VFAs bioproduction from waste activated sludge by coupling pretreatments with Agaricus bisporus substrates conditioning

Page 1

G

P

Vp

AWa

b

c

a

ARRA

KWAVHC

1

piVcvfic(sirc

EH

nC

l

1h

ARTICLE IN PRESS Model

RBI-9996; No. of Pages 7

Process Biochemistry xxx (2013) xxx– xxx

Contents lists available at ScienceDirect

Process Biochemistry

journa l h om epage: www.elsev ier .com/ locate /procbio

FAs bioproduction from waste activated sludge by couplingretreatments with Agaricus bisporus substrates conditioning

ijuan Zhoua, Jingwen Dua, Cristiano Varronea,b, Youzhao Wanga, Aijie Wanga,c,∗,enzong Liuc,∗∗

State Key Laboratory of Urban Water Resource and Environment, Harbin Institute of Technology (SKLUWRE, HIT), Harbin, ChinaENEA e Italian Agency for New Technologies, Energy and Sustainable Development, ItalyResearch Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing, China

r t i c l e i n f o

rticle history:eceived 30 September 2013eceived in revised form 29 October 2013ccepted 5 November 2013

eywords:aste activated sludge (WAS)

a b s t r a c t

A novel strategy for improving volatile fatty acids (VFAs) bioproduction from waste activated sludge(WAS) was developed by coupling pretreatments with conditioning (CPC). Agaricus bisporus substrate(ABS) was used as external carbohydrate additive source of conditioning step. Pretreatment was studiedin three ways: alkaline, alkaline-thermal and ultrasonic. WAS hydrolysis and protein degradation was dis-tinctly improved by CPC treatments, resulting in a considerable enhancement of VFAs yield. The maximalVFAs yield was 614 ± 71, 712 ± 49 and 598 ± 19 mg COD/g VSS at pre-optimized alkaline, alkaline-thermal

garicus bisporus substrates (ABS)olatile fatty acids (VFAs)ydrolysisonditioning

and ultrasonic CPC treatments, respectively, with an increase of 35%, 50% and 38% compared to theyields of pretreated WAS fermentation. Fourier transformed infrared spectroscopy and three-dimensionalexcitation-emission matrix fluorescence spectroscopy indicated that a synergistic effect occurred in co-digesting WAS and ABS. The conditioning of carbohydrate with feasible pretreatment provided largeroom for the digestibility improvement and the operation cost reduction in the whole WAS treatment

system.

. Introduction

Nowadays, waste activated sludge (WAS), the main solid wasteroduced in the wastewater treatment process, is considered with

ncreasing attention as a cheap but valuable energy resource.olatile fatty acids (VFAs) are important high added-value greenhemicals and preferred substrates for many bioprocesses. Pre-ious researches showed that VFAs can be successfully producedrom pretreated WAS by anaerobic digestion [1,2]. Notwithstand-ng that, VFAs yield is still limited by the insufficient carbohydrateontent in WAS. To solve this problem, the carbon-to-nitrogenC/N) ratio of WAS was balanced by use of external carbon-richubstrate additive conditioning [3]. Lignocellulosic biomass, for

Please cite this article in press as: Zhou A, et al. VFAs bioproduction from wasubstrates conditioning. Process Biochem (2013), http://dx.doi.org/10.1016

nstance, obtained as agriculture byproducts, forest and industrialesidues is an abundant, inexpensive, and renewable source ofarbohydrates [4]. Rughoonundun et al. (2012) investigated that

∗ Corresponding author at: State Key Laboratory of Urban Water Resource andnvironment, Harbin Institute of Technology (SKLUWRE, HIT), P.O. Box 2614, 202aihe Road, Harbin 150090, China. Tel.: +86 451 86282195; fax: +86 451 86282195.

∗∗ Corresponding author at: Research Center for Eco-Environmental Sciences, Chi-ese Academy of Sciences, 18 Shuangqing Road, Haidian District, Beijing 100085,hina. Tel.: +86 451 86282110; fax: +86 451 86282110.

E-mail addresses: [email protected], [email protected] (A. Wang),[email protected] (W. Liu).

359-5113/$ – see front matter. Crown Copyright © 2013 Published by Elsevier Ltd. All rittp://dx.doi.org/10.1016/j.procbio.2013.11.005

Crown Copyright © 2013 Published by Elsevier Ltd. All rights reserved.

co-digestion of mixed sewage sludge and bagasse achieved higheracid yields than each substrate fermented on its own [5]. Feng et al.(2009) and Zhou et al. (2013) reported that the bioconversion ofWAS to VFAs can be enhanced by adding rice straw and corn stoveras carbohydrate substrate, respectively [6,7]. Jia et al. (2013) usedperennial ryegrass as carbon source to adjust C/N radio in WAS toenhance VFAs production [8].

Agaricus bisporus is an edible mushroom, mostly consumed inthe western world, accounting for 98% of all mushrooms producedin the US [9]. The production of each kilogram of mushroom gen-erates approximately five kilograms of Agaricus bisporus substrate(ABS) [10]. About 400,000 tons fresh Agaricus mushrooms wereconsumed in 2011 in the US, which translated into more than 2 mil-lion tons ABS [9]. The US mushroom industry incurs an estimatedcost of US$7 million annually for residual substrates disposal [11].Commercial mushrooms are cultivated using a substrate consistingof wheat straw, corncobs, seed hulls, horse manure, and so on. Thus,it represents an ideal feedstock for the carbohydrate substrate inthe conversion of WAS to VFAs.

