Methane production from food waste via mesophilic anaerobic digestion with ethanol pre-fermentation: Methanogenic pathway and microbial community analyses

Journal Pre-proofs Methane production from food waste via mesophilic anaerobic digestion with ethanol pre-fermentation: Methanogenic pathway and microbial community analyses Hui Zou, Ming Gao, Miao Yu, Wenyu Zhang, Shuang Zhang, Chuanfu Wu, Yukihiro Tashiro, Qunhui Wang PII: DOI: Reference:

S0960-8524(19)31680-3 https://doi.org/10.1016/j.biortech.2019.122450 BITE 122450

To appear in:

Bioresource Technology

Received Date: Revised Date: Accepted Date:

6 October 2019 17 November 2019 18 November 2019

Please cite this article as: Zou, H., Gao, M., Yu, M., Zhang, W., Zhang, S., Wu, C., Tashiro, Y., Wang, Q., Methane production from food waste via mesophilic anaerobic digestion with ethanol pre-fermentation: Methanogenic pathway and microbial community analyses, Bioresource Technology (2019), doi: https://doi.org/10.1016/ j.biortech.2019.122450

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2019 Published by Elsevier Ltd.

1

Methane production from food waste via mesophilic anaerobic

2

digestion with ethanol pre-fermentation: Methanogenic pathway and

3

microbial community analyses

4 5

Hui Zou a, c #, Ming Gao a, b #, Miao Yu e, Wenyu Zhang c, Shuang Zhanga, Chuanfu Wu a,

6

b, Yukihiro

Tashiro d, Qunhui Wang a, b, *

7 8

a

9

Engineering, University of Science and Technology Beijing, Beijing 100083, P. R.

Department of Environmental Engineering, School of Energy and Environmental

10

China

11

b

12

Pollutants, University of Science and Technology Beijing, Beijing 100083, P. R. China

13

c

14

Institute of Environmental Protection, Beijing 100037, P. R. China

15

d Department

16

Kyushu University, Fukuoka 812-8581, Japan

17

e China

18

※ Hui

19

*

20

of Science and Technology Beijing, 30 Xueyuan Road, Haidian District, Beijing

21

100083, China. Tel/Fax: +86 (0)10- 6233 2778. E-mail address: [email protected]

22

(Q. Wang)

Beijing Key Laboratory on Disposal and Resource Recovery of Industry Typical

Institute of Soil Environment and Pollution Remediation, Beijing Municipal Research

of Bioscience and Biotechnology, Faculty of Agriculture, Graduate School,

Enfi Engineering Corporation, Beijing 100038, P. R. China

Zou and Ming Gao contributed equally to this work

Corresponding author. School of Energy and Environmental Engineering, University

23

1

1

Abstract: To investigate the methanogenic pathway and microbial community in a

2

mesophilic anaerobic digestion (AD) system with food waste (FW) ethanol pre-

3

fermentation (EP), two semi-continuous AD systems were operated by feeding FW with

4

(PSR) and without EP (control). In this study, δ13C-ethanol was supplemented as solo

5

substrate for AD sludge when the reactors operation stabilized to analyze the

6

methanogenic pathways. The results suggested that approximately 59.3% of methane

7

was produced from acetotrophic methanogens, while 40.7% was formed by

8

hydrogenotrophic methanogens in the PSR group. On the other hand, compared with

9

control, methane produced via CO2 reduction pathway was increased by 4.70%.

10

Meanwhile, the composition variations of the microbial community in AD supported

11

the

12

Methanobacterium were enhanced by 7.6% and 10.2%, respectively in PSR reactor.

13

These results provided a theoretical basis for AD applications and biogas yield

14

improvements with EP process.

above

conclusion,

since

the

relative

abundances

of

Clostridium

and

15 16

Keywords: Food waste; Stable isotope; High-throughput sequencing; Methanogenic

17

pathway; Microbial community

2

1

1. Introduction

2

According to the Ministry of Housing and Urban-Rural Development of China

3

(2012), more than 51 million tonnes of food waste (FW) is produced annually in the

4

country, and most of it is sent directly to landfills. Owing to its high moisture content,

5

and rich organic matter content, FW is considered as a potentially valuable material, and

6

it has been referred to as a ‘misplaced’ biomass resource (Ma et al., 2011; Kim et al.,

7

2010; Sun et al., 2016). Conversion of FW to fuel could potentially help alleviate

8

energy shortages in China (Ren et al., 2018). However, the low C/N ratio and high

9

biodegradability of food waste lead to rapid acidification in the process of anaerobic

10

digestion (AD) (Conrad, 2005; Polag et al., 2015; Zhao et al., 2016). In recent years,

11

biological pre-treatments—as a new part of the AD pre-treatment process—have

12

become a popular research topic. These pre-treatments involve the inoculation of

13

microorganisms and production of enzymes that can promote the hydrolysis of

14

substrates, alleviate acidification, and increase the anaerobic digestion rate (Ren et al.,

15

2018).

16

Ethanol pre-fermentation (EP) in FW via AD as a biological pre-treatment has

17

gained considerable attention. Zhao et al. (2016) studied the effect of EP on methane

18

yield. They indicated that, compared with the control (without EP), the concentrations

19

of volatile fatty acids, propionate, and acetate in the EP group were lower, with no

20

acidification of the system. At the same time, methane yields were 49.6% higher than

21

those in the control group. In addition, the research also showed that inoculation of yeast

22

(Saccharomyces cerevisiae) inhibited the presence of pathogens, including Escherichia

23

coli and Staphylococcus aureus. Wu et al. (2015) analysed the influence of EP with

24

different inoculum-to-substrate ratios (ISRs) before AD on FW and distillers’ grains.

