Investigation of the potential effect of encapsulated metal nanoparticles on enhancement of thermophilic anaerobic digestion

: The present work investigates the enhancement effect of seven different catalysts made of Cu/SiO 2 , Pd/SiO 2 , Pt/SiO 2 , Ni/SiO 2 , Co/SiO 2 , Ag/SiO 2 and Fe/SiO 2 nanoparticles (NPs) on methane production during thermophilic anaerobic digestion. The tested NPs were synthesized by the sol-gel


Introduction
Increased demand and inevitable depletion of fossil fuels, together with worldwide concerns about greenhouse gas emissions, have resulted in the development of promising technologies for renewable energy production [1].Therefore, potential alternatives for energy generation are intensively studied.One option is the use of biomass feedstock for the production of biogas through anaerobic digestion [2,3].This process is a biochemical technological process for the treatment of a wide range of feed-stocks (e.g., organic fraction of municipal waste, animal manure and slurry, agricultural crops, etc.) to produce methane-rich biogas which can be used as a replacement for fossil fuels in both heat and power generation and as a vehicle fuel [4,5].
However, there are critical issues, which need to be addressed to make the production of biomethane a techno-economically viable and ecologically acceptable renewable substitute.One of the most important issues is the effect of trace metals addition on anaerobic digestion.These metals can be stimulatory, inhibitory or even toxic for biochemical reactions, depending on their concentrations [6,7].
As reported in the literature, nickel [8][9][10][11][12], cobalt [8,11,13,14], iron [15][16][17][18] and molybdenum [19] are all involved in the methane production biochemical process and serve as cofactors in enzymes which are involved in the biochemistry of methane formation [7,20].In addition, a number of microbes are able to use some metals or metalloids as electron donors or acceptors in energy metabolism.They include eubacteria and archaea.Depending on the element, the metal species may be in simple ionic form or in the form of oxyanions.As energy sources, oxidizable metals or metalloids may satisfy the entire energy demand of an organism [21].
Due to nanoparticles' (NPs') (particles of 1-100 nm in diameter) unique electronic, optical, photonic and catalytic properties [22], which are primarily based on their small size and their high surface to bulk ratio, and because of their great potential in application to many science fields, the last decades have been marked by the increased attention of scientists to NPs [23,24].
Recently, enormous interest has been focused on biological application of metal NPs since they have shown interactions with microorganisms.Nevertheless, the influence of metal NPs on anaerobic digestion has seldom been investigated.Some publications discussed the effect of NPs' addition on anaerobic digestion.For example, Gonzalez-Estrella et al. [25] reported that Ag 0 , Al2O3, Ce2O3, Mn2O3, Fe 0 , Fe2O3, SiO2 and TiO2 NPs did not show any inhibitory effect on methanogens' activities even at concentrations of 1500 mg L -1 .Likewise, nano-TiO2, Al2O3 and SiO2 in doses up to 150 mg g -1 -TSS (total soluble sugars)) showed no inhibitory effect, whereas nano-ZnO showed inhibitory effect with its dosage increased [26].Luna-delRisco et al. [27] reported that 10.7 mg L -1 CuO and 57.4 mg L -1  ZnO NPs inhibited methane production from cattle manure.Other authors investigated the enhancement effect of NPs at very low concentration on bio-hydrogen anaerobic fermentation, including silver NPs [14,28], hematite Fe2O3 NPs [29,30], gold NPs [14,31], iron NPs [24,32], etc.At present, little is known about how NPs affect microorganisms in an anaerobic environment.Therefore, it is attractive to investigate the catalytic effect of NPs on methane production.
The purpose of this study was to investigate the influence of metal NPs (Cu/SiO2, Pd/SiO2, Pt/SiO2, Ni/SiO2, Co/SiO2, Ag/SiO2 and Fe/SiO2) on methane generation during anaerobic digestion and to suggest the mechanisms.The novelty of this study resides in the use of metallic NPs that are encapsulated in porous silica, allowing a constant release of the metal and avoiding agglomeration of the catalysts, to enhance bio-methane production.First, the effects of these seven NPs on methane generation were studied with acetate substrate as the final intermediate used to produce methane, and the experiments were carried out in batch tests (125 mL vials).In the second step, the effects of the most efficient NPs resulting from previous assays were investigated with complex substrates such as starch and glucose with three different NP concentrations, 10 -4 , 10 -5 and 10 -6 mol L -1 .Gas and liquid chromatography were used to monitor the intermediates and final metabolites, such as volatile fatty acids (VFAs) produced from the initial substrate of each stage.

