Content uploaded by Wangwang Yan
Author content
All content in this area was uploaded by Wangwang Yan on Nov 07, 2019
Content may be subject to copyright.
1
Unveiling the role of activated carbon on hydrolysis process in anaerobic digestion
1
Wangwang Yan a, Liang Zhanga, Surya Maitri Wijayab, Yan Zhoua,b*
2
3
aAdvanced Environmental Biotechnology Centre, Nanyang Environment and Water Research
4
Institute, Nanyang Technological University, 637141, Singapore
5
b School of Civil and Environmental Engineering, Nanyang Technological University, 639798,
6
Singapore
7
8
*Corresponding Author: Yan Zhou
9
Address: 50 Nanyang Avenue, Singapore 639798
10
E-mail: Zhouyan@ntu.edu.sg
11
Tel: (+65) 67906103
12
Fax: (+65) 67910676
13
14
15
16
17
18
19
20
21
2
Abstract:
22
Conventionally, activated carbon is widely applied in water treatment systems due to its
23
capability of adsorbing inhibitors or stimulating methanogenesis rate. This study demonstrates
24
that powder activated carbon (PAC) also stimulate hydrolysis in anaerobic digestion (AD) of
25
thermal hydrolysis pretreated sludge. This is evidenced with 0.95 ̶ 1.42 times higher methane
26
generation, 12.46% ̶ 20.06% higher volatile solids removal and greater refractory compounds
27
degradation stimulated by PAC. Functional prediction reveals that genes coding hydrolytic
28
enzymes and xenobiotics metabolism were highly expressed with the presence of PAC.
29
Furthermore, the stimulated hydrolysis activity was effectively maintained at PAC concentration
30
as low as 0.125 g/L, though methanogenesis rate reduced by 80.30% compared to 1g/L case.
31
This study reports the role of activated carbon on the hydrolysis which has been ignored
32
previously and the impact of PAC on AD performance in long-term operation. The results
33
improve understanding on the true function of PAC in AD system.
34
Keywords: Thermal hydrolysis pretreated sludge; powder activated carbon; hydrolysis; organic
35
compounds transformation; functional prediction
36
37
38
39
40
41
42
3
1. Introduction
43
In municipal wastewater treatment plant, the disposal cost of waste activated sludge (WAS)
44
accounts for 30-60% of operational fee (Appels et al., 2008; Yu et al., 2016). Anaerobic
45
digestion (AD) is widely accepted as one of the most economical and energy-efficient strategies
46
to treat WAS, as it can effectively reduce the sludge volume accompanied with biogas generation
47
(Appels et al., 2008). Anaerobic digestion of WAS is mainly composed of three stages: (1)
48
solubilization/ hydrolysis of organic particles and biological polymers with the generation of
49
VFAs and H2 (2) degradation of VFAs to H2 and acetate by syntrophic bacteria and (3)
50
conversion of H2 and acetate to methane (Shimizu et al., 1993). Among them, hydrolysis is
51
widely recognized as the rate-limiting step in the anaerobic sludge digestion (Ge et al., 2011;
52
Kim et al., 2003). In order to improve the degradability of WAS and AD efficiency, various
53
sludge pre-treatment methods have been developed to facilitate the solubilization and
54
acidification of WAS, such as such as thermal hydrolysis (Lu et al., 2018), ultrasonic treatment
55
(Xie et al., 2016), chemical oxidation (Kim et al., 2016). Among them, the thermal hydrolysis
56
pretreatment (THP) has been extensively implemented in industrial scale given this technology
57
can greatly increase the solubilization of the organic matter and offer an additional benefit of
58
pathogens removal (Ruiz-Espinoza et al., 2012; Tanaka et al., 1997). However, the remaining
59
chemical oxygen demand (COD) in the THP-AD effluent are generally higher than that in
60
conventional AD systems leaving a higher organic loading on the downstream treatment systems
61
(Lu et al., 2018). In addition, microbial activities in the THP-AD system may be also affected by
62
the accumulated ammonia or other inhibitors during the operation (Bonmati et al., 2001;
63
Schwede et al., 2013). These would further reduce organic matter conversion efficiency.
64
4
Activated carbon is a remarkable adsorbent with high specific surface area and pore volume. In
65
traditional AD technology, a high concentration of activated carbon (generally > 10 g/L) is
66
widely applied to adsorb toxic compounds to maintain the operational stability of systems
67
(Bhatnagar et al., 2013). Recently, it is found that low concentration of activated carbon (ca.
68
1g/L) can also promote AD. The enhanced AD performance is due to the fact that activated
69
carbon could act as conduits to accelerate direct interspecies electron transfer (DIET) among
70
syntrophic microbes (Lee et al., 2016; Yan et al., 2017). Although no one examined the
71
feasibility of applying activated carbon to improve the performance of AD system that is fed
72
with THP treated WAS, it could expect that activated carbon may have similar benefits in THP-
73
AD system.
74
Besides the stimulated methanogenesis rate, it is often observed that the total methane production
75
is also enhanced with the presence of activated carbon (Lee et al., 2016; Wang et al., 2018;
76
Zhang et al., 2018). For example, Zhang et al. (2018) reported that the supplement of activated
77
carbon increased methane yield by 41% in a conventional pilot-scale AD system. The greater
78
methane production could be attributed from the presence of more biodegradable carbon sources.
79
As such, we hypothesize that activated carbon may facilitate the hydrolysis process, which can
80
disintegrate more organic particles or colloids and in turn release more bioavailable carbon
81
source to support methanogenic activity for methane generation. However, previous studies often
82
argue that the accelerated rates of organic conversion and methanogenesis are due to the
83
stimulated DIET pathway (Cruz Viggi et al., 2014; Yin et al., 2018; Zhao et al., 2017).
