrogen fermentative production. Similarly, for the entire
duration of the assays, the pH of the broth was not adjusted, allow-
ing it to reach indigenous chemical equilibrium. According the design, an appropriate amount of feeding sub- strate was added up to obtain the operating volume. The bottles
were flushed with nitrogen gas for 60 s to establish anaerobic con-
ditions. They were sealed using screw caps with rubber septa
tightly fitted to sample gas bags (SKC, Flexfoil) to collect the biogas
produced by each reactor. The bottles were placed in a Dubnoff
shaker bath at 55 ± 1 °C operated at a reciprocating frequency of
100 cycles per minute. The reactors were fed by supplying the substrate every 12 h, ex- cept for assays with HRT shorter of 2 d which were fed every 8 h.
Feeding of reactors was accompanied by the concomitant with-
drawal of an equal amount of digested effluent. The effluent was
periodically sampled for chemical analysis aimed to characterize
the fermentation broth and process operation. After startup, each reactor was allowed to reach steady condi- tions, and since then, it was operated during a period of four HRTs
(preliminary set A) or six HRTs (set B). After each run, every reactor was cleaned and freshly re-inocu- lated for new assays. 2.5. Measurements and analytical methods Feeding mixtures and reactors'' effluents were characterized according to Standard Methods (APHA, 1998) in terms of TS and
VS, chemical oxygen demand (COD), pH, total volatile fatty acids
(TVFA) and total alkalinity (TA) content as shown in Table 2. Biohydrogen production was calculated from volume measure- ments of gas accumulated in sample bags and by measuring its
hydrogen content. Biogas composition was determined with a
gas chromatograph (Agilent, Micro GC 3000A) equipped with two
thermal conductivity detectors and two different columns. Hydro-
gen and methane were analyzed using a Molesieve/5A Plot column,
with nitrogen as the carrier gas at a flow rate of 30 mL/min. The
carbon dioxide content was analyzed using a different column
(Alltech HP-Plot U), with helium as the carrier gas at a flow rate Table 1
The experimental design used for the study. Treatment Coded levels for factors Actual values of factors TS (g kgÿ 1) SV (g kgÿ 1) OLR (gVS Lÿ 1 dÿ1) Substrates ratio HRT FVMW(% w/w) SM (% w/w) HRT (d) Preliminary set A
1A ÿ p 2 0 10 90 3 22.3 a 17.3 5.8 2A ÿ1 ÿ1 20 80 2 34.6 26.4 13.2 3A ÿ1 +1 20 80 4 34.6 26.4 6.6 4A 0 ÿ p 2 40 60 1.5 59.2 44.8 29.9 5A ÿ1 0 40 60 3 59.2 44.8 14.9 6A ÿ1 + p 2 40 60 4.5 59.2 44.8 10.0 7A +1 ÿ1 60 40 2 83.8 63.1 31.6 8A +1 +1 60 40 4 83.8 63.1 15.8 9A + p 2 0 70 30 3 96.1 72.3 24.1 Set B
1B ÿ p 2 0 15 85 2 30.0 (1.5)b 26.0 (1.0) 13.0 (0.5) 2B ÿ1 ÿ1 22.5 77.5 1.25 41.0 (2.0) 36.0 (1.5) 28.8 (1.2) 3B ÿ1 +1 22.5 77.5 2.75 41.0 (2.0) 36.0 (1.5) 13.1 (0.5) 4B 0 ÿ p 2 35 65 1 58.0 (4.0) 52.0 (3.0) 52.0 (3.0) 5B 0 0 35 65 2 58.0 (4.0) 52.0 (3.0) 26.0 (1.5) 6B 0 + p 2 35 65 3 58.0 (4.0) 52.0 (3.0) 17.3 (1.0) 7B +1 ÿ1 47.5 52.5 1.25 73.5 (5.5) 66.5 (5.0) 53.2 (4.0) 8B +1 +1 47.5 52.5 2.75 73.5 (5.5) 66.5 (5.0) 24.2 (1.8) 9B + p 2 0 55 45 2 83.0 (7.0) 76 (5.0) 38.0 (2.5) a Calculated from average composition of the two co-substrates.
b Data measured from three replicates. standard deviations in brackets. 8584 A. Tenca et al. / Bioresource Technology 102 (2011) 8582''8588 of 30 mL/min. The operational temperature of the injection port
was 100 °C, while those of Molesieve/5A and Plot U columns were
maintained at 100 and 55 °C, respectively. Biohydrogen production rate (P) was measured daily for each reactor, and for clarity, the values were normalized to the fermen-
tation broth volume and then expressed as LH2 Lÿ 1 dÿ1. Biohydrogen yield was calculated as the specific production per VS mass added in each treatment and then expressed as
mLH2 gÿ1 VS added. An index for describing the production stability of a reactor was defined by considering the ratio of the standard devi-
ation and mean of daily biohydrogen production and expressed as: Stability ¼ 1 ÿ sdðP' P ð1' According to Eq. (1), a reactor with constant hydrogen produc- tion has stability = 1, while production deviations as large as the
average value are represented with stability = 0. Reactors produc-
ing a methane content in biogas higher than hydrogen were classi-
fied with a stability = 0. In this study the terminal values measured during the last two HRTs of the specific treatment considered were assumed to be rep-
resentative of the operative conditions investigated, i.e., the pre-
sented results of average and deviation for biohydrogen production rate, yield and stability of each assay were calculated
considering the data measured during the last two experimented
HRTs. 3. Results and discussion 3.1. Biohydrogen production rate and yield: effects of FVMW/SM ratio
and HRT The overall results obtained with all the experimental condi- tions considered in the study are summarized in Table 2. As a general result, independent from the adopted HRT, low or no biohydrogen production was obtained in all the assays fed with
a substrate having a FVMW content of 20% or less (i.e., SM P 80%).
In these cases (Treatments 1A''3A and 1B''3B), the average produc- tion rate of hydrogen was indeed always lower than 0.3 LH2 Lÿ 1 dÿ1, with specific yields ranging from 0.5 mL H2 g ÿ1 VS added to 40 mLH2 gÿ1 VS added. Moreover, all these assays produced an increasing amount of methane, with terminal production rates
even exceeding 1 LCH4 Lÿ 1 dÿ1, indicating that a poor content of carbohydrate-rich material mixed with swine manure associated
with a corresponding neutral pH of the feeding co-substrate, estab-
lished fermentative conditions that are gradually far away from
those required for optimal hydrogen production, but they are suit-
able for methanogenic bacteria development. On the other hand, treatment 9A with 70% FVMW mixed with SM quickly resulted in large production instabilities, accompanied
by drops of pH as low as 4.2 ± 0.2 (data not shown). In all the repli-
cates of this treatment, the fermentation process resulted in a com-
plete blockage before entering the two terminal HRTs, indicating an
organic overload in the feeding. Low production also resulted in
operating conditions corresponding to the longest tested HRT, that
is, 4.5 d (treatment 6A), with less than 0.2 LH2 Lÿ 1 dÿ1 for all the rep- licates and an average yield of 15 mLH2 gÿ1 VS added. In Fig. 1 the contour line charts of the experimental results for biohydrogen production rates of preliminary set A and set B,
respectively, are plotted. With reference to preliminary set A, signif-
icant biohydrogen production rates were obtained in the experi-
mental domain included in the HRT range 1.5''3 d and FVMW
content in the co-substrate mixture in the range of 25''60%. Treat-
ment 4A corresponded to conditions allowing the highest produc-
tion rate of 2.82 ± 0.67 LH2 Lÿ 1 dÿ1, reaching a peak of 3.21 ± 0.27 LH2 Lÿ 1 dÿ1 for the best-producing replicate. This treat- ment, with a substrate composed of 40% FVMW and 60% SM, was
associated with a fast HRT of 1.5 d, and it also produced the highest
hydrogen content in biogas (51.9%), with an average yield of
95 mLH2 gÿ1 VS added. Again, the same co-substrate but fermented with a longer HRT of 3 d resulted in the highest yield of 149 mLH2 gÿ1 VS added, indicating the optimal balance in the mixture between carbohydrates from FVMW and the alkali/nutrients supply
from SM at this mixing ratio. As the highest producing condition [FVMW = 40%, HRT = 1.5 d] identified within preliminary set A resulted located on the border
of the experimental domain, a second set B was designed with Table 2
Main experimental results for biohydrogen production and process stability. Treatment Actual values of factors Biohydrogen production Methane production FVMW
(% w/w) SM
(% w/w) HRT
(d) Production rate
(LH2 Lÿ 1 dÿ1) Yield
(mLH2 gÿ1 VS added ) H2 content
(%) Production stability
index a Production rate
(LCH4 Lÿ 1 dÿ1) Preliminary set A
1A 10 90 3 0.13 (0.02) 22.8 (3.9) 0.5 (0.5) 0 b 0.17 (0.03) 2A 20 80 2 0.24 (0.08) 18.4 (5.7) 13.2 (2.7) 0 0.65 (0.16) 3A 20 80 4 0.26 (0.04) 39.8 (5.9) 20.9 (1.3) 0 0.05 (0.04) 4A 40 60 1.5 2.82 (0.67) 94.5 (22.4) 51.9 (2.9) 0.61 (0.33) 0 5A 40 60 3 2.23 (0.07) 149.5 (6.4) 33.5 (1.4) 0.71 (0.03) 0 6A 40 60 4.5 0.14 (0.09) 14.5 (8.2) 14.5 (2.3) 0.25 (0.25) 0 7A 60 40 2 1.36 (0.50) 43.0 (15.9) 31.5 (5.9) 0.91 (0.04) 0 8A 60 40 4 0.35 (0.35) 22.0 (22.0) 5.1 (0.0) 0.22 (0.02) 0 9A 70 30 3 0.00 (0.00) 0.0 (0.0) 0.0 0.26 (0.07) 0 Set B
1B 15 85 2 0.08 (0.05) 6.4 (3.6) 2.1 (1.5) 0 0.55 (0.05) 2B 22.5 77.5 1.25 0.02 (0.01) 0.5 (0.5) 0.1 (0.1) 0 1.20 (0.49) 3B 22.5 77.5 2.75 0.28 (0.14) 21.6 (10.4) 13 (5.0) 0 0.25 (0.05) 4B 35 65 1 1.38 (0.22) 26.6 (4.3) 17.7 (3.2) 0.55 (0.24) 0 5B 35 65 2 3.27 (0.51) 125.8 (22.4) 42.1 (5.1) 0.86 (0.04) 0 6B 35 65 3 2.06 (0.59) 119.3 (34.1) 37.6 (4.8) 0.77 (0.05) 0 7B 47.5 52.5 1.25 1.74 (0.82) 32.7 (15.5) 23.0 (8.2) 0.27 (0.04) 0 8B 47.5 52.5 2.75 2.39 (0.69) 98.9 (28.4) 41.3 (4.2) 0.67 (0.11) 0 9B 55 45 2 1.23 (0.63) 32.4 (16.6) 25.8 (8.0) 0.48 (0.41) 0 a See Eq. (1) for definition.
b Assays producing methane were assigned with a stability = 0. A. Tenca et al. / Bioresource Technology 102 (2011) 8582''8588 8585 the aim of investigating deeper insight into the neighborhood of
that condition, as well as extending the study toward new condi-
tions where possible high production could be found. Overall biohydrogen production obtained with set B fairly over- lapped with the results of set A (Fig. 1C). Nevertheless, the second
design allowed for the identification of optimal conditions with a
higher production rate than those found with set A. Treatment
5B (Fig. 2), with a FVMW content of 35% in co-substrate and a
HRT of 2 d, resulted in 3.27 ± 0.51 LH2 Lÿ 1 dÿ1, with a peak of 3.76 ± 0.51 LH2 Lÿ 1 dÿ1 for the best-producing replicate and an average hydrogen content of 42 ± 5% in the biogas. The overall highest yields in the study (120''150 mLH2 gÿ1 VS added) were obtained with treatments 5A, 5B and 6B, i.e., in a close range
of conditions corresponding to FVMW content of 35''40% and HRT
of 2''3 d. 3.2. Response surface analysis of biohydrogen production The results of biohydrogen production rate (P) were evaluated with a response surface analysis, a technique that determines
whether the experimental response exhibits a significant curvature
in its pattern and is therefore likely to have stationary points,
where the optimal response is expected to occur. To this aim, a second order model was introduced: P ¼ b0 þ b1  FVMW þ b2  HRT þ b3  FVMW  HRT þ b4  FVMW 2 þ b 5  HRT 2 ð2' as a simplified surface function to locally approximate the pro- duction rate P obtained with a combination of the two independent
variables (controllable factors), mixing ratio and HRT. Coefficients bi have to be determined by regression of experi- mental data, and once the adequacy of the fitting has been evalu-
ated, the second order function (2) can be used to find the
stationary points, representing a maximal (or minimal) response. Statistical calculations have been conducted on centered vari- ables fvwm' and hrt', instead of using natural variables FVMW
and HRT. Centered values were calculated by subtracting the mean
value of the corresponding natural variable and normalizing to its
step change (standard deviation). The results are expressed here in
terms of natural variables, obtained after back conversion. The analysis of variance for the fitting to Eq. (2) of experimental data of biohydrogen production rates, obtained with replicates of
the two experimental sets, indicates that for both preliminary set
A and set B, there is a significant curvature in the response surface.
Indeed, all the quadratic coefficients (b4 and b5) for the two regres-
sions have p-values < 0.001, while interaction between mixing ra-
tio and retention time (b3) does not have a significant effect on
production with p-values much larger than 0.1. In particular, for set B, the multiple correlations between all predicted and experimental production data resulted in a determi-
nation coefficient of R 2 = 0.82, with a RMSE = 0.50 L H2 L ÿ1 dÿ1. For this set, the best quadratic fit obtained after rejecting the non-
significant interaction term (i.e., by setting b3 = 0) was identified
by the equation: P ¼ ÿ14 :13 þ 0 :52  FVMW þ 7 :01  HRT ÿ 0 :01  FVMW 2 ÿ 1 :67  HRT 2 ð3' In Fig. 3, the experimental values of set B are plotted with the fitting function (3). By calculating the zeros of the derivatives,
the location of the local maximum of the quadratic model can be
found in correspondence to a FVMW content of 38.5% in the feed-
ing mixture and a HRT = 2.1 d. Regression of Eq. (2) was also studied when considering all the experimental data obtained with both preliminary set A and set B
together. In this case, the multiples correlation between predicted
and experimental production rates resulted in a lower determina-
tion coefficient of R 2 = 0.69, with a RMSE = 0.69 L H2 L ÿ1 dÿ1. The AN- OVA for the fitting to Eq. (2) indicates that the linear dependence
from HRT (b2), as well as from interaction between mixing ratio
and retention time (b3) do not have a significant impact on produc-
tion (p-values much larger than 0.05). Nevertheless, the quadratic
model showed a lack of fit with the experimental data especially
around the optimal range, with inacceptable underestimations of
production rate larger than 1 LH2 Lÿ 1 dÿ1, which excluded any use of the model to retrieve qualitative information about the process. Fig. 1. Experimented conditions (s) and biohydrogen production rates contours obtained with data of preliminary set A (left), set B (middle) and overlapping set A + set B
(right). Production data were normalized to the fermentation broth volume and expressed as LH2 Lÿ 1 dÿ1. 1 2 3 4 5 6 7 0 2 4 6 8 10 12 14 Time (d) H 2 p ro d u c ti o n r a te ( L H 2 L -1 d -1 ) 3.0 3.5 4.0 4.5 5.0 5.5 6.0 p H Fig. 2. Biohydrogen production rate (s) and pH of digested effluent (h) of the
highest producing treatment (5B: FVMW = 35%, HRT = 2 d), mean and standard
deviations for six replicates. 8586 A. Tenca et al. / Bioresource Technology 102 (2011) 8582''8588 3.3. Biohydrogen production stability: effects of FVMW/SM ratio and
HRT The overall biohydrogen production stability summarized in Ta- ble 2 and illustrated in Fig. 4A and B. As they produced significant
amounts of methane, treatments 1A''3A and 1B''3B (FVMW < 23%,
regardless of HRT) were assigned with a stability = 0, even though
some of them (2A and 3A) showed limited variability in their (low)
hydrogen production, and (1A) showed a nearly constant lack of
hydrogen production. In the range of the experimental domain where significant bio- hydrogen production levels were obtained, the stability pattern
overlapped with the production pattern. Indeed, the process exhib-
ited a stability index greater than 0.7 for operating conditions cor-
responding to a substrate mixture containing 35''40% of FVMW
and a HRT range of 1.5''3 d. Among highest production stability
was obtained with the replicates of treatment 5B [FVMW = 35%,
HRT = 2 d], with an index of 0.86 ± 0.04. A second peak was ob-
tained with treatment 7A [FVMW = 60%, HRT = 2 d] which ex-
ceeded in stability the optimal range with an index of
0.91 ± 0.04. The replicates run under these conditions resulted then
in the production of a stable and moderate (1.36 LH2 Lÿ 1 dÿ1 on average) biohydrogen volume. 3.4. Chemical parameters and process efficiency Chemical characterization of set B took in consideration both the substrates and the digested effluents of the co-fermentation;
Table 3 summarizes the obtained results. Substrate degradation
accompanied the hydrogen production evidencing that the sub-
strate had participated as a primary carbon source in the metabolic
reactions involved in hydrogen generation. Effluent''s COD ranged
between 29 gCOD kgÿ 1 and 85 g COD kg ÿ1 with a COD removal effi- ciency varying between 17% and 40%. The most productive exper-
imental condition 5B showed an average efficiency in COD
reduction of 32%, close to the highest value achieved by treatments
2B and 4B, both characterized by short HRT (1''1.25 d). COD con-
sumption was relatively high also in treatments with a low H2 pro-
duction, but with a considerable CH4 and/or CO2 productions (1B,
2B, 4B). Fig. 3. Measured biohydrogen production rates (s) and the best fitting quadratic
surface for experimental set B. Fig. 4. Experimented conditions (s) and production stability contours obtained with data of preliminary set B (left), and overlapping set A + set B (right). Assays with a
constant production of biohydrogen have stability = 1, while production deviations as large as the average value are represented with stability = 0. Reactors producing
methane are classified with a stability = 0. Table 3
Chemical characterization of substrates (IN) and digested effluents (OUT). Trial pH_IN feeding
substrate pH_OUT
effluent CODIN
(gCOD kgÿ 1) CODOUT
(gCOD kgÿ 1) COD
(%) TVFAIN (gAcetic Acid kg ÿ1) TAIN
(gCaCO3 kgÿ 1) TVFA/
(gAcetic Acid kg ÿ1) TAOUT
(gCaCO3 kgÿ 1) TVFA/
TAOUT 1B 7.71 (0.10) 5.64 (0.07) 41.3 (4.1) 28.9 (2.7) 30 1.76 (0.04) 9.21 (0.11) 0.19 11.2 (0.5) 8.3 (0.4) 1.35 2B 7.60 (0.15) 5.63 (0.32) 55.3 (3.7) 33.2 (6.2) 40 1.91 (0.07) 8.42 (0.98) 0.23 10.5 (410) 9.1 (0.3) 1.15 3B 7.38 (0.12) 5.40 (0.12) 55.3 (3.7) 45.9 (7.3) 17 1.91 (0.07) 8.42 (0.98) 0.23 11.4 (0.5) 8.4 (0.3) 1.36 4B 7.48 (0.14) 5.07 (0.23) 78.7 (5.2) 48.0 (4.4) 39 2.15 (0.07) 7.32 (0.12) 0.29 8.5 (1.1) 6.9 (0.7) 1.23 5B 7.35 (0.09) 4.88 (0.05) 78.7 (5.2) 53.5 (6.3) 32 2.15 (0.07) 7.32 (0.12) 0.29 12.2 (1.5) 5.9 (1.4) 2.07 6B 7.15 (0.13) 4.74 (0.05) 78.7 (5.2) 60.6 (5.8) 23 2.15 (0.07) 7.32 (0.12) 0.29 12.9 (0.8) 8.2 (1.3) 1.57 7B 7.21 (0.15) 4.43 (0.30) 95.9 (9.6) 70.0 (6.0) 27 2.41 (0.11) 6.85 (0.09) 0.35 8.3 (1.0) 7.1 (1.3) 1.17 8B 7.04 (0.09) 4.34 (0.12) 95.9 (9.6) 76.8 (4.9) 20 2.41 (0.11) 6.85 (0.09) 0.35 11.8 (2.4) 7.2 (0.3) 1.64 9B 7.00 (0.05) 4.57 (0.03) 110.2 (14.2) 84.9 (10.6) 23 2.73 (0.04) 5.66 (0.07) 0.48 11.2 (1.3) 6.5 (2.1) 1.72 A. Tenca et al. / Bioresource Technology 102 (2011) 8582''8588 8587 Effluents''s pH (pHOUT in Table 3) were found in the range 4.2''5.7 and fairly correlated to SM content in the substrate (R 2 = 0.87, p < 0.05). In particular, treatments 1B''3B showed the highest pH
values (5.4''5.7) commonly assumed as sufficiently low for inhibit-
ing methanogenic bacteria (Liu et al., 2008; Hwang et al., 2009), but
in this study that pH range resulted associated to a moderate meth-
anogenesis and a very poor H2 production. Moreover, all the hydro-
gen-producing conditions (4B''9B) showed pH levels below 5.1 and,
in particular, the highest producing treatment (5B) had a pHOUT of
4.88 ± 0.05 (Table 3). Even if such a pH value is relatively lower than
the optimum typically indicated by literature (Van Ginkel et al.,
2001; Liu et al., 2008; Hwang et al., 2009), the process resulted in
relevant hydrogen production rates and yields, remarkably without
using any external alkali for pH control. Effluents'' pH values were also fairly related to the endogenous alkalinity (TAIN) and the TVFA/TAIN ratio of substrates, indicating
that organic acids production was successfully equilibrated by
the alkaline species initially contained in SM. On the other hand,
feeding substrate''s TAIN as low as 7 gCaCO3 kgÿ 1 resulted in efflu- ents'' pHOUT lower than 4.6, likely contributing to reduce the biohy-
drogen fermentation efficiency. The TVFA concentration in digested effluents ranged between 8.3 gAcetic Acid kgÿ 1 and 12.9 g Acetic Acid kg ÿ1, a range which overlaps the values found in similar conditions by Hwang et al. (2010), after
18 h of ripened fruit fermentation. Lower TVFAOUT concentrations
were found when shorter HRTs were applied (2B, 4B and 7B), likely
due to more intense metabolites wash-out. As expected, the most
producing treatments (5B and 6B) showed the highest TVFAOUT
concentrations, with a TVFA/TAOUT ratio even exceeding 2. For
these operating conditions the chemical endogenous equilibrium
between the initial TA and the TVFA produced during the fermen-
tation was able to maintain stable and remarkable hydrogen
production. 4. Conclusions The experimental design identified as optimal operating param- eters a mixing ratio for substrate composition of 65% SM and 35%
FVMW and a HRT of 2 d. At these conditions, the natural buffer
capacity of SM was able to avoid pH drops and to maintain an opti-
mal environment for high (3.27 LH2 Lÿ 1 dÿ1, 126 mL H2 g ÿ1 VS added) and stable (deviations from daily average less than 14%) biohydrogen
production, showing the feasibility of fermenting carbohydrate-
rich substrates while avoiding the need for external alkali in possi-
ble implementations at pig farms. Acknowledgements The authors are grateful to Regione Lombardia, General Direc- torate of Agriculture, for its financial support to project AGRIDEN. References Aceves-Lara, C., Latrille, E., Bernet, N., Buffiere, P., Steyer, J., 2008. A pseudo- stoichiometric dynamic model of anaerobic hydrogen production from
molasses. Water Resour. 42, 2539''2550. American Public Health Association, 1998. Standard Methods for the Examination of Water and Wastewater, 20th ed., Washington DC, USA. Dabrock, B., Bahl, H., Gottschalk, G., 1992. Parameters affecting solvent production by Clostridium pasteurium. Appl. Environ. Microbiol. 58, 1233''1239. Dong, L., Zhenhong, Y., Yongming, S., Xiaoying, K., Yu, Z., 2009. Hydrogen production characteristics of the organic fraction of municipal solid wastes by anaerobic
mixed culture fermentation. Int. J. Hydrogen Energy 34, 812''820. Edwards, P.P., Kuznetsov, V.L., David, W.I., 2007. Hydrogen energy. Philos. Trans. A Math. Phys. Eng. Sci. 365, 1043''1056. Fang, H.H.P., Liu, H., 2002. Effect of pH on hydrogen production from glucose by a mixed culture. Bioresour. Technol. 82, 87''93. Gomes Antunes, J.M., Mikalsen, R., Roskilly, A.P., 2008. An investigation of hydrogen-fuelled HCCI engine performance and operation. Int. J. Hydrogen
Energy 33, 5823''5828. Hawkes, F.R., Dinsdale, R., Hawkes, D.L., Hussy, I., 2002. Sustainable fermentative biohydrogen: challenges for process optimization. Int. J. Hydrogen Energy 27,
1339''1347. Hwang, J.H., Choi, J.A., Abou-Shanab, R.A.I., Bhatnagar, A., Min, B., Song, H., Kumar, E., Choi, J., Lee, E.S., Kim, Y.J., Um, S., Lee, D.S., Jeon, B.H., 2009. Effect of pH and
sulfate concentration on hydrogen production using anaerobic mixed
microflora. Int. J. Hydrogen Energy 34, 9702''9710. Hwang, J.H., Choi, J.A., Abou-Shanab, R.A.I., Min, B., Song, H., Kim, Y., Lee, E.S., Jeon, B.H., 2010. Feasibility of hydrogen production from ripened fruits by a
combined two-stage (dark/dark) fermentation system. Bioresour. Technol.
102, 1051''1058. Kotsopoulos, T.A., Fotidis, I.A., Tsolakis, N., Martzopoulos, G.G., 2009. Biohydrogen production from pig slurry in a CSTR reactor system with mixed cultures under
hyperthermophilic temperature. Biomass Bioenergy 33, 1168''1174. Lay, J.J., 2000. Modeling and optimization of anaerobic digested sludge converting starch to hydrogen. Biotechnol. Bioeng. 68, 269''278. Lee, Y.J., Miyahara, T., Noike, T., 2002. Effect of pH on microbial hydrogen fermentation. J. Chem. Technol. Biotechnol. 77, 694''698. Li, M., Zhao, Y., Guo, Q., Qian, X., Niu, D., 2008. Bio-hydrogen production from food waste and sewage sludge in the presence of aged refuse excavated from refuse
landfill. Renew. Energy 33, 2573''2579. Lin, C.Y., Cheng, C.H., 2006. Fermentative hydrogen production from xylose using anaerobic mixed microflora. Int. J. Hydrogen Energy 31, 832''840. Liu, D., Min, B., Angelidaki, I., 2008. Biohydrogen production from household solid waste (HSW) at extreme-thermophilic temperature (70 °C) '' Influence of pH
and acetate concentration. Int. J. Hydrogen Energy 33, 6985''6992. Mu, Y., Wang, G., Yu, H.Q., 2006. Kinetic modeling of batch hydrogen production process by mixed anaerobic cultures. Bioresour. Technol. 97, 1302''1307. NIST/SEMATECH e-Handbook of Statistical Methods. div898/handbook/, accessed 20 February 2010. Okamoto, M., Noike, T., Miyahara, T., Mizuno, O., 2000. Biological hydrogen potential of materials characteristic of the organic fraction of municipal solid
wastes. Water Sci. Technol. 41, 25''32. Shin, H.S., Youn, J.H., 2005. Conversion of food waste into hydrogen by thermophilic acidogenesis. Biodegradation 16, 33''44. Van Ginkel, S., Logan, B.E., 2005. Inhibition of biohydrogen production by undissociated acetic and butyric acids. Environ. Sci. Technol. 39, 9351''9356. Van Ginkel, S., Sung, S., Lay, J.J., 2001. Biohydrogen production as a function of pH and substrate concentration. Environ. Sci. Technol. 35, 4726''4730. Venetsaneas, N., Antonopoulou, G., Stamatelatou, K., Kornaros, M., Lyberatos, G., 2009. Using cheese whey for hydrogen and methane generation in a two-stage
continuous process with alternative pH controlling approaches. Bioresour.
Technol. 100, 3713''3717. Venkata Mohan, S., Babu, L.M., Reddy, V.M., Mohanakrishna, G., Sarma, P.N., 2009a. Harnessing of biohydrogen by acidogenic fermentation of Citrus limetta
peelings: effect of extraction procedure and pretreatment of biocatalyst. Int. J.
Hydrogen Energy 34, 6149''6156. Venkata Mohan, S., Mohanakrishna, G., Kannaiah Goud, R., Sarma, P.N., 2009b. Acidogenic fermentation of vegetable based market waste to harness
biohydrogen with simultaneous stabilization. Bioresour. Technol. 100, 3061''
3068. Vijayaraghavan, K., Ahmad, D., Bin Ibrahim, M.K., 2006. Biohydrogen generation from jackfruit peel using anaerobic contact filter. Int. J. Hydrogen Energy 31,
569''579. Wagner, R.C., Regan, J.M., Oh, S.E., Zuo, Y., Logan, B.E., 2009. Hydrogen and methane production from swine wastewater using microbial electrolysis cells. Water
