Review TRENDS in Biotechnology Vol.22 No.

9 September 2004

Production of bioenergy and

biochemicals from industrial and
agricultural wastewater
Largus T. Angenent1,3, Khursheed Karim1, Muthanna H. Al-Dahhan1, Brian A. Wrenn2,3
and Rosa Domguez-Espinosa4
1
Department of Chemical Engineering, Washington University in St. Louis, One Brookings Drive, St. Louis, MO 63130, USA
2
Department of Civil Engineering, Washington University in St. Louis, One Brookings Drive, St. Louis, MO 63130, USA
3
Environmental Engineering Science Program, Washington University in St. Louis, One Brookings Drive, St. Louis, MO 63130, USA
4
Department of Chemical Engineering, University of Yucatan, Av. Juarez 421 Cd Industrial, Merida Yucatan, 97288 Mexico

The building of a sustainable society will require agricultural wastewater, with the generation of valuable
reduction of dependency on fossil fuels and lowering products. Three of these bioprocessing strategies result in
of the amount of pollution that is generated. Waste- the production of bioenergy (methane, hydrogen, electricity),
water treatment is an area in which these two goals can and the fourth processing strategy produces valuable
be addressed simultaneously. As a result, there has been biochemicals by fermentation. Although phototrophic
a paradigm shift recently, from disposing of waste to processes can also be used to produce hydrogen and
using it. There are several biological processing strat- valuable products, this review focuses exclusively on
egies that produce bioenergy or biochemicals while chemotrophic (i.e. dark) processes; in particular, on recent
treating industrial and agricultural wastewater, includ- technological developments, but barriers to implemen-
ing methanogenic anaerobic digestion, biological hydro- tation and unresolved scientific questions will also be
gen production, microbial fuel cells and fermentation for discussed. For each of the bioprocessing strategies, we will
production of valuable products. However, there are consider whether the technology is ready for full-scale
also scientific and technical barriers to the implemen- implementation, whether the products can be easily
tation of these strategies. separated from the treated wastewater, whether mixed,
pure or well-defined co-cultures are preferable, and
The bioprocesses that will be used to treat wastewater in whether the products have sufficient value to justify the
the future will be chosen as they have been in the past: added complexity over conventional wastewater treat-
according to technical feasibility, simplicity, economics, ment processes.
societal needs and political priorities. The needs and
political priorities of a sustainable society, however,
Biological methane production from organic material in
provide pressure that will shift the focus on wastewater
industrial and agricultural wastewater
treatment from pollution control to resource exploitation.
Methanogenic anaerobic digestion of organic material in
Many bioprocesses can provide bioenergy or valuable
wastewater (Table 1, reaction 11) has been performed for
chemicals while simultaneously achieving the objective of
about a century and is advantageous over aerobic active
pollution control. Industrial wastewaters, for example,
sludge systems because of its high organic removal rates,
from food-processing industries and breweries, and agri-
low energy-input requirement, energy production
cultural wastewaters from animal confinements are ideal
(i.e. methane), and low sludge production. The food web
candidates for bioprocessing because they contain high of anaerobic digestion is reasonably well understood (Box 1).
levels of easily degradable organic material, which results An important breakthrough was made w30 years ago,
in a net positive energy or economic balance, even when with the development of the upflow anaerobic sludge
heating of the liquid is required. In addition, they have a blanket [UASB] reactor [1], which efficiently retains the
high water content, which circumvents the necessity to complex microbial consortium without the need for
add water. Such wastewaters are potential commodities immobilization on a carrier material (for example, as a
from which bioenergy and biochemicals may be produced. biofilm) by formation of biological granules (i.e. granula-
Recovery of energy and valuable materials might reduce tion; self-immobilization) with good settling character-
the cost of wastewater treatment, and somewhat reduce istics. The mean cell residence time, that is, the average
our dependence on fossil fuels. time a typical microbial cell remains in the reactor, of
This review describes four different bioprocessing UASBs is much longer than the hydraulic residence time
strategies that can be used to treat industrial and (the average time the wastewater remains in the reactor),
Corresponding author: Largus T. Angenent ([email protected]). due to this self-immobilization process. Performance
Available online 23 July 2004 depends on the mean cell residence time and reactor
www.sciencedirect.com 0167-7799/$ - see front matter Q 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.tibtech.2004.07.001
478 Review TRENDS in Biotechnology Vol.22 No.9 September 2004

Table 1. Reactions for the different (bio) process strategies

Biotic or abiotic process Reaction Number (letter, if
shown, is that in
Figure I in Box 1)
Hydrogen fermentation to acetic acid C6H12O6C2H2O $ 4H2C2CH3COOHC2CO2 1 (a)
Hydrogen fermentation to butyric acid C6H12O6$2H2CCH3CH2CH2COOHC2CO2 2 (a)
Fermentation to ethanol C6H12O6 $ 2CH3CH2OHC2CO2 3 (a)
Propionic acid production with hydrogen C6H12O6C2H2 $ 2CH3CH2COOHC2H2O 4 (a)
Ethanol production with hydrogen CH3COOHCH2 $ CH3CH2OHCH2O 5
Syntrophic propionic acid oxidation CH3CH2COOHC2H2O $ CH3COOHC3H2CCO2 6 (b)
Syntrophic butyric acid oxidation CH3CH2CH2COOHC2H2O $ 2CH3COOHC2H2 7 (b)
Syntrophic acetic acid oxidation CH3COOHC2H2O $ 4H2C2CO2 8 (c)
Hydrogenotrophic methanogenesis 4H2CCO2 $ CH4C2H2O 9 (d)
Acetoclastic methanogenesis CH3COOH $ CH4CCO2 10 (e)
Methane formation from glucose C6H12O6 $ 3CH4C3CO2 11 (a,b,c,d,e)
Catalytic methane conversion to syngas CH4CH2O $ 3H2CCO 12
Catalytic gas-shift reaction COCH2O $ H2CCO2 13
Hydrogen fuel cell 2H2CO2 $ 2H2OCelectricity 14
Methane fuel cell CH4C2O2 $ CO2C2H2OCelectricity 15
MFC C6H12O6C6O2 $ 6CO2C6H2OCelectricity 16
Cellulose bioconversion [C6H11O6]nCaH2O $ bCH3COOHCcCH3CH2OHCdCO2CeH2 17
Polyhydroxyalkanoates formation aC6H12O6CbO2 $ [COOH(CH2)3COO]nCcH2O 18
Dyes formation [C6H11O6]nCaNHK
3 CbH2O $ cC23H26O5CdC22H27O5 NCeC21H22O5 19