84
Theoretically, the hydrolysis activity, which is driven by extracellular hydrolytic enzymes, is not
85
directly linked with the DIET pathway, as DIET mainly functions via accelerating interspecies
86
5
electron transfer. To date, there is no study to experimentally examine the effects of activated
87
carbon on hydrolysis process in AD systems.
88
During the AD operation, the concentration of dosed activated carbon would decrease along with
89
time owing to the loss of activated carbon during sludge discharging process. In addition, new
90
feeding sludge into the systems would further dilute its concentration. Whether such dilution
91
effect would reduce activated carbon’s impact on the AD system is not clear. Therefore, this
92
study aims to (a) investigate the effects of different concentration of powder activated carbon
93
(PAC) on all AD reaction steps with the focus on hydrolysis and the activities of hydrolytic
94
enzymes; (b) characterize the carbon transformation in each AD step using a size exclusion
95
chromatography (SEC) equipped with LC-OCD-OND system to better understand the effects of
96
PAC on AD process; and (c) elucidate the evolution of microbial community, metabolism and
97
DIET related functional genes using Illumina MiSeq sequencing and PICRUSt (Phylogenetic
98
Investigation of Communities by Reconstruction of Unobserved States) analysis.
99
2. Materials and methods
100
2.1 Seed and feed sludge source
101
The inoculum of the anaerobic reactors was collected from a local wastewater treatment plant.
102
The volatile solids (VS) concentration of the inoculum was 9.44 g/L. Thermal hydrolysis
103
pretreated waste activated sludge (THP sludge) was adopted as the feed. The preparation of THP
104
sludge was according to our previous study (Lu et al., 2018). Briefly, the WAS was firstly
105
warmed up at 70 oC for 30 mins, followed by thermal hydrolysis treatment at 121 oC for 30 mins
106
in a vertical pressure steam sterilizer. The VS concentration of THP sludge was adjusted to 9.27
107
g/L, so that the VS ratio of seed and feed sludge was approximately equal to 1.
108
6
2.2 Long-term effects of PAC on anaerobic digestion
109
This experiment was carried out using an automatic methane potential test system (AMPTSII,
110
Bioprocess Control Company, Sweden). Six anaerobic reactors each containing a mixture of 180
111
mL inoculum and 180 mL THP sludge were employed in this test, in which three of the six
112
reactors were supplemented with 1 g/L PAC (PAC group). PAC was supplied by Sigma-Aldrich.
113
The grain size d50, specific surface area, and H of point of zero charge of PAC were 15 µm, 1150
114
m2· g-1, and > 7.3, respectively. The rest three reactors without PAC were used as a reference
115
(Control group). All the reactors were purged with nitrogen for 8 min and sealed immediately to
116
create anaerobic environment. All the AD reactors were operated at 35 ºC with a continuous
117
stirring speed of 60 rpm. The entire operational period (120 days) was divided into four phases,
118
and each phase lasted for 30 days. Before entering into the next phase, 50% of the sludge (180
119
mL) in each reactor was replaced with 180 mL THP sludge. Thus, PAC concentration in PAC
120
group reactors from the first to fourth phase was 1 g/L, 0.5 g/L, 0.25 g/L and 0.125 g/L,
121
respectively. During the operation, the supernatant was periodically taken out for total organic
122
carbon (TOC), volatile fatty acids (VFA) and organic compounds composition analysis. At the
123
end of each phase, sludge samples were also collected to determine volatile solids (VS) removal.
124
The hydrolysis capability of the sludge in the last phase was characterized by analyzing the two
125
representive hydrolytic enzymes - protease and α-glucosidase.
126
2.3. The correlation between PAC dosage and AD performance
127
To explore the possible correlation between PAC dosage and AD performance, five sets of batch
128
groups dosed with different concentration of PAC (i.e. 1 g/L, 0.5 g/L, 0.25 g/L and 0.125 g/L
129
and 0 g/L) were set-up. Triplicates were performed for each PAC concentration. The other
130
experimental conditions, including the amount of seed, fed sludge and the setup procedures, were
131
7
the same as described in section 2.2. The methane generation profile of each batch reactor was
132
closely monitored for 30 days to investigate the correlation between PAC abundance and
133
methanogenesis.
134
2.4 Hydrolytic activity of sludge with presence of PAC
135
In this experiment, four sets batch groups were carried out in 500 mL reactors which were all
136
incubated with 180 mL of inoculum and 30 mL of THP sludge. Each batch group had three
137
replicates. 1 g/L PAC was dosed into six identical reactors. To inhibit the methanogenic
138
activities, 40 mmol/L sodium 2-bromoethanesulfonate (BES) was spiked into three parallel
139
reactors. The other six reactors without PAC addition were used as the Controls. The experiment
140
conditions were maintained the same as those in section 2.2. The batch tests lasted for 1 month.
141
To compare the hydrolytic performance of each group, TOC was measured on day 0, 3, 6 10, 16
142
and 30, and the electron transport activity was determined on day 0, 9, 18, and 27.
143
2.5 Analytical methods
144
2.5.1 The activity of hydrolytic enzymes
145
The activity of hydrolytic enzymes was indicated by the activity of protease and a-glucosidase.
146
The mixed liquor samples were taken from the end of the phase 4 from the long term study
147
(section 2.2). After filtered through the 0.45 um filter, the liquid samples were collected for
148
enzymes activity analysis. Protease was analyzed using a universal protease activity assay kit
149
(P23236, Thermo Fisher). The α-glucosidase activity was measured by using α-glucosidase
150
activity assay kit (MAK 123, Sigma). The detailed protocols were employed as per the
151
instructions of manufacturers.
152
2.5.2 Electron transport system activity
153
8
Electron transport system activity (ETS) of AD sludge is proportional to substrate removal rate
154
and found to be an effective indicator of overall microbial respiration activity in the AD process
155
(Wang et al., 2016; Yin et al., 2018). This method is based on measuring the activity of
156
dehydrogenase, which is involved in transferring electrons from substrates to electron carriers.
