Biochemical Methane Potential of a Bioreﬁnery’s Process-Wastewater and its Components at Different Concentrations and Temperatures

: A sustainable circular bioeconomy requires the side streams and byproducts of bioreﬁneries to be assimilated into bioprocesses to produce value-added products. The present study endeavored to utilize such a byproduct generated during the synthesis of 5-hydroxymethylfurfural as a potential feedstock for biogas production. For this purpose, biochemical methane potential tests for the full process-wastewater, its components (5-hydroxymethylfurfural, furfural, levulinic acid, and glycolic acid), together with furfural’s metabolites (furfuryl alcohol and furoic acid), and phenols (syringaldehyde, vanillin, and phenol), were conducted at mesophilic and thermophilic temperatures to assess their biodegradability and gas production kinetics. 0.1, 0.2, 0.3, and 0.4 g COD of the test components were added separately into assays containing 35 mL of inoculum. At their lowest concentrations, the test components, other than the process-wastewater, exhibited a stimulatory effect on methane production at 37 ◦ C, whereas their increased concentrations returned a lower mean speciﬁc methane yield at either temperature. For similar component loads, the mesophilic assays outperformed the thermophilic assays for the mean measured speciﬁc methane yields. Components that impaired the anaerobic process with their elevated concentrations were phenol, vanillin, and 5-hydroxymethylfurfural. Poor degradation of the process-wastewater was deduced to be linked to the considerable share of 5-hydroxymethylfurfural in the process-wastewater governing its overall characteristics. With excessive recalcitrant components, it is recommended to use such waste streams and byproducts as a substrate for biogas plants operating at moderate temperatures, but at low rates.


Introduction
The use of non-renewable fossil resources in current production practices has led to serious environmental issues. A probable solution to offset such concerns lies in a gradual transition from petroleum-based economies to biobased economies, with regenerative resources being the primary production input [1]. By encompassing different conversion processes and techniques, biorefineries convert the biobased feedstocks into biofuels, biobased materials, and equally important platform chemicals [2].
Platform chemicals are intermediates that function as building blocks to manufacture a variety of value-added products including biochemicals. A prime example of such an intermediate is 5-hydroxymethylfurfural (5-HMF), which is perceived as a versatile platform compound due to its broad industrial application spectrum; for instance, in pharmaceuticals, agrochemicals, textiles, plastics, resins, paints, and adhesives to name a few [1,3,4]. In addition, the US Department of Energy rates 5-HMF and its derivatives, levulinic acid, 2,5-Furandicarboxylic acid (FDCA), and 2,5-diformylfuran (DFF) as the topmost promising biomass-derived chemicals [5,6].
The synthesis of 5-HMF from biobased feedstock is usually carried out via a thermochemical conversion (TCC) process (200 • C, 18 min, 24 bar [7]), which, in general, initiates with the acid hydrolysis of cellulose and hemicellulose [8], followed by the isomerization of glucose [9], and concludes with the acid-catalyzed dehydration of fructose [10]. Coupled with the end bioproduct (5-HMF), a hazardous byproduct (process-wastewater) concurrently rich in organics [11] and moderately in inorganics is also produced [12]. For a circular bioeconomy with a zero-waste approach, residues and waste streams generated in biorefineries have to be transformed into value-added products as well [13]. Hence, biochemical conversion, in particular anaerobic digestion (AD), is the paradigm that can contribute toward the valorization of such biorefineries' byproducts [14,15]. AD is defined as the biodegradation of organic wastes in the absence of oxygen to produce a renewable energy source (biogas) that can be converted by a combined heat and power unit (CHP) into heat and electricity [16].
Process-wastewater or aqueous materials produced during different TCC techniques (e.g., hydrothermal carbonization (HTC), hydrothermal liquefaction (HTL), and pyrolysis) form various biobased feedstock such as agricultural [12] and agroindustrial wastes [17], forestry residues [18,19], food [20] and animal waste [21], lignocellulosic biomass [22], and municipal waste [23] were investigated for their biogas potential. The process parameters and the feedstock nature were determined to be the decisive factor for the ultimate composition of the wastewaters produced [24], thereby resulting in distinctive methane yields [25,26]. By elevating the process severity (temperature, retention time, catalyst concentration), significant amounts of soluble organic substances are released in comparison to the biomass treated with low severity parameters [18,26]. Following the extraction of the target components, the undesired constituents and side streams, comprising mainly of refractory and recalcitrant substances (e.g., acidic [24], aromatics [27], and nitrogenated compounds [28]) are washed down as process-wastewater, consequently upsurging the complexity of the generated byproduct [26]. Moreover, regardless of the pretreatment (e.g., steam explosion, hot water extraction, wet oxidation), or the treatment procedures (as in TCC) applied to the lignocellulosic feedstock, the aqueous phase or the process-wastewater produced as a result, will constitute components of furan and phenol origin together with weak acids [27,29].
As improper management or disposal of such hazardous materials may lead to environmental nuisances on account of their noxious characteristics, perhaps with anaerobic technologies, such byproducts can be detoxified while simultaneously producing biogas. Provision of such materials in fermenters does prompt gas production, but only to a certain extent (in terms of their loading rates and the subsequent microbial tolerance to it), whereafter they may exert a converse effect on the anaerobic consortia, thus leading to partial inhibition or deterioration of the process [25,30]. Complete inhibition of the fermentation process is linked to the furans and the phenols forming reactive oxygen species (ROS), disrupting or inactivating essential enzymes, distorting genetic materials and cell organelles, and weakening cell membranes, among others, resulting in the onset of cell apoptosis [31,32]. Similarly, inhibition due to weak acids, according to Palmqvist et al. [33], is the outcome of intracellular accumulation of anions, along with the increased level of protons (H + ) in the cytosol.
Thus, the present study aims to assess the feasibility of a biorefinery's processwastewater as a potential biogas plant feedstock by investigating the biochemical methane potential (BMP) of its typical constituents such as furans, weak acids, and phenols, plus the full 5-HMF process-wastewater, separately at different concentrations and temperatures through batch fermentation (i.e., the Hohenheim biogas yield test (HBT)). With the application of the modified Gompertz model (MGM), an insight into their biodegradation kinetics at different digestion conditions will be provided. The potency or the half maximal inhibitory concentration (IC 50 ) of the components at 37 • C and 53 • C can be evaluated through the estimated parameter maximum methane production rate of the MGM. Such tests will predict the suitability of utilizing the biorefineries' side streams on a practical scale for biogas production, as well as their fate (concerning either their degradation or hampering of the process) in the biogas plants.
5-HMF process-wastewater was selected as a representative biorefinery byproduct in the current study. The individual test component representing the class furans included 5-HMF, furfural, and furfuryl alcohol, class weak acids comprised furoic acid, levulinic acid, and glycolic acid, whereas class phenols contained syringaldehyde, vanillin, and phenol. Similarly, the technique executed (HBT) is considered to be a high-efficiency batch anaerobic process employed for evaluating the BMP of different substrates and organic waste streams [34,35].