3

1

Compared with the control group, EP effectively alleviated acidification, greatly

2

reduced the lag period, and significantly stimulated the growth of methanogens. Yu et al.

3

(2018) evaluated the effects of EP on fermentation and found that more of the carbon

4

source was converted to ethanol in the EP group than in the control. Furthermore, the

5

dominant genera of methanogenic bacteria—including Methanobacterium and

6

Methanosarcina—were promoted. Yet, the above results cannot explain how the

7

ethanol produced in the EP stage metabolizes to methane in the AD process.

8

Based on previous research findings (Karakashev et al., 2006; Sasaki et al., 2011),

9

stable isotopes (δ13C) offer a new method for studying the methanogenic pathway in

10

AD (Conrad, 1999; Qu et al., 2009). Zou et al. (2019) studied the metabolic changes

11

from AD after EP by using carbon isotope labelling. The EP process was found to

12

alleviate acidification, while channelling an extremely high carbon flux towards ethanol

13

formation (43.7%). An efficient acetogenesis phase was also observed in the EP group,

14

because of the high carbon conversion rate from ethanol to acetate. However, the effect

15

of EP in the methanogenesis stage was not evident. In addition, from the results of

16

microbial community analyses of AD after EP, Zhao et al. (2016) clarified that EP

17

could inhibit the presence of pathogens, and Yu et al. (2018) showed that in the EP

18

group, Methanobacterium became the dominant species as fermentation proceeded.

19

However, to our knowledge, there is no report on the methanogenic pathway in AD

20

after EP, especially in combination with the effects on microbial communities.

21

Ethanol produced in EP stage would be metabolized to acetate during AD. Then

22

formed acetate could be consumed as the main precursor for methane generation during

23

AD stage. Therefore, to investigate the underlying mechanisms of AD with EP, the

24

carboxyl- (CH313CH2-OH) and methyl- (13CH3CH2-OH) labelled ethanol was

4

1

supplemented as solo substrate respectively to evaluate the methanogenic pathways in

2

batch fermentation mode.

3

In the present study, a laboratory-scale mesophilic semi-continuous AD reactor was

4

operated with FW after EP as the substrate. Then, δ13C-ethanol was added to the

5

fermentation liquid obtained from the stable AD reactor, and labelled substances were

6

calculated

7

Furthermore, bacterial communities present in the stable AD reactor were analysed by

8

Illumina MiSeq high-throughput sequencing. This study demonstrates the methanogenic

9

pathway of AD after EP, providing a reference for further studies and potential

10

using gas chromatography-mass spectrometry

(GC-MS)

analysis.

engineering applications.

11 12

2. Materials and methods

13

2.1. Feedstock and inocula

14

The FW was taken from the students’ dining hall in the University of Science and

15

Technology of Beijing. After the non-biodegradable contaminants were removed, the

16

FW was homogenised with a macerating grinder (S52/010 Waste Disposer, IMC Ltd.,

17

UK), packed into plastic storage bags, and frozen at −20 °C. The AD sludge (ADS) was

18

acclimatised for 1 month, with continuous addition of FW before the experiments. The

19

chemical characteristics of FW and ADS are shown in Table 1.

20 21

2.2. Ethanol pre-fermentation methods

22

Alcohol active dry yeast (Angel Yeast, Hubei, China) was used as the inoculum for

23

EP. To recover the activity, the dry yeast powder was inoculated into 2% (M/V) sucrose

24

liquid with a mass ratio of 2.5% (W/V), and cultivated anaerobically in a rotary shaker

5

1

at 35 °C for 2 h with 150 rpm. The activated yeast broth was used as the inoculum for

2

EP process with a mass inoculation ratio of 2.5% (W/W, dry basis) in 5000-mL jar

3

fermenter containing food waste medium (FW). The EP process was conducted

4

anaerobically at 35 °C for 24 h, with stirring at 150 rpm in the rotary shaker. The yeast

5

was inoculated only once at the initial of EP. Then, the EP broth was used as the feeding

6

substrate in the following semi-continuous AD experiments.

7 8

2.3. Semi-continuous experimental methods

9

Semi-continuous experiments were performed with two of the same reactor types.

10

The reactor was comprised of a 2500-mL fermentation jar with 2000 mL working

11

volume, one 2000-mL biogas collection bag, and an electric constant temperature air

12

bath rotary shaker. In the first reactor, AD was fed with FW after EP for 24 h. The feed

13

loading rate was at 1.0 g of VSsubstrate/g of VSinoculum every two days. Samples were

14

collected after 2 h, and the experiment group was designated as the PSR group. The

15

second reactor was fed with FW without EP, under the same conditions; this was

16

designated as the control group. The AD was conducted at 35 °C with agitation at 80

17

rpm. The biogas produced from the inoculum was excluded.

18 19

2.4. Carbon isotope experimental methods

20

To evaluate the methanogenic pathway in AD after EP, 10 mL of broth was taken

21

from the above stable reactors and transferred to 25 mL vials. The vials were sealed

22

with a butyl rubber stopper and an aluminium cap; these were then flushed with high

23

purity nitrogen by using 0.45 µm filter membranes. The vials were pre-incubated for 5

24

days at 35 °C at a shaking speed of 100 rpm to equalise the degradation activity. After

6

1

pre-cultivation, the vials were flushed with nitrogen again; after this, 100 mM of 2-13C-

2

ethanol or 1-13C-ethanol (99 atom %; Cambridge Isotope Laboratories, USA) was added

3

to the pre-cultured vials. After 24 h of incubation with shaking, the biogases produced

4

were analysed via the selected ion monitoring method using gas chromatography-mass

5

spectrometry (GC-MS). Each set of experiments had three replicates. The other tests

6

without yeast, in which the same amount of 13C-ethanol was added during AD, served

7

as the control.

