Acidic and alkaline shock pretreatment to enrich acidogenic biohydrogen producing mixed culture: long term synergetic evaluation of microbial inventory, dehydrogenase activity and bio-electro kinetics

Abstract

Comprehensive experiments were designed to evaluate the function of acid-shock (pH 3; orthophosphoric acid; 24 h) and alkaline-shock (pH 11; NaOH; 24 h) pretreatment methods for the selective enrichment of an acidogenic culture to enhance H₂ production efficiency of mixed anaerobic consortia. Long term (520 days) operation in suspended-batch mode bioreactors illustrated the relative efficiency and feasibility of redox pretreated cultures against an untreated parent culture in enhancing H₂ production. Relatively higher H₂ production was observed with an acid pretreated mixed culture (15.78 mol kg⁻¹ COD_R) over alkaline pretreated (9.8 mol kg⁻¹ COD_R) and untreated mixed cultures (3.31 mol kg⁻¹ COD_R). On the contrary, substrate degradation was higher with untreated culture (ξ_COD, 62.86%; substrate degradation rate (SDR), 1.10 kg COD_R/m³-day) and alkaline-shock pretreated mixed culture (ξ_COD, 59.93%; SDR, 1.22 kg COD_R/m³-day) compared to the acid-shock culture (ξ_COD, 53.4%; SDR, 0.705 kg COD_R/m³-day). Synergetics of microbial inventory, dehydrogenase activity and bio-electro kinetics in association with H₂ production and substrate degradation were also evaluated in detail throughout the operation. Acid pretreatment of the parent culture has resulted in a shift in the fermentation pathway towards acetic acid production, while alkaline pretreatment showed a mixed type fermentation (acetic, butyric, propnoic acids) similar to an untreated parent mixed culture. Dehydrogenase activity of the biocatalyst showed a significant improvement after applying acid pretreatment indicating the increased redox inter-conversion reactions leading to the higher proton gradient in the cell that resulted in higher H₂ production. The redox catalytic currents observed from the cyclic voltammograms (CV) and the output from the Tafel analysis also strongly supported the increased biocatalyst performance after pretreatment, especially at acid-shock. The shift in oxidative and reductive Tafel slopes towards a lower value after applying acid-shock treatment supports the redox inter-conversion reactions required for proton conservation. Microbial profiling revealed that the pretreatment method in the long term operation substantially affected the species composition of microbial communities. Dominance of Clostridia and Bacilli classes were observed in the pretreated culture and indicates their positive role in the H₂ production process. This study shows the feasibility of controlling microbial metabolic functions by repeated application of the pretreatment to the reactor native microflora (in situ) during operation whenever required to regain or modify the process performances.

---

1 Introduction

Biological hydrogen (H₂) production through dark fermentation of wastewater is considered to be an influential and sustainable technology that has garnered much attention in recent years.^1–7 Wastewater fermentation to H₂ is manifested by a diverse group of bacteria employing a series of complex biochemical reactions. Selection of an appropriate biocatalyst or inoculum significantly influences the start-up, overall efficiency and the stability of the H₂ production system. H₂ production from wastewater has more significance when a mixed culture is used as the biocatalyst because of operational flexibility, diverse biochemical functions, process stability and the possibility of using a wider range of substrates/feed-stocks^8–10 and restricts the need of sterile conditions.¹ Typical mixed anaerobic consortia affect the H₂ yield as it gets rapidly consumed by hydrogenotrophic or H₂ consuming methane-producing bacteria (MB) including homoacetogens, and sulfate reducing bacteria.

Regulating the metabolic pathway towards acidogenesis by simultaneously inhibiting methanogenesis to allow H₂ to become a metabolic end product can facilitate good H₂ production.^1,2,11 Pretreatment of the parent culture (biocatalyst) plays an important role in the selective enrichment of H₂-producing acidogenic bacteria (AB). It can facilitate a shift in the metabolic function of a biocatalyst towards acidogenesis from methanogenesis and simultaneously prevent the function of MB without effecting the activity of the H₂-producing bacteria.^6,12,13 Physiological differences between AB and H₂-consuming bacteria (MB) forms the main basis for the preparation of biocatalyst for H₂ production.^6,14 H₂-producing bacteria can form spores in adverse environmental conditions viz., high temperature, extreme redox conditions, etc., which is specifically absent in methanogens.¹⁴ For biocatalyst preparation, various pretreatment methods were reported.⁴ The degree of success mainly associates with the nature of seed/inoculums and the pretreatment method applied. Heat-shock pretreatment was reported to be the most suitable method to enrich H₂ producing bacteria,^15,16 while others reported the effectiveness of loading-shock,¹⁷ chloroform¹⁸ and 2-bromoethanesulfonic acid (BESA)¹¹ treatment methods. These contradicting results may be due to different fermentation conditions, such as temperature, pH, inoculum sources, substrate types, and concentration.¹⁹ Besides, many of these studies were conducted as a single batch cultivation and the long-term effects were not disclosed. Short- and long-term effects might also influence the process viability¹⁹ and process sustainability. Systematic and long-term evaluation of various biocatalyst pretreatment methods on biohydrogen production may provide a more meaningful understanding of the process for up-scaling.

Henceforth, an attempt was made in this communication to evaluate the long term influence of acid and alkaline shock inoculum pretreatment methods on both H₂ production and substrate degradation in comparison with an untreated parent anaerobic culture. Methanogenic activity is generally limited to a relatively narrow pH range (6.8 to 7.2), while most H₂ producing acidogens can grow over a broader pH range.^{1,3–5,8,20,21} Application of acid-shock pretreatment suppresses methanogenic activity of the cultures by simultaneously protecting the spore-forming bacteria.^4,13 Alkaline-shock pretreatment was also reported to suppress the growth of methanogens.²² To enumerate a comprehensive understanding of the metabolic process and shift as a function of pretreatment application, microbial inventory, dehydrogenase activity and bio-electro kinetics analysis were carried out.

2 Experimental Methodology

2.1 Anaerobic mixed culture and pretreatment

An anaerobic consortium sampled from a full scale anaerobic reactor treating composite wastewater was used as the parent inoculum. After sampling, the parent culture was sieved to separate the grit using a nylon filter and the resulting thick sludge (3.6 g VSS/l) was used for the inoculum preparation by applying acid-shock and alkaline-shock pretreatment methods separately. Parent anaerobic inoculum (140 ml) was subjected to acid-shock treatment by adjusting pH to 3 using orthophosphoric acid (30%). The inoculum was maintained under anaerobic conditions for 24 h. Subsequently, the pH of acid pretreated inoculum was re-adjusted to 6 using 1 M NaOH. Alkaline-shock treatment was applied by adjusting the parent inoculum pH to 11 using 1 M NaOH and maintained for 24 h under anaerobic conditions. Subsequently, alkaline pretreated inoculum was re-adjusted back to pH 6 with the help of orthophosphoric acid (30%). Both the pretreatment methods were performed at room temperature (28 ± 1 °C) by providing continuous mixing by a magnetic stirrer (100 rpm). The parent inoculum without applying any pretreatment was used and evaluated as the control.

2.2 Experimental design

Experimental variations were evaluated in three identical bioreactors under similar operating conditions to study the relative efficiency of redox pretreated inoculums based on H₂ production and substrate degradation in comparison with the control. A bench scale anaerobic sequencing batch reactor (AnSBR) with a suspended growth configuration was fabricated with a total/working volume of 1.2/0.84 l and a gas holding capacity of 0.36 l fabricated with borosilicate glass was used in this study. Among them, two reactors were operated with acid-shock and alkaline-shock treated inoculum, whereas the third one was operated with untreated inoculum which represents a control. Proper care was taken to make bioreactor leak proof with proper inlet and outlet arrangements. All the bioreactors were operated in sequencing/periodic-discontinuous batch mode with a total cycle period of 48 h [hydraulic retention time (HRT)] consisting of 20 min of FILL, 47 h of REACT (anaerobic), 20 min of SETTLE and 20 min of DECANT phases. The reactor was kept in suspension during the REACT phase. At the beginning of each cycle, immediately after withdrawal, a predefined volume (0.72 l) was fed to the reactor during the FILL phase. Initially during the start up phase, each reactor was fed with 140 ml of designate inoculum along with 720 ml of designed synthetic wastewater [DSW; glucose 3 g l⁻¹, NH₄Cl 0.5 g l⁻¹, KH₂PO₄ 0.25 g l⁻¹, K₂HPO₄ 0.25 g l⁻¹, MgCl₂ 0.3 g l⁻¹, CoCl₂ 25 mg l⁻¹, ZnCl₂ 11.5 mg l⁻¹, CuCl₂ 10.5 mg l⁻¹, CaCl₂ 5 mg l⁻¹, MnCl₂ 15 mg l⁻¹, NiSO₄ 16 mg l⁻¹, FeCl₃ 25 mg l⁻¹]. Prior to operation, the pH of the reactor contents was adjusted to pH 6.0 using orthophosphoric acid (30%). During operation the reaction mixture was subjected to continuous mixing by a magnetic stirrer at 100 rpm. All the experiments were performed at a constant organic loading rate (OLR) of 2.14 kg COD/m³ day at ambient temperature (29 ± 2 °C). During the operation an anaerobic microenvironment was maintained in the bioreactor by sparging with N₂ gas for 4 min after every feeding and sampling event. Biogas volume and composition was monitored by employing the water displacement method and electrochemical sensor/gas chromatography respectively. Once the H₂ production performance of the bioreactor decreased, native microflora present in the respective bioreactors was subjected to repeated pretreatments in situ to restrain the process performance.

