Chemosphere 286 (2022) 131649 Contents lists available at ScienceDirect Chemosphere journal homepage: www.elsevier.com/locate/chemosphere Assessing main process mechanism and rates of sulfate reduction by granular biomass fed with glycerol under sulfidogenic conditions X. Zhou a, E. Fernández-Palacios a, A.D. Dorado b, X. Gamisans b, D. Gabriel a, * a GENOCOV Research Group, Department of Chemical, Biological and Environmental Engineering, Escola d’enginyeria, Universitat Autònoma de Barcelona, 08193, Bellaterra, Spain b Department of Mining, Industrial and ICT Engineering, Universitat Politècnica de Catalunya, Avinguda de les Bases de Manresa 61-73, 08242, Manresa, Spain 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 • The mechanisms of sulfate reduction using glycerol were established. • Ethanol and 1,3-propanediol were the main electron donors in sulfate reduction. • Incomplete-oxidizing SRB dominated under sulfidogenic conditions. • SRB preferentially used simple interme­ diate products except for acetate. A R T I C L E I N F O A B S T R A C T Handling Editor: Hyunook Kim Sulfate-reducing bioreactors for sulfide production are the initial stage of processes targeting elemental sulfur recovery from sulfate-rich effluents. In this work, the principal reactions involved in glycerol fermentation and Keywords: sulfate reduction using glycerol and its fermentation products as electron donors were assessed together with Glycerol fermentation their specific consumption/production rates. A battery of batch activity tests with and without sulfate were Sulfate-reduction mechanism performed with glycerol and with each fermentation product using a non-methanogenic but sulfidogenic gran­ Activity tests ular sludge from an up-flow anaerobic sludge blanket (UASB) reactor operated under long-term while fed with Intermediate products UASB crude glycerol. As a result, a mechanistic approach based on the experimental observations is proposed in this work. Glycerol was mainly fermented to 1,3-propanediol, ethanol, formate, propionate and acetate by fermen­ tative bacteria. All organic intermediates were found to be further used by sulfate reducing bacteria (SRB) for sulfate reduction except for acetate. The most abundant genus detected under sulfidogenic conditions were Propionispora (15.2%), Dysgonomonas (13.2%), Desulfobulbus (11.6%) and Desulfovibrio (10.8%). The last two SRB genera accounted for 22.4% of the total amount of retrieved sequences, which were probably performing an incomplete oxidation of the carbon source in the sulfidogenic UASB reactor. As single substrates, specific sulfate reduction rates (SRRs) using low molecular weight (MW) carbon sources (formate and ethanol) were 39% higher than those using high-MW ones (propionate, 1,3-propanediol and butanol). However, SRRs in glycerol-fed tests showed that 1,3-propanediol played a major role in sulfate reduction in addition to formate and ethanol. * Corresponding author. E-mail address: david.gabriel@uab.cat (D. Gabriel). https://doi.org/10.1016/j.chemosphere.2021.131649 Received 8 February 2021; Received in revised form 16 July 2021; Accepted 21 July 2021 Available online 23 July 2021 0045-6535/© 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). X. Zhou et al. Chemosphere 286 (2022) 131649 1. Introduction directly oxidized to acetate and bicarbonate in the sulfate reduction process (Bertolino et al., 2014; Dinkel et al., 2010; Qatibi et al., 1991a). Sulfate is an anion widely present in natural environments. However, However, the TOC imbalance found in these works was suggested to be high concentrations of sulfate can be toxic to aquatic life (Karjalainen caused by other intermediate products. Several studies have shown that et al., 2021). In aquatic environments, sulfate can be also reduced to other products are produced when glycerol is used as electron donor to hydrogen sulfide by sulfate-reducing bacteria (SRB) under anaerobic reduce sulfate. As examples, 3-hydroxypropionate (Qatibi et al., 1998) conditions generating further problems since hydrogen sulfide is and butyrate (Bertolino et al., 2014) were produced in glycerol poisonous and corrosive. Many industrial sites including pulp and paper fermentation coupled to sulfate reduction. It was also found that glyc­ and mining industries, tanneries, fermenting plants and thermal power erol was incompletely oxidized to acetate, lactate, and 1,3-propanediol plants generate flue gases or wastewaters containing large amounts of under low pH conditions (Santos et al., 2018). However, the mechanism sulfur, mainly as sulfate, that require further treatment. As an alterna­ of the sulfate reduction process, which is also culture dependent, is still tive to costly physical-chemical technologies, environmentally friendly, unclear as most of literature have only shown the oxidized products of biological processes arose recently, some considering the valorization of glycerol during this process. S-rich emissions into elemental sulfur (biosulfur) (Mora et al., 2020a), a The purpose of this work was to assess the main mechanism and the value-added product currently obtained from the petrochemical in­ process rates of glycerol fermentation and sulfate reduction of a sulfi­ dustry. Since microbial communities able to reduce sulfate directly to dogenic granular sludge enriched in an UASB fed with crude glycerol biosulfur have not been described yet, one biological-based alternative under long-term operating conditions without methane production. To relies on a two-stage process (Mora et al., 2020a). First, sulfate is this aim, microbial activity was assessed in batch tests with a range of reduced to sulfide in an upflow anaerobic sludge blanket (UASB) electron donors including glycerol, alcohols and volatile fatty acids bioreactor, and subsequently, sulfide is partially oxidized to elemental (VFAs) with and without the presence of sulfate. sulfur in a second bioreactor under microaerobic or anoxic conditions. The technical feasibility of biosulfur recovery in a two-stage bioscrubber 2. Materials and methods has been demonstrated previously in the SONOVA process (Mora et al., 2020a) despite some operational problems in the long-term leading to 2.1. Granular biomass growth conditions biomass losses in the anaerobic stage (Fernández-Palacios et al., 2021). However, further fundamental analysis is warranted to reveal the un­ A laboratory-scale UASB reactor as described in previous works derlying mechanisms in the UASB to develop a more robust, transferable (Fernández-Palacios et al., 2019; Mora et al., 2020a) was set up and technology to field applications. inoculated with granular sludge from a full-scale UASB from a paper UASB reactors are widely used in anaerobic digestion and have also recycling industry (Spain), which was mainly used for methane pro­ been reported as an appropriate technology for sulfate reduction using duction. The lab-scale UASB was operated to promote sulfidogenic, different carbon sources, including methanol (Weijma and Stams, 2001), non-methanogenic conditions as shown in supplementary material ethanol (Wu et al., 2018), sucrose (Weijma and Stams, 2001) and crude (Fig. S1). In order to minimize the effect of changing conditions, the glycerol (Fernández-Palacios et al., 2019). As an alternative to other sulfate inlet concentration was set to 250 mg S–SO2- 4 L − 1 while the crude expensive electron donors, crude glycerol is a by-product mainly pro­ glycerol to sulfate ratio (TOC/S–SO2-4 ) was steadily maintained at 1.49 ± duced in the biodiesel production industry with a high market potential. 