ww.sciencedirect.com

wat e r r e s e a r c h 4 6 ( 2 0 1 2 ) 5 3 9e5 4 6

Available online at w

journal homepage: www.elsevier .com/locate/watres

Modeling volatile organic sulfur compounds in mesophilicand thermophilic anaerobic digestion of methionine

Weiwei Du*, Wayne Parker

Dept. of Civil and Environmental Engineering, University of Waterloo, 200 University Ave. W., Waterloo, ON, Canada N2L 3G1

a r t i c l e i n f o

Article history:

Received 5 April 2011

Received in revised form

11 November 2011

Accepted 15 November 2011

Available online 25 November 2011

Keywords:

Methyl mercaptan

Dimethyl sulfide

Volatile sulfur compounds

Mesophilic

Thermophilic

Anaerobic digestion

* Corresponding author. Tel.: þ1 519 729 085E-mail address: wdu@engmail.uwaterloo

0043-1354/$ e see front matter ª 2011 Elsevdoi:10.1016/j.watres.2011.11.043

a b s t r a c t

Processes involved in volatile organic sulfur compound (VOSC) generation and degradation

in mesophilic and thermophilic digestion of methionine were identified, kinetically studied

and a mathematical model was established. MM was found to be the only VOSC directly

generated from methionine degradation. MM was methylated to form DMS and both MM

and DMS were subsequently degraded to H2S. Mixed-second order kinetics were found to

best fit the VOSC generation and conversion processes. The kinetic constants (average

values) for MM generation and methylation and MM and DMS degradation were estimated

to be 0.0032, 0.0047, 0.027, and 0.013 l g�1 h�1 respectively at 35 �C and 0.0069, 0.0012, 0.0083,

0.005 l g�1 h�1 respectively at 55 �C. More rapid MM release and slower VOSC decline at

thermophilic temperature implied that VOSC could be more problematic at thermophilic

temperatures as compared to mesophilic conditions.

ª 2011 Elsevier Ltd. All rights reserved.

1. Introduction anode of solid oxide fuel cell systems will be deactivated. In

Increasing complaints of odor problems at wastewater treat-

ment plants has aroused attention. Volatile organic sulfur

compounds (VOSC) and H2S from anaerobic digestion have

been identified as the most odorous compounds related to

sewage treatment because of their very low odor thresholds

and very negative hedonic values. Even a small amount of

VOSC and H2S can contribute to odor pollution (Smet and

Langenhove, 1998). In addition, VOSCs and H2S in biogas are

reactive and corrosive to metal pipes, biogas storage tanks,

and biogas utilization equipment such as biogas engines and

turbine generators. When biogas is used for more efficient

electricity generators such as fuel cells, it has to be cleaned-up

and reformed. VOSCs and H2S can poison the catalysts that

are used in both reforming and fuel cells. It has been reported

that when the total sulfur concentration is above 10 ppmv, the

0..ca (W. Du).ier Ltd. All rights reserved

addition, when the concentration of H2S was higher than

200 ppmv, activated carbon treatment could not effectively

accomplish sulfur removal to reach a required purity of biogas

(Wheeldon et al., 2007).

Proteineous materials and sulfate have been reported as

major sources of the volatile sulfur compounds that are

generated during anaerobic digestion of municipal sludge.

These sulfur containing compounds are converted to VOSC

and H2S under anaerobic conditions through mechanisms

such as biological reduction, hydrolysis, methylation, and

metal catalyzed oxidization (Drotar et al., 1987; Kadota and

Ishida, 1972; Lomans et al., 2002; Hullo et al., 2007;

Sreekumar et al., 2009). However, VOSC concentrations in

digesters are reduced by methanogens that mediate the

degradation of VOSC (Finster et al., 1992; Chen et al., 2005; De

Bok et al., 2006).

.

wat e r r e s e a r c h 4 6 ( 2 0 1 2 ) 5 3 9e5 4 6540

In anaerobic sludge digestion, biochemical generation of

H2S from sulfate reduction and pertinent kinetics has been

well established (Visser, 1995; Kalyuzhnyi et al., 1998;

Kalyuzhnyi and Fedorovich, 1998; Fedorovich et al., 2003)

while VOSC generation and conversion is notwell understood.

Temperature has been reported to affect VOSC generation and

release in both the start-up and steady state operation of

anaerobic digesters (Iranpour et al., 2005), however, quanti-

tative information on the effects of temperature on VOSC

behavior is limited. To better predict VOSC behavior in

anaerobic sludge digestion, the occurrence of the sulfur

related reactions in digestion ofmunicipal sludges needs to be

confirmed, corresponding kinetic information needs to be

quantified, and the effect of temperature should be assessed.