2.3 Microbial community

The microbial community was analyzed by denaturing gradient electrophoresis (DGGE) of the polymerase chain reaction (PCR) amplified V3 region on 16S rDNA. Microbial cultures from three bioreactors were sampled at different time intervals for identification of their organisms and shifts during long term operation. Genomic DNA of the microbial mixed culture was extracted and purified by the phenol–chloroform method. Purified DNA was amplified by PCR according to methods described previously²³ using nucleotide sequences of the universal primers [primer 341F, 5′-AGG CCT AAC ACA TGC AAG TC-3′; primer 517R, 5′-ATT ACC GCG GCT GCT GG-3′; GC; clamp added to primer 63GC, 5′-CGC CCG CCG CGC GCG GCG GGC GGG GCG GGG GCA CGG GGG GAG GCC TAA CAC ATG CAA GTC-3′]. The GC-rich sequence attached to the 5′-end of forward primer prevents the PCR products from completely melting during separation.²⁴ DGGE was performed for the amplified PCR product using the DCode™ (Bio-Rad Laboratories Inc., USA) system. Denaturation was performed with an optimized polyacrylamide gradient (30–70%) at 60 V for 18 h. Gels were stained with ethidium bromide and images were captured by a molecular imager (G:BOX EF System, Syngene). The middle portions of the predominant selected DGGE bands were excised with a sterile razor blade. The excised gel was incubated in 50 μl of sterile distilled water overnight at 4 °C. 3 μl of eluted DNA was used as the template for PCR amplification. PCR products were sequenced (Eurofins Genomics India Pvt. Ltd.) with same primers used for the amplification. All the 16S rDNA partial sequences were aligned with the GenBank database using the BLASTN facility and were also tested for possible chimera formation with the CHECK CHIMERA program. These sequences were further aligned with the closest matches found in the GenBank database with the CLUSTALW function of the Molecular Evolutionary Genetics Analysis package (MEGA). Neighbor-joining phylogenetic trees were constructed with the MEGA version 4.1.

2.3.1 Nucleotide sequence accession numbers. Sequences were submitted to the Nucleotide Sequence Database to the Gene Bank public database under the accession numbers from HE648271to HE648295.

2.4 Bioprocess monitoring

Apart from the water displacement method, H₂ gas generated during the experiments was also estimated by a microprocessor based pre-calibrated H₂ sensor (ATMI GmbH Inc., Germany). The separation and quantitative determination of the composition of volatile fatty acids (VFA) was performed by high performance liquid chromatography [HPLC LC20AD; UV-VIS detector; C18 reverse phase column 250 × 4.6 mm and 5 μm particle size; flow rate 0.6 ml h⁻¹; wavelength 210 nm; mobile phase 40% of acetonitrile in 1 mM H₂SO₄ (pH 2.5 to 3.0); sample injection 20 μl]. The performance of the bioreactors, based on treatment efficiency, was monitored by assessing the chemical oxygen demand (COD-potassium dichromate closed refluxing titrimetric method), pH and VFA by standard methods.²⁵ Dehydrogenase enzyme activity of the pretreated and control biocatalyst was estimated by the 2,3,5-triphenyltetrazolium chloride (TTC) method.

2.4.1 Bio-electrochemical evaluation. The bio-electrochemical behaviour of the biocatalyst was periodically evaluated in each reactor by cyclic voltammetry (CV), linear sweep voltammetry (LSV) and chronoamperometry (CA) employing a potentiostat-galvanostat system (Autolab-PGSTAT12, Ecochemie). CV was performed to identify the change in redox catalytic currents during oxidation and reduction reactions of the biocatalyst. LSV and CA were performed to obtain the changes in oxidation reduction catalytic currents and maximum feasible currents from the biocatalyst with the function of pre-treatment method applied under varying and constant applied potentials. At defined time intervals 150 ml of suspension (biocatalyst) from the bioreactor was transferred to an electrochemical cell consisting of three electrode assemblies. CV was recorded at a potential ramp of applied voltage range between +0.5 and −0.5 V at different scan rates of 10, 20, 30, 50 and 100 mV s⁻¹. All the bio-electrochemical assays were carried out with a whole cell biocatalyst in DSW (as the electrolyte). CV profiles were evaluated by Tafel analysis using GPES (version 4.0) software based on slope and polarization resistance. LSV (polarizing between −0.5 and +0.5 V) was studied to evaluate oxidation and reduction potentials for all the three reactors during operation. LSV was extended to negative potential to exploit the redox phenomenon of anaerobic bacteria with the function of the bio-anode. CA analysis was performed for 900 s at a poised potential of 1.2 V at the end of the experiment to evaluate the sustainable current generation in all the three experimental variations. All the electrochemical assays were performed using platinum wire and graphite rod as the working and counter electrodes respectively against a Ag/AgCl(S) reference electrode.

3 Results and discussion

Three bioreactors were operated simultaneously to assess long term performance of selectively enriched anaerobic mixed consortia by acid-shock and alkaline-shock pretreatment methods in comparison with an untreated (control) parent culture on both H₂ production and wastewater treatment efficiency. The performance of the bioreactors was studied continuously for 260 cycles accounting for a total operation period of 520 days with each cycle being 48 h. After a drop in H₂ production, repeated application of the pretreatment [acid-shock (87th and 151th cycles) and alkaline-shock (78th and 154th cycles)] was applied to the respective native microflora in situ to the bioreactors to restrain the process performance.

3.1 Biohydrogen production

Experimental data documented an apparent variation in the process performance and bio-electrochemical behavior of the biocatalyst with the function of the pretreatment method applied. Both the redox pretreatment strategies applied documented a relatively efficient H₂ production compared to the untreated parent inoculum (control) (Fig. 1). A distinct metabolic pattern was observed with the function of an individual pretreated biocatalyst signifying the change in the behavior of a selectively enriched mixed culture compared to the control operation. Variation in H₂ production observed might be due to the alteration of the microbial composition induced by the pretreatment methods. Acid-shock pretreated culture showed comparatively higher and sustained H₂ production followed by alkaline-shock and an untreated culture (control). During the initial five cycles of operation all the cultures showed more or less similar H₂ production. The untreated culture recorded relatively higher H₂ production (2.8 mmol) up to the 16th cycle of operation followed by a drop (21st cycle) and remained almost consistent until the 41st cycle (∼ 1.8 mmol). A similar performance was maintained till the end of experiments with an intermittent increment in H₂ production (2.9 ± 0.2 mmol) for a few cycles followed by the stabilization at lower values (∼ 1.8 mmol). This might be due to the prevailing function of methanogens in the parent culture catalyzing H₂ conversion to CH₄.

Fig. 1 Biohydrogen production of acid-shock, alkaline-shock and untreated (control) anaerobic inoculum with the function of reactor operation time (↑ indicated application of pretreatment).

In the case of acid-shock pretreated culture, a gradual improvement in H₂ production was observed up to the 6th cycle (5.8 mmol), which remained more or less the same up to the 14th cycle (6.1 mmol) prior to approaching a maximum at the 61st/68th cycle (15.84 mmol). Further operation showed almost stabilized performance up to the 70th cycle (14.82 ± 1.2 mmol). Later, a gradual drop in H₂ production was observed which approached a lower value (1.23 mmol) at 86th cycle. Acid-shock treatment suppresses the growth of MB resulting in enhancing metabolic activities of acidogenic mixed culture facilitating higher H₂ production compared to the control operation. The observed drop in H₂ production after a certain period of operation might be attributed due to long term operation under uncontrolled redox conditions which resulted in regaining the methanogenic activity by the reactor native mixed culture. Repeated loading of wastewater for a number of cycles under unsterilized conditions might also be the probable reason for regaining the methanogenic activity due to the adding up of methanogens associated with wastewater. At this stage of operation, the reactor was again subjected to acid-shock pretreatment to retain the performance, which showed a rapid improvement in H₂ production within two cycles of operation (88 and 89 cycles; 7.5 to 9.1 mmol) and then gradually increased with each additional cycle event and approached a maximum value during the 108th cycle (15.82 mmol). This phase was continued until the 111th cycle (14.6 ± 0.7 mmol) and dropped to 10.1 mmol during the 112th cycle operation and remained the same until the 118th cycle (10.26 mmol). Again a peak (14.1 mmol) prior to the drop at the 124th cycle (5.94 mmol) was recorded. A gradual increment in H₂ production was observed after the 124th cycle until the 149th cycle (13.32 mmol) and significantly dropped to its lowest value at the 154th cycle (0.04 mmol). Again, the native microflora was subjected to acid-shock pretreatment, which also resulted in the rapid improvement in H₂ production from the 161st cycle (5.58 mmol) to the 170th cycle (13.86 mmol) and then decreased to 8.86 mmol (181st cycle) followed by a gradual increment reaching a maximum value at the 195th cycle (15.86 mmol). Stable reactor performance was observed from the 201st cycle onwards (12.78 mmol) until the end of the experiment. Acid-shock culture has taken about 140 days for the regaining methanogenic activity after first and second pretreatment applications. However, after the third pretreatment application, the H₂ production remained more or less stable even after 200 days of operation.