157
The ETS activity of sludge samples from section 2.4 was determined for exploring hydrolytic
158
activities. ETS was quantified by using a modified 2-(p-iodophenyl)-3- (p-nitrophenyl)-5-
159
phenyltetrazolium chloride (INT) method (Yin et al., 2005) to compare the whole metabolism
160
activity of AD sludge impacted from PAC and BES. Detailed procedures can be found in our
161
previous study (Yan et al., 2018).
162
2.5.3 Characterization of the organics involved in different AD steps
163
Anaerobic digestion (AD) can be divided into four steps, and each step is characterized with
164
distinct substrates and products. Based on the characterization of the organics, high molecular
165
weight (HMW) protein, HMW polysaccharides and Humic-like substances (HS) are classified as
166
the sludge solubilization products, in other words, hydrolytic products released from insoluble
167
organics. LMW protein, LMW polysaccharides and building blocks are assigned to hydrolysis
168
products from broken down macromolecules. Volatile fatty acids belong to acidogenic products,
169
and acetic acid is classified as the products of acetogenesis process. The effects of PAC on each
170
AD step was explored based on transformation performance of different organic compounds.
171
The size exclusion chromatography (SEC) equipped with LC-OCD-OND system (DOC-
172
LABOR, Karlsruhe, Germany) was adopted to quantify the abundance of different types of
173
organic compounds in liquid collected during the experiment described in section 2.2. The
174
detailed analytical procedures were described in our previous study (Yan et al., 2018).
175
9
2.5.4 The quantification of methane generation
176
A modified Gompertz model was applied to quantitatively analyze the methanogenesis
177
performance.
178
𝑀(𝑡)= 𝑃 × 𝑒{−𝑒 [𝑅𝑚𝑎𝑥 ×𝑒
𝑃 × (𝜆− 𝑡)+ 1]} Eq.1
179
R = Rmax ÷ (VS× V) Eq.2
180
Where, M(t) is the accumulated methane yield at time t (h); P is the potential methane generation
181
(mL); e is the exponential constant; Rmax is the maximum methane generation rate (mL/day); λ
182
is the lag time (h). R is the specific methane generation rate (mL/g VSS/day). VS is the volatile
183
solids of THP sludge (9.27 g/L); V is the volume of feed THP sludge (180 mL).
184
2.5.5 Other analysis
185
The VS, total solids (TS) and volatile suspended solids (VSS) were measured following the
186
standard methods (APHA, 2012).
187
2.6 Microbial community analysis and function prediction
188
Biomass from the end of the first and fourth phase was collected for DNA extraction using
189
FastDNA Spin kit for soil (MP Biomedicals, USA). The variable region V4-V5 of the 16s RNA
190
gen was chosen for constructing the bacterial community for MiSeq sequencing. The Illumina
191
Miseq sequencing service were supplied by Novengene Corporation ( Beijing, China).
192
Operational taxonomic units (OTUs) were defined by clustering at a 0.03 distance (97%
193
similarity) threshold. Final OTUs were taxonomically classified using BLASTN against a
194
curated database derived from RDPII, NCBI and Green Genes.
195
10
The metabolic dynamics of each samples were predicted by applying Phylogenetic investigation
196
of communities by Reconstruction of Unobserved States (PICRUSt) analysis, which was
197
performed using the Galaxy platform against the Kyoto Encyclopedia of Gene and Genome
198
(KEGG) Orthology database (Langille et al., 2013). The metabolism difference was analyzed by
199
using the model KEGG after normalizing OUT table. Detailed analysis of functional genes was
200
conducted on the sequences relating to metabolism, hydrolysis and the direct interspecies
201
electron transport related adjuncts.
202
3. Results and discussion
203
3.1. The effects of PAC on methanogenesis
204
3.1.1 The methanogenic performance influenced by PAC
205
Methanogenesis is a key index to assess the AD performance and the related microbial activity.
206
As shown in Fig. 1, the presence of PAC significantly accelerated the methane generation rate.
207
During the first phase, the specific methane generation rate in PAC group (49.81 mL· g-1 VS · d-
208
1) was approximately three times higher than that in Control group (13.64 mL· g-1 VS · d-1 ),
209
indicating that the presence of PAC triggered DIET in the reactors (Lee et al., 2016; Liu et al.,
210
2012; Wang et al., 2018). Then, the difference of methane generation rate between PAC group
211
and Control group narrowed down in the following three phases. Specifically, the difference
212
decreased to 1.10, 0.76 and 0.45 times, at the 2nd , 3rd and 4th phase. This is because of the
213
gradually decreased specific methane generation rate in PAC group, while the corresponding
214
value in Control group enhanced gradually as the process continued. The phenomenon is likely
215
due to the reduction of PAC amount, as a strong positive correlation is shown between PAC
216
abundance and the specific methane generation rate (r=17.925, R2=0.9902). Since the methane
217
11
generation rate increased with the elongated operational time in Control group, it is concluded
218
that the methanogenic activity was not inhibited during the operation. Thus, the attenuated
219
influence of PAC is a predominated factor to the declined specific methanogenesis rate.
220
Besides the accelerated methanogenesis rate, total methane production was almost doubled with
221
the presence of PAC compared with Control group. It should be noted that the amount of
222
residual dissolved organic compounds exhibited little difference between the two groups at the
223
end of each phase. Thus, the greater methane production in PAC groups is likely attributed to
224
more bioavailable carbon sources, which was released from hydrolysis of organic particles or
225
colloids. This is demonstrated by higher VS reduction rate in PAC groups. The VS reduction
226
rates were 23.88% to 40.96% higher in PAC group compared with that of Control group during
227
the four phases (Table 1). Thus, PAC not only promoted the acidogenesis and methanogenesis as
228
reported previously (Cruz Viggi et al., 2014; Zhao et al., 2016), but also enhanced the hydrolysis
229
of organic particles and/or colloids. The mechanism of such phenomena is further explored in the
230
following sections. It is noteworthy that, although the specific methane generation rate of the
231
PAC group declined by 30.21% from the first phase to the fourth phase, the total methane
232
production only reduced by 9.48% (Fig. 1). Thus, PAC at low very concentration can still
233
effectively stimulate the hydrolysis.