Synthesis of NPs
Metallic salts of Cu, Pd, Pt, Ni, Co, Ag and Fe were used for the synthesis of encapsulated NPs inside a porous silica matrix, denoted as Cu/SiO2, Pd/SiO2, Pt/SiO2, Ni/SiO2, Co/SiO2, Ag/SiO2 and Fe/SiO2.The sol-gel process was used as described by [54,55].This process consists of the following: (i) First, there is the preparation in one step (called the "cogelation" method), based on the concomitant hydrolysis and condensation of tetraethoxysilane (TEOS, Si(OC2H5)4) with a modified alkoxide of formula (RO)3Si-XL, in which the ligand L is able to form a complex (LM) n m+ with a metallic cation M m+ (such as Pd 2+ , Ag + , Pt 2+ , ...) and is connected to the alkoxide (RO)3Si-moiety through an inert and hydrolytically stable organic group, X.By this method, materials in which the metallic cation is anchored into the matrix were obtained.(ii) Then, there is drying under vacuum to remove solvent, calcination under air to burn organic moieties and reduction under hydrogen to reduce metallic oxides into metals.The samples were characterized by nitrogen adsorption-desorption isotherms, electron microscopy and X-ray diffraction, as previously described in [54,55].For the clarity of this work, NPs are defined as metallic (dried), metallic oxide (calcined) and zero-valent metal (reduced) nanoparticles highly dispersed inside the silica matrix, whereas catalyst is used to refer the combination between NPs and silica.
Concentrated suspensions in water of these samples were prepared in 50 mL bottles by finely pounding (at micrometer-size) and weighing some catalysts.Based on the mass suspended in the bottles, the metallic mass loading in the catalyst and the metal atomic weight, a defined volume of homogenized suspension was transferred in the culture medium prior to sterilization in order to reach a final concentration of 10 -4 mol metal L -1 .Therefore, all the tests had the same NP concentration, but the total mass of catalyst (i.e., NPs + SiO2) differed from one test to the others, because the metallic mass loading differs between the investigated catalysts.

Characterizations of NPs
Nitrogen adsorption-desorption isotherms of all NPs samples were determined in an ASAP multisampler device from Micromeritics at -196 °C.
Transmission electron microscopy (TEM) pictures were obtained on a Tecnai G2 TWIN device from FEI.
The actual composition of the metal in the SiO2 materials was determined by inductively coupled plasma atomic emission spectroscopy (ICP-AES), with an ICAP 6500 THERMO Scientific device.The mineralization is fully described in [56], with HF used instead of HNO3.

Experimental conditions
The batch tests were performed in three experiments.The first experiment was aimed to study the effect of 10 -5 mol L -1 of Cu/SiO2, Pd/SiO2, Pt/SiO2, Ni/SiO2, Co/SiO2, Ag/SiO2 and Fe/SiO2 NPs on methane production by acetoclastic methanogens.Sodium acetate monohydrate (5 g L -1 ) was used as a substrate.In the second experiment, soluble starch (5 g L -1 ) was used as substrate with 10 -5 mol L -1 of Ni, Co and Cu NPs.In the third experiment, the effects of different concentrations and different valence states of Ni/SiO2 and Co/SiO2 NPs were investigated using glucose monohydrate as the substrate.Batch studies were performed to analyze methane production of each combination during treatment.Control trials were performed without NPs.
All batch tests were conducted in triplicate in 125 mL glass serum vials containing 65 mL of nutrients (modified MDT medium) and specified concentrations of NPs (mol L -1 ).The vials were inoculated by 5 mL of anaerobic sludge obtained from a full-scale anaerobic reactor treating biohydrogen effluent.The vials were capped with butyl rubber and degassed with nitrogen for 2 min to remove oxygen from the headspace and maintain an anaerobic environment.The tests were performed under thermophilic conditions (55 °C).

