ARTICLE IN PRESS

Available at www.sciencedirect.com

WAT E R R E S E A R C H 4 2 ( 2 0 0 8 ) 1 6 7 5 – 1 6 8 3

0043-1354/$ - see frodoi:10.1016/j.watres

�Corresponding auE-mail addresses1 Current addres

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

Acidogenic sequencing batch reactor start-up proceduresfor induction of 2,4,6-trichlorophenol dechlorination

Cheok Hong Mun, Wun Jern Ng�,1, Jianzhong He

Division of Environmental Science & Engineering, National University of Singapore, 10 Kent Ridge Crescent, Singapore 119260, Singapore

a r t i c l e i n f o

Article history:

Received 2 June 2007

Received in revised form

14 October 2007

Accepted 16 October 2007

Available online 18 October 2007

Keywords:

Acidogenic

SBR

Start-up

pH

Chlorophenol

Dechlorination

nt matter & 2007 Elsevie.2007.10.019

thor. Tel.: +65 6790 6813; f: g0304861@nus.edu.sg (s: Nanyang Technologica

a b s t r a c t

Dechlorination of 2,4,6-trichlorophenol to 4-chlorophenol under acidogenic conditions (pH

5.6–6.5) was successfully induced by manipulating the start-up procedure of an acidogenic

sequencing batch reactor (SBR). A stepwise pH reduction from neutral to acidic level during

start-up was crucial for inducing dechlorination. Once induced, dechlorination can proceed at

pH as low as 5.6 before inhibition occurrs. Optimum pH for maximum dechlorination rate

ranged from 6.0 to 6.3. High primary (sucrose) to secondary (2,4,6-trichlorophenol) substrate

ratio failed to induce dechlorination. Instead, dechlorination occurred at primary to

secondary substrate ratios of less than 103 M/M. A specific maximum trichlorophenol loading

rate of 60mmol/g MLVSS d was achieved before inhibition appeared with onset of acidogenic

reactor failure. T-RFLP profile analysis gave evidence that the start-up procedure resulted in

the selection of an appropriate microbial community, which resulted in the successful

development of an acidogenic consortium capable of degrading 2,4,6-trichlorophenol.

& 2007 Elsevier Ltd. All rights reserved.

1. Introduction

Chlorophenols represent a major class of chlorinated pollu-

tants in industrial wastewaters and leachates generated from

landfill (Ozkaya, 2005; Savant et al., 2006), which can be highly

toxic to biological treatment systems due to their biocidal effect

(Magar et al., 1999). To date, anaerobic reductive dechlorination

is the preferred treatment method for chlorophenols. Previous

studies have indicated that a group of dechlorinators (e.g.

Desulfitobacterium sp., Desulfomonile sp.) utilize chlorophenols as

an electron acceptor and volatile fatty acids, simple organics

(lactate, pyruvate), or H2 as the electron donor (Bouchard et al.,

1996; Perkins et al., 1994; Villemur et al., 2006).

However, methanogens typically coexist with these de-

chlorinators in an anaerobic process and are in direct

competition for the electron donors (i.e. acetate or H2). It

was also found the methanogens often outcompete the H2-

consuming dechlorinators, which resulted in incomplete

r Ltd. All rights reserved.

ax: +65 6791 0756.C.H. Mun), wjng@ntu.edul University, 50 Nanyang

dechlorination (Yang and McCarty, 1998). In an attempt to

enrich the H2-consuming dechlorinators, researchers have

found that a low dose of H2 could selectively encourage and

prevent the growth of dechlorinators and methanogens,

respectively. This was due to the dechlorinators’ lower H2

threshold (Fennell et al., 1997; Loffler et al., 1999). However,

applying such a strategy in the treatment of industrial

wastewaters or leachate is not feasible due to the constant

influx of high-strength organic waste into the bioreactor. In

addition, the low H2 feed concentration would also mean that

dechlorinators will grow slowly and can be washed out easily

from the bioreactor. Moreover, there are industry situations

where acidic effluents containing chlorophenols are gener-

ated (Ozkaya, 2005) and conventional anaerobic processes

(which are sensitive to low pH conditions) may not then

adequately remove the pollutants.

In view of this limitation, the authors have proposed an

alternative strategy: culturing the anaerobic mixed culture in

.sg (W.J. Ng), esehj@nus.edu.sg (J. He).Avenue, Block N1-B3b-29, Singapore 639798, Singapore.

ARTICLE IN PRESS

WAT E R R E S E A R C H 4 2 ( 2 0 0 8 ) 1 6 7 5 – 1 6 8 31676

an acidic environment (i.e. acidogenic condition) for dechlor-

ination. This offers two main advantages: (1) the lower pH

would inhibit methanogen growth and (2) the fermentative

microorganisms can establish a symbiotic relationship with

the dechlorinators in which organics are fermented into

volatile fatty acids and H2 for dechlorinator consumption. The

dechlorinators could then be selectively enriched, assuming

that dechlorinators are not inhibited at low pH. Although pure

culture studies of Desulfitobacterium sp. showed that the

optimum pH for growth had been at slightly alkaline

conditions, there had also been indication that dechlorination

had been possible at slightly acidic pH (pH o7.0) (Bouchard

et al., 1996; Sanford et al., 1996). More recently, Piringer and

Bhattacharya (1999) tried to use such an approach for

treatment of pentachlorophenol (PCP). Unfortunately, they

reported failure of the system due to PCP toxicity but did

suggest that modification in their operating protocol may be

required to achieve dechlorination of chlorophenol under

acidogenic condition.

In this study, the authors investigated the possibility of

2,4,6-trichlorophenol (TCP) dechlorination under acidogenic

condition and have reported the protocol for successful start-

up of an acidogenic bioreactor for treatment of TCP. TCP was

used as the model compound to initiate the study due to its

lower toxicity on acidogens as compared to PCP, if successful

the study can then be extended to include PCP.

