Bioresource Technology 372 (2023) 128632 Contents lists available at ScienceDirect Bioresource Technology journal homepage: www.elsevier.com/locate/biortech Magnetite-based nanoparticles and nanocomposites for recovery of overloaded anaerobic digesters Raquel Barrena a, *, María del Carmen Vargas-García b, Paula Catacora-Padilla a, Teresa Gea a, Ahmad Abo Markeb a, Javier Moral-Vico a, Antoni Sánchez a, Xavier Font a, Thomas J. Aspray c, d a GICOM Research Group Department of Chemical, Biological and Environmental Engineering, Universitat Autònoma de Barcelona Edifici Q, Carrer de les Sitges 08193 Bellaterra (Cerdanyola del Vallès), Barcelona, Spain b Department of Biology and Geology, CITE II-B Universidad de Almería CEIMAR Marine Campus of International Excellence, 04120 Almería, Spain c School of Energy, Geoscience, Infrastructure and Society, Heriot-Watt University, Edinburgh EH14 4AS, Scotland, UK d Solidsense Ltd, Bearsden, East Dunbartonshire G61 3BA, Scotland, UK H I G H L I G H T S G R A P H I C A L A B S T R A C T • Iron-based nanomaterials were tested in anaerobic granular and flocculated sludge. • Magnetite-based nanomaterials were able to recover overloaded anaerobic reactors. • VFA inhibition were reversed in 30–50 days using nanomaterials. • Microbiome in granular and flocculated sludge were reestablished with modifications. • Promoted microorganisms changed the dominance towards hydrogenotrophic species. A R T I C L E I N F O A B S T R A C T Keywords: The effect of magnetite nanoparticles and nanocomposites (magnetite nanoparticles impregnated into graphene Iron-oxide nanoparticles oxide) supplement on the recovery of overloaded laboratory batch anaerobic reactors was assessed using two Granular sludge types of starting inoculum: anaerobic granular sludge (GS) and flocculent sludge (FS). Both nanomaterials Flocculent sludge recovered methane production at a dose of 0.27 g/L within 40 days in GS. Four doses of magnetite nanoparticles Hydrogenotrophic methanogens VFA inhibition from 0.075 to 1 g/L recovered the process in FS systems between 30 and 50 days relaying on the dose. The presence of nanomaterials helped to reverse the effect of volatile fatty acids inhibition and enabled microbial communities to recover but also favoured the development of certain microorganisms over others. In GS reactors, the methanogenic population changed from being mostly acetoclastic (Methanothrix soehngenii) to being domi­ nated by hydrogenotrophic species (Methanobacterium beijingense). Nanomaterial amendment may serve as a preventative measure or provide an effective remedial solution for system recovery following overloading. * Corresponding author. E-mail address: raquel.barrena@uab.cat (R. Barrena). https://doi.org/10.1016/j.biortech.2023.128632 Received 18 November 2022; Received in revised form 12 January 2023; Accepted 13 January 2023 Available online 16 January 2023 0960-8524/© 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by- nc-nd/4.0/). R. Barrena et al. Bioresource Technology 372 (2023) 128632 1. Introduction inocula, which provide different microbial communities. However, to our knowledge, studies that merge both factors together do not exist. For Anaerobic digestion (AD) is a well-established resource recovery this reason, the aim of this work was to study the influence of the type of technology used commercially around the world. This is especially the inoculum in the recovery of overloaded reactors when magnetite-based case in Europe where AD is playing an increasingly crucial role in the NPs are added. Two different type of starting inoculum, granular sludge energy transition, diversifying the energy mix with renewable energy (GS) and flocculent sludge (FS), commonly used in suspended and sources (Commission and Communication, 2021). Despite its potential, immobilised biomass of AD, were used and the shift in microbial com­ many industrial plants do not operate to their maximum capacity in munities assessed. Instead of commercial magnetite NPs, we carefully order to avoid operational problems. For example, operational insta­ synthesized the NPs and characterized their size distribution. Simple bility due to fluctuations in organic loading is a common issue. High magnetite NPs were compared with magnetite NPs impregnated into organic loading rates (OLR) and high content of easily biodegradable graphene oxide due to the promising effects of this material previously substrates can lead to rapid acidification and accumulation of interme­ described. The improvement in VFA removal efficiency could be a sig­ diate volatile fatty acids (VFA). The monitoring of VFA to prevent nificant factor to enhance process stability. This, coupled with an in­ accumulation in anaerobic digesters is a fairly standard practice. crease in methane yield when using iron-based nanomaterials, opens a The use of conductive iron oxides, particularly magnetite, has been new strategy to have a ‘doubling effect’ when working at high OLR. reported as a strategy to enhance the methanogenic degradation of VFA (digestion intermediates) under ammonia stressed conditions (Lee et al., 2. Materials and methods 2019) leading to an increase in methane production. For example, Baek et al. (2017) found that the addition of magnetite (particle size 100–700 2.1. Synthesis and characterization of nanoparticles nm) resulted in a more resistant and resilient operation to process im­ balances, especially effective in avoiding propionic acid accumulation. Magnetite NPs were synthesized by the coprecipitation method using The enhanced biomethanation was attributed to the facilitated direct tetramethylammonium hydroxide (TMAOH) with slight modifications interspecies electron transfer (DIET) between the acetogens and (Abo Markeb et al. 2019). TMAOH was used to increase the colloidal methanogens. DIET between syntrophic microbes associated directly stability of the NPs as well as improve the degree of crystallinity. Briefly, with the anaerobic degradation of VFA were also reported by Yan et al. 40 mmol of iron (III) chloride hexahydrate (FeCl3⋅6H2O), and 20 mmol (2017) using conductive materials. The use of magnetite nanoparticles of iron (II) chloride (FeCl2) were dissolved in 200 mL of deoxygenated (NPs) has also been associated with an increase in both VFA formation water for 20 min. The iron salts solution mixture was added dropwise and subsequent degradation. Zhang et al. (2020) reported a 2.5-fold into 200 mL of 1 M deoxygenated TMAOH at room temperature until the increase in the formation of VFA and an enhancement in acetoclastic colour of the suspension changed to black. After this, the mixture sus­ methanogenesis when 100 mg/L of magnetite was used. pension was vigorously stirred (600 rpm) for 30 min under an inert In general, the use of conductive additives has been proven to be nitrogen gas stream. Lastly, the Fe3O4 NPs were washed three times effective in improving AD performance promoting DIET (Zhang et al., using magnetic decantation and then redissolved into 1 mM of TMAOH 2018). Of the conductive additives tested to date, graphene has been to obtain the final suspension of Fe3O4 NPs. shown to cause higher electron transfer efficiency in DIET because of its The magnetite nanoparticles impregnated into graphene oxide, higher conductivity and smaller size than other carbon-based conduc­ Fe3O4-rGO nanocomposites (NCs), were prepared