1. Appl Microbiol Biotechnol (2004) 64: 421–428
DOI 10.1007/s00253-003-1430-4
ENVI RON MENTA L BIOTECHNOLO GY
J. Sipma . R. J. W. Meulepas . S. N. Parshina .
A. J. M. Stams . G. Lettinga . P. N. L. Lens
Effect of carbon monoxide, hydrogen and sulfate on thermophilic
(55°C) hydrogenogenic carbon monoxide conversion in two
anaerobic bioreactor sludges
Received: 14 April 2003 / Revised: 21 July 2003 / Accepted: 26 July 2003 / Published online: 11 October 2003
# Springer-Verlag 2003
Abstract The conversion routes of carbon monoxide Introduction
(CO) at 55°C by full-scale grown anaerobic sludges
treating paper mill and distillery wastewater were Carbon monoxide (CO) can support a complex microbial
elucidated. Inhibition experiments with 2-bromoethane- food chain of a variety of trophic groups. CO can be
sulfonate (BES) and vancomycin showed that CO con- metabolized by methanogens or by sulfate reducers,
version was performed by a hydrogenogenic population possibly via H2 and CO2 as intermediates (Mörsdorf et
and that its products, i.e. hydrogen and CO2, were al. 1992). Hydrogenogens (Svetlichnyi et al. 2001) are
subsequently used by methanogens, homo-acetogens or capable of converting CO into H2 and CO2. Furthermore,
sulfate reducers depending on the sludge source and acetogens are reported to convert CO into acetate
inhibitors supplied. Direct methanogenic CO conversion (Mörsdorf et al. 1992) or ethanol and butanol (Bredwell
occurred only at low CO concentrations [partial pressure et al. 1999). The latter compounds can be utilized by other
of CO (PCO) <0.5 bar (1 bar=105 Pa)] with the paper mill anaerobes. Table 1 presents reactions that are possibly
sludge. The presence of hydrogen decreased the CO involved in the anaerobic conversion of CO and
conversion rates, but did not prevent the depletion of CO summarizes their stoichiometry and Gibbs free energy
to undetectable levels (<400 ppm). Both sludges showed under standard conditions (25°C) and 55°C.
interesting potential for hydrogen production from CO, In view of the current interest in biohydrogen, the
especially since after 30 min exposure to 95°C, the biological conversion of CO with production of hydrogen
production of CH4 at 55°C was negligible. The paper mill is a very promising reaction. Biohydrogen production can
sludge was capable of sulfate reduction with hydrogen, support the transition of the current unsustainable fossil
tolerating and using high CO concentrations (PCO>1.6- fuel based energy economy into a hydrogen economy.
bar), indicating that CO-rich synthesis gas can be used Furthermore, H2 has a great potential in (bio)chemical
efficiently as an electron donor for biological sulfate processes as an electron donor in various chemical and
reduction. biotechnological reductive processes, e.g. in biological
sulfate reduction processes (van Houten et al. 1994).
A cheap source of hydrogen -rich gas is synthesis gas,
J. Sipma (*) . R. J. W. Meulepas . G. Lettinga . P. N. L. Lens which is produced by, for example, steam reforming of
Sub−department of Environmental Technology, Wageningen natural gas, or thermal gasification of coal, oil, biomass or
University, other organic matter. The composition of the H2-rich
Bomenweg 2, synthesis gas produced varies greatly and depends on the
P.O. Box 8129, 6700 EV Wageningen, The Netherlands source material gasified and the gasification method
e-mail: jan.sipma@wur.nl
Tel.: +31-317-485098 employed (Perry et al. 1997). The major restriction of
Fax: +31-317-482108 synthesis gas utilization is the presence of CO, which can
range from 5% to over 50% (Perry et al. 1997). Sulfate
S. N. Parshina reduction with synthesis gas is hampered by the presence
Laboratory of Microbiology of Anthropogenic Environments,
Institute of Microbiology Russian Academy of Sciences, of CO concentrations of as low as 5% (van Houten et al.
Prosp. 60 let Oktyabrya, 7-2, 1996a). In proton exchange membrane fuel cells, less than