2. Materials and methods

2.1. Setup of sequencing batch reactor

Four SBRs (2-L Quickfit culture vessel) with 1-L working

volume each were set up. The mixed liquor in each reactor

was homogeneously mixed with a magnetic stirrer and kept

at ambient room temperature (24–32 1C but typically at 29 1C).

The reactors were wrapped with aluminum foil to prevent

photolytic degradation of the chlorophenol. Biogas was

collected using gas bags (Tedlar bag, SKC). All the reactors

were operated on a 12-h cycle consisting of 10 min fill, 10 h

40 min react, 1 h settle and 10 min decant. The hydraulic

retention time was 2 days.

The basic feed consisted of (units in mg/L) sucrose (9000),

NaHCO3 (4000), CaCl2 � 2H2O (110), MgCl2 � 6H2O (125), NH4Cl

(430), K2HPO4 (90), KH2PO4 (30), Na2SO4 (66), FeCl3 � 6H2O (7.5),

and a trace elements supplement (1 mL/L). The trace ele-

ments supplement contained (units in g/L) CoCl2 � 6H2O

(0.125), H3BO3 (1.25), MnCl2 � 2H2O (2.5), NaMoO3 (0.1), NiCl2 �

6H2O (1.25), ZnCl2 (1.25), and thiamine (2.0). Subsequently, a

vitamin supplement was also used, which contained (units in

mg/L) biotin (20), folic acid (20), nicotinic acid (50), panthothe-

nic acid (50), p-aminobenzoic acid (50), pyridoxine HCl (100),

riboflavin (50), and vitamin B12 (1). TCP would first be dosed at

20 mg/L (100 mM).

2.2. Start-up procedure

Seed sludge was obtained from an anaerobic digester at a

local sewage treatment plant receiving a combination of

domestic and industrial wastewaters. The seed sludge storage

container headspace was first purged with N2 before being

refrigerated at 4 1C. This sludge was used for the seeding

of the reactors in experiments 1, 2, 3, and 4. During and after

the transfer of the seed sludge, the storage container head-

space was continuously purged with N2 to minimize its

exposure to oxygen. Before seeding the reactor, the mixed

liquor was filtered through a 600mm sieve. The reactor was

purged with N2 after seeding. pH was monitored daily and

corrected using 1 M sodium hydroxide solution to the desired

pH range. The initial pH of the mixed liquor in the storage

container prior to seeding was approximately 7.6. Periodically,

reactor effluents were sampled and analyzed for TCP and its

metabolites. The start-up procedure was deemed to be

inappropriate if there was no sign of dechlorination after

three times the average mean cell residence time (MCRT) and

if the reactor already has a stable performance in terms of

stable residual TCP concentration in the effluent, VFAs

production, and stable pH. The MLSS at steady state typically

ranged from 5100 to 6700, averaging around 6000 mg/L.

Initial MLVSS/MLSS ratio was approximately 0.67 and subse-

quently it increased and stabilized to an average of 0.8470.4

after 16 days.

2.3. Experimental phase

The study was divided into four parts as shown in Table 1.

During experiment 1, reactor 1 was operated at pH 5.5 right

from the start and fed with the basic medium (i.e. sucrose

concentration of 25 mM, TCP concentration of 100mM). During

experiment 2, a fresh sample of sludge was obtained from the

same anaerobic digester that was collected at the start of

experiment 1. The feed medium in experiment 2 was

modified and had a reduced sucrose concentration (i.e. a

10-fold decrease to 2.5 mM) but an additional vitamin

supplement. Mixed liquor pH was allowed to drop in a

stepwise manner from 7.6 to 5.6 at a rate of 0.5 units per

week. Reactor 2 was operated in the same manner except that

TCP was first dissolved in methanol and then fed into the

reactor. Methanol concentration in the feed was 0.1%.

Reactors 3 and 4 were operated similarly to reactors 1 and 2,

respectively, except that pH was adjusted immediately to pH

5.5 at day 0. During experiment 3, reactor 1 was seeded with

TCP-dechlorinating acidogenic sludge obtained from experi-

ment 2 and had its specific TCP loading rate increased to

60 mmol/g MLVSS d. Reactors 2 and 3 were seeded with the

seed sludge collected at the start of experiment 1 to

investigate the effect of pH on the start-up procedure. Reactor

2 was used as a positive control with a stepwise reduction in

pH, whereas reactor 3 had its pH controlled at 6.0 from the

onset. In experiment 4, reactors 2–4 were again seeded with

the same seed sludge collected at the start of experiment 1 to

investigate the effect sucrose-loading rate and vitamin

supplementation had on dechlorination. The pH was allowed

to drop in a stepwise manner for all the reactors in

experiment 4. Reactor 2 was used as a positive control,

reactor 3 was fed with high sucrose concentration (i.e. basic

feed medium with the vitamin supplementation), and reactor

4 was fed with the reduced sucrose feed (i.e. 2.5 mM) and

without vitamin supplementation.

ARTICLE IN PRESS

Table 1 – Experimental protocol for investigating factors affecting TCP dechlorination

Reactor Experiment 1 Experiment 2 Experiment 3 Experiment 4

1 pH 5.5, high

sucrose loada

Stepwise reduction in pH, low

sucrose loadb

Increased TCP loading

to 300 mM/d

–

2 – Stepwise reduction in pH, low

sucrose load, TCP dissolved in

methanol

Stepwise reduction in

pH, low sucrose load

Stepwise reduction in

pH, low sucrose load

3 – pH 5.5, low sucrose load pH 6.0, low sucrose load Stepwise reduction in

pH, high sucrose load

4 – pH 5.5, low sucrose load, TCP

dissolved in methanol

– Stepwise reduction in

pH, low sucrose load,

no vitamin supplement

a High sucrose loading is equal to 12.5 mM/d.b Low sucrose loading is equal to 1.25 mM/d.