Batch Assays
The BMP of the test components was conducted via the HBT as per VDI-4630 standards [36]. In such a fermentation test, a 100 mL calibrated glass syringe, with a plunger and a capillary extension at either end, acts as a bioreactor, as shown in Figure 1.
plus the full 5-HMF process-wastewater, separately at different concentrations and temperatures through batch fermentation (i.e., the Hohenheim biogas yield test (HBT)). With the application of the modified Gompertz model (MGM), an insight into their biodegradation kinetics at different digestion conditions will be provided. The potency or the half maximal inhibitory concentration (IC50) of the components at 37 °C and 53 °C can be evaluated through the estimated parameter maximum methane production rate of the MGM. Such tests will predict the suitability of utilizing the biorefineries' side streams on a practical scale for biogas production, as well as their fate (concerning either their degradation or hampering of the process) in the biogas plants.
5-HMF process-wastewater was selected as a representative biorefinery byproduct in the current study. The individual test component representing the class furans included 5-HMF, furfural, and furfuryl alcohol, class weak acids comprised furoic acid, levulinic acid, and glycolic acid, whereas class phenols contained syringaldehyde, vanillin, and phenol. Similarly, the technique executed (HBT) is considered to be a high-efficiency batch anaerobic process employed for evaluating the BMP of different substrates and organic waste streams [34,35].

Batch Assays
The BMP of the test components was conducted via the HBT as per VDI-4630 standards [36]. In such a fermentation test, a 100 mL calibrated glass syringe, with a plunger and a capillary extension at either end, acts as a bioreactor, as shown in Figure 1. The test substrate, along with the inoculum, is brought into the syringe sampler and is promptly sealed with the application of an inert lubricating material (Korasilon by Kurt Obermeier, Bad-Berleburg, Germany) to the plunger and by fastening a clamp to the silicone pipe, fitted onto the capillary extension. A thermostat-controlled climatic chamber (INE 700 by Memmert, Schwabach, Germany) equipped with a rotating drum is utilized for incubation, where the filled syringe samplers are fitted horizontally for proper mixing of the substrate with the inoculum. The produced biogas displaces the plunger, and the volume is read on the graduated scale daily. The said biogas is manually directed from the syringe sampler into a calibrated gas transducer (AGM 10 by Pronova, Berlin, Germany) for measuring the methane content. After the complete release of the biogas, the clamp is fastened and the sampler is brought back into the climatic chamber for further incubation.

Inoculum
Two distinctive seeding materials, thermophilic and mesophilic inocula, were utilized for the current study, where the former was obtained from a farm-scale thermophilic biogas plant in Isny, Germany, and the latter from a 400 L lab-scale mesophilic anaerobic reactor (37 °C) at the State Institute of Agricultural Engineering and Bioenergy in Hohenheim, Germany. The test substrate, along with the inoculum, is brought into the syringe sampler and is promptly sealed with the application of an inert lubricating material (Korasilon by Kurt Obermeier, Bad-Berleburg, Germany) to the plunger and by fastening a clamp to the silicone pipe, fitted onto the capillary extension. A thermostat-controlled climatic chamber (INE 700 by Memmert, Schwabach, Germany) equipped with a rotating drum is utilized for incubation, where the filled syringe samplers are fitted horizontally for proper mixing of the substrate with the inoculum. The produced biogas displaces the plunger, and the volume is read on the graduated scale daily. The said biogas is manually directed from the syringe sampler into a calibrated gas transducer (AGM 10 by Pronova, Berlin, Germany) for measuring the methane content. After the complete release of the biogas, the clamp is fastened and the sampler is brought back into the climatic chamber for further incubation.