0.31 g C/g S–SO2- 4 . These conditions were selected according to the It was estimated that world glycerol production would reach 2.66 period of maximum sulfate reducing and organic removal efficiencies, million tons in 2020 (Kumar et al., 2019). The composition of crude according to Fernández-Palacios et al. (2019). The mineral medium glycerol varies depending on the process, containing around 70–98% of (MM) composition fed to the UASB was (in g L-1): NH4Cl (0.2), K2HPO4 glycerol plus some impurities including water, long-chain fatty acids, (3), and Na2SO4 (1.15) dissolved in tap water. The pH of the MM was fatty acid methyl esters, salts, methanol, soap, and ashes (Angeloni et al., adjusted to 8.5–8.8 using 2 M NaOH. The MM flow was mixed with the 2016; Viana et al., 2012; Vivek et al., 2017). carbon source (crude glycerol) flowrate and pumped into the UASB from Metabolic pathways of glycerol fermentation in anaerobic digestion the bottom of the reactor. The temperature was controlled at 35 ◦ C with have been clearly described by a combination of reductive and oxidative a water bath and the up-flow velocity was set at 0.25 m h− 1. The UASB reactions (Viana et al., 2012). From one side, glycerol is first dehydrated reactor was operated stably for 639 days under the conditions described to 3-hydroxypropionaldehyde that is further reduced to 1,3-propanediol above. Also, the redox potential (ORP) in the reactor was monitored by the enzyme 1,3-propanediol dehydrogenase. On the other hand, via throughout the operation with an ORP probe (pH 5350, Crison, Spain) an oxidative pathway, glycerol is also converted to phosphoenolpyr­ on top of the reactor close to the liquid outlet. The ORP varied from uvate, which subsequently produces propionate and pyruvate. Pyruvate − 460 mV to − 540 mV along the operation, indicating that dissolved can further produce other compounds such as n-butanol, 2,3-butanediol, oxygen did not penetrate into the UASB. ethanol, lactate, butyrate, formate, acetate, hydrogen and CO2 depending on the microbial cultures and environmental conditions. In 2.2. Batch activity tests addition to H2, most of the abovementioned organic compounds can be used as electron donors by SRB to reduce sulfate to sulfide. In order to identify the main mechanisms of the granular sludge The reduction of sulfate to sulfide by SRB follows both assimilative using glycerol as electron donor to reduce sulfate in the sulfidogenic and dissimilative metabolisms. Dissimilatory sulfate reduction is the UASB, granular sludge was taken from the UASB after 431, 549 and 578 main pathway to produce hydrogen sulfide (Madigan et al., 1997) by days of operation, when no methane was produced. The sludge with­ SRB using sulfate as the terminal electron acceptor either through drawn from the UASB on different days was tested with different carbon autotrophic or heterotrophic metabolisms (Lens and Kuenen, 2001). sources. Pure glycerol, ethanol, 1,3-propanediol, butanol and 2,3-buta­ Autotrophic SRB use CO2 as carbon source and H2 as electron donor to nediol were tested with the sludge collected on day 431. Formate was reduce sulfate while heterotrophic SRB can use a wide variety of organic tested with the sludge collected on day 549, while acetate and propio­ compounds as electron donor including organic acids (formic, acetic, nate were tested with the sludge collected on day 578. Sludge samples propionic, butyric, pyruvic, lactic, etc.) and alcohols (methanol, ethanol, were first rinsed with MM (without sulfate) and then transferred into butanol etc.). Heterotrophic SRB include incomplete oxidizers (that 250 ml serum bottles, which were afterwards fed with 150 ml of MM. In partly oxidize organic compounds to acetate), and complete oxidizers all cases, a control group without sulfate and an experimental group (that entirely oxidize organic compounds to CO2). were evaluated. The control group consisted of MM (without sulfate) Previous studies have reported that glycerol as electron donor can be supplied with different carbon sources only, while in the experimental 2 X. Zhou et al. Chemosphere 286 (2022) 131649 group sulfate was added to the MM and the carbon source. The initial 3. Results and discussion carbon concentration ranged from 318 to 372 mg C L− 1 in both the control and the experimental groups. The initial sulfate concentration of 3.1. Activity tests with glycerol as sole external carbon source the experimental group was set to 240 ± 4 mg S–SO2- 4 L − 1 , and the TOC/ S–SO2- 4 ratio was 1.4 ± 0.1 g C/g S–SO 2- 4 . After adding MM and the Long-term operation of the UASB reactor was performed under granular sludge, the pH was adjusted to 8–8.5 by adding either 2 M constant sulfate and organic loading rates for 639 days. The experi­ NaOH or 2 M HCl. The pH was not controlled during activity tests, but it mental results and discussion of the UASB performance and the evolu­ was measured at the beginning and at the end of each test with a pH tion of the microbial diversity can be found elsewhere probe (pH 5333, Crison, Spain). The gas phase was exchanged with N2 (Fernández-Palacios et al. (2021). In short, after operating the UASB and bottles were instantly capped with rubber stoppers and aluminum reactor for 200 days, methane production decreased significantly caps. Bottles were cultivated in a shaker (NB-T205, N-Biotek, Korea) at a concomitantly with VFAs production. At a TOC/S–SO2- 4 ratio of 1.49 ± constant temperature of 35 ± 1 ◦ C and a stirring rate of 150 rpm. The 0.31, the reactor had undergone a progressive washout of methanogens, incubation time was between 96 and 120 h depending on the test. including Methanosaeta, Methanobacteria, and Methanomicrobiales. SRB Volatile suspended solids (VSS) in batch tests were obtained at the end of completely outcompeted methanogens after 230 days of operation, in each experiment and ranged from 125 to 390 mg VSS L − 1. All batch tests which Desulfobulbus and Desulfovibrio were found the main were carried out with two replicates. sulfate-reducing genus. Through the treatment of sulfate-rich waste­ water, the methanogenic sludge was shifted into a sulfidogenic sludge. 2.3. Analytical methods Activity tests presented herein were performed to study the sulfate reduction mechanism using glycerol as carbon source using sludge Sulfate and thiosulfate were analyzed by ion chromatography (ICS- collected from the reactor when it was performing under sulfidogenic 2000 system, Dionex, USA) with a suppressed conductivity detector conditions. From day 200 to day 400, only VFAs were considered as the using an IonPac AS18-HC column (4 × 250 mm, Dionex). Total dissolved intermediate products of glycerol degradation, but a significant carbon sulfide (TDS) was analyzed by a sulfide selective electrode (9616BNWP, imbalance was found during the process of glycerol fermentation. Thermo Scientific, MA, USA) connected to a benchtop meter (Sym­ Therefore, other intermediate metabolites were considered in batch tests phony, VWR, Germany). Before TDS measurement, samples were after day 400, such as ethanol and 1,3-propanediol. diluted 20 times with a sulfide antioxidant buffer (SAOB) containing (g L− 1): ascorbic acid (35), EDTA (67) and NaOH (80). 3.1.1. Glycerol fermentation process Volatile fatty acids (VFAs) and alcohols were measured by high- Profiles for carbon and sulfur species using pure glycerol as the only performance liquid chromatography (HPLC, Ultimate 3000, Dionex, carbon source externally added to the serum bottles with and without USA) equipped with an ICSep ICE-CPREGEL 87H3 column (7.8 mm × the presence of sulfate are shown in Fig. 1. In the control group (Fig. 1A 150 mm) and a variable wavelength detector at 210 nm with a 6 mM and B), only glycerol was added with an initial granular sludge con­ H2SO4 mobile phase at a flow rate of 0.5 mL min− 1. Total organic carbon centration of 230 mg VSS L− 1. The 120-h culture in the control test was (TOC), total inorganic carbon (TIC) and total carbon (TC) were deter­ divided into three phases: a lag phase (0–10 h), a quick glycerol uptake mined in a TOC analyzer (multi N/C 2100S, Analytikjena, Germany). All phase (10–48 h), and a glycerol-free phase mediated by several in­ samples were filtered by 0.22 μm (Millipore, USA) before analysis. termediates (48–120 h). After the lag phase (10 h), glycerol began to be CH4, CO2 and H2 from test bottles headspace were analyzed by gas consumed while ethanol, 1,3-propanediol, formate, acetate and propi­ chromatography (7820A, Agilent Technologies, USA) equipped with a onate accumulated. Along this second period, glycerol was consumed at detector (TCD) and a capillary