Contrary to the acid-shock culture, relatively fewer fluctuations and comparatively less H₂ production was observed with the alkaline-shock pretreated biocatalyst in spite of relatively lower yields. The alkaline-shock treated culture showed a gradual improvement in H₂ production and approached a maximum during the 68th cycle (8.64 mmol). Later, a drop in H₂ production was observed and approached a minimum value (78th cycle; 0.9 mmol). The drop in H₂ production might be due to the higher methanogenic activity. To retain the performance, the reactor was again subjected to pretreatment, which resulted in a rapid improvement in H₂ production within ten cycles of operation (79 to 89 cycles; 1.8 to 5.12 mmol) and then gradually increased with each additional cycle until the 120th cycle (7.43 mmol) and dropped (154th cycle, 0.04 mmol). Here the reactor was again subjected to the alkaline-shock pretreatment, wherein, the H₂ production showed a gradual increase up to the 205th cycle (7.4 mmol) later with a gradual drop (213th cycle, 4.45 mmol).

Untreated mixed culture generally has higher bacterial population with a wide variety of biochemical functions facilitating diverse metabolic activities. On the contrary, pretreatment facilitates the selective enrichment of the bacterial population leading to less diversity in their biochemical functions specific towards acidogenesis.^7,14,26–28 The acid-pretreated culture showed a relatively higher H₂ production than that of the culture pretreated with alkaline-shock as well as the untreated culture. Methanogens growth can be repressed by controlling the cultivation conditions at low pH.^8,11,29,30 AB can survive in a highly acidic environment. The acid-pretreated mixed culture reported significantly higher H₂ production than that of the culture pretreated with heat-shock, freezing and thawing, alkaline-shock as well as the control.³¹ Alkaline pretreatment performed in the pH range between 8.5 and 12.0 showed the suppression of methanogens.²² Alkaline pretreatment cannot completely suppress the methanogenic activity.^1,14,28 Alkaline-tolerant H₂ producing bacteria showed the maximum H₂ yield at an initial pH of 11.0.³² Methanogenesis could also be eliminated by maintaining short retention times (2–10 h) during reactor operation as H₂-producing bacteria grow faster than the methanogens.^7,27,29 Untreated culture showed relatively poor H₂ yields compared to the alkaline-pretreated mixed culture. Experimental data showed the feasibility of controlling microbial metabolic function by repeated application of pretreatment to the reactor native microflora (in situ) during operation when ever required to modify or regain the performances.

3.2 Substrate degradation

Substrate degradation (based on COD removal) showed distinct variation with the function of individual pretreated biocatalysts (Fig. 2). The pattern of substrate degradation was almost contrary to the H₂ production pattern. Substrate degradation efficiency was higher with the alkaline pretreated biocatalyst over acid-pretreated biocatalyst. The untreated culture showed relatively higher substrate removal efficiency (ξ_COD, 62.86%; SDR, 1.10 kg COD_R/m³- day) followed by alkaline-shock (ξ_COD, 59.93%; SDR, 1.22 kg COD_R/m³-day) culture (Fig. S1, ESI†). The untreated mixed culture has a higher bacterial population with a wide variety of biochemical functions facilitating diverse metabolic activities while pretreatment facilitates selective enrichment of bacterial population leading to less diversity specific towards acidogenesis. The acid-shock biocatalyst documented the lowest substrate degradation (ξ_COD, 53.4%; SDR, 0.705 kg COD_R/m³-day) among the other two cultures. All the three cultures showed a rapid increase in the substrate degradation during the initial 30 cycles of operation, however, the efficiency varied with the type of pretreatment applied. The untreated culture showed a gradual increase with each cycle and reached a maximum during the 39th cycle and continued with almost similar efficiency (58 ± 4%; 1.10 ± 0.8 kg COD_R/m³-day). This might be due to the prevalence of higher methanogenic activity. The acid pretreated culture showed a gradual increment in substrate degradation till the 53rd cycle (51.56%) followed by a drop that stabilized for few cycles and dropped to a lower value prior to the increment at the 75th cycle (18.86%). After applying an additional pretreatment, a rapid drop in substrate degradation was observed till the 86th cycle which increased (33.2 ± 0.6%) followed with a gradual increment till the 156th cycle followed by a marked drop and gradual increment till the 170th cycle. Similarly, after applying the pretreatment for the third time, a marked reduction in substrate degradation was observed till the 206th cycle (25.46%) and then increased (222nd cycle; 48.86%). Further operation showed a series of drops and increments in the substrate degradation till the end of cycle operation. Improvement in substrate degradation and corresponding reduction in H₂ production with time might be attributed to the methanogenic activity. Repeated feeding of wastewater over a number of cycles under unsterilized conditions might be the probable reason for the regaining of the methanogenic activity. The alkaline pretreated biocatalyst showed a gradual improvement initially and then stabilized at the 79th cycle (ξ_COD, 55.16%; SDR, 1.18 kg COD_R/m³-day). An almost similar trend was observed as H₂ production after first, second and third pretreatments application. The alkaline-shock pretreated culture regained performance by suppression of the methanogenic bacteria only for few cycles of operation. On the contrary, acid-shock culture performance persisted for long operation periods. Particularly, H₂ production decreased along with the increment in substrate removal efficiency after few cycles. Higher substrate degradation and early regaining of methanogenic activity observed in this case compared to acid-shock culture might be due to the lower efficacy of the alkaline-shock pretreatment method in suppressing the methanogenic bacteria. However, the time taken for the regaining was almost the same after the first and second pretreatments. Specific H₂ yield (SHY) produces a similar trend, like H₂ production where higher values were registered with acid-shock (15.78 mol kg⁻¹ COD_R) followed by alkaline-shock (7.41 mol kg⁻¹ COD_R) and control (3.31 mol kg⁻¹ COD_R) operations (Fig. 2).

Fig. 2 Substrate degradation pattern as substrate degradation rate and specific hydrogen yield (SHY) of acid-shock, alkaline-shock and untreated (control) anaerobic inoculum with the function of reactor operation time (↑ indicated application of pretreatment).

3.3 Acid metabolites

Fermentative H₂ production associates with metabolic conversion of organic substrate to acid metabolic intermediates i.e. volatile fatty acids (VFA). Marked variation in VFA production was recorded with the function of the pretreatment method applied (Fig. 3). Comparatively higher VFA production was observed with the acid pretreated culture followed by alkaline pretreated culture and the control, which correlated well with the H₂ production pattern. Operation with untreated culture showed a gradual increment in VFA generation up to 160 cycles (876–1034 mg l⁻¹) and remained stable up to 193 cycles. Further operation showed a drop in production (875 mg l⁻¹; 195th cycle) and remained more or less stable (925 ± 40 mg l⁻¹) till the 243rd cycle. Rapid improvement was observed from the 243rd cycle (1322 mg l⁻¹) to the 259th cycle (1432 mg l⁻¹). Acid and alkaline pretreatments showed almost similar pattern of VFA generation throughout the operation. However the acid pretreated biocatalyst showed higher VFA generation over the alkaline pretreated biocatalyst. In both the conditions, VFA production showed a rapid increment till 20 cycles (acid, 1365 mg l⁻¹; alkaline, 1298 mg l⁻¹) and it was almost stable with acid pretreatment (1436 mg l⁻¹) by the end of the 77th cycle and at the 68th cycle (1276 mg l⁻¹) in the case of alkaline pretreatment. Further operation showed a marked drop in VFA concentration in both the cases till the 82nd cycle (832 mg l⁻¹) and 80th cycle (723 mg l⁻¹) in the case of acid and alkaline biocatalysts, respectively. This event was strongly supported by the rapid drop in H₂ production at respective cycles. After the application of repetitive pretreatment, a rapid increment in VFA generation was observed till the 89th cycle (acid, 1465 mg l⁻¹; alkaline, 1376 mg l⁻¹) and stabilized till the 164th (1466 mg l⁻¹; acid) and 156th cycles (1302 mg l⁻¹; alkaline). A drop in VFA generation was observed thereafter in both the cases which strongly support the prevailed methanogenic activity associated with lower H₂ production. However, the drop in VFA generation was lower after the second pretreatment compared to the first pretreatment. Re-treatment showed again a gradual increment in VFA production followed by stabilization (acid, 1466 mg l⁻¹; alkaline, 1318 mg l⁻¹) till the end of the experiment.