Inoculum
Two distinctive seeding materials, thermophilic and mesophilic inocula, were utilized for the current study, where the former was obtained from a farm-scale thermophilic biogas plant in Isny, Germany, and the latter from a 400 L lab-scale mesophilic anaerobic reactor (37 • C) at the State Institute of Agricultural Engineering and Bioenergy in Hohenheim, Germany.
Prior to the trial initiation, the two inocula were mixed at a 1:1 ratio and were placed in water baths heated to 37 • C and 53 • C for two weeks to develop the respective microbial consortia and to degas the inoculum simultaneously. Moreover, to ensure a sufficiently active microbial population, a specific daily ration consisting of wheat meal, soybean extract, dry grounded maize silage, and rapeseed oil, was added into the inocula at an organic loading rate of 0.5 gVS/L d (Volatile Solid, in grams). A brief overview of the chemical and physical characteristics of the inocula, after their acclimatization at thermophilic and  Table 1. The inoculum was filtered through a 1 mm sieve before its incorporation into the HBTs with the test components.

Stock Solution Preparation
Based on the composition of the full process-wastewater reported by Khan et al. [25], 5-HMF, furfural, levulinic acid, glycolic acid, and the process-wastewater were selected as test components for the current study along with intermediates of furfural decomposition, namely, furoic acid and furfuryl alcohol [37,38]. Phenolic compounds such as syringaldehyde, vanillin, and phenol were additionally selected. When using lignocellulosic biomass for 5-HMF synthesis, phenolics may be present in wastewater, originating from the thermochemical breakdown of lignin [39,40].
Owing to the reduced solubility of 6.67 gCOD in 100 mL distilled water, syringaldehyde (98% purity, Sigma-Aldrich, Steinheim am Albuch, Germany) and vanillin (99% purity, Sigma-Aldrich, Steinheim am Albuch, Germany) were brought into the assays in their original form (solid). The properties of these components examined for their BMP are given in Table 2.

Experimental Design and Procedure
Four different concentrations (i.e., 2, 4, 6, and 8 gCOD/L) of the test components were investigated for their biodegradability at 37 • C and 53 • C for a period of sixty days. Each combination of temperature and component concentration was tested in triplicate.
In addition, to assess the accuracy of the current trial, hay and feed-concentrate (standard substrates), with known methane yields, were also subjected to their biodegradation under the aforementioned digestion conditions. To compute the methane produced from the test components and the standard substrates, the thermophilic and mesophilic inoculum, as a stand-alone variant (zero sample), was put through a similar process according to VDI 4630.
From the prepared stock solutions, 1.5, 3, 4.5, and 6 mL dilutions, containing 0.1, 0.2, 0.3, and 0.4 gCOD of the individual test components, were separately added to syringes containing 35 mL of either a mesophilic or thermophilic inoculum. For a uniform working volume of 50 mL per batch-test, the assays were balanced by adding distilled water, as specified in Table 3. The test concentrations of syringaldehyde and vanillin were brought directly (as solids) to the mesophilic and thermophilic inoculum-holding assays, along with 15 mL of distilled water. The reference assays were prepared in a similar manner by incorporating 0.32 gVS of hay and feed-concentrate with 30 mL each of mesophilic and thermophilic inoculum.
The biogas measurements were performed when its amount exceeded 20 mL in the syringe sampler. For the quality measurement of the produced gas, the sampler was connected to the methane analyzer (AGM 10 by Pronova, Berlin, Germany) via a filter with Sicapent (Merck, Darmstadt ≥ 50% H 2 O absorption capacity). The analyzer, having the capacity to measure methane content in biogas within the range of 0% to 100 %, was calibrated before the measurements with test gas (60% CH 4 + 40% CO 2 , Westfalen AG, Weissenhorn, Germany), and later flushed with ambient air.