Resour. 43, 1480''1488. Wu, X., Yao, W., Zhu, J., 2010. Effect of pH on continuous biohydrogen production from liquid swine manure with glucose supplement using an anaerobic
sequencing batch reactor. Int. J. Hydrogen Energy 35, 6592''6599. Wu, X., Zhu, J., Dong, C., Miller, C., Li, Y., Wang, L., Yao, W., 2009. Continuous biohydrogen production from liquid swine manure supplemented with glucose
using an anaerobic sequencing batch reactor. Int. J. Hydrogen Energy 34, 6636''
6645. Ye, N.F., Lu, F., Shao, L.M., Godon, J.J., He, P.J., 2007. Bacterial community dynamics and product distribution during pH-adjusted fermentation of vegetable wastes.
J. Appl. Microbiol. 103, 1055''1065. Yokoi, H., Maki, R., Hirose, J., Hayashi, S., 2002. Microbial production of hydrogen from starch-manufacturing wastes. Biomass Bioenergy 22, 389''395. Zheng, X.J., Yu, H.Q., 2005. Inhibitory effects of butyrate on biological hydrogen production with mixed anaerobic cultures. J. Environ. Manage. 74, 65''70. Zhu, J., Li, Y., Wu, X., Miller, C., Chen, P., Ruan, R., 2009. Swine manure fermentation for hydrogen production. Bioresour. Technol. 100, 5472''5477. 8588 A. Tenca et al. / Bioresource Technology 102 (2011) 8582''8588 _ II_ Looking for practical tools to achieve next-future applicability of dark
fermentation to produce bio-hydrogen from organic materials
in Continuously Stirred Tank Reactors A. Tenca a, A. Schievano b, S. Lonati b, L. Malagutti c, R. Oberti a, F. Adani b,'' a Dipartimento di Ingegneria Agraria, Università degli Studi di Milano, Via Celoria 2, 20133 Milano, Italy
b Dipartimento di Produzione Vegetale, Università degli Studi di Milano, Via Celoria 2, 20133 Milano, Italy
c Dipartimento di Scienze Animali, Università degli Studi di Milano, Via Celoria 2, 20133 Milano, Italy a r t i c l e i n f o Article history:
Received 22 March 2011
Received in revised form 26 May 2011
Accepted 30 May 2011
Available online 24 June 2011 Keywords:
Dark fermentation
Organic waste a b s t r a c t This study aimed at finding applicable tools for favouring dark fermentation application in full-scale bio-
gas plants in the next future. Firstly, the focus was obtaining mixed microbial cultures from natural
sources (soil-inocula and anaerobically digested materials), able to efficiently produce bio-hydrogen by
dark fermentation. Batch reactors with proper substrate (1 g L ÿ 1 glucose) and metabolites concentrations, allowed high H2 yields (2.8 ± 0.66 mol H2 mol ÿ 1 glucose ), comparable to pure microbial cultures achieve- ments. The application of this methodology to four organic substrates, of possible interest for full-scale
plants, showed promising and repeatable bio-H2 potential (BHP = 202 ± 3 NLH2 kg ÿ 1 VS ) from organic frac- tion of municipal source-separated waste (OFMSW). Nevertheless, the fermentation in a lab-scale CSTR
(nowadays the most diffused typology of biogas-plant) of a concentrated organic mixture of OFMSW
(126 gTS L ÿ 1) resulted in only 30% of its BHP, showing that further improvements are still needed for future full-scale applications of dark fermentation. ' 2011 Elsevier Ltd. All rights reserved. 1. Introduction Hydrogen is widely recognized as a clean and efficient energy resource with high potential for the future. Instead of common
chemical or electrochemical hydrogen production processes and
among the less energy intensive biological approaches (Oh et al.,
2003) the dark fermentation has high H2-production rate (Ueno
et al., 2001) and relies on renewable resources. Many studies dem-
onstrated the possibility of coupling hydrogen production with the
utilization of a variety of organic substrates including waste mate-
rials, such as municipal solid waste, industrial wastewaters and
agro-industrial waste and this may simultaneously provide eco-
nomic and environmental benefits, meeting the growing demand
for renewable energy (Guo et al., 2010; Lay et al., 1999; Oh et al.,
2003). In particular, a number of authors reported that the two-
stage anaerobic digestion (AD) process, as compared to traditional
single-stage AD, would drive to differentiate the biofuel production
(bio-hydrogen and bio-methane), improve the overall biogas pro-
duction yields and allow higher CH4 concentrations in the biogas
produced in the second stage, decreasing the biogas purification
costs. However, today most of full-scale biogas plants in Europe rely on single-stage process and the two-stage technology remains
unproven in the field (Liu et al., 2006; Fantozzi and Buratti, 2009).
This is mainly because dark fermentation process stability and the
maximization of bio-hydrogen production yields in the first stage
are still uncertain. Rarely maximized and repeatable yields were reported even in pilot-scale, especially when complex and varied substrates (i.e.
waste, residues, etc.) are fermented in relatively highly-concen-
trated organic mixtures (>100 gVS kg ÿ 1 wet weight) and in simple, low-cost and readily applicable reactor designs, such as the contin-
uously stirred tank reactor (CSTR). Several studies on hydrogen
production from wastewaters or solid waste have been made
under different conditions in batch or continuous bioreactors,
investigating both physical and biological parameters involved in
the process, to optimize H2-production yields and rates. Hydrogen
production performances were reported to be variable because of
fermentation conditions such as pH, hydraulic or solid retention
time, hydrogen gas partial pressure, concentration of acids, micro-
bial community of hydrogen-producing bacteria and presence of
methane-producing microorganisms (Khanal et al., 2004; Lin
et al., 2007; Nandi and Sengupta, 1998; Van Ginkel et al., 2001). More recently, important efforts were focused, also, on reactor design and operational strategies for improving the H2 yields of
continuously-fed systems, looking for future applications of this
bioprocess (Ding et al., 2010; Wang et al., 2010). 0960-8524/$ - see front matter ' 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biortech.2011.05.088 '' Corresponding author. E-mail addresses: (A. Tenca), (A. Schievano), (F. Adani). Bioresource Technology 102 (2011) 7910''7916 Contents lists available at ScienceDirect Bioresource Technology j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / b i o r t e c h However, the vast majority of literature contributions do not deal yet with technical solutions readily transferable to full-scale
systems, but they focus on introducing improved strategies relying
on selected-microbial cultures, on feeding approaches based on
low-concentrated and/or completely soluble organic mixtures, or
on innovative reactor designs. In particular, regarding the microbial community, most studies were carried out on pure cultures of isolated strains of anaerobic
bacteria, such as Clostridium sp. Because of the ubiquitous nature
of hydrogen consumers and interspecific hydrogen transfer reac-
tions (Lay et al., 1999), Clostridium sp. or other pure strains gave
relatively higher hydrogen yield than natural mixed cultures, that
makes the former particularly suitable for maximizing hydrogen
production in laboratory test. Nevertheless, since dark fermenta-
tion has been shown to have great potential as applicable process
to produce biohydrogen from a variety of organic materials (Levin
et al., 2004), the use of natural mixed microbial communities
instead of pure ones appear to be the most appropriate choice
for future real implementation of this bioprocess. Indeed, mixed
inocula do not require substrate-sterilization or other procedures
that permit to maintain pure cultures active and, additionally, they
are easily adaptable to different organic substrates (Ueno et al.,
2001). Previous authors demonstrated that H2-producing consortia
can be obtained from various environmental sources, such as soil,
compost, sewage sludge and various fermented organic materials
(Kyazze et al., 2006; Li and Fang, 2007), but few investigated the
effectiveness of such natural inocula on hydrogen productions
(Akutsu et al., 2009; Morimoto et al., 2004; Van Ginkel et al.,
2001) as compared to pure/selected cultures. Furthermore, the the-
oretical H2 yield obtainable by standard substrates (typically glu-
cose with a stoichiometric yield of 4 mol H2 mol ÿ 1 glucose) were rarely closely achieved and hardly repeatable (Fig. 1), even if
pure/selected and genetically modified microbial strains were
used. Additionally, most studies were focused on mesophilic
microbial communities, while fewer on thermophilic ones, even
if these latter were reported to show more promising H2 produc-
tion yields, near to maximal theoretical values (Schröder et al.,
1994; Van Niel et al., 2002). Besides the type of microbial community, also, the substrate/ metabolites concentration is a factor particularly worthy of atten-
tion, affecting the efficiency of both substrate utilization and
hydrogen production activity (Lin et al., 2007) by microbial com-
munities suffering of both substrate and product inhibition. In
mesophilic batch tests, hydrogen-producing bacteria growth,
hydrogen yield and hydrogen production rate were all reported to decrease with increasing added ethanol, acetic acid, propionic
acid and butyric acid concentration from 0 to 300 mmol L ÿ 1 (Wang et al., 2008). This may be a problem when relatively concentrated
organic mixtures (more than 100 gVS kg ÿ 1) are used, such as typi- cally happens in full-scale biogas plants. In addition regarding bio-
reactor typology and process design/operation, readily applicative
perspectives were rarely addressed in literature. Most contribu-
tions have dealt with interesting innovative processes (such as
up-flow anaerobic sludge bed, expanded granular sludge bed,
internal circulation reactors, etc.), which are normally used with
relatively low-concentrated (<10 gVS kg ÿ 1 wet weight) and completely soluble carbon sources (glucose, molasses, etc.). Contrarily, today
the most diffused full-scale AD process design is the continu-
ously-stirred tank reactor (CSTR) operated in ''''wet'''' or ''''semi-dry''''
conditions to produce bio-methane, i.e., one AD stage fed with
solid''liquid complex mixtures of 50''200 gVS kg ÿ 1 wet weight (Fantozzi and Buratti, 2009). In this framework, this research addressed the viability of immediate applications of dark fermentation by using existing
full-scale bio-technological solutions. Specifically, a first objective
was to define a systematic and easily applicable lab-scale proce-
dure for maximizing H2-yields obtained with mixed microbial cul-
tures, possibly approaching the results reported for pure cultures.
Secondly, to apply this procedure on different organic substrates
which may be of practical interest for H2 production at full-scale
and to assess their bio-H2 potential production (BHP) with a lab-
scale simple test. Finally, to verify to what extent the BHP can be
achieved by the currently most diffused full-scale biogas plant de-
sign, namely the CSTR. This with a view of the possibility to shift to
two-stage AD process for producing bio-hydrogen and bio-meth-
ane, by adding an appropriate module to the classical biogas plant. 2. Methods 2.1. Efficient hydrogen-producing inocula preparation from natural
sources The study was divided into two subsequent steps based on two different procedures of harvesting microbial consortia and prepar-
ing the fermentation environment. In the first step, three different soils were used as sources of seed microorganisms: a rice soil (Inoculum A), a green urban soil
(Inoculum B) and a vegetables-cultured soil (Inoculum C). These
soils were dried for 24 h at 80 °C, shredded in a blender to pass
through a mesh of about 2 mm and stored at 4 °C (APHA, 1998).
The soils were suspended in water, getting a total solid (TS) con-
tent of the slurry of 70 g kg ÿ 1 wet weight (w.w.) (Table 1), then heat shocked at 100 °C for 2 h, in order to select spore-forming
microorganisms and inhibit hydrogen consumers (Van Ginkel
and Logan, 2005) and maintained for 4 weeks at 55 °C, under strict
anaerobic conditions. The pH, total solids (TS) and the total volatile
fatty acids (TVFA) were measured in all slurries, in order to charac-
terize the fermentation environment. Batch anaerobic tests (55 °C
incubation) were carried out with glucose as standard substrate
(concentrations of 1, 3, 5 and 7 g kg ÿ 1), both to compare the effec- tiveness of Inocula A, B, C and to find the ideal substrate concentra-
tion to maximize H2 production yields (mol H2 mol ÿ 1 glucose). In the second step, a digested slurry (collected from a full-scale biogas plant treating household source-separated bio-waste mixed
with agro-industrial by-products) was used as new source of
microbial consortia. This anaerobically fermented material was
treated, firstly, to inhibit hydrogen consumers (heat-shock at
100 °C for 2 h) and, secondly, to enrich the environment of fer-
menting microbial consortia. For this second purpose, the digestate
was acclimated for a 4-weeks period in a laboratory-scale 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0 10 20 30 40 50 60 Glucose concentration (g L -1 ) Y ie ld ( m o l H 2 m ol -1 gl u cos e) Fig. 1. Effect of substrate concentrations and type of inoculum on hydrogen yield:
this work in comparison with literature results. (N) acclimated inocula (this work),
(d) soil inocula (this work), (s) literature results with naturally-sourced mixed
microbial cultures, () literature results with pure/selected wild type microbial
cultures and (+) literature results with genetically modified pure cultures (see Table
S1). A. Tenca et al. / Bioresource Technology 102 (2011) 7910''7916 7911 anaerobic bioreactor, fed with glucose solution (30 g L ÿ 1) in semi- continuous mode (twice a day), with a hydraulic retention time
(HRT) of 3 days and an organic loading rate (OLR) of 10 g glu-
cose L ÿ 1 dÿ1. This bio-reactor had a working volume of 0.6 L and was operated at a temperature of 55 ± 1 °C at a constant pH of
5.5 ± 0.5. The process showed a H2 production of 0.96 LH2 L ÿ 1 dÿ1 and H2 concentration in the biogas of 45 ± 5% (v/v). No methane
was detected during the test. The digester output was then mixed with a blend of Inocula A, B and C (1:1:1 on w.w.), in order to decrease the high concentrations
of metabolites observed in the digestate (TVFAs concentration of
2.1 ± 0.5 gACETIC ACID L ÿ 1 during the 4-weeks observation) and at the same time to enrich soil inocula. Three different mixing ratios
(1:1, 2:1, and 3:1 w/w between soils mixture and acclimated slur-
ry) gave three new inocula (Inocula D, E and F, respectively) which
were compared in batch bio-reactors, by feeding them with an
equal substrate concentration (3 g glucose L ÿ 1). The best perform- ing inoculum was then tested with 1, 5 and 7 g glucose L ÿ 1, in or- der to find the best methodology to test the bio-hydrogen
potential of an unknown organic substrate through a batch test. 2.2. Bio-H2 potential production (BHP) of selected organic substrates Four selected substrates were tested by batch experiments to assess their bio-H2 potential production (BHP): market bio-wastes
(MBW), the organic fraction of municipal source-separated wastes
(OFMSW), maize silage (MS) and swine slurry (SS). The batch tests
were performed using the best inoculum harvested from natural
sources and the same procedure developed in the previous tests. The organic materials were collected from a full scale biogas plant near Milan (Italy), dried, shredded in a blender to pass into
a 1 mm mesh and added to the bottles at the concentration of
1 g L ÿ 1. Reference tests performed with 1 g Lÿ1 glucose and blanks (no substrate added) were run as control. All batch tests were per-
formed in duplicate. 2.3. Anaerobic batch tests procedure Batch tests used to select natural-sourced inocula and to carry out the BHP assays were conducted in 500 mL glass bottles with
a working volume of 300 mL under thermophilic temperature.
For each bottle, 1.5 mL of nutrient stock solution, i.e. 166 g of
(NH4)2HPO4, 100 g of KH2PO4, 10 g of MgSO47H2O, 1 g of NaCl,
1 g of Na2MoO42H2O, 1 g of CaCl22H2O, 0.6 g of FeCl36H2O and
1.18 g of MnSO42H2O (Lay et al., 1999), was added to meet the
requirement for microbial growth. Initial pH was adjusted to of
5.5 ± 0.2 using NaOH or HCl 1 mol L ÿ 1 and no buffer solution was added, as indicated by various authors (Liu et al., 2006; Van Ginkel
et al., 2005). All bottles were flushed with nitrogen gas, capped tightly with butyl rubber and incubated at 55 ± 1 °C, until no further bio-hydro-
gen production was detected (normally within 10 days). For each
inoculum/initial substrate concentration tested, the experiment was conducted in duplicate under identical experimental condi-
tions. Batch tests were periodically analyzed for both quantitative
and qualitative determination of biogas production. Quantitative
production was estimated by withdrawing the extra-pressure gas
with syringes of 10''100 mL. The biogas production of blank con-
trols was subtracted from the biogas production of each sample.
Hydrogen, methane and carbon dioxide concentrations in biogas
were measured by a gas chromatograph (Agilent, Micro GC
3000A) equipped with two thermal conductivity detectors (TCD)
and two different columns. H2 and CH4 concentrations were mea-
sured using a Molesieve/5A Plot column with N2 as carrier gas at a
flow rate of 30 mL/min. CO2 content in the biogas was analyzed
using a different column (Alltech HP-PLOT U) and He as carrier
gas at a flow rate of 30 mL min ÿ 1. The operational temperature of the injection port was 100 °C, while that of Molesieve/5A and PLOT
U columns was maintained at 100 and 55 °C, respectively. 2.4. H2 productivity in continuously-fed bioreactor As the best performing organic material found in BHP tests, OFMSW was used to feed a lab-scale continuous flow stirred tank
reactor (CSTR) with a working volume of 2.3 L. The feedstock mate-
rial was mixed with SS (2:3 on wet weight), according the blending
procedure used in the anaerobic digestion plant where OFMSW
was sampled. The prepared feeding mixture had a TS content of 126 ± 2 g kg ÿ 1 and a VS content of 108 ± 3 g kg ÿ 1 and was intermittently supplied to the reactor every 3 h by peristaltic pump (Masterflex L/S, Cole-
Parmer), withdrawing an equal amount of effluent from the reac-
tor. An HRT of 3 days was established according to previous expe-
riences on similar organic materials (Liu et al., 2006), which
resulted in an OLR of 36 ± 2 gVS L ÿ 1 reactor d ÿ 1. As traditional CSTR plants, the digester was continuously stirred by an impeller with a constant speed of 100 rpm to ensure mixing
the feedstock and fermentation broth and therefore favoring bio-
gas production. The internal temperature was kept constantly at
55 ± 1 °C by water jacket piping surrounding the reactor. During the measurement observation period which lasted 35 days, pH and temperature of the fermentative broth were con-
tinuously acquired by an electrode probe (InPro 3253/225/pt1000,
Mettler''Toledo). A gas flowmeter sensor (ADM 2000, Agilent Tech-
nologies) measured with a frequency of 1 Hz the biogas flow at
reactor''s outlet. Measured data and actuators control was managed by an indus- trial PC equipped with an input/output board and a purposely-
developed software. The produced biogas was daily sampled for composition analy- sis by GC as previously described. The bioreactor start up phase
was carried out using the best natural-sourced inoculum selected,
sparged for 30 min with N2 at a flowrate of 100 mL min ÿ 1, to remove dissolved O2 and to establish strict anaerobic conditions.
Liquid samples were withdrawn from the fermentation broth every
3 days for chemical analyses. Table 1
Naturally-sourced inocula characterization. Inocula Source pH TS (g kg ÿ 1) TVFA (gACETIC ACID L ÿ 1) A Rice field soil 6.7 70 <0.01 B Soil from green urban area 7 70 <0.01 C Soil from vegetable-cultured 7.2 70 <0.01 D Harvested slurry
(soil:acclimated slurry 1:1 w/w) 5.5 55 1.05 E Harvested slurry
(soil:acclimated slurry 2:1 w/w) 5.8 61 0.63 F Harvested slurry
(soil:acclimated slurry 3:1 w/w) 6 62.5 0.53 7912 A. Tenca et al. / Bioresource Technology 102 (2011) 7910''7916 2.5. Analytical methods Total solids (TS), volatile solids (VS), COD, total alkalinity (TA), total Kjeldahl nitrogen (TKN) and ammonium nitrogen (N-NH þ
4 ) were determined according to the standard procedures (APHA,
1998). Total volatile fatty acids (TVFA) and total alkalinity (TA) in
the bulk samples were performed on a 5-times-diluted solution
of 2.5 g of wet sample filtered to 0.45 nm. TVFAs were determined
according to the acid titration method (Lahav et al., 2002). TA was
determined in liquid phase by titration with HCl to a pH endpoint
of 4.3, as suggested by APHA (1998). Specific VFAs determination
(acetic, propionic and n-butyric) in the fermentation broth was
performed using a different gas chromatograph (Varian, CP-3800)
with a capillary column of 25 m  0.32 mm diameter and flame
ionization detector (FID). He at 20 kPa pressure was used as carrier
gas, and the temperatures of injector and FID were 220 °C and
240 °C, respectively. 3. Results and discussion 3.1. Naturally-sourced inocula performances The three inocula obtained from soil (Inocula A, B and C) displayed different behaviors with regard to H2 production, as
reported in Table 2. Inoculum A showed H2 yields that decreased
with increasing substrate concentrations. Inoculum C showed
opposite trends (H2 yields increased with increasing substrate
concentrations), while there were no differences in H2 yields for
Inoculum B with different glucose concentrations. In general, no
relevant differences were found between Inocula A, B and C in
terms of H2 production efficiencies (Table 2), which in all cases
resulted relatively low (0.47''1.35 mol H2 mol ÿ 1 glucose, average of 0.91 mol H2 mol ÿ 1 glucose), as compared to the best results reported in the literature for mixed microbial cultures (e.g.
2.63 mol H2 mol ÿ 1 glucose) and for pure/selected/GM ones (e.g. 2.76 mol H2 mol ÿ 1 glucose) (Fig. 1, Table S1). VFAs concentrations in the aqueous media consistently increased during fermentation,
even if no clear dependency on substrate concentrations and H2
yield could be found. The specific contributions of acetic, n-butyric and propionic acids to the total final VFAs contents were very var-
iable, although acetic and n-butyric acids were predominant in all
trials and propionic acid was rarely detected (Table 2). The pre-
dominance of acetate and butyrate was reported by several
authors to be ideal for efficient H2 productions (Antonopoulou
et al., 2008). However, in this case, the partial inefficiency of Inoc-
ula A, B and C could have been caused by partial incompleteness of
acid fermentation (formation of valerate, caproate and other or-
ganic acids which were not measured) or due to other metabolic
pathways that might have been developed, such as solventogenesis
(Hawkes et al., 2002). This led to conclude that, although H2-pro-
ducing microbial consortia can be easily obtained from natural
sources (soil, in this case), H2 yields optimization must rely on fur-
ther developed and specialized microbial consortia, able to follow
ideal organic matter fermentation pathways. 3.2. Enhancement of bio-hydrogen yields with acclimated inocula Inocula D, E and F, fed with glucose at concentration of 3 g glucose L ÿ 1, gave higher H 2 yields, as compared to Inocula A, B and C (Table 2). In particular, Inoculum F achieved 2.02 ±
0.05 mol H2 mol ÿ 1 glucose (Table 2), i.e. more than double of those obtained with Inocula A, B and C at the same glucose concentration
(3 g glucose L ÿ 1). The H2 yields obtained by Inocula D, E and F were significantly correlated (r = 0.91, P < 0.05) to the mixing ratio of the soils mix-
ture to the acclimated slurry, used as source of microbial fermen-
tation consortia. At the same time, the H2 yields were inversely
correlated to the TVFA concentrations both at the beginning
(r = 0.90, P < 0.05) and at the end (r = 0.97, P < 0.05) of the test
(Table 2). This led to conclude that TVFA concentrations in culture
media were responsible for either direct inhibition of hydrogen-
producing bacteria or indirect effects due to pH reduction below
optimal levels, as indicated by various authors (Khanal et al.,
2004; Van Ginkel et al., 2001). For these reasons, Inoculum F
(mixing ratio of 3:1 soil:slurry) resulted in the most performing
H2 production. For all trials, acetic and n-butyric acids were the predominant VFAs produced, in agreement with the work of other researchers Table 2
Bio-hydrogen yields achieved from glucose by mixed microbial cultures, obtained from natural sources. Volatile fatty acids concentrations measured
at the end of the tests. Inocula Substrate concentration
(g glucose L ÿ 1) Hydrogen yield
(mol H2 mol ÿ 1 glucose) TVFA at the end of the
process (gACETIC ACID L ÿ 1) Acetic acid
(g L ÿ 1) Propionic acid
(g L ÿ 1) n-Butyric
acid (g L -1) STEP 1: SOIL INOCULA
A 1 1.3 ± 0.13 1.23 c 1.23 '' a '' a 3 0.89 ± 0.14 2.78 2.32 '' a 0.67 5 0.47 ± 0.17 3.83 3.10 0.27 0.75 7 0.52 ± 0.38 2.32 1.96 '' a 0.53 B 1 0.88 ± 0.27 0.55 0.55 '' a '' a 3 0.99 ± 0.18 2.63 1.89 '' a 1.09 5 d.l. b 7 0.72 ± 0.29 6.26 4.62 0.62 1.67 C 1 0.78 ± 0.21 1.07 1.07 '' a '' a 3 0.85 ± 0.61 3.24 2.84 '' a 0.59 5 1.35 ± 0.16 3.72 2.96 '' a 1.11 7 1.05 ± 0.28 3.11 1.88 '' a 1.81 STEP 2: ACCLIMATED INOCULA
D 3 1.66 ± 0.08 11.41 c 5.72 0.41 7.86 E 3 1.81 ± 0.02 7.09 4.30 '' a 4.09 F 3 2.02 ± 0.05 4.26 2.82 '' a 2.11 F 1 2.8 ± 0.66 3.49 2.39 '' a 1.61 5 1.85 ± 0.03 3.85 1.89 '' a 2.88 7 1.78 ± 0.08 3.32 2.45 '' a 1.28 a Under detection limit.
b d.l. = data lost.
c Final concentrations include VFAs contained in the inocula before fermentation and VFAs formed during glucose fermentation. A. Tenca et al. / Bioresource Technology 102 (2011) 7910''7916 7913 (Hawkes et al., 2002), while propionic acid was only detected for