Although it has been observed that VFAs production from WAScan be enhanced by pretreatment or co-digesting with carbon-

ste activated sludge by coupling pretreatments with Agaricus bisporus/j.procbio.2013.11.005

rich substrates, WAS pretreatment combined with conditioninghas been seldom investigated. The main objective of this studywas to evaluate the feasibility of improving WAS hydrolysis andacidification by coupling pretreatment with conditioning (CPC).

ghts reserved.

Page 2

ING Model

P

2 ochem

Ecmsdicowsep

2

2

tCbtwudTtAwto

2

csh2[C

dpSppatgUuctU

teflcawt

ARTICLERBI-9996; No. of Pages 7

A. Zhou et al. / Process Bi

xternal ABS addition was taken as the carbohydrate source ofonditioning step. For this reason, three commonly used pretreat-ents, alkaline, alkaline-thermal and ultrasonic, were tested in this

tudy [12]. Optimal conditions were obtained, using the uniformesign (UD). Semi-continuous flow anaerobic co-digestion exper-

ment was conducted for model verification. The presence andhange of main adsorption functional groups in the compositionf the co-digesting WAS and ABS, under the three CPC treatments,as analyzed by Fourier transform infrared (FTIR) spectra. The dis-

olved organic matter was characterized using three-dimensionalxcitation-emission matrix (EEM) fluorescence spectroscopy witharallel factor analysis (PARAFAC).

. Materials and methods

.1. Substrates

Raw WAS was collected directly from the secondary sedimenta-ion tank of Taiping Municipal Wastewater Treatment Plant (Harbinity, Heilongjiang Province, China). Sludge was firstly concentratedy settling for 24 h, and then stored at 4 ◦C in a refrigerator for lesshan 2 weeks. To prevent clogging problems, concentrated sludgeas screened with a 1 mm sieve to remove impurities prior to besed as feed. The main characteristics (average value plus standardeviation of three tests) of the concentrated WAS are displayed inable S1. ABS was obtained from Shuangcheng Mushroom Cultiva-ion base (Harbin City, Heilongjiang Province, China). The choppedBS was dried in the oven at 70 ◦C until constant weight. Then itas milled to 2–10 mm, before storing at room temperature prior to

ests. The main composition (average value plus standard deviationf three tests) of raw ABS is shown in Table S2.

.2. Optimization of process parameters

To increase the digestibility of WAS, three pretreatments wereonducted in this study. Alkaline was carried out by adding 6 Modium hydroxide (NaOH). For thermal treatment, sludge wasydrolyzed in a 10 L autoclave. Ultrasonic was performed with8 + 40 kHz ultrasonicator (Ningbo Scientz Biotechnology, China)13]. After pretreatment, crushed ABS was added for balancing the/N radio of WAS fermentation.

Two parameters, namely alkaline dosage (g/g TSS) and ABSosage (g/g VSS), were optimized in the alkaline pretreatment cou-ling ABS conditioning (AP-CPC) process, each at seven levels (Table3), changing one variable at a time (due to the low number ofarameters). The experiment for optimization of alkaline-thermalretreatment coupling ABS conditioning (ATP-CPC) process wasrranged by design taking into consideration four factors, namelyemperature (X1, ◦C), alkaline dosage (X2, g/g TSS), ABS dosage (X3,/g VSS) and time (X4, min), each at eight levels. The UD table16(164) was applied to arrange the experiments (Table S4). For theltrasonic pretreatment coupling ABS conditioning (UP-CPC) pro-ess, three experimental variables, i.e., energy density (A, kW/L),ime (B, min) and ABS dosage (C, g/g VSS), were optimized using12(123) design (Table S5).

Batch experiments were conducted for experiment optimiza-ion in a series of serum bottles, with 300 mL of the mixed substratesach. The control tests were fed with non-CPC treated sludge. Afterushing with nitrogen gas to remove oxygen, all bottles were

Please cite this article in press as: Zhou A, et al. VFAs bioproduction from wasubstrates conditioning. Process Biochem (2013), http://dx.doi.org/10.1016

apped, sealed, and stirred in an air-bath shaker (100 rpm/min)t 35 ± 2 ◦C. VFAs production with the fermentation time of 3 das taken as the response. All the experiments were performed in

riplicate and mean values were applied.

PRESSistry xxx (2013) xxx– xxx

2.3. Semi-continuous flow experiments for model verification

Semi-continuous flow experiment was conducted in a com-pletely stirred tank reactor for model verification (Fig. S1). Thereactor consisted of two Plexiglas cylinders with 120 mm and160 mm diameters. The inner cylinder, with a volume of 3.0 Land a headspace of 1 L, was used for digestion. The temperatureof 35 ± 2 ◦C was maintained by circulating water from a water-bath heating vessel with a thermostatic controller. The mixing wascarried out by a controlled mechanical agitator. The reactor was ini-tially seeded with the mixture of pretreated WAS and crushed ABSat previously optimized treatment conditions. During the first 10days fermentation, the reactor was operated in batch mode withoutfeeding (Phase 1, day 1–11). From day 11 afterward, the reactor wasfed with fresh mixed substrates, with sludge sampling just beforefeeding (Phase 2, day 12∼day 28). Each reactor was operated withthe same hydraulic retention time of 10 d. In order to minimize pHfluctuations within the digesters, the reactors were fed twice a day.