where p (hPa) and p w (hPa) are the measured biogas pressure and the estimated vapor pressure at digestion temperature T (K), respectively. The standard biogas volume is represented by V 0 (mL). When performing the measurements, the biogas in the samplers was assumed to have cooled down by~2 • C. Hence, the estimated vapor pressure for the assays at 37 • C and 53 • C were considered to be 58 hPa and 130 hPa, respectively.
To estimate the methane production kinetics, the modified Gompertz model (MGM), which describes the biodegradation of the organics as a function of microbial growth in the assays [42,43], as represented in Equation (2), was fitted to the measured cumulative methane yield of the test components.
where Y is cumulative methane produced (mL/gCOD) and t is time (d). a, b, and c are the kinetic parameters donating the maximum methane produced (mL/gCOD), the maximum daily methane production rate (mL/gCOD d), and the lag phase (d), respectively, with the Euler's constant represented by e (=2.7183).
For the test components that exhibited a diauxic behavior, a two-phase Gompertz model (TGM), as described by Gomes et al. [44], and given as Equation (3) below, was fitted to their measured cumulative methane yields.
The preceding equation specifies the summation of the Gompertz functions fitted to the degradation pattern of the test component, starting from the initial phase (i = 1) to the final phase (n).
Statistical analysis (two-way ANOVA) of the measured specific methane yields (SMYs) of the individual test components was carried out with the Post-Hoc Tukey test at a significance level of α = 0.05. A tabulated summary of such analysis is provided as (Supplementary Material Tables S1-S4). Data processing, analysis, and visualization were conducted by implementing Microsoft Excel 2016, OriginPro 9.7, and Rstudio 4.1.2. Figure 2 represents the SMY for the inocula, feed-concentrate, and hay utilized as zero samples and standard substrates in the current study at 37 • C and 53 • C, respectively. The expected amount of the gas yields establishes hay and feed-concentrate as ideal candidates for standard substrates to assess the quality and soundness of the inoculum and digestion process, respectively [42]. The methane yields from the test components and the standard substrates were acquired by subtracting the methane yields of the zero samples from the inoculum-substrate mix assays' gas yields. Under the given digestion conditions, 0.4 g hay and 0.4 g feed-concentrate, each equivalent to 0.32 gVS added, yielded 115.1 ± 3.3 mL and 115.9 ± 4.1 mL methane at mesophilic temperatures. In contrast, the measured cumulative methane yields for the said substrates at the end of the trial, in thermophilic conditions, were 87.3 ± 3.4 mL and 107.9 ± 10.2 mL for the hay and feed-concentrate, respectively. Similarly, the corresponding mean cumulative methane yields for the zero samples at 37 °C and 53 °C were observed  Under the given digestion conditions, 0.4 g hay and 0.4 g feed-concentrate, each equivalent to 0.32 gVS added, yielded 115.1 ± 3.3 mL and 115.9 ± 4.1 mL methane at mesophilic temperatures. In contrast, the measured cumulative methane yields for the said substrates at the end of the trial, in thermophilic conditions, were 87.3 ± 3.4 mL and 107.9 ± 10.2 mL for the hay and feed-concentrate, respectively. Similarly, the corresponding mean cumulative methane yields for the zero samples at 37 • C and 53 • C were observed to be 87.6 ± 4.9 mL and 98.3 ± 2.5 mL, respectively. The SMYs for the standard substrates and the zero samples are given in Table 4. Using a similar digestion technique (HBT), Huelsemann et al. [41] observed the SMYs at mesophilic temperature for feed-concentrate and hay to be in the range of 347-361 mL/gVS and 306-321 mL/gVS, respectively, by employing five different inocula sources individually. Hence, in the current study, the obtained SMYs for the aforementioned control substrates ratified the quality and reliability of the inocula used for the BMP of the test components via HBT. Figure 3 depicts the biodegradation kinetics of the test components 5-HMF, furfural, and furfuryl alcohol, classified as furanic compounds in the current study, at their aforementioned test concentrations at 37 • C and 53 • C for a hydraulic retention time (HRT) of 60 days.