234
The maintained efficiency of PAC on methanogenesis could be due to the long-term incubation
235
regardless of the difference of PAC abundance. To clarify this point, another batch study with
236
different PAC dosage was conducted to investigate the correlation between PAC concentration
237
and AD performance. As shown in Fig. 2, both the methane generation rate and total methane
238
production are positively correlated with the PAC concentration. Specifically, the methane
239
production in 1g/L PAC group (667.73 mL/g VS) is 1.35 times higher than that in Control group
240
12
(284.59 mL/g VS). Corresponding to the long-term performance, the methane production in
241
0.125 g/L PAC group (536.16 mL/g VS) also reaches 80.30% of 1g/L PAC group, suggesting
242
that the hydrolysis activity can also be effectively maintained at PAC concentration of 0.125 g/L
243
without long-term acclimation.
244
3.1.2 Organic compounds transformation during each AD reaction step
245
In the first phase (PAC: 1 g/L, Fig. 3), the concentration of all dissolved organic compounds (i.e.
246
HS, HMW polysaccharides, HMW protein, building blocks, LMW polysaccharides, LMW
247
protein, and VFA ) was significantly lower in PAC group than those in the Control group. The
248
HS and HMW polysaccharides, the solubilization products, in PAC group were only 65.22% and
249
60.36% of that in Control group at the accumulation point of day 8. The maximum amount of the
250
acidogenic substrates - building blocks and LMW polysaccharides – were also 79.65% and
251
35.65% lower in the presence of PAC. In addition, a consistently fast degradation rate of both
252
HMW protein and LMW protein was observed in PAC group. Furthermore, PAC group also had
253
faster degradation rate and less accumulation of the acidogenic and acetoclastic products (i.e.
254
VFA). Specifically, the amount of propionate peaked on day 8 in PAC group with a
255
concentration of 7.98 mM, while the highest concentration of propionate (11.08 mM) was found
256
on day 21 in Control group. Compared to Control group, the consumption rate of the
257
methanogenesis substrate, i.e. acetic acid, was much faster in PAC group (Fig. 2a). Therefore, all
258
the four AD reaction steps were stimulated by PAC in the first phase.
259
In the second phase (PAC: 0.5 g/L, Fig. 3a), the accumulation of the solubilization products was
260
also significantly lower in the presence of PAC. However, a narrowed discrepancy between PAC
261
and Control group was observed at this phase. Specifically, within the two groups, the maximum
262
variance of all solubilization products decreased to less than 15%; the degradation performance
263
13
of all hydrolysis products exhibited little difference; and the difference in accumulated VFA
264
reduced to 17.41%. Thus, the impact of PAC on the performance in this phase became weakened
265
compared to that in the first phase.
266
In the third phase (PAC: 0.25 g/L, Fig. 3), the accumulated solubilization products are
267
significantly more in PAC group than in Control group, which is contrary to the results in the
268
first and second phases. In particular, the peak amount of HMW protein, HMW PS and HS were
269
10% ̶ 20% higher in PAC group. The results suggest the solubilization rate of organic
270
particulars/colloids was relatively faster than the consumption/hydrolysis rate of these
271
compounds in PAC group. In other words, the solubilization process was more effectively
272
promoted than hydrolysis at PAC concentration of 0.25 g/L. Similar to the 2nd phase (PAC: 0.5
273
g/L), the degradation of the hydrolysis products exhibited little difference between Control and
274
PAC group. As for the acidification products, significant higher amount of total VFA and
275
propionate were observed in PAC group, while accumulated acetic acid was slightly lower,
276
suggesting that solubilization and hydrolysis rates exceeded acidogenesis/acetogenesis rate in
277
PAC group.
278
During the fourth phase (PAC: 0.125 g/L, Fig. 3), only the accumulated amount of HS and
279
propionate were much higher in PAC group. This phenomenon confirmed that PAC is more
280
effective in stimulating the solubilization and hydrolysis of organic particles, and relatively less
281
efficient in promoting the conversion of propionate, i.e. acetogenesis, at low PAC abundance.
282
Overall, the transformation profiles of the organic compounds clearly demonstrate the positive
283
correlation between PAC abundance and the AD performance. It is worth pointing that HS are
284
refractory compounds due to the chemical structure of aromatic nuclei with carboxylic and
285
phenolic substituents. The accumulated amount of HS was significantly lower in PAC group than
286
14
Control group at both 1 g/L and 0.5 g/L of PAC groups, while an opposite phenomenon was
287
observed at PAC ≤0.25 g/L. Such phenomenon suggests that the degradation efficiency of
288
refractory compounds could be more effective at high PAC concentration.
289
3.2 The effects of PAC on hydrolytic enzymes and activities
290
3.2.1 The activities of key hydrolytic enzymes
291
The extracellular protease and glycoside hydrolases are normally characterized to represent the
292
hydrolytic activity. As such, during the fourth phase, the enzymatic activities of protease and a-
293
glucosidase were investigated to figure out the role of PAC on hydrolysis activity. As shown in
294
Table 2, the protease activity increased greatly in the initial 17 days in both groups, followed by
295
a slight decrease of protease from day 18 to day 30, which could be due to the decrease of
296
bioavailable substrates as a function of time. It should be noted that the protease activity in PAC
297
group was always more than 20% higher than that in Control group. Similar to protease,
298
significantly higher activity of a-glucosidase was also always observed in PAC group. The
299
present results indicate that the more methane generation with the presence of PAC could result
300
from the stimulated activity of hydrolytic enzymes.