Inoculum D. Together with the highest final content of TVFAs,
the presence of propionic acid justifies the lowest H2 yield of Inoc-
ulum D, among the tested acclimated inocula. Based on the stoichi-
ometry of glucose fermentation, more hydrogen evolves from
acetate-producing than from butyrate-producing fermentation
(Lin et al., 2007), while no hydrogen is produced via propionic fer-
mentation (Li et al., 2008). Therefore, if the process could be biased
towards predominantly acetate production (excepting homoaceto-
genesis), the yield could be improved (Kyazze et al., 2006).
In agreement with this statement, the H2 yield achieved by
using Inocula D, E and F at the same substrate concentration
(3 gglucose L ÿ 1) was directly correlated to the increase of the acetate/butyrate ratio (r = 0.99, P < 0.05). The identified best performing inoculum (Inoculum F), was ap- plied to test the influence of substrate concentrations on H2 pro-
duction. Increased glucose concentrations (1''7 g L ÿ 1) caused a progressive decrease in H2 yield (from 2.8 mol H2 mol ÿ 1 glucose to 1.78 mol H2 mol ÿ 1 glucose) (Table 2). This trend agrees with previous studies that indicated that the increase of substrate concentrations
may cause higher levels of inhibitory metabolites and change
chemical equilibrium, depressing further hydrogen production
(Van Ginkel et al., 2001; Zhang et al., 2003). In this case, TVFAs
concentration measured for Inoculum F, even if tested at different
glucose concentrations, did not show considerable differences at
the end of the trials (Table 2). This revealed that there was a limit
on TVFAs concentration, above which further fermentation
resulted inhibited or not optimized. For this reason, the lowest
substrate concentration used (1 gglucose L ÿ 1) allowed obtaining the best H2 yield. At the same time, the lowest substrate concentration
allowed prevalent production of acetate (high acetate/butyrate
ratio of 1.48), i.e. an optimized hydrogen production, as reported
by various studies (Van Ginkel et al., 2001; Liu et al., 2006). In any case, the acclimation strategy (Inoculum F) resulted in harvesting considerably more efficient microbial consortia and in improving H2 yield, with respect to the soil inocula (Inocula A''C)
(Table 2). The best H2 yield obtained (2.8 ± 0.66 mol H2 mol ÿ 1 glucose) was comparable to the best results cited in literature for mixed
and also for both pure/selected and genetically modified microbial
cultures, as shown in Fig. 1 and reported in Table S1. 3.3. BHP tests on selected organic substrates The BHP tests on four selected organic substrates were performed following the methodology developed in the previous
section (Inoculum F, substrate concentration 1 gVS L ÿ 1). As demon- strated, this methodology allows the prompt presence of an
efficient microbial community and avoids microbiological or met-
abolic inhibiting conditions, so that it is ideal for measuring the po-
tential bio-H2 production of a whatever substrate. Representative trends of the H2 production obtained are showed in Fig. 2. The H2 production reached, in all cases, a plateau
after 5 days and the duplicates showed high repeatability. As
expected, the control test (fed with glucose) achieved almost
the same H2 production (2.9 ± 0.09 mol H2 mol ÿ 1 glucose) previously reached at the same substrate concentration. Among the tested
biomasses, OFMSW and MBW produced the highest total amount
of hydrogen, i.e. 202 ± 3 and 176 ± 2 NLH2 kg ÿ 1 VS , respectively, while MS reached 118 ± 2 NLH2 kg ÿ 1 VS and SS only 14 ± 1 NLH2 kg ÿ 1 VS (Table 3). While SS, as expected, showed low BHP, OFMSW, MBW and MS showed relatively interesting BHP, as compared to various
literature results (Lay et al., 2003; Kim et al., 2004). 3.4. Bio-hydrogen productivity in CSTR Hydrogen production in CSTR fermentation process was ana- lyzed to gain an insight into possible production rates of real scale
plant implementations by using as benchmark the potential values
obtained in BHP assays. Indeed, these former values are obtained in
most optimal environmental and operative conditions allowed.
However, a real-scale successful implementation of biohydrogen
production by dark fermentation will only rely on continuously-
fed co-fermentation of highly concentrated organic mixtures (typ-
ically 50''150 gVS L ÿ 1, for CSTR reactors) achieving H 2 yields not too far from those obtained with BHP tests. The CSTR was inoculated with Inoculum F and the start-up phase lasted till day 5, when stable biogas production and H2 content in
biogas were established (Fig. 3). The process was continuously oper-
ated for more than one month with stable production and pH as
shown by graphs in Fig. 3. The broth average pH (5.78 ± 0.10) re-
sulted closely stable around the optimum for hydrogen production
(Liu et al., 2006; Lin et al., 2008) without any addiction of chemical
agents for pH adjustments, which is quite relevant for practical
implementation of the process. The average hydrogen content in
biogas was of 35 ± 4% and methane was never found in the biogas,
likely due to favorable conditions created by the short HRT and the
relative acidity of the broth (Table 4). Fig. 2. Batch bio-H2 potential (BHP) tests applied on four different organic
substrates plus glucose (as reference). Cumulative bio-hydrogen production trends
(average of duplicates). Table 3
Bio-hydrogen potential (BHP) productions from four substrates studied and glucose (as reference). Substrate TS (g kg ÿ 1) VS (g kg ÿ 1) H2 produced Total production (Nml H2) NL H2 kg ÿ 1 w.w. NL H2 kg ÿ 1 VS Glucose 1000 1000 53.4 ± 0.5 356 ± 9 356 ± 9 2 :90  0 :09molH 2 mol ÿ 1 glucose Maize silage 335 308 17.7 ± 0.2 39.6 ± 0.5 118 ± 2 Swine slurry 30 24 2.2 ± 0.1 0.4 ± 0.02 14 ± 1 OFMSW 270 240 30.3 ± 0.5 54.5 ± 0.9 202 ± 3 Market bio-waste 112 104 26.4 ± 0.3 19.7 ± 0.2 176 ± 2 7914 A. Tenca et al. / Bioresource Technology 102 (2011) 7910''7916 During the operation the reactor specific production rate of hydrogen resulted 2.2 ± 0.4 NLH2 L ÿ 1 reactor d ÿ 1 with a yield of 60 ± 4 NLH2 kg ÿ 1 VS-added (Table 4). These results are comparable with or superior to other results achieved in previous laboratory CSTR tests:
for example, Liu et al. (2006) used household solid waste with HRT of
2 days, achieving 43 NLH2 kg ÿ 1 VS . This H2 production yield repre- sented only the 30% of the BHP measured for OFMSW
(202 ± 3 NLH2 kg ÿ 1 VS ). This was probably linked to the fact that high substrate concentrations induce high metabolites concentrations,
as demonstrated before (Fig. 1): TVFA concentrations were very high
in the digestate (Table 4), as compared to the batch tests (Table 2),
even if relatively high TA allowed the pH to remain stable (Fig. 3). Other authors reported very similar results of the CSTR perfor- mance with concentrated organic mixtures (Table S2). Perhaps,
the CSTR process, which was here intentionally studied because
it is actually the most diffused technology in full-scale biogas
plants, is not the best solution for achieving optimized dark fer-
mentation and, especially when the process is run with relatively
high organic matter concentrations (>100 gVS kg ÿ 1 wet weight), is still relatively far from meeting the maximized BHP. As demonstrated
by recent works, more satisfactory performances might be reached
by using different strategies, taking into account hydrodynamics and reaction kinetics models to improve the bioreactor design
(Ding et al., 2010; Wang et al., 2010), although all these experi-
ences were based on low-concentrated and promptly-soluble or-
ganic substrates such as glucose and molasses. Other process
types, such as leaching bed reactors (Han and Shin, 2004), should
be further investigated, for improving dark fermentation reliability
as a really applicable technology. 4. Conclusions It is possible to obtain efficient mixed microbial cultures from natural sources, by proper acclimation to organic substrates. Such
types of inoculum, used in batch reactors with proper substrate
(1 gglucose L ÿ 1) and metabolites concentrations, allowed H 2 yields comparable to pure/selected/GM microbial cultures achievements.
This methodology (BHP test), can be applied to organic substrates
of possible interest for future applications, to test their potentiality
of producing bio-H2. However, the CSTR fermentation of a concen-
trated mixture of OFMSW, resulted in around 30% of its BHP, sug-
gesting that further efforts are needed for future applications of
dark fermentation in full-scale plants. Acknowledgements The authors are grateful to Regione Lombardia, General Direc- torate of Agriculture, for its financial support to project AGRIDEN. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.biortech.2011.05.088. References Antonopoulou, G., Gavala, H.N., Ioannis, V., Skiadas, K., Angelopoulos, C., Lyberatos, G., 2008. Biofuels generation from sweet sorghum: fermentative hydrogen
production and anaerobic digestion of the remaining biomass. Bioresour.
Technol. 99, 110''119. Akutsu, Y., Lee, D.Y., Li, Y.Y., Noike, T., 2009. Hydrogen production potentials and fermentative characteristics of various substrates with different heat-
pretreated natural microflora. Int. J. Hydrogen Energy 34, 5365''5372. American Public Health Association (APHA), 1998. Standard Methods for the Examination of Water and Wastewater, 20th ed. Washington, DC, USA. Ding, J., Wang, X., Zhou, X.F., Ren, N.Q., Guo, W.Q., 2010. CFD optimization of continuous stirred-tank (CSTR) reactor for biohydrogen production. Bioresour.
Technol. 101, 7005''7013. Fantozzi, F., Buratti, C., 2009. Biogas production from different substrates in an experimental Continuously Stirred Tank Reactor anaerobic digester. Bioresour.
Technol. 100, 5783''5789. Guo, X.M., Trably, E., Latrille, E., Carrere, H., Steyer, J.P., 2010. Hydrogen production from agricultural waste by dark fermentation: a review. Int. J. Hydrogen Energy
35, 10660''10673. Han, S.K., Shin, H.S., 2004. Biohydrogen production by anaerobic fermentation of food waste. Int. J. Hydrogen Energy 29, 569''577. Hawkes, F.R., Dinsdale, R., Hawkes, D.L., Hussy, I., 2002. Sustainable fermentative hydrogen production: challenges for process optimisation. Int. J. Hydrogen
Energy 27, 1339''1347. Khanal, S.K., Chen, W.H., Li, L., Sung, S.W., 2004. Biological hydrogen production: effects of pH and intermediate products. Int. J. Hydrogen Energy 29, 1123''1131. Kim, S.H., Han, S.K., Shin, H.S., 2004. Feasibility of biohydrogen production by anaerobic co-digestion of food waste and sewage sludge. Int. J. Hydrogen
Energy 29, 1607''1616. Kyazze, G., Martinez-Perez, N., Dinsdale, R., Premier, G.C., Hawkes, F.R., Guwy, A.J., Hawkes, D.L., 2006. Influence of substrate concentration on the stability and
yield of continuous biohydrogen production. Biotechnol. Bioeng. 93, 971''979. Lahav, O., Morgan, B., Loewenthal, R.E., 2002. Rapid, simple and accurate method for measurement of VFA and carbonate alkalinity in anaerobic reactors. Environ.
Sci. Technol. 36, 2736''3274. Lay, J.J., Lee, Y.J., Noike, T., 1999. Feasibility of biological hydrogen production from organic fraction of municipal solid waste. Water Resour. 33, 2579''2586. Lay, J.J., Fan, K.S., Chang, J.L., Ku, C.H., 2003. Influence of chemical nature of organic wastes on their conversion to hydrogen by heat-shock digested sludge. Int. J.
Hydrogen Energy 28, 1361''1367. Levin, D.B., Pitt, L., Love, M., 2004. Biohydrogen production: prospects and limitations to practical application. Int. J. Hydrogen Energy 29, 173''185. Table 4
Operative parameters and process performances of the CSTR fermenter, fed with
organic fraction of municipal source-separated waste and swine slurry. Parameter Average in the
observation period Substrate TS (g kg ÿ 1) 126 ± 2 VS (g kg ÿ 1) 108 ± 3 COD (g kg ÿ 1) 151 ± 11 TVFA (gACETIC ACID L ÿ 1) 4.58 ± 0.2 TA (gCaCO 3 L ÿ 1) 7.34 ± 0.2 TVFA/TA 0.62 TKN (g kg ÿ 1) 4.4 ± 0.2 N-NH þ
4 (g kg ÿ 1) 2.2 ± 0.1 Process HRT (d) 3 OLR (gVS L ÿ 1 reactor d ÿ 1) 36 ± 2 Operational average pH 5.78 ± 0.1 H2 content in the biogas (%) 35 ± 8 H2 production rate (NL H2 L ÿ 1 reactor d ÿ 1) 2.2 ± 0.4 H2 yield (NL H2 kg ÿ 1 TS-added ) 52 ± 4 H2 yield (NL H2 kg ÿ 1 VS-added ) 60 ± 4 Digestate TS (g kg ÿ 1) 48 ± 7 VS (g kg ÿ 1) 38 ± 5 COD (g kg ÿ 1) 58 ± 6 TVFA (gACETIC ACID L ÿ 1) 6.45 ± 0.2 TA (gCaCO 3 L ÿ 1) 5.35 ± 0.6 TVFA/TA 1.20 TKN (g kg ÿ 1) 2.6 ± 0.4 N-NH þ
4 (g kg ÿ 1) 1.6 ± 0.3 Fig. 3. Trend of bio-hydrogen production and pH in the CSTR system studied. A. Tenca et al. / Bioresource Technology 102 (2011) 7910''7916 7915 Li, C., Fang, H.H.P., 2007. Fermentative hydrogen production from wastewater and solid wastes by mixed cultures. Crit. Rev. Environ. Sci. Technol. 37, 1''39. Li, J., Ren, N., Li, B., Qin, Z., He, J., 2008. Anaerobic biohydrogen production from monosaccharides by a mixed microbial community culture. Bioresour. Technol.
99, 6528''6537. Lin, P.Y., Whang, L.M., Wu, Y.R., Ren, W.J., Hsiao, C.J., Li, S.L., Chang, J.S., 2007. Biological hydrogen production of the genus Clostridium: metabolic study and
mathematical model simulation. Int. J. Hydrogen Energy 32, 1728''1735. Lin, C.Y., Chang, C.C., Hung, C.H., 2008. Fermentative hydrogen production from starch using natural mixed cultures. Int. J. Hydrogen Energy 33, 2445''2453. Liu, D., Zeng, R.J., Angelidaki, I., 2006. Hydrogen and methane production from household solid waste in the two-stage fermentation process. Water Resour. 40,
2230''2236. Morimoto, M., Atsuko, M., Atif, A.A.Y., Ngan, M.A., Fakhru''l-Razi, A., Iyuke, S.E., Bakir, A.M., 2004. Biological production of hydrogen from glucose by natural
anaerobic microflora. Int. J. Hydrogen Energy 29, 709''713. Nandi, R., Sengupta, S., 1998. Microbial production of hydrogen: an overview. Crit. Rev. Microbiol. 24, 61''84. Oh, S.E., Van Ginkel, S., Logan, B.E., 2003. The relative effectiveness of pH control and heat treatment for enhancing biohydrogen gas production. Environ. Sci.
Technol. 37, 5186''5190. Schröder, C., Selig, M., Schönheit, P., 1994. Glucose fermentation to acetate, CO2 and H2 in the anaerobic hyperthermophilic eubacterium Thermotoga maritima:
involvement of the Embden''Meyerhof pathway. Arch. Microbiol. 161, 460''470. Ueno, Y., Haruta, S., Ishii, M., Igarashi, Y., 2001. Microbial community in anaerobic hydrogen-producing microflora enriched from sludge compost. Appl. Microbiol.
Biotechnol. 57, 555''562. Van Ginkel, S., Sung, S., Lay, J.J., 2001. Biohydrogen production as a function of pH and substrate concentration. Environ. Sci. Technol. 35, 4726''
4730. Van Ginkel, S., Logan, B.E., 2005. Increased biological hydrogen production with reduced organic loading. Water Resour. 39, 3819''3826. Van Ginkel, S.W., Oh, S.E., Logan, B.E., 2005. Biohydrogen gas production from food processing and domestic wastewaters. Int. J. Hydrogen Energy
30, 1535''1542. Van Niel, E.W.J., Budde, M.A.W., De Haas, G.G., Van der Wal, F.J., Claassen, P.A.M., Stams, A.J.M., 2002. Distinctive properties of high hydrogen producing extreme
thermophiles, Caldicellulosiruptor saccharolyticus and Thermotoga elfii. Int. J.
Hydrogen Energy 27, 1391''1398. Wang, B., Wan, W., Wang, J., 2008. Inhibitory effect of ethanol, acetic acid, propionic acid and butyric acid on fermentative hydrogen production. Int. J. Hydrogen
Energy 33, 7013''7019. Wang, X., Ding, J., Guo, W.Q., Ren, N.Q., 2010. A hydrodynamics''reaction kinetics coupled model for evaluating bioreactors derived from CFD simulation.
Bioresour. Technol. 101, 9749''9757. Zhang, T., Liu, H., Fang, H.H.P., 2003. Biohydrogen production from starch in wastewater under thermophilic condition. J. Environ. Manag. 69,
149''156. 7916 A. Tenca et al. / Bioresource Technology 102 (2011) 7910''7916 Supplemental data
Looking for practical tools to achieve next-future
applicability of dark fermentation to produce bio-
hydrogen from organic materials in Continuously
Stirred Tank Reactors.
Tenca A. a '', Schievano A.b*, Lonati, S. b, Malagutti L.c, Oberti R.a, Adani F. b* aDipartimento di Ingegneria Agraria, Università degli Studi di Milano, Via Celoria 2, 20133
Milano, Italy.
bDipartimento di Produzione Vegetale, Università degli Studi di Milano, Via Celoria 2, 20133
Milano, Italy.
cDipartimento di Scienze Animali, Università degli Studi di Milano, Via Celoria 2, 20133
Milano, Italy.
Supplemental data contains: - Table S1 and S2 - Literature referred to tables S1 and S2. Table S1 '' Maximum bio-H2 yields obtained in batch experiments with
genetically modified, pure/selected and naturally-sourced inocula in
literature and in this study Source of
Operational conditions Substrate (glucose) concentration (g/L) Yield (mol H2/mol glucose) Reference Genetically
aerogenes HU-101
(mutant A-1) 37° - pH 6.8 20 0.84 Rachman et al. 1997 Enterobacter
aerogenes HU-101 37° - pH 6.8 20 0.83 Rachman et al. 1997 '' e-mail:,, (mutant HZ-3)
aerogenes HU-101
(mutant AY-2) 37° - pH 6.8 20 1.17 Rachman et al. 1997 Escherichia coli
BW25113 (mutant
hyaB hybC hycA
fdoG frdC ldhA
aceE) 37° 18 1.35 Maeda et al. 2007 Escherichia coli K-
12 strain W3110
(mutant SR15,
'' ldhA, ''frdBC) 37° - pH 6 10.8 1.82 Yoshida et al 2006 Escherichia coli K-
12 strain W3110
(mutant SR14
'' hycA, ''ldhA, '' frdBC) 37° - pH 6 10.8 1.87 Yoshida et al 2006 Escherichia coli
BL21 (DE3 ''iscR
pAF pYdbK) 37° - pH 7 5 1.88 Akhtar and
Jones 2009 Escherichia coli
BW535 (mutant
JW135 Hyd1-,
Hyd2-derivative) 37° 4 1.25 Bisaillon et al. 2006 Escherichia coli
BW535 (mutant
LJT135, idhA
mutant of JW135) 37° 1 1.55 Bisaillon et al. 2006 Escherichia coli
BW535 (mutant
FJT135, fhlA
mutant of JW135 ) 37° 5 1.4 Bisaillon et al. 2006 Escherichia coli
BW535 (mutant
DJT135, idhA, fhlA
mutant of JW135) 37° 0.3 1.98 Bisaillon et al. 2006 Escherichia coli
BW135 (mutant
DJT135, ''hya-Km,
'' hyb-Km, ''ldhA, 35° - pH 4.5 4.5 0.77 Gosh and Hallenbeck 2010 fhlA-C) Pure cultures w.t. Enterobacter
aerogenes HU-101 37° - pH 6.8 20 0.56 Rachman et al. 1997 Escherichia coli K-
12 strain W3110 37° - pH 6 10.8 1.08 Yoshida et al 2006 Escherichia coli
BW25113 37° 18 0.65 Maeda et al. 2007 Rhodopseudomonas
Palustris P4 30° - buffer solution 1 2.76 Oh et al. 2002 Rhodopseudomonas
Palustris P4 30° - buffer solution 5 1.3 Oh et al. 2002 Rhodopseudomonas
Palustris P4 30° - buffer solution 10 0.98 Oh et al. 2002 Rhodopseudomonas
Palustris P4 30° - buffer solution 20 0.7 Oh et al. 2002 Rhodopseudomonas
Palustris P4 30° - buffer solution 50 0.66 Oh et al. 2002 Citrobacter sp. Y19 36° - pH 6-7 1 2.49 Oh et al. 2003b Citrobacter sp. Y19 36° - pH 6-7 5 1.4 Oh et al. 2003b Citrobacter sp. Y19 36° - pH 6-7 10 1.05 Oh et al. 2003b Citrobacter sp. Y19 36° - pH 6-7 20 0.7 Oh et al. 2003b Citrobacter sp. Y19 36° - pH 6-7 50 0.7 Oh et al. 2003b Clostridium sp. No.
2 36° - pH 6 10 1.99 Taguchi et al. 1994 Enterobacter
aerogenes strain
HO-39 38° - pH 6-7 10 1 Yokoi et al. 1995 Enterobacter
cloacae IIT-BT 08 36° - uncontrolled pH (initial 6.0) 10 2.2 Kumar and Das 2000 Clostridium
beijerinckii AM21B 36° - uncontrolled pH 10 2 Taguchi et al. 1992 Citrobacter
Freundii - 7.7 1.29 Kumar and Vatsala 1989 Clostridium
pasteurianum - 7.6 1.5 Brosseau et al. 1982 Citrobacter
intermedius - 7.6 1 Brosseau et al. 1982 Citrobacter
intermedius 35° - pH 4.5 22.5 0.35 Brosseau et al. 1982 Citrobacter
intermedius 35° - pH 8.5 4.5 1.54 Brosseau et al. 1982 Citrobacter
intermedius 35° - pH 8.5 22.5 0.29 Brosseau et al. 1982 Citrobacter
intermedius 25° - pH 6.5 4.5 1.34 Brosseau et al. 1982 Citrobacter
intermedius 25° - pH 6.5 22.5 0.23 Brosseau et al. 1982 Citrobacter
intermedius 45° - pH 6.5 4.5 0.91 Brosseau et al. 1982 Citrobacter
intermedius 45° - pH 6.5 22.5 0.23 Brosseau et al. 1982 Citrobacter
intermedius 25° - pH 4.5 13.5 0.18 Brosseau et al. 1982 Citrobacter
intermedius 25° - pH 8.5 13.5 0.43 Brosseau et al. 1982 Citrobacter
intermedius 45° - pH 4.5 13.5 0.29 Brosseau et al. 1982 Citrobacter
intermedius 45° - pH 8.5 13.5 0.56 Brosseau et al. 1982 Citrobacter
intermedius 35° - pH 6.5 13.5 1.69 Brosseau et al. 1982 Klebsielle oxytoca
HP1 35° - initial pH 7.0 9 1.1 Minnan et al. 2005 Klebsielle oxytoca
HP1 35° - initial pH 7.0 18 0.36 Minnan et al. 2005 Klebsielle oxytoca
HP1 35° - initial pH 7.0 27 0.22 Minnan et al. 2005 Klebsielle oxytoca
HP1 35° - initial pH 7.0 36 0.13 Minnan et al. 2005 Ethanoligenens
harbinense B49 35° - uncontrolled pH (initial 14.5 2.2 Guo et al. 2009 6.0) (optimized nutrients concentration) Naturally-sourced
mixed cultures
Mixed bacterial
cultures from
(anaerobic granular
sludge from a
UASB reactor) 37° - pH 7.5 5 1.46 Davila- Vazquez et al. 2008 Heat-conditioned
anaerobic digested
sludge 35° - pH 6-7 2 1.4 Kawagoshi et al. 2005 Unconditioned
anaerobic digested
sludge 35° - pH 6-7 2 1.3 Kawagoshi et al. 2005 Refuse compost
with pH
conditioning 35° - pH 6-7 2 0.5 Kawagoshi et al. 2005 Kiwi soil heat
treated or with pH
conditioning 35° - pH 6-7 2 0.5 Kawagoshi et al. 2005 Lake sediment 35° - pH 6-7 2 0.9 Kawagoshi et al. 2005 Dewatered and
thickened sludge
from a wastewater
treatment plant 30° - pH 6.2 3.76 1.17 Salerno et al. 2006 Dewatered and
thickened sludge
from a wastewater
treatment plant 30° - initial pH 6.2 - CO2 scavenging (KOH) 2 2 Park et al. 2005 Dewatered and
thickened sludge
from a wastewater
treatment plant 30° - initial
pH 6.2 - no CO2 scavenging 2 1.4 Park et al. 2005 Anaerobic sludge
from a local
municipal sewage
treatment plant, 37° - pH 6.0 10 1.75 Zheng and Yu 2005 heat treated
Centrifugate of
digested sewage
sludge, heat
pretreated 60° - pH 7 10 1.8 Zurawski et al. 2005 Compost 37° - Initial pH 5.5 14.0 1.0 Van ginkel et al. 2001 Compost 37° - Initial pH 5.5 21.0 0.8 Van ginkel et al. 2001 Compost 37° - Initial pH 5.5 28.0 0.9 Van ginkel et al. 2001 Compost 37° - Initial pH 5.5 35.0 0.5 Van ginkel et al. 2001 Compost 37° - Initial pH 5.5 42.0 0.4 Van ginkel et al. 2001 Compost 37° - Initial pH 5.5 1.4 2.2 Van ginkel et al. 2001 Compost 37° - Initial pH 5.5 2.8 1.9 Van ginkel et al. 2001 Compost 37° - Initial pH 5.5 0.5 2.6 Van ginkel et al. 2001 Potato soil 37° - Initial pH 5.5 14.0 0.91 Van ginkel et al. 2001 Potato soil 37° - Initial pH 5.5 21.0 0.67 Van ginkel et al. 2001 Potato soil 37° - Initial pH 5.5 28.0 0.48 Van ginkel et al. 2001 Potato soil 37° - Initial pH 5.5 35.0 0.37 Van ginkel et al. 2001 Potato soil 37° - Initial pH 5.5 42.0 0.29 Van ginkel et al. 2001 Soybean soil 37° - Initial pH 5.5 14.0 0.84 Van ginkel et al. 2001 Soybean soil 37° - Initial pH 5.5 21.0 0.50 Van ginkel et al. 2001 Soybean soil 37° - Initial pH 5.5 28.0 0.46 Van ginkel et al. 2001 Soybean soil 37° - Initial pH 5.5 35.0 0.32 Van ginkel et al. 2001 Soybean soil 37° - Initial pH 5.5 42.0 0.40 Van ginkel et al. 2001 Tomato plants soil,
heat treated 26° - pH 6 0.9 0.92 Logan et al. 2002 Palm oil mill
effuent (POME)
sludge 50° - uncontrolled pH 10.0 0.38 Morimoto et al. 2004 Palm oil mill
effuent (POME)
sludge 60° - pH 7 10.0 0.43 Morimoto et al. 2004 Sludge compost
(from Malaysia) 50° - pH 7 10.0 0.85 Morimoto et al. 2004 Sludge compost
(from Malaysia) 60° - pH 7 10.0 0.93 Morimoto et al. 2004 CREST compost
from a compost
manufacturing plant
(Philippines) 50° - pH 7 10.0 0.50 Morimoto et al. 2004 CREST compost
from a compost
manufacturing plant
(Philippines) 60° - pH 7 10.0 0.96 Morimoto et al. 2004 CREST compost
from a compost
manufacturing plant
(Philippines) 60° - pH 7 10.0 1.25 Morimoto et al. 2004 Dewatered
anaerobic sludge
(heat treated) 25° - 6.2 2.8 0.97 Oh et al. 2003a Anaerobic sludge at
a local cattle
manure treatment
plant (acid
treatment) 35.5° - pH initial 7 21.3 0.90 Cheong and Hansen 2006 Our results Mixed culture form
soils - Inoculum A 55° - pH uncontrolled (initial 5.5) 1 1.30 This study Mixed culture form
soils - Inoculum A 55° - pH uncontrolled (initial 5.5) 3 0.89 This study Mixed culture form 55° - pH 5 0.47 This study soils - Inoculum A uncontrolled (initial 5.5) Mixed culture form
soils - Inoculum A 55° - pH uncontrolled (initial 5.5) 7 0.52 This study Mixed culture form
soils - Inoculum B 55° - pH uncontrolled (initial 5.5) 1 0.88 This study Mixed culture form
soils - Inoculum B 55° - pH uncontrolled (initial 5.5) 3 0.99 This study Mixed culture form
soils - Inoculum B 55° - pH uncontrolled (initial 5.5) This study Mixed culture form
soils - Inoculum B 55° - pH uncontrolled (initial 5.5) 7 0.72 This study Mixed culture form
soils - Inoculum C 55° - pH uncontrolled (initial 5.5) 1 0.78 This study Mixed culture form
soils - Inoculum C 55° - pH uncontrolled (initial 5.5) 3 0.85 This study Mixed culture form
soils - Inoculum C 55° - pH uncontrolled (initial 5.5) 5 1.35 This study Mixed culture form
soils - Inoculum C 55° - pH uncontrolled (initial 5.5) 7 1.05 This study Mixed culture
(Inoculum F) 55° - pH uncontrolled (initial 5.5) 1 2.8 This study - Acclimation strategy Mixed culture
(Inoculum F) 55° - pH uncontrolled (initial 5.5) 3 2.02 This study - Acclimation strategy Mixed culture
(Inoculum F) 55° - pH uncontrolled (initial 5.5) 5 1.85 This study - Acclimation strategy Mixed culture
(Inoculum F) 55° - pH uncontrolled (initial 5.5) 7 1.78 This study - Acclimation strategy Table S2 '' Dark fermentation performances with concentrated organic
waste mixtures in Continuously Stirred Tank Reactors (CSTR) at different
process conditions, in literature and in this study
T Fed subst. concentr ation H
T OLR Prod. rate Max. yield Ref. °C gVS L -1 gVS L -1 d-1 LH2 L -1 d-1 LH2 kgVS -1 OFMSW 55 126 3 d 36 2.2 60 This study slaughterh
ouse and
waste 55 28.2 2 d 14.1 0.2 16.5 Karlsson et al.,
2008. food
waste 37 40 5 d 8 1 125 Shin and Youn, 2005. pretreated
manures 36 70 32 h 46.5 0.8 31.5 Yan et al., 2010. pig slurry 70 33.3 1 d 33.3 0.1 4 Thomas et al.,
2009. palm oil
effluent 55 27.5 2 d 13.8 1 77 Isnazunit a et al., 2011. Beet
er 35 10 10 h 24 1.1 44.5 Zhu et al., 2009. Househol
d solid
waste 37 75 2 d 37.5 1.6 43 Liu et al., 2006. cheese
whey 36 10 1 d 10 0.5 45 Yang et al., 2007. References cited in Table S1 and S2