Furanic Compounds
Under mesophilic conditions, 5-HMF and furfural at 2 gCOD/L concentration and 5-HMF at 4 gCOD/L concentration, exhibited an 'overshoot' where the mean SMYs surpassed their stoichiometric theoretical limits (i.e., 350 mL/gCOD). Such a trend continued to repeat for all the test concentrations of furfuryl alcohol at 37 • C and 53 • C (except for 6 gCOD/L at 53 • C). Nielfa et al. [45] attributed such tendencies to the synergistic effect that the test substrate and the inoculum jointly have upon each other when co-digested and can be determined when the ratio of the experimental yield to the theoretical yield of the test substrate exceeds the unity (>1). It can be assumed that the metabolites formed as a result of the test components (at low concentrations) and the inoculum interaction prompted such overshoots [46].
Moreover, the conversion of the furanic compounds to methane (in terms of SMY), at both operating temperatures, declined significantly when their per batch-test added amounts were increased (i.e., from 0.1 gCOD to 0.4 gCOD). Similar outcomes for the biodegradation of such toxicants were reported by a handful of researchers where reduced concentrations of the target components were effectively degraded by the microbial consortia while their increased concentrations either hampered or completely inhibited the anaerobic process [40,47]. Comprehensive mechanisms for the toxicity of the furans are still unknown; however, the presence of the aldehyde groups is presumed to cause or initiate their inhibitory traits [4,48].  Under mesophilic conditions, 5-HMF and furfural at 2 gCOD/L concentration and 5-HMF at 4 gCOD/L concentration, exhibited an `overshoot' where the mean SMYs surpassed their stoichiometric theoretical limits (i.e., 350 mL/gCOD). Such a trend continued to repeat for all the test concentrations of furfuryl alcohol at 37 °C and 53 °C (except for 6 gCOD/L at 53 °C). Nielfa et al. [45] attributed such tendencies to the synergistic effect that the test substrate and the inoculum jointly have upon each other when co-digested and can be determined when the ratio of the experimental yield to the theoretical yield of the test substrate exceeds the unity (>1). It can be assumed that the metabolites formed as a As for the measured cumulative methane yield, BMP assays of furanic compounds put through mesophilic conditions outdid the thermophilic assays, apart from 5-HMF at 0.3 gCOD added ( Figure 3A-iii), and furfural at 0.4 gCOD added ( Figure 3B-iv). Ghimire et al. [18] and Ghasimi et al. [40] have described this phenomenon as being the result of volatile fatty acid (VFAs) accumulations, particularly propionate at the terminal phase of thermophilic AD, giving rise to process instability. Even though furfural is described to be more inhibitory as compared to 5-HMF [49], under the present circumstances, it was found to be superior to 5-HMF in terms of its biodegradability and methane production potential. Such an outcome was in accordance with the observations made by Khan et al. [25], where the percent methane conversion for similar dosages of 5-HMF and furfural in a continuous process was reported to be around 59% and 73%, respectively. Table 5 summarizes the mean methane yields for the given concentrations of 5-HMF, furfural, and furfuryl alcohol at 37 • C and 53 • C along with the estimated kinetic parameters and the model's goodness of fit. Contrary to furfural and furfuryl alcohol, 5-HMF with its increased concentrations initially inhibited the AD which is characterized by longer lag phases estimated through the parameter 'c of the MGM. Complete inhibition of the process by the furanic compounds at different test concentrations and temperature regimes did not occur. However, as observed, 5-HMF stood out to be a more recalcitrant compound in the current class of toxicants where the mean percent degradation for its highest loads tested (5.2 gVS/L or 8 gCOD/L) at 37 • C and 53 • C were 49.5% and 44.5% whereas for furfural (4.8 gVS/L) and furfuryl alcohol (4.4 gVS/L), they were 66% and 74%, and 114.8% and 104.8% at mesophilic and thermophilic temperatures respectively.
While investigating the inhibitory threshold of the furanic compounds on the methanogenesis via specific methanogenic activity (SMA) assays, Ghasimi et al. [40] found furfural and 5-HMF to be completely hampering the methanogenic archaea at individual concentrations of 2 gVS/L (furfural: 3.33 gCOD/L, 5-HMF: 3.05 gCOD/L), at thermophilic and mesophilic temperatures. Likewise, for Malik Badshah [50], 4 gVS/L furfural at 37 • C partially inhibited methanogenic activity (58% methanized), whereas 5-HMF at 6 gVS/L completely inhibited the AD process [49]. Similarly, Park et al. [51] observed retarded degradations of 5-HMF at 35 • C at concentrations of 5 gCOD/L, with complete inhibition occurring at 10 gCOD/L. The retarded conversion was improved by increasing the cell biomass or the inoculum concentration in the bioreactors, but a complete hampering by the component at 10 gCOD/L persisted.
AD of furfural initiates with its reduction to furfuryl alcohol, which is documented to be less toxic in comparison to its parent material [11,52]. In the current trial, its rapid and robust degradation under the conditions provided, in conjunction with the overshoots, signified furfuryl alcohol as a docile component. Nonetheless, the percent deviation of SMYs from the theoretical limit decreased as their concentration per assay increased, which was calculated to be 70.6%, 30.7%, 19.0%, and 14.8% at 37 • C, and 20.4%, 6.6%, −6.1%, and 4.8% at 53 • C for 0.1, 0.2, 0.3, and 0.4 gCOD of added amounts of furfuryl alcohol per batch-test respectively.
During the anaerobic fermentation of furfural and furfuryl alcohol in a mesophilic CSTRs (Continuous Stirred Tank Reactors) system with furfural-acclimated inoculum, Rivard et al. [37] observed the bioconversion efficiency of furfural and furfuryl alcohol to be 78.5% and 93.5%, respectively, from their individually administered amounts of 0.735 g/d with 10 mL/d pretreated biomass liquor as base-feed. Such outcomes signify the anaerobic consortia's tolerance to the furanic compound metabolites. Table 5. Mean methane yields for the given concentrations of the furanic compounds under mesophilic and thermophilic conditions with kinetic parameters (a: max. methane produced, b: max. methane production rate, and c: lag phase) from the modified Gompertz model.  Figure 4 represents the biodegradation kinetics of the acidic compounds, which include furoic acid, levulinic acid, and glycolic acid, at test concentrations of 2, 4, 6, and 8 gCOD/L under mesophilic and thermophilic temperatures for 60 days.  The mean measured SMYs at 37 °C and 53 °C for all the acidic compounds corresponding to their per batch-test added amount of 0.1 gCOD, besides glycolic acid at 53 °C, surpassed the theoretical methane production limit, as was the case for 5-HMF and furfural at 37 °C, and furfuryl alcohol at 37 °C and 53 °C. Such a degradation course affirms the effective detoxification of the hazardous components at their reduced concentrations than the increased concentrations [47]. The mean measured SMYs at 37 • C and 53 • C for all the acidic compounds corresponding to their per batch-test added amount of 0.1 gCOD, besides glycolic acid at 53 • C, surpassed the theoretical methane production limit, as was the case for 5-HMF and furfural at 37 • C, and furfuryl alcohol at 37 • C and 53 • C. Such a degradation course affirms the effective detoxification of the hazardous components at their reduced concentrations than the increased concentrations [47].