Monitoring and analytical methods
The biogas production was measured by releasing the gas pressure in the vials using appropriately sized glass syringes in the 2.5-60 mL range to equilibrate with the ambient pressure.Water containing 9 M KOH in 100 ml gas replacement equipment was used to monitor the biogas production and composition, and the absorption potential of the KOH solution was regularly measured using gas mixtures containing 0, 20, 35, 80 and 100% carbon dioxide.The hydrogen (for the first 5 days of fermentation), methane and carbon dioxide contents were determined for each gas sample using the procedure published by Hiligsmann et al. [57] but adapted for anaerobic digestion.
Hydrogen, methane and carbon dioxide were determined using a method described by Hamilton et al. [58].The separation was achieved using a Hewlett Packard 5890 Series II gas chromatograph (GC; Agilent Technologies, Santa Clara, CA, USA) equipped with a 30 m long, 0.32 mm id Alltech GAS PRO GSC column (Grace, Deerfield, IL, USA) in series with a 20 m long, 0.25 mm id Chrompack CARBOPLOT P7 column (Agilent Technologies) and a thermal conductivity detector.The carrier and reference gas was He, and a mixture of nitrogen (15 %), carbon dioxide (35 %) and methane (50 %) was used to calibrate the instrument for determining the proportions of methane and carbon dioxide in the biogas.A mixture of hydrogen (80 %) and carbon dioxide (20 %) was used to allow the fraction of hydrogen in the biogas produced to be determined.The GC injection port, the thermal conductivity detector chamber and the oven were maintained at 90, 110, and 55 °C, respectively.VFAs' concentrations in the culture medium were determined using an Agilent 1110 series high performance liquid chromatograph (HPLC; Agilent Technologies) equipped with a Supelcogel C-610H column (Sigma-Aldrich, St Louis, MO, USA) preceded by a Supelguard H precolumn (Sigma-Aldrich).The columns were kept at a temperature of 40 °C, and the isocratic mobile phase was 0.1 % H3PO4 (in ultrapure, ''milliQ'' water), at a flow rate of 0.5 ml min -1 .A differential refraction index detector, kept at 35 °C, was used.This analysis took 35 min at a maximum pressure of 60 bar.The instrument was calibrated using solutions of glucose of 0.125, 0.25, 0.5, 1, 2 and 4 g L -1 , acetate, ethanol, propionate and butyrate.The liquid samples were collected and centrifuged at 13000 g for 10 min, and the obtained supernatants were filtered through a 0.2 mm cellulose acetate membrane (Minisart Sartorius).

Kinetic modeling
Mathematical modeling of microbial growth was been used to estimate various parameters.There are many models, such as the Gompertz, logistic, Richards, Stannard and Schnute models, that have been used for modeling somatic growth and population dynamics and therefore could be applied to microbial growth [59].In this study, to describe the effect of NPs on anaerobic digestion and methane production, cumulative methane production curves with respect to time were initially obtained from the methane production experiments.In the next step, the modified equation of Gompertz [60] was applied to determine the methane production potential (A), maximum rate of methane production (µ m ) and the duration of the lag phase (λ).The Modified Gompertz equation (Eq 1) has been widely used to predict rates of fermentative gas production processes [61].
where y is cumulative methane production (mL), λ is lag time (day), A is methane production potential (mL), μ m is the maximum methane production rate (mL day -1 ), and e = 2.718281828.Values of y, λ , A and μ m for each batch were estimated using Origin 9.0 for nonlinear regression analysis.

Metallic NPs characterizations
The compositions of the NPs are given in Table 1.For each metallic NP, the same amount of metal is obtained around 1 wt% (Table 1).The NP samples were measured by XRD, but due to the low amount of metal, no peak associated with the metallic species can be detected [51].An example is given for Pt/SiO2 NPs in Figure 1.Only the large peak corresponding to amorphous silica is noticed [51].The specific surface areas of all samples are also summarized in Table 1.All NP samples are in the range of 300 m 2 /g, as previously observed with this amount of dopant in silica [52,54,62], and only the Pt/SiO2 sample presents a higher specific surface area of 560 m 2 /g. Figure 2 presents the TEM images of all NP samples.It is observed that each sample has the same morphology: (i) They are composed of large silica nanoparticles around 20-50 nm in lighter color, and (ii) small metallic nanoparticles in dark color around 1-3 nm are dispersed in the lighter ones.This typical morphology had been observed with this specific sol-gel synthesis where small metal nanoparticles are dispersed in a porous silica matrix [52,54,62].
From these characterizations, it can be summarized that small metal nanoparticles have been highly dispersed in a porous silica matrix, allowing access to these metallic species while avoiding their agglomeration and securing a progressive metal release in water [51].