volume depends on the hydraulic residence time, there- horizontal-flow bioreactor that incorporated a migrating
fore, UASBs can efficiently convert wastewater organic blanket within a compartmentalized reactor (a multi-
compounds into methane in small high-rate reactors. vessel; anaerobic migrating blanket reactor [AMBR]) [4].
Approximately 60% of the thousands of anaerobic full- Due to the AMBRs inherent dynamic conditions, the
scale treatment facilities worldwide are now based on the organic removal rates are higher than those in UASB
UASB design concept, treating a diverse range of reactors [5,6].
industrial wastewaters [2,3]. A significant limitation of UASB reactors is the
It was originally thought that a continuous upflow- interference of suspended solids in the incoming waste-
hydraulic pattern was required for granulation, but this water with granulation and reactor performance [7].
phenomenon was recently observed in a continuously-fed Hence, other high-rate systems, such as the anaerobic

Box 1. Food web of methanogenic anaerobic digestion

Anaerobic bioconversion of complex organic material to methane
requires four major steps and five physiologically distinct groups
of microorganisms. Elements of the food web of methanogenic POLYMERS
anaerobic digestion are expected to occur also for biological Proteins, polysaccharides, lipids
hydrogen production, microbial fuel cells and biochemical pro-
duction. As shown in Figure I, complex organic polymers (i) Hydrolysis a
(e.g. proteins, polysaccharides) are hydrolyzed to monomers by
fermentative bacteria (a), which ferment the monomers to a MONOMERS & OLIGOMERS
mixture of low-molecular-weight organic acids and alcohols.
amino acids, sugars, fatty acids
These fermentation products are further oxidized to acetic acid
and hydrogen by obligatory hydrogen-producing acetogenic a
(ii) Fermentation
bacteria (b) through a process called acetogenesis. Acetogenesis
also includes acetate production from hydrogen and carbon a
a INTERMEDIATES
dioxide by acetogens and homoacetogens (c). Hydrogen-produ-
cing acetogenic bacteria (b) grow in syntrophic associations with Propionate, butyrate, alcohols
hydrogenotrophic methanogens (d), which keep the hydrogen
partial pressure low enough to allow acetogenesis to become
thermodynamically favorable (this process is referred to as b b
interspecies hydrogen transfer) [58]. Finally, acetoclastic methano- (iii) Acetogenesis
gens (e) convert the acetate to methane and carbon dioxide H2 + CO2 ACETATE
(methanogenesis). Although w70% of methane produced in many c
natural and engineered systems is due to acetoclastic methano-
(iv) Methanogenesis e
gens, it is increasingly clear that many stressed and thermophilic d
systems use an alternative pathway: syntrophic oxidation of
acetate to carbon dioxide and hydrogen by acetogenic or CH4 + CO2
homoacetogenic bacteria (c) coupled to hydrogen consumption
by hydrogenotrophic methanogens [8,59]. TRENDS in Biotechnology

Figure I. Intricate food web of methanogenic anaerobic digestion. Several

trophic groups of microorganisms work together to convert complex organic
material into methane and carbon dioxide

www.sciencedirect.com
 Review TRENDS in Biotechnology Vol.22 No.9 September 2004 479