WA T E R R E S E A R C H 4 2 ( 2 0 0 8 ) 1 6 7 5 – 1 6 8 3 1677

2.4. Batch test

The effects of pH and specific inhibitors on the dechlorination

rates were investigated using 20 mL sample vials (Agilent) as

described in Mun et al. (2007). Acidogenic sludge that already

has TCP dechlorinating activity was used for all batch test

experiments. The sludge was obtained from reactor 1 in

experiment 2 after the reactor had recovered from the pH

inhibition at 5.3 and reached steady state at pH 5.9 after day

145. The aim of the pH variation experiment was to determine

the optimum pH for dechlorination and to investigate

whether sharp deviation from the acclimation pH in the

SBR will affect the rate of dechlorination. Since the acclima-

tion pH in the SBR may or may not be the optimum pH, the

result from the batch test study will provide further evidence

on the optimum TCP dechlorination at acidic pH. pH values

of 5.0, 5.3, 5.5, 5.8, 6.0, 6.3, 6.5, 6.8, and 7.0 were tested.

4-Morpholine ethanesulfonic acid (MES) (1 mM) and sodium

bicarbonate were used as the pH buffer for the range of

5.0–7.0. The solution pH in the acidogenic culture was

adjusted using 0.1 M of NaOH and 0.1 M HCl before the start

of the experiment in a 100% N2-filled anaerobic glove box

(Plas Labs, Michigan) and the sample vials were sealed with

Teflon-lined butyl rubber stopper and aluminum crimp cap.

The culture was mixed using a swivel roller mixer and

sampled periodically. Inhibitors, 10 mM bromoethanesulfonic

acid (BES) for methanogens, 2 mM molybdate for sulfate-

reducing bacteria, and 0.14 mM vancomycin for Gram-posi-

tive bacteria, were tested on the acidogenic culture to

determine their effect on dechlorination activity. Autoclaved

sludge was used as a control. There was no loss of

chlorophenol due to abiotic reasons.

2.5. Theoretical calculation of changes in Gibbs free energy

A theoretical calculation on the changes in Gibbs free energy

of formation (DGo) under acidic condition was performed to

determine if dechlorination at acidic pH was possible. Due to

deprotonation of TCP and 4-chlorophenol (4-CP), there are

three possible reactions that could govern DGo (as listed in

Eqs. (1)–(3)). By taking into account the relative distribution

(Eqs. (4a) and (4b)) of the different forms of TCP and 4-CP at

different pH ranges, the total change in Gibbs free energy of

formation (DGo) could be calculated by the summation of DGo

of either Eqs. (1) and (2) or Eqs. (2) and (3) (Eq. (5)). Details of

the calculation are found in the supplementary information.

DGo is calculated based on the following conditions: H2 as the

sole electron donor; H2 is used because of its representative-

ness in calculation, other electron donors will also yield the

same relationship, all reactants and products except for H+

concentrations have a concentration of 1 M and temperature

set at 25 1C. Values of Gibbs free energy formation were

obtained from Dolfing and Harrison (1992) and Madigan et al.

(2003).

TCPþ 2H2 ¼ 4CPþ 2Hþ þ 2Cl�; (1)

TCP� þ 2H2 ¼ 4CPþHþ þ 2Cl�; (2)

TCP� þ 2H2 ¼ 4CP� þ 2Hþ þ 2Cl�; (3)

a0 ¼½Hþ�

½Hþ� þ Ka, (4a)

a1 ¼Ka

½Hþ� þ Ka, (4b)

DGo¼ a0DGo

1 þ a1DGo2. (5)

2.6. Analytical methods

TCP and its metabolites were identified and quantified using

gas chromatography mass spectrometry equipped with a

single quadrupole analyzer (QP 2010, Shimadzu, Japan). Solids

were removed from samples by centrifuging at 10,000 rpm for

10 min. The resulting supernatant was then subjected to

liquid–liquid extraction (ratio of 1:1) using methylene chloride

after acidification with 10mL of pure methanoic acid. A

capillary column, DB-5 (30 m�0.25 mm i.d. and 0.25mm film

thickness), was used for separation. The GC–MS temperature

profile used was 40–200 1C at 15 1C/min and held for 1 min

at 200 1C. Helium at 1.92 mL/min was used as the carrier gas.

2,6-Dibromophenol served as the internal standard. Mass

ARTICLE IN PRESS

WAT E R R E S E A R C H 4 2 ( 2 0 0 8 ) 1 6 7 5 – 1 6 8 31678

spectra and the retention time of individual mono- and

dichlorophenol isomers were used to identify the metabolites

produced. The chemicals used were all purchased from Sigma

Aldrich. Detection methods for biogas and VFA have been

described in Mun et al. (2007) while hydrogen sulfide was

detected by gas chromatography with a flame photometric

detector.

2.7. DNA extraction and T-RFLP analysis

One milliliter of mixed liquor was obtained from each reactor

after 60 days of operation. The sampling criterion was based

on the volumetric turnover rate, over which there was more

precise control. When sampling, all the reactors’ performance

was stable (Table 2). The criterion based on MCRT was not

used because of the variability in the MCRT during the start-

up phase—with a standard deviation of 4 days, making

determining of when to sample difficult. This problem is

further complicated by the dynamic and transient nature of

microbial community composition even when the reactor

performance is stable (Fernandez et al., 1999; Saikaly and

Oerther, 2004; Zumstein et al., 2000). The sludge was washed

with 1� phosphate-buffered saline and stored at �20 1C prior

to extraction. DNA extraction was carried out in accordance

with Godon et al. (1997). The 16S rRNA gene was amplified

using the universal forward primer 27F (50-Cy5-AGA GTT TGA

TCC TGG CTC AG-30) and reverse primer 1510R (50-GGT TAC

CTT GTT ACG ACT T-30) (Lane, 1991). Each reaction was

performed with 40 ng template DNA in a total volume of 100 ml

containing 1� PCR buffer with 0.25 mM MgCl2, 0.2 mM of each

primer, 0.2 mM of dNTP and 2 U Hot start DNA polymerase

(DyNAzyme II, Finnzymes). PCR amplification was performed

with initial hot-start denaturation for 10 min at 95 1C followed

by 30 cycles with denaturation for 30 s at 95 1C, annealing for

45 s at 55 1C and extension for 1 min at 72 1C, with a final

extension for 7 min at 72 1C. PCR products were cleaned up by

Table 2 – Factors affecting the TCP dechlorination under acido

Experiment/reactora

Factors No. ofruns

Dec

E2/R1, E3/R2, E4/R2 Positive controlc 3

E4/R3 High P/S 1

E3/R3 pH 6.0 from day 0 1

E2/R3 pH 5.5 from day 0 1

E4/R4 No vitamin

supplement

1

E2/R2 Methanol as

cosolvent

1

a E represents the experiment phase and R represents the reactor numbb Positive sign means complete transformation of TCP to 4-CP and nega