117811 Moscow, Russia 20 ppm CO can be tolerated (Otsuka et al. 2002).
Microorganisms capable of converting CO into hydrogen
A. J. M. Stams
Laboratory of Microbiology, Wageningen University, could represent an interesting biological alternative to the
Hesselink van Suchtelenweg 4, currently employed chemical catalytic water−gas shift
6703 CT Wageningen, The Netherlands reaction (reaction 1, Table 1), especially since biological

2. 422
Table 1 Stoichiometry and Gibbs free energy changes of reactions ΔH° standard enthalpy change. CO conversion routes, except
involved in the anaerobic degradation of CO. ΔG°′Apparent reaction 4, are described in detail in Mörsdorf et al. (1992)
standard free energy change; ΔG′ apparent free energy change;
Reactionsa ΔG°′ (kJ per reaction)a ΔH° (kJ per reaction)b ΔG′ 55°C(kJ per reaction)c
Direct
1 CO+H2O→CO2+H2 −20.0 2.9 −22.3
2 4CO+2H2O→3CO2+CH4 −210.9 −241.6 −207.8
3 4CO+4H2O→CH3COO−+2HCO3−+3H+ −165.4 −282.4 −153.6
Via H2/CO
4 CO+3H2→CH4+H2O −150.8 −250.2 −140.8
Via H2/CO2
5 2HCO3−+4H2+H+→CH3COO−+4H2O −104.6 −246.7 −90.3
6 HCO3−+4H2+H+→CH4+3H2O −135.6 −241.3 −125.0
7 4H2+SO42−+H+→HS−+4H2O −152.2 −252.2 −142.1
Via acetate
8 CH3COO−+4H2O→2HCO3−+4H2+H+ 104.6 246.7 90.3
9 CH3COO−+H2O→CH4+HCO3− −31.0 5.5 −34.7
10 CH3COO−+SO42−→2HCO3−+HS− −47.6 −5.4 −51.8
a
ΔG°′at 25°C according to Thauer et al. (1977)
b
ΔH° was calculated according to Chang (1981). ΔH is assumed to be independent of temperature change when the temperature change is
relatively small (<50°C; Chang 1981). The change in standard enthalpy of formation for HS− was obtained from Lide (2001)
c
ΔG′ values at 55°C were calculated using the ΔG° at 25°C and in the equation: (T is temperature)
Experimental set-up
reactions occur under mild temperatures and at atmo-
spheric pressure. Moreover, biological catalysts are highly Tests for assessing the routes of microbial CO utilization at 55°C
specific and often tolerate sulfur contamination better than under anaerobic conditions were performed as previously described
(Sipma et al. 2003). The basal medium contained (mM): NH4Cl 5.6,
inorganic catalysts (Bredwell et al. 1999). CaCl2.2H2O 0.7, MgCl2.6H2O 0.5, NaCl 5.1, Na2S.9H2O 0.3, yeast
A screening of CO conversion in full-scale grown extract 500 mg l−1, and 1 ml l−1 of an acid and alkaline trace element
anaerobic sludges showed that CO conversion capacity is solution according to Stams et al. (1993). The medium was buffered
ubiquitous in anaerobic bioreactor sludges both at ambient at pH 7.0 using 8.2 mM KH2PO4 and 11.4 mM Na2HPO4.2H2O.
To determine the routes of CO conversion, inhibition studies were
and elevated temperatures (Sipma et al. 2003). Moreover, performed using 25 mM BES to inhibit methanogens or 0.07 mM
at elevated temperatures (55°C) H2 was found to be an vancomycin to inhibit bacteria (Oremland and Capone 1988). In
important intermediate. However, CO conversion routes these experiments CO, acetate, H2/CO2 or H2/CO were used as the
were not elucidated in detail, nor was the effect of substrates. Incubations supplemented with CO as the sole substrate
substrate (CO) and product (H2) concentrations deter- were performed at high (PCO ≥1.6 bar; 1 bar=105 Pa), medium
(PCO±0.4 bar) and low (PCO≤0.06 bar) CO partial pressure to
mined. Therefore, two sludges capable of H2 production determine whether the biochemical routes are CO concentration
from CO in the presence of 2-bromoethanesulfonate (BES) dependent. Sulfate reduction with CO as the sole substrate was
were selected to study the CO conversion routes at 55°C in assessed at high PCO only, by supplementing the basal medium with
detail. Potential substrate (CO) and product (H2) inhibition 20 mM sodium sulfate. Controls consisted of: (1) sterilized biomass
(20 min at 121°C) and (2) basal medium without sludge.
of the biological water−gas shift reaction was evaluated as Granular sludge was crushed as described previously (Sipma et al.
well as the fate of sulfate in the presence of CO. 2003) to study the effect of disruption of the granular structure.
Furthermore, the effect of heat treatment (30 min exposure time) at
different temperatures (75, 95, 100, 105, 110 and 115°C) on CO
conversion by crushed sludge was investigated at high PCO
Materials and methods (>1.6 bar) and 55°C.