sequencing batch reactor (ASBR), were developed to equivalents that are present in wastewater. For example,
better handle high-suspended solids in wastewater. optimization of hydrogen production from related hexoses,
ASBRs are single-vessel bioreactors that operate in a at best, results in production of four moles of hydrogen per
four-step cycle: (i) wastewater is fed into the reactor with mole of hexose, because two moles of acetate are also
settled biomass; (ii) wastewater and biomass are mixed formed (Table 1, reaction 1). Complete oxidation to carbon
intermittently; (iii) biomass is settled and; (iv) effluent is dioxide and hydrogen, however, would produce 12 moles of
withdrawn from the reactor. ASBRs are particularly hydrogen per mole hexose (reaction 1 plus reaction 8 in
useful for agricultural waste, and it has recently been Table 1). Actual yields are even lower than the four moles
scaled up for on-farm treatment of dilute swine waste [8]. of hydrogen that are theoretically possible, typically
The methane that is produced by anaerobic digestion ranging from !1 to w2.5 moles hydrogen per mole hexose
has traditionally been used as a fuel source, usually, for (Table 2). When butyric acid is produced as a major
on-site heating or electricity production. Recently, fermentation product, only two moles of hydrogen can be
methane has also been converted to other useful products, produced (Table 1, reaction 2). Hydrogen yield is even
such as methanol for use in production of biodiesel, for lower when more reduced organic compounds, such as
example, by production of syngas (a mixture of hydrogen lactic acid, propionic acid, and ethanol, are produced as
and carbon monoxide; Table 1, reactions 12 and 13) in fermentation products, because these represent end
downstream chemical processes [9]. Production of syngas products of metabolic pathways that bypass the major
requires the removal of impurities, such as hydrogen hydrogen-producing reaction in carbohydrate fermenta-
sulfide, in the digester biogas, which can poison the tions (Table 1, reaction 3). Thermodynamic limitations for
catalyst. In addition, direct conversion of methane to hydrogen production are explained in Box 2.
electricity in solid-oxide fuel cells after a single-step It appears that, even under optimized conditions, one
anaerobic digester might soon become feasible (Table 1, cannot expect to recover more than w15% of the electron
reaction 15) [10]. equivalents in a high-carbohydrate wastewater as hydro-
gen, thus, it is not surprising that several research groups
Biological hydrogen production are considering implementing two-step processes, invol-
Much recent interest has been expressed in the biological ving biohydrogen production followed by methanogenic
production of hydrogen from wastewater by dark fermen- anaerobic digestion to increase the energy yield of the
tation, due to its potential importance in our economy overall process [11,21]. As described previously, methano-
[1114]. Biological hydrogen production shares many genic anaerobic digestion is a mature, reliable technology
common features with methanogenic anaerobic digestion, that has been demonstrated in thousands of full-scale
especially the relative ease with which the two gaseous facilities worldwide. Catalytic conversion of methane to
products can be separated from the treated wastewater. hydrogen gas is also a well-developed and reliable process
The mixed communities involved in both bioprocesses share (Table 1, reactions 12 and 13) [9]. Therefore, direct
some common elements but with one important difference: biological production of hydrogen through dark fermenta-
successful biological hydrogen production requires inhi- tion appears to be restricted to a pre-treatment step in a
bition of hydrogen-using microorganisms, such as homo- larger bioenergy or biochemical production concept.
acetogens (see Figure I in Box 1, group c) and methanogens Another anticipated disadvantage of large-scale hydrogen
(see Figure I in Box 1, group d). Inhibition is commonly production that needs to be addressed during scale-up is
accomplished by heat treatment of the inoculum to kill all the escape of hydrogen through large plastic enclosures
microorganisms except for spore-forming fermenting bac- and thin metal sheets that might occur due to the high
teria (for example, species from the families Clostridiaceae, diffusivity of hydrogen.
Streptococcaceae, Sporolactobacillaceae, Lachnospiraceae,
and Thermoanaerobacteriacea [1517]) (see Figure Ia in Biological electricity production
Box 1). Other methods that have been used include the Microbial fuel cells (MFCs) have also been suggested as an
operation of reactors at high dilution rates [18] or low pH [19]. alternative to follow biohydrogen production in a two-step
Considerable effort has been devoted to optimizing treatment process [22]. Many researchers, however, have
operational environmental conditions to maximize hydro- successfully generated electricity biologically in a single-
gen production (examples of optimization efforts are given step process (Table 1, reaction 16) [2326]. Figure 1 shows
in Table 2). Conceptually, important efforts are those that a generic schematic of how an MFC works. In principle,
prevent consumption of hydrogen by, for example, propio- MFCs are similar to hydrogen fuel cells, which channel
nic acid-producing bacteria, ethanol-producing bacteria, protons from an anode compartment to a cathode
and homoacetogens (Table 1, reaction 4, reaction 5, and compartment through an electrolyte membrane
the reverse of reaction 8, respectively); and those that (i.e. electronically insulated proton-exchange membrane)
channel more reducing equivalents towards reduction of with electrons going in the same direction via a conductive
protons by hydrogenases (see Box 2). Operating bio- wire. A hydrogen fuel cell oxidizes hydrogen to electrons
reactors at low hydrogen partial pressure, perhaps by and protons on the anode and reduces oxygen to water on
sparging with nitrogen gas to strip hydrogen from the the cathode (Table 1, reaction 14). Gas-permeable noble
solution as fast as it is produced [19,20], accomplishes both metals are used as electro-catalysts on the anode and
efforts simultaneously. cathode sides [27]. In an MFC, conversely, anaerobic
Unfortunately, optimization of biohydrogen production microorganisms oxidize organic material in the anode
focuses on a relatively small fraction of the total hydrogen chamber and they transfer the derived reducing
www.sciencedirect.com
480 Review TRENDS in Biotechnology Vol.22 No.9 September 2004

Box 2. Physiological limitations for biological production of hydrogen

Hydrogen production from organic substrates is limited by the ferredoxin to oxidized ferredoxin is greater than one. The free
thermodynamics of the hydrogenase reaction, which involves the energy change of the pyruvate-ferredoxin oxido reductase reaction
enzyme-catalyzed transfer of electrons from an intracellular electron (DG o; 0 ZK2.1 kcal/mole) is sufficient to allow the reaction to
carrier molecule to protons. Unfortunately, protons are poor electron proceed with more than a tenfold excess of products over
0
acceptors (EH2 ZK414 mV); so, the electron donor must be a strong reactants. Hence, the frequent observation of more than 30%
0
reducing agent. Ferredoxin is a low-potential (EoFd zK400 mV, depend- hydrogen in reactor headspace is not unexpected.
ing on source) iron-sulfur protein that is capable of reducing protons to In most systems for biological production of hydrogen, all of the
hydrogen [60]. Another important intracellular electron carrier, NADH, observed hydrogen can be attributed to electrons derived from a single
0
has a higher redox potential (EoNADH ZK320 mV). The ability of reduced reaction: oxidative decarboxylation of pyruvate by pyruvate:ferredoxin
ferredoxin and NADH to reduce protons is determined by the redox oxidoreductase (Figure II). Hexoses can be metabolized to pyruvate
potential of the net reaction under actual conditions. Assuming the through several pathways, often involving the Embden-Meyerhoff-
intracellular concentrations of the oxidized and reduced forms of Parnas (i.e. glycolysis) or the Entner-Doudoroff pathways. Both of
ferredoxin and NADH are about equal, hydrogen production becomes these pathways produce two moles of pyruvate and two moles of
thermodynamically unfavorable at hydrogen partial pressures greater NADH for every mole of hexose that is transformed [60]. Therefore,
than: hexose metabolism by bacteria that contain pyruvate:ferredoxin
( 0 )
0
2F EHo2 K Exo oxidoreductase can result in formation of 2 moles of hydrogen per
PH2 ;max % exp mole of hexose. If the hydrogen partial pressure is sufficiently low
RT
(!60 Pa), the NADH that is produced may also be used to generate
0
where EoX is the redox potential of the electron donor, F is Faradays hydrogen (at best, an additional 2 moles of hydrogen per mole of
constant, R is the ideal gas constant, and T is the absolute temperature. hexose), but most of the NADH will probably be oxidized through other
For ferredoxin, hydrogen production can continue as long as the fermentation pathways, such as butyrate fermentation (Figure II,
hydrogen partial pressure is less than w0.3 atm (3!104 Pa); for NADH, step 4). Some fermentation products (e.g. ethanol and lactate)
the partial pressure of hydrogen must be less than w6!10K4 atm represent the operation of alternative pathways for pyruvate metab-
(60 Pa). Note that these values assume equal concentrations of the olism that compete with pyruvate:ferredoxin oxidoreductase. As such,
oxidized and reduced forms of the electron donors. Higher they will usually be associated with systems that produce less than two
hydrogen partial pressures can be achieved if the ratio of reduced moles of hydrogen per mole of hexose.