in two steps. Reduced tive materials (Zhang et al., 2018). Other novel materials with electron graphene oxide (rGO) was prepared in the first step as follows: 1 g of donating capacities such as biochar are currently being investigated graphite powder was mixed with 0.5 g of sodium nitrate in 23 mL of with promising initial results (Sun et al., 2022). H2SO4 in an ice bath. Then, 3 g of potassium permanganate (KMnO4) The effect of the incorporation of NPs in the biochemical aspects of was added dropwise after the mixture was stirred for 30 min. Next, the the process has been related to the affectation of the microbiome among mixture solution was continuously stirred for 4 h in an ice bath (<10 ℃). other reasons (Wang et al., 2021). Several factors of the AD process have Afterwards, the reaction was terminated by the addition of 4 mL of been investigated in terms of their influence on the microbial commu­ H2O2, and 146 mL of deionized water. Hence, the graphene oxide pro­ nity (Kong et al., 2018). The relative abundance of methanogens, their duced was separated by centrifugation at 9000 rpm for 20 min, followed composition in terms of acetogenic and hydrogenotrophic species, and by washing with HCl (0.01 M), and deionized water. Lastly, the gra­ the presence of syntrophic bacteria, are basic aspects in the control of phene oxide (GO) was dried at 60 ℃ for 24 h. Next, 0.5 g of the dried GO AD. It has been shown that the microbial community structure can was dispersed in 100 mL of deionized water by ultrasonication for 2 h. change depending on the presence and characteristics of the NPs, Next, 2 mL of ascorbic acid solution (2.2 mM) was added dropwise into favouring an increase or decrease in the diversity of both eubacteria and the GO suspension. The resulting rGO was washed three times with archaea (Wang et al., 2021). In the case of magnetite, the results point to deionized water by centrifugation at 9000 rpm for 20 min, followed by low toxicity, mainly due to the stimulating effect of the presence of drying at 60 ℃ for 24 h. In the second step, magnetite nanoparticles Fe2+/Fe3+ ions on microbial activity (Arif et al., 2018) that promote coated with cetyltrimethylammonium bromide (CTAB) were synthe­ DIET. Ma et al. (2021) analysed the changes in microbial community sized in presence of the previously described rGO in order to obtain dynamics and found that magnetite caused different syntrophic in­ Fe3O4 NPs impregnated into rGO by using the in-situ co-precipitation teractions depending on the OLR. However, this effect may be regulated slightly modified method of the previously reported study (Abo Markeb by the concentration of NPs added (Salama et al., 2021). et al., 2019). Briefly, FeCl2⋅4H2O (25 mM) and FeCl3⋅6H2O (50 mM) in Besides the properties of the NPs, the use of different starting inocula 1:2 M ratio of Fe2+/Fe3+, were dissolved in pre-sonicated and deoxy­ can influence the response of the various microbial communities present genated 100 mL of ultrapure water (Milli-Q) containing 0.1 g of CTAB, in the process; however, this has been scarcely studied. Studies have 0.1 g of rGO. Then, the mixture solution was incubated for 1 h at 40 ◦ C found greater process stability in prokaryotic populations associated under inter atmosphere by purging with nitrogen. Next, 0.6 M of NH4OH with aerobic granular sludge than of flocculent activated sludge, which solution, previously purged with N2 for 1 h, was added dropwise into the translates into a lower affectation of biochemical and metabolic events mixed solution under agitation and nitrogen atmosphere until a pH of (Daraei et al., 2019). Moreover, the composition of the initial micro­ 9.0 was achieved. During the titration process, the black colour biome is already different between GS and FS (De Vrieze et al., 2015). confirmed the formation of Fe3O4 NPs. Then, the Fe3O4-rGO NCs sus­ The potential effect of NPs on the different microbiome has not been yet pension was stirred continuously under a nitrogen atmosphere for 1 h at compared. Therefore, there are studies that report the influence on AD of 40 ◦ C. Afterwards, the NPs were washed three times using ultrapure either the incorporation of magnetite NPs or the use of different types of water followed by ethanol using magnetic a neodymium magnet and, 2 R. Barrena et al. Bioresource Technology 372 (2023) 128632 then dried at 60 ◦ C for 12 h. S.A. (Valencia, Spain). PowerSoil™ DNA Isolation kit (MoBio Labora­ The synthesized nanomaterials were characterized in terms of size, tories Inc., Carlsbad, CA, USA) was used for the extraction of DNA as per distribution and morphology using a Zeiss Merlin Scanning Electron the manufacturers’ instructions. Sequencing data obtained from the Microscope (SEM) and using a JEM-2011/JEOL, High-Resolution Illumina MiSeq platform (Illumina, Ic., San Diego, CA, USA) were ana­ Transmission Electron Microscope (HR-TEM) equipped with Energy- lysed according to our previous works (Barrena et al., 2021). Dispersive Spectroscopy (EDS). Besides, selected areas electron diffrac­ tion images (SAED) were used to study the microstructure of the ma­ 2.6. Statistical analysis terial with high resolution images. ImageJ free software was used to obtain the size distribution histogram of magnetic NPs from TEM Statistical analysis was based on one-way ANOVA (p < 0.05 confi­ images. dence) with the Tukey test. Significant differences were analysed using Minitab 16 (Minitab Inc.). The alpha diversity indices of the different 2.2. Inoculum source microbial communities were obtained from the EzBioCloud microbio­ logical research platform (https://www.ezbiocloud.net), while the Two types of different mesophilic inoculum from full-scale meso­ SIMPER and Indicator Species analyses, as well as the dendrogram, were philic anaerobic digesters were used in batch experiments: anaerobic performed using the PAST software for statistical analysis of biological granular sludge (GS) obtained from an up-flow anaerobic sludge blanket data (PAST 4.05). The phylogenetic tree showing the evolutionary (UASB) digester treating wastewater of a local whisky distillery in relationship of the dominant amplicon sequence variants (ASVs) was Edinburgh (Scotland, UK); and anaerobic flocculent sludge (FS) from the generated from MEGA version 10.1.6 software, using the Neighbour- mesophilic anaerobic digester of Riu Sec municipal wastewater treat­ Joining method combined with the Maximum Composite Likelihood ment plant (WWTP) (Sabadell, Barcelona, Spain). The total solids (TS) method to compute evolutionary distances. content and volatile solids (VS) content of GS were 3.30 ± 0.01 % and 70.1 ± 0.1 %, respectively, while for FS were 2.70 ± 0.06 % and 64.11 3. Results and discussion ± 0.08 %, respectively. 