Sludge samples
Analysis and chemicals
Granular methanogenic sludge samples were obtained from a full-
scale anaerobic reactor treating wastewater from a paper mill
(Industriewater Eerbeek, Eerbeek, The Netherlands) and from an The pressure in the bottles was determined using a portable
alcohol production plant (Nedalco, Bergen op Zoom, The Nether- membrane pressure unit, WAL 0–4 bar absolute (Wal Mess und
lands). These two sludges were selected from a screening, which Regelsysteme, Oldenburg, Germany). The headspace gas composi-
revealed that Eerbeek sludge rapidly converted CO to methane via tion was measured on a gas chromatograph HP 5890 (Hewlett
H2/CO2, whereas Nedalco sludge produced H2 and CO2 but not Packard, Palo Alto, USA). The detection limit for CO, with the used
methane (Sipma et al. 2003). These sludges were originally settings, was 400 ppm. Volatile fatty acids (VFA) were analyzed on
cultivated between 30 and 35°C, and were not adapted to 55°C a HP 5890A gas chromatograph (Hewlett Packard) according to
prior to the experiments. Weijma et al. (2000). Sulfide was measured according to Trüper and

3. 423
Schlegel (1964). Total suspended solids and volatile suspended
solids (VSS) were analyzed according to standard methods (APHA
1995).
Sulfate was measured on a DX-600 IC system (Dionex, Salt Lake
City, USA). The columns used were IonPac AG17 and AS17 4 mm
operated at a temperature of 30°C and a flow rate of 1.5 ml min−1.
The injection volume was 25 µl. The eluent was made on-line using
the EG40 Eluent Generator (Dionex) equipped with a KOH
cartridge (Dionex P/N 053921) and deionized water as the carrier.
The KOH concentration of the eluent varied during one run as
follows: 1.0 mM from t=0 to t=5 min, 20.0 mM from 5 to 7 min,
40.0 mM from 7 to 7.1 min and 1.0 mM from 7.1 to 10 min. An
ASRS-ULTRA, 4 mm, auto-suppression recycle mode was used.
The detection was based on suppressed conductivity. Prior to
analysis, samples were centrifuged and diluted 100 times.
Solubilities of CO, CO2 and CH4 were calculated using data from
Lide (2001), solubility of H2 was calculated according to Perry et al.
(1997), and the amounts produced or consumed were calculated by
taking into account both gas and liquid phases.
All chemicals used were of analytical grade and purchased from
Merck (Darmstadt, Germany). CO (purity 99.997%) was supplied
by Hoek Loos (Rotterdam, The Netherlands).
Results
CO conversion at high PCO
The microbial population present in Eerbeek sludge
rapidly converted CO into methane (Fig. 1A) via H2 as
an intermediate, which accumulated temporarily to con-
centrations exceeding 1.0 mmol per bottle (Fig. 1A). In the
presence of 25 mM BES, amounts of H2 that were nearly
stoichiometric were formed (Fig. 1B), suggesting that
Fig. 1 CO conversion at high partial pressure of CO by granular
direct conversion of CO into CH4 did not occur at a high Eerbeek sludge at 55°C in the absence of any inhibitor (A) and in
PCO. Support for this was found in the incubations the presence of 25 mM 2-bromoethanesulfonate (BES) (B). The
supplemented with 0.07 mM vancomycin, where no CO symbols indicate; CO (■), hydrogen (○) and methane (x). CO2 was
conversion was observed (Table 2). The small amounts of only measured at the end of the incubations in order to determine the
carbon balance and is therefore not included in the figure
methane that accumulated in the gas phase of these bottles
most probably originated from endogenic substrates, since
in control bottles without CO a similar amount of methane maximal amount of H2 nearly equals the initial amount of
was produced. In each case, the carbon recovery at the end CO supplied. The slight decrease of H2 in the BES-
of the incubations was slightly higher than 100% supplemented incubations (Fig. 1B) was due to the
(Table 2), most probably because of methane production formation of small amounts of acetate (data not shown).