(a) PH < 60 Pa:

2
2 CO2

Glucose 2 pyruvate 2 acetate

1 2
2 NAD+ 2 NADH 2 Fdox 2 Fdred

3 3
2 H2 2 H+ 2 H2 4 H+

(b) PH2> 60 Pa:

2 CoASH 2 CO2 2 CoASH

Glucose 2 pyruvate 2 acetyl-CoA butyrate

1 2 4
2 NAD+ 2 NADH 2 Fdox 2 Fdred 2 NADH 2 NAD+

3
2 H2 4 H+
TRENDS in Biotechnology

Figure II. Effect of hydrogen partial pressure on biological production of hydrogen. (a) Oxidation of NADH by production of hydrogen is thermodynamically favorable only
when the hydrogen partial pressure is less than 60 Pa, otherwise; (b) other fermentation products must be formed. Reactions [(a) and (b)]: 1, glucose metabolism through
glycolysis or the Entner-Doudoroff pathway; 2, oxidative decarboxylation of pyruvate by pyruvate:ferredoxin oxidoreductase; 3, formation of hydrogen by hydrogenase;
4, butyrate fermentation.

equivalents (electrons) to an electrode rather than to an Shewanella spp., can transfer electrons to solid acceptors
electron-acceptor molecule. through either direct contact or by production of soluble
Several mechanisms have been described by which quinones that act as electron-shuttling compounds
electrons can be transferred to metals or electrodes. For [31,32]. For organisms of this type, direct contact with
example, the dissimilatory metal-reducing bacterium the electrode surface is not required.
(DMRB) Geobacter spp. makes direct physical contact In addition to microorganisms that can transfer
with solid electron acceptors and uses periplasmic c-type electrons to the anode, the presence of other organisms
cytochrome proteins as the metal reductase [28,29]; appears to benefit MFC performance. One research group
hence, these organisms must grow as a biofilm on the found that a mixed culture generated a current that was
electrode surface [30]. Other types of DMRB, such as sixfold higher than that generated by a pure culture [33].
www.sciencedirect.com
 Review TRENDS in Biotechnology Vol.22 No.9 September 2004 481

Table 2. Maximum hydrogen yields achieved from organic material by a mixed culture performing dark fermentation during
optimization efforts
Optimization effort Reactor type Substrate Max. Hydrogen yield References
(mol/mol) based on
hexose
Initial pH and acetic/butyric acid Batch Sucrose/starch 1.8 [61]
Reactor configuration Fluidized bed reactor Sucrose 1.3 [62]
Hydrogen partial pressure CSTR Wheat starch 1.9 [20]
Inhibition of acetic/butyric acid Batch-fed Glucose 2.0 [63]
Reactor operation, temperature Upflow reactor Wastewater 2.1 [13]
Immobilized biomass Batch Sucrose 2.0 [64]
Immobilized biomass, granules Fermentor Sucrose 2.1 [16]
pH Fermentor Glucose 2.1 [65]
Hydrogen partial pressure, substrate Batch Sucrose, lactate 0.5 [11]
Hydraulic retention time CSTR Glucose, sucrose 2.2 [66]
Peptone addition Batch/chemostat Cellulose 2.0 [15]
Nitrogen source Batch Glucose 2.4 [67]
pH and substrate levels Batch Sucrose 2.5 [21]
Hydrogen partial pressure CSTR Glucose 1.4 [68]

Anodophilic biofilm
e
e

Effluent H2O
H+

H+

H+
Wastewater

O2
Anode Membrane Cathode
TRENDS in Biotechnology

Figure 1. Schematic of a dual-chamber microbial fuel cell (MFC). Anaerobic microorganisms oxidize organic material in the anode chamber. Some microorganisms attach to
the electrode as a biofilm and the electrons are transferred directly to the electrode and then through a wire to the cathode where the electrons reduce oxygen to water.
Protons are transferred in the same direction through a proton-exchange membrane. Mixing is required in the anode chamber, whereas aeration is required in the cathode
chamber.

Hence, the microbial communities that develop in the Aeromonadaceae, and Comamonadaceae were able to
anode chamber may have a similar function as those found transfer electrons to electrodes [24,26,30,3338].
in methanogenic anaerobic digesters, except that micro- The major limitations to implementation of MFCs for
organisms that can transfer electrons to the electrode treatment of wastewater are that their power density is
surface replace methanogens. Rabaey and co-workers [23] still relatively low and the technology is only in the
referred to such microbial communities as adapted laboratory phase. Based on the potential difference, DE,
anodophilic consortia. Anodophilic bacteria from between the electron donor and acceptor, a maximum
different evolutionary lineages from the families of potential of w1 V can be expected in MFCs, which is not
Geobacteraceae, Desulfuromonaceae, Alteromonadaceae, much greater than the 0.7 V that is currently being
Enterobacteriaceae, Pasteurellaceae, Clostridiaceae, produced [39]. However, by linking several MFCs

Table 3. Maximum power densities achieved in MFCs using various substrates during optimization efforts
Optimization effort Chamber Inoculum Substrate Max. power densitya References
(mW/m2)
Cathode performance Single Wastewater Wastewater 26 [40]
Anodophile consortium selection Dual Anaerobic sludge Glucose 3,600 [23]
MFC configuration, electrodes Single Anaerobic sludge Glucose 788 [39]
Specific inoculum Dual Geobacter sp. Acetic acid 15 [30]
Electrode surface area, substrate Dual Rhodoferax sp. Glucose 8.2 [26]
Electrode material Single Anaerobic sludge Lactate 10.2 [33]
Concentration of bacteria Dual Shewanella sp. Lactate 3.6 [37]
Electron shuttles, physiological state Dual Escherichia coli Glucose 0.5 [36]
a
Expressed as power density, which is power per unit area of anode electrode surface.