column (Al2O3 PLOT: 50 m × 0.53 mm). a rate of 49.9 ± 13.7 mg C g VSS− 1 h− 1 and ethanol, 1,3-propanediol, H2S was analyzed by gas chromatography (HP 5890A GC, Hewlett formate, acetate and propionate were accumulated at rates of 2.1 ± Packard, USA) equipped with a thermal conductivity detector and a 2.9, 30.3 ± 8.2, 1.6 ± 2.4, 6.2 ± 3.0 and 0.5 ± 0.7 mg C g VSS− 1 h− 1, Porapak Q column. Before measuring the gas composition in the head­ respectively. At the end of the second phase (48 h), glycerol had already space of each bottle, pressure was measured by a manometer (SMC been completely converted to ethanol, 1,3-propanediol, formate, ace­ ISE30A-01-P, Japan). tate, propionate and inorganic carbon, accounting for 4.5 ± 0%, 57.5 ± Volumetric consumption/production rates of the measured species 0.6%, 3.9 ± 1.1%, 12.2 ± 2.7%, 1.1 ± 1.5%, and 4.0 ± 0.9% of the initial were calculated by the concentration changes between two consecutive glycerol content, respectively. During the last phase (48–120 h), 1,3-pro­ sampling events and, then, averaged for the whole length of the specific panediol and inorganic carbon were maintained at 195 and 23 mg C L− 1, period evaluated. VSS were measured according to Standard Methods respectively. All formate (53.6 mg C L− 1) was completely consumed (APHA, 2012) to calculate specific rates. while butyrate increased from 0 to 7.3 mg C L− 1 and propionate increased from 3.5 to 17.6 mg C L− 1. The pH decreased from 8.4 to 7.2 2.4. Illumina sequencing analysis concomitantly with VFAs production. Results obtained through Illumina sequencing analysis of the 16S Identification of the microbial population was performed using rRNA with the sludge sample collected on day 538 from the UASB are Illumina platform of samples collected on day 538 of the long-term presented in supplementary material (Fig. S2). The most abundant genus operation of the UASB. Genomic DNA was extracted by applying the detected were Propionispora (15.2%), Dysgonomonas (13.2%), Desulfo­ protocol of PowerSoil™ DNA isolation kit (MoBio Laboratories, USA) bulbus (11.6%) and Desulfovibrio (10.8%). Bacteria from the genus Dys­ following the supplier’s instructions. The quantity and quality of the gonomonas are obligately anaerobic also known for their fermentative extracted DNA were assessed by using a NanoDrop 1000 Spectropho­ activity while Propionispora is a genus of anaerobic fermentative bacteria tometer (Thermo Fisher Scientific, USA). Then, DNA samples were that typically use glycerol, and other carbohydrates, fermenting them to preserved at − 20 ◦ C for further analysis. Sequencing analyses were produce propionic, acetic acid, CO2 and H2 (Abou-Zeid et al., 2004). performed by “Genomic and Bioinformatics service” at the Universitat Major fermentation products are usually butyric and acetic acid while Autònoma de Barcelona. Amplicon sequencing that targets the V3–V4 propionic, isovaleric, isobutyric and phenylacetic acids may also be hypervariable regions (HVRs) of the 16S rRNA gene on Illumina MiSeq produced (Sakamoto, 2014). Additionally, Klebsiella (2.1%) was also platform were carried out using the universal primers 341F (5′ - CCT detected. The metabolism of Klebsiella pneumoniae with glycerol has also ACG GGN GGC WGC AG-3′ ) and 805R (5′ - GAC TAC HVG GGT ATC TAA been described as follows: a) glycerol is converted to ethanol and TCC -3′ ). The database used for the classification of organisms is based formate (Jarvis et al., 1997); b) glycerol is degraded into acetate and on the Greengenes database (http://greengenes.lbl.gov/). formate (Zeng et al., 1993); c) lactate, acetate, ethanol and 1, 3 X. Zhou et al. Chemosphere 286 (2022) 131649 Fig. 1. Anaerobic bioconversion of pure glycerol in serum bottles without sulfate (A, B) and with sulfate (C, D, E) showing the glycerol consumption and the concentration profiles for alcohols (A, C), volatile fatty acids and inorganic carbon (B, D) and sulfur species (E). Error bars represent standard deviations of duplicate tests. TIC represents total inorganic carbon. 3-propanediol are produced as end products (Cheng et al., 2007). Ac­ h− 1, respectively. Sulfate was reduced at a rate of 6.3 ± 3.4 mg S g VSS− 1 cording to the products of glycerol shown in Fig. 1A and B and the h− 1 along this phase. At the end of the second phase (33 h), glycerol was metabolic pathways of glycerol mentioned above, the stoichiometry of completely consumed while ethanol, 1,3-propanediol, formate, acetate, the reactions involved in glycerol fermentation are shown in Table 1. propionate and inorganic carbon accumulated, accounting for 17.5 ± 0.8%, 32.3 ± 8.0%, 7.4 ± 2.2%, 16.7 ± 3.2%, 2.3 ± 0.0%, and 4.5 ± 3.1.2. Sulfate reduction in glycerol-fed batch tests 2.5% of the initial glycerol content, respectively. In the third phase, The profiles obtained for the experimental group where glycerol and when there was no glycerol, a net alcohols and VFAs consumption was sulfate were added are shown in Fig. 1C for alcohols, in Fig. 1D for VFAs observed, coupled to sulfate reduction and a significant accumulation of and in Fig. 1E for sulfur species. The granular sludge concentration in acetate and inorganic carbon. From 33 to 96 h, the sulfate reduction rate the experimental bottles was 390 mg VSS L− 1. The 120-h culture in the (SRR) gradually decreased from 10.4 mg S g VSS− 1 h− 1 (33–48 h) to 2.6 experimental test were also divided into three phases: the lag phase mg S g VSS− 1 h− 1 (82–96 h). Ethanol, 1,3-propanediol, formate and (0–10 h), the glycerol fermentation phase (10–33 h) and a secondary propionate were completely consumed at a rate of 8.9 ± 0.6, 4.4 ± 1.5, carbon sources consumption/production phase (33–120 h). Similar to 1.9 ± 1.2 and 1.2 ± 1.0 mg C g VSS− 1 h− 1, respectively. Almost no ac­ the control group, the granular sludge in the experimental group also tivity was observed after 96 h of test. experienced a 10 h lag phase with neither glycerol fermentation nor Bertolino et al. (2014) found that acetate and carbonate were the sulfate reduction. After the lag phase, the specific rates along the test terminal products of glycerol oxidation under sulfidogenic conditions at with sulfate were assessed. In the second phase from 10 to 33 h, glycerol a sulfate loading rate of 4.7 kg m− 3 d − 1 but pointed out that some in­ was consumed at a rate of 33.1 ± 12.1 mg C g VSS− 1 h− 1, while ethanol, termediate products in the degradation process of glycerol were not 1,3-propanediol, formate, acetate and propionate accumulated at rates measured but probably consumed during sulfate reduction. In addition, of 5.0 ± 7.9, 11.9 ± 2.1, 2.0 ± 4.5, 5.9 ± 2.7 and 1.0 ± 1.4 mg C g VSS− 1 the production of formate, ethanol and 1,3-propanediol have also been 4 X. Zhou et al. Chemosphere 286 (2022) 131649 Table 1 Fermentation and sulfidogenic reactions involved in anaerobic digestion. Stoichiometric equation Eq. Reference Fermentation reaction Zeng et al. (1993) C3 H8 O3 (glycerol) + H2 O + 2NAD+ →C2 H4 O2 (acetic ​ acid) + HCOOH + 2NADH + 2H+ (1) Schauder and Schink (1989) C3 H8 O3 (glycerol) → C3 H6 O2 (propionic ​ acid) + H2 O (2) Sittijunda and Reungsang (2017) C3 H8 O3 (glycerol) + NADH + H+ →C3 H8 O2 (1, 3 − propanediol) + NAD+ + H2 O (3) Sittijunda and Reungsang (2017) C3 H8 O3 (glycerol) + NAD+ →C2 H6 O(ethanol) + HCOOH + NADH + H+ (4) Bijmans et al. (2008) HCOOH → CO2 + H2 (5) Sulfidogenic reaction Muyzer and Stams (2008) 4H2 + SO2−4 + H+ →HS− + 4H2 O (6) Liamleam and Annachhatre (2007) 4HCOO− + SO2−4 + H+ →HS− + 4HCO−3 (7) Luis (2018) C3 H5 O−2 (propionate) + 0.75SO2−4 →C2 H3 O−2 (acetate) + HCO−3 + 0.75HS− + 0.25H+ (8) Wu et al. (2018) C2 H6 O(ethanol) + 0.5SO2−4 →0.5HS− + C2 H3 O−2 (acetate) + H2 O + 0.5H+ (9) Qatibi et al. (1991b) C3 H8 O2 (1, 3 − propanediol) + SO2−4 →C2 H3 O−2 (acetate) + HS− + HCO−3 + H2 O + H+ (10) previously reported in the fermentation of glycerol (Jarvis et al., 1997; Desulfobulbus can oxidize propionate to acetate in the presence of sulfate Katarzyna et al., 2011; Sittijunda and Reungsang, 2017). At the same and ferment pyruvate and lactate to a mixture of acetate and propionate time, these compounds