8 9

2.5. Analysis of the reactor performance

10

A standard method (Standard Methods for the Examination of Water and

11

Wastewater, 1998) was used to measure the quantity of volatile solids (VS) and total

12

solids (TS). Protein content was measured by using an automatic Kieldahl apparatus

13

(KDN-2C, Xinrui Instruments & Meters Co., Ltd., Shanghai, China). The total sugar

14

concentration of FW was assayed by using the method of Miller (1959) after hydrolysis

15

of a FW sample with 25% HCl. The pH of FW was measured by using a pHS-3C type

16

digital acidity instrument. The volatile fatty acids (VFAs) were determined using the

17

colorimetric method (Wang et al., 2009). Alkalinity, chemical oxygen demand (COD),

18

ammonia nitrogen, and total nitrogen were determined using the standard methods of

19

the American Public Health Association (Standard Methods for the Examination of

20

Water and Wastewater, 1998).

21 22

2.6. Analysis of the composition content

23

δ13C-Ethanol, δ13C-acetate, δ13CH4(CH4), and δ13CO2 (CO2) were determined using

24

a GC-MS (GC: 6890 series; MS: 5973 series; Agilent, USA) and quantified with

7

1

standard substances (Dr. Ehrenstorfer, Germany). δ13C-Ethanol and δ13C-acetate were

2

separated on a DB-FFAP column (0.32 mm × 0.25 μm × 30 m; Agilent, California,

3

USA). The oven temperature was programmed to maintain a temperature of 45 °C for 5

4

min, then increase at 10 °C/min to 240 °C, and finally hold at 240 °C for 10 min. The

5

injector and ion source temperatures were set to 220 and 240 °C, respectively. Helium

6

was the carrier gas and set to a flow rate of 1.0 mL/min. δ13CH4 (CH4) and δ13CO2 (CO2)

7

were separated on a HP-PLOT-Q capillary column (0.32 mm × 40 μm × 30 m; Agilent,

8

California, USA), with a split ratio of 25:1. The oven, injector, and ion source

9

temperatures were each set to 35 °C. Argon was used as the carrier gas.

10 11

2.7. Analysis of the microbial community

12

High-throughput sequencing analysis was performed whilst the PSR and control

13

groups were stable, and the types of bacteria and archaebacteria were determined. The

14

bacterial primers were 515F (5′-GTGYCAGCMGCCGCGGTA-3′) and 907R (5′-

15

CCGTCAATTCMTTTRAGTTT-3′). The archaebacterial primers were Arch344F (5′-

16

ACGGGGYGCAGCAGGCGCGA-3′)

17

GTGCTCCCCCGCCAATTCCT-3′). The specific steps used are described in detail in

18

Yu et al. (2018).

and

Arch915F

(5′-

19 20

3. Results and discussion

21

3.1. Reactor operation and performance

22

Two litres of AD sludge were cultivated at 35 °C in two 2.5 L jar fermenters with

23

agitation at 80 rpm. One reactor was fed with FW after EP and designated as the PSR

24

group; the other was fed with FW under the same organic loading rate (OLR) and was

8

1

called the control group. The indexes of biogas, VFAs, and pH were measured; these

2

results are shown in Fig. 1 and Table 2.

3

After AD for 40 days (Fig. 1), the daily amount of biogas was stable at 38.45 and

4

31.85 mL/g-VS for the PSR and control groups, respectively; the pH was also stable at

5

7.52 and 7.49, respectively. The VFAs gradually increased in the early stage and

6

reached a maximum of 12.42 and 16.54 g/L on the 10th day in the PSR and control

7

groups, respectively; then, the values gradually decreased until a condition of

8

exhaustion was reached. Furthermore, Table 2 shows that after 40 days of fermentation,

9

TS, VS, and COD values demonstrated different degrees of reduction. The removal

10

rates of TS, VS and COD in the PSR group were all exceeded over 59%, which were

11

1.25-, 1.47-, and 1.18-folds higher than that of the control, respectively.

12

Table 2 summarizes the average values of reactor performance from 0 to 40 days.

13

The TS, VS, and COD of the two groups were consumed by more than 50%, thus

14

indicating that the two reactors had reached steady state; this is the same as that shown

15

in Fig. 1, and such results have been described previously (Karakashev et al., 2005).

16

The removal rates and biogas yield in the PSR group were higher than those in the

17

control group; this finding is likely related to the feeding substrate (EP of FW).

18

Moreover, based on the research of Wu et al. (2015), EP could increase the hydrolysis

19

rate and accelerate the substrate conversion to biogas. Zhao et al. (2016) also verified

20

that the methane yield with EP was higher than that without EP by 49.6%. In addition,

21

Yu et al. (2018) confirmed that AD after EP resulted in more of the carbon source being

22

converted to ethanol, with Methanobacterium as the predominant species. Zou et al.

23

(2019) also proved there was a higher carbon conversion rate from ethanol to acetate in

9

1

the acetogenesis phase in the EP group, thereby making the reactor more stable and less

2

acidified.

3

The above results indicated that the EP could influence the carbon flow and

4

microbial community in AD, but there have been no more detailed studies on the

5

methanogenic pathway in AD after EP, especially the effects on the microbial

6

community. Hence, based on the results described above, the two reactors were

7

stabilised, and isotope experiments were performed to study the methanogenic pathway

8

and microbial communities.