Fig. 3 VFA and pH variation of acid-shock, alkaline-shock and untreated (control) anaerobic inoculum with the function of reactor operation time (↑ indicated application of pretreatment).

The VFA profile reflects the changes occurring in the metabolic process and helps to provide information to facilitate favourable conditions for H₂ production. VFA composition by chromatography revealed the presence of acetic acid, butyric acid and propionic acid (Table 1). Among them, acetic acid was a major metabolite formed during H₂ production in both pretreatment methods compared with untreated culture. Acid pretreated culture showed acetic acid (62.85%) as a major metabolite followed by butyric acid (23.51%) and propionic acid (34.05%). Maximum acetic acid was observed in both pretreated cultures during the 108th cycle (acid, 62.85%; alkaline, 47.94%). On the contrary, control operation showed the presence of acetic acid only during the initial phase of operation (5th cycle, 20.41%), which strongly supports the observed lower H₂ production compared to acid and alkaline pretreated cultures. On the other hand, the presence of propionic acid was higher with the control (195th cycle, 64.27%) throughout the operation followed by alkaline (195th cycle, 30.79%) and acid (5th cycle, 27.27%) pretreated cultures. The presence of propionic acid was only observed during the initial phase of operation in the case of acid pretreated biocatalyst, which strongly supports the observed higher H₂ production. Butyric acid was intermittent to the acetic and propionic acids where its concentration was higher in acid and alkaline pretreated cultures compared to the control operation.

Table 1 Soluble metabolic product profiles during the pretreatment experiment with respect to cycle operation

Cycles	Pretreatment method	Acetic acid (%)	Butyric acid (%)	Propionic acid (%)
2	Control	20.41	—	24.38
	Acid	42.5	25.42	27.27
	Alkaline	42.34	12.61	29.04
37	Control	—	41.29	58.7
	Acid	48.04	36.49	15.45
	Alkaline	45.88	33.51	20.59
78	Control	17.9	54.4	—
	Acid	25.49	34.05	8.01
	Alkaline	30.56	48.44	12.98
168	Control	—	26.01	51.07
	Acid	62.85	37.59	1.55
	Alkaline	47.94	25.87	26.24
221	Control	—	4.15	63.54
	Acid	40.39	59.60	—
	Alkaline	44.51	55.51	—
224	Control	––	26.17	64.27
	Acid	54.74	15.29	20.26
	Alkaline	22.82	45.22	30.79

---

Accumulation of the volatile acids was observed in all the experimental variations studied, which resulted in a change in the system redox conditions (Fig. 3). However, the variation in pH was higher with the acid-pretreated operation compared to alkaline and control operations. The functional role of acidogenic mixed culture might be the reason for this and correlated well with the observed H₂ production. The differences in VFA production profile with different pretreated operations suggested that the metabolic pathways of the biocatalyst might have shifted and henceforth, these shifts were likely to impact on the H₂ consumption and production in the reactor. In general, the production of acetic acid and butyric acid favours H₂ production, while propionic acid consumes H₂ gas.³³ The presence of a higher concentration of propionic acid was not considered to be favourable for effective H₂ production. Generation of higher concentrations of acetic acid followed by butyric acid in experiments performed with acid pretreated cultures signified a favourable microenvironment for acidogenic activity. Distribution of soluble acid metabolites also suggested the dominance of an acetogenic metabolic pathway which was considered as an optimum microenvironment for effective H₂ generation. The trend in the formation of soluble metabolites suggested the effectiveness of the applied pretreatment method in manifesting effective acidogenesis particularly with the acid-shock method compared to alkaline-shock and control operations.

3.4 Dehydrogenase activity

Redox reactions for the inter-conversion of metabolites and transferring protons (H⁺) between metabolic intermediates are catalyzed by the dehydrogenase (DH) enzyme using several mediators (NAD⁺, FAD⁺, etc.). These redox mediators are capable of carrying H⁺ and electrons (e⁻) and are known as energy carriers as they are involved in biological energy generation (ATP). DH facilitates the availability of H⁺ through redox reactions to make H₂.³⁴ The acid pretreated biocatalyst showed higher DH activity followed by the alkaline and control operations in correlating with H₂ production (Fig. 4). A gradual increment in DH activity was observed in all the three cases till the 36th cycle from where the control operation showed a drop, while pretreated operations showed a rapid increase. Except in a few cycles, the control operation showed almost stable and relatively low DH activity (3.34 μg ml⁻¹) throughout the operation associated with lower H₂ production. In the case of acid pretreatment, DH activity registered a higher activity during the 74th cycle (9.76 μg ml⁻¹) followed by a decrement at the 85th cycle (3.12 μg ml⁻¹). After applying the pretreatment to retain the performance, improvement in DH activity was observed depicting a maximum value during the 130th cycle (10.87 μg ml⁻¹) followed by a decrement till the 154th cycle (6.23 μg ml⁻¹). After application of a third pretreatment a rapid increment in DH activity was also observed till the 161st cycle (10.67 μg ml⁻¹) followed by a drop till the 175th cycle and continued at an almost stable value (6.4 ± 0.9 μg ml⁻¹). Compared to the acid pretreated culture, the alkaline operation showed a relatively low DH activity with less fluctuation throughout the operation. This corroborates well with the observed H₂ production profiles of acid and alkaline treatment operations. Overall, the DH activity showed decrements in every additional pretreatment indicating the decrement in H⁺ transfer ability between the intermediates. Higher DH activity was observed during the 51st cycle (7.61 μg ml⁻¹) in the case of alkaline pretreatment operation.

Fig. 4 Dehydrogenase activity of acid-shock, alkaline-shock and untreated (control) anaerobic inoculum with the function of reactor operation time (↑ indicated application of pretreatment).

3.5 Microbial inventory analysis

Most studies on the pretreatment effects have focused on H₂ production, substrate utilization and product formation. There are very few studies on the composition change of microbial community responsible for affecting H₂ production after different pretreatments. In addition, knowledge of the microbial community structure is critical for improving hydrogen production by mixed culture fermentation.⁴ The microbial population and its variation with the function of operating time after application of pretreatment methods are important to visualize the metabolic changes with respect to the untreated culture. The microbial community analysis of long term operated selectively enriched mixed cultures by acid-shock and alkaline-shock pretreatment methods in comparison with untreated parent cultures was evaluated by DGGE analysis at V3 region of 16S rDNA with the function of operation time. The stained gel image of DGGE demonstrated a clear banding pattern (12 samples; 0, 65, 170, 210 cycles for each operation) of the microbial community structure. Distinct shifts in the microbial population were observed with the function of experimental variations (Fig. 5). Some bands migrated at a similar position suggesting the existence of a common bacterial population. Bands that show different migration behavior were sequenced and phylogenetically classified (Fig. 6). In total, 25 dominant true operational taxonomic units (OTUs) with 7 distinct classes were identified (Table 2). The results showed a close correlation between H₂ production with the bacterial community structure. Major bands were phylogenetically related to classes Clostridia, Bacilli, Bacteroidia, uncultured bacteria, β, γ, δ, ε-proteobacteria. The uncultured/parent culture showed dominance of the proteobacteria class followed by firmicutes, uncultured bacteria, actionobacteria and bacteroidia. The microbial community showed that dominance of non-H₂ producing microorganisms accounted for a major part of the population. Organisms belong to the class of proteobacteria (34.6%) viz., Janthinobacterium sp, Alcaligenes faecalis, Geobacteraceae bacterium, Alpha proteobacterium, delta proteobacterium were found to be dominant in the parent culture. The second dominant fraction of the microflora belongs to the phylum firmicutes (23.07%; uncultured firmicutes, uncultured Bacillus sp, Lactococcus lactis, Staphylococcus sp). A small fraction of the microbial community were related to Bacteroidia (7.68%, uncultured Bacteroidetes), these are non-spore forming obligate, gram-negative anaerobes, actinobacteria (11.52%, Corynebacterium pseudotuberculosis) and uncultured bacteria (23.07%; IICTSVMH22, IICTSVMH23) were also observed.