could also act as intermediate products under (El Houari et al., 2017). This genus can also use lactate, pyruvate, sulfidogenic conditions. Ethanol, 1,3-propanediol, formate and propio­ ethanol or propanol as carbon sources and also as electron donors for nate have been reported as suitable electron donors in biological sulfate anaerobic respiration, oxidizing them incompletely to acetate. Accord­ reduction processes applied in different bioreactors, such as UASB and ing to Rabus et al. (2013), Desulfobulbus and Desulfovibrio are not able to expanded granular sludge bed reactors (De Smul and Verstraete, 1999; completely oxidize organic electron donors, therefore they perform an Qatibi et al., 1991b; Vallero et al., 2004a, 2004b; Vallero et al., 2004a, incomplete oxidation to acetate as end-product. Desulfobulbus and 2004b; Wu et al., 2018). The stoichiometry of the reactions involved in Desulfovibrio do not utilize acetate as electron donor even if some species sulfate reduction by non-methanogenic but sulfidogenic biomass is of both genera are able to use H2 as electron donor to reduce sulfate to shown in Table 1. Results in Fig. 1 confirmed that intermediate reduced sulfide (Rabus et al., 2013). Desulfovibrio spp. only carry out an incom­ organic carbon sources were used as electron donors in sulfate reduction plete oxidation of substrates and they are able to excrete acetate as processes. However, data was not sufficient to clearly identify which end-product (Rabus et al., 2013). This could indeed explain the accu­ specific electron donors were used and their SRRs. As an example, when mulation of acetate in the experimental group indicating that SRB found there was no glycerol fermentation, formate was consumed with and in the UASB reactor performed an incomplete oxidization of the carbon without the presence of sulfate (Fig. 1B and D). In the absence of sulfate, sources used. formate was consumed because of anaerobic fermentation. However, in By comparing the consumption rate of glycerol during the glycerol the presence of sulfate, the consumption of formate cannot be inter­ fermentation phase in the control group (10–48 h) and the experimental preted as the direct attribution to anaerobic fermentation or its use as an group (10–33 h), it was found that the consumption rate of glycerol in electron donor by SRB. Consequently, separate experiments with indi­ the control group was 29.4 ± 8.1 mg C L− 1 h− 1, similar to that in the vidual organic compounds were required to confirm the metabolic experimental group (32.7 ± 12.0 mg C L− 1 h− 1). This indicates that pathway of sulfate reduction using ethanol, 1,3-propanediol, formate glycerol was not probably used by SRB for sulfate reduction but that and propionate (see section 3.2). glycerol was first fermented by granular sludge to produce secondary The TOC removal (0 h–120 h) in the absence and presence of sulfate carbon substrates, and then SRB used these carbon sources to reduce were 5.7% and 19.0%, respectively, indicating that SRB played a major sulfate. Therefore, it was necessary to explore the sulfate reduction role in removing organics. The percentage of relative abundance of SRB mediated by other electron donors formed during glycerol fermentation on day 538 in the UASB could be considered as the sum of genus in order to understand the mechanism of sulfate reduction by the Desulfobulbus and Desulfovibrio accounting for a 22.4% of the total granular sludge. amount of sequences detected. Most bacteria that belong to the genus 5 X. Zhou et al. Chemosphere 286 (2022) 131649 3.1.3. Mass balances It has been reported that VOSC are produced in the anaerobic digestion The carbon mass balance of the glycerol-added serum bottle exper­ of organic wastes (Papurello et al., 2012) as well as propylene glycol and iments is presented in the supplementary material section (without glycerol (Trabue et al., 2007). Compounds such as methanethiol and sulfate, Fig. S3A, and with sulfate, Fig. S3B). The sum of carbon was dimethyl sulfide are formed by the degradation of sulfur-containing calculated as the sum of volatile fatty acids and alcohols. In the absence amino acids or sulfide methylation in anaerobic digestion (Lomans of sulfate, the carbon imbalance was 10.1 ± 3.0% between TOC and the et al., 2002). As shown in Fig. 1E, the concentration of H2S in the gas sum of carbon species measured (Fig. S3A) from 10 to 120 h. This could phase measured at 120 h was 20.1 mg S L− 1. Compared with the TDS be explained by the presence of some other organic carbon species not concentration measured in the liquid, the H2S concentration in the gas monitored as well as to biomass growth. Butyrate and propionate phase accounted only for 6.6% of the total S fed. The sulfur mass balance accumulation started only after glycerol depletion, which may be due to was calculated by monitoring the sulfate, thiosulfate and TDS content in the accumulation of phosphoenolpyruvate during the fermentation of the liquid and the hydrogen sulfide content in the gas phase at the glycerol (Viana et al., 2012). Phosphoenolpyruvate was subsequently beginning and at the end of each test. On average, a 4.7% sulfur oxidized after 48 h. Moreover, via a reductive pathway, glycerol not imbalance was observed between the S mass at 0 h and 120 h, which only produces 1,3-propanediol but 3-hydroxypropionate, a metabolite may be caused by the formation of organic sulfur compounds. of glycerol fermentation (Qatibi et al., 1998). Therefore, these possible The sulfur and carbon mass balances of the UASB reactor are shown intermediate metabolites of glycerol that were not measured may in Fig. S1C while mass balance calculations are detailed in supplemen­ contribute to the 10.1% carbon imbalance. In the presence of sulfate, the tary material (section S1). Considering the H2S stripping in the UASB, carbon imbalance between TOC and the sum of all the rest of carbon the sulfur imbalance along the operation oscillated between − 10% and species was 18.5 ± 6.3% (Fig. S3B). Compared with the carbon imbal­ +20%. The sulfur imbalance was attributed to the production of unde­ ance without sulfate (10.1%), the increase in the imbalance was prob­ tected sulfur species in the rector. As reported by Mora et al. (2020b), ably due to the formation of volatile organic sulfur compounds (VOSC). the sulfur imbalance was not caused by the metallic sulfide precipitating Fig. 2. Carbon and sulfate conversions in the experimental bottles feeding sulfate plus propionate (A), ethanol (B), 1,3-propanediol (C), formate (D) and butanol (E). Error bars represent standard deviations of duplicate tests. TIC represents total inorganic carbon. 6 X. Zhou et al. Chemosphere 286 (2022) 131649 but due to the production of organic sulfur compounds as intermediate H2S inhibition could be another reason that lead to incomplete oxidizing compounds of the biodegradation. The negative imbalance was attrib­ SRB to dominate among SRB species. Maillacheruvu and Parkin (1996) uted to the standard deviation in the measurement of TDS (8%). reported that acetotrophic SRB and propionate fermenters were more Considering CH4 and CO2 stripping on the gas phase of UASB, the carbon sensitive to sulfide toxicity compared to hydrogenotrophic SRB and imbalance was mainly between − 5% and 30%, which is in concordance incomplete propionate-utilizing SRB. Thus, the mechanism of sulfate with the sulfur imbalance. The positive imbalance was attributed to the reduction using propionate was considered to be partial oxidation, growth of biomass in the reactor according to previous references represented by Eq. (8). This was confirmed by comparing the stoichio­ (Fernández-Palacios et al., 2019). metric rates and the experimental rates (Table S1). Their calculation process is shown in supplementary material (section S3.1). 