Biological Hydrogen Production Through Dark Fermentation with High-Solids Content: An Alternative to Enhance Organic Residues Degradation in Co-Digestion with Sewage Sludge

Abstract

Adequate treatment of the organic fraction of municipal solid waste (OFMSW) in co-digestion with sewage sludge (SS) through dark fermentation (DF) technologies has been widely studied and recognized. However, there is little experience with a high-solids approach, where practical and scalable conditions are established to lay the groundwork for further development of feasible industrial-scale projects. In this study, the biochemical hydrogen potential of OFMSW using a 7 L batch reactor at mesophilic conditions was evaluated. Parameters such as pH, redox potential, temperature, alkalinity, total solids, and substrate/inoculum ratio were adjusted and monitored. Biogas composition was analyzed by gas chromatography. The microbial characterization of SS and post-reaction percolate liquids was determined through metagenomics analyses. Results show a biohydrogen yield of 38.4 NmLH₂/gVS OFMSW, which forms ~60% of the produced biogas. Aeration was proven to be an efficient inoculum pretreatment method, mainly to decrease the levels of methanogenic archaea and metabolic competition, and at the same time maintain the required total solid (TS) contents for high-solids conditions. The microbial community analysis reveals that biohydrogen production was carried out by specific anaerobic and aerobic bacteria, predominantly dominated by the phylum Firmicutes, including the genus Bacillus (44.63% of the total microbial community), Clostridium, Romboutsia, and the phylum Proteobacteria, with the genus Proteus.

1. Introduction

Traditional fossil fuel energy sources have been causing significant environmental impacts and depleting finite fuel sources for over a century [1]. Simultaneously, more than 930 million tons of food and organic waste are generated worldwide in cities [2]. These waste materials are primarily deposited in dumpsites and traditional landfills, leading to problems such as odors, fires, groundwater contamination, and the emission of greenhouse gases (GHG) [3]. In light of these challenges, there is a pressing global demand for developing and disseminating new and existing technologies. For example, the biological transformation of the organic fraction of municipal solid waste (OFMSW) through dark fermentation (DF) technologies emerges as a promising solution offering not only adequate waste stabilization and treatment but also the potential for fertilizers and renewable hydrogen (H₂) generation [4,5].

Hydrogen has a heating value of 142 MJ/kg, which is over 2.5 times higher than that of many hydrocarbon fuels, such as methane, natural gas, or gasoline. Currently, its primary production method is the steam reforming of fossil fuels, which is costly and environmentally harmful; alternative H₂ production methods are being explored worldwide [1]. Lately, interest in low-emission H₂ production has increased since conventional H₂ production methods (e.g., water electrolysis or chemical cracking of hydrocarbons) require electricity, which is mainly derived from fossil fuels [6]. Low-emission H₂ production is yet to develop as an industry; if the announced projects are realized, annual production could reach 38 Mt in 2030, and there is still a long road to meet the steadily growing need for H₂ and climate ambitions. Moreover, the estimated demand for H₂ in 2050 could be up to 600 megatons (Mt) [7]. In this context, OFMSW has the potential to produce competitive amounts of H₂ (biohydrogen) at a low-emission or even negative-emission intensity [8] via highly efficient biological technologies.

Biohydrogen is a carbon-impartial energy carrier known for its potential to replace some fossil fuels. It can be generated by photofermentation and dark fermentation. DF is considered a profitable approach to treating OFMSW, together with its conversion into other valuable metabolites of commercial interest, such as organic acids, solvents, or enzymes. Hydrogen production through DF is a biological process where microorganisms break down organic matter in the absence of oxygen to produce hydrogen gas (H₂). This process is carried out by diverse anaerobic bacteria that do not need light. They ferment organic compounds (typically carbohydrates like sugars, starches, and even complex organic waste) through metabolic pathways. Thus, DF was preferred over photofermentation for this study due to the following reasons: (1) No light requirement, considering also that OFMSW is typically dark and opaque itself. (2) Broader substrate versatility, where DF can utilize a wider variety of complex organic substrates, and photofermentation is generally limited [9]. (3) Photofermentation usually requires conditions that add significant capital and operational costs, and are often complex to implement for large-scale waste treatment [10].

The high carbohydrate composition of the OFMSW can promote biohydrogen production. However, its protein and lipid contents present a challenge since they influence the efficiency of the process. Even when their contents contribute to increasing the diversity of the organic compounds, as well as hydrogen production [11], as pointed out by Capson-Tojo et al., 2018 [12], high protein proportions may cause nutrient imbalance due to a low C/N ratio. Additionally, it has already been proven that co-digestion of food waste and sewage sludge is suitable as substrate mixtures for H₂ production [6], making them two possible optimal co-sources, which are explored in this research. According to Kothari et al., 2014 [13], digested sludge is the most suitable inoculum source for anaerobic fermentation to treat organic waste with a high concentration of solid substrate. Other key parameters that influence hydrogen production through fermentation include substrate concentration, pretreatment pathways, pH, temperature, inoculum characteristics, hydraulic retention time, the presence of inhibitors, and reactor configuration.

Several studies have investigated various conditions for biological hydrogen production from OFMSW. Nevertheless, pragmatic and scalable pilot tests with high-solids conditions and low-energy inputs have been little explored. In this regard, DF biohydrogen production requires more advanced studies on a medium and large scale to overcome technological and economic barriers to become a competitive technology [14]. Industrial-scale projects have also not been developed because there are diverse challenges to scaling up these processes, and their efficient performance is unclear [1].