9 10

3.2. Determination of the methanogenic pathway using δ13C-ethanol

11

Based on the results described in Zou et al. (2019), the carbon flow distributions in

12

AD after EP were calculated using the δ13C6-glucose method; approximately 60% of the

13

carbon was converted to ethanol in the hydrolysis acidification stage. Ethanol is not a

14

precursor for methanogens; therefore, it is a matter of speculation as to how it

15

metabolizes to methane in the methanogenesis stage. Carboxyl- and methyl- labelled

16

ethanol was used to study the methanogenic pathway in the AD process after EP in

17

terms of the acetate metabolism pathway in AD (Karakashev et al., 2006; Shigematsu et

18

al., 2004). The experiments were carried out when the reactors reached a steady state, as

19

mentioned in Section 3.1. The δ13C-labelled products were detected using GC-MS;

20

results from the blank experiment (when labelled ethanol was not added) were excluded.

21

Based on the GC-MS results (Fig. 2), only methyl-labelled acetate was observed

22

following the addition of [1-13C] ethanol (13CH3CH2OH) to the liquid samples, and only

23

carboxyl-labelled acetate was detected following the addition of [2-13C] ethanol

24

(CH313CH2OH) to the liquid. In addition, both δ13CH4 and δ13CO2 were detected in the

10

1

gaseous samples following the addition of [1-13C] ethanol or [2-13C] ethanol. Thus,

2

from the observed results, it could be inferred that there was no homo-acetotrophic

3

pathway in mesophilic AD after EP. The conversion of ethanol to methane mainly

4

involves four reactions: (1) methyl-labelled ethanol is metabolized to only methyl-

5

labelled acetate in the acetotrophic pathway; (2) acetate degradation by acetotrophic

6

methanogens causes methyl-labelled acetate to form only labelled methane; (3) both

7

carbon atoms in acetate are converted to carbon dioxide during syntrophic acetate

8

oxidation; and (4) some of the carbon dioxide is subsequently reduced to methane via

9

the CO2 reduction pathway (Liu and Whitman, 2008).

10

13CH

3CH2OH

+ H2O → 13CH3COO− + H+ + 2H2

11

13CH

3COO

−

+ H2O → 13CH4 + HCO3−

(2)

12

13CH

3COO

−

+ 4H2O → H13CO3− + HCO3− + H+ + 4H2

(3)

13

H13CO3− (or HCO3−) + H+ + 4H2 → 13CH4 (or CH4) + 3H2O

(1)

(4)

14

According to the peak area determined by GC-MS (Table 3), the concentrations of

15

δ13C-ethanol, δ13C-acetate, δ13CH4 (δ12CH4), and δ13CO2 (δ12CO2) were calculated using

16

standard substances. The methanogenic pathway distribution in the AD with EP was

17

also analysed.

18

In the PSR group (AD with EP), the δ13CH4 produced in the methyl-labelled ethanol

19

reaction minus the δ13CH4 produced in the carboxyl-labelled ethanol reaction was the

20

δ13CH4 derived only from the acetotrophic pathway. On the basis of reactions (1) and

21

(2), the acetotrophic pathway accounted for 59.3% of the methanogenic pathway.

22

According to Liu and Whitman (2008), the main methanogenic pathways in AD are the

23

acetotrophic and CO2 reduction pathways. Thus, the CO2 reduction pathway in the PSR

24

group accounted for 40.7%. After 24 h of fermentation following the addition of [2-13C]

11

1

ethanol, the δ13CH4 from the carboxyl-labelled ethanol reaction amounted to 45.6 mM

2

derived from the acetic oxidation pathway (3). On the basis of reaction (3), the acetic

3

oxidation pathway accounted for 34.2%. Results are shown in Fig. 2.

4

Accordingly, in the control group, δ13CH4 produced in methyl-labelled ethanol and

5

carboxyl-labelled ethanol reactions amounted to 79.1 mM and 37.3 mM, respectively.

6

Therefore, the acetotrophic pathway accounted for 66.1% and the CO2 reduction

7

pathway accounted for 33.9%. Meanwhile, the δ13CO2 produced in the carboxyl-

8

labelled ethanol reaction amounted to 39.1 mM, and that of the acetic oxidation

9

pathway accounted for 26.8% in the acetotrophic pathway. Results are also shown in

10

Fig. 2.

11

In general, approximately 70% of biomethane comes from the acetotrophic pathway,

12

with the CO2 reduction pathway accounting for 30% (Hickey et al., 1987; Liu and

13

Whitman, 2008). As shown in Table 3, the acetotrophic pathway in the control group

14

(66.1%) was higher than that in the PSR group (59.3%). In addition, the CO2 reduction

15

pathway decreased by 6.80% in the control. These results indicate that the EP affected

16

the methanogenic pathway during mesophilic AD. Notably, methanogenic pathway

17

determinations through stable isotope analysis have already been reported for a wide

18

range of anaerobic ecosystems. For example, Shigematsu et al. (2004) studied the

19

methanogenic pathway by the addition of δ13C-acetate and found that the acetotrophic

20

pathway was the main pathway at a low dilution rate. Karakashev et al. (2006)

21

supplemented a small portion of δ14C-acetate in AD and confirmed that the

22

methanogenic pathway was preferred to the CO2 reduction pathway, given a high

23

concentration of inhibitors (such as ammonia nitrogen and VFAs). The balance of δ13C-

12

1

labelled carbon was also determined in the two groups; the results, focusing on the

2

δ13C-labelled carbon only, are shown in Table 4.