www.sciencedirect.com
482 Review TRENDS in Biotechnology Vol.22 No.9 September 2004

together, the voltage can be increased. Currents and Strategies to enhance bioconversion of organic material
power densities, however, are lower than what is theor- from waste streams are the main focus of the studies
etically possible, and system performance varies consider- reported in the literature. The poor conversion yield is
ably (Table 3). The maximum power density reported in perhaps the greatest barrier, and therefore much effort
the literature (3600 mW/m2) was observed in a dual- has been devoted to enhancing the amount of product
chamber fuel-cell treating glucose with an adapted formed per reactor volume, per time period. The main
anaerobic consortium in the anode chamber and a approaches to achieving this goal have been discovery of
continuously aerated cathode chamber containing an novel microorganisms, metabolic engineering for heter-
electrolyte solution that was formulated to improve ologous gene expression, protein engineering by mutagen-
oxygen transfer to the cathode. Costs were minimized by esis and molecular evolution, manipulation of cell
using plain graphite electrodes without exogenous elec- genomes, and control of carbon fluxes by the addition of
tron shuttles and a commercially available exchange external substrate and inhibitors [4548].
membrane [Ultrexe ion-exchange membrane (Membranes Bioconversion may also be enhanced by process
International, Inc.; http://www.membranesinternational. modification, such as culture immobilization [49,50] or
com/)] [23]. Further improvement in power density is coupling two separate bioreactors. For example, anaerobic
required, and the rates of electron transport to the anode (hydrogen) fermentation and aerobic conversion of volatile
electrode are thought to be a major limiting factor [23]. fatty acids by Ralstonia eutropha were combined to
Optimization of MFCs involves investigation of a variety enhance the efficiency of polyhydroxyalkanoate pro-
of aspects of their operation, as given in Table 3. duction from food wastes [51]. Another promising strategy
For wastewater treatment in MFCs to be feasible, the for increasing waste bioconversion rates is the use of co-
construction and operating costs must be reduced [23,40]. cultures of pure microbial strains in a single process.
The requirement for expensive noble metals in electrodes, Within this strategy, one species performs most of the
and soluble or electrode-bound electron shuttles are two complex nutrient hydrolysis, and, in turn, provides its
important cost elements that are being addressed by metabolic byproducts to the second species, which forms
current researchers [23,39,40]. Single-chamber MFCs the desired product [52]. This study achieved high acetic
with air as the cathode chamber circumvents the cost of acid yields during bioconversion of milk permeate by
aeration [39,40]. In addition, the rate of electron transport combining cells of Clostridium thermolacticum, Moorella
must be improved; this may be achieved by selecting a thermoautotrophica, and Methanothermobacter thermo-
well-adapted anodophilic microbial community and opti- autotrophicus in a consortium. Addition of the hydrogeno-
mizing the MFC operating conditions. Optimization can trophic methanogen decreased the hydrogen partial
be conducted in a systematic fashion only when the pressure, which increased the acetic acid production.
mechanisms of electron transfer from microorganism to Another potential advantage of co-cultures is that they
electrode are better understood. Even when optimization might be less sensitive than single cultures to changes in
is achieved, it remains to be seen whether reactor size can composition of industrial and agricultural wastewaters
be small enough to make direct bioelectricity production (i.e. the wastewater may contain multiple organic com-
by MFCs economically viable. pounds, the relative concentrations of which may change
on a diurnal cycle). Stability in biological waste treatment
Biological chemical production systems has been linked to increased biodiversity of the
A major limitation to wider application of the bioenergy microbial community [53] and to an increased ability of
technologies described in the previous sections is the the community to use alternative removal pathways for
relative low cost of the current non-renewable energy the same substrate [54].
sources. Government subsidies, or a direct local need to In bioconversion of agricultural and food wastes to
save on energy costs (for example, biogas that is used valuable chemicals, separation and purification of the
directly on-site as a fuel), are necessary to make those products from the bulk liquid represents the highest
processes economically viable. In addition, although percentage of the manufacturing cost. Therefore, the
bioenergy production may reduce the cost of wastewater economic feasibility of reusing wastes will strongly depend
treatment, it cannot entirely satisfy the energy demand of on the downstream processing efficiency. In recent years,
our society. Therefore, the production of high-value considerable advances have been achieved in biomolecule
chemicals from organic material in wastewater might be purification technologies; one of the key goals is to achieve
more feasible than bioenergy production [41]. Industrial more selective, more efficient, and shorter separation
wastes can become inexpensive raw materials for inte- routes [55]. For example, a single-step direct lactic acid
grated fermentation processes [4244]. High-carbo- separation from fermentation broths has been success-
hydrate wastewaters that are unsuitable for animal or fully implemented using anionic fluidized-bed columns
human feeding are particularly appropriate for conversion [56]. Supercritical fluid technology (i.e. the use of a fluid at
to valuable products in pure-culture or co-culture pro- a temperature and pressure that are greater than its
cesses (e.g. Table 1, reactions 1719). It is the cost critical levels) has also proven to be a useful tool for
efficiency of the bioconversion process that ultimately achieving separation of compounds directly from cultures,
determines whether a specific waste stream is suitable and is an attractive option because separation is usually
for production of a specific product [42]. Examples of performed at ambient temperatures using non-toxic,
biochemicals that are obtained via fermentation of organic non-flammable solvents, while achieving product
material in waste streams are shown in Table 4. crystallization [57].
www.sciencedirect.com
 Review TRENDS in Biotechnology Vol.22 No.9 September 2004 483