301
3.2.2 The effects of PAC on hydrolytic activity
302
To further determine whether the promoted hydrolysis is an independent process, or it is
303
attributed to the faster intermediate’s removal in the down-stream steps. BES was added into
304
reactors to inhibit the methanogenic activity, which could prevent the consumption of the
305
acetogenesis products. According to the TOC profile, the concentration of accumulated dissolved
306
TOC in the presence of PAC was significantly higher than that in the absence of PAC after
307
methanogenesis was inhibited by BES (Fig. 4a). Specifically, the amount of the dissolved TOC
308
15
in PAC-BES group was 28.76%, 26.19%, 14.92%, 30.51% and 27.07% higher than that in
309
Control-BES group on day 3, day 6, day 10, day 16, and day 30, respectively. This fact suggests
310
that PAC promoted hydrolysis did not rely on the driving force from the faster intermediate
311
products removal. Besides the higher accumulation of TOC was observed in PAC-BES group
312
compared to Control-BES group (Fig. 4a), the poor adsorption capability of the 1 g/L PAC was
313
also observed. These facts confirm that the physical absorption capability of PAC was not a
314
dominate attributor to its positive effects on the digestion of THP sludge.
315
The overall metabolic performance was characterized by measuring the ETS activity. As shown
316
in Fig. 4b, the activity of ETS was 33.2%, 66.1% and 34.7% higher in PAC-BES compared to
317
Control-BES on day 9, day 18 and day 27, respectively. The observed ETS activity mainly
318
attributed to the hydrolysis and acidogenesis processes since the methanogenesis was inhibited
319
by BES. The results demonstrate that the existence of PAC could stimulate the hydrolysis step,
320
even when the down-stream process, i.e. methanogenesis, was inhibited.
321
3.3 Microbial Community analysis
322
Principal Coordinates Analysis (PCoA) based on weighted UniFrac distance shows that only
323
high PAC concentration (1 g/L) significantly influenced the microbial community structures,
324
while low PAC concentration microbial population were grouped together with control cases.
325
The influence of PAC on microbial community was further investigated at the phylum and genes
326
level. The top ten phyla, accounting for more than 95% of the total microbial community, was
327
selected for further discussion (Fig. 5a). Firmicutes was the most predominant phylum in the
328
inoculum (58.4%). Its relative abundance decreased significantly to 42.9% and then returned to
329
60.3% in PAC group, while, in Control group, its abundance remained stable and was much
330
higher than that in PAC groups during the whole operational period. In contrast, the relative
331
16
abundance of Bacteroidetes in PAC group was much higher than that in Control group, which is
332
in agreement with the findings in Venkiteshwaran et al. (2015), who reported that the hydrolysis
333
activity of biomass was positively correlated with the abundance of Bacteriodietes. Similar to
334
Bacteroidetes, the relative abundance of Chloroflexi also increased greatly in the first phase, then
335
fell back in the fourth phase. The genera Longilinea, under the phylum of Chloroflexi, was
336
reported with the capability in hydrolyzing diverse types of carbohydrates and proteins
337
(Hatamoto et al., 2007). At the end of the first phase, the abundance of Longilinea was one time
338
higher in PAC (6.68%) than Control group (2.83%), but no significant difference between the
339
two groups was shown in the fourth phase. Different from Bacteroidete and Chloroflexi, the
340
relative abundance of phylum Spirochaetes was significantly higher in PAC group in both the
341
first and fourth phases. Spirochaetes are potential involved in DIET process, as they are
342
chemotrophs and can obtain energy by the oxidation of extracellular electron donors in
343
circumstances (Margulis et al., 1993). In addition, the strains affiliated with the Spirochaetes
344
phylum are able to grow on a variety of carbohydrates via the hydrolytic breakdown of high-
345
polymeric cellulose and hemicellulose (Bryant, 1952). Overall, the presence of PAC suppressed
346
the growth of genera affiliated with Firmicutes phylum, and alternatively promoted the growth of
347
hydrolytic and acidogenic bacteria under other phyla.
348
Methanosaeta, Methanosarcina and Methanospirillum were the most abundant archaea,
349
accounting for more than 98% of the Euryarchaeota OTUs. Species under both Methanosaeta
350
and Methanosarcina have widely been reported to possess the capability of accepting electrons
351
and participating in the DIET pathway (Liu et al., 2012; Rotaru et al., 2014b) (Fig. 5b).
352
Compared to Control group, the relative abundance of Methanosaeta was around nine (first
353
phase) and two folds (fourth phase) higher in PAC group (Fig. 5b), while the relative abundance
354
17
of Methanosarcina was significantly lower in PAC group (Fig. 5b). Therefore, it is predicted the
355
electro-active methanogens involving in DIET pathway was mainly ascribed to the
356
Methanosaeta rather than Methanosarcina. Methanospirillum is a kind of hydrogenotrophic
357
methanogens, which produce energy via the reduction of carbon dioxide with hydrogen
358
(Gunsalus et al., 2016). In this study, its relative abundance was always more than two times
359
higher in PAC group. The hydrolytic process of organic molecules is often accompanied with the
360
production of hydrogen, which could further stimulate the growth of hydrogenotrophic
361
methanogens. For example, the growth of hydrogenotrophic methanogens was greatly
362
stimulated when co-culture with Longilinea, which always showed much higher abundance in
363
PAC group due to its activity in degrading polymeric substances (Hatamoto et al., 2007). In PAC
364
groups, more organics underwent acidogenesis, thus providing the substrate. H2, for the growth
365
of hydrogenotrophic methanogens.