The predominant amino acids in proteins that contain

sulfur are methionine (C5H11NO2S, CH3eSe(CH2)2eCH(NH2)e

COOH) and cysteine (C3H7NO2S, HSeCH2eCH(NH2)eCOOH)

and both have been reported to be VOSC precursors in

municipal sludge digestion. Methionine has been reported to

be degraded through different pathways under different

conditions to produce either methyl mercaptan (MM),

dimethyl sulfide (DMS) or H2S. Cysteine normally is consid-

ered as an organic precursor of only H2S under anaerobic

conditions (Kadota and Ishida, 1972; Derbali et al., 1998).

While the pathways for VOSC generation have been

documented, kinetic information on VOSC formation and

conversion in the degradation of methionine is not available

and hence prediction of the VOSC behavior in anaerobic

sludge digestion is challenging. This paper describes a study of

the kinetics of VOSC generation from methionine and their

subsequent degradation in mesophilic and thermophilic

anaerobic sludge digestion. The development of a mathemat-

ical model that describes the conversions is presented.

2. Materials and methods

Two groups of batch tests were conducted in the present

study. The first group of tests included anaerobic digestion of

methionine with and without methanogen inhibition, using

digested municipal wastewater treatment sludge as the

inocula. In methionine digestion without methanogen inhi-

bition, the types of VOSCs that are generated and subse-

quently transformed were identified. The VOSC responses in

these tests were also subsequently employed for the purpose

of model verification. The biodegradation of VOSCs was

eliminated when methanogens were inhibited and hence

VOSC accumulation in methionine digestion with inhibition

of methanogens was employed for estimation of the kinetics

of VOSC formation from methionine. The second group of

batch tests involved incubations that were dosed with indi-

vidual VOSCs. In these tests, the VOSCs that were observed to

be formed frommethionine were dosed into a digested sludge

inocula and their degradationwasmonitored for development

of kinetic rate models.

Batch incubations were conducted in non-stirred, 328 ml

serum bottles that were sealed with screw-caps which were

equipped with bromobutyl rubber septa. The temperatures of

the mesophilic and thermophilic incubations were 35 and

55 �C, respectively. Digested sludges that were obtained from

pilot-scale mesophilic (35 �C) and thermophilic (55 �C)digesters were pre-incubated at the target temperatures for

48 h, and then used as inocula. The pilot-scale digesters were

operated at an average hydraulic retention time of 22 days and

fed with a co-thickened sludge that was the underflow of

primary settler collected from Waterloo Wastewater Treat-

ment Plant. The yearly average flow rate of wastewater

influent, waste activated sludge (WAS), and primary under-

flowwas 36,655, 753, and 282m3/day respectively. The average

total suspended solids concentration (TSS) of theWAS and co-

thickened sludge was about 0.45% and 3% (VSS/TSS w 70%).

The values for VSS destruction across the pilot-scale meso-

philic and thermophilic digesters were approximately 60%

and 64% respectively. The total sulfur concentrations in the

pilot-scale mesophilic and thermophilic digesters were

between 20 and 36 mg S/g VSS while the organic sulfur

concentrations in the pilot-scale digesters were between 9 and

20mg S/g VSS. The total volume of the liquid phasewas 200ml

with initial volatile suspended solids (VSS) varying between

1.8 and 9.3 g/l. VSS concentrations were employed as a surro-

gate for the concentration of biomass which mediated bio-

logical sulfur related reactions. The specific reaction rates for

methionine and VOSC degradation at 35 �C and DMDS

degradation at 55 �Cwere found to be relatively fast and hence

dilute inocula with an initial VSS between 1.8 and 3.0 g/l was

used. For most of the tests at 55 �C, undiluted inocula with

higher initial VSS levels, between 6.4 and 9.3 g/l, was used due

to the slow degradation rates of MM and DMS. All the serum

bottles were purged with pure nitrogen before incubation.

In methionine-dosed tests a 1000 mM methionine stock

solution was added to achieve concentrations of 2e5 mM

(equivalent to 6.9e70.2 mg S/g VSS when normalized by VSS).

When dilute inocula were used at 55 �C, MM accumulated and

minimal decay was observed within 5 days. Therefore, undi-

luted inocula were used for methionine incubations at 55 �C,which led to a reduced normalized initial S concentration. In

bottles where methanogens were inhibited, a 2-bromoethane-

sulfonic acid (BES) stock solution was dosed to a final concen-

tration of 15 mM as the methanogen inhibitor. Methionine

incubations, with and without methanogen inhibition, were

carried out with 5 replicates respectively. The replicate bottles

were gradually sacrificed for total Kjeldahl nitrogen and

ammonia analysis tomonitormethionine hydrolysis with time

(data not presented in this paper). At the same time, the repli-

cate bottles ensured replicate headspace samples for VOSC

analysis. In the tests that were dosed with VOSCs which were

detected in the methionine incubations (MM and DMS), 2e5

different initial VOSC dosages were employed. The initial VOSC

dosages were in the range of 0.92e10.57 mg S/g VSS. The

analytical grade VOSCwas dosed into serum bottles directly by

syringes.Duplicatebottleswerepreparedforeach initialdosage.