MCRT.c Positive control run is defined as the acidogenic reactor that was fed w

stepwise manner.d Experience settling problem during the start-up of the reactor.

utilizing a QIAquick PCR purification kit (Qiagen, Chatsworth,

CA). Subsequently, the PCR product was digested separately

using MspI and RsaI restriction enzymes at 37 1C for 3 h

according to the manufacturer’s instruction (New England

Biolabs, United States). The digested products were desalted

using ethanol precipitation with glycogen as a carrier. The

size and intensity of the terminal fragments were determined

using capillary electrophoresis (CEQ 8000 automated sequen-

cer, Beckman Coulter) as described by Pang and Liu (2006).

The peak heights of each distinct fragment were used to

calculate relative abundances. Relative abundance of any

terminal fragment less than 1% would not be reflected in the

microbial community profile. A cluster analysis of the T-RFLP

fingerprint was performed using MINITAB Statistical soft-

ware. Euclidean distance was calculated after square root

transformation of the relative abundance for each terminal

fragment and a dendrogram was generated using Ward’s

hierarchical-clustering method.

3. Results

3.1. Changes in Gibbs free energy formation under acidiccondition

The relationship between DGo (H2 as the electron donor and

TCP as the electron acceptor) and pH is shown in Fig. 1. It is

found that lowering the pH from 9.0 to 5.0 would only have

resulted in an energy reduction of 20 kJ/mol (i.e. approxi-

mately 6% of the potential amount of energy that could have

been derived from the metabolic reaction shown earlier in

Section 2). Thus dechlorination can proceed because there

would have been little loss of energy due to lowering pH. This

hypothesis was further investigated in the subsequent

experimental work.

genic condition

hlorination activityTCP to 4-CPb

Lag time fordegradation

of TCP (days)

AverageMCRT (days)

+ 14–23 24

� – 27

� – 21

� – 13d

+ 16 30

+ 45 26

er.

tive sign means no sign of dechlorination even after three times the

ith a low P/S ratio and vitamin supplement, and pH was reduced in a

ARTICLE IN PRESS

WA T E R R E S E A R C H 4 2 ( 2 0 0 8 ) 1 6 7 5 – 1 6 8 3 1679

3.2. Stepwise pH reduction to induce 2,4,6-trichlorophenoldechlorination under acidic condition

In experiment 1, an acidogenic reactor (reactor 1) was

acclimated with TCP at pH 5.5 from day 0. The sharp drop

in pH from onset inhibited methanogenesis quickly (Oh et al.,

2003). However, reactor 1 could not dechlorinate TCP after

nearly 3 months of operation (results not shown). From these

preliminary results and the literature (Armenante et al., 1993),

a shock lowering of pH could also have inhibited dechlorina-

tion. The start-up procedure was, subsequently, modified and

3 4 5 6 7 8 9 10 11 12 13

-360

-340

-320

-300

-280

-260

-240

-220

-200

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4TCP to 4CP TCP- to 4CP TCP- to 4CP-

[TCP]

[TCP-]

[4CP]

[4CP-]

ΔG°

pH

ΔG°

(kJ

/mol)

Spe

cie

s d

istr

ibutio

n

Fig. 1 – Relationship between pH and DGo of TCP and H2.

0 7 14 21 28 35 42 49 56 63

0

20

40

60

80

100

120

0

20

40

60

80

100

Time (d)

2,4

,6 -

TC

P (

μM)

Me

tab

olit

es (

μM)

0 7 14 21 28 35 42 49 56 63

5.0

5.5

6.0

6.5

7.0

7.5

8.0

pH

2,4,6 -TCP 2,4- DCP 4-CP

Time (d)

Fig. 2 – Effect of pH on acidogenic dechlorination. (A) Stepw

(B) inhibition of dechlorination at pH 5.3.

in experiment 2, the pH was decreased in a stepwise manner

(pH value of 0.5 per week) from 7.6 to 6.5 (Fig. 2A). Reactor 1

then showed dechlorination after the 14th day of operation at

pH 6.5 (Fig. 2A). 2,4-Dichlorophenol (2,4-DCP) was the

dominant metabolite within 28 days, which was completely

transformed to 4-CP within 50 days. The reactor pH was then

operating at 5.8. The sludge was sampled after 10.5 h of

reaction—near the end of the react phase—at pH 5.8 and

was subjected to solid-phase extraction to determine if TCP or

2,4-DCP had adsorbed onto the biomass in accordance with

the protocol (in preparation for publication). TCP was not

detected on the biomass, although approximately 0.3 and

0.5 mg/g of 2,4-DCP and 4-CP, respectively, were found

adsorbed onto the biomass obtained from reactor 1, experi-

ment 2. Similar solid-phase extractions were done for rest of

the acidogenic SBRs that had different start-up procedures

(results not shown). It was found that for an acidogenic SBR

that developed TCP dechlorinating activity, removal of TCP

and 2,4-DCP via adsorption was negligible.