Akhtar, M.K., Jones, P.R., 2009. Construction of a synthetic YdbK-
dependent pyruvate - H2 pathway in Escherichia coli BL21 (DE3).
Metab. Eng. 11, 139-147.

Bisaillon, A., Turcot, J., Hallenbeck, P.C., 2006. The effect of
nutrient limitation on hydrogen production by batch cultures of
Escherichia coli. Int. J. Hydrog. Energy 31, 1504-1508.

Brosseau, J.D., Magaritis, A.J., Zajic, E., 1982. The effect of
temperature on the growth and hydrogen production by Citrobacter
intermedius. Biotechnol. Lett. 3, 307-312.

Cheong, D.Y., Hansen, C.L., 2006. Acidogenesis characteristics of
natural, mixed anaerobes converting carbohydrate-rich synthetic
wastewater to hydrogen. Process Biochem. 41, 1736-1745.

Davila-Vazquez, G., Alatriste-Mondragòn, F., de Leòn-Rodrìguez,
A., Razo-Flores, E., 2008. Fermentative hydrogen production in
batch experiments using lactose, cheese whey and glucose: Influence
of initial substrate concentration and pH. Int. J. Hydrog. Energy 33,

Ghosh, D., Hallenbeck, P.C., 2010. Response surface methodology
for process parameter optimization of hydrogen yield by the
metabolically engineered strain Escherichia coli DJT135. Bioresour.
Technol. 101, 1820-1825.

Guo, W.Q., Ren, N.Q., Wang, X.J., Xiang, W.S., Ding, J., You, Y.,
Liu, B.F., 2009. Optimization of culture conditions for hydrogen
production by Ethanoligenens harbinense B49 using response
surface methodology. Bioresour. Technol. 100, 1192-1196.

Isnazunita, I., Mohd, A.H., Nor''Aini, A.R., Chen, S.S., 2011. Effect
of retention time on biohydrogen production by microbial consortia
immobilised in polydimethylsiloxane. African J. of Biotechnol., 10,
601 '' 609.
Karlsson, A., Vallin, L., Ejlertsson, J., 2008. Effects of temperature,
hydraulic retention time and hydrogen extraction rate on hydrogen
production from the fermentation of food industry residues and
manure. Int. J. Hydrogen Energy 33(3), 953 - 962

Kawagoshi, Y., Hino, N., Fujimoto, A., Nakao, M., Fujita, Y.,
Sugimura, S., Furukawa, K., 2005. Effect of inoculum conditioning
on hydrogen fermentation and pH effect on bacterial community
relevant to hydrogen production. J. Biosci. Bioeng. 100, 524-530.

Kumar, N., Das, D., 2000. Enhancement of hydrogen production by
Enterobacter cloacae IIT-BT 08. Process Biochem. 35, 589-593.

Kumar, G.R., Vatsala, T.M., 1989. Hydrogen production from
glucose by Citrobacter Freundii. Ind. J. Exp. Biol. 27, 824-825.

Logan, B.E., Oh, S.E., Van Ginkel, S., 2002. Biological hydrogen
production measured in batch anaerobic respirometer. Environ. Sci.
Technol. 36, 2530-2535.

Maeda, T., Sanchez-Torres, V., Wood, T.K., 2007. Escherichia coli
hydrogenase 3 is a reversible enzyme possessing hydrogen uptake
and synthesis activities. Appl. Microbiol. Biotechnol. 76, 1035-1042.

Minnan, L., Jinli, H., Xiaobin, W., Huijuan, X., Jinzao, C.,
Chuannan, L., Fengzhang, Z., Liangshu, X., 2005. Isolation and
characterization of a high H2-producing strain Klebsiella oxytoca
HP1 from a hot spring. Res. Microbiol. 156, 76-81.

Morimoto, M., Atsuko, M., Atif, A.A.Y., Ngan, M.A., Fakhru''l-
Razi, A., Iyuke, S.E., Bakir, A.M., 2004. Biological production of
hydrogen from glucose by natural anaerobic microflora. Int. J.
Hydrog. Energy 29, 709-713.

Noike, T., Mizuno, O., 2000. Hydrogen fermentation of organic
municipal wastes. Water Sci. Technol. 42,155 - 162.
Oh, S.E., Van Ginkel, S., Logan, B.E., 2003a. The Relative
Effectiveness of pH Control and Heat Treatment for Enhancing
Biohydrogen Gas Production. Environ. Sci. Technol. 37, 5186-5190.

Oh, Y.K., Seol, E.H., Lee, E.Y., Park, S., 2002. Fermentative
hydrogen production by a new chemoheterotrophic bacterium
Rhodopseudomonas Palustris P4. Int. J. Hydrog. Energy 27, 1373-

Oh, Y.K., Seol, E.H., Kim, J.R., Park, S., 2003b. Fermentative
biohydrogen production by a new chemoheterotrophic bacterium
Citrobacter sp. Y19. Int. J. Hydrog. Energy 28, 1353-1359.

Park, W., Hyun, S.H., Oh, S.E., Logan, B.E., Kim, I.S., 2005.
Removal of Headspace CO2 Increases Biological Hydrogen
Production. Environ. Sci. Technol. 39, 4416-4420.

Rachman, M.A., Furutani, Y. Nakashimada, Y., Kakizono, T.,
Nishio, N., 1997. Enhanced Hydrogen Production in Altered Mixed
Acid Fermentation of Glucose by Enterobacter aerogenes. J.
Ferment. Bioeng. 83, 358-363.

Salerno, M.B., Park, W., Zuo, Y., Logan, B.E., 2006. Inhibition of
biohydrogen production by ammonia. Water Res. 40, 1167-1172.

Shin, H., Youn, J., 2005. Conversion of food waste into hydrogen by
thermophilic acidogenesis. Biodegradation. 16, 33 '' 44.

Taguchi, F., Chang, J.D., Takiguchi, S., Morimoto, M., 1992.
Efficient Hydrogen Production from Starch by a Bacterium isolated
from termites. J. Ferment. Bioeng. 73, 244-245.

Taguchi, F., Mizukami, N., Hasegawa, K., Saito-Taki, T., 1994.
Microbial conversion of arabinose and xylose to hydrogen by a
newly isolated Clostridium sp. No. 2. Can. J. Microbiol. 40, 228-233.

Kotsopoulos, T.A., Fotidis, I.A., Gerassimos, N.T., Martzopoulos,
G. 2009. Biohydrogen production from pig slurry in a CSTR reactor system with mixed cultures under hyper-thermophilic temperature,
Biomass and Bioen., 33: 1168-1174

Van Ginkel, S., Sung, S., Lay, J.J., 2001. Biohydrogen production as
a function of pH and substrate concentration. Environ. Sci. Technol.
35, 4726-4730.

Xing, Y., Li, Z., Fan, Y., Hou, H., 2010. Biohydrogen production
from dairy manures with acidification pretreatment by anaerobic
fermentation. Environ Sci. Pollut. Res. 17, 392 '' 399.

Yang, P., Zhang, R., McGarvey, J.A., Benemann, J.R., 2007.
Biohydrogen production from cheese processing wastewater by
anaerobic fermentation using mixed microbial communities. Int. J.
Hydrogen Energy 32, 4761 '' 71.

Yokoi, H., Ohkawara, T., Hirose, J., Hayashi, S., Takasaki, Y., 1995.
Characteristics of hydrogen production by aciduric Enterobacter
aerogenes strain HO-39. J. Ferment. Bioeng. 80, 571-574.

Yoshida, A., Nishimura, T., Kawaguchi, H., Inui, M., Yukawa, H.,
2006. Enhanced hydrogen production from glucose using ldh- and
frd-inactivated Escherichia coli strains. Appl. Microbiol. Biotechnol.
73, 67-72.

Zheng, X.J., Yu, H.Q., 2005. Inhibitory effects of butyrate on
biological hydrogen production with mixed anaerobic cultures. J.
Environ. Manag. 74, 65-70.

Zhu, G., Chaoxiang, L., Guihua, X., Jianzheng, L., Yanli, G., Lijun,
C., Haichen, L., 2009. Simultaneous Biohydrogen Production and
Wastewater Treatment in Continuous Stirred Tank Reactor (CSTR)
Using Beet Sugar Wastewater. Proceedings of the 2009 International
Conference on Energy and Environment Technology, ICEET '09,
Vol. 2: 737-740, IEEE Computer Society Washington, DC, USA

Zurawski, D., Meyer, M., Stegmann, R., 2005. Fermentative
production of biohydrogen from biowaste using digested sewage sludge as inoculum. Proceedings Sardinia 2005, Tenth International
Waste Management and Landfill Symposium, S. Margherita di Pula,
Cagliari, Italy.

_ III_ Two-stage vs. single-stage fermentation process:
of energetic performances and chemical characterization Tenca A. a, Schievano A.b, Scaglia B.b, Oberti R.a, Adani F.b,*
aDipartimento di Ingegneria Agraria, Università degli Studi di Milano, Via Celoria
2, 20133 Milano, Italy.
bDipartimento di Produzione Vegetale, Università degli Studi di Milano, Via Celoria
2, 20133 Milano, Italy.
* Corresponding Author: e-mail: Keywords:
Two-stage anaerobic digestion, Hydrogen, Energy recovery, Gas
Mass Spectrometry Abstract
Two-stage anaerobic digestion (AD) is claimed to be an innovative
biological strategy to gain environmental-friendly energy vector (H2)
from waste biomasses and to improve traditional AD process in
terms of waste stabilization efficiency and net energy recovery. A
two-stage laboratory-scale CSTR digester, fed with a mixture of
agricultural and livestock residues, was successfully run for 700
hours and compared to a traditional single-stage reactor. High
hydrogen yields (140 Ndm 3 H 2 kg -1 VS-added) were reached, with subsequent methane production of 351 Ndm 3 CH 4 kg -1 VS-added. Higher CH4 yields were achieved in single stage reactor (404 Ndm 3 CH 4 kg -1 VS-added), therefore almost the same overall energy recovery was produced by the two processes (13-14 kJ kg -1 VS-added). Even slightly lower biodegradation efficiencies were shown by the two-stage
process, thus partial inhibition of the methanogenic reactor of the
two-stage process is assumed. Nevertheless the two-stage produced
biogas with high CH4 content (''70%), advantageous for lowering the
biogas upgrading cost.
Furthermore, this study also propose the GC-MS technique as a
novel diagnostic instrument to explore in depth the digestion process and the variety of biochemical reactions actually occurring into the

1. Introduction
Anaerobic digestion is one of the effective technologies used to
recover energy resources from organic wastes and it is a simple and
effective biotechnological mean of reducing and stabilizing organic
wastes (Mata-Alvarez et al., 1993; Ueno et al., 2007). This process
has been applied to an increasing number of complex feedstock like
municipal wastewater sludges, chemical and industry wastewaters
and sewage (Demirel and Yenigun, 2002).
Conventional fermentation involves different consortia of bacteria:
acidogenic bacteria break down the substrates into mainly H2, acetic
acid and CO2, while the methanogenic bacteria convert these
products to methane gas. A variety of higher organic acids, such as
propionic, butyric and lactic, as well as alcohols and ketones, are also
commonly formed during the breakdown of the organic substrates by
the acidogens, but, in a well operating process, these products are
mostly converted to acetic acid and H2 to be further consumed by
methanogenic organisms (Cooney et al., 2007; Van Ginkel et al.,
2001; Kraemer and Bagley, 2005).
Digestion process is commonly made in single-stage reactor, but
considering the delicate balance between these two groups of
microorganisms, which differ in terms of physiology, nutritional
needs, growth kinetics and sensitivity to environmental conditions
(Demirel and Yenigun 2002), a two-stage approach has been
proposed to improve the process. Splitting hydrolysis/acidogenesis
and methanogenesis and optimizing each phase, could enhance the
overall reaction rate and the biogas yield and make the process control easier (Blonskaja et al., 2003; Liu et al., 2006; Mata- Alvarez et al., 1993). The two-stage approach has also been used to produce hydrogen and
methane separately from each phase with two separate bioreactors in
series (Lee et al., 2010; Ting and Lee, 2007; Ueno et al., 2007).
Due to its high energy content (122 kJ/g), hydrogen is a promising,
clean and sustainable energy carrier and with its higher gravimetric
energy density it can be used in electrochemical (fuel cell) and
combustion processes or blended with other fuels such as methane
(CH4) to increase the combustion efficiencies (Venkata Mohan et al.,
However, typically only 15% of the energy from the organic source
is obtained from the first stage in the form of H2 (from 1 to 2 mol of
H2/mol of glucose) and this results in 80''90% of the initial chemical
oxygen demand (COD) remaining in the wastewater as volatile
organic acids (VFAs) and solvents (Das and Veziroglu, 2001;
Kraemer and Bagley, 2005; Van Ginkel and Logan, 2005). The
addition of a methanogenic reactor in series after the hydrogen-
producing reactor is one alternative promising solution for the
exploitation of the remaining 85% of the unused substrate and for
concomitant removal of organic pollutants (De Vrije and Claassen,
2005; Koutrouli et al., 2009; Logan, 2004; Ueno et al., 2007).
Some authors reported advantages of the two-stage hydrogen-
methane process over the conventional single-stage process. The
two-stage process enriches different bacteria in each anaerobic
digester (Ghosh and Class, 1978) and this brings to extend
processable waste species, in particular for the effective utilization of
high water content organic waste, to upgrade percent energy
recovery (AIST, 2004), to enhance substrate conversion producing a
lower chemical oxygen demand effluent (Azbar and Speece, 2001)
and to enhance volume reduction of wastes (Cuetos et al., 2007). The
process separation in two-stage could also reduce the overall
processing time, because the retention time of the hydrogenogenic
operation is relatively shorter than conventional methanogenic
reactors (Ueno et al., 2007), and permit to have reduced dimension of the H2-reactor if a conventional slow-rate methanogenic process is
connected as the second stage (Ueno et al., 2007). Moreover this
process could increase the stability of the overall process by
controlling the acidification phase in the first digester and hence
preventing the overloading and/or the inhibition of the methanogenic
population in the second digester (Koutrouli et al., 2009).
On the other hand the two-stage process arouses doubts as it adds to
complexity and methanogenic and hydrogenogenic baceria are both
liable to inhibition by different factors (pH, organic loading rate,
temperature) and could independently lower the whole process
For example if low cost substrates rich in carbohydrates, such as
organic wastes/wastewater or agricultural residues, are particularly
suitable for fermentative hydrogen production (Benemann, 1996;
Han and Shin, 2004; Venkata-Mohan et al. 2009), reaching the
highest amount of hydrogen per mole of substrate (de Vrije and
Claassen, 2005), the simultaneous build-up of undissociated volatile
fatty acids, the favourite substrates for methanogenesis, could deplete
the buffering capacity of the medium and decrease the pH, inhibiting
microbial growth, hydrogen production and the overall process (Van
Ginkel et al., 2001).
In this work the same feeding mixture, made up of market bio-waste
and animal slurry, was used to compare a two-stage hydrogen-
methane process with the traditional single-stage anaerobic digestion
process, both run without pH adjustment.
The performance of the two processes was evaluated not only from
the perspectives of total energy recovery and biogas production, but
also of organic matter degradation yield, digestate chemical
characterization and volatile organic compounds (VOCs) present in
the reactors gas phase. The content of soluble intermediates like
VFA reflects the changes in the metabolic process involved and may
influence the variability of hydrogen and methane production
(Venkata-Mohan et al., 2009). For the same purpose a specific GC/MS approach was adopted to describe the process through the
VOCs profile, a large group of anthropogenic (xenobiotic) or
biogenic organic compounds. Moreover, as the separation of the
acidogenesis and methanogenesis processes is said to negatively
affect syntrophic association and to prevent interspecies hydrogen
transfer (Reith et al., 2003), a microbial community analysis of
microflora of the two different reactors of the two-stage process was
performed and compared to that of the single-stage.
To our knowledge such a complete comparison between a two stage
and a single-stage fermentation system operated with the same
process parameters and without pH adjustment of culture media has
not been yet reported.

2. Materials and methods 2.1. Inocula and substrates Hydrogen-producing inoculum consisted in a digested material of a
full-scale biogas plant, treating household source-separated bio-
waste and agro-industrial by-products. Before the use the digestate
was heat-shocked at 100 ºC for 2 h in order to inactivate
hydrogenotrophic bacteria and to harvest anaerobic spore-forming
bacteria (Liu et al., 2006; Van Ginkel et al., 2005).
The same digestate, without being heat-shocked, was used as
inoculum for the methanogenic stage of both two- and single-stage
systems. After inoculation, each reactor was sparged for 30 min at
0.1 L min -1 with N 2 to remove dissolved O2 and to obtain anaerobic conditions.
The feeding substrate was represented by a mixture of swine manure
and market bio-waste. Swine manure was collected from 4 different
private farms near Milan (Italy) and then filtered through a stainless
steel sieve (US Mesh No. 10). Market bio-waste consisted of fruits
and vegetables residues was obtained from the municipal fruits and
vegetable market of Milan (Italy). Before the use, bio-waste was shredded by a blender and then stored at -20 °C. Total (TS) and
volatile solids (VS) detected for the swine manure were of 10 ± 1
and 8.1 ± 0.5 g L -1, respectively, while raw market bio-waste had a TS and VS content of 133 ± 8 and 99.8 ± 4 g L -1. Before feeding the reactors the bio-wastes were mixed with swine
manure in a 25:75 weight/weight (w/w) ratio, getting a TS content of
39.5 ± 2.5 g kg -1 (Tab.1). The average chemical oxygen demands (COD) value of the feeding solution measured over the duration of
the experiment was of 86 gO2 kg -1. Table 1 shows the characterization of the input mixture in terms of
TS, VS, COD, N-NH4 +, total nitrogen content (Khielldal method) (TKN) content, pH, total volatile fatty acids (TVFA), alkalinity
content (TALK) and their ratio (TVFA TALK -1), acetate, propionate and butyrate content, biomethane potential (Schievano et al., 2008),
and cumulative biological oxygen demand (OD20) (Schievano et al.,
2010) .

2.2. Apparatus and process operation
Three continuous flow stirred tank reactors (CSTR) were used in this
study and the reactor designs are reported in Fig. 1. The two-stage
process consisted of a 3 L hydrogen-producing reactor with 2 L
working volume (R1) and a 18 L reactor with 14.7 L working
volume for methane production (R2). Similarly, the single-stage
process consisted of a 18 L reactor with 14.7 L working volume (R3). The same feeding mixture was added twice a day both to R1 and R3 after the removal of an equal amount (measured as wet
weight) of effluent from the reactors. Peristaltic pumps (Masterflex
L/S, Cole-Parmer, Mississauga, ON, Canada) were used to supply
intermittently the feeding at the predetermined OLR and HRT and
both to transfer effluent from R1 to the methane reactor (R2) and to
remove effluents of R2 and R3 to a disposal tank. Hydraulic retention time (HRT) and the corresponding organic
loading rates were of 3, 22 and 25 days, and of 13.3, 2.3 and 1.6 gTS
L -1 day-1, for the reactor R1, R2 and R3, respectively. Low HRT for R1 was chosen on the basis of the feeding
composition, because market bio-wastes are easily hydrolysable and
rich in carbohydrate (Venkata-Mohan et al., 2009), and according to
previous experiences about optimized bio-hydrogen production from
organic waste materials reported in literature (Ueno et al., 1995,
1996; Tenca et al., 2011). Longer HRT were chosen for R2 and R3
(22 and 25 days, respectively), as methanogenic bacteria have higher
growth rates (Conklin et al., 2006).
The overall HRT of two- and single-stage processes were equal (25
d), in order to make them comparable. The three digesters were
simultaneously and continuously mixed for 15-seconds every 45-
seconds and kept at a temperature of 55 ± 2 °C via water bath
through water jackets surrounding the reactors.
During the trial period, the pH in the three reactors was not actively
controlled or adjusted and was dependent on the process natural
conditions. pH and temperature of the fermentative broth were
measured in continuous by three different InPro 3253/225/pt1000
electrodes (Mettler-Toledo international inc.). Gas flow-meters (adm
2000 model, Agilent technologies) were installed in each reactor to
record automatically the gas production. Biogas volumes were
registered as cumulated every minute and, daily, the average (over 24
h) was reported.