The volume of biogas produced in the serum bottles was

measured by release into a manometer, until the pressure in

the serum bottle was equal to atmosphere pressure. The

volume of water in the manometer which was replaced by the

released gas was then recorded. The frequency of biogas

release was gradually reduced from once every day to once

a week over the duration of incubation. Methane generation

was used to monitor the methanogen activity. The methane

and carbon dioxide content of biogas samples were analyzed

wat e r r e s e a r c h 4 6 ( 2 0 1 2 ) 5 3 9e5 4 6 541

by an SRI 310C gas chromatograph (GC) equipped with

a thermal conductivity detector and a 60 � 1/800OD Porapak

column (80/100 mesh). The VOSC content of the biogas was

analyzed by a GC equippedwith a capillary column (Valcobond

VB-1, 100% dimethyl-polysiloxane, 30 m � 0.32 mm, 0.4 mm)

and a pulsed flame photometric detector (PFPD). For methio-

nine incubation tests, the gas phase of all the bottles was

sampled 2 times every day for the first 3 days, then once every

1e3 days until the end of the incubation. For the VOSC dosed

tests, the gas phase was sampled with a frequency from 1 to 4

times a day. Total suspended solids/volatile suspended solids

(TSS/VSS) of the inocula were measured following Standard

Methods 2540 D and E respectively (APHA, AWWA, and WEF,

1992).

3. Results and discussion

To better reflect the sulfur transformations duringmethionine

degradation and VOSC conversion, the masses of methionine

and the VOSCs were represented by their S equivalent and

normalized by VSS mass in the serum bottles. VOSC concen-

trations in the headspace were measured by the GC-PFPD.

Concentrations of VOSC in the liquid phase were quantified

based on their respective liquid-gas partitioning coefficient

(Du, 2010) with an assumption of instantaneous equilibrium

for VOSC distribution between the liquid and gas phases. The

total VOSC mass was the sum of VOSC masses in the two

phases.

3.1. Methionine incubations

MM and DMS have been reported to be generated through

methionine degradation in anoxic saltmarsh sediments (Kiene

and Visscher, 1987), in treated biosolids (Higgins et al., 2006),

and in food fermentation (Bonnarmeet al., 2000). In thepresent

tests, VOSC concentrations were monitored over time in the

headspaces of methionine incubations without methanogen

inhibition. Consistent with the literature, MM and DMS were

detected in the sludge digestionswith dosedmethionine. Their

concentrations (average of duplicates) with time are presented

in Fig. 1. The variability of measurements was less than 10% of

the means and hence was not presented in the figure.

From Fig. 1 it can be seen that at 35 �C, with an initial

methionine dosageof 70.2mgS/gVSS,MMrapidly increased to

Fig. 1 e MM and DMS concentrations versus time in

methionine incubation without methanogen inhibition.

2.5 mg S/g VSS after 25 h of incubation and then gradually

declined to less than 0.1 mg S/g VSS after 300 h of incubation.

The concentration of DMS increased with MM concentrations,

peaked at 0.9 mg S/g VSS after 100 h of incubation, and

decreased to less than 0.1mg S/g VSS after 200 h of incubation.

Similarly at 55 �C, with an initial methionine concentration of

6.9 mg S/g VSS MM rapidly increased and maintained a high

concentration of 1.2 mg S/g VSS for the first 20 h of incubation

and then gradually declined to less than 0.01 mg S/g VSS after

78 h. The concentration of DMS increased with MM concen-

trations, reached a peak value of 0.5 mg S/g VSS after 54 h of

incubation, anddeclined tobelow thedetection limit after 78h.

Kiene and Visscher (1987) reported that DMS was gener-

ated from MM and that MM peak concentrations were 6e30

fold of the peak DMS concentrations during methionine

degradation in sediments. The observations of the present

study with sludge digestion were consistent with the previous

study at both temperatures that were examined. Without

methanogen inhibition MM generation started immediately

and the MM concentrations reached peak values earlier than

the DMS values. The MM peak concentrations were much

higher than DMS peak concentrations at both temperatures.

The results suggest that MM was the direct VOSC product of

methionine degradation while DMS was generated from MM

as shown in Fig. 2. The disappearance of MM and DMS will be

subsequently demonstrated to be mediated by methanogens.

The time-course of the VOSC resultswill be subsequently used

to validate the models that have been developed to simulate

VOSC formation and decay during methionine incubation.

Fig. 3 presents the concentrations of MM that were

observed when methanogens were inhibited in the methio-

nine incubations. The initial VSS concentrations in the bottles

were utilized to normalize the sulfur mass. At both tempera-

tures there was a significant lag in MM generation from

methionine at 35 �C (Fig. 3), however, after about 120 h

of incubation, themass ofMM increased rapidly. After 400 h of

incubation at 35 �C, the mass of MM accumulated to a level of

65.0 mg S/g VSS, which represented 92.5% of the sulfur

available from the dosedmethionine (original sulfur dose was

70.2 mg S/g VSS). After 340 h of incubation at 55 �C, the MM

concentration approached a constant level that was main-

tained until the end of the batch test. The generated MM

accumulated to a level of 45.0 mg S/g VSS, which represented

Fig. 2 e Conceptual model of VOSC generation and

degradation pathways in anaerobic digestion of

methionine.