2.4. Additional experiments on individual WAS and ABSfermentation

In order to better investigate the synergistic effect of WAS andABS (in co-digestion) on enhanced VFAs production in the threeCPC processes, individual WAS and ABS fermentation tests wereconducted in semi-continuous flow mode. All experiments werecarried out in 500 mL glass serum bottles with a working volumeof 300 mL. Some of the fresh sludge was maintained under aera-tion at room temperature overnight and then used as inoculum.The pretreatments of WAS were carried out at previously opti-mized treatment conditions. All other operations were the same asdescribed above. All the experiments were performed in triplicate.

2.5. Analytical methods and data analysis

Sludge samples were centrifuged at 10,000 rpm after fermenta-tion, then filtered through a 0.45 �m cellulose nitrate membranefilter and finally stored at 4 ◦C prior to analysis. The determinationsof SCOD, TCOD, TSS, VSS, carbohydrates and proteins have beenperformed as previously described [7,14]. Analysis of ammonia(NH4

+-N) and phosphorus (PO43−-P) were conducted in accordance

with standard methods [15]. Lipid and oil was determined gravi-metrically after Soxhlet extraction with petroleum ether for 6 h at80 ◦C. A gas chromatography (GC) was utilized to analyze the VFAs[16]. VFAs production was calculated as the sum of the measuredacetic (HAc), propionic (HPr), n-butyric (n-HBu), iso-butyric (iso-HBu), n-valeric (n-HVa) and iso-valeric (iso-HVa) acids. The activityof �-glucosidase and protease was measured according to Mironet al. [17]. The COD conversion factors are 1.50 g COD/g protein(assumed as (C4H6.1O1.2N)x), 1.06 g COD/g carbohydrate (assumedas C6H12O6), 1.07 g COD/g HAc, 1.51 g COD/g HPr, 1.82 g COD/g HBu,and 2.04 g COD/g HVa.

The infrared spectra were recorded using 2–5 mg powder of thefreeze-dried raw/digested substrates with 250 mg of dry potassiumbromide. The instrument used was a PerkinElmer Spectrum One BFTIR spectrometer (Waltham, MA, USA), scanning from 4000 cm−1

to 450 cm−1. The spectra were baseline corrected and normal-ized to 1.0 for comparison. The fluorescence excitation-emissionmatrix (EEM) was measured using a fluorescence spectrometry(FP-6500, Jasco, Tokyo, Japan). Parallel factor analysis (PARAFAC)was used to model EEM fluorescence data in this study [18]. BeforePARAFAC analysis, the Raman scattering was removed by subtrac-

ste activated sludge by coupling pretreatments with Agaricus bisporus/j.procbio.2013.11.005

ting the pure distilled water spectrum from the sample spectrum,and the Rayleigh scattering was overcome by inserting a seriesof zero values in the region of no fluorescence (excitation wave-length � emission wavelength).

Page 3

IN PRESSG Model

P

chemistry xxx (2013) xxx– xxx 3

ess

3

3

ASetet

Y

Y

sFfad

Table 1Optimal parameters for VFAs production from ABS and WAS co-digestion under CPCtreatments.

AP-CPC ATP-CPC UP-CPC

Temperature (◦C) – 81.0 –Alkaline dosage (g/g TSS) 0.075 0.105 –ABS dosage (g/g VSS) 0.90 0.82 0.57Time (min) – 20.0 30.0Energy density (kW/L) – – 0.98

Fa

ARTICLERBI-9996; No. of Pages 7

A. Zhou et al. / Process Bio

The software Matlab 2012a (MathWorks Inc., USA) wasmployed for handling EEM data. The Design Expert software (Ver-ion 7.4.1.0, Stat-Ease Inc., Minneapolis, USA) was used for thetatistical and stepwise regression analyzes of the data obtained.

. Results

.1. Optimization of the CPC parameters for VFAs production

The results for AP-CPC and the design of experiments forTP-CPC and UP-CPC treatments are shown in Table S3, S4 and5, respectively. By applying multiple regression analysis on thexperimental data obtained from ATP-CPC and UP-CPC processes,he corresponding two factor interactive second-order polynomialquations, relating the coefficients obtained for VFAs production tohe experimental variable, were as follows:

VFAS(ATP-CPC) = −2673.7 + 103.3X1 + 45701.4X2 + 1150.5X3

− 928.8X1X2 − 74.9X1X3 + 95014X2X3 − 925.0X2X4 (1)

VFAS(UP-CPC) = 4559.7 + 70.1B + 343.4C − 111.1AB + 1101.0BC

(2)

The analysis of variance (ANOVA) was conducted to test theignificance of the fit of the second-order polynomial equations.