2.3. Analysis Total and volatile solids (TS and VS), Chemical oxygen demand
(COD), total (TKN) and ammonium nitrogen were determined
according to Standard Methods (APHA, 1998).
Biogas composition was determined using a gas chromatograph
(Agilent, Micro GC 3000A) equipped with two thermal conductivity detectors (TCD) and two different columns. Hydrogen and methane
were analyzed using a Molesieve/5A Plot column with nitrogen as
the carrier gas at a flow rate of 30 mL/min. The carbon dioxide
content was analysed using a different column (Alltech HP-PLOT U)
with helium as the carrier gas at a flow rate of 30 mL/min. The
operational temperature of the injection port was 100 °C, while that
of Molesieve/5A and PLOT U columns was maintained at 100 and
55 °C, respectively.
The analysis of volatile fatty acids in the fermentation broths were
performed using a different gas chromatograph (Varian, CP-3800)
with a flame ionization detector (FID) and a capillary column 25 m x
0.32 mm in diameter. Helium at 20 kPa pressure was used as the
carrier gas, and the temperatures of injector and FID were 220 °C
and 240 °C, respectively.
Gaseous emission produced from the three reactors were caught up
into a Nalophan TM bags (3 liters volume) connected to the reactor headspaces. Volatile organic compounds (VOCs) from gas samples
were analyzed by SPME/GC-MS. A manual SPME device and
divinylbenzene (DVB)/Carboxen/polydimethylsiloxane (PDMS) 50-
30 µm fiber - Supelco, Bellefonte, PA, USA) was used. The
compounds were adsorbed from the gas samples by exposing the
fiber, preconditioned for 3 h at 250°C as suggested by the supplier,
in Nalophan bags for 30 min at room temperature. A solution of
deuterated p-xylene in methanol was used as internal standard (IS)
for quantitative analysis. VOC analysis was performed using an
Agilent 5975C Series GC/MSD. Volatiles were separated using a
capillary column for VOC (HP 5MS, Agilent Technologies, Santa
Clara, CA, United States) of 30 m x 0.25 mm (ID) and a film
thickness of 0.25 µm. Carrier gas was helium at a flow rate of 1 ml
min -1. VOC were desorbed exposing the fiber in the GC injection port for 600 s at 250 °C. A 0.75 mm i.d. glass liner was used and the
injection port was in splitless mode. The temperature program was
isothermal for 3 min at 35 °C, raised to 200°C at a rate of 8 °C/ min. The transfer line to the mass spectrometer was maintained at 250 °C.
The mass spectra were obtained by electronic impact at 70 eV, a
multiplier voltage of 1294 V and collecting data at a m/z range of
33''300. Compounds were tentatively identified by comparing their
mass spectra with those contained in the NIST (USA) 98 library. A
semi-quantitative analysis, for all the identified compounds, was
performed by direct comparison with the internal standard. Results
were expressed as µg m -3. The biodegradability of the organic matter (OM) contained in the
digestates was determined by both short-term and long term
biological tests: the specific oxygen uptake rate (SOUR) and the
biochemical methane potential (BMP) tests, respectively.
Specific Oxygen Uptake Rate (SOUR test) for all the samples was
determined by using a standardized method reported in Schievano et
al. (2010). Briefly, the cumulative oxygen demand during 20-hours
test (OD20: gO2 kg -1 FM 20h -1) is measured in a water solution during the microbial respiration in degrading a suspended solid matrix. The
microbial respiration works out in standardized moisture conditions,
and in maximized conditions of both oxygenation and bacteria-
substrate interaction, amplifying the differences among the different
samples. This test provides a measure of the short-term
biodegradability (putrescibility) of the organic matter.
The Biochemical Methane Potential (BMP test) was performed such
as reported by Schievano et al. (2008). In brief, organic matrices
were incubated with inoculum at a ratio of 1:2 (substrate:inoculum
on a TS basis), for 60 days in batch 100-ml serum bottles under
thermophilic conditions. According to our previous approach
(Schievano et al., 2008), the test was performed under standardized
conditions and the total biogas was reported as parameter to evaluate
the organic matter performance under anaerobic condition.
Effluents from the reactors R1, R2 and R3 were sampled four times
for a period of one months (one sample/week) during steady state
processes (i.e. 700 hours). Steady states were assumed to have been reached when gas evolution rate and the concentration of H2 or CH4
(v/v) in biogas, were constant over 15 days.

3. Results 3.1. Biogas production and energetic performance The two-stage and single-stage processes were operated for about 3
month, anyway in this work we show a period of 700 h during
which, processes showed a steady state .
The biogas production during the hydrogen stage (R1) was of 3.5 L
L -1 d -1 with an hydrogen content in biogas around 45% (volume/volume) (v/v) similar to that reported in literature (Chu et
al., 2008), using food wastes to produce biogas. No CH4 was
detected during the trial (Tab. 3). Hydrogen production registered
was high (1.59 ± 0.27 LH2 L -1d-1) but the production was discontinuous showing a great variability i.e., 2.7 LH2 L -1d-1 as maximum and 0.15 LH2 L -1d-1 as minimum (Fig. 2). This discontinuity could be correlated to the semi-continuous feeding
approach and could be further reduced with continuous management
of the process.
The average hydrogen yield reached the value of 140 LH2 kg -1 VS added, remarkably higher than the results obtained by other Authors (Liu et
al., 2006) using similar wastes, i.e., 43 LH2 kg -1 VS added and 96 LH2 kg - 1 VS added, respectively. Biogas production for R2 was of 0.7 L L -1 d-1 with a methane concentration in the biogas of 68% v/v, higher than that reported for
similar digestion processes by Lane (1984). Methane produced
corresponded, as average, to 0.48 ± 0.07 LCH4 L -1d-1. On the other hand the single-stage reactor (R3) produced 1 L L -1 d-1 of biogas with an average methane percentage of 54.5%, that means a methane
production rate of 0.53 ± 0.04 LCH4 L -1 d-1. A further comparison between the two processes was performed on
an energetic basis through hydrogen and methane conversion to normalized energy units (MJ kg -1 VS added). The total energy yield was outwardly similar, i.e. 13.10 and 13.54 MJ kg -1 VS added for the two- stage and the single-stage reactor, respectively. The first-stage
influenced just for nearly the 13% of the total energy produced by
the two-stage system (Tab. 3).

3.2. Removal efficiency The AD processes determined important reductions of the TS, VS
and COD contents of the feeding mixtures (Table 4). In the two-stage
reactor VSs decreased from 854 ± 26 gVS kg -1 TS (influent) to 750 ± 81 gVS kg -1 TS (R1 effluent) and further to 605 ± 21 gVS kg -1 TS (R2 effluent), resulting in a VS reduction after first stage and after the
overall process of 25% and 63%, respectively.
The single-stage process, that received the same influent of R1,
showed a VS content in the digestate of 610 ± 11 gVS kg -1 TS, that means a VS reduction of 69%, very close to that of the two-stage
COD analyses performed on both ingestate and digestates, indicated
that 69% and 76% of COD was removed during the two and the one
stage process, respectively.

3.3. Digestate characterization
Table 4 reported the characteristics of the digested materials, as
average of the observation period. pH and VFA concentrations are
two of the main environmental factors that regulate the metabolic
pathways of anaerobic digestion (Liu et al., 2006). The feeding pH
was sub-alkaline (7.2), but due to the chosen operating parameters
and to a stable VFA production during acidogenic phase, pH values
for R1 were constantly around 5.5, that was in the optimum range for
specific hydrogen production (pH of 5.5-5.7) (Khanal et al., 2004;
Liu et al., 2006; Van Ginkel et al., 2001). Due to monomers
conversion to hydrogen, CO2 and volatile fatty acids, R1 digestate showed, for the whole period, an high total VFAs content (VFA of
3840 ± 745 mgCH3COOH kg -1), more than double than the amount in the fed material (VFA of 1600 ± 115 mgCH3COOH kg -1). In particular, acetic acid was the main VFA specie with more than 2500 mg kg -1, propionate approximately double its concentration while butyrate,
which was under detection limit in the fed mixture, showed the
highest increase reaching the average concentration of 960 mg kg -1 (Tab. 4).
On the contrary, the methanogenic reactors R2 and R3 were
characterized by high alkalinity content (5050 and 6480 mgCaCO3 kg -1, respectively) and by methanogenic bicarbonate production, which
acts to buffer the organic acids and to keep the process pH stably on
sub-alkaline vales and in the optimal range for the methanogenic
activity (Tab. 5) (Cheong, 2005; Pind et al., 2003).
In R2 the VFAs present in R1 digestate were consumed to produce
methane and CO2, as expected during the methanogenic process, and
their total content in R2 digestate was drastically reduced to the
amount of 756 ± 410 mgCH3COOH kg -1. In particular, acetate content was always below the theoretical inhibiting limits for methanogenesis (Hill, 1982), and butyrate was not found (Tab. 4).
Differently, R3 digestate showed a further lower total VFAs content
(VFA of 75 ± 40 mgCH3COOH kg -1) with acetate as the main VFA and butyrate and propionate under detection limit.
Total VFAs content and TVFA/Total alkalinity ratio are sensitive
and diagnostic parameters for system imbalance and R2 and R3
showed values of both parameters compatible with stable
methanogenic conditions and below the limits commonly indicated
for process inhibition (VFA content over 6000 mg L -1 and TVFA TALK -1 ratio over 0.4) (Chen et al., 2008; Pind et al., 2003). Notwithstanding, R2 parameters values were constantly higher than
R3 values (Table 4).
Finally, the two processes final digestates show similar TKN and
ammonia content, this latter under the inhibiting level of 3000 mg N-NH4 + kg-1 (Chen et al., 2008). The mineralization of the organic nitrogen led to the increase of the N-NH4 +/TKN * 100 ratio from the 64% for the fed mixture to over 70% for the digestates of both the
processes (Massé et al., 2007).

3.4. Volatile Organic Compounds characterization The GC-MS characterization of the biogas produced by the three
bioreactors is reported in Table 5. R1 showed a net preponderance of
carboxylic acids (over 70% of the total VOCs) followed by aromatic
compounds and alcohols (Table 5). In particular, VFAs were the 96
± 2 % of total carboxylic acids, while long chain fatty acids (LCFAs)
and other compounds represented the rest (Table 6). Among VFAs, a
severe prevalence of hexanoic acid was revealed (64 ± 10% of
tVFAs), followed by butyric and acetic acids. Aromatic compounds
detected were mainly benzene compounds (more than 50% of the
total; cymene was the predominant), plus thiazoles, thiophenes,
toluene and furans (Table 5).
R2 showed a wider variety of similarly concentrated VOCs, with a
prevalence of ketones (nearly 30% of total VOCs), carboxylic acids
(15 ± 5%) and aromatic compounds (14 ± 4%). VFAs represented
the large majority of carbossylic compounds (92 ± 3%), again with
the prevalence of hexanoic acid (nearly 60% of tVFAs) and the
presence of acetic, propionic and butyric acids. Similarly to R,
among the aromatic compounds cymene represented more than half
of them, followed by naphthalene and phenols (Table 5).
R3 showed VOCs relative content in the biogas similar to R2, except
a marked predominance of aromatic compounds (more than 30% of
total VOCs), lower concentrations of carbossylic acids and alcohols,
and LCFA slightly more concentrated than in R2 (Table 5). Once
more, hexanoic acid was the main VFA detected, followed by
propionic and butyric acids (acetic acid was undetectable; Table 5).
4. Discussion The present work demonstrates the feasibility of anaerobic digestion
separation in two stages, which allowed the simultaneous production
of hydrogen and methane.
Furthermore, the two stage process showed a higher methane content
in the biogas than that shown by the parallel single-stage process
(Tab. 3). This is probably due to the release of carbon dioxide during
the first acidogenic phase, allowing for less carbon dioxide within
the methane-rich biogas. This could bring many practical advantages
for decreasing gas conditioning requirements of methane (Azbar and
Speece, 2001).
The small contribution of the hydrogenesis stage to the total energy
yield of the two-stage reactor (13% of the total energy produced) was
in agreement with other works which demonstrated that hydrogenesis
contributes very little to the total energetic potential, and reported,
for organic wastes, 15% maximum recovery as hydrogen (Van
Ginkel and Logan, 2005).
From the energetic point of view (energy balance; Table 3) the two
systems were quite similar, with the single-stage process that
exhibited an energetic yield that was 3.9% higher than that of the
two-stage system, which was characterized by a methane production
slightly lower than single-stage reactor (R2: 0.48 LCH4 L -1 d-1; R3: 0.53 LCH4 L -1 d-1). This was in contrast with what found by other authors which showed a total methane production in a two-stage
system 19-21% higher than single-stage process (Liu et al., 2006;
Mata-Alvarez et al., 1993).
The different energy recovery between the two systems was
supported by the higher COD removal efficiency of the single-stage
process (> 6.9% with respect to R1 + R2 reactors) (Table 4). These
data lead to the hypothesis that R2 digestate was less degraded and
therefore that it could contain an unexploited energetic potential. To
prove this hypothesis two biological test (BMP and SOUR tests) were applied to the digestaes to detect the residual biodegradability
of the digestate (SOUR-OD20) and to quantify the unexpressed
energetic potential (BMP test) held in the digestates.
The BMP tests performed on R2 and R3 digestates, indicated a
higher biogas production for R2 digestate (+ 21%) than R3 digestate,
i.e., 1.89 Ndm 3 CH4 kg -1 and 1.56 Ndm3 CH4 kg -1, respectively. The digestates degradability measured under aerobic conditions
showed, as expected, a severe reduction of the easily biodegradable
fractions contained in the fed organic matter after both one and two
stages processes. More interestingly, SOUR test indicated for the R2
digestate a biodegradability of the OM double than that of R3
digestate (117 g O2 kg -1 TS for R2 versus 64 g O 2 kg -1TS for R3; Table 4). Considering that the OD20 measured the biodegradability of
easily degradable OM (Schievano et al., 2008), these results agree
with the higher TVFAs content found for R2 digestate than the R3
digestate. These results, together with high VFA content
(accumulation of product coming from the first stage high-acid
content effluent - 4000 mgCH3COOH kg -1 -) and low pH in R2 reactor seem to indicate a slowing/partial inhibition of the methanogenic
process in the two-stage system. The residual energy recovery shown
by BMP test (0.066 MJ kg -1 added) counterbalances the slightly surplus of energy produced by the single-stage when compared with the two-
stage process, such as previously discussed.
In order to better clarify the process condition, we further
investigated the process by means of GC-MS technique. As VOCs
represents the main products (VFA, ketone and alcohols) of the
anaerobic digestion, their investigation can be used to better describe
the process and to get a better understanding of it. GC/MS was used
to detect the types of VOCs present in gas sample of each reactor, as
average of 4 samples/reactor. In R1 and R2 samples 120 different
compounds were detected on average, while a statistically relevant
lower amount of compound (11 ± 6) was found for R3 (Tab. 5).
Despite the difference in the number of detected compounds, their total amounts in ppbv were similar between the three reactors. Due to
carbohydrates and proteins degradation and to lipids anaerobic
oxidation (Elefsiniotis and Oldham, 1994; Horiuchi et al., 2002), the
72.6% of the total compounds of R1 were carboxylic acids. On the
contrary, in R2 the carboxylic acids are considerably lesser (15%)
than in R1, because they were consumed by acetogens and
methanogens for methane production. Other headspace R2
compounds were ketones (29%) and aromatics (14. If compared to
R2, R3 shows a different distribution of VOCs between the classes
and the most abundant classes are aromatic compounds (34%),
ketones (26%) and siloxanes (14%).
Table 6 shows the relative composition of the carboxylic acids class
found in R1, R2 and R3. Short-chain fatty acids (VFA, from 1 to 6
atoms of carbon) constitute nearly the totality of the carboxylic acids
compounds of R1 and R2 (96% and 92% respectively) while in R3
they are just the 15% of the class. The analysis of VFA is of
particular interest, as they are important intermediate metabolites of
anaerobic digestion and known inhibitors in excess concentrations
for degradative microorganisms. Moreover, high concentration of a
specific VFA is the result of different parameters like the feedstock
composition (Dawson and Glenn, 2001), the HRT and OLR adopted
for the fermenter, the operational temperature and pH (Banerjee et
al., 1998; Rittman and McCarty, 2001; Wang et al., 2005) and
therefore could be indicative for specific metabolic pathway.
In particular, hexanoic acid was the highest amongst other acids
found in R1 and R2 samples: this data confirms the study of Levy et
al. (1981) who reported that the interruption of anaerobic digestion,
producing aliphatic acids instead of methane and CO2, brings to
higher levels of caproic acid and that caproate, ethanol and carbon
dioxide are the typical products of cellulose-rich substrates
fermentation. R1 gas sample shows a butyrate content less than half
of hexanoate and other VFAs below the 5%. Differently in R2
sample was detected just the 5% of butyrate, the 13% of propionate and the 16% of acetate, the main precursor for methane production
via anaerobic digestion (Mountfort and Asher 1978; Lata et al.
2002). On the other hand in R3 hexanoic acid, as other VFAs, were
under the detection limits of the analysis.
As for the aromatics, phenol, indole and benzene, with and without
substituents, were the most abundant compounds detected (data not
shown). Together with benzene, phenol is one of the largely present
organic compounds in the environment and it has been found
frequently in zootechnical slurries, which are the 75% (w/w) of this
study''s fed mixture. Phenols are the results of tyrosine metabolic
pathway (Jacobs et al., 1981; Spoelstra, 1978) and could be further
anaerobically degraded by microorganisms both to VFA, with
cyclohexanone, caproic acid and propionic acid production, and to
methane (Healy and Young, 1979; Fedorak and Hrudey, 1984;
Young and Rivera, 1985). Indole could be found in zootechnical
slurries, too (Ramadori and Tandoi, 1993) and is produced in rumen
and colon of monogastric by microbic degradation of tryptophan
(Yokoyama and Carlson, 1979).

5. Conclusion A two-stage laboratory-scale CSTR digester was successfully run for
700 hours: hydrogen yields averaged 140 Ndm 3 H 2 kg -1 VS-added, followed by methane production in the second stage of 351 Ndm 3 CH4 kg -1 VS-added. Differently from previous works, no energy surplus (counting the production of both H2 and CH4) was produced by the
two-stage reactor, since higher CH4 yields were achieved in single
stage reactor (404 Ndm 3 CH 4 kg -1 VS-added), therefore producing almost the same amount of overall energy (13-14 kJ kg -1 VS-added). However, biological analysis together with deeper process characterization
(GC-MS technique), suggest partial inhibition of the methanogenic
reactor of the two-stage process. Nevertheless the study support the
higher CH4 content (''70% on the total biogas volume) produced through the two-stage reactor, and invite to develop GC-MS
technique as specific diagnostic instrument for the process.

6. References AIST, 2004. World First Biogas Plant to Recover Hydrogen and
Methane Quickly from Kitchen Waste. Translation of the AIST press
released on July 14, 2004.

American Public Health Association, 1998. Standard Methods for the
Examination of Water and Wastewater, 20th ed., Washington DC,

Azbar, N.; Speece, R.; 2001. Two-phase, two-stage, and single-stage
anaerobic process comparison. J. Environ. Eng. 2001, 127, 240-247.

Banerjee, A; Elefsiniotis, P; Tuhtar, D; 1998. Effect of HRT and
temperature on acidogenesis of municipal primary sludge and
industrial wastewater. Water Sci. Technol. 38(8-9), 417-423.
Benemann J.R., 1996. Hydrogen biotechnology: Progress and
prospects. Nat. Biotechnol. 14, 1101-1103.

Blonskaja, V.; Menert, A.; Vilu, R.; 2003. Use of two-stage
anaerobic treatment for distillery waste. Adv. Environ. Res. 7(3),

Chen, Y.; Cheng, J.J.; Creamer, K.S. 2008. Inhibition of anaerobic
digestion process: a review. Biores. Technol. 99, 4044-4064.

Cheong, D.Y.; 2005. Studies of high rate anaerobic bio-conversion
technology for energy production during treatment of high strength
organic wastewaters. Ph.D. Dissertation, Utah State University,
Logan, UT.
Chu, C.F.; Li, Y.Y.; Xu, K.Q.; Ebie, Y.; Inamori, Y.; Kong, H.N.;
2008. A pH- and temperature-phased two-stage process for hydrogen
and methane production from food waste. Int. J. Hydrog. Energy,

Conklin, A.; Stensel, H.D.; Ferguson, J.; 2006. Growth Kinetics and
Competition Between Methanosarcina and Methanosaeta in
Mesophilic Anaerobic Digestion. Water Environ. Res. 78(5), 486-

Cooney, M.; Maynard, N.; Cannizzaro, C.; Benemann, J.; 2007.
Two-phase anaerobic digestion for production of hydrogen''methane
mixtures. Biores. Technol. 98, 2641-2651.

Cuetos, M.J.; Gòmez, X.; Escapa, A.; Moràn, A.; 2007. Evaluation
and simultaneous optimization of bio-hydrogen production using 32
factorial design and the desirability function. J. Power Sour., 169,

Das, D.; Veziroglu, T.N.; 2001. Hydrogen production by biological
processes: survey of literature. Int. J. Hydrog. Energy 26, 13-28.

Dawson, T.E.; Glenn, B.P.; 1996. Influence of barley, corn or oats on
batch in vitro ruminal volatile fatty acid production. J. Sci. Food

De Vrije, T.; Claassen, P.A.M.; 2005. Dark hydrogen fermentations.
In: Bio-methane & Bio-hydrogen. Status and perspectives of
biological methane and hydrogen production. Edited by Reith J.H.,
Wijffels R.H., Barten H., pp 103-123. Dutch Biological Hydrogen
Foundation Publishing, Petten, The Netherlands.

Demirel, B.; Yenigun, O.; 2002. Two-phase anaerobic digestion
processes: a review. J. Chem. Technol. Biotechnol. 77, 743-755.

Elefsiniotis, P., Oldham, WK; 1994. Influence of pH on the
acidphase anaerobic digestion of primary sludge. J. Chem. Technol.
Biotechnol. 60, 89-96.
Fedorak, P.M.; Hrudey, S.E.; 1984. The effects of phenol and some
alkyl phenolics on batch anaerobic methanogenesis. Water Res.
18(3), 361-367.

Ghosh, S.; Klass, D.L.; 1978. Two-phase anaerobic digestion.
Process Biochem. 1978, 13, 15-24.

Han S.K., Shin H.S. 2004a. Biohydrogen production by
anaerobic fermentation of food waste. Int. J. Hydrog. Energy
29(6), 569-577.

Healy, J.B.J.; Young, L.Y.; 1979. Anaerobic biodegradation of
eleven aromatic compounds to methane. Appl. Environ. Microbiol.
38, 84-89.

Hill, D.T.; 1982. A comprehensive dynamic model for animal waste
methanogenesis. Transact. ASAF 25, 1374-1380.

Horiuchi, J.I.; Shimizu, T.; Tada, K.; Kanno, T.; Kobayashi, M.;
2002. Selective production of organic acids in anaerobic acid reactor
by pH control. Biores. Technol. 82, 209-213.

Jacobs, L.W.; Phillips, J.H.; Zabik, M.J.; 1981. Final report to
Michigan Department of Natural Resources, Lansing, Mich., USA.

Khanal, S.; Chen, W.H.; Li, L.; Sung, S.; 2004. Biological hydrogen
production: effects of pH and intermediate products. Int. J. Hydrogen
Energy 29, 1123-1131.

Koutrouli, H.C.; Kalfas, H.; Gavala, H.N.; Skiadas, I.V.;
Stamatelatou, K.; Lyberatos, G.; 2009. Hydrogen and methane
production through two-stage mesophilic anaerobic digestion of
olive pulp. Bioresour. Technol. 100(15), 3718-3723.
Kraemer, J.T; Bagley, D.M; 2005. Continuous Fermentative
Hydrogen Production Using a Two-Phase Reactor System with
Recycle. Environ. Sci. Technol. 39, 3819-3825.

Lata, K.; Rajeshwari, K.V.; Pant, D.C.; Kishore, V.V.N.; 2002.
Volatile fatty acid production during anaerobic mesophilic digestion
of tea and vegetable market wastes. J. Microbiol. Biotechnol. 18,589-

Lay, J.J.; Lee, Y.J.; Noike, T.; 1999. Feasibility of biological
hydrogen production from organic fraction of municipal solid waste.
Water Res. 33(11), 2579-2586.

Lee, Y.; Chung J.; 2010. Bioproduction of hydrogen from food waste
by pilot-scale combined hydrogen/methane fermentation. Int. J.
Hydrog. Energy 35(21), 11746-11755.

Levy, PF; Sanderson, JE; Kispert, RG; Wise, DL; 1981. Biorefining
of biomass to liquid fuels and organic chemicals. Enzyme Microb.
Technol. 3,207-215.

Liu, D.W.; Liu, D.P.; Zeng, R.J.; Angelidaki, I.; 2006. Hydrogen and
methane production from household solid waste in the two-stage
fermentation process. Water Res. 40(11), 2230-2236.

Logan, B.E.; 2004. Biologically extracting energy from wastewater:
biohydrogen production and microbial fuel cells. Environ. Sci.
Technol. 38, 160A''167A.

Mata-Alvarez, J.; Mtz.-Viturtia, A., Llabrés-Luengo, P., Cecchi, F.;
1993. Kinetic and performance study of a batch two-phase anaerobic
digestion of fruit and vegetable wastes. Biomass Bioenergv 5(6),

Mountfort, D.O.; Asher, R.A.; 1978. Changes in the proportions of
acetate and carbon dioxide used as methane precursors during
anaerobic digestion of bovine waste. Appl. Environ. Microbiol. 35,
Owen, M.F.; Stuckey, D.C.; Healy, J.B.; Young, L.Y., McCarthy,
P.L.; 1979. Bioassay for monitoring biochemical methane potential
and anaerobic toxicity. Wat. Res. 13, 485-492.

Pind, P.F.; Angelidaki, I.; Ahring, B.K.; 2003. Dynamics of the
Anaerobic Process: Effects of volatile fatty acids. Biotechnol.
Bioengineer. 2003, 82 (7), 791-801.