366
3.4 Microbial functional prediction
367
3.4.1 Identification of the genes related to metabolism, hydrolysis and DIET
368
There were more than 6900 unique protein genes identified from microbial metagenome of each
369
sample. As shown in Fig. 6a, the metabolism can be classified into 10 sub-categories, i.e. Amino
370
Acid Metabolism, Carbohydrate Metabolism, Energy metabolism, Glycan Biosynthesis and
371
Metabolism, Lipid Metabolism, Metabolism of Cofactors and Vitamins, Metabolism of other
372
Amino Acids, Metabolism of Terpenoids and Polyketides, Xenobiotics Biodegradation and
373
Metabolism, and Nucleotide Metabolism, Among all metabolism pathways, the Xenobiotics
374
Biodegradation and Metabolism pathway exhibited the most pronounced variation among all
375
samples. Specifically, the gene sequences relating to Xenobiotics Biodegradation and
376
Metabolism pathway was 8.11% ̶ 14.86% higher in PAC group than that in Control group (Fig.
377
18
6a). The present result is supported by the finding of significant higher removal rate of HS in the
378
presence of PAC (Fig. 3a). The higher abundance of Xenobiotics Biodegradation and
379
Metabolism pathway related genes verified that the presence of PAC can trigger the capability of
380
microbial community to degrade those refractory compounds, which may not occur in the control
381
group.
382
As the hydrolysis of organic compounds was greatly stimulated by PAC, the functional genes
383
that are involved in biosynthesis of extracellular hydrolysis enzymes were investigated to figure
384
out the mechanism behind this phenomenon. More than 400 hydrolytic enzymes are identified.
385
Based on their catalyzed reactions, the hydrolytic enzymes mainly compose of seven types
386
(Sikora et al., 2018) (Fig. 6b). It is interesting to find that the presence of PAC maintained the
387
relatively higher abundance of the Halide bonds by 12.92% and 30.73% during the first phase
388
and fourth phase, respectively (Fig. 6b). The halide bonds are difficult to be biodegraded due to
389
the short distance between halogen and lewis base (Metrangolo et al., 2005). The higher
390
abundance of genes relating to the hydrolysis of Halide bonds suggests that the presence of PAC
391
could greatly stimulate bacterial capability in catalyzing the cleavage of the strong chemical
392
bonds.
393
3.4.2 Identify the genes coding for DIET-related proteins in methanogens
394
Previous transcriptomic studies demonstrated that F420 H2 dehydrogenase is the key enzyme for
395
electron transfer into the cytoplasm in DIET pathway (Holmes et al., 2018). As such, the relative
396
abundance of F420 H2 dehydrogenase was analyzed. The results reveal that the relative
397
abundance of F420 H2 dehydrogenase was 48.03% higher in PAC (1g/L) than that in Control
398
group after 30 days incubation. In addition, its relative abundance at 1 g/L PAC was much
399
19
higher than that at 0.125 g/L PAC, suggesting that DIET pathway toward methane generation is
400
closely related with the PAC concentration.
401
The electrically conductive pili (pilA-N, GSU1496) and outer membrane cytochromes (omcS,
402
GSU2504) are required for extracellular electron transfer between G. sulfurreducens and G.
403
metalliteducens (Rotaru et al., 2014a; Summers et al., 2010). The pilA-N gene was much higher
404
in PAC groups compared to Control groups. However, it is difficult to compare the abundance of
405
omcS due to the extremely low relative abundance. This is because the reported electroactive
406
adjuncts are limited to certain strains, such as Geobacter (Rotaru et al., 2014a), which are rarely
407
found in many conductive materials supplemented reactors (Lü et al., 2016; Luo et al., 2015)
408
including this study. Therefore, more genes related to code the electroactive adjuncts should be
409
investigated for further understanding the mechanism of direct interspecies electron transfer
410
pathway.
411
4. Conclusions
412
This study demonstrates that PAC could greatly enhance the degradation of refractory
413
compounds and the hydrolysis of organic particles or colloids via promoting the hydrolytic
414
enzymes activity. The metabolic function prediction reveals that PAC enriched the genes related
415
to hydrolysis and the Xenobiotics Biodegradation and Metabolism. By tracking the
416
transformation of organic compounds, it is found that PAC could maintain its superiority in
417
stimulating hydrolysis at very low concentration, i.e. 0.125 g/L. This study thoroughly
418
investigates the long-term effects of PAC on AD performance fed with THP sludge, the result of
419
which could provide important information for designing the strategy of dosing activated carbon
420
in future full-scale application.
421
20
Supplementary data associated with this article can be found in the online version.
422
References:
423
1. APHA. 2012. Standard methods for the examination of water and wastewater. American
424
Public Health Association Washington, DC.
425
2. Appels, L., Baeyens, J., Degrève, J., Dewil, R. 2008. Principles and potential of the anaerobic
426
digestion of waste-activated sludge. Prog. Energy. Combust. Sci., 34(6), 755-781.
427
3. Bonmati, A., Flotats, X., Mateu, L., Campos, E. 2001. Study of thermal hydrolysis as a
428
pretreatment to mesophilic anaerobic digestion of pig slurry. Water Sci. Technol., 44(4), 109-
429
116.
430
4. Bryant, M. 1952. The isolation and characteristics of a spirochete from the bovine rumen. J.
431
Bacteriol., 64(3), 325.
432
5. Cruz Viggi, C., Rossetti, S., Fazi, S., Paiano, P., Majone, M., Aulenta, F. 2014. Magnetite
433
particles triggering a faster and more robust syntrophic pathway of methanogenic propionate
434
degradation. Environ. Sci. Technol., 48(13), 7536-43.
435
6. Gunsalus, R.P., Cook, L.E., Crable, B., Rohlin, L., McDonald, E., Mouttaki, H., Sieber, J.R.,
436
Poweleit, N., Zhou, H., Lapidus, A.L. 2016. Complete genome sequence of
437
Methanospirillum hungatei type strain JF1. Standards in genomic sciences, 11(1), 2.
438
7. Hatamoto, M., Imachi, H., Fukayo, S., Ohashi, A., Harada, H. 2007. Syntrophomonas
439
palmitatica sp. nov., an anaerobic, syntrophic, long-chain fatty-acid-oxidizing bacterium
440
isolated from methanogenic sludge. Int. J. Syst.Evol. Microbiol., 57(9), 2137-2142.