3.2. Sulfate reduction with single organic compounds 3.2.2. Sulfate reduction with ethanol and 1,3-propanediol During the sulfate reduction with glycerol, it was found that SRB In the control bottles fed with ethanol as carbon source, the ethanol metabolized simple organic carbon intermediate products (Fig. 1C and concentration remained constant and no other products were produced E). Combined with the description of the fermentation mechanism of (Fig. S4B). However, in the presence of sulfate, sulfate was reduced to glycerol described by Viana et al. (2012), the intermediate products sulfide while ethanol was converted to acetate (Fig. 2B). When sulfi­ observed in Fig. 1 were used as individual carbon sources to further dogenesis dominated in a long-term UASB treatment with sulfate-rich study the mechanism and rates of the sulfate reduction process. wastewater, SRB oxidized ethanol into acetate for sulfidogenesis ac­ When no sulfate was added, the carbon content remained constant cording to Eq. (9) (Wu et al., 2018). The experimental results of Fig. 2B when sludge was fed with propionate (Fig. S4A), ethanol (Fig. S4B), 1,3- are well-explained through Eq. (9), which was confirmed by the stoi­ propanediol (Fig. S4C) and butanol (Fig. S4E). This indicates that chiometric calculation in Table S1. granular sludge was not capable of using those carbon sources without Similarly, results showed that 1,3-propanediol was not degraded to sulfate. However, in the test that used formate as the carbon source VFAs or inorganic carbon in the absence of sulfate (Fig. S4C). However, (Fig. S4D), and after 40 h of lag time, this compound was consumed at a in the presence of sulfate, 1,3-propanediol was oxidized to acetate, rate of 2.8 ± 0.8 mg C g VSS− 1 h− 1 while inorganic carbon was accu­ propionate, and bicarbonate (Fig. 2C). Qatibi et al. (1991b) reported mulated at a rate of 3.6 ± 1.3 mg C g VSS− 1 h− 1. Fig. 2 shows the profiles that 1,3-propanediol was oxidized to acetate and bicarbonate according of C and S species when single electron donors produced during glycerol to Eq. (10) during sulfate reduction (Table 1). They found that 3-hydrox­ fermentation were fed in batch tests to the UASB sludge. As can be ypropionate, which was further oxidized to acetate, was the interme­ observed, propionate (Fig. 2A), ethanol (Figs. 2B), 1,3-propanediol diate of the degradation of 1,3-propanediol by Desulfovibrio (Fig. 2C), formate (Fig. 2D) and butanol (Fig. 2E) were consumed in alcoholovorans. Interestingly, Liu and Liu (2016) reported that propio­ the sulfate-added tests. Coupled to sulfate reduction, propionate was nate was produced from 3-hydroxypropionate by Metallosphaera sedula degraded to acetate and inorganic carbon according to Eq. (8) (Table 1), through the autotrophic carbon dioxide assimilation cycle through ethanol was degraded to acetate according to Eq. (9) (Table 1), formate 3-hydroxypropionyl-CoA synthesis, 3-hydroxypropionyl-CoA dehydra­ was degraded to inorganic carbon, and butanol was oxidized to butyrate. tion, acryloyl-CoA reduction, and CoA hydrolysis (Fig. 3). Then, 1,3-pro­ When 1,3-propanediol was added, sulfate was reduced to sulfide and the panediol (C3H8O2) oxidation to propionate (C3H5O−2 ) was lumped into consumption of 1,3-propanediol was accompanied by the accumulation Eq. (11) in this work. of acetate, propionate and inorganic carbon (Fig. 2C). Despite the VFA C3 H8 O2 + SO2−4 + 3NADH + 3H+ →C3 H5 O−2 + HS− + 4H2 O + 3NAD+ production, the buffering capacity of the mineral medium maintained the pH above 7.5. (11) Consequently, both Eqs. (10) and (11) were used in the work herein 3.2.1. Sulfate reduction with propionate to describe the mechanism for the oxidation of 1,3-propanediol coupled According to literature, there are two main pathways described for to the sulfate reduction following the pathway described in Fig. 3. the sulfate reduction process using propionate as electron donor. One is acetogenesis, in which propionate is degraded into acetate and hydrogen 3.2.3. Sulfate reduction with formate and butanol by acetogens (Li et al., 2017), and then sulfate is reduced both by Despite some extra inorganic carbon was produced based on the hydrogenotrophic and acetotrophic SRB (Muyzer and Stams, 2008). The formate consumed, which was attributed to some remaining COD in the other is the propionate oxidation pathway by propionate-degrading granules before the test, formate was oxidized to produce inorganic sulfate reducers, which directly uses propionate as electron donor to carbon in the absence of sulfate (Fig. S4D). In addition, 2.6 mg L− 1 of H2 reduce sulfate. The oxidation pathway includes partial oxidation of were detected in the headspace of serum bottles at the end of the test. propionate to form acetate and bicarbonate and complete oxidation of Thus, formate was converted into CO2 and H2 according to Eq. (5) propionate to bicarbonate (Luis, 2018). In the absence of sulfate, acetate (Table 1). In the presence of sulfate, sulfate was reduced to TDS, while and bicarbonate were not produced in this work, showing that there was formate was oxidized to carbonate (Fig. 2D). Similarly, 7.6 mg L− 1 of H2 no acetogenesis from propionate. However, propionate was incom­ were also detected after cultivation. It has been reported that formate is pletely oxidized to acetate in the presence of sulfate while acetate could rapidly oxidized into CO2 and H2; then, hydrogen can be used as electron not be further fermented to produce methane nor used to reduce sulfate, donor for further sulfate reduction by hydrogenotrophic SRB (Bijmans as shown in supplementary material (Fig. S5). Although many studies et al., 2008) according to Eq. (6) (Table 1). It is worth mentioning that have shown that acetate can be oxidized to CO2 by SRB, including formate can also be used as electron donor to directly reduce sulfate by Desulforhabdus amnigenus, Desulfobacca acetoxidans, Desulfarculus baarsii SRB (De Smul and Verstraete, 1999; Liamleam and Annachhatre, 2007). and Desulfatitalea (Higashioka et al., 2013; Oude Elferink et al., 1998; If all formate was oxidized and the H2 produced was used for sulfate Widdel and Pfennig, 1977), they were not present in this work as reduction only according to Eqs. (5) and (6), respectively, the corre­ confirmed by 16S rRNA sequence analysis. In previous works in the sponding SRR calculated by the stoichiometric equations would be 3.8 same UASB reactor fed with crude glycerol as studied herein, it was ± 1 mg S g VSS− 1 h− 1. Instead, the actual experimental SRR was 5.2 ± reported that acetate accumulated in the long-term after increasing the 0.9 mg S g VSS− 1 h− 1. Consequently, formate was converted into CO2 sulfate loading rate or after a 24 h pH shock (Fernández-Palacios et al., and H2 as well as used as electron donor for sulfate reduction (Eq. (7)). In 2019; Mora et al., 2020b), since methanogens were probably the only addition, hydrogen was used by SRB as electron donor for sulfate microorganisms that consumed acetate. After methanogens wash out, reduction as represented by Eq. (6). acetate cannot be further oxidized due to the lack of acetotrophic SRB. 7 X. Zhou et al. Chemosphere 286 (2022) 131649 Fig. 3. The metabolic pathway for the conversion of 1,3-propanediol into propionate, acetate and bicarbonate proposed in this work based on the mechanisms described in Qatibi et al. (1991b) (vertical pathway) and Liu and Liu (2016) (horizontal pathway). 