3.1. Nanomaterials characterisation 2.3. Anaerobic digestion (AD) batch experimental set-up and procedure The size of synthesised magnetite ranged from 6 to 13 nm, with 60 % Biochemical methane potential (BMP) tests were conducted using falling between 8 and 10 nm (Fig. 1A, B). As seen in Fig. 1D and 1E the glass serum bottles (250 mL) containing inoculum and glucose with/ synthetized materials are highly crystalline in nature. In particular, without magnetite-based NPs. The volume of liquid was adjusted in all Fig. 1C shows a high-resolution TEM image in which crystalline planes bottles with water as necessary. pH was adjusted to ~8.0 using 0.1 mL of of a single nanoparticle are clearly described, which is confirmed by the citric acid (300 g/L) as appropriate. Glucose was added in sufficient SAED image of Fig. 1E, where the typically reported pattern for amount to inhibit methanogenesis and induce acidification. The inoc­ magnetite can be distinguished (Cai and Wan, 2007). This pattern was ulum to substrate ratio (ISR) of 2:1 based on vS content was enough to less clear when NPs are analysed in a sample taken from the anaerobic obtain the inhibitory conditions in GS (6.4 g/L). For FS, a higher amount digester (Fig. 1F) assumed due to the presence of inoculum and of glucose was needed to inhibit the process and an ISR of 1:1 was substrate. required (21.3 g/L). Previous experiments showed that these concen­ The presence of graphene cannot be confirmed with this analysis due trations were suitable for sustained inhibition of the complete AD pro­ to its amorphous structure, but as observed in Fig. 1C, the impregnation cess. The headspace of each bottle was purged with nitrogen gas for one of NPs with rGO causes the image of NPs not to be as clear as non- minute to ensure anaerobic conditions. The bottles were incubated at impregnated NPs (Fig. 1A) and new structures corresponding to gra­ 37 ◦ C with continuous mixing for experiments using GS (orbital incu­ phene oxide appear. SEM images were analysed to determine the bator, 110 rpm) and intermittent mixing (turn up down twice/three interaction established between nanomaterials and inoculum (see sup­ times a week) for FS. plementary material). Blank assays containing inoculum without glucose were used to subtract the background methane production from the inocula. Positive 3.2. Recovery of overloaded batch reactors inoculated with granular controls using microcrystalline cellulose were also established to vali­ sludge (GS) date inocula activity (Holliger et al., 2016). Each experiment treatment/ control was setup in triplicate as a minimum. For GS, Fe3O4 NPs and The specific methane production in the experiment using GS is Fe3O4-rGO NCs were added to glucose-fed reactors in a concentration of shown in Fig. 2A. As expected, the amount of glucose added to the 0.27 g/L as previously used by the same authors (Casals et al., 2014). reactor was enough to inhibit the process. Methane content in biogas BMP of TMAOH at the same concentration as the final suspension of was <30 %, suggesting the inhibition of methanogenic microorganisms. Fe3O4 NPs was also evaluated in glucose-fed reactors to discard any Low methane content was maintained during all the process in control effect of this additive on the process. For FS, four doses of magnetite NPs reactors with glucose. However, as observed in Fig. 2A, reactors sup­ from 0.075 to 1 g/L were tested in glucose-fed reactors. The experiments plemented with magnetite NPs recovered from the inhibition after 40 were carried out for at least 100 days irrespective of inoculum type. days. The specific methane production rose and the concentration of methane in biogas increased to 60–65 %. Even though this is a signifi­ 2.4. Analytical methods cant period in a discontinuous digestion, these results are encouraging as per the implications of the use on NPs or NCs at a commercial scale. This TS and vS content and pH were determined according to standard suggests the hypothesis that a continuous digester microbiome adapted methods. VFA samples were analysed by high-pressure liquid to the presence of NPs would potentially have the ability to overcome chromatography-ultraviolet detection after filtering (0.22 μm) as pre­ drastic changes in OLR without the need of long adaptation periods, to viously described in Barrena et al. (2021). Biogas composition was be confirmed in future experiments. analysed by gas chromatography as described in Casals et al. (2014). Fig. 2B shows VFA content in reactors 45 days after NP supplemen­ tation and comparative controls. As observed, the total VFA in control 2.5. Microbial community analysis reactors was 3100 mg/L, almost entirely composed of acetic and butyric acid (each 1500 mg/L), and more than double the threshold (1500 mg/L Sequencing and bioinformatics were performed by Life Sequencing total VFA) considered to guarantee a stable digestion process 3 R. Barrena et al. Bioresource Technology 372 (2023) 128632 Fig. 1. Nanoparticles characterisation: TEM images: A) magnetite nanoparticles, B) distribution size histogram of magnetite NPs obtained from TEM images, C) Magnetite NPs impregnated with rGO, D) high resolution image of magnetite nanoparticles, E) SAED pattern of magnetite nanoparticles, F) SAED pattern of magnetite nanoparticles after being added to an anaerobic digester. (Angelidaki et al., 2005). Siegert and Banks (2005) reported that 100 NL methane kg− 1 VS) being clearly affected by the inhibitory con­ fermentation of glucose was inhibited at VFA concentrations above ditions. No signs of recovery were observed for more than 100 days. In 4000 mg/L. The accumulation of VFA is thought to explain the failure of contrast, the recovery of supplemented reactors was accompanied by the control reactors. They presented a very low methane production (around reduction of total VFA, presenting a total amount lower than 80 mg/L 4 R. Barrena et al. Bioresource Technology 372 (2023) 128632 A) B) Fig. 2. A) Evolution of methane production of overloaded reactors with and without iron-based NPs (NPs and NCs) at a dose of 0.27 g/L in granular sludge; B) Volatile fatty acid (VFA) content (mg/L) after the recovery of reactors supplemented with NPs. 5 R. Barrena et al. Bioresource Technology 372 (2023) 128632 (Fig. 2B). After the recovery, reactors with NPs presented a specific enhancement effect of Fe3O4-rGO NCs was superior. Concerning the methane production close to the theoretical production for a carbohy­ recovery of the AD systems studied in this work, no significant differ­ drate (415 NL methane kg-1VS). ences were observed for time recovery or methane production. Thus, the The methane production profile obtained in the reactors supple­ addition of graphene did not improve the effect of magnetite NPs con­ mented with NCs is also shown in Fig. 3a. As observed, the addition of firming the potential use of this more simple and lower cost material. NCs had a similar effect to that of the NPs in the recovery of methano­ The microbiological study confirmed these minimal differences between genesis. Other authors pointed out differences in the effects of using the NCs and NPs (see section 3.4). To author’s knowledge, the effect of NCs same type of NPs and NCs in biological processes. For instance, Sun et al. on the anaerobic digestion performance has been scarcely studied and it (2021) observed that both nanomaterials could intensify butyrate-type is interesting to have new data for further applications. fermentation and other positive aspects in dark fermentation, but the Methane production of the assay containing TMAOH (1 mM) did not Fig. 3. A) Evolution of methane production of overloaded reactors with and without NPs at different doses in flocculent sludge; B) Volatile fatty acid (VFA) content (mg/L) after the recovery of reactors supplemented with NPs. 6 R. Barrena et al. Bioresource Technology 372 (2023) 128632 show any differences in methane production in comparison with the concentrations via the supplementation of zero