Effect of metallic NPs on acetate degradation and methane production
In a first step, seven encapsulated NPs of Cu/SiO2, Pd/SiO2, Pt/SiO2, Ni/SiO2, Co/SiO2, Ag/SiO2 and Fe/SiO2 with diameters of 2-3 nm [54,55] were added in a biochemical methane potential (BMP) test (120 mL batch serum bottles with working volume of 70 mL).Sodium acetate monohydrate (5 g L -1 ) was used as the substrate.The NP concentration was adjusted to 10 -5 mol L -1 in the culture medium.The porous silica matrix used to encapsulate the NPs maintains the stability of the metal NPs, which can very easily coagulate in solution to form aggregates, which are less effective as catalysts than the original ones [63].
Biogas production was used as an indicator of anaerobic digestion imbalance [64].The volume of biogas was measured every 15 days during 90 days after the inoculation.Cumulative methane production during the incubation period from the control and test samples with different NPs is illustrated in Figure 3.In this experiment, a slight delay in methane production occurred in all samples during the first days of fermentation due to high substrate/microorganism ratio (i.e., more substrate and less active bacteria).Hence, microbial consortia had to acclimatize to this new environment for their optimum metabolism [65].Once the acclimatization had elapsed, the methane production started increasing in all samples and reached peak rates at 30 days specifically in the samples containing Ni/SiO2 and Co/SiO2 NPs, where the methane production increased significantly by 25.5 and 22.5%, respectively, compared to the controls with substrate only.The results indicated that the addition of Ni/SiO2 and Co/SiO2 markedly stimulated methane production and resulted in the complete conversion of the added acetate to methane.This is in agreement with results obtained by Fathepure [66], who described the bioavailability of nickel and cobalt as an important factor for enhancement of methane production.Cu/SiO2 and Fe/SiO2 NPs showed a slight improvement effect on methane production, which was evident after 45 days of fermentation.Previous results [66] indicated that addition of iron at a concentration as low as 10 -4 mol L -1 stimulates the conversion of acetate to methane.Furthermore, the results showed that the addition of Pt/SiO2 and Cu/SiO2 NPs did not have any significant effect on the production of methane and acetate consumption compared with control.The effect of copper on fermentative methane production has been documented in the literature [67,68], with a highly toxic effect on methane.However, the toxic effect of Cu/SiO2 NPs did not appear in our experiment since NPs are encapsulated inside the porous silica matrix, which limits the metal ions' liberation.The used concentration was at the same order as that reported by Karri et al. [68], who found partial inhibition of methane production by acetoclastic methanogens at 1.6×10 -4 -3.2×10 -4 mol L -1 of Cu 2+ and complete inhibition at a concentration of 4.7×10 -4 mol L -1 of Cu 2+ .Despite extensive use of silver NPs for antimicrobial applications, in the bottles containing Ag/SiO2 NPs, methane production rate was a little bit slower than in control bottles, especially from the beginning until 30 days of fermentation, where methane production started to increase significantly to produce a volume of methane close to the control.The results of [69] showed that silver up to 9.10 -4 mol L -1 did not adversely affect the methanogenic cultures.In contrast, it was found that Pd/SiO2 NPs showed a clearly inhibitory effect on methane production and bacterial activity.
In order to evaluate the effects of Cu/SiO2, Pd/SiO2, Pt/SiO2, Ni/SiO2, Co/SiO2, Ag/SiO2 and Fe/SiO2 NPs on methane production by acetotrophic methanogens' activity along with acetate degradation, the experimentally obtained data were fitted to the modified Gompertz model.The estimated values of the parameters are summarized in Table 2.The correlation coefficients of nonlinear analysis for all metals were above 0.99.According to this model, the values of the maximum cumulative methane productions (mL) for all samples were convergent at the end of the experiment, i.e., 95.8 ± 1.3 ml.However, the values of the maximum methane production rate (MPR) were significantly differentiated.They were 70%, 48%, 7% and 6% higher with Ni/SiO2, Co/SiO2, Fe/SiO2 and Pt/SiO2 NPs, respectively, than the control without NPs.In contrast, a significant inhibitory effect was found with Pd/SiO2 NPs, with the maximum being 39% lower than in the control.Again, no significant inhibitory effect was found with Ag/SiO2 and Co/SiO2 NPs.The calculated lag time ranged between 8.03 and 16.65 days.The maximum of 16.65 days obtained for Pd/SiO2 NPs confirms the inhibitory effect.Note: λ = Lag phase duration (days), μm = Maximum methane production rate (mL day -1 ), A = Total methane production (mL), R² = correlation coefficient.Standard deviations are calculated based on the triplicate experiments made for each condition.All the R 2 values for the Gompertz model were higher than 0.99.
In our experiment, the production of methane is accompanied by acetate degradation.Thus, acetate concentration is a useful indicator for monitoring methane production.Figure 4 showed a significant variation in the rate of acetate degradation in the presence of NPs.The highest acetate removal rate was between 15 and 45 days, where 60-79% of acetate was converted to methane and carbon dioxide, except for the Pd/SiO2 NPs, which showed an inhibitory effect on acetate degradation.Results indicate that Ni/SiO2 NPs had the best effect on acetate consumption, since the acetate was depleted after 60 days.The other NPs, except Co/SiO2 NPs, showed no significant effect on acetate degradation.
After the 90 days of culture and total arrest of methane production in all bottles, a new culture phase was carried out by adding 10 mL of acetate solution up to 5 g L -1 of acetate monohydrate in all serum bottles.The purpose of this experiment was to investigate the long-term effect of nanoparticles on methane production from acetate and to confirm the previous results.The results (Figure 5) showed that except for the Pd/SiO2 NPs, the curve profiles and the volumes of methane produced during both culture phases are similar.This was confirmed when comparing the parameters of the Gompertz model (Table 3) since no significant difference was observed between the tested NPs and the control.The values of the maximum methane production rates and lag times for all conditions were slightly lower than obtained from the first culture phase, probably due to the large inoculum in the second phase.The highest performances for methane production were recorded with Ni/SiO2 NPs, with maximum methane production rates 69% higher than the control without NPs.3. Kinetic parameters and Gompertz coefficient adjusted on the profiles of volumetric methane production curves for the BMP tests with 10 -5 mol L -1 of NPs and the addition of 5 g L -1 acetate monohydrate.