Acidic Compounds
Similar to furfuryl alcohol digested at 37 • C and 53 • C, furoic acid treated under mesophilic conditions displayed an exorbitant mean percent degradation for all the test concentrations i.e. 151.5%, 119%, 114.3%, and 105.7% for 2, 4, 6, and 8 gCOD/L concentrations, respectively, whereas under thermophilic conditions, it was calculated to be 109.4%, 82.2%, 74.1%, and 77.5% for the abovementioned concentrations, respectively. As stated earlier, metabolites of furfural are known to be less refractory than its precursor, and for furoic acid, it was observed to be as such, except on a couple of occasions where the furfural's mean percent conversion to methane at 53 • C, for the concentrations at 2 gCOD/L and 4 gCOD/L were recorded to be 94.2% and 79.1%, respectively, compared with 82.2% and 74.1% for furoic acid. A likely cause for the low conversion of the furoic acid in the said batch assays could be the poor development and performance of the microbial aggregates responsible for converting the acid to its respective end intermediate acetate, or the final product methane.
According to Ran et al. [53], microbes display a greater affinity for furfuryl alcohol and HMF alcohol (2,5-Furandimethanol) than furoic acid and HMF acid (5-Hydroxymethyl-2-furoic acid). This was evident from the lower SMYs and longer lag phases of furoic acid (Table 6) in comparison to furfuryl alcohol at similar concentrations and temperatures (Table 5). Nonetheless, several bacterial species (e.g., Bacillus, Clostridium, and Desulfovibro genera) are identified to metabolize furoic acid and its precursors adequately [37,54]. Table 6 provides the mean methane yields of the acidic compounds for their test concentrations 2, 4, 6, and 8 gCOD/L at 37 • C and 53 • C, together with the MGM's estimated kinetic parameters, the coefficient of determination (R 2 ), and the mean absolute error (MAE). As estimated by the parameter 'b , the maximum daily methane yields in the thermophilic assays parallel to the mesophilic assays of levulinic acid were two to threefold higher, as specified in Table 6. Consequently, the degradation kinetics at 37 • C and 53 • C, for all the investigated concentrations of levulinic acid, appeared dissimilar from one another, as shown in Figure 4E-i-E-iv. Certain bacterial strains are known to utilize levulinic acid as a sole carbon source producing sugars, acetic acid, and propionic acid along their metabolic pathways [55,56]. Park et al. [57] employed these VFAs at 3 gCOD/L as feedstock separately with varying inoculum concentrations to investigate the inhibitory effect of levulinic acid at 35 • C and concluded the conversion of the target acid to be more sensitive in the presence of propionic acid than acetic acid. In the current scenario, the delayed conversion of levulinic acid can be attributed to immoderate propionate generation at 37 • C rather than 53 • C, which is known to have a slower degradation rate among short-chained VFAs [58].
Similar degradation patterns (inconsistent and delayed) for levulinic acid, in a continuous process at 43 • C, were also reported by Khan et al. [25], where the mean SMY for 0.5 gCOD of the component was recorded to be 274.67 mL/gCOD, accounting for 78.47% of methane conversion. Irrespective of the degradation routes, high removal percentages of levulinic acid (e.g., 89.2% and 96.3% for their highest loads at 37 • C and 53 • C, respectively) were also observed in the current trial. Table 6. Mean methane yields for the given concentrations of the acidic compounds under mesophilic and thermophilic conditions with kinetic parameters (a: max. methane produced, b: max. methane production rate and c: lag phase) from the modified Gompertz model. Glycolic acid, chemically categorized as alpha-hydroxy acid, displayed a diauxic criterion at 53 • C, revealing a two-phase decomposition of the component for each test concentration [36]; therefore, a TGM was fitted to such digestion curves. In contrast, reproducible degradation patterns at 37 • C for the test concentrations above 2 gCOD/L, similar to levulinic acid, could not be achieved, as seen in Figure 4F-ii-F-iv. Nevertheless, MGM was fitted to their degradation kinetics. Such disparities observed within the replicates of the kinetics for similar test concentrations under the present conditions can be attributed to the slow and stunted microbial development in the assays. Krümpel et al. [59] have described alpha-hydroxy acid to be a slow methane-yielding substrate due to its association with a low microbial growth rate. As for the gas production potential under the present conditions, the SMY for glycolic acid dropped by 40.32% and 18.27% at 37 • C and 53 • C for the per batch-test added amounts of 0.1 gCOD and 0.4 gCOD respectively.