valent iron NPs. The control (Fig. 2A). The process did not recover over the time the test was effective use of NPs according to the level of inhibition should be carried out (100 days). The effect of stabilizers used in NPs synthesis is carefully addressed to obtain a reliable dose in the recovery of the not usually described in works dealing with magnetite NPs and AD. process. Dosage adjustment has been previously pointed out as a strat­ However, it has been previously reported that TMAOH could be egy to maintain the positive effects of iron NPs in semicontinuous anaerobically degraded (Whang et al., 2014), which could imply an anaerobic digesters, where higher doses sustained positive effects for a increase in biogas production or also present other side effects. How­ longer time (Barrena et al., 2021). ever, to discount other effects of TMAOH, the microbial community of As observed in this work, the use of magnetite-based nanoparticles these bioreactors were also analysed and compared with other samples allowed for the recovery of the process under conditions of overloading as described below (section 3.4). for both GS and FS inoculants. In that sense, the study of the microbial After 100 days of process, NPs and NCs were added to glucose population structure of both inocula before and after the recovery ap­ inhibited reactors (data not shown). Interestingly, when NPs and NCs pears interesting to know the applicability of nanomaterials as an ad­ were added, anaerobic digesters recovered from failure within 40 days, ditive to correct and prevent process imbalances of different types of the same time observed in all previously recovered processes (Fig. 2A). anaerobic digesters. As no difference was observed between supplementing the magnetite- based NPs at the start or during the test, the use of magnetite NPs and 3.4. Effect of process conditions on microbial community: granular sludge NCs could have potential application as a preventative additive or remedy following failure due to overloading. The results derived from the biodiversity analysis for GS (Fig. 4) To demonstrate the preventive effect of NPs and NCs against over­ indicate the loss of diversity in inhibited reactors compared to the loading, recovered reactors were newly fed with a pulse of glucose control. As observed, the microbiome associated with AD subjected to (using the same amount initially fed, 6.5 g/L). In this case, no inhibition VFA inhibition was recovered, at least partially, with the incorporation was observed after the feed (Fig. 2A, New pulse). Biogas production of NPs. Specifically, the alpha diversity values obtained showed the resumed immediately after the pulse without presenting a pronounced similarity in this respect between the control and the samples supple­ lag phase. Methane content in biogas was also in an appropriate range, mented with NPs. Thus, the values found in the GS processes were around 60 %, indicating a non-inhibited process. In this sense, as pre­ around 700–800, 500–600 and 4, for CHAO1, number of ASVs and viously pointed out, the presence of NPs endows the system with certain Shannon, respectively, except for the process inhibited, for which the characteristics that promote microbial activity in relation to substrate respective levels were 420, 294 and 2.40. These values are similar to degradation, such as increases in surface area, reactivity and the number those described in literature in relation to the microbial community of active sites (Ellacuriaga et al., 2021). In particular, magnetite NPs are present at the end of a batch AD (Wang et al., 2021). Slight differences postulated as accelerators of methane production, by favouring the are likely due to the nature of the inoculum and the way the process is transfer of electrons between the species that supply the compounds operated (Li et al., 2020). required by methanogens (Zhong et al., 2020). Thus, the effect caused by The results from the study of beta diversity also support the hy­ the NPs seems to favour a more adapted and resilient microbial com­ pothesis that the use of NPs in inhibited anaerobic digesters allowed a munity composition (Lim et al., 2022), able to better withstand further certain degree of recovery of the microbiome, increasing the similarity perturbations. of the communities present, both qualitatively and quantitatively, with respect to those associated with non-inhibited processes. Thus, the 3.3. Recovery of overloaded batch reactors inoculated with flocculent values found for the Sorensen-Dice index (Fig. 4b) showed higher levels sludge (FS) of homology between the final microbiome of the control process and that of the processes with NPs presence compared to the inhibited pro­ Fig. 3A shows methane evolution in the batch test using FS. In this cesses. Microbial homogeneity levels were also increased in presence of experiment different doses of NPs (from 0.075 g/L to 1 g/L) were the stabilizer (TMAOH), although in in a lower extend than in presence studied. As observed in Fig. 3B, reactors supplemented with magnetite of/with NPs. This lower capacity of the stabilizer to return the microbial NPs recovered from inhibition within 30 and 50 days depending on the community to the population levels of the control may explain to a large dose. Significant differences were observed both in time and biogas extent the non-recovery of the process (in biogas production terms), an production depending on the dose evaluated. The highest dose (1 g/L) aspect discussed in section 3.2. These patterns observed for the different showed faster recovery and higher methane production than lower types of AD were repeated both in terms of species identity and relative doses. In addition, as observed in Fig. 3B, the dose of 1 g/L after the abundance, although the similarity values were always higher in the first recovery showed the minimum VFA levels of all samples evaluated. The case. This may be a consequence of a redistribution of the relative second highest dose (0.5 g/L) was also effective at recovering the pro­ abundances of species, so that the presence of the NPs favours the cess although more time was needed (around 40 days). VFA reduction development of certain microorganisms over others that, without being was also evident in comparison with control reactors with no NPs. inhibited or disappearing, show a lower tolerance to the NPs, as sug­ However, only one triplicate reactor of dose 0.25 g/L successfully gested by Eduok et al. (2017). recovered the process. This replicate also showed a reduction of VFA in The analysis performed on the most dominant microbial community Fig. 3B (25_R) whereas the others had the same amount of VFA as the members (relative abundance greater than 1 % in at least one of the control reactors with no NPs. In fact, lower doses were not effective in processes) showed similar conclusions (Fig. 5A) to those described for the recovery. Different results obtained using the dose of 0.25 g/L in­ the entire microbiome (Fig. 4). Thus, the inhibited process was clearly dicates the threshold of applicability of this dose according to the level differentiated from the others, all of them grouped in the same cluster, of inhibition. VFA concentrations in experiments inoculated with FS which confirms the recovery action of the NPs. Comparing the relative were much higher than the ones observed in experiments using GS, abundances