Effect of Ni/SiO2, Co/SiO2 and Cu/SiO2 NPs on anaerobic digestion of starch and methane production
In consecutive experiments, Ni/SiO2, Co/SiO2 and Cu/SiO2 NPs were added at the same concentration (10 -5 mol L -1 ) in batch cultures with 5 g L -1 of soluble starch as substrate.In the first 5 days of incubation, starch degradation to VFAs was accompanied by hydrogen and carbon dioxide production.Total biogas produced during the acidogenic phase (Figure 6a) was, on average, 55 mL, including hydrogen production ranging between 40 and 44 mL.This volume of hydrogen, which was eliminated, might produce 10 mL of methane by carbon dioxide reduction.However, the results displayed in Figure 6b confirmed the large influence of Ni/SiO2 and Co/SiO2 NPs on methane production compared to the control with substrate only.It was found that after 42 days, the addition of Ni/SiO2 and Co/SiO2 NPs increased the methane production by 47% and 22%, respectively, compared to the control with substrate only.Meanwhile, the addition of Cu/SiO2 NPs showed consistently negative responses, including decrease in methane production by 6% compared with the control.The results of this work could be compared to those obtained by Pobeheim et al. [8], who found that nickel at 10.6×10 -6 mol L -1 enhanced methane production by 25%, and cobalt at 2×10 -6 mol L -1 increased the methane production by 10% approximately from substrate consisting of xylan and starch.The maximum methane production rates calculated from the Gompertz model (Table 4) were 70% and 33% higher with Ni/SiO2 and Co/SiO2 NPs, respectively, and 11% lower with Cu/SiO2 NPs than the control without NPs.Again, comprehensively, no significant differences were observed for the maximum cumulative methane productions (mL) at the end of the experiment (i.e., 119.6 ± 2.5 ml), and the calculated lag time ranged between 8.6 and 12.1 days.
The addition of the NPs in the culture batch did not modify the metabolites profile (Figure 7a-d).After one day of incubation, 61-67% of the starch was converted into VFAs and ethanol.Meanwhile, 8%, 12%, 16% and 17% of the starch were converted into lactate, formate, acetate and butyrate, respectively, in the presence of Ni/SiO2 NPs (Figure 7c), compared with 13%, 11%, 17% and 13% in the presence of Co/SiO2 NPs (Figure 7d); 27%, 7%, 7% and 16% in the presence of Cu/SiO2 NPs (Figure 7b); and 11%, 11%, 13% and 20% in the control (Figure 7a).Therefore, copper is associated with some differences in the general metabolites.The metabolite distribution profiles (Figure 7a-d) showed that the addition of Ni/SiO2 and Co/SiO2 NPs, respectively, accelerated the conversion of metabolite into acetate.Then the production of methane that was clear after 15 days of incubation with Ni/SiO2 NPs (Figure 6).It results a total conversion of metabolites into methane after about 72 days of fermentation (Figure 6).Table 4. Kinetic parameters and Gompertz coefficient adjusted on the profiles of volumetric methane production curves for the BMP tests with 10 -5 mol L -1 of Ni/SiO2, Co/SiO2 and Cu/SiO2 NPs and 5 g L -1 soluble starch.