Fig. 3 eMMgenerationwith time inmethionine incubation

with methanogens inhibited.

Fig. 4 e MM and DMS concentrations versus time (in

mesophilic incubations and when initial masses of MM

and DMS were less than 2 and 3 mg S/g VSS respectively in

themophilic incubations).

wat e r r e s e a r c h 4 6 ( 2 0 1 2 ) 5 3 9e5 4 6542

essentially all the sulfur available from the dosed methionine

(original sulfur dose was 45.9 mg S/g VSS). Zinder and Brock

(1978) reported that MM was the direct product of methio-

nine degradation in anaerobic sediments. The present study

conducted with municipal digestion sludge confirmed this

conclusion. In the methionine incubations with methanogens

inhibited, MM was the only VOSC which was detected.

It was previously observed that at both temperatures MM

and DMS were detected in anaerobic incubations of methio-

nine without methanogen inhibition and both VOSCs even-

tually disappeared. In both mesophilic and thermophilic

incubations with methanogen inhibition, MM was the domi-

nant sulfur containing product generated. Its concentration

rapidly accumulated with time after the acclimation period

and then maintained at a high concentration level until the

end of incubation. These results confirmed that the degrada-

tion of both MM and DMS is mediated by methanogens.

Pianotti et al. (1986) reported that inhibition of methionine

degradation occurred when an initial methionine concentra-

tion which was higher than 2 mM was employed in the

incubation and the inhibition led to a postponed sulfur

release. In the present study, with an initial methionine

concentration of 2e5 mM, the MM generation from methio-

nine was only delayed in the methanogen-inhibited incuba-

tions. There was no lag in the methionine incubation without

methanogen inhibition at both mesophilic and thermophilic

temperatures. Therefore, high initial methionine concentra-

tion (>2 mM) was not sufficient to explain the time lag which

only occurred in incubations with methanogen inhibition.

Baena et al. (1998) reported that methionine degradation

depended on the removal of its fermentation products by

methanogens. Accumulation of fermentation products of

methionine could be inhibitory for continuous methionine

degradation. In the present experiment, no ammonia accu-

mulation was detected (ammonia detection limit, 0.3 mg/l

with 95% confidence) during the lag period of the inhibited

bottles (data not presented). With an assumption that

ammonia was one of the fermentation products of methio-

nine, slow ammonia generation suggested an inhibited

fermentation. Therefore, the temporarily inhibited methio-

nine degradation was not caused by the accumulation of

fermentation products.

It is typically assumed that the inhibitor BES only affects

methanogens (Thauer, 1998). However, the lag in generation

of ammonia and MM in the methionine incubations with BES

addition suggested that BES might also inhibit methionine

hydrolysis. The data suggest that the biomass was able to

acclimate to the inhibitory effect after 120 h and methionine

was totally hydrolyzed at the end of the tests.

3.2. VOSC incubations

Batch tests that were dosedwith individual MMandDMSwere

conducted to facilitate estimation of degradation kinetic rate

constants for these substances. Different initial dosages were

employed in these tests to assess the effect of the concen-

tration of VOSCs on their degradation. The estimation of the

kinetic coefficients of VOSC degradation based on the time-

course of VOSC variation will be discussed in a subsequent

section. The effect of the temperature was assessed by

comparing VOSC degradation rates at temperatures of 35 and

55 �C.Fig. 4 presents total MM and DMS masses (normalized by

VSS) present in serum bottles versus incubation time in the

mesophilic batch tests with an initial MM mass of 3.11 mg S/g

VSS and an initial DMS mass of 2.92 mg S/g VSS respectively.

The responses observed in Fig. 4 were representative of the

trends in VOSC degradation that were observed in all

the mesophilic batch tests and representative of the trends in

the thermophilic batch tests when the initial masses of MM

and DMS were less than 2 mg S/g VSS and 3 mg S/g VSS

respectively. The decline of the mass of VOSCs in the serum

bottles was attributed to biodegradation that wasmediated by

methanogens. The role of methanogens in VOSC degradation

was observed in themethionine incubation tests and has been

reported in the literature (Lomans et al., 1999).

At both mesophilic and thermophilic temperatures the

mass of DMS declined at a considerably slower rate as

compared to MM. For instance, at 35 �C with an initial MM

concentration of 3.11 mg S/g VSS about 55 h was required to

achieve 95% reduction in the mass of MM while 120 h was

required to obtain the same reduction of DMSwhen the initial

concentration of DMS was 2.92 mg S/g VSS (Fig. 4). At 55 �Cwith an initial concentration of 1.84 mg S/g VSS, 60 h was

Fig. 6 e DMS and MM concentrations versus time in MM-

dosed mesophilic incubation.

wat e r r e s e a r c h 4 6 ( 2 0 1 2 ) 5 3 9e5 4 6 543

required to achieve a 95% reduction of MM while DMS

required 175 h to achieve the same extent of degradation

when an initial concentration of 2.13 mg S/g VSS was

employed.