from endogenic substrates. Table 2 further shows that the With Nedalco sludge, CO was rapidly converted to H2,
conversion of CO to H2 is stoichiometric according to the but methane was not formed, not even in the absence of
water−gas shift reaction (reaction 1; Table 1), as the inhibitors (Table 2). Table 2 shows that amounts of H2 that
Table 2 Effect of inhibitors on CO conversion at 55°C. At the start concentration of 85–95% in the gas phase (remainder: N2). VSS
of the incubations CO was supplied as sole substrate at a partial Volatile suspended solids; BES 2-bromoethanesulfonate
pressure (PCO) of 1.5−1.6 bar (1 bar=105 Pa) resulting in a CO
Incubation Start CO End CO Max H2 End H2 End CH4 Acetate produced VSS CO−carbon
in the presence of: millimoles per bottle (grams per bottle) recovery (%)
Eerbeek sludge (incubation period 12 days)
No inhibitor 4.1±0.03 0 1.1±0.05 0 1.2±0.05 n.d. 0.068±0.004 116±2
BES 4.1±0.04 0 3.9±0.2 3.6±0.05 0.2±0.04 n.d. 0.074±0.001 109±1
Vancomycin 4.1±0.06 4.1±0.05 0 0 0.03±0.01 n.d. 0.066±0.004 106±1
Nedalco sludge (incubation period 9 days)
No inhibitor 3.8±0.1 0 3.8±0.2 3.4±0.2 0.01±0.01 0.1±0.1 0.054±0.001 104±1
BES 4.2±0.2 2.4±0.6 n.d. 1.5±0.7 0.01±0.01 0.1±0.03 0.027±0.002 100±3
Vancomycin 4.1±0.05 4.0±0.05 0 0 0 0 0.026±0.005 97±1

4. 424
were nearly stoichiometric accumulated in the gas phase, CO conversion at medium and low PCO
indicating that CO was solely converted to H2 and CO2. At
the end of the experiments, the amount of H2 decreased At PCO of 0.3–0.5 bar, CO was converted to methane by
somewhat as a result of acetate production (Table 2). The Eerbeek sludge in the absence of inhibitors. The amount of
carbon recovery from each incubation was nearly 100%. methane produced was slightly higher than expected based
As methane formation was negligible in the absence of on the supplied CO (according to reaction 1 followed by
any inhibitor, methanogens are presumed not to be reaction 6 of Table 1) because of endogenic methane
involved in CO conversion by Nedalco sludge. This is production (Table 3). Acetate accumulation was compar-
further supported by the absence of CO conversion in the able in the presence and absence of CO (Table 3),
presence of vancomycin (Table 2). suggesting that acetate formation was due to conversion of
endogenous substrates. In the presence of BES, a
considerable production of H2 was observed, although
Conversion of acetate and H2/CO2 small amounts of methane were formed as a result of
in the absence of CO incomplete inhibition of methanogens by BES (Table 3).
The difference between endogenous acetate production
Acetate supplied as sole substrate was not converted to (0.21 mmol) and acetate production in the BES-supple-
methane at 55°C by either sludge after 20 days of mented incubations (Table 3) accounts for the loss of H2.
incubation (data not shown), indicating the absence of Direct conversion of CO to acetate seems unlikely, as the
acetotrophic methanogens. Furthermore, as no H2 accu- uninhibited incubations showed an acetate production of
mulation was observed in acetate-supplemented incuba- 0.21 mmol, equal to the incubations without added
tions, oxidation of acetate into H2 and CO2 did not occur. substrates (Table 3). In the presence of vancomycin, a
H2 and CO2 supplied as a substrate to Eerbeek sludge 25% decrease in the added CO was found and a small
resulted, after a lag phase of 2 days, in complete amount of methane was formed suggesting that direct
conversion to methane within 2 days, and acetate methanogenic CO conversion may be involved at
concentrations were negligible. However, in the presence PCO<0.5 bar (Table 3). However, not all the depleted
of BES, H2 and CO2 were converted to acetate after a lag CO was recovered as methane (Table 3). The CO
phase of about 7 days at a considerably lower rate than in conversion at a low PCO (0.05–0.09 bar) showed the
case of methane production (data not shown). These same trend (Table 3), suggesting that the routes at PCO
results suggest that hydrogenotrophic methanogens in values up to 0.5 bar are not affected by the CO
Eerbeek inoculum sludge probably outcompete homo- concentration.
acetogens competing for H2 and CO2 as substrate. H2 and Both at medium PCO (0.3–0.4 bar) and low PCO (0.04–
CO2 supplied as substrates to Nedalco sludge resulted in a 0.06 bar), small amounts of methane were formed by
rapid production of methane, indicating the presence of Nedalco sludge (Table 3). However, the amounts were
hydrogenotrophic methanogens (data not shown). equal to those in the endogenic incubations. The majority
of CO was recovered as H2, although the amount of H2
recovered could not account for all of the CO converted
(Table 3). Part of the H2 and CO2 produced was converted
Table 3 CO conversion by Inhibitor PCO start (bar) CO start CO end H2 end CH4 end Acetate formed
Eerbeek and Nedalco sludges
at 55°C in the absence and millimoles per bottle
presence of inhibitors at medi-
um and low PCO. The amount of Eerbeek sludge, medium PCO (0.3–0.5 bar CO; 20–25% CO in the gas phase)
biomass was 0.07±0.003 and None 0.30±0.01 0.81±0.03 0 0 0.27±0.01 0.20±0.02
0.03±0.005 g VSS per bottle for
BES 0.37±0.05 0.96±0.11 0.07±0.07 0.62±0.04 0.04±0.00 0.25±0.02
Eerbeek and Nedalco sludges,
respectively. The incubation Vancomycin 0.40±0.05 1.07±0.16 0.76±0.10 0 0.05±0.01 0.02±0.01
time was 9 days Eerbeek sludge, low PCO (0.05–0.09 bar CO; 4–6% CO in the gas phase)
None 0.06±0.02 0.17±0.04 0 0 0.07±0.03 0.21±0.01
BES 0.08±0.01 0.22±0.02 0 0.20±0.02 0.04±0.00 0.18±0.01
Vancomycin 0.07±0.01 0.20±0.03 0.14±0.03 0 0.03±0.01 0.02±0.00
Nedalco sludge, medium PCO (0.3–0.4 bar CO; 20–25% CO in the gas phase)
None 0.31±0.03 0.82±0.09 0 0.60±0.12 0.02 0.16±0.01
BES 0.35±0.02 0.89±0.05 0 0.79±0.20 0 0.16±0.02
Vancomycin 0.37±0.05 0.97±0.09 0.90±0.10 0 0 0.02±0.01
Nedalco sludge, low PCO (0.04–0.06 bar CO; 3–4% CO in the gas phase)
None 0.05±0.01 0.13±0.03 0 0.13±0.00 0.01±0.00 0.14±0.01
BES 0.05±0.00 0.13±0.00 0 0.17±0.00 0 0.15±0.01
Vancomycin 0.05±0.01 0.11±0.01 0.08±0.00 0 0 0.01±0.00

5. 425
to acetate, as the acetate production was slightly higher the gas phase was H2 or N2. To maintain a high
than the 0.13 mmol measured in the control incubations concentration of H2 during CO conversion, BES was
(Table 3). Direct conversion of CO to acetate at the lowest added to some of the bottles with a H2/CO atmosphere.