Evaluation of the optimum concentration of nickel NPs on methane production
In a number of serial batch experiments, Ni/SiO2 NPs at different valence states (i.e, calcined, reduced and dried) and concentrations (10 -6 , 10 -5 and 10 -4 mol L -1 ) were tested using glucose monohydrate (5 g L -1 ).The volume of biogas was measured every 15 days during 105 days after the inoculation, with the exception of the first 5 days, where the volume of biogas was measured after 1, 3 and 5 days, since glucose degradation to VFAs is associated with hydrogen production that can lead to inhibition of acetoclastic and hydrogenotrophic methanogens.The results showed that hydrogen production ranged between 40 and 55 mL, with improvement in the presence of calcined Ni/SiO2 NPs at a concentration of 10 -4 mol L -1 up to 40% above control.Methane production profiles during the incubation period with different Ni/SiO2 NP concentrations are illustrated in Figures 8a-c.Batch trials with all valence states of Ni/SiO2 NPs showed higher methane performances than without NPs.After 15 days and whatever the Ni/SiO2 NP concentration, the methane production was about 15.3 ± 2.0 mL.
During the exponential growth phase, it increased appreciably with increasing Ni/SiO2 NP concentration up to 10 -4 mol L -1 .At day 60, the methane potential was already produced with the calcined Ni/SiO2 NPs at a concentration of 10 -4 mol L -1 , whereas 13%, 21% and 25% were still to be produced at concentrations of 10 -5 and 10 -6 mol L -1 and in the control, respectively.Results were similar with dried Ni/SiO2 NPs and about 12% lower for reduced Ni/SiO2 NPs.Batch tests with the addition of NiCl2 were also carried out to compare the influence of different concentrations of NiCl2 with those of Ni/SiO2 NPs on methane production.The results displayed in Figure 8d confirm the large influence of nickel on methane production and showed that the improvement effect of NiCl2 was higher than that of NPs at 10 -6 and 10 -5 mol L -1 concentration.Previous studies showed that methane production was enhanced by addition of nickel in concentrations ranging between 10 -7 and 10 -4 mol L -1 [70][71][72][73].Murray et al. [74] reported that nickel addition (10 -7 mol L -1 ) and especially in combination with cobalt and molybdenum increased total biogas and methane production from food processing waste by 42%, while Kumar et al. [75] found that total biogas in anaerobic digestion of potato waste and cattle manure mixture was increased 71% by the addition of nickel (4.26×10 -5 mol L -1 ).Work by Qing-Hao Hu et al. [73] has highlighted the possibility of increasing the methane production in anaerobic digestion of synthetic wastewater up to 45% with increasing nickel concentration up to 2×10 -4 mol L -1 .Similar results were obtained by Lo et al. [76], who found that nickel at concentrations varying between 10 -5 and 10 -4 mol L -1 enhances biogas production.
The results obtained by Gompertz modeling (Table 5) confirm the trends, with the highest maximum methane production (MPR) (3.38 ml/day) achieved with the addition of NiCl2 at a concentration of 10 -4 mol L -1 .The MPR was 86% higher than in the control.Using NPs, the MPR compared to control conditions was higher by 72.1%, 26.3% and 8.5% for calcined Ni/SiO2 NPs at concentrations of 10 -4 , 10 -5 and 10 -6 mol L -1 .Lower effects were achieved with dried and reduced Ni/SiO2 NPs, varying from 6.3% to 48.6% MPR increases depending on NP concentration (10 -6 -10 -4 mol L -1 ).Table 5. Kinetic parameters and Gompertz coefficient adjusted on the profiles of volumetric methane production curves for the BMP tests with different valence states of Ni/SiO2 NPs and NiCl2 at different concentrations (10 -6 , 10 -5 and 10 -4 mol L -1 ).
Metal concentration (mol L -1 ) λ (days) µm (mL day -1 ) A (mL) R Regarding metabolites produced from glucose degradation, Figure 10 shows that the main intermediate products during the acidogenic step were lactate, formate, acetate and butyrate.The accumulation of VFA leads to a decrease of medium pH down to the values of 5.2-5.4,which are unfavorable for the methanogens.In the course of the following consumption of VFAs, the medium pH value recovered to 7.2.Lactate, formate and ethanol were the first products converted in acetate (Figure 10b).The BMP tests showed that the presence of Ni/SiO2 NPs in the medium did not quite modify the metabolites profile but accelerated the consumption of acetate and butyrate by the methanogens and acetogens, respectively.The carbon mass balance calculated on the basis of detected substances at the end of experiment showed that 79±2% of glucose was converted into methane and carbon dioxide for all cases, which was close to results from [77].