Please cite this article in press as: Zhou A, et al. VFAs bioproduction from wasubstrates conditioning. Process Biochem (2013), http://dx.doi.org/10.1016

-values of 5.34 and 24.28 and p-value (Prob > F) of 0.039 and 0.002or Eqs. (1) and (2), respectively, implied a significant model fit. Inddition, the models did not show lack of fit and presented highetermination coefficients (R2 0.934 and 0.967), explaining 93.4%

ig. 1. Time-course profiles of soluble carbohydrate (A), soluble protein (B) and VFAs concnalysis when maximum VFAs yields reached (D) (Note: error bars represent standard de

Predicted maximum VFAs yield(mg COD/L)

3901 5118 4064

and 96.7% of the variability in the responses. Adequate precision(7.56 and 13.94) measured the signal-to-noise ratio, and a ratiogreater than 4 was generally desirable. Optimal conditions and thepredicted VFAs production for the AP-CPC, ATP-CPC and UP-CPCprocesses are summarized in Table 1.

3.2. Changes in organic matter content during WAS and ABSco-digestion

WAS was made up of various complex organic compounds. Ascan be seen in Table S1, carbohydrates and proteins were the mainconstituents of sludge organic matter, which accounted for 12%and 67% of TCOD, respectively. The trend of soluble carbohydrateand protein concentration in CSTR are shown in Fig. 1A and B.A large amount of carbohydrates and proteins were observed in

ste activated sludge by coupling pretreatments with Agaricus bisporus/j.procbio.2013.11.005

WAS solution after AP, ATP and UP pretreatments. During the batchoperation (Phase 1), these soluble organic compounds sharplydeclined, being transformed into VFAs. From day 11 onward, theaverage consumption of carbohydrates was 399, 601 and 282 mg

entration (C) under CPC treatments during entire digestion time; VFAs compositionviation).

Page 4

ARTICLE ING Model

PRBI-9996; No. of Pages 7

4 A. Zhou et al. / Process Biochem

Fe

Ct2tVgassP5Cobevpds(f(tFtaA

CC21T3cow

3

(t

ig. 2. Hydrolytic enzyme activities when maximum VFAs yields reached (Note:rror bars represent standard deviation).

OD/L under AP-CPC, ATP-CPC and UP-CPC treatments, respec-ively, while the corresponding proteins consumption was 1883,150 and 801 mg COD/L. Consumption of carbohydrates and pro-eins in the control tests was only 8 and 121 mg COD/L respectively.FAs concentrations rapidly increased from the begging and thenradually decreased with time extension (Fig. 1C), suggesting that

methanation process occurred in each reactor. Furthermore, theelf-degradation of mixed substrates in Phase 1 enabled a stabletart-up of the reactor for the subsequence semi-continuous run inhase 2, without inoculum [19]. VFAs yields reached to 4103 ± 111,602 ± 73 and 3728 ± 16 mg COD/L under AP-CPC, ATP-CPC and UP-PC treatments, respectively, within the same fermentation timef optimization experiment (3 d). An excellent correlation existedetween predicted (Table 1) and measured VFAs yield, standardrrors were ±5%, 9% and 8% under the three-CPC treatments, thusalidating the model and confirmed the existence of an optimaloint. Concerning the continuous operation (Phase 2), VFAs pro-uction, soluble carbohydrates and proteins first increased, thentabilized, and finally decreased (Fig. 1A–C). Maximum VFAs yieldcorresponding fermentation time) showed a 1.81-fold (21 d), 2.26-old (21 d) and 1.73-fold (21 d) increase over that in the control3062 ± 27 mg COD/L, 17 d) under AP-CPC, ATP-CPC and UP-CPCreatments, respectively. The VFA composition analysis is shown inig. 1D. With the CPC treatment, all the considered VFAs increased,hough to a different extent. HAc had a relatively large increase,ccounting for 43 ± 1.7%, 46 ± 0.9% and 40 ± 3.2% of total VFAs withP-CPC, ATP-CPC and UP-CPC treatments, respectively.

The removal of organic solids is schematically shown in Fig. S2.PC treated samples showed higher removal efficiency. Taking ATP-PC as example, the TSS and VSS removed were about 17 ± 3% and5 ± 1% during phase 1, while that of control was reduced by about1 ± 2% and 14 ± 1%, respectively. During the whole 28 d digestion,SS and VSS reduction for ATP-CPC treated sludge was 33 ± 4% and5 ± 2%, respectively, compared to the 26 ± 4% and 27 ± 2% of theontrol. Although external ABS additive conditioning added morerganic solids into the WAS fermentation, the reduction efficienciesere still higher than in the control.

.3. Hydrolytic enzymes activation

Please cite this article in press as: Zhou A, et al. VFAs bioproduction from wasubstrates conditioning. Process Biochem (2013), http://dx.doi.org/10.1016

Fig. 2 shows the activities of two types of hydrolytic enzymesprotease and �-glucosidase) under different CPC treatments at aime of 21 days. Since proteins and carbohydrates represented the

PRESSistry xxx (2013) xxx– xxx

major organic fraction of WAS and ABS (Table S1, S2), the activitiesof �-glucosidase and protease were quite important in the hydroly-sis process. The results showed that the two enzymes were clearlyactive after CPC treatments. Taking ATP-CPC treatment as exam-ple, the activity of protease and �-glucosidase were 95 ± 20 and115 ± 16 Eu, which enabled a 6.3-fold and 0.9-fold increase overthat in the control, respectively.