Ramadori, R.; Tandoi, V.; 1993. Depurazione biologica delle acque
di scarico. In: Ecologia Applicata. Edited by R. Marchetti; Città
Studi, Milano, ITALY.

Reith, J.H.; Wijffels, R.H.; Barten, H.; 2005. Introduction: the
perspectives of biological methane and hydrogen production. In:
Bio-methane & Bio-hydrogen. Status and perspectives of biological
methane and hydrogen production. Edited by Reith J.H., Wijffels
R.H. and Barten H., pp 103-123. Dutch Biological Hydrogen
Foundation Publishing, Petten, The Netherlands.

Rittman, B.E.; McCarty, P.L.; 2001. Environmental Biotechnology:
Principles and Applications. McGraw Hill Inc., New York, USA.

Schievano, A.; Pognani, M.; D´Imporzano, G., Adani, F.; 2008.
Predicting anaerobic biogasification potential of ingestates and
digestates of a full-scale biogas plant using chemical and biological
parameters. Bioresour. Technol. 99, 8112-8117.

Schievano, A.; Scaglia, B.; D´Imporzano, G.; Malagutti, L.; Gozzi,
A.; Adani, F; 2010. Prediction of biogas potentials using quick
laboratory analyses: Upgrading previous models for application to
heterogeneous organic matrices. Bioresour. Technol. 100, 5777-

Spoelstra, S.F.; 1978. Degradation of Tyrosine in Anaerobically
Stored Piggery Wastes and in Pig Feces. Appl. Environ. Microbiol.,
Tenca, A.; Schievano, S.; Perazzolo, F.; Adani, F.; Oberti, R.; 2011.
Biohydrogen from thermophilic co-fermentation of swine manure
with fruit and vegetable waste: Maximizing stable production
without pH control. Biores. Technol. 102, 8582-8588.

Ting, C.H.; Lee, D.J.; 2007. Production of hydrogen and methane
from wastewater sludge using anaerobic fermentation. Int. J. Hydrog.
Energy 32, 677-682.

Ueno, Y.; Kawai, T; Sato, S.; Otsuka, S.; Morimoto, M.; 1995.
Biological Production of Hydrogen from Cellulose by Natural
Anaerobic Microflora. J. Ferment. Bioeng. 79(4), 395-397.

Ueno, Y.; Otsuka, S.; Morimoto, M.; 1996. Hydrogen Production
from Industrial Wastewater by Anaerobic Microflora in Chemostate
culture. J. Ferment. Bioeng 82( 2), 194-197.

Ueno, Y.; Fukui, H.; Goto, M.; 2007. Operation of a Two-Stage
fermentation process producing hydrogen and methane from organic
waste. Environ. Sci. Technol. 41, 1413-1419.

Van Ginkel, S.; Logan, B.; 2005. Increased biological hydrogen
production with reduced organic loading. Wat. Res. 39, 3819-3826.

Van Ginkel, S.; Sung, S.; Lay, J.J., 2001. Biohydrogen production as
a function of pH and substrate concentration. Environ. Sci. Technol.
35, 4726-4730.

Van Ginkel, S.; Oh, S.E.; Logan, B.E.; 2005. Biohydrogen
production from food processing and domestic wastewaters. Int. J.
Hydrog. Energy 30, 1535-1542.

Venkata Mohan, S.; Mohanakrishna, G., Kannaiah Goud, R.; Sarma,
P.N., 2009. Acidogenic fermentation of vegetable based market
waste to harness biohydrogen with simultaneous stabilization.
Biores. Technol. 100, 3061-3068.
Wang, G.; Mu, Y.; Yu, H.Q.; 2005. Response surface analysis to
evaluate the influence of pH, temperature and substrate concentration
on the acidogenesis of sucrose-rich wastewater. Biochem. Eng. J.
23(2), 175-184.

Yokoyama, M.T.; Carlson, J.R.; 1979. Microbial metabolites of
tryptophan in the intestinal tract with special reference to skatole.
Am. J. Clin.Nutr. 32, 173-178.

Young, L.Y.; Rivera, M.D.; 1985. Methanogenic degradation of four
phenolic compounds. Water Res. 19, 1325-1332.

Tables and Figures Fig. 1 - a) Schematic diagram of two-stage (a) and single stage (b) process. (1-Hydrogen
reactor, 2-Effluent bottle, 3-Methane reactor, 4-Gas meter and counter, 5-Mixer).

Tab. 1 - Characteristics of the input mixture. TS g kg -1 39.5 ± 2.5 VS g kg -1 TS 854 ± 26 COD gO2 kg -1 85.9 ± 8.4 TKN g kg -1 2.4 ± 1.1 NH4 +-N g kg -1 1.5 ± 0.7 NH4 +-N TKN-1 % 64 ± 4 pH 7.2 ± 0.1 TVFA mgCH3COOH kg -1 1600 ± 115 Acetic acid mg kg -1 675 ± 520 Propionic acid mg kg -1 172 ± 145 Butirric acid mg kg -1 u.d.l.* TALK mgCaCO3 kg -1 4708 ± 10 TVFA TALK -1 kgCH3COOH kg -1 CaCO3 0.32 BMP N dm 3 CH 4 kg -1 TS 388.1 ± 47.0 Ndm 3 CH 4 kg -1 15.53 ± 1.88 OD20 gO2 kg C -1 org h -1 297.85 ± 16.06 SOUR mgO2 g C -1 org 54.61 ± 8.6 *u.d.l. = under detection limit

Tab. 2 - Process Parameters. Single-stage R1 R2 R3 V working L 2.3 14.7 14.7 V headspace L 1.1 5 5 Hydraulic Retention Time (HRT) d 3 22 25 Operational Temperature °C 55 ± 2 55 ± 2 55 ± 2 TSin g L -1 40 ± 1 50 ± 15.7 40 ± 1 Organic Loading Rate (OLR as TS ) gTS L -1 d -1 13.3 2.3 1.6 Two-stage
Tab. 3 - Biogas, biohydrogen and biomethane production rates and yields for the two-stage
and single-stage processes. Single-stage R1 R2 R3 Volumetric biogas production rate Nm 3 m3dig d-1 3.5 ± 0.58 0.7 ± 0.1 1.00 ± 0.08 Volumetric hydrogen production rate Nm 3 H2 m 3dig d-1 1.59 ± 0.27 0 0 Volumetric methane production rate Nm 3 CH4 m 3dig d-1 0 0.48 ± 0.07 0.53 ± 0.04 Hydrogen content in biogas % 44.9 ± 5.5 0 0 Methane content in biogas % 0 68.2 ± 1.7 54.5 ± 1.9 Specific hydrogen or methane production on
VS basis Ndm 3 H2/CH4 kg -1 VS added 140.00 324.90 387.90 Volumetric energy production kJ L -1 d-1 19.0 ± 4.9 17.4 ± 2.2 18.5 ± 1.3 Energetic yield MJ kg -1 VS added 1.67 ± 0.37 11.42 ± 1.19 13.54 ± 0.84 Specific total energy production on TS basis MJ kg -1 VS added 13.54 ± 0.84 13.10 ± 1.24 stage reactors Two-stage
Tab. 4 -
Characterization of the digestates of the three reactors observed (R1, R2, R3). Data
reported as average of 4 samples during the observation period. Single-stage R1 R2 R3 TS g kg -1 34.0 ± 14.7 21.9 ± 3 17.5 ± 3 VS g kg -1 TS 750 ± 81 605 ± 21 610 ± 11 COD gO2 kg -1 58.3 ± 7.4 26.8 ± 3.2 20.9 ± 3.2 TKN g kg -1 2.85 ± 0.5 2.2 ± 0.3 1.9 ± 0.2 NH4 +-N g kg -1 1.7 ± 0.3 1.6 ± 0.2 1.4 ± 0.1 NH4 +-N TKN-1 % 60 ± 4 72 ± 3 72 ± 6 pH 5.52 ± 0.14 7.61 ± 0.06 7.94 ± 0.25 TVFA mgCH3COOH kg -1 3840 ± 745 756 ± 410 75 ± 40 Acetic acid mg kg -1 2511 ± 345 455 ± 155 53 ± 35 Propionic acid mg kg -1 318 ± 140 299 ± 120 u.d.l.* Butirric acid mg kg -1 958 ± 270 u.d.l.* u.d.l.* TALK mgCaCO3 kg -1 4050 ± 980 5050 ± 430 6480 ± 976 TVFA TALK -1 kgCH3COOH kg -1 CaCO3 0.95 ± 0.06 0.15 ± 0.03 0.01 ± 0.01 BMP Ndm 3 CH 4 kg -1 TS 318.9 ± 61.5 101.7 ± 8.1 102.8 ± 9.3 Ndm 3 CH 4 kg -1 15.27 ± 2.71 1.89 ± 0.05 1.56 ± 0.04 OD20 gO2 kg C -1 org h -1 270 ± 111 111 ± 40 64 ± 22 SOUR mgO2 g C -1 org 91 ± 33 19 ± 9 10 ± 1 *u.d.l. = under detection limit Two-stage Tab. 5 - Relative area counts (in %) for grouped VOCs found in the reactors headspaces on the
total compounds amount in ppbv. Single-stage R1 R2 R3 Alcohols 4.6 ± 2.4a 5.0 ± 4.6a 0.9 ± 0.9a Aldehydes 1.1 ± 0.7a 1.9 ± 0.5a u.d.l.* Alkanes 3.5 ± 2.4a 7.2 ± 3.7a 7.2 ± 3.2a Alkenes 0.7 ± 0.5a 2.2 ± 0.8b 1.5 ± 1.4a Aromatic compounds 7.5 ± 1.1a 13.6 ± 2.9a 33.6 ± 6.7b Carboxylic acids 72.6 ± 2.9c 15.0 ± 4.9b 7.3 ± 5.4a Cycloalkanes 0.2 ± 0.1a 1.3 ± 1.1a 2.3 ± 1.9a Cycloalkenes 0.1 ± 0.1a 1.3 ± 0.7b u.d.l.* Esters 1.2 ± 1.0a u.d.l.* u.d.l.* Ethers 0.2 ± 0.2a 1.1 ± 1.1a u.d.l.* Halogenated compounds 0.2 ± 0.2a 0.5 ± 0.1a u.d.l.* Ketones 1.2 ± 0.2a 29.3 ± 6.5b 25.6 ± 8.1b Nitrogen compounds 1.1 ± 0.9a 3.9 ± 2.0a 3.9 ± 3.0a Siloxanes 2.0 ± 1.6a 8.9 ± 0.9ab 14.0 ± 9.5b Sulphur compounds 0.7 ± 0.7a 1.6 ± 1.9a u.d.l.* Terpenes 3.0 ± 0.9a 8.8 ± 3.0b 7.3 ± 2.7ab Total compounds amount (ppbv) 27167 ± 6332a 23510 ± 10415a 23353 ± 15168a Total compounds number 123 ± 3b 119 ± 37b 11 ± 6a * u.d.l. = under detection limit a number followed by the same letter in the same column are not statistically different (Test Tukey,
p<0.05) Two-stage

Tab. 6 -
VFA, with their relative partition into single species, and other carboxylic acids
compounds percentage amount on the total carboxylic acids. Single-stage R1 R2 R3 Volatile Fatty Acids 96 ± 2 92 ± 3 15 ± 3 Formic acid u.d.l.* u.d.l.* u.d.l.* Acetic acid 5 ± 2 16 ± 8 u.d.l.* Propionic acid 2 ± 0.5 13 ± 11 u.d.l.* Butyric acid 25 ± 9 5 ± 5 u.d.l.* Pentanoic acid 3 ± 3 u.d.l.* u.d.l.* Hexanoic acid 64 ± 10 58 ± 26 u.d.l.* Others 4 ± 2 8 ± 3 85 ± 2 Two-stage

0% 10% 20% 30% 40% 50% 60% 70% 80% 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 0 100 200 300 400 500 600 700 H y d ro g e n c o n te n t in b io g a s P ro d u c ti o n ( N d m 3 _ H 2 /L _ d ig d ) Tempo (h) Average H2 production % H2 0% 10% 20% 30% 40% 50% 60% 70% 80% 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 0 100 200 300 400 500 600 700 M e th a n e c o n te n t in b io g a s P ro d u c ti o n ( N d m 3 _ C H 4 /L _ d ig d ) Time (h) Average CH4 pro d uction % CH4 0% 10% 20% 30% 40% 50% 60% 70% 80% 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 0 100 200 300 400 500 600 700 M e th a n e c o n te n t in b io g a s P ro d u c tio n ( N d m 3 _ C H 4 /L _ d ig d ) Time (h) Average CH4 production % CH4 Fig. 2 - Biogas production and its composition in hydrogen (for R1) and methane (for R2 and
R3), monitored during the reactors operation (700 h) for R1 (a), R2 (b) and R3 (c). _ IV_ MEC and Anaerobic Digestion performance
comparison with complex industrial wastewater
Tenca A. a, Cusick R.b, Logan B.E.b,* a Department of Agricultural Engineering, Università degli Studi di Milano, Via
Celoria 2, 20133 Milano, Italy
b Department of Civil and Environmental Engineering, The Pennsylvania State
University, University Park, 212 Sackett Building, PA 16802, USA

* Corresponding author. Tel.: þ1 814 863 7908; fax: þ1 814 863 7304.
E-mail address: (B.E. Logan). To be submitted to: Biotechnology & Bioengineering Highlights + MEC and Anaerobic Digestion comparison + Methanol exploitation + Molybdenum disulfide as valid alternative catalyst Abstract Microbial electrolysis cells (MECs) are a novel bio-technology that
can be used to recover energy as hydrogen from organic matter.
Through MEC membraneless architecture and unbuffered wastewater use as feeding substrate, methane production was
allowed in MEC and organic removal efficiencies and the rate and
yield of biogas produced by MEC were compared with simulated
Anaerobic Digestion (AD) process. Both the process were fed with
an actual industrial wastewater with high methanol content. MEC
energy recovery was positive (3.76 and 3.38 kWh/kg TCOD removed with
platinum and molybdenum disulfide cathode, respectively) and 14-
16% higher than that of AD simulation. Also the MEC removal
efficiency was high (85-87% TCOD removal, up to 20% higher than in AD) with complete degradation of methanol, a chemical never
reported to be exploited in MEC. MEC emerged to be competitive
with the AD process, especially using cheaper alternative catalysts
such as molybdenum disulfide (MoS2).

Keywords Anaerobic Digestion; MEC; Methanol; Wastewater; Molybdenum
disulfide 1. Introduction Due to depletion of oil reservoirs and climate changes, it is currently
of major concern to develop sustainable, environment-friendly and
worldwide competitive new technologies for energy production.
The exploitation of biomass, in the case of organic wastes, allow to
extract bioenergy and/or biochemicals while treating wastes: in
particular, industrial and agricultural wastewaters are ideal
candidates because they contain high levels of easily degradable
organic material, which results in a net positive energy or economic
balance (Angenent et al. 2004).
Several strategies can be used for bioprocessing, including
methanogenic anaerobic digestion, biological hydrogen production,
ethanol fermentation and fermentation for production of high-value
by-products (Angenent et al. 2004, Pham et al. 2006). Recently,
bioelectrochemical systems (BESs) for producing hydrogen and
electricity have also been developed as a novel biotechnology to
harvest energy from soluble biomass. BESs offer the option of
electricity production in microbial fuel cells (MFCs) or hydrogen
production in microbial electrolysis cells (MECs) (Pham et al. 2006;
Thygesen et al. 2010).
Each of the above-mentioned bioprocessing technologies shows
complexities and issues that make them advantageous over the other
processes in specific fields. Therefore, these bioprocesses are often combined in parallel with others to maximize the overall energy
output (Thygesen et al. 2010) or presented as alternative to the
It is well known that anaerobic digestion (AD) is a mature process
that allows for the intake of both high and low concentration COD
biomass: carbohydrates are particularly well suited, but almost any
type of bioavailable substrate is exploitable by AD. Moreover, AD is
characterized by high organic removal rates, low energy-input
requirements, low sludge production, with final production of energy
in the form of methane (Angenent et al. 2004). However, despite
AD is a well established technology, it requires meso- to thermo-
philic temperatures to achieve sufficient turnover and limited
methane solubility and it produces methane which needs to be
upgraded from a biogas mix, full of undesirable compounds like
H2S, siloxanes, etc.
Among the BESs, both in MFCs and MECs anode respiring bacteria
oxidize organic compounds with the anode electrode acting as
electron acceptor. In a MEC the electrons produced are later
consumed at the cathode by endothermal reduction of H + to H 2, thus requiring an addition of power from an external electrical source
(Thygesen et al. 2010).
MECs show high hydrogen yields with high efficiency relative to the
electrical input (up to over 400%; Lalaurette et al. 2009), which is
typically between +0.2 and +0.9 V applied, an amount much less
than that used for water electrolysis (>1.6''1.8 V applied) (Call and
Logan, 2008). Moreover, MECs have been proved to efficiently
convert into hydrogen a wide range of simple organic materials (such
as acetic acid, butyric acid, lactic acid, glucose and cellulose), even if
few tests have been conducted using complex mixtures of substrates
or actual wastewaters (Lalaurette et al. 2009; Wagner et al. 2009;
Cusick et al. 2010). While treating the wastewaters and providing
energy in the form of hydrogen, this technology can also reduce
solids production and sludge handling cost, and can possibly limit the release of odors (Logan et al. 2008). Due to their similarity with
MFC, it is also possible to assume that higher performance could be
reached with MEC treating organic matters with readily available
soluble COD, also occurring at ambient temperature (25 °C or less)
where anaerobic digestion generally fails due to low reaction rates
and high solubility of the methane produced (Pham et al. 2006).
On the other hand, the relatively low power density of MECs, their
architecture and especially the requirement for expensive noble
metals in the electrodes (typically platinum) are the major limitations
that make this technology not mature for real-scale applications.
Aiming at minimizing the costs and at scaling up of the technology
to real scale, the ion-selective membranes, which are typically used
to obtain relative pure cathodic hydrogen gas, could be removed.
Anyway different researchers (Call and Logan 2008; Wagner et al.
2009) reported that in membraneless MECs the reduction of ohmic
voltage loss in the cell and of a bulk pH gradient in the liquid, is
accompanied by the production of appreciable amount of methane
gas. Clauwaert and Verstraete (2008) demonstrated that methanogenesis can easily become dominant in membraneless MECs
and that the 65 ± 13% of the influent acetate removed was
transformed in methane. In addition, methane production in MEC is
more abundant with the use of wastewaters as feeding, as they
contain endogenous hydrogen-consuming microorganisms, further
favored by relatively long operation cycles for slow degradation of
complex organic matter (Logan et al. 2008).
Rather than to consider CH4 as an undesired by-product, methane
production in a MEC could be a valuable result. Indeed, MECs
methane production may be stronger than hydrogen production
(Clauwaert and Verstraete 2008) and compared to anaerobic
digesters MECs don't need to use some of the produced energy for
heating the reactor to the temperatures needed for efficient
methanogenesis (Logan et al. 2008). Therefore, it must be evaluated
if the process performance and the value of the gas produced can compensate for the electrical energy requirements and the more
complex design of MECs compared to ADs.
In the present study instead of searching for strategies to inhibit the
methanogenesis in membraneless MECs (Call and Logan 2008;
Rozendal et al. 2008), their biogas, and especially methane,
production and wastewater treatment efficiency were monitored.
Process performances during consecutive fed-batch cycles were
compared in membraneless MECs operated in presence of an
external applied cell voltage and with no voltage applied. The latter
condition intended to simulate an anaerobic digestion process.
Furthermore, as most MEC researches have been done using
Platinum as catalyst, which accounts for the greatest percentage of
the cost of the MEC and can also be negatively affected by
components often present in waste streams (Jeremiasse 2010),
alternative catalysts are needed, especially for minimizing real scale
application cost. Two alternatives (stainless steel and MoS2) have
been used and compared with Pt performance in this study.
Finally, high-COD industrial wastewater, with a significant content
of methanol, was used as feeding substrate. Methanol exploitation in
MEC has never been previously reported.

2. Materials and methods 2.1 Wastewater Industrial wastewater was collected from the wastewater treatment
system at Allentown, USA. Samples were placed on ice and shipped
overnight to the laboratory and stored at 4° C. Wastewater served
both as inoculum and substrate in all experiments and its full
characterization is shown in Table 1. 2.2 Reactors construction and operation A common practice for enriching a bacterial community in a MEC is
to operate a MFC for several fed-batch cycles and then transfer the anode into a MEC. This procedure ensures biofilm formation on the
anode and preselects an exoelectrogenic community for subsequent
MEC operation (Logan et al. 2008). To this aim, six single-chamber
MFCs were built, having a cylindrical chamber 4 cm long by 3 cm in
diameter (empty volume = 28 mL) and an anaerobic culture tube
glued to fit on top of the reactor (1.6 cm inner diameter and 6 cm
length; 12 mL capacity). The adopted anodes were heat-pretreated
graphite fiber brush electrodes (PANEX 33 160K, Gordon Brush,
OD = 2.5 cm, L = 2.5 cm), while cathodes were flat carbon cloth
(Type B-1B, E-TEK, 3.8 cm diameter) added with a Pt catalyst (10%
Pt/C) on the anode-facing side of the electrode (Call and Logan
2008). Cathodes used in MFCs were also treated to have four
diffusion layers (PTFE) applied to their airfacing side (Cheng et al.
The six reactors were operated as MFCs for two months (details
given in supporting information) in order to enrich biofilm on the
anodes surface, with consequent current production, directly using
the wastewater without any dilution or amendment.
After evidence of relatively stable exoelectrogenic activity, the
reactors were converted to operate as MECs as described by Call and
Logan (2008), both by replacing the cathodes and by covering them
with a plate to exclude air and eliminating the oxygen reduction at
the cathode.
Three different cathode types were used, in order to compare the
performance of different catalysts: Platinum (Pt), Molybdenum
disulfide (MoS2) and stainless steel (SS). All MEC cathodes had a
total surface area of 12 cm2, with only 7 cm2 exposed to the
solution, and no diffusion layers.
Carbon cloth cathodes with a Pt catalyst were constructed using a
mixture 10/90 of platinum powder and carbon black (E-TEK, C1-10,
10 wt.% Pt on Vulcan XC-72).
Carbon cloth cathodes with MoS2 catalyst were prepared using a
MoS2 powder (Aldrich, 99%, particle size < 2 mm), mixed with 5 mg/cm 2 carbon black (Cabot, VULCAN XC-72R, > 99%) at mass loading ratios of 33.3% and with 50 µL/cm 2 of a 2:1 volume solution of Nafion polymer (Aldrich, 5 wt%) and iso-propanol. After
vortexing the mixture for 15 s at 3200 rpm (VWR Vortex Mixer) this
was applied just to the solution side of the carbon cloth (E-TEK, B-
1/B/30WP, 30% by weight PTFE Wet-Proofed).
For SS cathodes, flat sheets of type 304 stainless steel (Trinity Brand
Industries, Inc.) were sanded smooth with silicon carbide sand paper,
then ultrasonically washed in deionized water and rinsed with
acetone, with a final rinse in deionized water and drying overnight
(>12 h) before testing.
A positive voltage (Eap) of 0.7 V was applied to the MECs by
connecting the negative pole of an external power supply (model
3646A; Circuit specialists, Inc.) to a resistor (10') and then to the
cathode, and the positive pole to the anode. Instead, for anaerobic
digestion simulation, no voltage was applied (Open Circuit Voltage)
to the cells (OCVCs). When the complete gas production cycle
ended, as indicated by zero gas production rate for one hour or more,
both the MECs and the OCVCs were drained, refilled with fresh
substrate, and flushed with ultra high-purity nitrogen gas (99.998%)
for 15 minutes.
All the tests were run in fed batch mode with duplicate reactors at
30° C in a constant temperature room.

2.3. Measurements and chemical analyses Reactor voltage was measured across an external resistor (MFC: Rex
= 1 k', MEC: Rex = 10 ') every 20 min using a multimeter (2700;
Keithley, United States) connected to a personal computer. Current
and power generation were calculated using I = E/R and P = IE
respectively, where I (A) is the current, P (W) the power, E (V) the
voltage and R (') the resistance. During each fed-batch cycle the volume of gas produced was
recorded using a respirometer (AER-200; Challenge Environmental)
and the evolved gas was collected in air-tight gas bags (0.2 L
capacity; Cali-5-Bond, Calibrated Instruments, Inc.). Gas from the
gas bag and the reactor headspace was sampled using a gas-tight
syringe (200 µL injection volume) and analyzed using gas
chromatography (Models 310 & 8610B; SRI Instruments, Torrence,
The influent and the effluent of the MECs were characterized for
each batch cycle. The concentrations of solvents, alcohols, and
organic acids (acetone, methanol, ethanol, propanol, butanol, acetate,
propionate, and butyrate) were measured by gas chromatography
(Varian Star 3400) with injector and flame ionization detector
temperatures of 250 °C. Total and soluble COD were quantified
through HACH method 8000 (HACH COD system, HACH
Company, Loveland, CO). Probes were used to measure pH (Mettler
Toledo Seven Multi; Model: pH; S/N: 290843) and conductivity
(Mettler Toledo Seven Multi; Model: Cond.; S/N: 291048).