In order to determine the lower pH limit for dechlorination,

the pH was later decreased to 5.3 on day 91. At this point,

inhibition occurred and 2,4-DCP accumulated, following by

accumulation of TCP (Fig. 2B). Although pH was adjusted back

to 5.9 on day 97, dechlorination activity could not be

recovered and had completely ceased on day 100. After

maintaining reactor 1 at pH 5.9 for 30 days at the same

operating condition, it fully recovered from the pH inhibition

Time (d)

2,4

,6 -

TC

P (

μM)

Me

tab

olit

es (

μM)

5.0

5.5

6.0

6.5

7.0

7.5

8.0

pH

84 88 92 96 100 104 108 112 116 120 124

0

20

40

60

80

1002,4,6-TCP 2,4- DCP 4-CP

0

20

40

60

80

100

Lowering of pH to 5.3

84 88 92 96 100 104 108 112 116 120 124

Time (d)

ise reduction in pH coupled with TCP dechlorination and

ARTICLE IN PRESS

0 10 20 30 40 50 60

0

5

10

15

20

25

30

35

40

2,4,6-TCP removal

50

60

70

80

90

100

110

4-CP produced

Specific Loading rate of 2,4,6-TCP

(μmoles /g MLVSS. d)

4-C

P p

roduced

(μm

ole

s/g

ML

VS

S.d

)

2,4

,6-T

CP

rem

oval effic

iency (

%)

Fig. 4 – Specific loading rate of TCP on acidogenic bioreactor.

0 4 8 12 16 20 24

0

5

10

15

20

25

30

2,4,6-TCP

2,4-DCP

4-CP

30

40

50

60

70

80

90

100

Time (hrs)

2,4

,6-T

CP

and 2

,4,-

DC

P (

μM)

4-C

P (

μM)

Fig. 3 – Kinetics of TCP degradation to 4-CP at pH 6.0.

WAT E R R E S E A R C H 4 2 ( 2 0 0 8 ) 1 6 7 5 – 1 6 8 31680

on day 120. The presence of intermediate 2,4-DCP can,

therefore, be used as indicator in accessing acidogenic

dechlorination performance.

Apparently, pH strongly influenced the acidogenic bio-

mass’s ability to dechlorinate TCP. Thus, the working pH

range for TCP dechlorination was determined using the batch

serum bottle test. The maximum dechlorination rates were

found to occur at pH 6.0 and 6.3. For example, Fig. 3 illustrates

the TCP degradation kinetics at pH 6.0, fed with 25mM of TCP

and 658mM of sucrose. After 8 h of treatment, it was found

that TCP was completely transformed to 4-CP. Dechlorination

did not occur at pH 5.0, whereas at pH 5.3, the reaction was

approximately five times slower than the maximum dechlor-

ination rate. Dechlorination was also found to proceed at

neutral pH although the rate was then approximately two

times slower. Since the seed sludge was obtained from reactor

1, experiment 2 after day 145, it was likely that both the

acidogens and the dechlorinators has been enriched and

acclimated to the pH of 5.8 and 6.0. With the necessary

dechlorination enzymes present, the pH inhibition at 5.3 or at

7.0 was a reflection on how pH variations can affect

dechlorination activity.

3.3. Start-up procedure favorable for acidogenic TCPdechlorination

In experiment 2, acidogenic TCP dechlorination was success-

ful by incorporating success factors drawn from experiment

1. These were stepwise reduction in pH, lower sucrose

concentration, and addition of a vitamin supplement. Experi-

ments 2–4 were conducted to determine which factor was

most influential (Table 2). In the three positive control track

runs with stepwise pH reduction (experiment 2/reactor 1;

experiment 3/reactor 2 and experiment 4/reactor 2), the total

lag time for the degradation of TCP ranged from 14 to 23 days

(Table 2). However, for reactors that were operated at pH 6.0

and 5.5 from the beginning (2.5 mM sucrose), no sign of

dechlorination was noted even after 90 days of operation,

indicating that stepwise reduction in pH was crucial in

inducing TCP dechlorination. This is so despite the fact that

maximum dechlorination rate was noted to occur at pH

between 6.0 and 6.3.

The primary to secondary substrate ratio (P/S ratio) was

also an important factor. In experiment 4, when reactor 3 was

fed with a high P/S ratio (25 mM of sucrose with respect to

100mM of TCP) and when pH decreased in a stepwise manner,

dechlorination did not take place. Subsequently, when the

influent sucrose concentrations were reduced (20, 14, 5.6 mM),

dechlorination was only noted at 5.6 mM sucrose even though

the same start-up procedure was used. Thus, a high primary

to secondary substrate ratio was not favorable for induction

of TCP dechlorination. In addition, it was found that

the maximum and minimum P/S ratio, which induced

dechlorination during start-up, was 103 and 17.2 M/M, re-

spectively. When the start-up feed TCP was 150 mM (17.2 P/S

ratio), it required nearly twice as long as compared to the

reactor fed with 100mM TCP (26.3 P/S ratio) to achieve

dechlorination.

Later, when the specific loading rate of TCP in reactor 1 in

experiment 3 was increased to determine its maximum

loading (while keeping sucrose concentration constant), the

reactor successfully removed nearly 98% of TCP up to a

specific loading rate of 39mmol/g MVLSS d (Fig. 4). However,

when 50mmol/g MLVSS d was applied, TCP removal declined

to about 80%. Reactor failure became apparent (indicated by

increasing amount of 2,4-DCP from nearly zero to about 93mM

in the effluent) and TCP accumulated rapidly from 20 to

180mM when TCP was increased to 60mmol/g MLVSS d.