441
8. Holmes, D.E., Rotaru, A.-E., Ueki, T., Shrestha, P.M., Ferry, J.G., Lovley, D.R. 2018.
442
Electron and proton flux for carbon dioxide reduction in methanosarcina barkeri during direct
443
interspecies electron transfer. Front. Microbiol., 9, 3109.
444
9. Kim, M.S., Lee, K.-M., Kim, H.-E., Lee, H.-J., Lee, C., Lee, C. 2016. Disintegration of waste
445
activated sludge by thermally-activated persulfates for enhanced dewaterability. Environ. Sci.
446
Technol., 50(13), 7106-7115.
447
10. Langille, M.G., Zaneveld, J., Caporaso, J.G., McDonald, D., Knights, D., Reyes, J.A.,
448
Clemente, J.C., Burkepile, D.E., Thurber, R.L.V., Knight, R. 2013. Predictive functional
449
profiling of microbial communities using 16S rRNA marker gene sequences. Nat.
450
Biotechnol., 31(9), 814.
451
11. Lee, J.Y., Lee, S.H., Park, H.D. 2016. Enrichment of specific electro-active microorganisms
452
and enhancement of methane production by adding granular activated carbon in anaerobic
453
reactors. Bioresour. Technol., 205, 205-12.
454
12. Liu, F., Rotaru, A.-E., Shrestha, P.M., Malvankar, N.S., Nevin, K.P., Lovley, D.R. 2012.
455
Promoting direct interspecies electron transfer with activated carbon. Energy Environ. Sci.,
456
5(10), 8982.
457
13. Lu, D., Sun, F., Zhou, Y. 2018. Insights into anaerobic transformation of key dissolved
458
organic matters produced by thermal hydrolysis sludge pretreatment. Bioresour. Technol.,
459
266, 60-67.
460
14. Lü, F., Luo, C., Shao, L., He, P. 2016. Biochar alleviates combined stress of ammonium and
461
acids by firstly enriching Methanosaeta and then Methanosarcina. Water Res., 90, 34-43.
462
21
15. Luo, C., Lü, F., Shao, L., He, P. 2015. Application of eco-compatible biochar in anaerobic
463
digestion to relieve acid stress and promote the selective colonization of functional microbes.
464
Water Res., 68, 710-718.
465
16. Margulis, L., Ashen, J.B., Sole, M., Guerrero, R. 1993. Composite, large spirochetes from
466
microbial mats: spirochete structure review. PNAS, 90(15), 6966-6970.
467
17. Metrangolo, P., Neukirch, H., Pilati, T., Resnati, G. 2005. Halogen bonding based
468
recognition processes: a world parallel to hydrogen bonding. Accounts Chem. Res., 38(5),
469
386-395.
470
18. Rotaru, A.-E., Shrestha, P.M., Liu, F., Markovaite, B., Chen, S., Nevin, K.P., Lovley, D.R.
471
2014a. Direct interspecies electron transfer between Geobacter metallireducens and
472
Methanosarcina barkeri. Appl. Environ. Microbiol., 80(15), 4599-4605.
473
19. Rotaru, A.-E., Shrestha, P.M., Liu, F., Shrestha, M., Shrestha, D., Embree, M., Zengler, K.,
474
Wardman, C., Nevin, K.P., Lovley, D.R. 2014b. A new model for electron flow during
475
anaerobic digestion: direct interspecies electron transfer to Methanosaeta for the reduction of
476
carbon dioxide to methane. Energy Environ. Sci., 7(1), 408-415.
477
20. Ruiz-Espinoza, J.E., Méndez-Contreras, J.M., Alvarado-Lassman, A., Martínez-Delgadillo,
478
S.A. 2012. Effect of low temperature thermal pre-treatment on the solubilization of organic
479
matter, pathogen inactivation and mesophilic anaerobic digestion of poultry sludge. J.
480
Environ. Sci.Health, Part A, 47(12), 1795-1802.
481
21. Schwede, S., Rehman, Z.-U., Gerber, M., Theiss, C., Span, R. 2013. Effects of thermal
482
pretreatment on anaerobic digestion of Nannochloropsis salina biomass. Bioresour. Technol.,
483
143, 505-511.
484
22. Sikora, A., Detman, A., Mielecki, D., Chojnacka, A., Błaszczyk, M. 2018. Searching for
485
Metabolic Pathways of Anaerobic Digestion: A Useful List of the Key Enzymes. in: Biogas,
486
IntechOpen.
487
23. Summers, Z.M., Fogarty, H.E., Leang, C., Franks, A.E., Malvankar, N.S., Lovley, D.R.
488
2010. Direct exchange of electrons within aggregates of an evolved syntrophic coculture of
489
anaerobic bacteria. Science, 330(6009), 1413-1415.
490
24. Tanaka, S., Kobayashi, T., Kamiyama, K.-i., Bildan, M.L.N.S. 1997. Effects of
491
thermochemical pretreatment on the anaerobic digestion of waste activated sludge. Water
492
Sci. Technol., 35(8), 209-215.
493
25. Venkiteshwaran, K., Bocher, B., Maki, J., Zitomer, D. 2015. Relating anaerobic digestion
494
microbial community and process function: supplementary issue: water microbiology.
495
Microbiol. insights, 8, MBI. S33593.
496
26. Wang, G., Li, Q., Gao, X., Wang, X.C. 2018. Synergetic promotion of syntrophic methane
497
production from anaerobic digestion of complex organic wastes by biochar: Performance and
498
associated mechanisms. Bioresour. Technol., 250, 812-820.
499
27. Wang, W., Wang, S., Zhang, J., Hu, Z., Zhang, X., Sierra, J.M. 2016. Degradation kinetics of
500
pentachlorophenol and changes in anaerobic microbial community with different dosing
501
modes of co-substrate and zero-valent iron. Int. Biodeterior. Biodegr., 113, 126-133.