3.4. FTIR spectroscopic analysis

The presence and changes of main adsorption functional groupswas analyzed by FTIR spectra (Fig. 3). A number of absorption peakswas revealed in the spectra of the control and CPC samples, indi-cating the complex nature of WAS and ABS. The interpretationof the main bands was based on literature [20,21]. The spectrumshowed a very broad and strong absorption band at 3300 cm−1,which could be assigned to the OH in the carboxyl group andthe stretching band of N H, while bands in the region between2880 and 3000 cm−1 being due to aliphatic C H stretching. Twopartially overlapping bands at 1720 and 1658 cm−1 were observedin all cases. The former was assigned to C O of esters and acids,and the latter to N H bending band of amide-I. Peaks at 1550 andat 1230 cm−1 could also be due to the NH2 or N H bending andC N stretching bands of amide-II and amide-III peaks, respectively.Effective intensity enhancement for amide groups after digestionsuggested the degradation of hydrolyzed proteins. A strong bandwas also observed at 1064 cm−1, which could have been caused bythe stretching vibrations of the C O and C O C groups of carbohy-drates, aromatic ethers. Evident intensity reduction at 1064 cm−1

was shown after digestion, which was indicative of the degradationof carboxylic compounds.

3.5. Characterization of co-digesting WAS and ABS by EEMspectra with PARAFAC

Three-dimensional EEM spectroscopy was applied for charac-terizing the dissolved organic matter (DOM) of the co-digestingWAS and ABS (Fig. 4). Four DOM-EEM fluorescence spectra weresimilar in the peak locations, but had different fluorescence inten-sities (FI). By separating the spectra of the main components fromoverlapped EEM spectra with PARAFAC approach, three compo-nents were found to be appropriate, which were identified ashumic-acid-like substances (Com.1, Ex/Em 340/430), protein-likesubstances (Com.2, Ex/Em 280/340) and fulvic-acid-like substances(Com.3, Ex/Em 260/450) [18,22,23] (Fig. 5). The excitation andemission spectra (loadings) of the three organic fluorphores over-lapped slightly (Fig. S3). Additionally, excitation spectra had one ormore maxima, but emission spectra exhibited only one emissionmaximum, which fitted well with the spectral properties expectedfor fluorophores [24].

Protein-like substances were the primary components of theco-digesting solution, with a FI of 773, 960 and 476 under AP-CPC, ATP-CPC and UP-CPC treatments, respectively, compared toonly 197 in the control (Fig. S4). Results were in accordancewith practical measurements (Fig. 1B). Humic-acid-like and fulvic-acid-like substances, mostly provided by external ABS addition,represented the second and third most abundant compounds inall samples.

4. Discussion

WAS hydrolysis was strengthened by CPC treatment. This was

ste activated sludge by coupling pretreatments with Agaricus bisporus/j.procbio.2013.11.005

shown by the increased soluble organics (Fig. 1A and B) andthe active hydrolytic enzymes (Fig. 2). A previous study reportedthat a variable fraction (17% of l-Leu-aminopeptidase, 5% of �-glucosidase, 23% of protease and 44% of �-amylase) of enzyme

Page 5

ARTICLE IN PRESSG Model

PRBI-9996; No. of Pages 7

A. Zhou et al. / Process Biochemistry xxx (2013) xxx– xxx 5

tock: C

apstlutata

6CwaC

Fig. 3. FTIR spectra of the raw and digested feeds

ctivities were associated to the easily extractable extracellularolymeric substances (EPS) from sludge flocks [25]. Pretreatmentteps can disintegrate the EPS matrix and release extracellular pro-eins, polysaccharides and enzymes in different sludge layers toiquid phase [26], which further contributed to the increase of sol-ble organic concentration and enzyme activity. As also observed inhe EEM analysis, the FI of CPC treated protein-like, humic-acid-likend fulvic-acid-like substances were much higher than in the con-rol. This step would provide more hydrolysates for subsequencecidogenesis.

VFAs yields obtained with co-digesting WAS and ABS were14 ± 71, 712 ± 49 and 598 ± 19 mg COD/g VSS under AP-CPC, ATP-

Please cite this article in press as: Zhou A, et al. VFAs bioproduction from wasubstrates conditioning. Process Biochem (2013), http://dx.doi.org/10.1016

PC and UP-CPC treatments, respectively (Table S6). Comparedith the overall VFAs production by individually pretreated WAS

nd ABS fermentation (Fig. S5), those obtained by AP, ATP and UPPC treatments increased by 19%, 33% and 21%, respectively (Table

Fig. 4. EEM fluorescence spectra of DOM obtained from the co-digesting ABS and W

ontrol and AP-CPC (A); ATP-CPC and UP-CPC (B).