3

With regard to the results shown in Table 4, the samples of the PSR and control

4

groups were pre-cultured. Then δ13C-ethanol was added and metabolic activities were

5

allowed to proceed. After AD for 24 h, the δ13C-ethanol in the two groups was mostly

6

consumed. The conversion rate of δ13C-ethanol to δ13CH4 and δ13CO2 in the PSR group

7

was approximately 92.3%–94.9%, and that in the control group was 80.4%–90.5%. The

8

labelled carbon recovery in the two groups was > 96%; it did not reach 100% (Table 4),

9

probably because a few of the carbon flow distributions (CFDs) in δ13C-ethanol were

10

converted to other substances (or test errors). However, considering that the flow was

11

relatively small, these could be ignored. The difference in CFDs before and after AD for

12

24 h was not significant; they were all close to 100%, thus indicating that the carbon

13

flow was basically balanced.

14

As mentioned above, although the CO2 reduction pathway increased in the PSR

15

group, the acetotrophic pathway was still the primary metabolism pathway. The

16

increase in the CO2 reduction pathway could be attributed to the supplement of FW with

17

EP, that is, the composition of the substrate leads to the change of the methanogenic

18

pathway. Therefore, it is speculated that EP facilitates more substrate conversion to

19

ethanol, based on the comparison of the Gibbs free energy (ΔG) of ethanol, lactate, and

20

propionate, and the conversion of butyrate to acetate (Refai et al., 2014). The

21

conversion of ethanol to acetate is faster than that of other acids. As a result, a large

22

amount of acetate is produced and is partly metabolized to methane via the acetotrophic

23

pathway, while excess acetate promotes the acetic oxidation pathway, resulting in a

24

large amount of carbon dioxide and hydrogen produced. This further exacerbates the

13

1

CO2 reduction pathway. To verify these findings, it is necessary to study the changes in

2

the microbial community; the results are presented in Section 3.3.

3 4

3.3. Changes in the microbial community in the PSR reactor

5

Based on the results in Section 3.2, the ethanol produced in AD after EP could be

6

metabolized to acetate. Acetate is an important intermediate during biogas production

7

and can be converted into methane through acetoclastic methanogenesis and syntrophic

8

acetate oxidation coupled with hydrogenotrophic methanogenesis. Biogas fermentation

9

is a microbiological process involving bacterial and archaeological microorganisms.

10

Therefore, to analyse the microbial community in AD after EP, the changes in bacterial

11

and archaebacterial communities were investigated. DNA used for the amplification and

12

subsequent sequencing of the 16S rRNA gene was extracted from the two stable semi-

13

continuous reactors (on day 40 of AD), and the samples were designated PSR and

14

control; the results are shown in Fig. 3.

15 16 17

3.3.1. Changes in bacterial communities As shown in Fig. 3(a), on day 40 of methane fermentation, only seven genera of

18

bacteria with concentrations greater than 5% were classified. Compared with the control,

19

the concentration of Clostridium significantly increased by 7.6% in the PSR group,

20

followed by that of Sphaerochaeta, VadinBC27, and Treponema, which increased by

21

2.8%,

22

Christensenellaceae decreased by 2.2% and the changes in the remaining bacteria were

23

in the range of 0.5%.

1.3%,

and

1.1%,

respectively.

14

In

contrast,

the

concentration

of

1

Two mechanisms for methane formation from acetate have been described, that

2

is,the first one is acetoclastic (2) and is carried out by Methanosarcinaceae or

3

Methanosaetaceae. The second mechanism encompasses a two-step reaction in which

4

acetate is first oxidized to H2 and CO2 and subsequently converted, with these products,

5

to methane (3 & 4). This reaction is performed by acetate-oxidizing bacteria (often

6

Clostridium spp.) in a syntrophic association with hydrogenotrophic methanogens

7

(often Methanomicrobiales or Methanobacteriales) (Conrad, 1999; Hattori et al., 2000).

8

The concentration of Clostridium, the syntrophic acetate oxidation bacteria (SAOB),

9

increased greatly in AD after EP. This increase might cause more acetate to be

10

converted into H2 and CO2 and eventually produce methane by the CO2 reduction

11

pathway, enhancing the CO2 reduction pathway in the PSR group. In addition,

12

Sphaerochaeta, VadinBC27, and Treponema, which are related to hydrolytic activity for

13

substrates, also increased to some extent in the PSR group. The increase represented a

14

rise in the hydrolysis ability of the substrate, similar to what has been previously

15

described (Rui et al., 2015). Hence, it can be concluded that EP caused the conversion

16

of more carbon to ethanol—a neutral substance—thus providing a suitable environment

17

for AD that in turn resulted in a greater concentration of hydrolysis bacteria and faster

18

hydrolytic metabolism of the substrate. Meanwhile, more ethanol could be metabolized

19

to acetate quickly, which increased the concentration of SAOB and enhanced the acetic

20

oxidation pathway. This is consistent with the conclusion in Section 3.2.

21

The above-mentioned research indicates that EP can alter the structure of some

22

bacterial communities. Thus, the core microorganisms involved in AD can be increased.

23

Further research is needed to determine the archaebacterial communities.

24

15

1 2

3.3.2. Changes in archaebacterial communities As shown in Fig.3(b), on day 40 of AD, the primary methanogens were

3

Methanobacterium

(hydrogenotrophic

methanogenesis)

and

Methanosarcina

4

(acetoclastic methanogenesis); the relative concentrations of both were greater than

5

65% in the two groups. The relative concentration of Methanosarcina in the PSR group

6

(10.6%) was lower than that in the control group (16.5%), while the relative

7

concentration of the hydrogen-consuming Methanobacterium in the PSR group (58.9%)

8

was much higher than that in the control group (48.7%). Compared with the control, the

9

relative concentrations of Methanobacterium, Methanosaeta, and Methanoculleus in the

10

PSR group increased by 10.2%, 1.4%, and 1.9%, respectively. In addition, the relative

11

concentrations of Methanosarcina and Methanobrevibacter decreased by 5.9% and

12

0.7%, respectively.