Table 4. Biochemical production from agricultural and food industry waste streams by different microbial strains
Waste stream Industry Biochemical Function Microbial strain(s)a Refer-
ences
Corn steep liquor Corn wet milling Lipase Enzyme Galactomyces geotrichumb [69]
Whey Wheat milling Protease Enzyme Mucor spp.b [70]
Gluten free effluent Wheat milling Glycerol Solvent Pichia farinosab [43]
Sugar cane bagasse Sugar Xylitol Sweetener Pichia guilliermondiib [49]
Starch-processing Starch Biomass protein Feed 6 microfungi (Rhizopus sp.) [71]
Salad oil Vegetable oil Biomass protein Feed 5 yeasts [72]
Gluten free effluent Wheat milling Monascus pigment Dye Monascus purpureusb [73]
Milk permeate Dairy Acetic acid Deicing salt Clostridium stercorarium subsp. [52]
Thermolacticumc plus Moorella
thermoautotrophicac plus
Methanothermobacter thermoautotrophicusd,e
Gluten free effluent Wheat milling Lactic acid Bioplastic Lactobacillus bulgaricusc [43]
Food Agricultural Polyhydroxyalkanoates Bioplastic Ralstonia eutrophac [51]
Whey Dairy Polyhydroxyalkanoates Bioplastic Recombinant E. colic [47]
Mussel-processing Fish Nisin, Pediocin Antibacterials Lactococcus lactisc, Pediococcus acidilacticic [74]
a
According to current nomenclature
b
Fungal strain
c
Bacterium
d
Archaeaon
e
In a co-culture

Table 5. Comparison of the four bioprocessing strategies for wastewater treatment

Bioprocess strategy Level of maturity Separation of products Culture Value added
Anaerobic digestion Mature, operational Easy, gas Mixed Low
Hydrogen fermentation Laboratory phase Easy, gas Mixed Low to medium
MFC Laboratory phase Easy, electricity Mixed Low
Biochemical production Scale-up phase Hard, soluble products Pure or co Medium to high

Outlook soluble products from the fermentation broth, and the

Anaerobic digestion of industrial and agricultural waste- stability of pure- or co-culture fermentation processes
water to methane is a mature process that is being used remain to be addressed. Furthermore, the effect of
within full-scale facilities worldwide. Although methane is bioreactor design configuration and operating conditions
a low-value product (Table 5), methanogenic anaerobic (e.g. biomass retention, mixing) need to be more widely
digestion may still represent the most economically viable studied and optimized before the processes can be scaled
technology, because catalytic conversion of biogas up. The biological oxygen demand in the effluent, which is
(i.e. methane) to syngas (i.e. hydrogen and carbon an indication of how well wastewater is treated, will be too
monoxide) is relatively simple. Syngas can be used to high in all four different bioprocessing strategies, and
produce liquid fuel and high-value products through thus, post-treatment with, for example, activated sludge,
conventional chemical manufacturing processes. Hydro- is an anticipated requirement. Hence, post-treatment
gen production via dark fermentation has the greatest facilities must to be integrated into the design of full-
potential as a pre-treatment step that can be followed by a scale bioprocessing operations.
suitable secondary process step, such as bioconversion of
volatile fatty acids to other products, such as polyhydroxy-
Acknowledgements
alkanoates. The possibility of direct conversion of organic We acknowledge the support provided by DOE to Dr. Khursheed Karim as
material in wastewater to bioelectricity is exciting, but part of the grant DE-FC3601GO11054. In addition, we acknowledge the
fundamental understanding of the microbiology and helpful input from anonymous reviewers.
further development of the technology is required.
Methanogenic anaerobic digestion, hydrogen fermenta-
References
tion and bioelectricity production all use a mixed microbial 1 Lettinga, G. et al. (1980) Use of the upflow sludge blanket (USB)
community that is selected according to function. This is reactor concept for biological wastewater treatment, especially for
well-suited to the non-sterile, ever-changing, complex anaerobic treatment. Biotechnol. Bioeng. 22, 699734
environment of wastewater treatment. Also, the products 2 Jantsch, T.G. et al. (2002) Anaerobic biodegradation of spent sulphite
liquor in a UASB reactor. Bioresour. Technol. 84, 1520
from these bioprocesses can be easily separated as gases
3 Karim, K. and Gupta, S.K. (2003) Continuous biotransformation and
or bioelectricity (Table 5). However, in our current, removal of nitrophenols under denitrifying conditions. Water Res. 37,
cheap-energy economy, such large-scale implementation 29532959
may not be economically feasible without further conver- 4 Angenent, L.T. et al. (2004) Formation of granules and Methanosaeta
sion to more valuable products. Fortunately, specialized fibres in an anaerobic migrating blanket reactor (AMBR). Environ.
Microbiol. 6, 315322
high-value biochemicals might soon be produced from 5 Angenent, L.T. and Sung, S. (2001) Development of anaerobic
wastewater to elevate bioprocessing as an economically migrating blanket reactor (AMBR), a novel anaerobic treatment
viable choice. Issues concerning the separation of the system. Water Res. 35, 17391747
www.sciencedirect.com
484 Review TRENDS in Biotechnology Vol.22 No.9 September 2004