S6). This was probably due to a synergistic effect that occurredwhen co-digesting WAS and ABS together, because the additionof carbohydrate balanced the C/N ratio of WAS anaerobic fermen-tation system. A comparison of different VFAs yield from WASfermentation (by pretreatment and/or co-digesting carbon-richsubstrates) is given in Table 2 [27]. In the present study, VFAs yieldobtained under CPC treatments was higher than those obtainedby pretreatment or co-digesting carbon-rich substrates from WAS.Based on the analysis of VFAs composition, HAc was the dominantshort-chain fatty acids under CPC treatment, which is beneficialfor subsequent bioprocesses, such as biological nutrient removal[28], biogas and biopolymer production [29,30] (Fig. 1D). This result

ste activated sludge by coupling pretreatments with Agaricus bisporus/j.procbio.2013.11.005

was consistent with individual WAS digestion [13,31] and otherco-digesting studies, such as WAS + rice straw [6,32], WAS + cornstover [7], mixed sludge + bagasse [5] and WAS + perennial ryegrass[8].

AS under AP, ATP and UP pretreatments when maximum VFAs yields reached.

Page 6

ARTICLE IN PRESSG Model

PRBI-9996; No. of Pages 7

6 A. Zhou et al. / Process Biochemistry xxx (2013) xxx– xxx

F ) hum( d from

dituobattiscrtmtspp

TC

ig. 5. Three components of DOM decomposed by the PARAFAC approach: (Com.1280/340); (Com.3) fulvic-acid-like substances, Ex/Em (260/450). DOM was obtaine

The degradation of proteins in WAS was stimulated by the con-itioning step of CPC treatment. This was proved by an increase

n ammonia, a by-product of protein degradation, during the reac-ion. Much higher concentration of soluble ammonia was achievednder CPC treatment, which was in good agreement with previ-us findings [6] (Fig. S6). FTIR results also showed that adsorptionands for amide groups were enhanced after digestion. A certainmount of phosphate was also released in the liquid with CPCreatments. However, the observed ammonia release was higherhan phosphorus in all cases. The same results were observed dur-ng primary sludge [33] and WAS fermentation [34]. It has beenuggested that, before VFAs-rich fermentation supernatant is recir-ulated back to be used as carbon source for the main-stream, theeleased phosphorus and ammonia should be removed. Otherwise,his would increase the load of phosphorus and ammonia to the

ain-stream [35,36]. Previous researches investigated that simul-

Please cite this article in press as: Zhou A, et al. VFAs bioproduction from wasubstrates conditioning. Process Biochem (2013), http://dx.doi.org/10.1016

aneous recover of phosphorus and ammonia in the form of struvitehowed a distinct advantage over other methods because therecipitate can be used as fertilizer [35]. It was reported that phos-horus concentration was the limiting factor for struvite recovery,

able 2omparison of VFAs yield from WAS fermentation by pretreatment and/or co-digesting c

Pretreatments Carbon-rich substrates

Thermo- or ultrasonic-alkaline –

Adjusting pH (8) Thermal-treated rice straw

– Perennial ryegrass

– Lime-treated bagasse

Alkaline-thermal ABS

ic-acid-like substances, Ex/Em (340/430); (Com.2) protein-like substances, Ex/Em the co-digesting ABS and WAS under AP-CPC, ATP-CPC and UP-CPC treatments.

the minimal phosphorus concentration via crystallizer was 60 mg/L[37]. In the present study, maximum phosphorus concentrationwas 147 ± 30, 182 ± 33 and 172 ± 7 mg/L under AP, ATP and UPCPC treatments, respectively. This suggested that the fermentationsupernatant obtained from CPC treatments would be a good candi-date for struvite recovery when the mixing of the fermenting sludgeis implemented.

It is well known that pretreatment costs represent the majorbottleneck for WAS treatment. Alkaline, thermal and ultra-sonic pretreatments were commonly used for WAS solubilization[12]. However, additional costs were still entailed, includingcost of chemicals, steam, electricity and corrosion-resistant pre-treatment reactors. The results of this study demonstratedthe feasibility of CPC treatment for the conversation of VFAsfrom WAS fermentation. The introduction of such a condition-ing step in the WAS treatment system could help reducing

ste activated sludge by coupling pretreatments with Agaricus bisporus/j.procbio.2013.11.005

the operation costs of a primary pretreatment step, due tothe better performance obtained. This would also foster theintroduction of a new recycling technology for lignocellulosicbiomass.

arbon-rich substrates.

VFAs yield References

224 mg VFAs/g VS [27]520.1 mg COD/g VSS [6]369 ± 18 mg COD/g TS [8]360 mg carboxylic acid/g VS [5]712 ± 49 mg COD/g VSS 489 ± 34 mg COD/g TSS This study

Page 7

ING Model

P

chem

5

tdsama6atosbfric

A

tbNRCoToPdH(

A

fp

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[phorus removal. Chemosphere 2010;78:29–34.