13

From the analysis of archaebacterial communities, the Methanobacterium in the

14

PSR group increased by 10.2% over that in the control. The reason was that EP caused a

15

reduction in the hydrogen partial pressure, thereby promoting the conversion of ethanol,

16

propionate, and butyrate to acetate, consistent with the results of previous studies

17

(Hagen et al., 2014; Yu et al., 2018). Then, some of the acetate goes through the

18

acetotrophic pathway, while the excess promotes syntrophic acetate oxidation, leading

19

to a large increase in carbon dioxide and hydrogen, enhancing the CO2 reduction

20

pathway. This is consistent with the conclusion in Section 3.2. Methanobacterium is one

21

of the hydrogenotrophic methanogenesis actors, and is the dominant species, so its

22

relative concentration was increased in the PSR group. In addition, the relative

23

concentration of acetoclastic methanogens was comparatively lower in the PSR group.

24

However, according to the results in Section 3.2, the acetotrophic pathway is still the

16

1

main metabolic pathway, and only Methanosarcina and Methanosaeta can metabolize

2

acetate to methane by the acetotrophic pathway in nature (Angelidaki et al., 2006).

3

Therefore, it is speculated that Methanosarcina is more efficient than hydrogenotrophic

4

methanogenesis when metabolizing acetate to methane in the PSR group.

5

The findings detailed in Sections 3.2 and 3.3 indicate that the CO2 reduction

6

pathway increased by 4.7% in the PSR group compared to the control, the reason for the

7

enhancement of CO2 reduction pathway in PSR group was suggested as following:

8

Firstly, much more neutral substrate ethanol was generated in EP stage, which provided

9

suitable neutral environment for AD (Zhao et al., 2016; Wu et al., 2015). Further, the

10

thermodynamic analysis (ΔG) indicates that the conversion of ethanol to acetate was

11

faster than that of other organic acids, thereby increasing the SAOB (e.g., Clostridium)

12

and enhancing the acetic oxidation pathway (Zou et al., 2019). Secondly, the variation

13

of microbial community in AD system suggested that the core bacteria in the PSR group,

14

Clostridium, was increased by 7.6%, and that the relative concentration of the

15

hydrogen-consuming Methanobacterium was increased by 10.2%. Therefore, the

16

increase of hydrogenotrophic methanogenesis accelerated the conversion of H2 and CO2

17

to CH4, which reduced the hydrogen partial pressure in the AD reactor, thereby

18

increasing CO2 reduction.

19 20

4. Conclusions

21

This work used δ13C-ethanol as a substrate in mesophilic anaerobic digestion (AD)

22

after ethanol per-fermentation (EP) to study the methanogenic pathway. The results

23

indicate that the acetotrophic pathway is still the main metabolism pathway (59.3%) in

24

the PSR group. EP caused the relative concentration of Clostridium and

17

1

Methanobacterium to increase by 7.6% and 10.2% over the control group, respectively.

2

Finally, the CO2 reduction pathway was enhanced by 4.70%. The results of this study

3

provide a theoretical foundation for regulating the methanogenic pathway and

4

improving the biogas yield.

5 6

Acknowledgements

7

This study was financially supported by the National Key R&D Program

8

(2018YFC1900903), the Beijing Natural Science Foundation Program (8192028), the

9

National Natural Science Foundation of China (51578063), and the Foundation of the

10

Committee on Science and Technology of Tianjin (18YFHBZC00020).

11 12

References

13

Angelidaki, I., Chen, X.X., Cui, J.B., Kaparaju, P., Ellegaard, L., 2006. Thermophilic

14

anaerobic digestion of source-sorted organic fraction of household municipal solid

15

waste: start-up procedure for continuously stirred jar reactor. Water Res. 40, 2621–

16

2628.

17 18

Appels, L., Baeyens, J., Degreve, J., Dewil, R., 2008. Principles and potential of the anaerobic digestion of waste-activated sludge. Prog. Energy Combust. 34, 755–781.

19

Conrad, R., 1999. Contribution of hydrogen to methane production and control of

20

hydrogen concentrations in methanogenic soils and sediments. FEMS Microbiol.

21

Ecol. 28, 193–202.

22 23

Conrad, R., 2005. Quantification of methanogenic pathways using stable carbon isotopic signatures: a review and a proposal. Org. Geochem. 36, 739−752.

18

1 2

Deublein, D., Steinhauser, A., 2008. Biobiogas from waste and renewable resources: An introduction. John Wiley & Sons, Inc., Weinheim, Germany.

3

Ferry, J., 1993. Fermentation of acetate, in: Ferry, J.G. (Ed.), Methanogenesis. Ecology,

4

physiology, biochemistry and genetics. Chapman and Hall, New York, N.Y., pp.

5

305–334.

6

Hagen, L.H., Vivekanand, V., Linjordet, R., Pope, P.B., Eijsink, V.G.H., Horn, S.J.,

7

2014. Microbial community structure and dynamics during co-digestion of whey

8

permeate and cow manure in continuous stirred jar reactor systems. Bioresour.

9

Technol. 171, 350–359.

10

Hattori, S., Kamagata, Y., Hanada, S., Shoun, H., 2000. Thermacetogenium phaeum gen.