Fig. 5 Variation of DGGE profiles with different pretreatment methods [Lane (acid shock) A1, A2, A3, A4-1.Lysinibacillus xylanilyticus; 2.Bacillus cereus; 3.Clostridium cellulosi; 4.delta proteobacterium; 5.Bacillusthuringiensis; 6.Uncultured Ruminococcaceae; 7.Enterobacter sp; 8. Uncultured Geobacteraceae bacterium: Lane (control) C1, C2, C3, C4- 1. Lactococcus lactis; 2. Janthinobacterium sp; 3. Geobacteraceae bacterium; 4. Lysinibacillus xylanilyticus; 5.Uncultured Ruminococcaceae; 6.Alpha proteobacterium; 7. Delta proteobacterium; 8. Uncultured Firmicutes; 9. Staphylococcus arlettae; 10. Alcaligenesfaecalis; 11. Uncultured cyanobacterium; 12.Bacillus cereus; 13. Corynebacterium pseudotuberculosis; 14. Staphylococcus sp; 15.Uncultured Bacillus sp.: Lane (alkaline shock) Al1, AL2, AL3, AL4- 7. Uncultured cyanobacterium; 8. Bacillus boroniphilus; 9.Staphylococcus sp; 10.Lysinibacillus xylanilyticus; 11. Delta proteobacterium; 12.Lactococcus lactis; 13.Alpha proteobacterium.

Fig. 6 Neighbor-joining tree constructed using Mega 4.0 showing phylogenetic relationships of 16S rDNA sequences from closely related sequences from Gen Bank; phylogenetic community comparison with the function of experimental variations studied.

Table 2 Phylogenetic sequence affiliation and similarity to the closest relative of amplified 16S rDNA sequence excised from DGGE

DGGE band	Closest match	Acid				Alkaline				Control				Identity (%)	Accession no.	Phylogenetic affiliation (Class)	Phylogentic affiliation (Phylum)
		BP	65	170	210	BP	65	170	210	BP	65	170	210
IICTSVMH1	Clostridium sp HM801879.1	✓	✓	–	✓	✓	✓	✓	–	✓	–	✓		93	HE648271	Clostridia	Firmicutes
IICTSVMH2	Uncultured Clostridiumu sp FR836437.1	✓	–	✓	✓	–	–	–	–	–	–	–	–	92	HE648272	Clostridia
IICTSVMH3	Clostridium cellulosi FJ465164.1	–	✓	✓	✓	–	–	–	–	–	–	–	–	93	HE648273	Clostridia
IICTSVMH4	Uncultured Ruminococcaceae GU939475.1	✓	✓	✓	✓	–	–	–	–	✓	–	–	–	92	HE648274	Clostridia
IICTSVMH5	Bacillus cereus JN717163.1	✓	✓	✓	✓	✓	✓	✓	–	✓	–	–	–	100	HE648275	Bacilli
IICTSVMH6	Bacillus thuringiensis HQ873480.1	✓	✓	✓	✓	✓	✓	–	–	–	–	–	–	100	HE648276	Bacilli
IICTSVMH7	Staphylococcus sp JN695717.1	–	–	–	–	–	–	–		✓		✓	✓	100	HE648277	Bacilli
IICTSVMH8	Staphylococcus arlettae HQ234350.1	✓	✓	–	–	✓	✓	–	✓	✓	✓	✓	✓	99	HE648278	Bacilli
IICTSVMH9	Lactococcus lactis JN194197.1	–	–	–	–	✓	✓	✓	✓	✓		✓	✓	99	HE648279	Bacilli
IICTSVMH10	Lysinibacillus xylanilyticus JN644616.1	✓	✓	✓	✓	✓	✓	✓	✓	✓	✓	–	–	100	HE648280	Bacilli
IICTSVMH11	Staphylococcus sp HM171269.1	–	–	–	–	✓	–	✓	✓	–	–	–	–	98	HE648281	Bacilli
IICTSVMH12	Staphylococcus sp.JN695717.1	–	–	–	–	–	–	–	–	–	–	–	–	100	HE648282	Bacilli
IICTSVMH13	Uncultured Bacillus sp FR751013.1	–	–	–	–	✓	✓	✓	–	–	–	–	–	90	HE648283	Bacilli
IICTSVMH14	Uncultured Bacillus sp EU360150.1	✓	✓	–	✓	✓	✓	–	✓	–	–	–	–	96	HE648284	Bacilli
IICTSVMH15	Uncultured Bacillus sp EU360150.1	✓	✓	–	✓	–	–	–	–	✓	✓	✓	✓	96	HE648285	Bacilli
IICTSVMH16	Uncultured Firmicutes CU923711.1	✓	✓	–	✓	✓	–	–	–	✓	–	✓	✓	92	HE648286	Clostridia/Bacilli
IICTSVMH17	Janthinobacterium sp JN032578.1	✓	–	–	–	✓	✓	✓	✓	✓	–	✓	✓	79	HE648287	Beta Proteobacteria	Proteobacteria
IICTSVMH18	Alcaligenes faecalis AJ748207.1	✓	–	–	–	✓	–	–	–	✓	–	–	–	91	HE648288	Beta Proteobacteria
IICTSVMH19	Enterobacter sp FJ608232.1	✓	–	✓	✓	–	–	–	–	✓	–	–	–	86	HE648289	Gammaproteobacteria
IICTSVMH20	Uncultured Geobacteraceae bacterium EF668817.1	✓	✓	✓	✓	✓	✓		✓	✓	–	–	✓	96	HE648290	Deltaproteobacteria
IICTSVMH21	Alpha proteobacterium HM163276.1	✓	✓	✓	–	✓	–	–	–	✓	✓	✓	–	96	HE648291	Alphaproteobacteria
IICTSVMH22	Uncultured bacterium FJ162550.1	✓	–	–	–	✓	–	–	–	✓	–	–	–	95	HE648292	–	–
IICTSVMH23	Uncultured bacterium HE589881.1		–	–	–	✓	–	–	–	✓	–	–	–	98	HE648293	–	–
IICTSVMH24	Uncultured Bacteroidetes CU917906.1	✓	–	–	–	✓	–	–	–		–	–	–	82	HE648294	Bacteroidia	Bacteroidetes
IICTSVMH25	Corynebacterium pseudotuberculosis GU818733.1	✓	–	–	–		–	–	–	✓	✓	–	–	93	HE648296	–	A̲c̲t̲i̲n̲o̲b̲a̲c̲t̲e̲r̲i̲a̲

---

The acid-pretreated culture showed dominance of firmicutes followed by proteobacteria, uncultured bacterium (IICTSVMH22, IICTSVMH23), bacteroidia and actinobacteria. Firmicutes are composed mainly of the Clostridia class (Clostridium cellulosi, Uncultured Ruminococcaceae) and Bacilli class (Bacillus cereus, Lysinibacillus xylanilyticus, Bacillus thuringiensis). The abundance of firmicutes was found to increase with every additional pretreatment event or with operating time [34.78% (65th cycle); 44.78% (170th cycle); 53.6% (210th cycle)]. The abundance in proteobacteria class (Delta proteobacterium, Enterobactre sp, Geobacteraceae bacterium) also increased with time [21.73% (65th cycle); 21.73% (170th cycle); 34.61% (210th cycle)]. Uncultured bacterium species composition was decreased with time [13.04% (65th cycle); 7.68% (170th cycle); 7.68% (210th cycle)]. The bacteroidia class vanished in the last cycles [8.63% (65th cycle); 8.63% (170th cycle); 0% (210th cycle)]. Actinobacteria species composition also disappeared in the last cycles [4.84% (65th cycle); 4.84% (170th cycle); 0% (210th cycle)]. Firmicutes became the dominant phylum after application of acid-shock pretreatment with Clostridium sp., Bacillus sp. and Enterobacter sp. Firmicutes are reported to be effective H₂ producers^35,36 and have the capability to grow in acidogenic microenvironment forming endo-spores in adverse conditions,^30,37,38 which substantiate its presence in the acidogenic reactors. Clostridia were found to have iron dependent hydrogenase ([FeFe]-hydrogenase) as the dominant functional gene, which has higher activity than [NiFe]-hydrogenase.³⁶

The alkaline-pretreated culture showed dominance of Proteobacteria class followed by Firmicutes, Uncultured bacteria and Actinobacteria/Bacteroidia which was more or less similar to the parent inoculum with additional presence of Clostridium (Uncultured Ruminococcaceae) and Bacillus (Lysinibacillus xylanilyticus, Uncultured Ruminococcaceae, Bacillus boroniphilus, Bacillus cereus, Bacillus thuringiensis, Uncultured Bacillus sp.) species. Proteobacteria is composed mainly of α-proteobacterium, uncultured Geobacteraceae bacterium, Alcaligenes faecalis and Enterobacter sp. The abundance of proteobacteria was found to increase with every additional pretreatment event or with operating time [34.61% (65th cycle); 32.61% (170th cycle); 32.61% (210th cycle)]. The fraction of firmicute class decreased with the application of alkaline pretreatment. Bacillus thuringiensis, Clostridium sp., Bacillus cereus and Uncultured Bacillus species were found to disappear with operation time [30.76% (65th cycle); 27.76% (170th cycle); 27.76% (210th cycle)]. Uncultured bacteria, Bacteroidia and Actinobacteria microbial community profile did not show much change till the end of the operation. In contrast, Lactobacillus sp. is dominant in the microbial community pretreated with alkaline-shock. Lactobacillus sp. is known to be the common co-existing bacteria in the fermentation process and has an inhibitory effect on H₂ production by secreting bacteriocins, which have deleterious effects on other bacteria.³⁹ In addition, the large amount of CO₂ produced by Lactobacillus sp. is also an adverse condition.⁴⁰

The microbial inventory/composition gradually changed with repeated applications of pretreatment events. Species shifts were well correlated with both H₂ production and substrate degradation with the function of biocatalytic nature. Bacterial species Corynebacterium pseudotuberculosis, Alcaligenes faecalis, Staphylococcus sp., Alpha proteobacterium and Janthinobacterium sp., existing with the acid pretreated culture during initial phase operations showed a gradual disappearance with time. Phylum Firmicutes (Clostridia and Bacilli) dominated in the microbial community of pretreated cultures while dominance of Clostridia was found to be dominant in acid pretreated cultures. Generally, Firmicutes are wide spread in the environment mostly with a diverse metabolic activity. It is evident from the microbial inventory data that the pretreatment method substantially affects the species composition and the composition of microbial community influenced the H₂ production.