6 Angenent, L.T. et al. (2002) Microbial community structure and 32 Newman, D.K. and Kolter, R. (2000) A role for excreted quinones in
activity in a compartmentalized, anaerobic bioreactor. Water Environ. extracellular electron transfer. Nature 405, 9497
Res. 74, 450461 33 Park, D.H. and Zeikus, J.G. (2002) Impact of electrode composition on
7 Kalogo, Y. and Verstraete, W. (1999) Development of anaerobic sludge electricity generation in a single-compartment fuel cell using
bed (ASB) reactor technologies for domestic wastewater treatment: Shewanella putrefaciens. Appl. Microbiol. Biotechnol. 59, 5861
motives and perspectives. World J. Microbiol. Biotechnol. 15, 523534 34 Bond, D.R. et al. (2002) Electrode-reducing microorganisms that
8 Angenent, L.T. et al. (2002) Methanogenic population dynamics harvest energy from marine sediments. Science 295, 483485
during startup of a full-scale anaerobic sequencing batch reactor 35 Kim, N. et al. (2000) Effect of initial carbon sources on the performance
treating swine waste. Water Res. 36, 46484654 of microbial fuel cells containing Proteus vulgaris. Biotechnol. Bioeng.
9 Witt, P.M. and Schmidt, L.D. (1996) Effect of flow rate on the partial 70, 109114
oxidation of methane and ethane. J. Catal. 163, 465475 36 Park, D.H. and Zeikus, J.G. (2000) Electricity generation in microbial
10 Tao, S. and Irvine, J.T. (2003) A redox-stable efficient anode for solid- fuel cells using neutral red as an electronophore. Appl. Environ.
oxide fuel cells. Nat. Mater. 2, 320323 Microbiol. 66, 12921297
11 Logan, B.E. et al. (2002) Biological hydrogen production measured in 37 Kim, H.J. et al. (2002) A mediator-less microbial fuel cell using a metal
batch anaerobic respirometers. Environ. Sci. Technol. 36, 25302535 reducing bacterium, Shewanella putrefaciens. Enzyme Microb. Tech-
12 Yokoi, H. et al. (2002) Microbial production of hydrogen from starch- nol. 30, 145152
manufacturing wastes. Biomass Bioenergy 22, 389395 38 Pham, C.A. et al. (2003) A novel electrochemically active and Fe(III)-
13 Yu, H. et al. (2002) Hydrogen production from rice winery wastewater reducing bacterium phylogenetically related to Aeromonas hydro-
in an upflow anaerobic reactor by using mixed anaerobic cultures. Int. phila, isolated from a microbial fuel cell. FEMS Microbiol. Lett. 223,
J. Hydrogen Energy 27, 13591365 129134
14 Zhang, T. et al. (2003) Biohydrogen production from starch in 39 Park, D.H. and Zeikus, J.G. (2003) Improved fuel cell and electrode
wastewater under thermophilic condition. J. Environ. Manage. 69, designs for producing electricity from microbial degradation. Biotech-
149156 nol. Bioeng. 81, 348355
15 Ueno, Y. et al. (2001) Microbial community in anaerobic hydrogen- 40 Liu, H. et al. (2004) Production of electricity during wastewater
producing microflora enriched from sludge compost. Appl. Microbiol. treatment using a single chamber microbial fuel cell. Environ. Sci.
Biotechnol. 57, 555562 Technol. 38, 22812285
16 Fang, H.H. et al. (2002) Characterization of a hydrogen-producing 41 Bungay, H.R. (2004) Confessions of a bioenergy advocate. Trends
granular sludge. Biotechnol. Bioeng. 78, 4452 Biotechnol. 22, 6771
17 Fang, H.H. et al. (2002) Microbial diversity of a mesophilic hydrogen- 42 Laufenberg, G. et al. (2003) Transformation of vegetable waste into
producing sludge. Appl. Microbiol. Biotechnol. 58, 112118 value added products: (A) the upgrading concept; (B) practical
18 Chen, C.C. et al. (2001) Kinetics of hydrogen production with implementations. Bioresour. Technol. 87, 167198
continuous anaerobic cultures utilizing sucrose as the limiting 43 Wang, R. et al. (2002) The application of a generic feedstock from
substrate. Appl. Microbiol. Biotechnol. 57, 5664 wheat for microbial fermentations. Biotechnol. Prog. 18, 10331038
19 Oh, S.E. et al. (2003) The relative effectiveness of pH control and heat 44 Montgomery, R. (2004) Development of biobased products. Bioresour.
treatment for enhancing biohydrogen gas production. Environ. Sci. Technol. 91, 129
Technol. 37, 51865190 45 Bramucci, M.G. and Nagarajan, V. (2000) Industrial wastewater
20 Hussy, I. et al. (2003) Continuous fermentative hydrogen production bioreactors: sources of novel microorganisms for biotechnology. Trends
from a wheat starch co-product by mixed microflora. Biotechnol. Biotechnol. 18, 501505
Bioeng. 84, 619626 46 Aldor, I.S. and Keasling, J.D. (2003) Process design for microbial
21 Ginkel, S.V. et al. (2001) Biohydrogen production as a function of pH plastic factories: metabolic engineering of polyhydroxyalkanoates.
and substrate concentration. Environ. Sci. Technol. 35, 47264730 Curr. Opin. Biotechnol. 14, 475483
22 Logan, B.E. (2004) Extracting hydrogen and electricity from renew- 47 Park, S.J. et al. (2002) Production of poly(3-hydroxybutyrate) from
able resources: a roadmap for establishing sustainable processes. whey by fed-batch culture of recombinant Escherichia coli in a pilot-
Environ. Sci. Technol. 38, 160A167A scale fermenter. Biotechnol. Lett. 24, 185189
23 Rabaey, K. et al. (2003) A microbial fuel cell capable of converting 48 van Hoek, P. et al. (2003) Fermentation goes large-scale. Chem. Eng.
glucose to electricity at high rate and efficiency. Biotechnol. Lett. 25, Prog. 99, 37S42S
15311535 49 Carvalho, W. et al. (2002) Metabolic behavior of immobilized Candida
24 Park, H.S. et al. (2001) A novel electrochemically active and Fe (III)- guilliermondii cells during batch xylitol production from sugarcane
reducing bacterium phylogenetically related to Clostridium butyr- bagasse acid hydrolyzate. Biotechnol. Bioeng. 79, 165169
icum isolated from a microbial fuel cell. Anaerobe 7, 297306 50 Huang, Y.L. et al. (2002) Production of carboxylic acids from
25 Kim, B.H. et al. (2003) Enrichment of microbial community generat- hydrolyzed corn meal by immobilized cell fermentation in a fibrous-
ing electricity using a fuel-cell-type electrochemical cell. Appl. bed bioreactor. Bioresour. Technol. 82, 5159
Microbiol. Biotechnol. 63, 672681 51 Du, G. and Yu, J. (2002) Green technology for conversion of food scraps
26 Chaudhuri, S.K. and Lovley, D.R. (2003) Electricity generation by to biodegradable thermoplastic polyhydroxyalkanoates. Environ. Sci.
direct oxidation of glucose in mediatorless microbial fuel cells. Nat. Technol. 36, 55115516
Biotechnol. 21, 12291232 52 Collet, C. et al. (2003) Improvement of acetate production from lactose
27 Hoogers, G. (2002) Fuel cell components and their impact on by growing Clostridium thermolacticum in mixed batch culture.
performance. In Fuel Cell Technology Handbook (Hoogers, G. ed.), J. Appl. Microbiol. 95, 824831
pp. 414427, CRC Press 53 Pynaert, K. et al. (2003) Characterization of an autotrophic nitrogen-
28 Nevin, K.P. and Lovley, D.R. (2000) Lack of production of electron- removing biofilm from a highly loaded lab-scale rotating biological
shuttling compounds or solubilization of Fe(III) during reduction of contactor. Appl. Environ. Microbiol. 69, 36263635
insoluble Fe(III) oxide by Geobacter metallireducens. Appl. Environ. 54 Hashsham, S.A. et al. (2000) Parallel processing of substrate
Microbiol. 66, 22482251 correlates with greater functional stability in methanogenic bio-
29 Magnuson, T.S. et al. (2001) Isolation, characterization and gene reactor communities perturbed by glucose. Appl. Environ. Microbiol.
sequence analysis of a membrane-associated 89 kDa Fe(III) reducing 66, 40504057
cytochrome c from Geobacter sulfurreducens. Biochem. J. 359, 147152 55 Karumanchi, R.S.M.S. et al. (2000) Field-assisted extraction of cells,
30 Bond, D.R. and Lovley, D.R. (2003) Electricity production by Geobacter particles and macromolecules. Trends Biotechnol. 20, 7278
sulfurreducens attached to electrodes. Appl. Environ. Microbiol. 69, 56 Sosa, A.V. et al. (2001) Fluidized bed design parameters affecting novel
15481555 lactic acid downstream processing. Biotechnol. Prog. 17, 10791083
31 Nevin, K.P. and Lovley, D.R. (2002) Mechanisms for accessing 57 Sarkari, M. et al. (2003) CO2 and fluorinated solvent-based technol-
insoluble Fe(III) oxide during dissimilatory Fe(III) reduction by ogies for protein microparticle precipitation from aqueous solutions.
Geothrix fermentans. Appl. Environ. Microbiol. 68, 22942299 Biotechnol. Prog. 19, 448454
www.sciencedirect.com
 Review TRENDS in Biotechnology Vol.22 No.9 September 2004 485