of the dominant ASVs present in each of the processes, it can where NPs concentration of 0.27 g/L was effective in the recovery of all be observed how the high addition of glucose decreased both the relative inhibited replicates. However, despite the high levels of total VFA found proportion and diversity of methanogens, and intensely enhanced the in FS reactors (around 18000 mg/L) the system recovered when NPs abundance of representatives of the genus Clostridium, especially were supplemented. As previously indicated, a higher amount of glucose C. butyricum, which reached almost 45 % of the total population. was necessary to inhibit FS reactors (section 2.3). This fact may lead to a Members of the phylum Firmicutes, practically absent in the control high VFA accumulation in reactors without NPs supplementation. process, were also detected in the processes with the presence of NPs, Recently, Lim et al. (2022) recovered reactors with similar VFA although the relative abundances found did not exceed 5 % in any case. 7 R. Barrena et al. Bioresource Technology 372 (2023) 128632 Fig. 4. Diversity indexes found for the different anaerobic digestions analysed; A) Alpha diversity: Chao1, number of ASVs and Shannon; B) Beta diversity: Sorensen- Dice (the qualitative index is shown in the upper diagonal and the quantitative index in the lower diagonal). Although a specific taxonomic affiliation could not be obtained for incorporation of NPs. This recovery was especially remarkable in the several of them, their evolutionary relationships place them in the case of NCs, not only in terms of the percentage of methanogens with clostridia group. Many clostridia are characterized as syntrophic acetate respect to the total population, but also in terms of the similarity of both oxidizers (Westerholm et al., 2019), capable of providing H2 and CO2 to communities, since it was the only case in which the quantitative ho­ the hydrogenotrophic methanogens (Werner et al., 2014). This is mology exceeded 50 %, as indicated by the Sorensen-Dice index values consistent with the redistribution observed in the methanogenic com­ (Fig. 5C). In addition to its contribution as an additional conductive munity (Fig. 5A-B), where the dominance of Methanothrix soehngenii in material, graphene has been reported to be able to recover processes the control process, a species widely recognized as an acetoclastic with high VFA production, characterized by the generation of exces­ methanogen, was replaced in the NPs-bearing processes by that of sively acidic pH values, not suitable for methanogenic communities (Wu Methanobacterium beijingense, which produces methane from H2. The et al., 2020). explanation for this effect seems to be related to the ability of some On the other hand, the analysis of the entire methanogenic com­ species of Methanobacterium to accept electrons directly, an activity munity also explains the lower recovery shown by the process that recently associated with species of this genus (Zheng et al., 2020), and incorporated the stabilizing agent. In this case, and although the ho­ the stimulus that NPs generates on this ability because of its conductive mology with respect to the control process was high (above 87 %) at the capability (Vu et al., 2020). This stimulating effect has been observed level of taxonomic affiliation of the ASVs present, the quantitative dis­ even in processes with high OLR, as in the present case (Zhao et al., tributions of such ASVs in both cases led to a decrease in the observed 2017). On the other hand, and although studies in this sense are scarce, similarity to values below 43 %. On the contrary, the degree of quan­ it is also necessary to consider the possibility that NPs with conductive titative homology with respect to the process supplemented with glucose behaviour directly affect the activity of methanogenic species, without reached a level close to 81 %. In both processes, the dominant metha­ the intervention of other species (Salvador et al., 2017). nogens were Methanobacterium beijingense and M. petrolearium, species The ability of NPs to partially recover the AD process can be that have been reported to be unable to utilize methyl (Ma et al., 2005), appreciated more directly by studying the entire methanogenic com­ or to do so without parallel methane production detected (Mori and munity (Fig. 5B-C), since this is the community most sensitive to the Harayama, 2011). occurrence of disruptive events such as substrate inhibition and VFA accumulation (De Vrieze et al., 2015). It can be seen how the abundance of species belonging to the methanogenic community partially recov­ ered from the drop experienced in the presence of glucose because of the 8 R. Barrena et al. Bioresource Technology 372 (2023) 128632 Fig. 5. Analysis of the prokaryotic populations associated with the different anaerobic digestion processes carried out; A) Evolutionary relationships between the species present with relative abundance levels above 1 % in at least one of the processes, relative abundances associated with each of them and dendrogram showing the grouping of the different digestions according to this dominant microbiota; B) Structure of the different methanogenic communities. The legend shows in brown the species associated to the processes with GS, in blue those associated to the processes with FS and in gray the shared ones. Those with relative abundances greater than 1 % are highlighted in bold; C) Sorensen-Dice index for the methanogenic communities Diversity indexes found for the different anaerobic digestions analysed; A) Alpha diversity: Chao1, number of ASVs and Shannon; B) Beta diversity: Sorensen-Dice (the qualitative index is shown in the upper diagonal and the quantitative index in the lower diagonal). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) 3.5. Effect of process conditions on microbial community: Flocculent methanogenesis (Sposob et al., 2021), during which it generates H2 in sludge addition to acetic and lactic acid. This favours the action of both ace­ togenic and hydrogenotrophic methanogens, as in the case of the process Inoculum type has been pointed out in numerous studies as the most with NPs. influential factor for the microbiota of AD processes, even above others Differences were also manifested in relation to the phylum Firmi­ such as the inoculum origin or the operating system. Although in the cutes, important in both types of processes, with dominance of ASVs long-term the composition of the microbial community tends to belonging to the Clostridia group. However, in the GS process most of converge (Peces et al., 2018), except for the methanogenic community the ASVs were identified as members of the genus Clostridium, contrary (Wojcieszak et al., 2017). The results presented herein support the to the FS process. In FS process without NPs the dominant species was importance that the inoculum has on the composition of the microbial Sedimentibacter hydroxybenzoicus. By comparison, in the FS process with community associated with AD. This is shown by the Sorensen beta di­ NPs, dominant species were Calaromator australicus and, to a greater versity index (Fig. 4), for which the values found between processes extent, Caldalkalibacillus thermarum, assigned to the phylum Bacilli. inoculated with GS and FS did not exceed 15 % and 6 % in any case, in S. hydroxybenzoicus is associated with the production of acids, especially qualitative and quantitative terms, respectively. The microbiome asso­ acetic acid (Klang et al., 2019), which supports the fundamentally ace­ ciated with the FS process was characterized by a greater diversity in toclastic behaviour of the control process. C. thermarum and terms of the number of phyla detected, with the presence of represen­ C. australicus, both thermophilic species, have been previously found in tatives assigned to Proteobacteria, Bacteroidetes and Chloroflexi, barely AD systems, the former being able to degrade carbohydrates and VFA significant in the case of the GS. According to the literature, these phyla (Yan et al., 2020), and the latter, detected in H2-producing environ­ appear to be common in processes seeded with FS (Ribera-Pi et al., ments, producing acetic acid and ethanol (Mohd Yasin et al., 2011). 