PCO with Nedalco sludge was most likely absent as the CO was nearly completely depleted from all bottles within
recovery of CO was stoichiometric in H2 (Table 3). Direct 7 days (Fig. 2A), although the differences in CO
conversion of CO into methane did not occur, as the conversion rates were considerable. At a CO concentration
incubations in the presence of vancomycin showed no of 24%, the replacement of nitrogen by hydrogen in the
methane formation (Table 3). absence of BES resulted in a 50% decreased CO
conversion rate (5.52±0.92 mmol (g VSS)−1 day−1). In
the presence of BES the rate was only 35% of the rate
Effect of H2 on CO conversion routes under a CO/N2 atmosphere (3.93
±0.46 mmol (g VSS)−1 day−1). The CO conversion rate,
The presence of H2 may enable an alternative methano- at 12% CO in the presence of BES (remainder H2), was
genic CO conversion route as illustrated by reaction 4 in 1.96 mmol (g VSS)−1 day−1, which is only 50% of that
Table 1. To test the occurrence of alternative metabolic with 22–25% CO. The effect of consumption of H2 during
routes of CO and H2 to CH4 by methanogens (reaction 4, CO conversion is rather limited, suggesting that the initial
Table 1), Eerbeek sludge was incubated with different CO/H2 ratio in the different bottles determines the CO
amounts of CO and H2 in the presence of vancomycin. conversion rate. Figure 2B shows the Gibbs free energy
The latter was added to eliminate the effect of CO change (ΔG), calculated for each measured point
conversion to H2 and CO2 by non−methanogenic micro- presented in Fig. 2A. At the start, the ΔG is highest in
organisms. The gas phase was initially free of CO2, thus the incubations with 24% CO and 76% N2. This probably
excluding hydrogenotrophic methanogenesis. Since Ne- results in a higher biomass yield and thus higher
dalco sludge did not produce CH4 in the presence of CO, conversion rate. Consumption of H2 during CO conver-
tests were exclusively performed with Eerbeek sludge. sion resulted in a slightly higher CO conversion rate,
In all bottles the H2 concentration decreased (Table 4) compared with the BES-supplemented incubation. The
and CH4 was produced. Table 4 shows, however, that only ΔG for 12 and 24% of CO in the presence of H2 and BES
at low CO concentrations, some CO conversion occurred showed only slight differences (Fig. 2B), although the
within 70 days. At PCO=0.08 bar, 60–70% of the CO was rates are lower with decreased CO concentrations
converted. At PCO=0.24 bar the conversion was only 20– (Fig. 2A). This is most probably due to a decreased
30%, whereas at PCO≥0.57 bar, no CO conversion was driving force for diffusion, which is largely determined by
observed. The amount of CH4 produced was considerably the concentration differences between the gas and liquid
higher than expected from the amount of CO converted phase. The presence of relatively large amounts of H2,
(according to reactions 2 and 4, and Table 1) and is nevertheless, did not result in complete inhibition of CO
assumed to be the result of endogenic sources within the conversion to H2 and CO2, and in each bottle the CO
sludge despite the washing procedure prior to incubation. concentration at the end of the experiment was below the
Thus, addition of H2 did not promote the direct metha- CO detection limit of 400 ppm.