3HPCS, 3-hydroxypropionyl-CoA synthetase; 3HPCD, 3-hydrox­ ypropionyl-CoA dehydratase; ACR, acryloyl-CoA reductase. Despite butanol was not found in the batch test with glycerol (Fig. 1A be further confirmed. Although sulfidogenic granular sludge can oxidize and C), it has been reported as an electron donor for SRB to reduce butanol to butyrate, no other organic compounds and inorganic carbon sulfate. Particularly, Desulfovibrio spp. are capable to incompletely were produced during the oxidation of butanol to butyrate, which means oxidize butanol to succinate coupled to sulfate reduction (Dowling et al., that SRB could not further use butyrate formed to reduce sulfate. As 1992). Sarti and Zaiat (2011) found that butanol was converted to acetic observed in Fig. 1C and D, glycerol was not converted into butanol or acid by incompletely oxidizing SRB in an anaerobic sequential batch butyrate in the sulfate reduction process. Similarly, no VFAs were pro­ reactor. Due to the significant presence of Desulfovibrio, it was decided to duced when only 2,3-butanediol was added (Fig. S6A). Similar results test butanol with the granular biomass from the UASB reactor. In the were obtained adding sulfate (Fig. S6B), thus indicating that there were work herein, butanol was not degraded in the absence of sulfate no butanediol-degrading anaerobes in this work. (Fig. S4E). When sulfate was added, all butanol was converted to butyrate while no acetate, other VFA nor alcohols were produced 3.2.4. The mechanism of biodegradation of glycerol and sulfate reduction (Fig. 2E). This observation has not been reported previously in literature Based on the batch tests results, a mechanism for sulfate reduction and the specific strain of SRB able to oxidize butanol to butyrate needs to using glycerol as electron donor was proposed in this work (Fig. 4). From Fig. 4. The main mechanism of biodegradation of glycerol and sulfate reduction without methane pro­ duction. The repeated degradation and sulfate reduc­ tion pathways are not listed, such as propionate produced by reaction 11 and formate produced by reaction 4. Black reactions and continuous lines represent the fermentation process, while red re­ actions and discontinuous lines represent the sulfido­ genic process. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) 8 X. Zhou et al. Chemosphere 286 (2022) 131649 the sulfate reduction process performed by each single electron donor under non-methanogenic conditions, it was concluded that the granular (Vallero et al., 2004a, sludge taken from the UASB operated under long-term sulfidogenic Wu et al. (2018) conditions while fed with crude glycerol was 1) able to ferment glycerol to acetate, formate, propionate, ethanol and 1,3-propanediol; 2) able to This study This study This study This study This study Reference use H2, formate, propionate, ethanol and 1,3-propanediol for sulfate reduction; 3) able to oxidize formate to CO2 and H2; and 4) there were no b) acetogens that used propionate, ethanol nor 1,3-propanediol directly. Despite other intermediates not considered in this mechanism play a role 1 Specific sulfate reduction rate (mg S g VSS− in the detailed metabolic pathways of the microbial cultures grown in the UASB, the approach proposed herein provides a simplified view of the granular sludge activity through the main VFA and alcohols involved in glycerol fermentation. Such an approach contributes to expand the knowledge of sulfate reducing bioreactors and clarifies the mechanism of sulfate reduction through intermediate products and their contribu­ tion to the sulfate reduction process. 3.3. Contribution of different electron donors to sulfate reduction a c c c c c 9.5 ± 1.9 5.2 ± 0.9 3.7 ± 0.8 5.8 ± 1.5 3.9 ± 1.1 4.3 ± 1.6 b 66.2 h− 1) Table 2 shows the specific rates of different electron donors con­ sumption and sulfate reduction in batch activity tests. Specific rates were obtained by dividing volumetric rates by VSS, while volumetric 1 Specific substrate consumption rate (mg C g VSS− rates were calculated based on data from Fig. S4 and Fig. 2 as described in section 2.3. In the present work, SRRs of tests performed with single electron donors were ranked as follows: ethanol > formate > butanol >1,3-propanediol > propionate. The specific SRRs using ethanol and formate were 5.8 and 5.2 mg S g VSS− 1 h− 1, respectively. Thus, the granular sludge cultivated under sulfidogenic anaerobic conditions of continuous feeding with crude glycerol promoted formate-utilizing and ethanol-utilizing SRB species. SRB tend to use low molecular weight (MW) organic compounds in the cultivation of organic matrix mixtures (Neculita et al., 2007), which also has the benefit of improving sulfate The specific rate was calculated by taking out the granular sludge from the UASB after 330 days of operation. The specific rate was calculated by taking out the granular sludge from the UASB after 431 days of operation. The specific rate was calculated by taking out the granular sludge from the UASB after 55 days of operation. reduction efficiency (Zhao et al., 2010). As can be observed from 92.8 ± 6.5 11.6 ± 4.4 6.9 ± 1.5 4.4 ± 0.5 9.3 ± 5.1 4.7 ± 0.5 Table 2, the SRR obtained using formate and ethanol (compounds con­ N.A. h− 1) taining two carbons or less) was 39% higher than that obtained with compounds containing three carbons or more on average (propionate, 1, 3-propanediol and butanol), which is in agreement with Neculita et al. (2007). Vallero et al. (2004a) reported a SRR of 9.5 mg S g VSS− 1 h− 1 in Final N.A. N.A. 7.8 7.5 7.6 7.8 7.5 pH a methanogenic sludge fed with formate while Wu et al. (2018) reported a SRR of 66.2 mg S g VSS− 1 h− 1 using ethanol as sole carbon source, the Specific rates obtained from the sulfidogenic anaerobic digestion in batch activity tests. later reported in a sulfidogenic dominant stage. The difference of SRRs Initial N.A. 7.8 7.9 7.9 8.4 8.4 8.4 between our work and previous studies was related to the development pH of microbial populations acclimated to a specific carbon source providing thus a larger affinity for such low-MW electron donors. (◦ C) 65 35 35 35 35 35 35 Table 3 shows the contribution of each electron donor to the overall T SRR from the batch tests in which pure glycerol and sulfate were fed. The calculation process is shown in supplementary material (section S–SO2- S3.2). From 33 to 96 h, sulfate was mainly reduced using ethanol and 4 TOC/ 0.37 0.27 1.52 1.41 1.35 1.43 1.40 1,3-propanediol as electron donors. The overall contribution of these two compounds to the total sulfate reduction along the test with glycerol plus sulfate accounted for 78.6%. However, within 33–48 h, SRB mainly Initial carbon concentration (mg C g used ethanol to reduce sulfate rather than formate or 1,3-propanediol since the concentration of ethanol fermented through glycerol was 2.4 times that of formate after 33 h (Fig. 1C and D). After ethanol was consumed in the sulfate reduction process (48 h), 1,3-propanediol became the main electron donor for the reduction of sulfate. Once glycerol was completely degraded after 33 h (Fig. 1C), the production of 1,3- propanediol was 4.4 times that of formate. According to Table 2, the specific SRR for formate was higher than that of 1,3-propanediol. Note: N.A. not applicable. 1500 L− 1) However, formate was not a key contributor to the overall sulfate 750 372 342 318 339 334 reduction, probably due to its low-production from glycerol fermenta­ tion. The predominance of ethanol and 1,3-propanediol as main electron Electron donor propanediol donors for sulfate reduction is consistent with previous works in which Propionate glycerol fermentation in anaerobic digestion lead also to a major pro­ Formate Formate Butanol Ethanol Ethanol Table 2 duction of both compounds (Metsoviti et al., 2012; Rossi et al., 2012; Wu 1,3- et al., 2011). However, it is worth mentioning that the contribution of b a c 9 X. Zhou et al. Chemosphere 286 (2022) 131649 Table 3 student. Stoichiometric specific SRRs according to Eqs. (1)–(11), proposed mechanism and percentage contribution of each electron donor to the observed SRR of Appendix A. Supplementary data granular sludge fed with pure glycerol. Time Formate Propionate 1,3- Ethanol Observed Supplementary data to this article can be found online at https://doi. (h) and H2 mg mg S g propanediol mg S g SRR mg S g org/10.1016/j.chemosphere.2021.131649. S g VSS− 1 VSS− 1 h− 1 mg S g VSS− 1 VSS− 1 VSS− 1 h− 1 h− 1 (%) (%) h− 1 (%) h− 1 (%) (%) Credit author statement 33–48 2.2 0 2.6 (25.3%) 5.9 10.4 (20.9%) (57.0%) 48–72 0.4 ± 0.57 1.1 ± 2.2 5.2 ± 1.2 0 6.7 ± 1.8 Xudong Zhou: Conceptualization, Methodology, Validation, Formal (6.0 ± (16.2 ± (77.7 ± analysis, Investigation, Data curation, Writing – original draft, Visuali­ 8.7%) 33.2%) 27.1%) zation Eva Fernández-Palacios: Validation, Formal analysis, Investiga­ 72–96 0 0.1 ± 1.1 3.3 ± 0.5 0 3.7 ± 0.6 tion, Data curation, Writing – review & editing, Visualization. Antoni D. (3.3 ± (88.8 ± Dorado: Writing – review & editing, Visualization, Supervision Xavier 31.9%) 20.5%) Gamisans: Resources, Writing – review & editing, Supervision, Funding acquisition. David Gabriel: Conceptualization, Methodology, Valida­ the different electron donors depends on several factors such as physi­ tion, Resources, Writing – review & editing, Visualization, Supervision, ological differences among microbial cultures, the type of fermentation, Project administration, Funding acquisition substrate concentration and possible inhibitions (Biebl et al., 1999). Wu et al. (2011) reported that an excess amount of glycerol may cause a References metabolic product shift from ethanol to 1,3-propanediol. Therefore, conditions can be accommodated to create an environment that is more Abou-Zeid, D.M., Biebl, H., Spröer, C., Müller, R.J., 2004. Propionispora hippei sp. nov., a novel Gram-positive, spore-forming anaerobe that produces propionic acid. Int. J. conducive to the growth of SRB, such as the production of simpler car­ Syst. Evol. Microbiol. 