Evaluation of the optimum concentration of Co/SiO2 NPs on methane production
The improvement effect of different valence states of Co/SiO2 NPs (i.e., calcined, reduced and dried) using the BMP tests were carried out at different concentrations i.e., 10 -6 , 10 -5 and 10 -4 mol L -1 .Experimental conditions were the same used with Ni/SiO2 NPs.The volume of biogas measured after 1, 3 and 5 days showed a limited effect of cobalt NPs on hydrogen production, while the hydrogen production was 43±4 mL in all conditions.The production of methane profiles during the incubation period from the control and test samples with different concentrations of cobalt NPs are illustrated in Figure 7a-c.No inhibitory effect was observed in all Co/SiO2 NP treatments.In contrast, the addition of Co/SiO2 NPs showed consistently positive responses, including increases in methane production.In this experiment, calcined Co/SiO2 NPs (Figure 9a) showed higher improvement effect on methane production compared with the other forms of Co/SiO2 NPs, and the results showed that 90% of methane potential was produced in the presence of 10 -4 mol L -1 calcined Co/SiO2 NPs after 60 days, with increases of 8%, 13% and 14% compared with the concentrations of 10 -5 and 10 -4 mol L -1 and the control, respectively.Slightly lesser improvement effects were detected with dried and reduced cobalt NPs.However, cumulative methane yield after 105 days was close for all cases, including the control (i.e., 348 ± 7 mL CH4 / g glucose), but the presence of Co/SiO2 NPs accelerated the methane production.
As done with Ni/SiO2 NPs, batch tests with the addition of CoCl2 were carried out using the same concentrations tested for Co/SiO2 NPs.Cumulative production of methane during the incubation period from the control and test samples with different concentrations is illustrated in Figure 9 d.The results showed that the stimulatory effect of CoCl2 on methane production was lower at a concentration of 10 -4 mol L -1 by 5% than calcined Co/SiO2 but still higher by 12% than in the control.

Discussion
The presence of heavy metal ions (i.e., Cu, Zn, Fe, Ni, Co, Mo) during anaerobic biodegradation of organic matter is known to be fundamental for numerous reactions.The present studies on anaerobic digestion with different metal NP additives showed that methane production was affected by different metal NPs and by NP concentration.Nickel addition had a clear stimulating effect on the methane production from all tested substrates.This can be explained by the fact that nickel is an essential cofactor for a number of reactions, and nickel-containing enzymes are involved in at least five metabolic processes, including methanogenesis [78].Ragsdale [79] reported that nickel can exist in oxidation states from 0 to + 4 in biology, and the nickel redox enzymes appear to access either the 2 + /1 + and/or the 3 + /2 + redox couples.In addition, the zero-valence state is important during catalysis by acetyl-CoA synthase, and the hydrolases appear to only use Ni 2+ , which is a Lewis acid.However, our results showed uneven positive effects of Ni/SiO2 and Co/SiO2 NPs at different oxidation states.
The effects of NPs on anaerobic digestion were studied by many authors [25,26,80,81].Most studies focus on the antimicrobial function of NPs on methanogenic activities, which is related to the liberation of metal ions.Studies of catalytic effect of NPs on anaerobic digestion are limited to enhancement of fermentative bio-hydrogen production [28,32], and mechanisms of the NP effects are not clear.Qilin et al. [82] suggested that NPs can be sorbed onto cell surfaces and thereby alter the membrane properties (e.g., increase membrane permeability,) which leads to more efficient electron transfer to an acceptor.Beckers et al. [32] suggested that NPs played a role of active catalytic site involved in the production of hydrogen and are used by the bacteria for oxidation/reduction chemical reaction to help the bacteria transfer its electrons faster without consuming or metabolizing the metal, as when it is added to the medium in a dissolved form.An investigation was carried out in our lab on iron NPs at concentrations varying between 10 -4 and 10 -3 M using ICP-AES (inductively coupled plasma atomic emission spectrometry), which showed that 10 -7 M of iron ions were released from autoclaved iron NPs, and this concentration increased with time.The advantage of using metal NPs in anaerobic bio-methane production is that the NPs are encapsulated inside the porous silica matrix, which limits the metal ions' liberation in the culture medium.Thus, bacteria can uptake metals gradually, attaining sufficiently high intracellular metal concentrations to meet the demand of the metal enzymes and avoiding toxicity or other problems commonly associated with higher concentration of metals in bacteria and also metal precipitation.This secures the bacteria's need of metal for a longer time compared with the addition of other forms of metal, in addition to the extracellular role that may be played by NPs as active catalytic sites, as mentioned by [32].
In Table 7, the present study is compared to the literature by summarizing the most important parameters and results of each study concerning the production of bio-methane using metallic NP materials as catalysts.
It can be observed directly that making a comparison is quite difficult because many conditions such as the type of NPs, their concentration or the substrate differ greatly from one study to another.Nevertheless, it is observed that the Ni NPs are materials that lead to high methane production in this present study but also in Gran et al. [30] and Abdelsalam et al. [83].In Zhang et al. [16], the Fe NPs are also encapsulated in a matrix but in chitosan, compared to porous silica used in this present study.It is also observed that not much information about the NPs is given, such as electron microscope image or specific surface area.