Test
To the best of the authors' knowledge, little to no information on the BMP of furoic acid and glycolic acid is available; therefore, further statements concerning its anaerobic digestion at this stage cannot be made. Although, stoichiometrically furoic acid can be transformed into 45% methane and 55% carbon dioxide, whereas glycolic acid into 37.5% methane and 62.5% carbon dioxide [60].

Phenolic Compounds
The temporal progression of methane production from syringaldehyde, vanillin, and phenol at 37 • C and 53 • C for the given test concentrations are shown in Figure 5, whereas their respective methane yields, kinetic parameters, and the model's goodness of fit are summarized in Table 7.
On this occasion, overshoots for the lowest amount of phenolic compounds examined were observed for the batch assays treated at 37 • C only where the mean supplementary methane produced for syringaldehyde, vanillin, and phenol were 17%, 25.5%, and 23% of their theoretical thresholds, respectively. On the contrary, the mean percent conversion at 53 • C for 2 gCOD/L syringaldehyde and vanillin were evaluated to be 95.7% and 71%, respectively, whereas phenol completely inhibited the inoculum.
As expected, the mean measured SMY of the aromatic aldehydes dropped remarkably with the rise in test concentrations at both operating temperatures and was observed to be prominent under thermophilic conditions. No inhibition for syringaldehyde for all of its test concentrations at 37 • C and 53 • C, was detected. The mean measured SMYs for its highest loads were evaluated to be 293.1 ± 15.2 mL/gCOD and 202.9 ± 31.4 mL/gCOD, representing a decrease of 28.3% and 39.4%, compared with its initial concentrations examined at mesophilic and thermophilic states, respectively. Moreover, at loading rates of 6 gCOD/L and 8 gCOD/L, vanillin at the onset produced minute amounts of methane under thermophilic conditions but entered an inhibitory mode at the end of the trial phase. Furthermore, 0.2 gCOD and 0.3 gCOD per batch-test added amounts of vanillin under mesophilic conditions displayed a two-phase degradation pattern. The initial digestion phase can be presumed to be the acidogenesis phase, where VFAs were accumulated, and the latter methanogenesis phase, where the accumulated acids were transformed into methane [44]. Therefore, it can plausibly be assumed that the 53.1 ± 36.2 mL/gCOD methane produced from 8 gCOD/L of vanillin originated from the initial digestion phase of the said component under mesophilic conditions, which accounts for 15.2% of its conversion to methane. To our knowledge, such biodegradation kinetics for vanillin have not been previously reported.
Inhibitory traits of phenolic compounds are more pronounced with lower molar mass and vice versa [49]. This was apparent in the findings of Barakat et al. [61] when syringaldehyde and vanillin, at 2 gVS/L each, produced 78% and 17% methane, respectively, in regard to their theoretical limits. Furthermore, the said author co-digested the aromatic aldehydes separately with xylose on a 1:1 mass-based ratio and observed the methane production of syringaldehyde + xylose to be approximately 1.4 times higher than the xylosecontained reference assay (290 mL/gVS). Conversely, for vanillin + xylose, the conversion was 225 mL/gVS, implying the hampering of xylose degradation by vanillin. In another study, Zaldivar et al. [62] examined the impacts of selected aldehydes on the growth and fermentation of E.coli by evaluating their potencies at different concentration levels (g/L). With an incubation period of 48 h, the IC 25 , IC 50 , and IC 100 (concentrations inhibiting 25%, 50%, and 100% of the population) for syringaldehyde were estimated to be 0.6, 1.2, and 2.5 against 0.5, 0.9, and 1.5 for vanillin, respectively. Toxicities of such components, according to Palmqvist et al. [33], are not only associated with their molar masses but with their hydrophobicity too. By measuring the hydrophobicity (LogP octanol/water ) of syringaldehyde (0.99) and vanillin (1.21), the outcomes by Zaldivar et al. [62] were in agreement with the aforesaid statement.
As for phenol, it was found to be the most potent component in the current trial as it inhibited the whole process under the given digestion conditions, other than the one described above. Fang et al. [63] highlighted two degradation routes that phenol adapts (i.e., via benzoate to VFAs and caproate to acetate), where the initial pathway occurs at mesophilic and thermophilic temperatures, and the latter at thermophilic temperatures. Necessary enzymes, including benzoyl-CoA, involved in ring reduction and the cleavage of phenol [64] are inactivated by higher temperatures [65], which, in the current experiment may have caused the full inhibition of the process at 53 • C for all of the phenol concentrations tested, assuming the initial degradation course in the assays. The bioconversion of phenol is recommended to be performed at a moderate temperature, as most of the substances degrading consortia are mesophilic [66]. Under the current circumstances though, this was apparent only for phenol examined at a 2 gCOD/L loading rate. Perhaps a prolonged acclimatization phase was required for the microbial aggregates to convert the increased concentration of phenols at 37 • C, judging by their slight recovery at day 50, as seen in Figure 5I-ii. Nevertheless, phenol is considered to be one of the most notorious inhibitors [67], and this study has demonstrated this accordingly.
Degradation dynamics of the increased phenol concentrations in the anaerobic process are presented by Chapleur et al. [68]. With an HRT of 140 days and 2.7 g/L cellulose as a test substrate, phenol, with its most reduced concentrations (0.01 g/L-0.1 g/L) was completely degraded within the first 20 days of incubation in batch assays, whereas the mid-range concentrations (0.5 g/L and 1 g/L) required almost 60 days. High concentrations, such as 2 g/L and 4 g/L, hampered the process, whereas for 1.5 g/L phenol, a lag phase of 40 days and a removal percentage of 82% were observed. Adopting a similar methodology on a different substrate (mashed biowaste), Simon et al. [69] observed the general bioconversion pattern and the arrest of the anaerobic process in the presence of 0.10 g/L-5 g/L phenol for 200 days to be somewhat alike to Chapleur et al. [68] and evaluated the IC 50 of the toxin at the mesophilic condition to be 1.25 g/L.

Process-wastewater
Degradation kinetics, methane yields, and the kinetic parameters for the 5-HMF process-wastewater digested at 37 • C and 53 • C are provided in Figure 6, and Tables 8  and 9, respectively. longed acclimatization phase was required for the microbial aggregates to convert the increased concentration of phenols at 37 °C, judging by their slight recovery at day 50, as seen in Figure 5I-ii. Nevertheless, phenol is considered to be one of the most notorious inhibitors [67], and this study has demonstrated this accordingly. Degradation dynamics of the increased phenol concentrations in the anaerobic process are presented by Chapleur et al. [68]. With an HRT of 140 days and 2.7 g/L cellulose as a test substrate, phenol, with its most reduced concentrations (0.01 g/L-0.1 g/L) was completely degraded within the first 20 days of incubation in batch assays, whereas the mid-range concentrations (0.5 g/L and 1 g/L) required almost 60 days. High concentrations, such as 2 g/L and 4 g/L, hampered the process, whereas for 1.5 g/L phenol, a lag phase of 40 days and a removal percentage of 82% were observed. Adopting a similar methodology on a different substrate (mashed biowaste), Simon et al. [69] observed the general bioconversion pattern and the arrest of the anaerobic process in the presence of 0.10 g/L-5 g/L phenol for 200 days to be somewhat alike to Chapleur et al. [68] and evaluated the IC50 of the toxin at the mesophilic condition to be 1.25 g/L.