3.6 Bio-electrochemical analysis

The bio-electrochemical activity of the biocatalyst in whole cell form was evaluated using cyclic voltammetry (CV) during the H₂ production process with the function of pretreatment methods applied. Protons (H⁺) and electrons (e⁻) generated during the substrate metabolism capitulates the fundamental information pertaining to the nature and the rate of metabolic activities.⁸ CV is a standard electrochemistry tool exploited more recently to study and characterize the electron (e⁻) transfer interactions between microorganisms or microbial metabolites by applying an external potential.^8,41 Voltammogram profiles visualized significant variation in the bio-electrochemical behavior with the function of pretreatment method applied (Fig. 7). Reduction current (catalytic) was relatively higher compared to the oxidation current (catalytic) with untreated culture (control) operation, indicating the feasible microenvironment for the protons (H⁺) reduction with the terminal electron accepting process (TEAP) of H₂-dependant methanogenesis. After applying the pretreatment, an increase in oxidation currents associated with the decrease in reduction currents was observed in both acid and alkaline shock treatments, indicating the marked change in metabolic activities of biocatalyst towards redox inter-conversion reactions where protons gets shuttled between intermediates and make them available for the H₂ production. The acid-shock culture showed comparatively higher oxidation current (1.91 μA) than the alkaline-shock (1.24 μA) and control (0.84 μA) operations. On the contrary, control operation showed higher reduction current (−3.96 μA) followed by acid-shock (−0.86 μA) and alkaline-shock (−0.82 μA) pretreatment cultures.

Fig. 7 Cyclic voltammograms recorded during the operation with acid-shock and alkaline-shock pretreated cultures in comparison with the untreated (control) operation.

The redox catalytic currents showed variation with the function of operating time also indicating the change in biocatalyst behavior with the change in metabolic activities (Fig. 7). The control operation showed a gradual increment in the reduction current with each additional cycle and depicted a maximum during the 121st cycle (−3.96 μA) and dropped at the 168th cycle (−0.48 μA), which again increased to its maximum value during the 224th cycle (−2.96 μA). However, the oxidation current was almost constant throughout the operation (0.75 ± 0.1) indicating the strong reductive behavior of the biocatalyst. Lower DH activity of the biocatalyst observed throughout the control operation strongly supports the relatively fewer inter-conversion reactions leading to lower H⁺ shuttling between the intermediates resulting in lower H₂ production.⁸ The formation of lower VFA quantities during this operation also supports the favorable environment for a strong reduction microenvironment that maintains lower proton concentration in the cell due to their consumption during methanogenesis.

Interestingly, oxidation currents dominated over the corresponding reduction currents after application of an acid-shock pretreatment indicating simultaneous redox reactions that help in maintaining the proton gradient in the cell which are precursor molecules for the H₂ production. Moreover, higher proton concentration in the cell also facilitates the activation of hydrogenase enzymes, which function towards reduction of these protons to H₂. Both the oxidation and reduction currents increased during initial cycles (2nd cycle, 0.86 μA, −0.41 μA; 37th cycle, 1.92 μA, −0.78 μA) and dropped at the 78th cycle (FS, 1.21 μA; RS, −0.54 μA), which is supported by the drop in H₂ production. After applying repetitive acid-shock pretreatment, the redox currents showed increment with each additional cycle and depicted maximum during 121 cycle (FS, 1.96 μA; RS, −0.92 μA) which is also supported by an increase in H₂ production. The redox currents dropped to a lower value during the 168th cycle (FS, 0.39 μA; RS, −0.26 μA), where a drop in H₂ production was again observed. The redox currents increased again after applying the pretreatment and reached a higher value at the 224th cycle (FS, 0.98 μA; RS, −1.5 μA). However, marginal variations in the oxidation and reduction currents were observed at this stage where reduction current was relatively higher than the oxidation unlike previous cycles. The H₂ production was almost stable but relatively lower than previous cycles during this phase (around the 224th cycle) supporting the higher reductive behavior of the biocatalyst. The DH activity also decreased during this cycle, supporting the lower inter-conversion reactions. Higher VFA production and accumulation after each pretreatment and a gradual decrease with time also supported the observed redox behavior of the biocatalyst.

The alkaline-shock pretreated culture also showed a marginally higher oxidation current compared to the control, which supports the observed H₂ production. Similar to the acid-shock treatment, the oxidation current was dominant during the initial cycles and with time almost similar redox currents were observed. Higher redox currents were recorded during the 37th (FS, 1.2 μA; RS, −0.42 μA) cycle which remained almost the same until the 78th cycle (FS, 1.14 μA; RS, −0.38 μA) and 121st cycle (FS, 1.19 μA; RS, −0.63 μA). The redox currents dropped during the 168th cycle (FS, 0.23 μA; RS, −0.64 μA) and marginally increased during the 224th cycle (FS, 0.48 μA; RS, −0.32 μA). DH activity in this case showed similar pattern like acid-shock treatment but the immediate DH activity after each pretreatment had decreased indicating the shift in the metabolic activities of biocatalyst towards strong reduction from the simultaneous redox reactions. This might have resulted in lower H⁺ shuttling between intermediates and further H₂ production, especially at the later phase of operation. H₂ production profiles were also observed to increase immediately after applying the pretreatment and gradually decreased with each cycle, which was supported by the observed redox catalytic currents. VFA accumulation showed an almost similar value after first pretreatment until the 238th cycle of operation followed by a decrease, supporting the observed redox behavior.

3.6.1 Tafel analysis. Tafel slope analysis was employed to understand the change/shifts in bioelectro-catalytic ability of the biocatalyst with the function of pretreatment method applied using the semi-empirical Tafel equation^42,43 relating current (i, A) and applied voltage (E, V) expressed by

ln i = i₀ + α_a nFE/RT (1)

This equation simplifies the kinetics of the electron transfer controlled process to two parameters, the exchange current density (i₀) and Tafel slope (β_a, α_anF/RT) to derive kinetic parameters viz., oxidative slope (β_a) and polarization resistance (R_p in Ω).⁴³ Tafel slopes and polarization resistance showed good correlation with the observed H₂ production profiles (Fig. S2;†Table 3). A marked decrement in the oxidative Tafel slope was observed after application of pretreatment in spite of a negligible variation in the reductive slope supporting the improved oxidation reactions along with the reduction (Fig. 8). This has resulted in the inter-conversion reactions leading to the increased proton gradient in the cell resulting in the improved H₂ production capability of the biocatalyst. The control operation showed a gradual increment in the oxidative slope with each cycle and reached a maximum during the 168th cycle (0.188 V/dec) followed by a drop till the 210th cycle (0.085 V/dec) which sustained till the 224 cycle (0.081 V/dec). On the contrary, the reduction slope was similar throughout the operation (0.49 ± 0.06 V/dec). The acid-shock culture depicted a drop in the oxidative Tafel slope (0.078 V/dec) and continued to decrease till the 37th cycle (0.056 V/dec). A gradual drop in H₂ production after the 37th cycle was accompanied by the gradual increase in the oxidative slope depicting higher values (78 ^th cycle; 0.089 V/dec). After applying the pretreatment, a marginal drop in the oxidative slope was observed till the 168th cycle (0.073 V/dec), which sustained a similar value thereafter till the 224th cycle (0.084 ± 0.001 V/dec) even after applying repetitive pretreatments. On the contrary, the reductive Tafel slope was almost similar to that of the control operation and throughout the operation indicating the possibility of similar reduction reactions along with the improved oxidation reactions within the biocatalyst. The reductive Tafel slope increased to a maximum value (37 ^th cycle; 0.782 V/dec) and decreased to the initial value (78 ^th cycle; 0.412 V/dec), which was almost stable thereafter (0.49 ± 0.02 V/dec). A lower and stable oxidative slope supports high favorability for the oxidation reactions, while stable and higher reductive slopes indicated lower reduction reactions. This has resulted in higher redox inter-conversion reactions and is supported by the observed higher DH activity leading to the higher H₂ production efficiency of the biocatalyst after the application of acid-shock treatment. Similarly, alkaline-shock pretreatment also showed a lower oxidative Tafel slope which gradually decreased till the 37th cycle (0.052 V/dec) and increased thereafter depicting a maximum value at the 168th cycle (0.094 V/dec) followed by a marginal drop at the 210th cycle, which was sustained afterwards (0.082 ± 0.001 V/dec). On the other hand, similar reductive slopes to acid-shock treatment were observed throughout the operation but with an increment in the reductive slope at the 224th cycle (0.632 V/dec). A gradual drop in overall H₂ production efficiency and DH activity with each pretreatment during the operation supports the observed increment in both the oxidative and reductive Tafel slopes at the end of operation.