Conclusions
In this work, seven types of encapsulated metal NPs were used as potential catalysts in thermophilic anaerobic digestion to produce methane.These NPs were produced thanks to a sol-gel method called cogelification.This process allows the dispersion of metallic NPs in a highly porous silica matrix, allowing a constant release of metal in water medium and avoiding the agglomeration of the NPs.All seven NPs showed this typical morphology.
Then, the seven metal NPs (Cu/SiO2, Pd/SiO2, Pt/SiO2, Ni/SiO2, Co/SiO2, Ag/SiO2 and Fe/SiO2) were used and compared for bio-methane production from acetate substrate.The cumulative methane production and production rate were investigated and modeled according to the Gompertz equation to identify the best catalysts.It was found that only Ni/SiO2, Co/SiO2 and Cu/SiO2 showed stimulatory effects on methane generation from acetate.
In the last part of this work, the two best catalysts, Ni/SiO2 and Co/SiO2, were used for biomethane production with two other carbon substrates: starch and glucose.The effects of NP concentration and the type of thermal pre-treatment of NPs on the methane production were investigated.The result was that the methane production was higher with Ni/SiO2 NPs.In addition, it was found that calcined NPs were the most efficient, almost evenly with dried ones.Similar profiles of intermediate fermentation products were recorded between Ni-and Co/SiO2 catalysts.The influence of those NPs on methane production rate was concentration dependent, increasing with the increase of metal NPs concentration from 10 -6 to 10 -4 mol L -1 .Finally, with the best catalyst, Ni/SiO2, a production of bio-methane 72.1% higher than the control was obtained.

Figure 3 .
Figure 3. Cumulative volume of methane produced during thermophilic anaerobic digestion of 5 g L -1 acetate monohydrate in batch test (serum vials 125 mL) with 10 -5 mol L -1 of encapsulated metal NPs of Cu/SiO2, Pd/SiO2, Pt/SiO2, Ni/SiO2, Co/SiO2, Ag/SiO2 and Fe/SiO2.The standard deviation bars are calculated based on the triplicate experiments made for each condition.

Figure 5 .
Figure 5. Cumulative volume of methane produced during second successive thermophilic anaerobic digestion of 5 g L -1 acetate monohydrate in batch test (serum vials 125 mL) with 10 -5 mol L -1 of encapsulated metal NPs of Cu/SiO2, Pd/SiO2, Pt/SiO2, Ni/SiO2, Co/SiO2, Ag/SiO2 and Fe/SiO2.The standard deviation bars are calculated based on the triplicate experiments made for each condition.

Figure 6 .
Figure 6.Cumulative volumes of (a) total biogas and (b) methane produced during thermophilic anaerobic digestion of 5 g L -1 soluble starch in batch test (serum vials 125 mL) with 10 -5 mol L -1 of encapsulated NPs of Ni/SiO2, Co/SiO2 and Cu/SiO2.The standard deviation bars are calculated based on the triplicate experiments made for each condition.

Figure 8 .
Figure 8. Catalytic effects of different valence states of nickel NPs -(a) calcined Ni/SiO2 NPs, (b) dried Ni/SiO2 NPs and (c) reduced Ni/SiO2 NPs at concentrations of 10 -6 , 10 -5 and 10 -4 mol L -1 -on methane production compared with the effect of (d) NiCl2 addition at the same concentrations.The standard deviation bars are calculated based on the triplicate experiments made for each condition.

Figure 9 .
Figure 9. Catalytic effects of different valence states of cobalt NPs-(a) calcined Co/SiO2 NPs, (b) dried Co/SiO2 NPs and (c) reduced Co/SiO2 NPs at concentrations of 10 -6 , 10 -5 and 10 -4 mol L -1 -on methane production, compared with the effect of (d) CoCl2 addition at the same concentrations.The standard deviation bars are calculated based on the triplicate experiments made for each condition.

Figure 10 .
Figure 10.The changes of metabolic concentration (mM) during thermophilic anaerobic digestion of 5 g L -1 of glucose monohydrate with 10 -4 mol L -1 of (b) calcined Ni/SiO2 and (c) calcined Co/SiO2 NPs compared with (a) control without NPs.

Table 1 .
Composition and specific surface area of NP samples.
SBET = specific surface area calculated with BET method from nitrogen adsorption-desorption isotherm data.

Table 2 .
Kinetic parameters and Gompertz coefficient adjusted on the profiles of volumetric methane production curves for the BMP tests with 10 -5 mol L -1 of NPs and 5 g L -1 acetate monohydrate.

Table 6 .
Kinetic parameters and Gompertz coefficient adjusted on the profiles of volumetric methane production curves for the BMP tests with different valence states of Co/SIO2 NPs and CoCl2 at different concentrations (10 -6 , 10 -5 and 10 -4 mol L -1 ).

Table 7 .
Literature comparison of methane production in anaerobic digestion catalyzed by metallic NPs.