54, 951–954. https://doi.org/10.1099/ijs.0.03054-0. bon sources that SRB would use preferentially over complex ones. Angeloni, M., Remacha, P., Martínez, A., Ballester, J., 2016. Experimental investigation of the combustion of crude glycerol droplets. Fuel 184, 889–895. https://doi.org/ 10.1016/j.fuel.2016.06.045. 4. Conclusions Apha, Awwa, 2005. Standard methods for the examination of water and wastewater. Wef 21, 5–72. https://doi.org/10.5209/rev_ANHM.2012.v5.n2.40440. The research in this work permitted to establish a simplified mech­ Bertolino, S.M., Melgaço, L.A., Sá, R.G., Leão, V.A., 2014. Comparing lactate and glycerol as a single-electron donor for sulfate reduction in fluidized bed reactors. anism of sulfate reduction by granular sludge from a long-term operated Biodegradation 25, 719–733. https://doi.org/10.1007/s10532-014-9694-1. sulfidogenic UASB using glycerol as electron donor. It also provides the Biebl, H., Menzel, K., Zeng, A.P., Deckwer, W.D., 1999. Microbial production of 1,3- experimental rates and stoichiometric equations considering the propanediol. Appl. Microbiol. Biotechnol. https://doi.org/10.1007/s002530051523. Bijmans, M.F.M., Peeters, T.W.T., Lens, P.N.L., Buisman, C.J.N., 2008. High rate sulfate contribution of the intermediate products produced by the fermentation reduction at pH 6 in a pH-auxostat submerged membrane bioreactor fed with of glycerol previous to the sulfate reduction step. Under non- formate. Water Res. 42, 2439–2448. https://doi.org/10.1016/j.watres.2008.01.025. methanogenic but sulfidogenic conditions, SRB did not directly used Cheng, K.K., Zhang, J.A., Liu, D.H., Sun, Y., Liu, H.J., Yang, M. De, Xu, J.M., 2007. Pilot- scale production of 1,3-propanediol using Klebsiella pneumoniae. Process Biochem. glycerol to reduce sulfate. Glycerol was firstly fermented by the granular 42, 740–744. https://doi.org/10.1016/j.procbio.2007.01.001. sludge to form simpler intermediates such as H2, formate, propionate, De Smul, A., Verstraete, W., 1999. Retention of sulfate-reducing bacteria in expanded ethanol, 1,3-propanediol, and then SRB reduced sulfate with such in­ granular-sludge-blanket reactors. Water Environ. Res. 71, 427–431. https://doi.org/ termediate products. The sulfate reduction process mainly used 1,3-pro­ 10.2175/106143097x122077. Dinkel, V.G., Frechen, F.B., Dinkel, A.V., Smirnov, Y.Y., Kalyuzhnyi, S.V., 2010. Kinetics panediol and ethanol as electron donors, which were produced through of anaerobic biodegradation of glycerol by sulfate-reducing bacteria. Appl. Biochem. glycerol fermentation. Butanol was not an intermediate product of Microbiol. 46, 712–718. https://doi.org/10.1134/s0003683810070069. glycerol fermentation, but it was capable to be oxidized to butyrate in Dowling, N.J.E., Brooks, S.A., Phelps, T.J., White, D.C., 1992. Effects of selection and fate of substrates supplied to anaerobic bacteria involved in the corrosion of pipe-line the presence of sulfate. SRB performing an incomplete oxidation of the steel. J. Ind. Microbiol. 10, 207–215. https://doi.org/10.1007/BF01569768. electron donors dominated in the granular sludge from the sulfidogenic El Houari, A., Ranchou-Peyruse, M., Ranchou-Peyruse, A., Dakdaki, A., Guignard, M., UASB reactor. SRB preferred to use simple intermediate products con­ Idouhammou, L., Bennisse, R., Bouterfass, R., Guyoneaud, R., Qatibi, A.I., 2017. Desulfobulbus oligotrophicus sp. Nov., a sulfate-reducing and propionate-oxidizing taining less than 2 carbons as electron donors, except in the case of bacterium isolated from a municipal anaerobic sewage sludge digester. Int. J. Syst. acetate that was not found to be used for sulfate reduction. Desulfobulbus Evol. Microbiol. 67, 275–281. https://doi.org/10.1099/ijsem.0.001615. and Desulfovibrio were found as the main sulfate reducing genus, ac­ Fernández-Palacios, E., Lafuente, J., Mora, M., Gabriel, D., 2019. Exploring the performance limits of a sulfidogenic UASB during the long-term use of crude glycerol counting for 22.4% of the total amount of sequences detected under as electron donor. Sci. Total Environ. 688, 1184–1192. https://doi.org/10.1016/j. sulfidogenic conditions. scitotenv.2019.06.371. Fernández-Palacios, E., Zhou, X., Mora, M., Gabriel, D., 2021. Microbial diversity dynamics in a methanogenic-sulfidogenic UASB reactor. Int. J. Environ. Res. Publ. Declaration of competing interest Health 18, 1–16. https://doi.org/10.3390/ijerph18031305. Higashioka, Y., Kojima, H., Watanabe, M., Fukui, M., 2013. Desulfatitalea tepidiphila gen. The authors declare that they have no known competing financial nov., sp. nov., a sulfate-reducing bacterium isolated from tidal flat sediment. Int. J. Syst. Evol. Microbiol. 63, 761–765. https://doi.org/10.1099/ijs.0.043356-0. interests or personal relationships that could have appeared to influence Jarvis, G.N., Moore, E.R.B., Thiele, J.H., 1997. Formate and ethanol are the major the work reported in this paper. products of glycerol fermentation produced by a Klebsiella planticola strain isolated from red deer. J. Appl. Microbiol. 83, 166–174. https://doi.org/10.1046/j.1365- Acknowledgments 2672.1997.00217.x. Karjalainen, J., Mäkinen, M., Karjalainen, A.K., 2021. Sulfate toxicity to early life stages of European whitefish (Coregonus lavaretus) in soft freshwater. Ecotoxicol. Environ. Authors are members of the GENOCOV research group from the Saf. 208, 111763. https://doi.org/10.1016/j.ecoenv.2020.111763. Department of Chemical, Biological and Environmental Engineering at Katarzyna, L., Katarzyna, C., Kamila, M., 2011. Biotechnological synthesis of 1,3-pro­ panediol using Clostridium ssp. Afr. J. Biotechnol. 10, 11093–11101. https://doi.org/ UAB (Universitat Autònoma de Barcelona), and the Department of 10.5897/ajb11.873. Mining, Industrial and ICT Engineering, Universitat Politècnica de Cat­ Kumar, L.R., Yellapu, S.K., Tyagi, R.D., Zhang, X., 2019. A review on variation in crude alunya. Authors acknowledge the Spanish Government, through the glycerol composition, bio-valorization of crude and purified glycerol as carbon source for lipid production. Bioresour. Technol. 293, 122155. https://doi.org/ project RTI2018-099362-B-C21 MINECO/FEDER, EU, for the financial 10.1016/j.biortech.2019.122155. support provided to perform this research. Authors thank the China Scholarship Council (CSC, 201706300052) for funding this doctoral 10 X. Zhou et al. Chemosphere 286 (2022) 131649 Lens, P.N.L., Kuenen, J.G., 2001. The biological sulfur cycle: novel opportunities for (Eds.), The Prokaryotes: Prokaryotic Physiology and Biochemistry. Springer Berlin environmental biotechnology. In: Water Science and Technology, pp. 57–66. Heidelberg, pp. 309–404. https://doi.org/10.2166/wst.2001.0464. Rossi, D.M., da Costa, J.B., de Souza, E.A., Peralba, M. do C.R., Ayub, M.A.Z., 2012. Li, Y., Sun, Y., Yang, G., Hu, K., Lv, P., Li, L., 2017. Vertical distribution of microbial Bioconversion of residual glycerol from biodiesel synthesis into 1,3-propanediol and community and metabolic pathway in a methanogenic propionate degradation ethanol by isolated bacteria from environmental consortia. Renew. Energy 39, bioreactor. Bioresour. Technol. 