502
28. Xie, G.-J., Liu, B.-F., Wang, Q., Ding, J., Ren, N.-Q. 2016. Ultrasonic waste activated sludge
503
disintegration for recovering multiple nutrients for biofuel production. Water Res., 93, 56-64.
504
29. Yan, W., Shen, N., Xiao, Y., Chen, Y., Sun, F., Kumar Tyagi, V., Zhou, Y. 2017. The role of
505
conductive materials in the start-up period of thermophilic anaerobic system. Bioresour.
506
Technol., 239, 336-344.
507
22
30. Yan, W., Sun, F., Liu, J., Zhou, Y. 2018. Enhanced anaerobic phenol degradation by
508
conductive materials via EPS and microbial community alteration. Chem. Eng. J., 352, 1-9.
509
31. Yin, Q., Yang, S., Wang, Z., Xing, L., Wu, G. 2018. Clarifying electron transfer and
510
metagenomic analysis of microbial community in the methane production process with the
511
addition of ferroferric oxide. Chem. Eng. J., 333, 216-225.
512
32. Yu, H., Wang, Z., Wu, Z., Zhu, C. 2016. Enhanced waste activated sludge digestion using a
513
submerged anaerobic dynamic membrane bioreactor: performance, sludge characteristics and
514
microbial community. Sci. Rep., 6, 20111.
515
33. Zhang, L., Zhang, J., Loh, K.-C. 2018. Activated carbon enhanced anaerobic digestion of
516
food waste–Laboratory-scale and Pilot-scale operation. Waste Management, 75, 270-279.
517
34. Zhao, Z., Li, Y., Quan, X., Zhang, Y. 2017. Towards engineering application: Potential
518
mechanism for enhancing anaerobic digestion of complex organic waste with different types
519
of conductive materials. Water Res., 115, 266-277.
520
35. Zhao, Z., Zhang, Y., Yu, Q., Dang, Y., Li, Y., Quan, X. 2016. Communities stimulated with
521
ethanol to perform direct interspecies electron transfer for syntrophic metabolism of
522
propionate and butyrate. Water Res., 102, 475-84.
523
524
525
526
527
528
529
530
531
532
533
534
535
23
Figures
536
537
538
Figure 1. Methane generation profile during the long-term operation.
539
540
541
542
543
544
545
546
547
548
549
550
24
0 5 10 15 20 25 30 35
0
100
200
300
400
500
600
700
Methane ( mL/g VSS)
Day
0 g/L
0.125 g/L
0.25 g/L
0.5 g/L
1 g/L
2 g/L
551
Figure 2. Methanogenic profile at different PAC concentration.
552
553
554
555
556
557
558
559
560
561
562
563
564
25
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
Figure 3. The transformation of (a) organic compounds and (b)VFA during the long-term operation
582
26
583
584
0 5 10 15 20 25 30
0
500
1000
1500
2000
2500
3000 (b)(a)
Control
PAC
Control+BES
PAC+BES
TOC (mg/L)
Day
Day-0 Day-9 Day-18 Day-27
0
10
20
30
40
50
60
70
80
ETS ( ug · min-1·mL-1)
Control
PAC
Control + BES
PAC + BES
585
Figure 4. Dissolved TOC profile (a) and electron transport system activity (b) with the presence of
586
methanogenic inhibitor̶ BES.
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
27
604
605
606
607
608
609
Figure 5. Microbial community at the phylum level (a) and the distribution of methanogens (b).
610
611
612
613
614
615
616
617
618
619
620
621
622
623
28
624
Figure 6. Bacterial metabolism profile analyzed by PICRUSt: the relative abundance of (a) metabolism
625
and (b) hydrolytic enzymes related sequences.
626
627
628
629
630
631
29
Tables
632
633
Table 1. The methanogenic and volatile solids reduction performance of different groups.
634
Phase
Reactors
Methane
production
potential
(mL· g-1 VS)
R
(mL· g-1 VS · day-1)
Actual Methane
production
( mL/g VS)
Volatile Solids
reduction ratio
(%)
1
Control
348.27 ± 22.78
13.64 ± 0.75
279.51 ± 22.85
48.92 ± 3.61
PAC
697.32 ± 8.05
49.81 ± 1.59
676.71 ± 27.89
68.98 ± 4.13
2
Control
377.27 ± 10.81
19.50 ± 1.04
343.81 ± 27.42
53.71 ± 3.83
PAC
728.16 ± 25.52
41.02 ± 2.39
650.38 ± 33.48
67.81 ± 4.51
3
Control
371.11 ± 5.51
20.12 ± 0.67
348.58 ± 26.29
51.46 ± 3.47
PAC
731.98 ± 19.50
35.42 ± 1.46
646.09 ± 32.08
67.64 ± 4.83
4
Control
339.52 ± 13.84
23.92 ± 0.94
346.69 ± 14.98
52.18 ± 2.57
PAC
624.27 ± 19.42
34.76 ± 1.12
612.59 ± 22.14
64.64 ± 4.47
635
636
637
638
639
640
641
642
643
644
645
646
30
Table 2. The activity of key extracellular enzymes
647
Time
(day)
Protease (U/g VSS)
a-glucosidase (U/g VSS)
PAC
Control
PAC
Control
0
6.01 ± 0.27*
5.06 ± 0.32
0.16 ± 0.02
0.13 ± 0.03
6
15.97 ± 0.74*
10.19 ± 0.23
0.38 ± 0.03*
0.26 ± 0.01
11
15.79 ± 0.62*
11.21 ± 0.14
0.37 ± 0.03*
0.31 ± 0.02
17
18.11 ± 0.63*
12.97 ± 0.41
0.36 ± 0.02*
0.22 ± 0.01
30
15.44 ± 0.27*
9.52 ± 0.18
0.33 ± 0.04
0.29 ± 0.02
648
Note: Asterisks represent statistically significant difference (p < 0.05) between PAC and Control groups
649
on the same day.
650
651
652
653
654
655