In the thermophilic batch tests, the degradation of MMwas

inhibited when the initial normalized mass of MM in the

serum bottles was higher than 2mg S/g VSS. Fig. 5 presents an

example of this type of response when the initial normalized

mass was 4.26 mg S/g VSS. From this figure it can be seen that

after 240 h of incubation the normalizedmass had reduced by

less than 5% of the initial value. Inhibition of methanogen

activity has been previously reported under conditions with

high concentrations of MM (Kiene et al., 1986; Van Leerdam

et al., 2006).

Elevated concentrations of DMS (3.96, 6.6 and 10.57 mg S/g

VSS) alsoappeared to inhibit thedegradationof this compound

in the thermophilic incubations. However, in this case the

decline in concentration of DMS exhibited a staged and slow

decline. As indicated in Fig. 5 the mass of DMS declined

approximately linearly for the first 70 h of incubation and then

the concentrations became essentially constant. By the end of

the incubation (170 h) less than 70% of the dosed DMS had

degraded (Fig. 5). Inhibition of DMS degradation was not

observed in tests with initial dosages less than 3 mg S/g VSS.

When the initial DMS dosages were low (<3 mg/g VSS), DMS

declined steadily (Fig. 4). The inhibited DMS degradation at

high initial concentrations was consistent with the results of

Kiene et al. (1986) where high DMS concentrations inhibited

methane generation. Inhibition was only observed at ther-

mophilic temperature in thepresent study. Considering that in

mesophilic and thermophilic municipal sludge digestion, the

typical accumulated MM and DMS mass was less than 2 and

3 mg S/g VSS respectively, the inhibition conditions were not

included in modeling VOSC conversions in sludge digestion.

In the MM dosed tests, there was a transient accumulation

and subsequent degradation of DMS in the serum bottles.

Fig. 6 presents the MM and DMS responses that were observed

in the mesophilic incubation of MM at an initial dose of

1.30 mg S/g VSS and is representative of the responses

observed for the other initial doses at 35 �C. From Fig. 6 it can

be seen that approximately 25 h after the addition of MM, DMS

was detected and its concentration slowly increased to a peak

at around 30 h, which corresponded with the time when MM

was exhausted. The transient accumulatedDMS subsequently

degraded below the detection limit within another 20e25 h.

Fig. 5 e MM and DMS concentrations versus time in

thermophilic incubations at elevated initial doses.

A similar response was observed in the thermophilic

incubation of MM. Fig. 7 shows DMS accumulation in the

thermophilic degradation of MM with an initial dose of

0.92 mg S/g VSS and is representative of the responses that

were observed when the initial dose was less than 2 mg S/g

VSS. From Fig. 7 it can be seen that after 5 h of incubation DMS

was detected and the DMS concentration slowly increased

with the decrease of MM. In contrast to the mesophilic tests,

the generated DMS did not disappear at 60 h of incubation

when the test was terminated.

The results suggest that methylation of MM to form DMS

occurred in the incubations at both temperatures. The

continuous accumulation of DMS in the thermophilic incu-

bation reflects a slower degradation rate for DMS at thermo-

philic temperatures as compared to mesophilic temperatures.

This behavior was consistent with that observed in the DMS

dosed tests.

However, methylation was not the only mechanismwhich

resulted in the reducedmass ofMM-S in the incubations. In all

the MM dosed tests at 35 �C, the peak masses of DMS-S were

about 20% of the initial dosedMM-Smasses. At 55 �C the DMS-

S mass slowly increased with the decrease of MM-S. When

more than 90% of the dosed MM-S had degraded, the amount

of the S associated with the generated DMS was less than 13%

of the initial dosed S in the MM. Hence, generation of DMS

could not represent the total amount of MM which was

removed. These results suggest that direct reduction of MM to

generate H2S was likely a competitive reaction with

methylation.

Fig. 7 e DMS and MM concentrations versus time in MM-

dosed thermophilic incubation.

Table 1 e Model of VOSC generation and conversion in methionine digestion.

Process C_met C_MM C_DMS C_H2S Rate

Degradation of methionine �1 1 �kmet$C_met$VSS

DMS degradation �1 1 �k_DMS$C_DMSliq$VSS

MM degradation �1 1 �k_MM$C_MMliq$VSS

MM methylation to DMS �1 1 �kmethy$C_MMliq$VSS

wat e r r e s e a r c h 4 6 ( 2 0 1 2 ) 5 3 9e5 4 6544

3.3. Modeling and kinetic parameter estimation

The pathways of generation and consumption of VOSCs in

methionine incubation presented in Fig. 2 were confirmed by

the methionine and VOSC responses in the dosed batch tests.