nogenic CO conversion according to reaction 4 (Table 4)
compared with direct CO conversion to methane in the
absence of H2 (Table 3). Effect of heat treatment of crushed sludge and sulfate
reduction capacity
Inhibition of CO conversion by hydrogen In crushed Eerbeek sludge, methane formation was
strongly inhibited (data not shown), whereas sulfate
The effect of the presence of H2 on CO conversion was reduction occurred (data not shown). Short-term
tested with Eerbeek sludge, i.e. to test the potential product (30 min) heat treatment of granular Eerbeek sludge at
inhibition by H2. Experiments were performed in duplicate 75°C also resulted in a strong but incomplete suppression
with 12 and 24% CO in the gas phase and the remainder of of methanogenic activity when incubated at 55°C and high
Table 4 Effect of H2 partial pressure on direct methanogenic CO incubation (0, 6, 17 and 34% CO supplied) was analyzed at the start
conversion with Eerbeek sludge in the presence of 0.07 mM and upon termination of the incubations (after 70 days)
vancomycin at 55°C. The composition of the gas phase of each
CO start H2 start CO start CO end H2 start H2 end CH4 formed
% PCO (bar) % millimoles per bottle
0 0 100 0 0 3.79±0.00 2.93±0.00 0.13±0.01
6 0.08 93 0.27±0.02 0.08±0.02 4.00±0.01 2.45±0.04 0.37±0.02
17 0.24 81 0.86±0.03 0.65±0.09 3.93±0.02 2.06±0.04 0.45±0.04
34 0.57 64 2.05±0.01 2.05±0.01 3.84±0.01 1.94±0.06 0.39±0.01

6. 426
measurements as only 20–30% of the sulfate was recov-
ered as sulfide in the medium. Nevertheless, the calculated
amount of H2 produced from CO corresponded with the
sum of H2 remaining at the end of the experiment, the
amount used for acetate formation, and the amount
required for the reduction of sulfate. Sulfide analysis of
the liquid medium including the crushed biomass revealed
that a large part of the sulfide was present as precipitates or
absorbed to the crushed sludge. Taking into account the
amount of biomass-associated sulfide, corrected for the
sulfide amount present in the original inoculum, approxi-
mately 85% of the sulfate reduced could be recovered as
sulfide.
Discussion
Effect of CO concentration on CO conversion route
Previous experiments with several anaerobic bioreactor
sludges showed that at 55°C and PCO>1.5 bar, CO
conversion proceeds via H2 and CO2 (Sipma et al. 2003).
Independent of the PCO, the majority of the CO was
Fig. 2A, B Effect of hydrogen on CO conversion using granular
Eerbeek sludge at 55°C. A CO depletion in incubations with 24%
CO, 70% H2 and 6% N2 with BES (■), 24% CO, 70% H2 and 6%
N2 without BES (ж), 12% CO, 80% H2 and 8% N2 with BES (△),
and 24% CO and 76% N2 without BES (●). B Gibbs free energy
change (ΔG) for reaction CO+H2O→H2+CO2 and apparent stan-
dard free energy (ΔG°′) = −20.0 kJ (mol CO)−1 at each
measurement, taking into account the detection limit of the CO
analysis of 400 ppm. In order to enable calculations at CO, CO2 or
H2 values below the detection limit, the minimal concentration for
each gas was set equal to the detection limit of CO (400 ppm)
PCO (>1.6 bar), whereas no effect on CO conversion was
observed (data not shown). A range of temperatures was
tested to evaluate the effect of this heat treatment on CO
conversion by crushed Eerbeek sludge at 55°C. After
exposure to 95°C for 30 min, methane formation was
absent and all H2 formed from CO was converted to
acetate (data not shown). Also after treatment temperatures
of up to 105°C, rapid CO conversion without methane
formation was observed. Acetate formation started clearly
after CO conversion was completed (data not shown). CO
conversion rates were the highest after heat treatment at
95°C, and were even higher than those of the nontreated
controls. Treatment at 115°C for 30 min resulted in a
complete loss of CO conversion.
After heat treatment at 95°C for 30 min in the presence
of sulfate, Eerbeek sludge was capable of reducing sulfate
with CO or its conversion products (Fig. 3A). This was not
observed with Nedalco sludge (Fig. 3B), although sulfate
reduction with H2/CO2 at 55°C occurred with both
Fig. 3 CO conversion at high PCO in sulfate supplemented heat-
Eerbeek and Nedalco sludges (data not shown). Based treated (30 min at 95°C) crushed Eerbeek (A) and crushed Nedalco
on sulfate measurements, the sulfate reduction seemed (B) sludge at 55°C. The symbols indicate: CO (■), hydrogen (○),
considerably better than expected on basis of sulfide methane (x), sulfide (ж), sulfate (◆) and acetate (▲)

7. 427
converted by hydrogenogens (Svetlichnyi et al. 2001) H2 in anaerobic granular sludge. The actual minimal
present within the sludge. At a low PCO, a small fraction of maintenance requirements will ultimately determine the
the CO was directly converted to methane with Eerbeek minimal CO concentration that can be reached, which still
sludge (Table 3). Complete inhibition of Methanobacte- needs to be established.