3.4. Dechlorination activity inhibitors

Although the reactor was operated in the acidogenic phase,

there was methane production (10–25% methane in the gas

phase). This was likely due to the long MCRT (420 days) and

the stepwise reduction in pH. The methanogens could have

acclimated to the low pH environment. In order to verify if

methanogens and other sulfate-reducing bacteria were not

responsible for the dechlorination observed, inhibitors on

specific microbial communities were applied. Dechlorination

was severely inhibited by the presence of vancomycin (78%

inhibition). Production of volatile fatty acids (e.g. acetic and

propionic acid) was also reduced drastically (Table 3). How-

ever, with BES, CH4 production was severely inhibited (96%

inhibition), while TCP degradation was unaffected. Molybdate

ARTICLE IN PRESS

Table 3 – Effect of inhibitors on the biodegradation of TCP

Inhibitors Chlorophenols Inhibitiona (%) Metabolic product

TCP (mM) 2,4-DCP (mM) 4-CP (mM) VFAb (mg/L) H2Sc (ppm) CH4d (%)

Vancomycin 21.8 (1.5) 3.2 (2.0) 52.8 (3.4) 78 112 (21) N.D. 0.5 (0.2)

Molybdate 16.5 (1.4) 10.9 (1.1) 53.9 (4.4) 54 255 (19) N.D. 6.8 (0.9)

BES 4.7 (1.5) 5.9 (2.1) 86.7 (3.5) 0.9 321 (27) N.D. 0.3 (0.4)

No addition of inhibitors 4.5 (1.6) 6.5 (0.7) 80 (5.7) – 286 (35) N.D. 7.59 (1.2)

Autoclaved sludge 26.6 (1.0) N.D. 47.8 (3.8) – 45 (34) N.D. N.D.

Note: Dosed TCP and sucrose concentrations are 25 mM and 0.625 mM, respectively (26.3 P/S ratio). Metabolites were measured at the end of 12 h.

Results are the average of triplicates and values in brackets are the standard deviations.a Inhibition calculated with respect to the control where no inhibitors were added.b VFA comprised acetic and propionic acids converted to COD mg/L.c N.D.—not detected.d Gas composition in the head space.

Dis

tance

19.13

12.76

6.38

0.00

Day 0 + Ctl

E4/R2

No vit pH 6

E2/R1

High P/S

E3/R2

+ Ctl

E4/R4 E3/R3

+ Ctl

E2/R2 E4/R3

Methanol

Fig. 5 – Cluster analysis of T-RFLP fingerprints of the

microbial community obtained from different start-up

procedure. Note: ‘‘Day 0’’ represents the seed sludge; E/R

represents the experiment phase and reactor number,

respectively, as described in Tables 1 and 2 in which the

sludge was obtained from; ‘‘+ Ctl’’ represents the three

positive control reactors; ‘‘No Vit’’ represents the reactor

that was not fed with vitamin supplement; ‘‘pH 6.0’’

represents the reactor that operated at pH 6.0 from Day 0,

‘‘Methanol’’ represents the reactor that was fed with TCP

dissolved in methanol with stepwise reduction in pH; ‘‘High

P/S’’ represents the reactor that was fed with high sucrose

to TCP ratio.

WA T E R R E S E A R C H 4 2 ( 2 0 0 8 ) 1 6 7 5 – 1 6 8 3 1681

was found to partially inhibit TCP degradation (54% inhibi-

tion). The acidogenic culture developed did not show any

sulfate-reducing activity (as hydrogen sulfide was not de-

tected in the headspace gas).

3.5. Microbial community profiling

It was hypothesized that the different start-up procedures

might have led to different microbial community structures

and hence the differences in the dechlorination activity. The

microbial community profiles obtained following the various

start-up procedures were compared using cluster analysis.

Cluster analysis revealed that the start-up procedure did lead

to a different microbial community composition even though

all the biomasses were developed from the same seed sludge

(Fig. 5). The first distinct cluster was formed solely by the seed

sludge at day 0. The second distinct cluster was formed by

two positive control reactors (experiment 3/reactor 2 and

experiment 4/reactor 2) and the reactor that had no vitamin

supplement (experiment 4/reactor 4) while the third distinct

cluster was formed by a positive control reactor (experiment

2/reactor 1) and a reactor which was fed with TCP dissolved in

methanol (experiment 2/reactor 2) and a reactor that was

operated at pH 6.0 from the onset (experiment 3/reactor 3).

The last distinct cluster was from the reactor fed with a high

P/S ratio.

4. Discussions

Even though an acidic environment could possibly allow for

enrichment of dechlorinators against methanogens, to date,

studies on anaerobic reductive dechlorination had focused

almost entirely on conditions at pH of 7.0–8.0. This is not

surprising as numerous studies on pure cultures showed the

optimum pH for dechlorination was typically under slightly

alkaline conditions and dechlorination slowed under acidic

conditions (Armenante et al., 1993; Chang et al., 1999;

Villemur et al., 2006). Such observations could have resulted

from the way experiments had been conducted to investigate

pH influence on dechlorination. Usually, biomass with

dechlorinating ability (at neutral pH) was exposed to various

pH environments in serum bottle tests. The sludge concei-

vably did not have adequate opportunity to acclimate to the

sudden pH change and this had led to a reduction in

dechlorination capacity. In this study, the acclimation process

to low pH conditions was found to be the critical element in

inducing dechlorinating activity. A seed sludge, which

demonstrated dechlorinating activity after appropriate start-

up, did not show such activity even after 90 days when

challenged with a sharp pH change to pH 6.0 during start-up.

Interestingly, from the T-RFLP cluster analysis, the microbial

ARTICLE IN PRESS

WAT E R R E S E A R C H 4 2 ( 2 0 0 8 ) 1 6 7 5 – 1 6 8 31682

profile of the reactor that was operated at pH 6.0 at the

beginning was clustered together with the positive control

reactor. As such, it was likely that the sudden pH drop had

inhibited dechlorination activity rather than change in the

microbial community structure.

The results of this study indicated that the pH range of

5.6–6.0 was most effective for inhibiting methanogenic

activity (CH4 composition was then 5–10% versus 60–70% at

neutral pH and at the same TCP exposure) while at the same

time maintaining dechlorination activity, thereby enriching

the dechlorinators. Results from the inhibition test (BES)

suggested that methanogens were not involved in the

dechlorination process. TCP dechlorination may be depen-

dent on fermentation activities as reductive dechlorination

was observed with reduction in VFA generation (results from

vancomycin inhibitor test). This suggested a symbiotic

relationship between the fermentative bacteria and dechlor-

inators, which was also suggested by Lanthier et al. (2005).