2.4. Calculations Biogas Production Rate
The volumetric production rate, Q (m 3 gas/m3reactor d), of both H 2, CH4 and CO2 was calculated based on the measured specific gas
produced normalized to the reactor volume. TCOD Removal
The ability of the process to be a feasible treatment technology was
expressed both as the total COD removal (% of the TCOD of the
feeding wastewater which has been removed at the end of the batch
cycle) and as the TCOD removal rate, r TCOD (kg TCOD removed/m 3reactor d). Hydrogen and Methane Yield
Being the wastewater used for the experiments a complex source of
organic matter, the yield (Y) of hydrogen and methane was
expressed on the basis of the COD removal (Nm 3 gas/kg TCOD removed) and calculated as = '' where Vgas is the total volume of a specific gas, hydrogen or
methane, produced, Vr the reactor volume and CODi and CODe the
COD concentrations of the wastewater at the beginning and end of
the batch test.
COD removal was calculated as the average from two to five batch
tests. Energy Recovery
Energy yield in the MEC or in OCVC was calculated as energy
production per unit of reactor volume (kWh/m 3) normalized to COD removal (kg TCOD/m 3). Energy recovered as gas was determined by multiplying the moles of
hydrogen and methane produced by their lower heat of combustion
(-286 and -891 kJ/mol respectively) and converting the value to
express the yield as kWh/kg TCOD removed.
As for the MECs, the net energy recovery (kWh/ kg TCOD removed) of
the process was calculated by subtracting the supplemental energy
required and added to overcome the potential for hydrogen evolution
from the value of energy produced as gas during each MEC batch-
cycle. The energy required by MEC system (Wap) is obtained by
converting the recorded voltage added by the power source (Eap)
taken over time intervals ''t = 20 min for n intervals as followed = '' 1 '' 1000 '' 3. Results 3.1 Reactors performance in MFC mode Four representative cycles of current generation for four cells later
converted to MEC are shown in Fig. S.1. Maximal voltage and
current density reached were 420 mV (1 k') and 0.59 A/m 2, respectively, similar to the results of other studies with actual
wastewaters; for instance winery wastewater with high TCOD (2200
± 510 mg/L) produced 441 ± 17 mV (1 k') (Cusick et al., 2010).
This shows that the wastewater provided a good source of
exoelectrogenic bacteria and that their enrichment on the anode was
During the experiments all the cells showed a similar gradual
decrease both for the batch cycle time and the maximal current
density reached, decreasing from 6 to 4 days and from 0.60 to 0.46
A/m 2, respectively (Fig. S.1). This, despite the coulombic efficiency remained stably low (7 ± 2%), but within a range typical for MFC
treating wastewaters (from 5 to 20%; Min and Logan, 2004;
Heilmann and Logan, 2006; Min et al. 2005), as well as the TCOD
removal resulted stably high (90%). Again, both the batch cycle
time and the TCOD removal are similar to those achieved by Cusick
et al. (2010) for winery wastewater, who reported a COD removal of
83 ± 10% in 6 days.
Assuming that the double-peak profile in the current production (Fig.
S.1) is indicating a different kinetics in the degradation of different
organic fractions, both TCOD removal and VFAs and alcohols
content in the liquid solution were analyzed after the end of the first
peak of current production ('' 2.5 days) and at the end of the cycle
(Tab. S.1). Anyway, the more abundant compounds in the
wastewater, i.e. methanol and acetate, 1.54 and 0.18 g/L respectively,
were equally degraded after the first current peak (78% and 75%,
respectively) and totally degraded at the end of the cycle (Tab. S.1).
3.2 MEC performance and comparison with OCV MEC duplicates exhibited reproducible results throughout the
experiments in terms of both current generation and biogas
production (Fig. S2).
At an applied voltage (Eap) of + 0.7 V, the amount of electrical
current produced in MEC resulted to depend on the cathode type: Pt
reached the maximal current density of 2.07 A/m 2, higher than MoS 2 and SS catalysts, with 1.4 and 0.95 A/m 2, respectively. Among the MECs, Pt-MEC produced also biogas at the highest average flow
rate (1.71 m 3 gas/m3reactor d), with the highest average H 2 content (32 ± 4%). MoS2-MEC and SS-MEC had lower performances both
in terms of flow rate and of hydrogen content (MoS2: 1.18 m 3 gas/m 3reactor d, 25 ± 5 % H 2; SS: 0.83 m 3 gas/m3reactor d, 16 ± 1 % H2) (Fig. 1). Methane was always found in the biogas, with high
relative amount of 55 ± 4 %, 62 ± 3 % and 70 ± 2 % for Pt, MoS2
and SS catalysts, respectively.
Coulombic Efficiency is low even if compared to other studies with
actual wastewaters (Ditzig et al. 2007; Wagner et al. 2009; Cusick et
al. 2010): Pt showed the same CE as in MFC mode (7 ± 1%), while
MoS2 and SS reached minimal higher values of 10% and 12.5%,
Finally, Pt had the shorter cycle time (31 hours), while MECs with
MoS2 and SS as catalysts showed similar and longer fed-batch cycle
time (37.5 and 45.5 hours, respectively).
Comparing the biogas production in MEC with that in OCV,
differences both for the total amount of biogas produced, production
flow rate and gas composition were detected. According to the data
averaged over different batch cycles, biogas produced by MEC
represented 156%, 155% and 116% of the total amount of biogas
produced in OCV with the same catalysts, Pt, MoS2 and SS
respectively. Considering the production rate (Q), both Pt-MEC and MoS2-MEC
had higher flow rate than their corresponding OCV, while due to its
longer batch cycle time SS-MEC had a lower Q than without added
voltage (Fig. 1). Lower amounts of hydrogen (Pt: 26 ± 3%; MoS2: 21
± 6%; SS: 14 ± 1%) and higher of methane (Pt: 65 ± 3%; MoS2: 69 ±
7%; SS: 74 ± 1%) were found in OCV biogas (Fig. 1).
The maximal hydrogen yield for this wastewater was obtained by Pt-
MEC (0.174 ± 0.001Nm 3 H 2/kg TCOD removed), while lower yields were achieved with MoS2 and SS as catalysts (0.117 ± 0.001 and 0.075 ±
0.002 Nm 3 H 2/kg TCOD removed, respectively) (Fig. 2). Previous MEC tests using domestic wastewater in a two-chamber MEC (Ditzig et al.
2007) resulted in hydrogen yields similar to those obtained here.
Both the biogas yield and the hydrogen yield were higher in MEC
than in OCV, in particular with Pt as the catalyst (Fig. 2).
Methane yield were higher than hydrogen yield and in particular they
were again higher in MEC than in OCV. MEC showed similar
methane yields between the different catalysts and a range between
0.286 Nm 3 CH 4/kg TCOD removed (SS-MEC) and 0.300 Nm 3 CH 4/kg TCOD removed (Pt-MEC) (Fig. 2). The methane yields in OCVCs were lower, with 0.260 ± 0.002, 0.245 ± 0.002, 0.277 ± 0.003 Nm 3 CH 4/kg TCOD removed for Pt-MEC, MoS2-MEC and SS-MEC respectively (Fig. 2).
More interestingly, Fig. 3 shows the overall net energy balance of the
processes, expressed as kWh/kg TCOD removed, considering the amount
of energy in the biogas produced (CH4 + H2) per kg COD removed.
Apart for similar energy recovery with SS, both Pt and MoS2
catalysts reached higher energy recovery in MEC than in OCV. Pt-
MEC got the highest recovery of 3.758 ± 0.003 kWh/kg TCOD removed,
while lower performances were reached in OCV (3.236 ± 0.003
kWh/kg TCOD removed); with MoS2 at the cathode 3.379 ± 0.003 and
2.971 ± 0.003 kWh/kg TCOD removed were respectively achieved in MEC
and OCV.
3.3 Treatment efficiency Consistent and effective treatment of the wastewater was achieved
both in OCV and with addition of external voltage of +0.7 V (MEC).
TCOD removal efficiency in MEC was always over 85%, with lower
performance for SS cathode and higher (89%) for Pt (Fig. 4, Tab. 2)
and with results similar to those obtained during the preliminary
study in MFC. Considering the time of the batch cycle (31 h, 37.5 h,
47.5 h for Pt, MoS2 and SS cathodes, respectively) and the relatively
high amount of TCOD in the influent (4070 ± 180 mg/L) with high
content of methanol (1540 ± 50 mg/L) these removals can be
considered quite encouraging positive results.
Ditzig et al. (2007) showed an overall higher COD removal of 95 ±
2% with MEC treating domestic wastewater, which however had a
lower initial COD, between 204 and 481 mg/L.
Summarily comparing the MECs with OCVCs, the latter always
showed a lower TCOD removal efficiency (between 74% and 81%,
Fig. 4, Tab. 2) than MECs. OCVCs lower performances were
especially observed with Pt and MoS2 cathodes. On the contrary, if
we consider the removal rate of TCOD (r TCOD, Fig. 4), MEC
achieved an higher rate than OCV just with MoS2 as catalyst (2.26
and 1.87 kg TCOD removed/m3reactor d for MoS2-MEC and MoS2-
OCV, respectively). However, the overall higher r TCOD was
reached by Pt-MEC (2.80 kg TCOD removed/m3reactor d) and Pt-
OCV (3.17 kg TCOD removed/m 3reactor d) (Fig. 4). Furthermore, the soluble part of COD (SCOD) is high and similar between the two
processes (MEC and OCV) and the different catalysts, being
comprised between 78% and 87% of the total COD (Tab. 2).
Tab. 2. shows the content of VFAs and alcohols in the liquid solution
of MEC and OCVC for all the cathode types. All the MECs showed
just traces of acetate (< 30 mg/L) within the liquid solution and a
complete consumption of the methanol present in the feeding, while
the OCVCs had a more complex profile with unexpected high amount of acetate (always higher than 130 mg/L) and uneven
detection of methanol, acetone, propionate and butyrate.

4. Discussion Like previously demonstrated by other researchers (Clauwaert and
Verstraete, 2008), persistent methanogenesis was obtained in this
study with the use of single chamber MECs, which have a simple
reactor design (no membrane) and have been fed with a raw actual
wastewater, with high methanol content and likely rich in complex
anaerobic microflora.
Moreover, the relatively long cycle time (Fig. S.2) required both for
a more complete degradation of the complex high-TCOD organic
substrate, and for avoiding kinetic roadblocks from hydrolysis and
fermentation in the MEC (Lee et al. 2009), probably resulted in
favouring a well established methanogenic community in the biofilm
on the electrodes. Therefore, even if this study didn't consider any
microbial community characterization, it is presumable the
simultaneous presence and activity of fermentative methanogenic
bacteria and exoelectrogenes, as hydrolysis and fermentation are
both preliminary steps necessary for the utilization of a complex
wastewater by exoelectrogenes.
Ditzig et al. (2007) already stated that, compared to the acetate or
non fermentable pure compounds, the use of complex wastewaters
could bring to a low efficiency of hydrogen recovery (H2 consumed
by other microorganisms) and to a low electron recovery (failure to
convert organic matter to current). With domestic wastewater (Ditzig
et al. 2007) maximum CE of 26% was reached, compared to CEs of
78% and 92% in previous MEC studies with acetate (Liu et al. 2005;
Rozendal et al. 2006). Therefore, in the present study a large
percentage of the electrons coming from the COD removed was
probably not successfully transferred into current and instead used for cell growth or for anaerobic digestion process, as indicated by
low CEs (7-12.5%) and by the amount of methane found in the cells.
However, the production of biogas with high relative content of
methane (always > 55%) gives opportunity for this technology to be
compared with classical anaerobic digestion process. Indeed, the
present study showed that both Pt and MoS2 as catalysts achieved
higher biogas production rate and biogas yield in MEC than in OCV.
At the temperature of 30 °C, maximal average methane production
rate of 0.9 m 3 CH 4 / m 3 reactor day was reached by Pt-MEC while with MoS2 at the cathode the CH4 production rate increased from
0.51 in OCV to 0.73 m 3 CH 4/m 3 reactor day in MEC, a rate similar to that shown by Clauwaert and Verstraete (2008) with graphite
granules at the cathode.
Hydrogen was found in biogas together with methane, but it was
generated at low rates, if compared to other studies with both pure
compounds in buffered environment and wastewaters (Cheng and
Logan 2007, Call and Logan 2008, Lalaurette et al. 2009, Wagner et
al. 2009). Hydrogen production rate was higher in MEC than in OCV
and the maximal rate in MEC was achieved with Pt catalyst (0.54 ±
0.08 m 3 H 2/m 3 reactor day). However it is possible to assume that the higher relative amount of
hydrogen produced in MEC (up to 32% of the biogas, v/v, for Pt-
MEC) could have been even higher, because an undefined additional
amount of H2 electrochemically produced by exoelectrogenes could
have been consumed by hydrogenotrophic bacteria to produce
methane, or other reduced compounds as acetate (Thygesen et al.
2010). Tartakovsky et al. (2008) gave evidence that hydrogenotrophic methanogens converted up to 50% of the H2
produced at the cathode into methane in a continuous-flow MEC.
In order to be really competitive, MEC needs to reach the anaerobic
digestion cost-effectiveness level and for this purpose the net
amount of energy extracted was calculated, since the energy content
in the biogas must offset the energetic cost of the electrical energy (i.e. applied voltage) consumed in the MEC. MEC energy recovery,
considering both the methane and the hydrogen gas produced, was
positive (3.76 and 3.38 kWh/kg TCOD removed for Pt- and MoS2-
respectively) and 14 - 16 % higher than that of OCV.
As previously stated, the capital cost for the electrodes is still a big
obstacle for an easy application of MECs to a large-scale facility, and
Tokash and Logan (2011) recently demonstrated that MoS2
composite cathodes perform similarly to Pt cathodes in terms of
current densities, hydrogen production rates and COD removal in
MEC fed with simple pure substrate (sodium acetate) and that 10
wt.% platinum composite cathodes are about five times more
expensive than similarly constructed composite MoS2 cathodes. The
good energy recovery and overall performance of MoS2-MEC
reached in this study demonstrated that also using wastewaters MoS2
is a valid alternative to Pt as catalyst, able to achieved the same
increase in performances, compared to its corresponding OCVC,
obtained with Pt.
Even from the organic removal efficiency standpoint, not only
microbial electrolysis system may be a promising alternative to AD
(higher TCOD removal) but again the use of MoS2 catalyst achieved
a TCOD removal rate in MEC 20% higher than in OCV.
The study of Ditzig et al. (2007) assumed that the initial COD was
affecting the reactor performances, which increased with wastewater
strength, suggesting that wastewaters with high organic matter
concentration could be useful in immediate applications of
bioelectrochemical systems. Here COD-rich wastewater was
efficiently treated in MEC (between 85 and 87 % TCOD removed),
showing higher performances than previous results with actual
wastewaters. Wagner et al. (2009) obtained swine wastewater COD
reduction up to 72 ± 4%; Cusick et al. (2010) a TCOD removal of 58
± 3% and 47 ± 3% with domestic and winery wastewater,
respectively. But this study also demonstrates that MEC system has
higher organic removal than anaerobic digestion process, simulated in OCV, therefore gaining another advantage in the comparison
between the two processes, since higher TCOD removal means
reduction of the effluent waste handling cost.
In MFC the presence of other organisms (mixed culture) than those
that can transfer electrons to the anode (pure culture) has been shown
to benefit the performance, generating a current six fold higher than
that produced by the pure culture (Angenent et al. 2004) and also
increasing the COD removal. Similarly, the biological process here
established could have been improved by the interaction between the
exoelectrogenes bacteria and the hydrogenotrophic methanogens
within the thick biofilm on the anode (Lee et al. 2009), increasing the
already high TCOD removal achieved in OCV (up to 81% with
stainless steel as catalyst). Moreover, also the removal of the high
amount of methanol, never before demonstrated in MEC, could have
profit by this cooperation. In this specific case of study, the feeding
wastewater had an high amount of methanol, which, as an important
material of chemical plants, often occurs in chemical wastewaters,
even in high concentration and contributed here to the 55% of the
total COD of the wastewater (Tab. 1). Previous studies evaluated the
exploitation of methanol both in a conventional fuel cell (Logan
2004) and in anaerobic degradation, achieving high removal
performance especially in upflow anaerobic sludge blanket reactors
(Chen et al., 2000; Paris and Blondeau, 1999; Woods et al., 1989).
The 99% of methanol contained in the condensate from a paper mill
was biodegraded in a UASB by Park and Park (2003).
In this study, methanol was completely degraded both in MEC and
OCVC, except for traces below 12 mg/L in Pt- and MoS2-OCVC
(Tab. 2).
At a pH close to 7.0, i.e. that of this experimentation, methanol could
either be directly converted to methane or through formation of
acetate or through a combination of both (Bhatti et al. 1996).
4CH3OH '' 3CH4 + H2O + HCO3 + H + (1) Generally, the former is the dominant reaction while the acetate-
producing relies on the existence of CO2 or HCO3- and trace
elements (Gonzalez-Gil et al. 1999; Weijma and Stams 2001).
Here, the simultaneous absence of methanol in the ffluent of the
processes and the increased content of acetate recorded in OCV seem
to suggest that the conversion of methanol to acetate via acetoclastic
methanogens, expressed as follows
4CH3OH + 2HCO3 '' 3CH3COO + H + + 4H 2O (2) is prevailing in this study.
The difference in the acetate content at the end of the process
between MEC and OCVC could be connected with the higher TCOD
removal and the higher amount of biogas produced in MEC,
assuming that additional acetate from methanol fermentation has
been efficiently used by the exoelectrogenic bacteria, favored by the
added voltage.
Further work on microbial communities characterization in these
systems could clarify better the interrelationships between the
bacteria involved in the degradation of such a type of industrial,
high-methanol content wastewater.

5. Conclusions A comparison between MEC and Anaerobic Digestion, simulated
without adding voltage to a MEC system, was made, regarding
energy balance and organic removal efficiency of the two processes.
It is known that MEC is not a technology ready for real-scale
application like AD and that its use for wastewater treatment will
depend on different factors (the cost of the materials, which for
large-scale treatment is yet to known, and especially of the catalysts,
the amount of energy needed, etc.).
However substantial methane production was found in single
chamber MEC, and with both platinum and molybdenum sulfide at the cathode higher energy recovery and TCOD removal than in
anaerobic fermentation was reached.
Therefore MoS2 proved to be a valid alternative to Pt at the cathode,
much more affordable for pilot- or real-scale appliances. Moreover
the MEC technology gave evidence to be a competitive method for
efficiently treating unbuffered, raw wastewaters, exploiting the
joined different skills of fermentative bacteria and exoelectrogenes to
efficiently treat complex wastewaters and specific contaminants,
such as methanol in this study.
In order to enhance the advantages toward the classical anaerobic
digestion process and to strengthen its application niche, further
studies and comparison could be made with MEC treating low
concentration COD substrates and at low temperatures (10-20 °C),
i.e., where AD does not function well (Pham et al. 2006). Also, as
proposed by Logan et al. (2008), research is needed on whether MEC
systems will be capable of stand-alone operation or if aerobic
effluent polishing must be coupled, as often required by AD.
Alternatively, as proposed by Zeeman et al. (2008), methane-
producing MECs could be used in combination with conventional
anaerobic digestion as a way to remove residual fatty acid and

Acknowledgements The authors thank Air Products and Chemicals, Inc. for providing
wastewater samples.

6. References Angenent LT, Karim K, Al-Dahhan MH, Wrenn BA, Domìguez-
Espinosa R. 2004. Production of bioenergy and biochemicals from
industrial and agricultural wastewater. TRENDS in Biotechnology
Pham TH, Rabaey K, Aelterman P, Clauwaert P, De Schamphelaire
L, Boon N, Verstraete W. 2006. Microbial Fuel Cells in Relation to
Conventional Anaerobic Digestion Technology. Eng Life Sci 6(3)

Liu H, Grot S, Logan BE. 2005. Electrochemically assisted microbial
production of hydrogen from acetate. Environ Sci Technol 39:4317-

Thygesen A, Thomsen AB, Possemiers S, Verstraete W. 2010.
Integration of Microbial Electrolysis Cells (MECs) in the Biorefinery
for Production of Ethanol, H2 and Phenolics. Waste Biomass Valor

Lalaurette E, Thammannagowda S, Mohagheghi A, Maness PC,
Logan BE. 2009. Hydrogen production from cellulose in a two-stage
process combining fermentation and electrohydrogenesis. Int J
Hydrogen Energy 34:6201-6210.

Call D, Logan BE. 2008. Hydrogen production in a single chamber
Microbial Electrolysis Cell Lacking a Membrane. Environ Sci
Technol 42:3401-3406.

Wagner RC, Regan JM, Oh SE, Zuo Y, Logan BE. 2009. Hydrogen
and methane production from swine wastewater using microbial
electrolysis cells. Water research 43:1480-1488.

Cusick RD, Kiely PD, Logan BE. 2010. A monetary comparison of
energy recovered from microbial fuel cells and microbial electrolysis
cells fed winery or . domestic wastewaters. Int J Hydrogen Energy

Logan BE, Call D, Cheng S, Hamelers HM, Sleutels THJA,
Jeremiasse AW, Rozendal RA. 2008. Microbial Electrolysis Cells for High Yield Hydrogen Gas Production from Organic Matter. Environ
Sci Technol 42:8630-8640.

Clauwaert P, Verstraete W. 2008. Methanogenesis in membraneless
microbial electrolysis cells. Appl Microbiol Biotechnol

Rozendal RA, Sleutels THJA, Hamelers HVM, Buisman CJN. 2008.
Effect of the type of ion exchange membrane on performance, ion
transport, and pH in biocatalyzed electrolysis of wastewater. Water
Science & Technology | 57.11 |:1757-1762.

Jeremiasse AW, Hamelers HVM, Saakes M, Buisman CJN. 2010. Ni
foam cathode enables high volumetric H2 production in a microbial
electrolysis cell. Int Journal Hydrogen energy 35:12716-12723.

Cheng SA, Liu H, Logan BE. 2006. Increased performance of single-
chamber microbial fuel cell using an improved cathode structure.
Electrochem. Commun. 8:489-494.

Min B, Logan BE. 2004 Continuous Electricity Generation from
Domestic Wastewater and Organic Substrates in a Flat Plate
Microbial Fuel Cell. Environ. Sci. Technol. 38(21):5809-5814.

Heilmann J, Logan BE. 2006. Production of Electricity from Proteins
Using a Microbial Fuel Cell. Water Environ. Res. 78(5):1716-1721.

Min B, Kim JR, Oh SE, Regan JM, Logan BE. 2005. Electricity
generation from swine wastewater using microbial fuel cells. Wat.
Res. 39(20):4961-4968.

Ditzig J, Liu H, Logan BE. 2007. Production of hydrogen from
domestic wastewater using a bioelectrochemically assisted microbial
reactor (BEAMR). Int J Hydrogen Energy 32:2296-2304.
Lee HS, Torres CI, Parameswaran P, Rittmann BE. 2009. Fate of H2
in an Upflow Single-Chamber Microbial Electrolysis Cell Using a
Metal-Catalyst-Free Cathode. Environ. Sci. Technol. 43:7971-7976.

Liu H, Grot S., Logan BE. 2005. Electrochemically Assisted
Microbial Production of Hydrogen from Acetate. Environ. Sci.
Technol. 39(11):4317-4320.

Rozendal RA, Hamelers HVV, Euverink GJW, Metz SJ, Buisman
CJN. 2006. Principle and perspectives of hydrogen production
through biocatalyzed electrolysis. Int J Hydrogen Energy 31:1632-

Cheng S, Logan BE. 2007. Sustainable and efficient biohydrogen
production via electrohydrogenesis. PNAS, 104(47):18871-18873.

Tartakovsky B, Manuel MF, Neburchilov V, Wang H, Guiot SR.
2008. Biocatalyzed hydrogen production in a continuous flow
microbial fuel cell with a gas phase cathode. J Power Sources

Tokash JC, Logan BE. 2010. Electrochemical evaluation of a
molybdenum disulfide catalyst for the hydrogen evolution reaction
under solution conditions applicable to microbial electrolysis cells.
Int. J. Hydrogen Energy 36(16):9439-9445.

Logan BE. 2004. Extracting Hydrogen and Energy from Renewable
Resources (Feature article). Environ. Sci. Technol. 38(9):160A-
Chen SH, Dong J, Berthouex PM, Boyle WC. 2000. Fate of
Pentachlorophenol (PCP) in an Anaerobic Digester. Water
Environment Research 72(2):201-206.

Paris D, Blondeau R. 1999. Isolation and characterization of
Arthrobacter sp. from activated sludge of a pulp and paper mill. Wat.
Res. 33(4):947-950.

Park JH, Park JK. 2003. Fate of Methanol in an Anaerobic Digester.
Korean J. Chem. Eng. 20(3):509-516.

Woods SL, Ferguson JF, Benjamin MM. 1989. Characterization of
Chlorophenol and Chloromethoxybenzene Biodegradation during
Anaerobic Treatment. Environ. Sci. Technol. 23(1):62-68.

Gonzalez-Gil G, Kleerebezem R, Van Aelst A, Zoutberg GR,
Versprille AI, Lettinga G. 1999. Toxicity effects from formaldehyde
on methanol degrading sludge and its anaerobic conversion in biobed
expanded granular sludge (EGSB) reactors. Wat. Sci. Tech.

Weijma J, Stams AJ. 2001. Stams Mehtanol conversion in high-rate
anaerobic reactors. Wat. Sci. Tech. 44(8):7-14.

Bhatti ZI, Furukawa K, Fujita M. 1996. Feasibility of Methanolic
Waste Treatment in UASB Reactors. Wat. Res. 30(11):2559-2568.