The stoichiometric relations between the sulfur containing

compounds are summarized in Table 1. Mixed-second order

kinetics (Equation (1)), were found to best depict anaerobic

biological methionine degradation, degradation of MM and

DMS, as well as the methylation of MM.

r ¼ dcdt

¼ �k$C$VSS (1)

Where r is the consumption rate of the sulfur containing

compound (mg l�1 h�1), C is the sulfur concentration (as

methionine or VOSC) in the liquid phase (mg/l), t is time (h), k

is the mixed-second order kinetic parameter (l g�1 h�1), and

VSS is the volatile suspended solids which represented the

biomass concentration (g/l). According to the mixed-second

order kinetics, the rates of VOSC generation/conversion

processes depend on the concentration of substrate as well as

the biomass concentration. In the present study, the biomass

concentration was assumed to be a constant value over the

incubation and represented by the initial VSS. The actual VSS

deduction rate was up to 20% in the methionine incubations

without methanogen inhibition and was about 10% in the

other batch incubations. The assumption of negligible

decrease of VSS through the digestion may result in slightly

conservative estimates of the kinetic constant values. The

specific mixed-second order kinetic expressions for each

conversion are summarized in Table 1. The MM and DMS

degradation kinetic constants were estimated with the data

collected from incubations that were not inhibited with high

MM and DMS concentrations.

Methionine concentrations (C_met) were not directly

measured in the experiments conducted in this study.

Assuming a direct conversion of methionine to MM (Table 1)

the rate of methionine degradation was determined from the

rate of MM generation. Therefore, kmet was estimated by

fitting MM accumulation in the methionine incubations.

Table 2 e Methionine degradation and VOSC conversion coeffi

Mesophilic

Average value Range (P ¼kMet-S (l g�1 h�1) 0.0032 0.0023e0.

k_DMS (l g�1 h�1) 0.013 0.010e0

k_MM (l g�1 h�1) 0.027 0.010e0

kmethy (l g�1 h�1) 0.0047 0.0038e0.

The estimated values of kmet at mesophilic and thermophilic

temperatures are presented in Table 2. The value of kmet for

thermophilic digestion was two times higher than that for

mesophilic digestion, reflecting the data that demonstrated

the release of MM from methionine was more rapid at 55 �C.Fig. 8 presents the fit of the model to the decline of DMS at

two different initial dosages in the thermophilic batch tests

and is representative of the model fit for DMS degradation

under both mesophilic and thermophilic conditions. From

Fig. 8 it can be seen that the calibrated model was able to

describe the decay of DMS over a range of DMS concentra-

tions. The estimated rate coefficients for DMS degradation at

the two temperatures are summarized in Table 2.

In the MM dosed batch tests, MM was converted to

generate both H2S and DMS and the generated DMS also

decayed. Fig. 9 presents representative responses for the

decline of MM and the formation and disappearance of DMS

with an initial MM mass of 4.67 mg S/g VSS at 35 �C. The

responses presented in Fig. 9 are representative of those

observed for MM degradation with different initial dosages in

both the mesophilic and thermophilic batch tests. The value

of k_DMS employed in these simulations was the value that

was estimated from for the previously described DMS degra-

dation tests. The values of k_MMand kmethy were estimated by

fitting the MM and DMS time-course curves observed in the

MM dosed tests (Table 2).

The decay rate coefficients for each process (MM and DMS

decay and MM methylation) were estimated for each initial

VOSC dosage and were found to be not statistically different

between the various doses (P ¼ 0.01). Comparing the average

values of the coefficients estimated at 35 �C and at 55 �C (Table

2), it can be seen that the MM and DMS decay rate constants

were about 3.4 times and 2.7 times higher at 35 �C as compared

to 55 �C. Hence, the values of the decay rate coefficients reflect

the trends in compound disappearance that were previously

discussed. The results suggest that methanogens have

a reduced capacity for MM and DMS degradation at 55 �C as

compared to that at 35 �C.TheMMmethylation rate constantwas also greater at 35 �C

as compared to its value at 55 �C with the average rate

cients.

Thermophilic

0.01) Average value Range (P ¼ 0.01)

0040 0.0069 0.0053e0.0085

.016 0.005 0.0040e0.0057

.047 0.0083 0.0053e0.013

0063 0.0012 0.0005e0.0041

Fig. 9 e Observed and predicted concentrations of MM and

DMS variation in MM-dosed mesophilic incubation.

Fig. 10 e Simulated and monitored VOSC concentrations

with time in mesophilic methionine digestion.

Fig. 11 e Simulated and monitored VOSC concentrations

with time in thermophilic methionine digestion.

Fig. 8 e Observed and predicted DMS concentrations versus

time in thermophilic incubations.

wat e r r e s e a r c h 4 6 ( 2 0 1 2 ) 5 3 9e5 4 6 545

coefficient at 35 �C about 3.9 times the value at 55 �C. Thisvalue was consistent with the faster decline of MM that was

observed at 35 �C. Considering the more rapid release of MM

from methionine at 55 �C that was previously described, it

could be speculated that VOSC persistence under thermo-

philic conditions would be longer and at higher concentra-

tions than that under mesophilic conditions, when similar

initial sulfur masses are present.