11

nov., sp. nov., a strictly anaerobic, thermophilic, syntrophic acetate-oxidizing

12

bacterium. Int. J. Syst. Evol. Microb. 50, 1601–1609.

13 14

Hattori, S., 2008. Syntrophic acetate-oxidizing microbes in methanogenic environments. Microbes Environ. 23, 118–127.

15

Hickey, R.F., Vanderwielen, J., Switzenbaum, M.S., 1987. The effects of organic

16

toxicants on methane production and hydrogen biogas levels during the anaerobic

17

digestion of waste activated sludge. Water Res. 21, 1417–1427.

18

Karakashev, D., Batstone, D.J., Angelidaki, I., 2005. Influence of environmental

19

conditions on methanogenic compositions in anaerobic biobiogas reactors. Appl.

20

Environ. Microbiol. 71, 331–338.

21

Karakashev, D., Batstone, D.J., Trably, E., Angelidaki, I., 2006. Acetate oxidation is the

22

dominant methanogenic pathway from acetate in the absence of Methanosaetaceae.

23

Appl. Environ. Microbiol. 72, 5138–5141.

19

1

Kim, M.D., Song, M., Jo, M., Shin, S.G., Khim, J.H., Hwang, S., 2010. Growth

2

condition and bacterial community for maximum hydrolysis of suspended organic

3

materials in anaerobic digestion of food waste-recycling wastewater. Appl.

4

Microbiol. Biotechnol. 85, 1611–1618.

5 6

Liu, Y., Whitman, W.B., 2008. Metabolic, phylogenetic, and ecological diversity of the methanogenic archaea. Ann. NY Acad. Sci. 1125, 171–189.

7

Ma, J., Duong, T.H., Smits, M., Verstraete, W., Carballa, M., 2011. Enhanced

8

biomethanation of food waste by different pre-treatments. Bioresour. Technol. 102,

9

592–599.

10 11

Miller, G.L., 1959. Use of dinitrosalicylic acid reagent for determination of reducing sugar. Anal. Chem. 31, 426–428.

12

Ministry of Housing and Urban-Rural Development of the People’s Republic of China,

13

2012. CJJ 184–2012: Technical code on food waste treatment. China Architecture &

14

Building Press.

15

Polag, D., May, T., Müller, L., König, H., Jacobi, F., Laukenmann, S., et al., 2015.

16

Online monitoring of stable carbon isotopes of methane in anaerobic digestion as a

17

new tool for early warning of process instability. Bioresour. Technol. 197, 161–170.

18

Qu, X., Mazeas, L., Vavilin, V., Epissard, J., Lemunier, M., Mouchel, J.-M., et al., 2009.

19

Combined monitoring of changes in δ13CH4 and archaeal community structure

20

during mesophilic methanization of municipal solidwaste. FEMS Microbiol. Ecol.

21

68, 236–245.

22

Refai, S., Wassmann, K., Deppenmeier, U., 2014. Short-term effect of acetate and

23

ethanol on methane formation in biogas sludge. Appl. Microbiol. Biotechnol. 98,

24

7271–7280.

20

1

Ren, Y., Yu, M., Wu, C., Wang, Q., Gao, M., Huang, Q., et al., 2018. A comprehensive

2

review on food waste anaerobic digestion: Research updates and tendencies.

3

Bioresour. Technol. 247, 1069–1076.

4

Rui, J., Li, J., Zhang, S., Yan, X., Wang, Y., Li, X., 2015. The core populations and co-

5

occurrence patterns of prokaryotic communities in household biobiogas digesters.

6

Biotechnol. Biofuels 8, 158.

7

Sasaki, D., Hori, T., Haruta, S., Ueno, Y., Ishii, M., Igarashi, Y., 2011. Methanogenic

8

pathway and community structure in a thermophilic anaerobic digestion process of

9

organic solid waste. J. Biosci. Bioeng. 111, 41–46.

10

Shigematsu, T., Tang, Y., Kobayashi, T., Kawaguchi, H., Morimura, S., Kida, K., 2004.

11

Effect of dilution rate on metabolic pathway shift between aceticlastic and

12

nonaceticlastic methanogenesis in chemostat cultivation. Appl. Environ. Microbiol.

13

70, 4048–4052.

14 15

Standard Methods for the Examination of Water and Wastewater, 1998. American Public Health Association, Washington, D.C.

16

Sun, C., Cao, W., Banks, C.J., Heaven, S., Liu, R., 2016. Biogas production from

17

undiluted chicken manure and maize silage: A study of ammonia inhibition in high

18

solids anaerobic digestion. Bioresour. Technol. 218, 1215–1223.

19

Wang, Y.Y., Zhang, Y.L., Wang, J.B., Meng, L., 2009. Effects of volatile fatty acid

20

concentrations on methane yield and methanogenic bacteria. Biomass Bioeng. 33,

21

848–853.

22

Wu, C., Wang, Q., Yu, M., Zhang, X., Song, N., Chang, Q., et al., 2015. Effect of

23

ethanol pre-fermentation and inoculum-to-substrate ratio on methane yield from

24

food waste and distillers’ grains. Appl. Energy 155, 846–853.

21

1

Yu, M., Wu, C., Wang, Q., Sun, X., Ren, Y., Li, Y.Y., 2018. Ethanol prefermentation of

2

food waste in sequencing batch methane fermentation for improved buffering

3

capacity and microbial community analysis. Bioresour. Technol. 248, 187–193.

4

Zhao, N., Yu, M., Wang, Q., Song, N., Che, S., Wu, C., et al., 2016. Effect of ethanol

5

and lactic acid pre-fermentation on putrefactive bacteria suppression, hydrolysis,

6

and methanogenesis of food waste. Energy Fuels 30, 2982–2989.