rium thermoautotrophicum, which is able to convert CO Conversion of gaseous substrates is usually limited by
directly to CO2 and CH4 according to reaction 2 (Table 1), the mass transfer of substrate from the gas phase to the
occurred at 60% CO (Daniels et al. 1977), whereas biomass in the aqueous phase (Bredwell et al. 1999). Since
Methanosarcina barkeri could be slowly adapted to thermodynamics affects the growth of the CO converting
growth at 100% CO (O’Brien et al. 1984). Despite the populations, especially in case of a conversion that yields
observed reversibility of CO as reported by Daniels et al. relatively little energy (Fig. 2B), high CO concentrations
(1977), methane production with CO-exposed Nedalco together with good biomass retention are favorable for the
sludge was not observed in this study. Reversibility of CO establishment of dense bacterial populations, resulting in a
toxicity could result in methane production with Eerbeek bioreactor with high CO conversion rates. Independently
sludge. However, the absence of methane production in of the reactor design, application of selective H2
crushed Eerbeek sludge after CO is depleted (Sipma et al. membranes (Perry et al. 1997) in combination with CO2
2003) also suggests that CO toxicity is not reversible in fixation will be very beneficial, as they decrease the ΔG
crushed Eerbeek sludge. More likely, the higher CO and result in higher growth and conversion rates.
tolerance of the methanogens present in Eerbeek sludge
results from the protection within sludge granules,
possibly as a result of CO scavenging by other Sludge conditioning by heat treatment
microorganisms at the surface of the granules.
Heat treatment of the inoculum sludge, prior to CO
exposure, results in elimination or at least a dramatic
Effect of the presence of H2 on CO conversion reduction of undesired side reaction, i.e. methane produc-
tion (Fig. 3). Both heat-treated (30 min at 95°C) Nedalco
Direct methanogenic CO conversion (reaction 2 or 4, and Eerbeek sludges are promising seed sludges for a CO-
Table 1) was not stimulated in the presence of H2, fed hydrogen producing reactor, whereas heat-treated
although it was not investigated whether CO conversion Eerbeek sludge is also suitable for start-up of thermophilic
with H2 is possible at all (reaction 4, Table 1). Never- (55°C) sulfate reducing reactors using synthesis gas as a
theless, the presence of H2 and CO2 as products of the cheap electron donor. Despite the fact that sulfate reducers
water−gas shift reaction (reaction 1, Table 1) decreases the are reported to be more vulnerable to CO toxicity than
ΔG and results in decreased CO conversion rates, but not methanogens (Mörsdorf et al. 1992), in crushed Eerbeek
in observable increased final CO concentrations (Fig. 2A). sludge methanogens were severely inhibited, whereas
CO is removed to concentrations below the detection limit sulfate reduction proceeded well even after heat treatment
of 400 ppm (e.g. Fig. 2A). CO conversion into H2 and at 95°C (Fig. 3B). Some sulfate reducing bacteria, e.g.
CO2 yields relatively small amounts of energy for ATP Desulfotomaculum spp., have been described to metabo-
formation (Table 1). Initial ΔG values are considerably lize CO at relative low CO concentrations (up to 20%;
lower when H2is present from the start, as would be the Mörsdorf et al. 1992). The occurrence of sulfate reduction
case when using synthesis gas (Fig. 2B). The differences at high CO concentrations is favorable for the direct use of
in ΔG values at the start of these incubations most synthesis gas as a cheap electron donor for sulfate
probably determine the conversion rates. Since the energy reduction.
gain of this reaction is rather low, it can be expected that The cost of short-term heat treatment is expected to be
the organisms only grow for a short time when the ΔG is rather limited, especially when taking into account the
sufficient. Thereafter, they convert CO uncoupled to unwanted loss of electron donor to methanogens during
growth. In this case, the initial thermodynamics determine reactor operation. Sulfate reducers are generally character-
the population size and ultimately the CO conversion ized by a higher specific growth rate combined with a
rates. lower affinity constant and higher yield compared to
At the detection limit of CO (400 ppm), the ΔG of methanogens (Oude Elferink et al. 1993). However,
reaction 1 (Table 1) is <−14 kJ mole−1 (Fig. 2B). The bioreactor studies show that the population dynamics are
minimal required ΔG for energy conservation is not much slower and that methanogens remain active in
known for the microorganisms involved. Hoehler et al. sulfate reduction bioreactor sludge for extended periods
(2001) reported an apparent minimum free energy (van Houten et al. 1996b). Therefore, elimination of
requirement (threshold ΔG) for methanogens and sul- methanogenesis by short-term heat treatment is expected
fate-reducing bacteria of −11 kJ (mol CH4)−1 and to decrease the long-term operational cost considerably.
−19 kJ (mol SO42−)−1, respectively. In chemostat and
field studies, catabolic activity was observed when the
associated ΔG was at least −10 kJ mole−1 (Seitz et al.