Fig. 8 Tafel analysis, oxidative slope (ba), reductive slope (bc) and polarization resistance (Rp) with respect to experimental variations studied.

Table 3 The Tafel slopes and polarization resistance values during the pretreatment experiment with respect to cycle operation

Cycles	Oxidation slope (V/dec)			Reduction slope (V/dec)			Polarization resistance (Ω)
	Acid	Alkaline	Control	Acid	Alkaline	Control	Acid	Alkaline	Control
2	0.078	0.08	0.07	0.436	0.561	0.551	123100	404100	513800
37	0.056	0.052	0.135	0.782	0.632	0.412	341100	633200	638300
78	0.089	0.074	0.146	0.412	0.505	0.421	1258000	644500	1072300
168	0.073	0.094	0.188	0.509	0.492	0.521	442500	347000	1114500
210	00.082	0.084	0.085	0.487	0.437	0.531	1590000	1343000	1491000
224	0.085	0.079	0.081	0.471	0.632	0.529	268300	122000	1115400

---

The resistance for the electron transfer from the biocatalyst at the solution–electrode interface can be understood through polarization resistance (R_p) which was calculated using Tafel analysis (Table 3). R_p also showed a similar pattern like oxidative Tafel slope where higher R_p was observed during control (untreated) operation followed by acid-shock and alkaline-shock cultures. Gradual increment in R_p was observed with each additional cycle and reached a maximum at the 78th cycle (1072.3 kΩ), which was sustained thereafter. Higher R_p throughout operation supports the lower electron transfer which might be due to their consumption in the metabolic pathway towards methane. Acid-shock treatment reduced the R_p significantly (2nd cycle, 123.1 kΩ), which showed a gradual increment with each cycle and reached its maximum (78th cycle; 1258 kΩ) along with a decrement in H₂ production efficiency. After applying the pretreatment, the R_p decreased (168th cycle; 442.5 kΩ) and increased again (210th cycle; 1590 kΩ) followed by decrement to a lower value (224th; 268.3 kΩ) where the H₂ production was also stable. On the contrary, alkaline-shock treatment showed a gradual increment in R_p and reached a maximum value at the 210th cycle (1343 kΩ) and sustained thereafter (224th cycle, 1220 kΩ). Overall decrement in H₂ production efficiency and decreased DH activity with each addition pretreatment application supports the increased R_p.

3.6.2 Electron transfer rate. Voltamograms recorded under varying scan rates were used to analyze the dependency of the peak currents and to deliberate the electron transfer rates from the biocatalyst based on the bio-electrochemical reversibility. The pretreated mixed culture showed a relatively higher catalytic current compared to the control in the forward sweep irrespective of the scan rates indicating the change in oxidation reactions (Fig. 9). The relationship between peak currents and scan rates exhibited a linear correlation [R², 0.887 (untreated); 0.971 (acid-shock); 0.963 (alkaline-shock)] representing a typical surface controlled electrochemical processes. In the control, the reduction catalytic currents shifted towards more negative values from linearity by increasing the scan rate that supports the observed reductive behavior. A marginal change in oxidation current observed with varying scan rates also supports the same. On the contrary, redox pretreated cultures showed shifts in both oxidation and reduction currents with increasing scan rate towards more positive and negative values respectively from linearity, which is evidence for the effective function of simultaneous redox reactions facilitating the development of a proton gradient in the cell towards H₂ production. Further, the dependence of peak potential on the scan rate (v) was evaluated based on the Laviron theory,^44,45 when ΔE_p ≥ 200 mV (ΔEp = E_pc − E_pa).

E_pc = E⁰_c –[RT/αnF] ln [αnF ν_c/RTk_app] (2)

E_pa = E⁰_a –[RT/(1-α)nF] ln [(1-α)nF ν_a/RTk_app] (3)

Fig. 9 Voltammogram profiles and peak currents as a function of logarithm of the scan rates for experimental variations studied.

A higher electron transfer rate (k_app) was observed with the acid pretreated mixed culture followed by alkaline-shock and untreated mixed culture. Higher proton gradient in the cell due to the simultaneous redox reactions might be the probable reason for the observed higher k_app value in the case of pretreated mixed culture (acid, 24.0 S⁻¹; alkaline, 17.34 S⁻¹) over untreated culture (12.14 S⁻¹). Higher H₂ production and the DH activity observed throughout operation with the pretreated mixed culture strongly support the same. A relatively higher k_app compared to the alkaline-shock mixed culture also supports the observed higher H₂ production.

On the other hand, LSV was also employed to identify the maximum feasible redox catalytic currents from the biocatalyst. The instantaneous current generated during oxidation and reduction will give a broad understanding on the bio-electrochemical activity of the biocatalyst. The LSV profile showed a higher instantaneous current with the acid-shock culture during oxidation (7.47 mA) followed by alkaline-shock culture (7.1 mA) and untreated culture (3.24 mA) (Fig. S3, ESI†). On the contrary, a higher reduction current was observed with the alkaline-shock culture (−6.2 mA) followed by untreated (−6.1 mA) and acid pretreated culture (−3.87 mA). A significant increment in oxidation current after applying the pretreatment supports the induced oxidative behavior of the biocatalyst leading to simultaneous redox inter-conversion reactions. Moreover, the reduced reductive current observed with acid-shock operation also supports the suppression of reductive behavior leading towards higher H⁺ availability. During the stable phase of operation, CA was recorded under constant applied potential (1.2 V) to understand the maximum feasible sustainable current that the biocatalyst can generate (Fig. S3, ESI†). The acid-shock operation recorded a higher current (32 μA) initially followed by a marked drop up to few seconds and then stabilized near 2.5 μA after 100 s. The alkaline-shock culture also showed a higher current (22.5 μA) initially and dropped by stabilizing at 2.03 μA. On the contrary, a control operation showed a relatively lower current (18 nA) initially, which dropped significantly within 10 s and reached the base line. Pretreatment facilitated higher proton availability unlike control operation. The available H⁺ with the acid-shock culture leads to H₂ production, while in alkaline-shock operation leads to H⁺ consuming processes, which are also evident from the higher reductive currents observed in CV and LSV. The differential behavior of CA and the LSV pattern correlates well with CV profiles and indicates the effective performance of pretreated cultures in conserving protons.

A mixed culture is a combination of wide and diverse microorganisms including the H₂ producers as well as H₂ consumers. The protons and electrons generated by the untreated mixed culture during the substrate metabolism will be consumed by different TEAP depending on the acceptor available such as hydrogenase, carbon dioxide, sulphate, nitrate, etc. and the nature of microorganism. The untreated biocatalyst might have reduced the protons and electrons in methanogenesis, instead of maintaining a gradient in the cell, which might have resulted in the lower electron transfer compared to the pretreated mixed culture resulting in a lower k_app. However, after pretreatment, the diversity of the bacterial population will be restricted to a lower number based on the specific function of pretreatment method applied. The acid pretreatment helps in restricting the microorganisms to only acid producers, while the alkaline treatment restricts the microorganisms that can survive at extreme alkaline conditions. However, both the processes help in suppressing the methanogens which are pH sensitive. The suppression of methanogens might have helped in conserving the protons in the cell which can further proceed towards H₂ production. However, the acid treated culture showed higher efficiency over the alkaline treated mixed culture due to the survival of acidogens. The higher VFA accumulated during the operation with acid pretreated culture strongly supports the same. The DH activity observed in combination with the redox catalytic currents also depicts the higher efficiency of the acid treated culture over the alkaline treated culture.