Verification of the combined model presented in Table 1

was conducted by comparing model simulations with the

Table 3 e Initial conditions and kinetic parameter valuesutilized in the simulation of methionine digestion.

Parameter values

Mesophilic Thermophilic

k_met (l g�1 h�1) 0.0034 0.0065

k_MM (l g�1 h�1) 0.069 0.017

kmehty (l g�1 h�1) 0.006 0.003

k_DMS (l g�1 h�1) 0.02 0.004

Initial mass/concentration

Mesophilic Thermophilic

Dosed methionine (mg S/l) 160 64

VSS (g/l) 2.28 9.30

VOSC quantities that were previously described for methio-

nine digestion without methanogen inhibition. The average

parameter values presented in Table 2 were employed as the

initial values for these simulations and were adjusted within

the range presented in Table 2 to optimize the model simu-

lations with the measured data. The initial methionine doses

and VSS concentrations and the adjusted values of the kinetic

constants employed in these simulations are listed in Table 3.

The observed and predicted VOSC quantities for methionine

digestion without methanogen inhibition are presented in

Figs. 10 and 11 for mesophilic and thermophilic digestion

respectively. From these figures it can be observed that the

simulated values exhibited good agreement with the moni-

tored data and hence validated the model structure and the

values of the kinetic coefficients.

4. Conclusions

The processes involved in VOSC generation from methionine

and the degradation of the byproducts were quantitatively

characterized in this study. At both mesophilic and thermo-

philic temperatures MM was the direct volatile sulfur con-

taining product of anaerobic methionine degradation. The

wat e r r e s e a r c h 4 6 ( 2 0 1 2 ) 5 3 9e5 4 6546

generated MM was demonstrated to be methylated to form

DMS. Both MM and DMS were found to be degraded by the

methanogenic population. At both temperatures, the degra-

dation of DMS was slower than that of MM. Inhibition was

observed in thermophilic VOSC degradation when initial

masses of MM was above 2 mg S/g VSS and DMS was above

3 mg S/g VSS. Mixed-second order kinetics were able to best

describe MM generation, MM methylation, and MM and DMS

degradation in anaerobic methionine incubation when

digested sludge was employed as the inocula. The kinetic

constant for MM generation at 55 �C was about 2.2 times that

of the kinetic constant at 35 �C, which suggested a more rapid

MM release at thermophilic temperatures. The kinetic

constants of MM methylation and MM and DMS degradation

at 35 �C were about 3.9, 3.4, and 2.7 times of those observed at

55 �C, which reflected the faster disappearance of MM and

DMS at mesophilic temperature. The model developed in the

present study which combined four VOSC conversion

processes was able to predict VOSC generation and subse-

quent degradation in methionine digestion at two different

temperatures. Inhibition of VOSC degradation whichmight be

caused by high initial VOSC masses was not included in the

present model because in typical sludge digestion systems

accumulated MM and DMSmasses were found to be less than

2e3 mg S/g VSS and hence would not be inhibitory.

r e f e r e n c e s

APHA, AWWA, WEF, 1992. Standard Method for the Examinationof Water and Wastewater, eighteenth ed..

Baena, S., Fardeau, M.L., Labat, M., Ollivier, B., Thomas, P.,Garcia, J.-L., Patel, B.K.C., 1998. Aminobacterium colombiense gen.nov. sp. nov., an ammonia acid-degrading anaerobe isolatedfrom anaerobic sludge. Anaerobe 4, 241e250.

Bonnarme, P., Psoni, L., Spinnler, H.E., 2000. Diversity of l-methionine catabolism in cheese-ripening bacteria. Appliedand Environmental Microbiology 66, 5514e5517.

Chen, Y., Higgins, M.J., Maas, N.A., Murthy, S.N., Toffey, W.E.,Foster, D.J., 2005. Roles of methanogens on volatile organicsulfur compound production in anaerobically digestedwastewater biosolids.Water Science andTechnology 52, 67e72.

De Bok, F.A.M., van Leerdam, R.C., Lomans, B.P., Smidt, H.,Lens, P.N.L., Janssen, A.J.H., Stams, A.J.M., 2006. Degradationof methanethiol by methylotrophic methanogenic archaea ina lab-scale upflow anaerobic sludge-blanket reactor. Appliedand Environmental Microbiology 72 (12), 7540e7547.

Derbali, E., Makhlouf, J., Vezina, L.-P., 1998. Biosynthesis of sulfurvolatilecompounds inbroccoli seedlingsstoredunderanaerobicconditions. Postharvest Biology and Technology 13 (3), 191e204.

Drotar, A., Burton Jr., G.A., Tavernier, J.E., Fall, R., 1987.Widespread occurrence of bacterial thiol methyltransferasesand the biogenic emission of methylated sulfur gases. Appliedand Environmental Microbiology 53 (7), 1626e1631.