2020). For example, Ottowia pentelensis, the major proteobacterium in all When the methanogenic community was considered exclusively, the three processes, is recognized for its ability to form flocs, which gives it a similarity between processes seeded with different inocula increased prominent role in this type of sludge. Moreover, its hydrolytic activity slightly, with maximum levels of 48 % and 18 %, respectively (Fig. 5c). (Feng et al., 2016) allows it to participate in the first stage of the AD However, the response caused by the incorporation of NPs to the process. Petrimonas sulfuriphila is also considered a hydrolytic species, inhibited process, concerning the recovery of biogas production capac­ with production of acetic acid, H2 and CO2. This bacterium of the ity, was similar to that described in the case of GS. On this occasion, the phylum Bacteroidetes has been detected in relatively high abundance in recovery seems to be a consequence of the reestablishment of part of the an AD process amended with magnetite (Ziganshina et al., 2021), as in methanogenic species, since it went from 21 % of similarity between the the present case. Among the representatives of the phylum Chloroflexi, inoculum and the inhibited process, to 51 % detected between the Levilinea saccharolytica was the most prominent species. This bacterium inoculum and the process with added NPs. In fact, the methanogenic seems to have a prominent role in the acidogenic stage of community in the inhibited process was practically non-existent, 9 R. Barrena et al. Bioresource Technology 372 (2023) 128632 accounting for<1 % of the total community. Methanothrix soehngenii, at of the inoculum. After the exposure to nanomaterials, the results were 0.32 % relative abundance, was the most abundant methanogenic spe­ also different. Specifically, the results obtained in relation to the cies. It was also dominant in the other two processes, but the quantified composition of the different microbial communities were studied by relative abundance was 3.34 % and 1.30 % for the control process and applying SIMPER and Indicator Species analyses, which report the with NPs, respectively. These results apparently differ from those dis­ percentage contribution of each species to the level of dissimilarity be­ cussed in the case of GS process, where the presence of NPs changed the tween processes and the identity of the most significant species in each nature of the methanogenic population from being mostly acetoclastic to case, respectively. As indicated by the values of the beta diversity being dominated by a hydrogenotrophic species. However, it should be indices, the most marked differences in the microbiome were ascribed to mentioned that the second most abundant species in the methanogenic the type of inoculum (Fig. 6). The number of species significantly community of the FS process recovered with NPs was Methanoculleus highlighted in terms of type of inoculum was clearly higher than that submarinus, a hydrogenotroph (Mikucki et al., 2003), not detected in the detected in the case of the presence of NPs and the recovery effect. In the control process. The presence of L. saccharolytica and, especially, GS process, the statistically dominant microbiota was mostly related to P. sulfuriphila in the NPs process, two H2-producing bacteria, could methanogenic species as well as representatives of the genus Clostridium, explain the detection of this methanogen. The low growth rate of the while inoculation with FS resulted in a much more diverse community in methanogenic community could explain the long recovery times terms of the taxonomic affiliation of the significant species. Particularly observed in both batch experiments. striking is the non-significance of methanogenic species in the FS pro­ cess, something already described by other authors, who associate this 3.6. Comparative analysis of the microbial community as a function of fact to the concurrence of better conditions for this prokaryotic group in the factors under study: SIMPER and Indicator species analyses GS (Winkler et al., 2012). The presence of NPs as a differentiating criterion led to the identi­ Both microbial communities for FS and GS processes were studied in fication of only one species as significant, Lutispora thermophila. This detail. The study involved the initial microbial communities and their representative of the phylum Firmicutes only appeared at prominent evolution following exposure to the nanomaterials. Differences were levels in the samples in which NPs were incorporated. This is a species found between the community at the start and end of the processes. In mostly related to amino acid fermentation (Venkiteshwaran et al., the case of the initial microbial communities, this was due to the source 2017), which may point to a possible influence of magnetite NPs on the Fig. 6. SIMPER and Indicator Species analyses for species showing relatives abundances higher than 1 % in at least one process, according to addition of NP, inoculum type and recovery effect. Significant species in each case according to Indicator Species analysis have been marked with a rectangle, while major con­ tributors to dissimilarity (>3%) according to SIMPER analysis have been highlighted according to the colour assigned to each specific effect. In red are highlighted those species that contribute in all cases. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) 10 R. Barrena et al. Bioresource Technology 372 (2023) 128632 presence of these compounds, as it has already been reported (Li et al., Appendix A. Supplementary data 2019). This same species was also the only one significantly differenti­ ated in the processes operated to recover methanogenic activity, while Supplementary data to this article can be found online at https://doi. Metanothrix soehngenii was significantly differentiated in the control org/10.1016/j.biortech.2023.128632. processes. In the case of the processes inhibited with glucose, three representatives of the Clostridia group stood out statistically with References respect to the rest of the species, which is common in processes over­ loaded with this sugar (Li et al., 2015). Abo Markeb, A., Llimós-Turet, J., Ferrer, I., Blánquez, P., Alonso, A., Sánchez, A., Moral- Vico, J., Font, X., 2019. The use of magnetic iron oxide based nanoparticles to Most of the species cited were those indicated by the SIMPER anal­ improve microalgae harvesting in real wastewater. Water Res. 159, 490–500. ysis as the main contributors