4 Conclusions

Performance of selectively enriched anaerobic mixed culture by acid-shock and alkaline-shock methods in comparison with an untreated parent culture was studied for a long term. Efficiency of the biocatalyst was evaluated based on H₂ production and wastewater treatment efficiency in association with bioprocess monitoring and electro-kinetic parameters. Experimental data documented the positive influence of acid pretreatment method on the H₂ production compared to the alkaline pretreated and untreated cultures. Acetate-type fermentation was observed with acid pretreated culture while alkaline pretreatment showed mixed type fermentation similar to the untreated parent mixed culture. Contrary to H₂ production, substrate degradation was found to be effective with untreated culture due to the presence of methanogenic bacteria. The study documented synergetic interaction of microbial inventory, dehydrogenase activity and bio-electro kinetics in association with H₂ production and substrate degradation throughout the operation. Dehydrogenase activity showed improvement after application of acid pretreatment indicating the increased redox inter-conversion reactions leading to the higher proton gradient in the cell. Cyclic voltammograms visualized the increased biocatalyst performance after acid-shock pretreatment. Bioelectro-catalytic ability of the biocatalyst by Tafel analysis also showed the efficiency of pretreatment application to the biocatalyst. Microbial inventory studies illustrated the effective suppression of methanogenic activity after applying acid-shock pretreatment. Changes in microbial communities indicated selective enrichment of Clostridium sp. and Bacillus sp. in pretreated cultures. Long term operation of the bioreactor illustrated the feasibility of controlling the microbial metabolic function by repeated application of pretreatment to the reactor native microflora (in situ) during operation whenever it was required to modify or regain the process performance.

Acknowledgements

The authors wish to thank the Director, CSIR-IICT for his encouragement in carrying out this work. The authors acknowledge the financial support from Ministry of New and Renewable Energy (MNRE, Project. No. 103/131/2008-NT), Government of India for the research grant on biohydrogen research. RKG acknowledge Council of Scientific and Industrial Research (CSIR) for providing research fellowship.

References

1. Y. Y. Wang, P. Ai, C.-X. Hu and Y.-L. Zhang, Int. J. Hydrogen Energy, 2011, 36, 390–396.
2. E. P. Leano, S. Babel,  DOI:10.1016/j.renene.2011.08.030..
3. S. Venkata Mohan, P. Chiranjeevi and G. Mohanakrishna, Int. J. Hydrogen Energy, 2011, 37, 3130–3141.
4. H. Liu, G. Wanga, D. Zhu and V. Pan, Int. J. Hydrogen Energy, 2009, 34, 9696–9701.
5. C. J. Ling, C. W. Guang, Y. C. Li, Z. H. U. DaLing, P. A. N. GuangHua, Chinese Science Bulletin, 2009,  DOI:10.1007/s11434-009-0399-5..
6. S. Venkata Mohan, Journal of Scientific and Industrial Research, 2008, 67(11), 950–961.
7. S. Venkata Mohan, Int. J. Hydrogen Energy, 2009, 34, 7460–7474.
8. S. Srikanth, S. Venkata Mohan, V. Lalit Babu and P. N. Sarma, Int. J. Hydrogen Energy, 2010, 35, 10693–10700.
9. P. Mohammadi, S. Ibrahim, M. Suffian, M. Annuar and S. Lawa, J. Clean. Prod., 2011, 19, 1654–1658.
10. S. Venkata Mohan, G. Mohanakrishna, S. Srikanth, in Biofuels, Elsevier Science and Technology, ed. Ashok Pandey, Christian Larroche, Steven C Ricke, Claude-Gilles Dussap, Edgard Gnansounou, Imprint, Burlington, Academic Press, 2011, pp. 499–524..
11. S. Venkata Mohan, V. Lalit Babu and P. N. Sarma, Bioresour. Technol., 2008, 99, 59–67.
12. R. Sparling, D. Risbey and H. M. Poggi-Varaldo, Int. J. Hydrogen Energy, 1997, 22, 563–566.
13. N. Q. Ren, W. Q. Guo and X. Y. Wang, Int. J. Hydrogen Energy, 2008, 33, 4318–4324.
14. H. Zhu and M. Beland, Int. J. Hydrogen Energy, 2006, 31, 1980–1988.
15. Y. Mu, H. Q. Yu and G. Wang, Enzyme Microb. Technol., 2007, 40, 947–953.
16. J. Wang and W. W. Wan, Int. J. Hydrogen Energy, 2008, 33, 2934–2941.
17. S. O-Thong, P. Prasertsan and N. K. Birkeland, Bioresour. Technol., 2009, 100, 909–918.
18. B. Hu and S. L. Chen, Int. J. Hydrogen Energy, 2007, 32, 3266–3273.
19. G. Luo, L. Xie, Z. Zou, W. Wang and Q. Zhou, Bioresour. Technol., 2010, 101, 959–964.
20. S. H. Kim and H. S. Shin, Int. J. Hydrogen Energy, 2008, 33, 5266–5274.
21. S. Venkata Mohan, S. Veer Raghavulu, R. Kannaiah Goud, S. Srikanth, V. Lalit Babu and P. N. Sarma, Int. J. Hydrogen Energy, 2010, 35, 12208–12215.
22. D. Y. Cheong and C. L. Hansen, Appl. Microbiol. Biotechnol., 2006, 72, 635–643.
23. R. Kannaiah Goud, S. Veer Raghavulu, G. Mohanakrishna, K. Naresh and S. Venkata Mohan, Int. J. Hydrogen Energy, 2012, 37, 4068–4076.
24. G. Muyzer, E. C. Waal de and A. G. Uitterlinden, Appl Environ Microbiol, 1993, 59, 695–700.
25. APHA, Standard Methods for the Examination of Water and Wastewater. American Public Health Association/American Water Works Association/Water Environment Federation, Washington, DC, USA, 1998.
26. J. T. Kraemer and D. M. Bagley, Biotechnol. Lett., 2007, 29, 685–695.
27. S. Venkata Mohan, in Sustainable Biotechnology: Renewable Resources and New Perspectives, ed. S. Om, and H. Steve, Springer, 2010, pp 129–164..
28. S. Venkata Mohan, G. Mohanakrishna, S. V. Ramanaiah and P. N. Sarma, Int. J. Hydrogen Energy, 2008, 33, 550–558.
29. H. H. P. Fang and H. Liu, Bioresour. Technol., 2002, 82, 87–93.
30. I. M. M. Moreno-Davila, L. J. Rios-Gon-Zalez, J. G. Gaona-Llozano, Y. Garza-Garcia and J. G. Rodriguez-Martinez, Applied Science, 2010, 5(6), 376–382.
31. J. Wang and W. Wan, Int. J. Hydrogen Energy, 2009, 34, 799–811.
32. M. Cai, J. Liu and Y. Wei, Environ. Sci. Technol., 2004, 38, 3195–3202.
33. G. Antonopoulou, H. N. Gavala, L. V. Skiadas, K. Angelopoulos and S. Lyberato, Bioresour. Technol., 2008, 99, 110–119.
34. S. Venkata Mohan and M. Lenin Babu, Bioresour. Technol., 2011, 102, 8457–8465.
35. C. Li and H. H. P. Fang, Crit. Rev. Environ. Sci. Technol., 2007, 37, 1–39.
36. S. Venkata Mohan, L. Agarwal, G. Mohanakrishna, S. Srikanth, A. Kapley and H. J. Purohit, Int. J. Hydrogen Energy, 2011, 36, 8234–8242.
37. F. R. Hawkes, I. Hussy, G. Kyazze, R. Dinsdale and D. L. Hawkes, Int. J. Hydrogen Energy, 2007, 32, 172–184.
38. Y. Tao, C. Yang, W. Yongqiang, H. Yanling and Z. Zhihua, Int. J. Hydrogen Energy, 2007, 32, 200–206.
39. T. Kawano, K. Wada, Y. Y. Li and T. Noike, J. Jpn. Soc. Water Environ., 2004, 27, 473–479.
40. T. Noike, H. Takabatake, O. Mizuno and M. Ohba, Int. J. Hydrogen Energy, 2002, 27, 1367–1371.
41. S. Venkata Mohan, M. Lenin Babu, M. Venkateswar Reddy, G. Mohanakrishna and P. N. Sarma, Int. J. Hydrogen Energy, 2009, 34, 6149–6156.
42. E. G. & G. Services Parsons, in Fuel Cell Handbook, U.S. Department of Energy Office of Fossil Energy National Energy Technology Laboratory, Morgantown, 2000, ch. 2, pp 2–23–2–26.
43. S. Veer Raghavulu, P. Suresh Babu, R. Kannaiah Goud, G. Venkata Subhash, S. Srikanth and S. Venkata Mohan, RSC Adv., 2012, 2, 677–688.
44. E. Laviron, J. Electroanal. Chem., 1974, 52, 355–393.
45. S. Venkata Mohan and K. Chandrasekhar, Bioresour. Technol., 2011, 102, 9532–9541.

---

Footnote

† Electronic Supplementary Information (ESI) available. See DOI: 10.1039/c2ra20526b/

---