245, 1022–1029. https://doi.org/10.1016/j. 223–227. https://doi.org/10.1016/j.renene.2011.08.005. biortech.2017.09.028. Sakamoto, M., 2014. The family Porphyromonadaceae. In: Rosenberg, E., DeLong, E.F., Liamleam, W., Annachhatre, A.P., 2007. Electron donors for biological sulfate reduction. Lory, S., Stackebrandt, E., Thompson, F. (Eds.), The Prokaryotes - Other Major Biotechnol. Adv. 25, 452–463. https://doi.org/10.1016/j.biotechadv.2007.05.002. Lineages of Bacteria and the Archaea. Springer Berlin Heidelberg, pp. 811–824. Liu, Z., Liu, T., 2016. Production of acrylic acid and propionic acid by constructing a Santos, S.C., Liebensteiner, M.G., van Gelder, A.H., Dimitrov, M.R., Almeida, P.F., portion of the 3-hydroxypropionate/4-hydroxybutyrate cycle from Metallosphaera Quintella, C.M., Stams, A.J.M., Sánchez-Andrea, I., 2018. Bacterial glycerol sedula in Escherichia coli. J. Ind. Microbiol. Biotechnol. 43, 1659–1670. https://doi. oxidation coupled to sulfate reduction at neutral and acidic pH. J. Gen. Appl. org/10.1007/s10295-016-1843-6. Microbiol. 64, 1–8. https://doi.org/10.2323/jgam.2017.02.009. Lomans, B.P., Van der Drift, C., Pol, A., Op den Camp, H.J.M., 2002. Microbial cycling of Sarti, A., Zaiat, M., 2011. Anaerobic treatment of sulfate-rich wastewater in an anaerobic volatile organic sulfur compounds. Cell. Mol. Life Sci. 59, 575–588. https://doi.org/ sequential batch reactor (AnSBR) using butanol as the carbon source. J. Environ. 10.1007/s00018-002-8450-6. Manag. 92, 1537–1541. https://doi.org/10.1016/j.jenvman.2011.01.009. Luis, C.R.A., 2018. Optimization of the Electron Donor Supply to Sulphate Reducing Schauder, R., Schink, B., 1989. Anaerovibrio glycerini sp. nov., an anaerobic bacterium Bioreactors Treating Inorganic Wastewater. CRC Press. https://doi.org/10.1017/ fermenting glycerol to propionate, cell matter, and hydrogen. Arch. Microbiol. 152, CBO9781107415324.004. 473–478. https://doi.org/10.1007/BF00446932. Madigan, M.T., Martinko, J.M., Parker, J., 1997. Brock Biology of Microorganisms, vol. Sittijunda, S., Reungsang, A., 2017. Fermentation of hydrogen, 1,3-propanediol and 11. Prentice hall, Upper Saddle River, NJ. ethanol from glycerol as affected by organic loading rate using up-flow anaerobic Maillacheruvu, K.Y., Parkin, G.F., 1996. Kinetics of growth, substrate utilization and sludge blanket (UASB) reactor. Int. J. Hydrogen Energy 42, 27558–27569. https:// sulfide toxicity for propionate, acetate, and hydrogen utilizers in anaerobic systems. doi.org/10.1016/j.ijhydene.2017.05.149. Water Environ. Res. 68, 1099–1106. https://doi.org/10.2175/106143096x128126. Trabue, S., Scoggin, K., Tjandrakusuma, S., Rasmussen, M.A., Reilly, P.J., 2007. Ruminal Metsoviti, M., Paramithiotis, S., Drosinos, E.H., Galiotou-Panayotou, M., Nychas, G.J.E., fermentation of propylene glycol and glycerol. J. Agric. Food Chem. 55 (17), Zeng, A.P., Papanikolaou, S., 2012. Screening of bacterial strains capable of 7043–7051. https://doi.org/10.1021/jf071076i. converting biodiesel-derived raw glycerol into 1,3-propanediol, 2,3-butanediol and Vallero, M.V.G., Camarero, E., Lettinga, G., Lens, P.N.L., 2004a. Thermophilic (55-65◦ C) ethanol. Eng. Life Sci. 12, 57–68. https://doi.org/10.1002/elsc.201100058. and extreme thermophilic (70-80◦ C) sulfate reduction in methanol and formate-fed Mora, M., Fernández-Palacios, E., Guimerà, X., Lafuente, J., Gamisans, X., Gabriel, D., UASB reactors. Biotechnol. Prog. 20, 1382–1392. https://doi.org/10.1021/ 2020a. Feasibility of S-rich streams valorization through a two-step biosulfur bp034329a. production process. Chemosphere 126734. https://doi.org/10.1016/j. Vallero, M.V.G., Sipma, J., Lettinga, G., Lens, P.N.L., 2004b. High-rate sulfate reduction chemosphere.2020.126734. at high salinity (up to 90 mS.cm-1) in mesophilic UASB reactors. Biotechnol. Bioeng. Mora, M., Lafuente, J., Gabriel, D., 2020b. Influence of crude glycerol load and pH 86, 226–235. https://doi.org/10.1002/bit.20040. shocks on the granulation and microbial diversity of a sulfidogenic Upflow Viana, M.B., Freitas, A.V., Leitão, R.C., Pinto, G.A.S., Santaella, S.T., 2012. Anaerobic Anaerobic Sludge Blanket reactor. Process Saf. Environ. Protect. 133, 159–168. digestion of crude glycerol: a review. Environ. Technol. Rev. 1, 81–92. https://doi. https://doi.org/10.1016/j.psep.2019.11.005. org/10.1080/09593330.2012.692723. Muyzer, G., Stams, A.J.M., 2008. The ecology and biotechnology of sulphate-reducing Vivek, N., Sindhu, R., Madhavan, A., Anju, A.J., Castro, E., Faraco, V., Pandey, A., bacteria. Nat. Rev. Microbiol. 6, 441–454. https://doi.org/10.1038/nrmicro1892. Binod, P., 2017. Recent advances in the production of value added chemicals and Neculita, C.-M., Zagury, G.J., Bussière, B., 2007. Passive treatment of acid mine drainage lipids utilizing biodiesel industry generated crude glycerol as a substrate – metabolic in bioreactors using sulfate-reducing bacteria. J. Environ. Qual. 36, 1–16. https:// aspects, challenges and possibilities: an overview. Bioresour. Technol. https://doi. doi.org/10.2134/jeq2006.0066. org/10.1016/j.biortech.2017.05.056. Oude Elferink, S.J.W.H., Luppens, S.B.I., Marcelis, C.L.M., Stams, A.J.M., 1998. Kinetics Weijma, J., Stams, A.J.M., 2001. Methanol conversion in high-rate anaerobic reactors. of acetate oxidation by two sulfate reducers isolated from anaerobic granular sludge. Water Sci. Technol. 44, 7–14. https://doi.org/10.2166/wst.2001.0452. Appl. Environ. Microbiol. 64, 2301–2303. https://doi.org/10.1128/aem.64.6.2301- Widdel, F., Pfennig, N., 1977. A new anaerobic, sporing, acetate-oxidizing, sulfate- 2303.1998. reducing bacterium. Desulfotomaculum (emend.) acetoxidans. Arch. Microbiol. 112, Papurello, D., Soukoulis, C., Schuhfried, E., Cappellin, L., Gasperi, F., Silvestri, S., 119–122. https://doi.org/10.1007/BF00446665. Santarelli, M., Biasioli, F., 2012. Monitoring of volatile compound emissions during Wu, J., Niu, Q., Li, L., Hu, Y., Mribet, C., Hojo, T., Li, Y.Y., 2018. A gradual change dry anaerobic digestion of the organic fraction of municipal solid waste by proton between methanogenesis and sulfidogenesis during a long-term UASB treatment of transfer reaction time-of-flight mass spectrometry. In: Bioresource Technology. sulfate-rich chemical wastewater. Sci. Total Environ. 636, 168–176. https://doi.org/ Elsevier Ltd, pp. 254–265. https://doi.org/10.1016/j.biortech.2012.09.033. 10.1016/j.scitotenv.2018.04.172. Qatibi, A.I., Bories, A., Garcia, J.L., 1991a. Sulfate reduction and anaerobic glycerol Wu, K.J., Lin, Y.H., Lo, Y.C., Chen, C.Y., Chen, W.M., Chang, J.S., 2011. Converting degradation by a mixed microbial culture. Curr. Microbiol. 22, 47–52. https://doi. glycerol into hydrogen, ethanol, and diols with a Klebsiella sp. HE1 strain via org/10.1007/BF02106212. anaerobic fermentation. J. Taiwan Inst. Chem. Eng. 42, 20–25. https://doi.org/ Qatibi, A.I., Cayol, J.L., Garcia, J.L., 1991b. Glycerol and propanediols degradation by 10.1016/j.jtice.2010.04.005. Desulfovibrio alcoholovorans in pure culture in the presence of sulfate, or in Zeng, A.P., Biebl, H., Schlieker, H., Deckwer, W.D., 1993. Pathway analysis of glycerol syntrophic association with Methanospirillum hungatei. FEMS Microbiol. Lett. 85, fermentation by Klebsiella pneumoniae: regulation of reducing equivalent balance and 233–240. https://doi.org/10.1111/j.1574-6968.1991.tb04729.x. product formation. Enzym. Microb. Technol. 15, 770–779. https://doi.org/10.1016/ Qatibi, A.I., Bennisse, R., Jana, M., Garcia, J.L., 1998. Anaerobic degradation of glycerol 0141-0229(93)90008-. by Desulfovibrio fructosovorans and D. carbinolicus and evidence for glycerol- Zhao, Y.G., Wang, A.J., Ren, N.Q., 2010. Effect of carbon sources on sulfidogenic dependent utilization of 1,2-propanediol. Curr. Microbiol. 36, 283–290. https://doi. bacterial communities during the starting-up of acidogenic sulfate-reducing org/10.1007/s002849900311. bioreactors. Bioresour. Technol. 101, 2952–2959. https://doi.org/10.1016/j. Rabus, R., Hanses, T.A., Widdel, F., 2013. Dissimilatory sulfate- and sulfur- reducing biortech.2009.11.098. prokaryotes. In: Rosenberg, E., DeLong, E.F., Lory, S., Stackebrandt, E., Thompson, F. 11