They studied the spatial distribution of Desulfitobacterium

hafniense on PCP-degrading granules and found that

D. hafniense was always found together with fermentative

bacteria on the outer layer of granules. The likely reason for

the close proximity between these two groups of microorgan-

isms is to allow rapid electron transfer.

The P/S ratio was the other factor that affected dechlorina-

tion activity during start-up. A high sucrose loading resulted

in a significant change in the microbial community profile

with the relative abundances of the dominant bacteria

different from communities subjected to a lower P/S ratio.

The high sucrose content could have selectively enriched the

fermentative bacteria, leading to their rapid growth and

dominance in the system (unpublished clone library results).

Furthermore, dechlorinators in the acidogenic environment

were found to have slow growth rates (unpublished results

from authors’ enrichment experiments), thus fermentative

bacteria can outcompete dechlorinators for nutrients leading

to unfavorable growth condition for the latter. Further to this,

the reason for the inactivity in dechlorination due to high P/S

ratio is not entirely clear at this stage. Nevertheless, the

results of this study has implications on dechlorination under

acidogenic conditions—there is a limit on the applied non-

chlorinated organic load (and this is also likely relative to the

chlorinated organic load) during start-up and perhaps during

operation as well.

The start-up procedure was compared with other studies

on acidogenic dechlorination of chlorinated organics. Piringer

and Bhattacharya (1999) conducted a treatability study on PCP

dechlorination under acidogenic conditions, in which PCP

toxicity was suggested to have an adverse effect on the

acidogenic biomass capability to dechlorinate the chlorinated

compounds. A closer look at their operating conditions

suggested that there could be other factors at play. The P/S

ratio of 528 M/M was higher than the working P/S ratio that

had induced dechlorination in this study. While their study

did not show clearly how acclimation of the seed sludge was

achieved from the neutral to acidic condition, it was likely

that an inappropriate start-up procedure and P/S ratio, in

addition to PCP toxicity, were the reasons for failure to enrich

the dechlorinating microorganisms under acidogenic condi-

tions. In another similar study, Chin et al. (2005) investigated

2,4-dichlorophenoxyacetic acid degradation at acidogenic

conditions of pH 4.5–5.0. 2,4-Dichlorophenoxyacetic acid

degradation occurred only after a long acclimation period of

100 days. Their operating P/S ratio of 57.9 was within the

working range defined in this study, which suggested the

importance of primary to chlorinated organic ratio. Likewise,

in their study, the acclimation process on how to achieve

dechlorination activity of the seed sludge from the neutral to

acidic condition was not clear. Perhaps, the lower pH range

could have been the reason for the longer acclimation period.

Performance of the acidogenic process was comparable to

the conventional anaerobic process. In this study, the TCP

degradation lag phase ranged from 14 to 45 days for start-up

when pH was reduced in a stepwise manner to pH 5.6–6.0,

whereas the lag time for the conventional anaerobic process

at pH 7.0 was reported to be as short as 5–20 days to as long as

190–215 days (Garibay-Orijel et al., 2005; Majumder and

Gupta, 2007; Ye and Shen, 2004). In terms of treatment

efficiency, the acidogenic reactor achieved 98% removal up to

200mM/d of TCP and started to fail at 300mM/d of TCP.

Maximum treatment efficiency achieved by the anaerobic

process for TCP was reported to an average of 99% for loading

of 198–400mM/d (Armenante et al., 1999; Garibay-Orijel et al.,

2005). The dechlorination pathway of 2,4,6-TCP to 4-CP via

2,4-DCP was similar in both the acidogenic and anaerobic

processes, indicating that the same chlorophenol reductase

was still active under acidic conditions.

5. Conclusion

(1)

This study has demonstrated the possibility of dechlor-

ination of 2,4,6-TCP to 4-CP under acidic conditions and

hence the use of the acidogenic process for treatment of

TCP-contaminated wastewater.

(2)

The start-up protocol is an important feature in successful

acidogenic dechlorination, which established guidelines

for appropriate start-up such as the stepwise reduction in

pH and the sucrose to chlorophenol feed ratio.

(3)

Acidogenic dechlorination offers an alternative strategy

for selection of dechlorinators against methanogens and

provides an alternative strategy for treatment of acidic

effluents with chlorinated organics.

Acknowledgment

The authors thank the two anonymous reviewers for their

valuable comments and suggestions. The research project

was kindly supported by the Academic Research Fund,

Ministry of Education, Singapore, under Project no. R288-

000-002-112.

Appendix A. Supplementary materials

Supplementary data associated with this article can be found

in the online version at doi:10.1016/j.watres.2007.10.019

ARTICLE IN PRESS

WA T E R R E S E A R C H 4 2 ( 2 0 0 8 ) 1 6 7 5 – 1 6 8 3 1683

R E F E R E N C E S

Armenante, P.M., Kafkewitz, D., Jou, C.J., Lewandowski, G., 1993.Effect of pH on the anaerobic dechlorination of chlorophenolsin a defined medium. Appl. Microbiol. Biotechnol. 39 (6),772–777.

Armenante, P.M., Kafkewitz, D., Lewandowski, G.A., Jou, C.-J.,1999. Anaerobic–aerobic treatment of halogenated phenoliccompounds. Water Res. 33 (3), 681–692.

Bouchard, B., Beaudet, R., Villemur, R., McSween, G., Lepine, F.,Bisaillon, J.G., 1996. Isolation and characterization of Desulfi-tobacterium frappieri sp. nov., an anaerobic bacterium whichreductively dechlorinates pentachlorophenol to 3-chlorophe-nol. Int. J. Syst. Bacteriol. 46 (4), 1010–1015.

Chang, B.V., Chiang, C.W., Yuan, S.Y., 1999. Microbial dechlorina-tion of 2,4,6-trichlorophenol in anaerobic sewage sludge.J. Environ. Sci. Health Part B 34 (3), 491–507.