58 Thauer, R.K. et al. (1977) Energy conservation in chemotrophic 67 Ueno, Y. et al. (2001) Characterization of a microorganism isolated
anaerobic bacteria. Bacteriol. Rev. 41, 100180 from the effluent of hydrogen fermentation my microflora. J. Biosci.
59 Schnurer, A. et al. (1999) Mesophilic syntrophic acetate oxidation Bioeng. 92, 397400
during methane formation in biogas reactors. FEMS Microbiol. Ecol. 68 Mizuno, O. et al. (2000) Enhancement of hydrogen production from
29, 249261 glucose by nitrogen gas sparging. Bioresour. Technol. 73, 5965
60 Gottschalk, G. (1986) Bacterial Metabolism, Springer-Verlag 69 Burkert, J.F. et al. (2004) Optimization of extracellular lipase
61 Khanal, S.K. et al. (2004) Biological hydrogen production: effects on production by Geotrichum sp. using factorial design. Bioresour.
pH and intermediate products. Int. J. Hydrogen Energy 29, 11231131 Technol. 91, 7784
62 Wu, S.Y. et al. (2003) Hydrogen production with immobilized sewage 70 Tubesha, Z.A. and Al-Delaimy, K.S. (2003) Rennin-like milk coagulant
sludge in three-phase fluidized-bed bioreactors. Biotechnol. Prog. 19, enzyme produced by a local isolate of Mucor. Int. J. Dairy Technol. 56,
828832 237241
71 Jin, B. et al. (1999) Screening and selection of microfungi for microbial
63 Chin, H.L. et al. (2003) Fedbatch operation using Clostridium
biomass protein production and water reclamation from starch
acetobutylicum suspension culture as biocatalyst for enhancing
processing wastewater. J. Chem. Technol. Biotechnol. 74, 106110
hydrogen production. Biotechnol. Prog. 19, 383388
72 Zheng, S.K. et al. (2002) Seed yeast cultivation for salad oil manu-
64 Wu, S.Y. et al. (2002) Microbial hydrogen production with immobilized
facturing wastewater treatment. J. Environ. Sci. (China) 14, 3943
sewage sludge. Biotechnol. Prog. 18, 921926
73 Domnguez-Espinosa, R.M. and Webb, C. (2003) Production of
65 Fang, H.H. and Liu, H. (2002) Effect of pH on hydrogen production
monascus pigments in wheat based substrates. World J. Microbiol.
from glucose by a mixed culture. Bioresour. Technol. 82, 8793 Biotechnol. 19, 329336
66 Chen, C.C. and Lin, C.Y. (2001) Start-up of anaerobic hydrogen 74 Guerra, N.P. and Pastrana, L. (2002) Nisin and pediocin pro-
producing reactors seeded with sewage sludge. Acta Biotechnol. 21, duction on mussel-processing waste supplemented with glucose
371379 and five nitrogen sources. Lett. Appl. Microbiol. 34, 114118

Getting animated with parasites!

Interested in the molecular cell biology of hostparasite interactions? Then take a look at the online animations produced by Trends
in Parasitology, one of our companion TRENDS journals. The pictures below are snapshots from two of our collection of animations
revealing the latest advances in understanding parasite life cycles. Check them out today!

Microsporidia: how can they invade other cells?

By C. Franzen [(2004) Trends Parasitol. 20, 10.1016/j.pt.2004.04.009]
http://archive.bmn.com/supp/part/franzen.html

Interaction of Leishmania with the host macrophage

By E. Handman and D.V.R. Bullen [(2002) Trends Parasitol. 18, 332334]
http://archive.bmn.com/supp/part/swf012.html

www.sciencedirect.com