7

Zou, H., Gao, M., Wang, Q., Zhang, W., Wu., C, Song, N., 2019. Metabolic analysis of

8

efficient methane production from food waste with ethanol pre-fermentation using

9

carbon isotope labeling. Bioresour. Technol., doi: 10.1016/j.biortech.2019.121849.

22

1

Figure captions

2

Fig. 1. VFAs, pH, and changes in biogas production in the experimental groups.

3

Fig. 2. Schematic diagram of methanogenic pathway.

4

Fig. 3. Changes of bacterial and archaebacterial community at subordinate level in the

5

PSR and control groups.

6

23

1

Table 1. Physicochemical characteristics of the substrates used in this study Parameters

Food waste

Anaerobic digestion sludge

TS (%a)

23.6

12.0

VS (%a)

22.9

6.89

VS/TS (-)

97.0

57.2

TCOD (%b)

42.8

26.9

(%b)

2.38

2.49

TN

C/N (-)

18.0

10.8

(%b)

13.9

13.1

Carbohydrate (%b)

48.6

63.7

pH

5.63

7.61

Protein

2

a

On a wet basis.

3

b

On a dry basis.

24

1

Table 2. Operational parameters and reactor performance during stable operations. TS

VS

COD

TN

NH4+-N

Alkalinity

(g/L)

(g/L)

(g/L)

(g/L)

(g/L)

(g/L)

Initial mean value

29.12

19.84

18.28

0.82

0.51

1.08

6.22

PSR

10.81

6.87

7.49

0.38

0.18

1.86

7.55

Control

11.84

8.25

8.37

0.54

0.43

1.57

7.51

Reactor

2

25

pH

1

Table 3. Distribution of δ13CH4, (δ12CH4), and δ13CO2 (δ12CO2) production in the PSR

2

group and control group (24 h). CH4 produced from labelled ethanol Group PSR

Substrate

δ12CH4

δ13CH4

13CH

3CH2OH

191810

280408

CH313CH2OH

343016

188103

13CH

259701

506472

Methyl group/total CH4

a

Control

59.3%

3CH2OH

b

66.1% 13CH

CH3 Blank

2OH

no addition

473149

209767

6011

4532

CO2 produced from labelled ethanol Group PSR

Substrate

δ12CO2

δ13CO2

13CH

3CH2OH

4336890

450595

CH313CH2OH

6059388

549894

13CH

3CH2OH

5370492

740950

CH313CH2OH

2420347

639587

no addition

142026

53264

Carboxyl group/total CO2

a

Control

Blank

13.4%

b

17.1%

3

*All values are the average of three individual experiments. The peak areas at an m/z value of 16, 17,

4

44, and 45 from CH4 and CO2 were removed as the background.

5

a

PSR refers anaerobic digestion with ethanol pre fermentation.

6

b

Control refers anaerobic digestion without ethanol pre fermentation.

26

1

Table 4. Concentrations of δ13C-labelled carbon in the PSR and control groups. Addition of 1-13C-ethanol Group

PSR

Control

Addition of 2-13C-ethanol

Time

0 h (C-mM)

24 h (C-mM) Time

13CH

3CH2OH

97.9

0.31

CH313CH2OH

96.4

2.78

13CH

3COOH

0.07

2.28

CH313COOH

0.15

4.05

13CH

4

0.12

88.1

13CH

4

0.14

46.7

13CO

2

0.08

6.81

13CO

2

0.11

45.6

∑ sum

98.2

97.5

∑ sum

96.8

99.1

13CH

3CH2OH

97.2

2.48

CH313CH2OH

95.8

5.29

13CH

3COOH

0.14

5.85

CH313COOH

0.08

12.2

13CH

4

0.09

79.1

13CH

4

0.11

41.3

13CO

2

0.11

11.4

13CO

2

0.13

39.1

97.5

98.8

∑ sum

96.1

97.9

∑ sum 2 3

27

0 h (C-mM) 24 h (C-mM)

1 2

(a)

3 4 5 6 7 8

(b) 9 10 11 12 13 14 15

Fig. 1

28

1

H2 13CH

13CH

3CH2OH

3COOH

①

H2

② 59.3%

13CH

4

③ 13CO

2/CO2

40.7% ④

13CH

H2

4

/CH4

PSR

H2 CH313CH2OH

CH313COOH

66.1%

CH4

H2 13CO

Control

2 3

2/CO2

Fig. 2

4 5

29

33.9%

13CH

H2

4

/CH4

1 2

(a)

3 4 5 6 7 8

(b)

9 10 11 12 13 14 15 16 17

Fig. 3

18 19 20 21 22 23 24

Highlights  δ13C-Ethanol in mesophilic anaerobic digestion after ethanol pre-fermentation (EP) was studied.  The results indicate that the acetotrophic pathway is still the main metabolism pathway (59.3%) in the anaerobic digestion (AD) with EP.

30

1 2

 EP caused the relative abundance of Clostridium and Methanobacterium in AD system to increase by 7.6% and 10.2% over the control group, respectively.

3 4 5

Author Contributions Section

6

Hui Zou and Ming Gao designed and performed the experiment. Miao Yu,

7

Wenyu Zhang and Shuang Zhang analyzed the data. Hui Zou, Chuanfu Wu and

8

Qunhui Wang drafted the manuscript. Ming Gao, Yukihiro Tashiro and Qunhui

9

Wang gave the suggestion for the experiment and helped to revise the paper. All

10

authors gave final approval for publication.

11 12

31