1990; Westermann 1994). This study shows that a ΔG of
−14 kJ (mol CO)−1 still sustains the conversion of CO to

8. 428
Acknowledgements The authors are grateful to Jing Zhang for her Oremland RS, Capone DG (1988) Use of “specific” inhibitors in
practical assistance in this research. This research was financially biogeochemistry and microbial ecology. Adv Microbiol Ecol
supported by a grant from the Technology Foundation STW (grant 10:285–383
STW-WBC 5280), applied science division of NWO (The Nether- Otsuka K, Shigeta Y, Takenaka S (2002) Production of hydrogen
lands) and Paques B.V. (Balk, The Netherlands). from gasoline range alkanes with reduced CO2 emission. Int J
Hydrogen Energy 27:11–18
Oude Elferink SJWH, Visser A, Hulshoff Pol LW, Stams AJM
(1993) Sulphate reduction in methanogenic bioreactors. FEMS
References Microbiol Rev 15:119–136
Perry RH, Green DW, Maloney JO (1997) Perry’s chemical
APHA (1995) Standard Methods for the examination of water and engineers’ handbook, 7th edn. McGraw-Hill, New York
wastewater, 19th edn. American Public Health Association, Seitz HJ, Schink B, Pfennig N, Conrad R (1990) Energetics of
Washington, D.C. syntrophic ethanol oxidation in defined chemostat cocultures. I.
Bredwell MD, Srivastava P, Worden RM (1999) Reactor design Energy requirement for H2 production and H2 oxidation. Arch
issues for synthesis−gas fermentations. Biotechnol Prog Microbiol 155:82–88
15:834–844 Sipma J, Lens PNL, Stams AJM, Lettinga G (2003) Carbon
Chang R (1981) Physical chemistry with applications to biological monoxide conversion by anaerobic bioreactor sludges. FEMS
systems. MacMillan, New York Microbiol Ecol 44:271–277
Daniels L, Fuchs G, Thauer RK, Zeikus JG (1977) Carbon Stams AJM, Dijk JB van, Dijkema C, Plugge CM (1993) Growth of
monoxide oxidation by methanogenic bacteria. J Bacteriol syntrophic propionate-oxidizing bacteria with fumarate in the
132:118–126 absence of methanogenic bacteria. Appl Environ Microbiol
Hoehler TM, Alperin MJ, Albert DB, Martens CS (2001) Apparent 59:1114–1119
minimum free energy requirements for methanogenic archaea Svetlitchnyi V, Peschel C, Acker G, Meyer O (2001) Two
and sulfate-reducing bacteria in an anoxic marine sediment. membrane-associated NiFeS-carbon monoxide dehydrogenases
FEMS Microbiol Ecol 38:33–41 from the anaerobic carbon-monoxide-utilizing eubacterium
Houten RT van, Hulshoff Pol LW, Lettinga G (1994) Biological Carboxydothermus hydrogenoformans. J Bacteriol 183 5134–
sulphate reduction using gas-lift reactors fed with hydrogen and 5144
carbon dioxide as energy and carbon source. Biotechnol Bioeng Thauer RK, Jungermann K, Decker K (1977) Energy conservation
44:586–594 in chemotrophic anaerobic bacteria. Bacteriol Rev 41:100–180
Houten RT van, van der Spoel H, van Aelst AC, Hulshoff Pol LW, Trüper HG, Schlegel HG (1964) Sulphur metabolism in Thiorho-
Lettinga G (1996a) Biological sulphate reduction using daceae. I. Quantitative measurements on growing cells of
synthesis gas as energy and carbon source. Biotechnol Bioeng Chromatium okenii. Antonie van Leeuwenhoek 30:225–238
50:136–144 Weijma J, Stams AJM, Hulshoff Pol LW, Lettinga G (2000)
Houten RT van, Yu Yun S, Lettinga G (1996b) Thermophilic Thermophilic sulfate reduction and methanogenesis with
sulphate and sulphite reduction in lab-scale gas-lift reactors methanol in a high rate anaerobic reactor. Biotechnol Bioeng
using H2/CO2 as energy and carbon source. Biotechnol Bioeng 67:354–363
55:807–814 Westermann P (1994) The effect of incubation temperature on
Lide DR (ed) (2001) CRC handbook of chemistry and physics, 82nd steady-state concentrations of hydrogen and volatile fatty acids
edn. CRC, Boca Raton, Fla. during anaerobic degradation in slurries from wetland sedi-
Mörsdorf G, Frunzke K, Gadkari D, Meyer O (1992) Microbial ments. FEMS Microbiol Ecol 13:295–302
growth on carbon monoxide. Biodegradation 3:61–82
O’Brien JM, Wolkin RH, Moench TT, Morgan JB, Zeikus JG (1984)
Association of hydrogen metabolism with unitrophic or mix-
otrophic growth of Methanosarcina barkeri on carbon monox-
ide. J Bacteriol 158:373–375