Process-wastewater
Degradation kinetics, methane yields, and the kinetic parameters for the 5-HMF process-wastewater digested at 37 °C and 53 °C are provided in Figure 6, and Table 8 and 9, respectively. The process-wastewater was observed to be the only test component that did not display a mean overshoot for its lowest load (2 gCOD/L or 28.5 gVS/L) under mesophilic conditions. Nevertheless, its average percent degradation to methane was evaluated to be 96.93%, next to 72.8% at 53 °C. Moreover, as anticipated, the conversion of the processwastewater to methane decreased with the increase in loading rate per HBT assay and was computed to be 71.1%, 58.9%, and 44.5% for 57.06, 85.6, and 114.12 gVS/L at 37 °C, respectively. As with furanic compounds, the thermophilic assays containing the process-  The process-wastewater was observed to be the only test component that did not display a mean overshoot for its lowest load (2 gCOD/L or 28.5 gVS/L) under mesophilic conditions. Nevertheless, its average percent degradation to methane was evaluated to be 96.93%, next to 72.8% at 53 • C. Moreover, as anticipated, the conversion of the process-wastewater to methane decreased with the increase in loading rate per HBT assay and was computed to be 71.1%, 58.9%, and 44.5% for 57.06, 85.6, and 114.12 gVS/L at 37 • C, respectively. As with furanic compounds, the thermophilic assays containing the process-wastewater underperformed relative to their counterparts for the SMYs at similar concentrations, apart from the HBT with the highest load. Ultimately, the measured SMYs for the thermophilic assays were shy of 27.2%, 38.7%, 58.1%, and 53.1% from their theoretical potential for the examined concentrations of 2, 4, 6, and 8 gCOD/L, respectively.
The present investigation into process-wastewater for its BMP concerned an amalgam of furans (5-HMF and furfural) and weak acids (formic acid, acetic acid, glycolic acid, lactic acid, and levulinic acid) and its limited biodegradation with increased concentrations may have originated from their synergistic effect (concerning inhibition) on the anaerobic process [11]. The potency of such aqueous materials on the anaerobic consortia elevates further when 5-HMF and furfural coexist in the fed substrate. This was evident in the findings of Taherzadeh et al. [70], wherein the sum of 5-HMF and the furfural concentration exceeding 2 g/L strongly reduced the fermentation rates. Furthermore, carboxylic acids (undissociated form) are known to function interactively with other inhibitors in suppressing the fermentation process [31,70]. Similarly, alpha-hydroxy acids are being described as slow methane-yielding components [59], which, in such a case, might be coupled to their lag phase, as prolonged lag phases, according to Park et al. [51], can eventually lead to the significant deterioration of the digestion process.
With 14.8 g/L, 5-HMF was identified to be the main component of the processwastewater, constituting 30.1% of its DOC (28.17 g/L), whereas the other components occupied 31% combined. As outlined by Ghasimi et al. [40], the conversion of 5-HMF proceeds after the complete degradation of furfural only when the latter is present at a lower concentration. Therefore, it can be postulated for the decomposition of the process-wastewater in the current setup to be initiated with the conversion of its minor-concentration constituents, followed by the major-concentration constituents. Thus, the reduced biodegradability of the process-wastewater, with extended lag phases at high loads under mesophilic and thermophilic conditions, could be at the behest of the increased concentration of 5-HMF in the assays.
5-HMF-containing aqueous material generated during the HTC of chicory roots and maize silage digestate were investigated for their BMP via HBT by Stökle et al. [12] and Cao et al. [43], respectively. With approximately 3 gVS per assay of the respective substrates, a SMY of 244 mL/gCOD, for either feedstock, at 37 • C was observed. For a similar concentration, a mean methane yield of 249 mL/gCOD for 5-HMF process-wastewater at 37 • C was recorded. 5-HMF though comprised a minor fraction (0.6 g/L-0.7 g/L) of the aqueous material of the abovementioned authors.
Based on the substantial proportion of 5-HMF in the process-wastewater and the degradation kinetics observed ( Figure 3A-i-A-iv), it is plausible to assume that the overall inhibitory characteristics of the said process-wastewater to be governed by the 5-HMF.
It is worth mentioning that the 5-HMF process-wastewater acquired from the biorefinery specializes in the synthesis of 5-HMF and its subsequent refinement. The excessive concentration of 5-HMF in the wastewater are retentates, residuals, or rejects generated during the filtration processes of the crude 5-HMF.

Conclusions
Batch anaerobic digestion of the typical constituents (furans, weak acids, and phenols) of a biorefinery's process-wastewater at their lowest concentrations displayed a stimulatory effect on methane production under mesophilic conditions. Increasing the test components' concentrations in the assays resulted in diminishing methane conversion at both operating temperatures of 37 • C and 53 • C. Moreover, the mesophilic assays containing the test components were not on par with their counterparts (thermophilic assays) for the mean measured SMYs at similar concentrations. Within the furanic compounds, 5-HMF appeared to be the most refractory component. Similarly, the degradation patterns of weak acids such as levulinic acid and glycolic acid were characterized by disparities and inconsistencies existing within the replicates for similar concentrations, which was profoundly true at 37 • C. Furthermore, two of the test components, glycolic acid and vanillin, exhibited a diauxic criterion elucidating the complexities in their anaerobic decomposition. Ultimately, test components that impaired the anaerobic process under mesophilic and thermophilic conditions with their increased concentrations were identified to be the two phenolic compounds, phenol and vanillin, and one furan, 5-HMF.
In the current study, the poor performance of the 5-HMF process-wastewater as a substrate is considered to be associated with the presence of the exceptional concentration of 5-HMF in the said process-wastewater, which is assumed to be dictating its overall characteristics. This feature might exclusively grade the 5-HMF process-wastewater as being a risky feedstock unless proper remedial measures are applied, such as increasing the microbial cell biomass concentration, bioaugmentation, or photodissociation, among others.