to the existing dissimilarities between https://doi.org/10.1016/j.watres.2019.05.023. processes grouped according to the aforementioned criteria (Fig. 6). Angelidaki, I., Boe, K., Ellegaard, L., 2005. Effect of operating conditions and reactor configuration on efficiency of full-scale biogas plants. Water Sci. Technol. a J. Int. Thus, Methanobacterium beijingense and Clostridium butyricum were Assoc. Water Pollut. Res. 52, 189–194. shown to be determinant in differentiating processes, regardless of the Arif, S., Liaquat, R., Adil, M., 2018. Applications of materials as additives in anaerobic criteria adopted. Moreover, Cluster 1165, evolutionarily close to the digestion technology. Renew. Sustain. Energy Rev. 97, 354–366. https://doi.org/ 10.1016/j.rser.2018.08.039. Clostridia group, Petrimonas sulfuriphila and Ottowia pentelensis, and Barrena, R., Vargas-García, M.D.C., Capell, G., Barańska, M., Puntes, V., Moral-Vico, J., Methanothrix soehngenii were for NPs, type of inoculum, and recovery Sánchez, A., Font, X., 2021. Sustained effect of zero-valent iron nanoparticles under effects, respectively. semi-continuous anaerobic digestion of sewage sludge: Evolution of nanoparticles and microbial community dynamics. Sci. Total Environ. 777 https://doi.org/ 10.1016/j.scitotenv.2021.145969. 4. Conclusions Cai, W., Wan, J., 2007. Facile synthesis of superparamagnetic magnetite nanoparticles in liquid polyols. J. Colloid Interface Sci. 305, 366–370. https://doi.org/10.1016/j. Anaerobic digestion supplemented with magnetite-based NPs could jcis.2006.10.023. Casals, E., Barrena, R., García, A., González, E., Delgado, L., Busquets-Fité, M., Font, X., be an effective strategy to operate at high OLR improving the stability of Arbiol, J., Glatzel, P., Kvashnina, K., Sánchez, A., Puntes, V., 2014. Programmed iron the AD. NPs addition shows the ability to recover the microbiome oxide nanoparticles disintegration in anaerobic digesters boosts biogas production. associated with AD subjected to VFA inhibition, especially the reestab­ Small 10 (14), 2801–2808. Commission, E., Communication, D.-G. for, 2021. Delivering the Green Deal : the role of lishment of the methanogenic communities but also changing the clean gases including hydrogen. Publications Office of the European Union. doi/ dominance favouring hydrogenotrophic species (Methanobacterium bei­ 10.2775/718801. jingense). Species related to amino acid fermentation (Lutispora thermo­ Daraei, H., Rafiee, M., Yazdanbakhsh, A.R., Amoozegar, M.A., Guanglei, Q., 2019. A comparative study on the toxicity of nano zero valent iron (nZVI) on aerobic phila) appear as the most differentiating criteria related to NPs presence, granular sludge and flocculent activated sludge: Reactor performance, microbial regardless the starting inoculum. Further experiments in continuous behavior, and mechanism of toxicity. Process Saf. Environ. Prot. 129, 238–248. conditions should be addressed to gain knowledge about both effective https://doi.org/10.1016/j.psep.2019.07.011. De Vrieze, J., Gildemyn, S., Vilchez-Vargas, R., Jáuregui, R., Pieper, D.H., Verstraete, W., operational use of NPs and economics of the process. Boon, N., 2015. Inoculum selection is crucial to ensure operational stability in anaerobic digestion. Appl. Microbiol. Biotechnol. 99, 189–199. https://doi.org/ CRediT authorship contribution statement 10.1007/s00253-014-6046-3. Eduok, S., Ferguson, R., Jefferson, B., Villa, R., Coulon, F., 2017. Aged-engineered nanoparticles effect on sludge anaerobic digestion performance and associated Raquel Barrena: Conceptualization, Investigation, Visualization, microbial communities. Sci. Total Environ. 609, 232–241. doi: 10.1016/j. Writing – original draft. María del Carmen Vargas-García: Formal scitotenv.2017.07.178. analysis, Visualization, Writing – original draft. Paula Catacora- Ellacuriaga, M., Cascallana, J.G., González, R., Gómez, X., 2021. High-solid anaerobic digestion: reviewing strategies for increasing reactor performance. Environ. – MDPI Padilla: Investigation, Validation. Teresa Gea: Validation, Writing – 8 (8), 80. review & editing. Ahmad Abo Markeb: Resources. Javier Moral-Vico: Feng, Q., Song, Y.C., Bae, B.U., 2016. Influence of applied voltage on the performance of Resources, Writing – original draft. Antoni Sánchez: Writing – review & bioelectrochemical anaerobic digestion of sewage sludge and planktonic microbial communities at ambient temperature. Bioresour. Technol. 220, 500–508. https:// editing. Xavier Font: Conceptualization, Writing – review & editing. doi.org/10.1016/j.biortech.2016.08.085. Thomas J. Aspray: Conceptualization, Supervision, Writing – review & Holliger, C., Alves, M., Andrade, D., et al., 2016. Towards a standardization of editing. biomethane potential tests. Water Sci. Technol. 74, 2515–2522. doi: 10.2166/ wst.2016.336. Klang, J., Szewzyk, U., Bock, D., Theuerl, S., 2019. Nexus between the microbial Declaration of Competing Interest diversity level and the stress tolerance within the biogas process. Anaerobe 56, 8–16. https://doi.org/10.1016/j.anaerobe.2019.01.003. Kong, X., Yu, S., Fang, W., Liu, J., Li, H., 2018. Enhancing syntrophic associations among The authors declare that they have no known competing financial Clostridium butyricum, Syntrophomonas and two types of methanogen by zero interests or personal relationships that could have appeared to influence valent iron in an anaerobic assay with a high organic loading. Bioresour. Technol. the work reported in this paper. 257, 181–191. https://doi.org/10.1016/j.biortech.2018.02.088. Lee, J., Koo, T., Yulisa, A., Hwang, S., 2019. Magnetite as an enhancer in methanogenic degradation of volatile fatty acids under ammonia-stressed condition. J. Environ. Data availability Manage. 241, 418–426. https://doi.org/10.1016/j.jenvman.2019.04.038. Li, Y.F., Abraham, C., Nelson, M.C., Chen, P.H., Graf, J., Yu, Z., 2015. Effect of organic Data will be made available on request. loading on the microbiota in a temperature-phased anaerobic digestion (TPAD) system co-digesting dairy manure and waste whey. Appl. Microbiol. Biotechnol. 99, 8777–8792. https://doi.org/10.1007/s00253-015-6738-3. Acknowledgments Li, S., Cao, Y., Zhao, Z., Zhang, Y., 2019. Regulating secretion of extracellular polymeric substances through dosing magnetite and zerovalent iron nanoparticles to affect anaerobic digestion mode. ACS Sustain. Chem. Eng. 7, 9655–9662. https://doi.org/ Raquel Barrena is grateful to TECNIOspring fellowship programme 10.1021/acssuschemeng.9b01252. (TECSPR15-1-0051) co-financed by the European Union through the Li, S., Li, D., Wang, Y., Zeng, H., Yuan, Y., Zhang, J., 2020. Startup and stable operation Marie Curie Actions and ACCIÓ (Generalitat de Catalunya). of advanced continuous flow reactor and the changes of microbial communities in aerobic granular sludge. Chemosphere 243, 125434. https://doi.org/10.1016/j. As Antoni Sánchez, a [co-]author on this paper, is an editorial board chemosphere.2019.125434. member of Bioresource Technology, he was blinded to this paper during Lim, E.Y., Lee, J.T.E., Zhang, L., Tian, H., Ong, K.C., Tio, Z.K., Zhang, J., Tong, Y.W., review, and the paper was independently handled by Samir Kumar 2022. Abrogating the inhibitory effects of volatile fatty acids and ammonia in overloaded food waste anaerobic digesters via the supplementation of nano-zero Khanal as editor. valent iron modified biochar. Sci. Total Environ. 