ARTICLERBI-9996; No. of Pages 7

A. Zhou et al. / Process Bio

. Conclusions

VFAs production from WAS was successfully improved by CPCreatment, which gave higher concentrations than those pro-uced by sole pretreatment or co-digestion with carbon-richubstrates. A synergistic effect occurred when co-digesting WASnd ABS. WAS hydrolysis was strengthened further by CPC treat-ent with dominant production of protein-like, humic-acid-like

nd fulvic-acid-like substances. The maximal VFAs yield was14 ± 71, 712 ± 49 and 598 ± 19 mg COD/g VSS at pre-optimizedlkaline, alkaline-thermal and ultrasonic CPC treatments, respec-ively, with an increase of 35%, 50% and 38% compared to the yieldsf pretreated WAS fermentation. Protein degradation was alsotimulated by the conditioning of CPC treatments, as demonstratedy the activity of protease and �-glucosidase, which enabled a 6.3-old and 0.9-fold increase over that in the control, respectively. Theesults indicated that the introduction of such a conditioning stepn the WAS treatment system could help reducing the operationosts of a primary pretreatment step.

cknowledgements

This research was supported by National Science Founda-ion for Distinguished Young Scholars (Grant No. 51225802),y National Natural Science Foundation of China (NSFC, Granto. 51111140388, No. 51208496), by Science Fund for Creativeesearch Groups of the National Natural Science Foundation ofhina (Grant No. 51121062), by National High-tech R&D Programf China (863 Program, Grant No. 2009AA062906), by National Keyechnology Research and Development Program of the Ministryf Science and Technology of China (2010BAC67B02), by Chinaostdoctoral Science Foundation (No. 2012M510574), by “Hun-red Talents Program” of the Chinese Academy of Sciences, and byeilongjiang Science Foundation for Distinguished Young Scholars

Grant No. JC201003).

ppendix A. Supplementary data

Supplementary data associated with this article can beound, in the online version, at http://dx.doi.org/10.1016/j.rocbio.2013.11.005.

eferences

[1] Morgan-Sagastume F, Pratt S, Karlsson A, Cirne D, Lant P, Werker A. Productionof volatile fatty acids by fermentation of waste activated sludge pre-treated infull-scale thermal hydrolysis plants. Bioresour Technol 2011;102:3089–97.

[2] Yuan H, Chen Y, Zhang H, Jiang S, Zhou Q, Gu G. Improved bioproduction ofshort-chain fatty acids (SCFAs) from excess sludge under alkaline conditions.Environ Sci Technol 2006;40:2025–9.

[3] Donoso-Bravo A, Fdz-Polanco M. Anaerobic co-digestion of sewage sludgeand grease trap: assessment of enzyme addition. Process Biochem 2013;48:936–40.

[4] Cao G, Ren N, Wang A, Lee D-J, Guo W, Liu B, Feng Y, Zhao Q. Acid hydrol-ysis of corn stover for biohydrogen production using Thermoanaerobacteriumthermosaccharolyticum W16. Int J Hydrogen Energy 2009;34:7182–8.

[5] Rughoonundun H, Mohee R, Holtzapple MT. Influence of carbon-to-nitrogenratio on the mixed-acid fermentation of wastewater sludge and pretreatedbagasse. Bioresour Technol 2012;112:91–7.

[6] Feng L, Chen Y, Zheng X. Enhancement of waste activated sludge protein con-version and volatile fatty acids accumulation during waste activated sludgeanaerobic fermentation by carbohydrate substrate addition: the effect of pH.Environ Sci Technol 2009;43:4373–80.

[7] Zhou A, Guo Z, Yang C, Kong F, Liu W, Wang A. Volatile fatty acids productiv-

Please cite this article in press as: Zhou A, et al. VFAs bioproduction from wasubstrates conditioning. Process Biochem (2013), http://dx.doi.org/10.1016

ity by anaerobic co-digesting waste activated sludge and corn straw: effect offeedstock proportion. J Biotechnol 2013;168:234–9.

[8] Jia S, Dai X, Zhang D, Dai L, Wang R, Zhao J. Improved bioproduction of short-chain fatty acids from waste activated sludge by perennial ryegrass addition.Water Res 2013;47:4576–84.

[

PRESSistry xxx (2013) xxx– xxx 7

[9] USDA. Mushrooms National Agricultural Statistics Service. Washington, DC:Agricultural Statistics Board; 2011. p. 1–13.

10] Finney KN, Ryu C, Sharifi VN, Swithenbank J. The reuse of spent mushroom com-post and coal tailings for energy recovery: comparison of thermal treatmenttechnologies. Bioresour Technol 2009;100:310–5.

11] Kapu NU, Manning M, Hurley TB, Voigt J, Cosgrove DJ, Romaine CP. Surfactant-assisted pretreatment and enzymatic hydrolysis of spent mushroom compostfor the production of sugars. Bioresour Technol 2012;114:399–405.

12] Appels L, Baeyens J, Degrève J, Dewil R. Principles and potential of the anaerobicdigestion of waste-activated sludge. Prog Energ Combust 2008;34:755–81.

13] Zhou A, Yang C, Kong F, Liu D, Chen Z, Ren N, Wang A. Improving the short-chainfatty acids production of waste activated sludge stimulated by a bi-frequencyultrasonic pretreatment. J Environ Biol 2013;34:381–9.

14] Zhou A, Yang C, Guo Z, Hou Y, Liu W, Wang A. Volatile fatty acids accumulationand rhamnolipid generation in situ from waste activated sludge fermentationstimulated by external rhamnolipid addition. Biochem Eng J 2013;77:240–5.