Chin, H., Elefsiniotis, P., Singhal, N., 2005. Biodegradation of2,4-dichlorophenoxyacetic acid using an acidogenic anaerobicsequencing batch reactor. J. Environ. Eng. Sci. 4 (1),57–63.

Dolfing, J., Harrison, B.K., 1992. Gibbs free energy of formation ofhalogenated aromatic compounds and their potential role aselectron acceptors in anaerobic environments. Environ. Sci.Technol. 26 (11), 2213–2218.

Fennell, D.E., Gossett, J.M., Zinder, S.H., 1997. Comparison ofbutyric acid, ethanol, lactic acid and propionic acid ashydrogen donors for the reductive dechlorination of tetra-chloroethene. Environ. Sci. Technol. 31 (3), 918–926.

Fernandez, A., Huang, S., Seston, S., Xing, J., Hickey, R., Criddle, C.,Tiedje, J., 1999. How stable is stable? Function versuscommunity composition. Appl. Environ. Microbiol. 65 (8),3697–3704.

Garibay-Orijel, C., Rios-Leal, E., Garcia-Mena, J., Poggi-Varaldo,H.M., 2005. 2, 4, 6-Trichlorophenol and phenol removal inmethanogenic and partially-aerated methanogenic conditionsin a fluidized bed bioreactor. J. Chem. Technol. Biotechnol. 80(10), 1180–1187.

Godon, J.J., Zumstein, E., Dabert, P., Habouzit, F., Moletta, R., 1997.Molecular microbial diversity of an anaerobic digestor asdetermined by small-subunit rDNA sequence analysis. Appl.Environ. Microbiol. 63 (7), 2802–2813.

Lane, D.J., 1991. 16S/23S rRNA sequencing. In: Stackebrandt, E.,Goodfellow, M. (Eds.), Nucleic Acid Techniques in BacterialSystematics. Wiley, New York, pp. 115–175.

Lanthier, M., Juteau, P., Lepine, F., Beaudet, R., Villemur, R., 2005.Desulfitobacterium hafniense is present in a high proportionwithin the biofilms of a high-performance pentachlorophenol-degrading, methanogenic fixed-film reactor. Appl. Environ.Microbiol. 71 (2), 1058–1065.

Loffler, F.E., Tiedje, J.M., Sanford, R.A., 1999. Fraction of electronsconsumed in electron acceptor reduction and hydrogenthresholds as indicators of halorespiratory physiology. Appl.Environ. Microbiol. 65 (9), 4049–4056.

Madigan, M.T., Martinko, J.M., Parker, J., 2003. Brock Biology ofMicroorganisms, 10th ed. Prentice-Hall, Englewood Cliffs, NJ.

Magar, V.S., Stensel, H.D., Puhakka, J.A., Ferguson, J.F., 1999.Sequential anaerobic dechlorination of pentachlorophenol:competitive inhibition effects and a kinetic model. Environ.Sci. Technol. 33 (10), 1604–1611.

Majumder, P.S., Gupta, S.K., 2007. Removal of chlorophenols insequential anaerobic–aerobic reactors. Bioresour. Technol. 98(1), 118–129.

Mun, C.H., Ng, W.J., He, J., 2007. Evaluation of the biodegradationpotential of carbon tetrachloride and chlorophenols underacidogenic condition. J. Environ. Eng., in press.

Oh, S.E., Van Ginkel, S., Logan, B.E., 2003. The relative effective-ness of pH control and heat treatment for enhancingbiohydrogen gas production. Environ. Sci. Technol. 37 (22),5186–5190.

Ozkaya, B., 2005. Chlorophenols in leachates originating fromdifferent landfills and aerobic composting plants. J. Hazard.Mater. 124 (1–3), 107–112.

Pang, C.M., Liu, W.T., 2006. Biological filtration limits carbonavailability and affects downstream biofilm formation andcommunity structure. Appl. Environ. Microbiol. 72 (9),5702–5712.

Perkins, P.S., Komisar, S.J., Puhakka, J.A., Ferguson, J.F., 1994.Effects of electron donors and inhibitors on reductive de-chlorination of 2,4,6-trichlorophenol. Water Res. 28 (10),2101–2107.

Piringer, G., Bhattacharya, S.K., 1999. Toxicity and fate ofpentachlorophenol in anaerobic acidogenic systems. WaterRes. 33 (11), 2674–2682.

Saikaly, P.E., Oerther, D.B., 2004. Bacterial competition in activatedsludge: theoretical analysis of varying solids retention timeson diversity. Microb. Ecol. 48 (2), 274–284.

Sanford, R.A., Cole, J.R., Loffler, F.E., Tiedje, J.N., 1996. Character-ization of Desulfitobacterium chlororespirans sp. nov., whichgrows by coupling the oxidation of lactate to the reductivedechlorination of 3-chloro-4-hydroxybenzoate. Appl. Environ.Microbiol. 62 (10), 3800–3808.

Savant, D.V., Abdul-Rahman, R., Ranade, D.R., 2006. Anaerobicdegradation of adsorbable organic halides (AOX) from pulpand paper industry wastewater. Bioresour. Technol. 97 (9),1092–1104.

Villemur, R., Lanthier, M., Beaudet, R., Lepine, F., 2006. TheDesulfitobacterium genus. FEMS Microbiol. Rev. 30 (5), 706–733.

Yang, Y., McCarty, P.L., 1998. Competition for hydrogen within achlorinated solvent dehalogenating anaerobic mixed culture.Environ. Sci. Technol. 32 (22), 3591–3597.

Ye, F.X., Shen, D.S., 2004. Acclimation of anaerobic sludgedegrading chlorophenols and the biodegradation kineticsduring acclimation period. Chemosphere 54 (10), 1573–1580.

Zumstein, E., Moletta, R., Godon, J.-J., 2000. Examination of twoyears of community dynamics in an anaerobic bioreactorusing fluorescence polymerase chain reaction (PCR) single-strand conformation polymorphism analysis. Environ. Micro-biol. 2 (1), 69–78.