817, 152968 https://doi.org/ 10.1016/j.scitotenv.2022.152968. 11 R. Barrena et al. Bioresource Technology 372 (2023) 128632 Ma, K., Liu, X., Dong, X., 2005. Methanobacterium beijingense sp. nov., a novel Wang, S., Chen, L., Yang, H., Liu, Z., 2021. Influence of zinc oxide nanoparticles on methanogen isolated from anaerobic digesters. Int. J. Syst. Evol. Microbiol. 55, anaerobic digestion of waste activated sludge and microbial communities. RSC Adv 325–329. https://doi.org/10.1099/ijs.0.63254-0. 11 (10), 5580–5589. Ma, K., Wang, W., Liu, Y., Bao, L., Cui, Y., Kang, W., Wu, Q., Xin, X., 2021. Insight into Werner, J.J., Garcia, M.L., Perkins, S.D., Yarasheski, K.E., Smith, S.R., Muegge, B.D., the performance and microbial community profiles of magnetite-amended anaerobic Stadermann, F.J., DeRito, C.M., Floss, C., Madsen, E.L., Gordon, J.I., Angenent, L.T., digestion: varying promotion effects at increased loads. Bioresour. Technol. 329, Spormann, A.M., 2014. Microbial community dynamics and stability during an 124928 https://doi.org/10.1016/j.biortech.2021.124928. ammonia-induced shift to syntrophic acetate oxidation. Appl. Environ. Microbiol. 80 Mikucki, J.A., Liu, Y., Delwiche, M., Colwell, F.S., Boone, D.R., 2003. Isolation of a (11), 3375–3383. methanogen from deep marine sediments that contain methane hydrates, and Westerholm, M., Dolfing, J., Schnürer, A., 2019. Growth characteristics and description of Methanoculleus submarinus sp. nov. Appl. Environ. Microbiol. 69, thermodynamics of syntrophic acetate oxidizers. Environ. Sci. Technol. 53, 3311–3316. https://doi.org/10.1128/AEM.69.6.3311-3316.2003. 5512–5520. https://doi.org/10.1021/acs.est.9b00288. Mohd Yasin, N.H., Rahman, N.A., Man, H.C., Mohd Yusoff, M.Z., Hassan, M.A., 2011. Winkler, M.K.H., Kleerebezem, R., Khunjar, W.O., de Bruin, B., van Loosdrecht, M.C.M., Microbial characterization of hydrogen-producing bacteria in fermented food waste 2012. Evaluating the solid retention time of bacteria in flocculent and granular at different pH values. Int. J. Hydrogen Energy 36, 9571–9580. https://doi.org/ sludge. Water Res. 46, 4973–4980. https://doi.org/10.1016/j.watres.2012.06.027. 10.1016/j.ijhydene.2011.05.048. Wojcieszak, M., Pyzik, A., Poszytek, K., Krawczyk, P.S., Sobczak, A., Lipinski, L., Mori, K., Harayama, S., 2011. Methanobacterium petrolearium sp. nov. and Roubinek, O., Palige, J., Sklodowska, A., Drewniak, L., 2017. Adaptation of Methanobacterium ferruginis sp. nov., mesophilic methanogens isolated from salty methanogenic inocula to anaerobic digestion of maize silage. Front. Microbiol. environments. Int. J. Syst. Evol. Microbiol. 61, 138–143. https://doi.org/10.1099/ https://doi.org/10.3389/fmicb.2017.01881. ijs.0.022723-0. Wu, B., Lin, R., Kang, X., Deng, C., Xia, A., Dobson, A.D.W., Murphy, J.D., 2020. Peces, M., Astals, S., Jensen, P.D., Clarke, W.P., 2018. Deterministic mechanisms define Graphene addition to digestion of thin stillage can alleviate acidic shock and the long-term anaerobic digestion microbiome and its functionality regardless of the improve biomethane production. ACS Sustain. Chem. Eng. 8, 13248–13260. https:// initial microbial community. Water Res. 141, 366–376. https://doi.org/10.1016/j. doi.org/10.1021/acssuschemeng.0c03484. watres.2018.05.028. Yan, M., Fotidis, I.A., Jéglot, A., Treu, L., Tian, H., Palomo, A., Zhu, X., Angelidaki, I., Ribera-Pi, J., Campitelli, A., Badia-Fabregat, M., Jubany, I., Martínez-Lladó, X., McAdam, 2020. Long-term preserved and rapidly revived methanogenic cultures: Microbial E., Jefferson, B., Soares, A., 2020. Hydrolysis and Methanogenesis in UASB-AnMBR dynamics and preservation mechanisms. J. Clean. Prod. 263 https://doi.org/ treating municipal wastewater under psychrophilic conditions: importance of 10.1016/j.jclepro.2020.121577. reactor configuration and inoculum. Front. Bioeng. Biotechnol. doi: 10.3389/ Yan, W., Shen, N., Xiao, Y., Chen, Y., Sun, F., Kumar Tyagi, V., Zhou, Y., 2017. The role of fbioe.2020.567695. conductive materials in the start-up period of thermophilic anaerobic system. Salama, A.M., Abedin, R.M.A., Elwakeel, K.Z., 2021. Influences of greenly synthesized Bioresour. Technol. 239, 336–344. https://doi.org/10.1016/J. iron oxide nanoparticles on the bioremediation of dairy effluent using selected BIORTECH.2017.05.046. microbial isolates. Int. J. Environ. Sci. Technol. 19, 7019–7030. https://doi.org/ Zhang, Z., Guo, L., Wang, Y., Zhao, Y., She, Z., Gao, M., Guo, Y., 2020. Application of iron 10.1007/s13762-021-03625-3. oxide (Fe3O4) nanoparticles during the two-stage anaerobic digestion with waste Salvador, A.F., Martins, G., Melle-Franco, M., Serpa, R., Stams, A.J.M., Cavaleiro, A.J., sludge: Impact on the biogas production and the substrate metabolism. Renew. Pereira, M.A., Alves, M.M., 2017. Carbon nanotubes accelerate methane production Energy 146, 2724–2735. https://doi.org/10.1016/j.renene.2019.08.078. in pure cultures of methanogens and in a syntrophic coculture. Environ. Microbiol. Zhang, J., Zhao, W., Zhang, H., Wang, Z., Fan, C., Zang, L., 2018. Recent achievements in 19 (7), 2727–2739. enhancing anaerobic digestion with carbon- based functional materials. Bioresour. Siegert, I., Banks, C., 2005. The effect of volatile fatty acid additions on the anaerobic Technol. 266, 555–567. digestion of cellulose and glucose in batch reactors. Process Biochem. 40, Zhao, Z., Li, Y., Quan, X., Zhang, Y., 2017. Towards engineering application: Potential 3412–3418. https://doi.org/10.1016/j.procbio.2005.01.025. mechanism for enhancing anaerobic digestion of complex organic waste with Sposob, M., Moon, H.-S., Lee, D., Yun, Y.-M., 2021. Microbiome of seven full-scale different types of conductive materials. Water Res. 115, 266–277. https://doi.org/ anaerobic digestion plants in south korea: Effect of feedstock and operational 10.1016/j.watres.2017.02.067. parameters. Energies 14 (3), 665. Zheng, S., Liu, F., Wang, B., Zhang, Y., Lovley, D.R., 2020. Methanobacterium capable of Sun, Z., Feng, L., Li, Y., Han, Y., Zhou, H., Pan, J., 2022. The role of electrochemical direct interspecies electron transfer. Environ. Sci. Technol. 54, 15347–15354. properties of biochar to promote methane production in anaerobic digestion. https://doi.org/10.1021/acs.est.0c05525. J. Clean. Prod. 362, 132296 https://doi.org/10.1016/j.jclepro.2022.132296. Zhong, D., Li, J., Ma, W., Qian, F., 2020. Clarifying the synergetic effect of magnetite Venkiteshwaran, K., Milferstedt, K., Hamelin, J., Fujimoto, M., Johnson, M., Zitomer, D. nanoparticles in the methane production process. Environ. Sci. Pollut. Res. 27, H., 2017. Correlating methane production to microbiota in anaerobic digesters fed 17054–17062. https://doi.org/10.1007/s11356-020-07828-y. synthetic wastewater. Water Res. 110, 161–169. https://doi.org/10.1016/j. Ziganshina, E.E., Belostotskiy, D.E., Bulynina, S.S., Ziganshin, A.M., 2021. Effect of watres.2016.12.010. magnetite on anaerobic digestion of distillers grains and beet pulp: Operation of Vu, M.T., Noori, M.T., Min, B., 2020. Conductive magnetite nanoparticles trigger reactors and microbial community dynamics. J. Biosci. Bioeng. 131, 290–298. syntrophic methane production in single chamber microbial electrochemical https://doi.org/10.1016/j.jbiosc.2020.10.003. systems. Bioresour. Technol. 296, 122265 https://doi.org/10.1016/j. biortech.2019.122265. 12