15] APHA. Standard Methods for the Examination of Water and Wastewater, Wash-ington, DC, USA. 20th ed; 1998.

16] Liu W, Huang S, Zhou A, Zhou G, Ren N, Wang A, Zhuang G. Hydrogen gener-ation in microbial electrolysis cell feeding with fermentation liquid of wasteactivated sludge. Int J Hydrogen Energy 2012;37:13859–64.

17] Miron Y, Zeeman G, van Lier JB, Lettinga G. The role of sludge retention timein the hydrolysis and acidification of lipids, carbohydrates and proteins duringdigestion of primary sludge in CSTR systems. Water Res 2000;34:1705–13.

18] Lu L, Xing D, Liu B, Ren N. Enhanced hydrogen production from waste activatedsludge by cascade utilization of organic matter in microbial electrolysis cells.Water Res 2011;46:1015–26.

19] Kobayashi T, Yasuda D, Li Y-Y, Kubota K, Harada H, Yu H-Q. Characterizationof start-up performance and archaeal community shifts during anaerobic self-degradation of waste-activated sludge. Bioresour Technol 2009;100:4981–8.

20] Socrates G. Infrared Characteristic Group Frequencies. 2nd ed. England, UK:John Wiley & Sons; 1994.

21] MacCarthy P, Rice JA. Spectroscopic methods (other than NMR) for determiningfunctionality in humic substances. New York, USA: John Wiley; 1985.

22] Chen W, Westerhoff P, Leenheer JA, Booksh K. Fluorescence excitation-emissionmatrix regional integration to quantify spectra for dissolved organic matter.Environ Sci Technol 2003;37:5701–10.

23] Wang Z, Wu Z, Tang S. Characterization of dissolved organic matter in asubmerged membrane bioreactor by using three-dimensional excitation andemission matrix fluorescence spectroscopy. Water Res 2009;43:1533–40.

24] Stedmon CA, Bro R. Characterizing dissolved organic matter fluorescence withparallel factor analysis: a tutorial. Limnol Oceanogr: Methods 2008;6:572–9.

25] Cadoret A, Conrad A, Block J-C. Availability of low and high molecular weightsubstrates to extracellular enzymes in whole and dispersed activated sludges.Enzyme Microb Technol 2002;31:179–86.

26] Yu G-H, He P-J, Shao L-M, Zhu Y-S. Extracellular proteins, polysaccharides andenzymes impact on sludge aerobic digestion after ultrasonic pretreatment.Water Res 2008;42:1925–34.

27] Liu X, Liu H, Chen J, Du G, Chen J. Enhancement of solubilization and acidificationof waste activated sludge by pretreatment. Waste Manag 2008;28:2614–22.

28] Li X, Chen H, Hu L, Yu L, Chen Y, Gu G. Pilot-scale waste activated sludgealkaline fermentation, fermentation liquid separation, and application of fer-mentation liquid to improve biological nutrient removal. Environ Sci Technol2011;45:1834–9.

29] Zhang D, Chen Y, Zhao Y, Zhu X. New sludge pretreatment method to improvemethane production in waste activated sludge digestion. Environ Sci Technol2010;44:4802–8.

30] Cai M, Chua H, Zhao Q, Sin N, Ren J. Optimal production of polyhydrox-yalkanoates (PHA) in activated sludge fed by volatile fatty acids (VFAs)generated from alkaline excess sludge fermentation. Bioresour Technol2009;100:1399–405.

31] Bolzonella D, Cavinato C, Fatone F, Pavan P, Cecchi F. High rate mesophilic,thermophilic, and temperature phased anaerobic digestion of waste activatedsludge: a pilot scale study. Waste Manag 2012;32:1196–201.

32] Kim M, Yang YN, Morikawa-Sakura MS, Wang QH, Lee MV, Lee DY, Feng CP, ZhouYL, Zhang ZY. Hydrogen production by anaerobic co-digestion of rice straw andsewage sludge. Int J Hydrogen Energy 2012;37:3142–9.

33] Banister SS, Pitman AR, Pretorius WA. The solubilisation of N and P duringprimary sludge fermentation and precipitation of the resultant P. Water SA1998;24:337–42.

34] Chen Y, Jiang S, Yuan H, Zhou Q, Gu G. Hydrolysis and acidification of wasteactivated sludge at different pHs. Water Res 2007;41:683–9.

35] Zhang C, Chen Y. Simultaneous nitrogen and phosphorus recovery from sludge-fermentation liquid mixture and application of the fermentation liquid toenhance municipal wastewater biological nutrient removal. Environ Sci Tech-nol 2009;43:6164–70.

36] Yuan Q, Oleszkiewicz JA. Biomass fermentation to augment biological phos-

ste activated sludge by coupling pretreatments with Agaricus bisporus/j.procbio.2013.11.005

37] Benisch M, Baur R, Britton A, Oleszkiewicz JA, Neethling JB. Utilizing phospho-rus recovery for optimization of the biological nutrient removal process. In:Procedure International Water Association 2nd Special Conference NutrientManagement in Wastewater Treatment Processes. 2009. p. 1005–16.