Zeeman G, Kujawa K, de Mes T, Hernandez L, de Graaff M, Abu-
Ghunmi L, Mels A, Meulman B, Temmink H, Buisman C, van Lier
J, Lettinga G. 2008. Anaerobic treatment as a core technology for
energy, nutrients and water recovery from source-separated domestic
waste(water). Water Sci Technol 57:1207-1212.
Tables and Figures Tab.1 - Wastewater characterization. Fig. S.1 - Current density for four representative MFC reactors fed with undiluted, not-
amended wastewater. Parameters pH 6.68 ± 0.30 Conductivity mS/cm 2.04 ± 0.02 TCOD mg/L 4070 ± 180 SCOD mg/L 3810 ± 160 BOD mg/L 800 TS mg/L 1340 ± 91 TSS mg/L 63 ± 18 Phosphorous (P) mg/L 8.9 Solfate (SO4) mg/L 55.5 Nitrate (NO3) mg/L < 5 Nitrogen Ammonia (NH3-N) mg/L 0.25 Total carbohydrates mgCOD/L 386 ± 7 Soluble carbohydrates mgCOD/L 240 ± 6 Acetone mg/L 52.85 ± 1.8 Methanol mg/L 1537.4 ± 48.6 Ethanol mg/L 18.3 ± 4.8 Propanol mg/L 2.1 ± 1.9 Butanol mg/L 0 Acetate mg/L 182.4 ± 34.4 Propionate mg/L 0 Butyrate mg/L 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 5 10 15 20 25 C u rr e n t D e n si ty ( A /m 2 ) Time (d) Tab. S.1 - Chemical characterization of the liquid phase of MFC reactors after the first peak of
current production and at the end of the batch cycle. Fig. S.2 - Illustrative cycle of current density (A) and cumulative biogas production (B) in the
single chamber MECs used, at 0.7 V applied voltage. Duplicate reactors for each cathode type
were used: 10% platinum on carbon cloth (blu lines); 33% MoS2 on carbon cloth (red lines);
bare stainless steel (green lines). (For interpretation of the references to colors in this figure
legend, the reader is referred to the web version of this article). Parameters Influent (wastewater) End of the first current peak (''2.5 d) End of the batch cycle (effluent) pH - 6.68 ± 0.30 - 7.25 ± 0.22 Conductivity mS/cm 2.04 ± 0.02 - 2.13 ± 0.13 TCOD mg/L 4070 ± 180 1460 ± 32 397 ± 11 SCOD mg/L 3810 ± 160 - 336 ± 20 SCOD/TCOD % 94 ± 4 - 84 ± 8 TCOD removal % - 64 90 Organic Compounds Acetone mg/L 52.85 ± 1.8 0 0 Methanol mg/L 1537.4 ± 48.6 323.4 ± 24.8 0 Ethanol mg/L 18.3 ± 4.8 0 0 Propanol mg/L 2.1 ± 1.9 0 0 Butanol mg/L 0 0 0 Acetate mg/L 182.4 ± 34.4 46.3 ± 14.4 0 Propionate mg/L 0 0 0 Butyrate mg/L 0 0 0 0.0 0.5 1.0 1.5 2.0 2.5 0 10 20 30 40 50 C ur re n t D e ns it y (A /m 2 ) Time (h) 0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 B io g a s p ro d u ct io n ( m L) Time (h) Fig. 1 - Production rate (Q) for specific biogas compound. Comparison between MEC and
OCV with different cathodes (data averaged over 5 and 2 cycles, respectively). Fig. 2 - Hydrogen and methane yield per TCOD removal. Comparison between MEC and
OCV with different cathodes (data averaged over 5 and 2 cycles, respectively). 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 Pt-MEC Pt-OCV MoS2-MEC MoS2-OCV SS-MEC SS-OCV Q ( m 3 - g a s/ m 3 re a c to r- d ) CO2 CH4 H2 0.00 0.10 0.20 0.30 0.40 0.50 0.60 Pt-MEC Pt-OCV MoS2-MEC MoS2-OCV SS-MEC SS-OCV G a s yi e ld ( N m 3 g a s/ kg T C O D r e m o ve d) H2 CH4 Fig. 3 - Net energy recovery comparison between MEC (grey) and OCV (black) process with
different cathodes. Fig. 4 - TCOD removal (%) and TCOD removal rate (kgTCOD removed/m 3 reactor-day). Comparison between MEC and OCV with different cathodes (data averaged over 5 and 2 cycles,
respectively). 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 Pt-MEC Pt-OCV MoS2-MEC MoS2-OCV SS-MEC SS-OCV E n e rg y b a la n ce (K W h /k g C O D r e m o ve d) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 0 10 20 30 40 50 60 70 80 90 100 Pt-MEC Pt-OCV MoS2-MEC MoS2-OCV SS-MEC SS-OCV r T C O D ( k g T C O D r e m o ve d/ m 3 re a ct o r- d a y ) T C O D r e m o v a l ( % ) Tab. 2 - Effluent characterization and TCOD removal efficiency comparison between MEC
and OCV with different cathodes (data averaged over 5 and 2 cycles, respectively). Parameters Pt-MEC Pt-OCV MoS2-MEC MoS2-OCV SS-MEC SS-OCV pH 6.38 ± 0.02 6.91 ± 0.10 6.32 ± 0.03 6.68 ± 0.21 6.57 ± 0.10 6.34 ± 0.31 Conductivity mS/cm 2.01 ± 0.04 2.71 ± 0.16 2.05 ± 0.0 2.67 ± 0.04 2.14 ± 0.07 2.16 ± 0.0 TCOD mg/L 453 ± 3 881 ± 18 513 ± 16 1035 ± 52 585 ± 88 754 ± 31 SCOD mg/L 356 ± 3 697 ± 35 422 ± 40 832 ± 34 480 ± 46 656 ± 8 TCOD removal % 89 ± 0 78 ± 0 87 ± 0 74 ± 1 85 ± 3 81 ± 0 Organic
Compounds Acetone mg/L 0 0 0 8.4 ± 1.3 0 0 Methanol mg/L 0 8.8 ± 3.1 0 12.4 ± 5.4 8.6 ± 0.8 0 Ethanol mg/L 0 0 0 0 0 0 Propanol mg/L 0 0 0 0 0 0 Butanol mg/L 0 0 0 0 0 0 Acetate mg/L 13.3 ± 1.8 284.0 ± 21.2 27.5 ± 1.5 585.2 ± 25.9 11.5 ± 2.7 131.2 ± 58.9 Propionate mg/L 0 20.0 ± 4.4 0 55.6 ± 7.2 0 7.2 ± 1.5 Butyrate mg/L 0 0 0 21.2 ± 3.9 0 7.4 ± 2.3 APPENDIX B Congress communications

Bio-hydrogen production from bio-waste: ready
for full-scale applications'
Schievano A. a, Consonni E.a, Tenca A.b, Oberti R.b, Adani F.a a Gruppo Ricicla, Dipartimento di Produzione Vegetale - Università di Milano, Via
Celoria 2, 20133 Milano, Italy
b Dipartimento di Ingegneria Agraria, Università di Milano, Via Celoria 2, 20133
Milano, Italy Abstract Bio-hydrogen production through fermentation of organic substrates
can be a good candidate for full scale application in the next future,
especially if the organic materials are by-products or residues of any
human activity. In this work, bio-hydrogen was produced from two
semi-continuous lab-scale digesters (D1 and D2). The substrates
used were household source-separated bio-waste collected in 3
municipalities in Lombardy (Italy) (D1) and the biowaste coming
from the green-market of the city of Milan (D2). The two semi-
continuous digesters were manually-fed once a day with a organic
loading rate of respectively 16.4 and 13.5 gVS d -1 l-1 and the hydraulic retention times were respectively 4 and 3 days.
Both digesters resulted in interesting bio-hydrogen productions. The
average bio-hydrogen production rates were respectively 0.85 and
1.70 Ndm 3 H 2 l digest d -1, while the maximum rates obtained were 1.59 and 3.15 Ndm 3 H 2 ldigest d -1. The average conversion yields were respectively 55 and 126 Ndm 3H 2 g -1 VS, while the maximum rates obtained were 105 and 227 Ndm 3H 2 g -1 VS. The produced biogas showed always complete absence of methane and an hydrogen
content around 50%. Because the digesters were manually and semi-
continuously loaded, the parameters were not optimized and the
producion showed some instability. On the other hand, the obtained results were indicative and very promising for exploiting this
technology in full scale biogas plants.

Keywords: bio-hydrogen, renewable fuel, waste, biogas

1 Introduction Hydrogen has been always recognized as an ideal alternative energy
source to substitute fossil fuels.
Hydrogen produced directly from organic materials by bacteria, i.e.
bio-hydrogen, has considerable potential in defining hydrogen''s
future use [1].
In anaerobic conditions, organic matter is converted to methane and
carbon dioxide via a series of interrelated microbial metabolisms,
including hydrolysis/fermentation, acetogenesis, and methanogenesis. Fermentative bacteria hydrolyze and ferment
carbohydrates, proteins, and lipids to volatile fatty acids, which are
further converted to acetate, and CO2/H2 by acetogenic bacteria. The
products of acetogenesis, i.e. acetate and CO2/H2, are finally
converted to methane by methanogenic bacteria [2]. A bioreactor
could possess significant capacity for the transformation of organics
into hydrogen gas when bioactivity of hydrogen consumers
contained in a bioreactor was inhibited [3-6]. Some methods have
been reported to inhibit methanogen bacteria and to harvest
anaerobic spore-forming bacteria such as Clostridium sp., capable to
produce hydrogen. One is a heat shock of the inoculum at 100º C for
2 hours, which favours only spore-forming microrganisms. Other
method is the pH control in the interval 5<pH<6, which has been
shown to be optimal for hydrogen-type fermentation and to inhibit
methanogenic activity [7-9]. In literature, the pH control has been
always achieved by the use of chemicals such as NaOH or KOH and
HCl [7-10]. On the other hand, the use of large amounts of reagents
wouldn''t be possible in a full scale-process. Besides, high concentrations in the digester of volatile fatty acids
(VFA), forming during fermentation, are responsible of both
inhibiting the hydrogen-producing bacteria and dropping the pH
below pH 5 [11]. The concentrations of VFA in the digester are
proportional to the organic loading rate (OLR) and at the same time
to the hydraulic retention time (HRT). The higher the OLR is, the
faster the wash-out of the VFA produced should be, i.e. low HRT.
Han et al. [11] suggested diluting the liquid phase of food waste
during its anaerobic fermentation by leaching pure water through the
solids (leaching''bed reactor), so that the high concentration of VFA
is washed out. The result was an optimal dilution D (D = flow rate /
operating volume) of 4.5 d -1 [11]. On the other hand, this solution, applied on full-scale, would imply huge water consumption and
problems in environmental and economical feasibility.
Many traditional biogas plants in Europe are connected to farms,
which produce large amounts of animal slurries. These pre-digested
liquid manures, normally have a pH of 6.5-8, a considerable alkaline
buffer-capacity and low concentrations of volatile fatty acids.
This paper aims to investigate the use of swine manure in co-
digestion with fresh biowaste materials, to control the hydrogen
production process in semi-continuous thermophilic bio-digesters.
Positive results would help in providing more informations for future
full-scale developments.

2 Methods 2.1. Seed microorganisms The seed sludge was the digestate taken from an anaerobic digester
biogas plant and boiled (100º C) for 2 hours to inactivate
hydrogenotrophic bacteria and to harvest anaerobic spore-forming
bacteria such as Clostridium sp. [7]. This procedure was carried out
for the start-up phase of reactors. The pH, alkalinity, total VFA concentration and volatile solids (VS) concentration of the sludge
were respectively 8.1, 7470 mgCaCO3 l -1, 1190 mg l-1 and 750 mg l- 1. 2.2. Feedstock for feeding The two considered substrates were respectively household source-
separated bio-waste collected in 3 municipalities in Lombardy (Italy)
(S1) and the biowaste coming from the green-market of the city of
Milan (S2). The substrates (S) were used for creating feeding
mixtures with swine manure (SM). The two mixtures (M1 and M2)
were made by mixing at the ratios S:SM of 1:1 and 1:2, respectively.
In Table 1, a characterization of the used materials is reported. Table 1 '' Characterization of the materials used 2.3. Experimental setup and procedure A completely mixed reactor with working volume of 600 ml was
operated in a semi-continuous mode by feeding once a day. The
digesters D1 and D2 were operated at a temperature of 55ºC
(thermophilic conditions) and Hydraulic Retention Times (HRTs) of
4 and 3 days, respectively (Table 2). The output digestate was withdrawn once a day before each feed. The digesters were fed with
organic loading rates (OLRs) of respectively 16.4 gVS l -1 d-1 and 13.5 gVS l -1 d-1 (Table 2). 2.4. Analytical procedures The biogas production was measured using a volumetric gas meter
column connected to the headspace of the digester. Biogas
composition was analyzed by a gas chromatograph (Agilent, Micro
GC 3000A) equipped with two thermal conductivity detectors (TCD)
and two columns, using Nitrogen and Helium as carriers.

3 Results and discussion 3.1. H2 productions The process performances in terms of H2 yield per S unit added to
the digester (Ndm 3 H 2 kg -1 VS) and the volumetric hydrogen production rates (Ndm 3 H 2 kg -1 VS) are reported in Table 2. The H2 to CO2 ratios in biogas were always found in the range 40 '' 60% in
both the digesters. Methane was never found in the biogas, indicating
complete inhibition of the hydrogen-consumers microrganisms.
These favourable conditions were created by the quick HRTs (3-4
days) and the relatively low pH operational conditions (Table 2).
The hydrogen production rates and yields showed noticeable
imbalances. Some intermediate periods, for both the digesters, were
characterized by a lag-phases, in which less biogas was produced
compared to the maximum productions. This drove to low hydrogen
yields and rates, if compared with the best results (Table 2). This was
the reason for the relatively big differences between the average and
the maximum values for both H2 production yields and rates reported
in Table 2. However, the semi-continuous mode, do not permit to
achieve stable productions and to optimize the results. On the other
hand, the results were quite satisfactory, if compared to the yields of 43 dm 3H 2 kg -1 VSadded, achieved by continuously stirred and fed reactors (HRT of 2 days) by Liu et al. [7].

4 Conclusions Bio-hydrogen production from different types of bio-waste was
found possible. With simple lab-equipment, satisfactory H2
productions were achieved. The use of swine manure in co-digestion
with the bio-waste favoured the process conditions, by buffering the
pH to ideal values and diluting the produced VFAs. Further research
should be carried out in improving and optimizing the process and
for assessing the methane yields, obtainable from the effluent of the
H2-production. This may easily lead to full-scale applications in a
close future.

Table 2 '' Operational conditions and bio-hydrogen productions References [1] ZAJIC, J.E., KOSARIC, N., BROSSEAU, J.D.: Microbial
production of Hydrogen. Adv Biochem Eng, 7, 57-109 (1978).

[2] BITTON, G.: Wastewater microbiology. New York: Wiley-Liss,

Hydrogen production from inhibited anaerobic composters. Int J
Hydrogen Energy, 22(6), 563''6 (1997).

[4] LAY, J.J., LEE, Y.J., NOIKE, T.: Feasibility of biological
hydrogen production from organic fraction of municipal solid waste.
Water Res, 33(11), 2579''86 (1999).

[5] LAY, J.J.: Modeling and optimization of anaerobic digested
sludge converting starch to hydrogen. Biotechnol Bioeng, 68, 269''
78 (2000).

[6] LAY, J.J.: Biohydrogen generation by mesophilic anaerobic
fermentation of microcrystalline cellulose. Biotechnol Bioeng, 74(4),
280''7 (2001).

[7] LIU, D., LIU, D., ZENG, R.J., ANGELIDAKI, I.: Hydrogen and
methane production from household solid waste in the two-stage
fermentation process, Water Research, 40, 2230-2236 (2006).

[8] CHANG, F.Y., LIN, C.Y.: Biohydrogen production using an
upflow anaerobic sludge blanket reactor. Int. J. Hydrogen Energy, 29
(1), 33''39 (2004).
[9] OH, S.E., VAN GINKEL, S., LOGAN, B.E.: The relative
effectiveness of pH control and heat treatment for enhancing
biohydrogen gas production. Environ. Sci. Technol., 37 (22), 5186''
5190 (2003).

[10] HANG-SIK, S., JONG-HO, Y.: Conversion of food waste into
hydrogen by thermophilic acidogenesis, Biodegradation, 16, 33''44

[11] HAN, S.K., SHIN, H.S.: Biohydrogen production by anaerobic
fermentation of food waste, International Journal of Hydrogen
Energy, 29, 569 '' 577 (2004).

An operational strategy to produce Bio-hydrogen:
the use of digestate for process control
A. Schievano a, A. Tenca b, R. Oberti b and F. Adani a a Dipartimento di Produzione Vegetale, University of Milano, Milano (I)
b Istituto di Ingegneria Agraria, University of Milano, Milano (I) Abstract A semi-continuous digester was fed twice a day with a concentrated
solution of glucose (100 g l -1) and monitored for a 30-days period, with the aim of testing the possibility of utilizing the digestate of a
traditional biogas plant, after a heat-shock at 100ºC, for controlling
process parameters (organic loading rate OLR, pH, volatile fatty
acids VFA concentration), by adding it to the fresh substrate at a
ratio R of the total feeding volume. The process resulted instable for
OLR=10 gVS L -1and R=0.7, while more stable for OLR of 5 g VS L - 1and R=0.85. The maximum bio-hydrogen production rate in stable
conditions was 100 NmLH2 h -1 and the conversion yields were 1.7 - 1.8 molH2 mol -1glucose. The produced biogas showed always complete absence of methane.

Keywords: bio-hydrogen, renewable fuel, waste, biogas

Introduction Hydrogen has been always recognized as an ideal alternative energy
source to substitute fossil fuels.
Hydrogen produced directly from organic materials by bacteria, i.e.
bio-hydrogen, has considerable potential in defining hydrogen''s
future use [1].
In anaerobic conditions, organic matter is converted to methane and
carbon dioxide via a series of interrelated microbial metabolisms, including hydrolysis/fermentation, acetogenesis, and methanogenesis. Fermentative bacteria hydrolyze and ferment
carbohydrates, proteins, and lipids to volatile fatty acids, which are
further converted to acetate, and CO2/H2 by acetogenic bacteria. The
products of acetogenesis, i.e. acetate and CO2/H2, are finally
converted to methane by methanogenic bacteria [2]. A bioreactor
could possess significant capacity for the transformation of organics
into hydrogen gas when bioactivity of hydrogen consumers
contained in a bioreactor was inhibited [3''6]. Some methods have
been reported to inhibit methanogens and to harvest anaerobic spore-
forming bacteria such as Clostridium sp., capable to produce
hydrogen. One is a heat shock of the inoculum at 100ºC for 2 hours,
which favours only spore-forming microrganisms. Other method is
the pH control in the interval 5<pH<6, which has been shown to be
optimal for hydrogen-type fermentation and to inhibit methanogenic
activity [7 - 9]. In literature, the pH control has been always achieved
by the use of chemicals such as NaOH or KOH and HCl [7-10]. On
the other hand, the use of large amounts of reagents wouldn''t be
possible in a full scale-process.
Besides, high concentrations in the digester of volatile fatty acids
(VFA), forming during fermentation, are responsible of both
inhibiting the hydrogen-producing bacteria and dropping the pH
below pH 5 [11]. The concentrations of VFA in the digester are
proportional to the organic loading rate (OLR) and at the same time
to the hydraulic retention time (HRT). The higher the OLR is, the
faster the wash-out of the VFA produced should be, i.e. low HRT.
Han et al. [11] suggested diluting the liquid phase of food waste
during its anaerobic fermentation by leaching pure water through the
solids (leaching''bed reactor), so that the high concentration of VFA
is washed out. The result was an optimal dilution D (D = flow rate /
operating volume) of 4.5 d -1 [11]. On the other hand, this solution, applied on full-scale, would imply huge water consumption and
problems in environmental and economical feasibility. Traditional biogas plants produce abundant effluents, i.e. digestates,
which have normally a pH of 7-8, a considerable alkaline buffer-
capacity and low concentrations of volatile fatty acids, as they were
transformed into methane.
This paper aims to investigate the use of pre-heated (100ºC)
digestates to control the hydrogen production process in a semi-
continuous thermophilic bio-digester, by diluting the liquid medium
and buffering the pH to the desired values.

2 Methods 2.1.Seed microorganisms The seed sludge was the digestate taken from an anaerobic digester
biogas plant and boiled (100ºC) for 2 hours to inactivate
hydrogenotrophic bacteria and to harvest anaerobic spore-forming
bacteria such as Clostridium sp. [7]. This procedure was carried out
twice a day, before feeding. The pH, alkalinity, total VFA
concentration and volatile solids (VS) concentration of the sludge
were respectively 8.1, 7470 mg CaCO3 l -1, 1190 mg l-1 and 750 mg l- 1. 2.2.Feedstock for feeding A solution of 100 g l -1 of pure glucose (99%) was used as feeding mixture, in order to represent an extreme condition for concentration
of sugar-type substrate in the feedstock. The glucose was also chosen
as known substrate for better understanding the process
performances. 2.3. Experimental setup and procedure A completely mixed reactor with working volume of 600 ml was
operated in a semi-continuous mode by feeding twice a day a mix of
the glucose solution and the heat-shocked digestate by a syringe. The digester was operated at a temperature of 55ºC (thermophilic
conditions) and a HRT of 3 days. Output digestate was withdrawn
twice a day before each alimentation. 2.4. Experimental conditions Two experiments were performed to test the process behaviour in
producing biohydrogen. The aim was to find the maximum OLR and
the minimum recirculation ratio (R), i.e. the digestate input volume
divided by the total input volume. The digester was fed in the first
two-week period with an OLR of 10 gVS l -1 d-1 and in the second period with an OLR of 5 gVS l -1 d-1. The recirculation ratios were of 0.7 and 0.85 respectively. 2.5. Analytical procedures The biogas production was measured using a volumetric gas meter
column connected to the headspace of the digester. Biogas
composition was analyzed by a gas chromatograph (Agilent, Micro
GC 3000A) equipped with two thermal conductivity detectors (TCD)
and two columns, using Nitrogen and Helium as carriers.

3 Results and discussion 3.1.H2 production In Figure 2 are reported the results of the process performance in
terms of H2 production per mole of glucose added and the hydrogen
production rate. Methane was never found in the biogas, indicating
complete inhibition of the hydrogen-consumers microrganisms.
Figure 3 shows the pH trend and the total VFA concentration in the
digester. 3.2. First two-week period (OLR=10 gVS l -1 d-1) The first period (hours 0 to 250) was characterized by a lag-phase
during the first week (hours 0 to 115), in which few biogas was
produced (Figure 1). This drove to low hydrogen yields, if compared
with the second week, when the process yielded 1.5-1.8 molH2 mol - 1glucose. This result was quite satisfactory, if compared to the
maximum yields of 2.45 and 2.6 molH2 mol -1glucose achieved in controlled batch cultures by Van Ginkel et al. and Taguchi et al. [12 -
13]. Hang-sik and Jong-Ho [10] obtained a calculated value of 2.2
molH2 mol -1exose, feeding mixed food waste continuously, with an operative HRT of 5 days and adjusting the pH with specific pure
reagents (KOH and HCl).
The hydrogen production rate showed notable imbalance, showing,
during the second week, maximum peaks around 150 and 200 mlH2
l -1 h-1and null values. The obtained maximum rates were four times higher than those measured by Hang-sik and Jong-Ho [10], in
fermenting mixed food waste, with similar OLR. This was probably
caused by the high availability of glucose to microrganisms, if
compared to more complex organic molecules that must be
hydrolyzed before fermentation.
On the other hand, the use of glucose meant higher shock due to the
high-loading. The fast production of VFA partially inhibited the
fermentation because of high concentrations (around 10 g l-1)
(Figure 1) and dropped twice the pH to 4.7 and 4.8 (Figure 2),
causing process imbalance. The addition of digestate in the feeding
with the ratio R=0.7, resulted non-sufficient to control the process
parameters. Because the digester was semi-continuously fed, this
effect was probably more evident than what would happen in a
continuously-fed system. 3.3.Second two-week period (OLR=5 gVS l-1 d-1) From hour 250 to 500, the OLR was lowered to 5 gVS l -1 d-1. The hydrogen production gave yields almost constantly around 1.7 '' 1.8
molH2 mol -1glucose and rates following a relatively stable trend, with maximum peaks around 80 mlH2 l-1 h -1 ( 2). The variations were probably caused by the semi-continuous feeding. The pH was
maintained always higher than in the first period (Figure 2), between
pH 5.4 and 6.2, meaning that the R ratio (0.85) was the upper limit
for a satisfactory control of the process pH. The VFA concentration
resulted in lower values than in the first period (Figure 2).
These results revealed that the digestate added to the feeding (at the
ratio R) has a remarkable effect on the process control, for both
diluting VFA concentration and buffering the pH to desired values
(5<pH<6). This strategy would probably work better in a
continuously-fed system.

4 Conclusions The studied semi-continuous process, with a fixed HRT of 3 days,
showed imbalance conditions for R between 0.7 and 0.85 and for
OLR between 5 and 10 gVS l -1 d-1. Comparing this process to the one proposed by Han et al. [11], the dilution ratio (D) for optimal
operation, in the present case, would be 0.3 d -1. Using pure water in a leaching-bed system, Han et al. found optimal Ds between 2 and 5 so
that high water consumptions were needed. The use of digestate
instead of water would probably give a better option to both control
the pH and dilute the VFAs, also in a leaching-bed system. Further
research should be carried out to test this strategy for fermenting
various kinds of organic substrates, with different operational
conditions. Fig. 1. Process yield and hydrogen production rate. Fig. 2. Trends of the pH and the total VFA concentrations, measured during the test. 0 50 100 150 200 250 0 200 400 600 hours m L H 2 L -1 h -1 0.0 0.5 1.0 1.5 2.0 m o l H 2 m o l -1 G lu co se H2 Rate Yield 1st period 2nd period 4.5 5 5.5 6 6.5 0 200 400 600 hours p H 0 2 4 6 8 10 12 14 g l -1 pH Total VFA 1st period 2nd period References [1] ZAJIC, J.E., KOSARIC, N., BROSSEAU, J.D.: Microbial
production of Hydrogen. Adv Biochem Eng, 7, 57-109 (1978).

[2] BITTON, G.: Wastewater microbiology. New York: Wiley-Liss,

Hydrogen production from inhibited anaerobic composters. Int J
Hydrogen Energy, 22(6), 563''6 (1997).

[4] LAY, J.J., LEE, Y.J., NOIKE, T.: Feasibility of biological
hydrogen production from organic fraction of municipal solid waste.
Water Res, 33(11), 2579''86 (1999).

[5] LAY, J.J.: Modeling and optimization of anaerobic digested
sludge converting starch to hydrogen. Biotechnol Bioeng, 68, 269''
78 (2000).

[6] LAY, J.J.: Biohydrogen generation by mesophilic anaerobic
fermentation of microcrystalline cellulose. Biotechnol Bioeng, 74(4),
280''7 (2001).

[7] LIU, D., LIU, D., ZENG, R.J., ANGELIDAKI, I.: Hydrogen and
methane production from household solid waste in the two-stage
fermentation process, Water Research, 40, 2230-2236 (2006).

[8] CHANG, F.Y., LIN, C.Y.: Biohydrogen production using an
upflow anaerobic sludge blanket reactor. Int. J. Hydrogen Energy, 29
(1), 33''39 (2004).
[9] OH, S.E., VAN GINKEL, S., LOGAN, B.E.: The relative
effectiveness of pH control and heat treatment for enhancing
biohydrogen gas production. Environ. Sci. Technol., 37 (22), 5186''
5190 (2003).

[10] HANG-SIK, S., JONG-HO, Y.: Conversion of food waste into
hydrogen by thermophilic acidogenesis, Biodegradation, 16, 33''44

[11] HAN, S.K., SHIN, H.S.: Biohydrogen production by anaerobic
fermentation of food waste, International Journal of Hydrogen
Energy, 29, 569 '' 577 (2004).

SAITO, T.; HARA, K.: J. Ferment. Bioeng., 82(1), 80 (1996).

[13] VAN GINKEL, S., SUNG, S., L A Y, J.J.: Biohydrogen
production as a function of pH and substrate concentration, Environ.
Sci. Technol., 35, 4726-4730 (2001).

© Eiom - All rights Reserved     P.IVA 00850640186