Du, W., 2010. Modeling volatile organic sulfur compounds inanaerobic digestion. Univeristy of Waterloo, Waterloo,Ontario, Canada. Ph.D Thesis.

Fedorovich, V., Lens, P., Kalyuzhnyi, S., 2003. Extension ofanaerobic digestion model no. 1 with processes of sulfatereduction. Applied Biochemistry and Biotechnology 109, 33e45.

Finster,K.,Tanimoto,Y.,Bak, F., 1992. Fermentationofmethanethioland dimethylsulfide by a newly isolated methanogenicbacterium. Archives of Microbiology 157, 425e430.

Higgins, M.J., Chen, Y.-C., Yarosz, D.P., Murthy, S.N., Maas, N.A.,Glindemann, D., Novak, J.T., 2006. Cycling of volatile organicsulfur compounds in anaerobically digested biosolids and itsimplications for odors. Water Environment Research 78 (3),243e253.

Hullo, M.-F., Auger, S., Soutourina, O., Barzu, O., Yvon, M., 2007.Conversion of methionine to cysteine in Bacillus subtilis and itsregulation. Journal of Bacteriology 189 (1), 187e197.

Iranpour, R., Cox, H.H., Fan, S., Abkian, V., Kearney, R.J.,Haug, R.T., 2005. Short-term and long-term effects ofincreasing temperatures on the stability and the production ofvolatile sulfur compounds in full-scale thermophilic anaerobicdigesters. Biotechnology & Bioengineering 91 (2), 199e212.

Kadota, H., Ishida, Y., 1972. Production of volatile sulfurcompounds by microorganisms. Annual Review ofMicrobiology 26, 127e138.

Kalyuzhnyi, S., Fedorovich, V., Lens, P., Pol, L.H., Lettinga, G.,1998. Mathematical modeling as a tool to study populationdynamic between sulfate reducing and methanogenicbacteria. Biodegradation 9, 187e199.

Kalyuzhnyi, S., Fedorovich, V., 1998. Mathematical modeling ofcompetition between sulfate reduction and methanogenesisin anaerobic reactors. Bioresource Technology 65 (3),227e242.

Kiene, R.P., Oremland, R.S., Catena, A., Miller, L.G., Capone, D.G.,1986. Metabolism of reduced methylated sulfur compounds inanaerobic sediments and by a pure culture of an estuarinemethanogen. Applied and Environmental Microbiology 52 (5),1037e1045.

Kiene, R.P., Visscher, P.T., 1987. Production and fate ofmethylated sulfur compounds from methionine anddimethylsulfoniopropionate in anoxic salt marsh sediments.Applied and Environmental Microbiology 53 (10), 2426e2434.

Lomans, B.P., Op den Camp, H.J.M., Pol, A., Van der Drift, C.,Vogels, G.D., 1999. Role of methanogens and other bacteria indegradation of dimethyl sulfide and methanethiol in anoxicfreshwater sediments. Applied and EnvironmentalMicrobiology 65 (5), 2116e2121.

Lomans, B.P., Pol, A., Op den Camp, H.J.M., 2002. Microbial cyclingof volatile organic sulfur compounds in anoxic environments.Water Science and Technology 45 (10), 55e60.

Pianotti, R., Lachette, S., Dills, S., 1986. Desulfurization of cysteineand methionine by Fusobacterium nucleatum. Journal of DentalResearch 65 (6), 913e917.

Smet, E., Langenhove, H.V., 1998. Abatement of volatile organicsulfur compounds in odorous emissions from the bio-industry. Biodegradation 9, 273e284.

Sreekumar, R., Al-Attabi, Z., Deeth, H.C., Turner, M.S., 2009.Volatile sulfur compounds produced by probiotic bacteria inthe presence of cysteine or methionine. Letters in AppliedMicrobiology 48 (6), 777e782.

Thauer, R.K., 1998. Biochemistry of methanogenesis: a tribute toMarjory Stephenson. Mcrobiology 144, 2377e2406.

Van Leerdam, R.C., de Bok, F.A.M., Lomans, B.P., Stams, A.J.M.,Lens, P.N.L., Janssen, A.J.H., 2006. Volatile organic sulfurcompounds in anaerobic sludge and sediments:biodegradation and toxicity. 25 (12), 3101e3109.

Visser, A. 1995. The Anaerobic Treatment of Sulfate ContainingWastewater. Ph. D. thesis, Wageningen Agriculture University,Netherlands.

Wheeldon, I., Caners, C., Karan, K., Peppley, B., 2007. Utilization ofbiogas generated from Ontario wastewater treatment plantsin solid oxide fuel cell systems: a process modeling study.International Journal of Green Energy 4 (2), 221e231.

Zinder, S.H., Brock, T.D., 1978. Methane, carbon dioxide, andhydrogen sulfide production from the terminal methiol groupof methionine by the anaerobic lake sediments. Applied andEnvironmental Microbiology 35 (2), 344e352.