The potential of biological hydrogen production from OFMSW has been evaluated in diverse studies, mainly using traditional mesophilic continuous stirred tank reactors; however, there has been a growing interest in exploring high-solids dark fermentation (HSDF) for the treatment of these wastes, studied, for example, by Elsamadony and Tawfik, 2015 [15], who proved the efficient digestion of OFMSW and hydrogen production from OFMSW in high solids, and its dependence on the HRT and OLR. Or Capson-Tojo et al., 2018 [12], who suggested that food waste with high total-solids (TS) contents allows for the operation under high-solids conditions, requiring smaller reactor volumes and producing less digestate than traditional dark fermentation, among other advantages. Also, Yeshanew et al., 2018 [16], efficiently co-produced hydrogen and methane from OFMSW using a two-stage AD system in a pilot-scale batch dark fermenter. Thus, the cited studies highlight the viability of batch HSDF to treat OFMSW. However, there has been little attention paid to the actual feasibility and microbial interactions involved in this type of anaerobic reaction.

Additionally, this research advances biological hydrogen production knowledge because—even when multiple studies have been published on biohydrogen production through dark fermentation—it replicates a real-life scenario in a large-scale digester by treating the OFMSW under a high concentration of solid substrates, requiring no extra water addition, minimal substrate and inoculum pretreatment, or avoiding the incorporation of additives. Also, unlike other studies, this research focuses on biohydrogen production potential and microbial community analyses in co-digestion with sewage sludge. Thus, the significance of this study lies in its ability to contribute to the technical and realistic implementation of such projects, which is frequently relegated to a secondary position, and also in simplifying the process and replicating a real-life scenario without increasing operational costs. Therefore, the goals of this work are (1) to assess the potential for hydrogen production when OFMSW is co-digested with sewage sludge under high-solids conditions, and (2) to identify the essential microorganisms involved in biohydrogen production with aeration pretreatment and determine the functional activity and symbiotic or competing relationships between them; thus, promoting beneficial biological interactions that optimize biohydrogen production in HSDF systems.

Therefore, this research aims to test the hypothesis that a pilot standard reactor with high-solids conditions (in this study, TS = 24.5%) and minimal pretreatment is sufficient for producing valuable biohydrogen, considering the indigenous microbiota in the OFMSW and the mixed microbial consortium in the sewage sludge. Testing this hypothesis can potentially contribute to unlocking a more efficient and sustainable way of producing biohydrogen from OFMSW. Overall, this research proposes an alternative solution for biohydrogen production that could have appropriate implications for large-scale bioreactors and contribute to optimizing the treatment of OFMSW.

2. Materials and Methods

2.1. Substrate and Inoculum

2.1.1. Organic Fraction Municipal Solid Waste (OFMSW)

The OFMSW samples were collected directly at the landfill site in Saltillo, Coahuila, Mexico, using the manual sampling method [17]. The samples mainly comprised peels and post-consumption residues of fruits and vegetables, derivates from corn and wheat, and other food residues generated by households and small businesses. The residues were transported to the laboratory and stored at −20 °C until utilization.

Before the experimental analyses, as the only pretreatment, the substrate was trimmed to reach a particle size +/−5 cm, which promotes a larger contact area (microorganisms/feedstock). This particle size was selected to imitate, as much as possible, conditions in a large-scale case where this is a viable and realistic pretreatment. Prior to this, the physicochemical characteristics of the OFMSW were determined following local and US National Renewable Energy Laboratory (NREL) norms, as described in detail in Silva-Martínez et al., 2024 [18]. The corresponding parameters are presented in

Table 1. Physicochemical characteristics of the OFMSW.

Parameter	Value	Parameter	Value
pH	5.33	N (%TS)	2.74
Humidity (%)	74.99	S (%TS)	0.097
TS (%)	25.01	P (%TS)	0.3872
VS (%)	22.09	C/N Ratio	16.45
VS (%TS)	88.17	Fats (%)	12.27
Ash (%TS)	11.83	Proteins (%)	16.74
C (%TS)	44.72	Carbohydrates (%)	70.98
H (%TS)	6.03	Raw fiber (%)	43.93
O (%TS)	32.55

.

2.1.2. Inoculum

The batch reaction was initiated with fresh inoculum from the activated effluent of a municipal wastewater treatment plant. From this, the inoculum was prepared for use in the experiment as a liquid sewage sludge, which was also used as the recirculated percolate (with a 98% water content). Additionally, the co-substrate, solid sewage sludge (with a water content of 77%), served as the inoculum, which was mixed with the substrate at the beginning of the experiment.

To achieve these conditions, the water content of the solid sludge was lowered from 93.74 to 77%, by bypassing the sludge through a centrifuge Rotina 380 (Hettich, Tuttlingen, Germany) at 4000 rpm for 15 min. The remains were aerated for one day with the two purposes as follows: to permeate the remaining liquid through filter paper and reach high-solids conditions, and as a pretreatment to inhibit some metabolic competitors and H₂ consuming microbes. The high-solids sludge was then stored in airtight bottles to maintain as close as possible to the original mesophilic temperatures of 35–38 °C and incubated for seven days before utilization. This was to degas the inoculum and minimize the effect of any possible methane production.

Table 2. Physicochemical characteristics of the sewage sludge of the wastewater treatment plant.

Parameter	Value
Floating material	Absence
pH	6.98 ± 0.18
Turbidity (NTU)	88.96 ± 41.41
Fats, oils, and grease (mg/L)	12.05 ± 10.95
Total Suspended Solids (mg/L)	91.10 ± 50.36
Settleable solids (mg/L)	0.54 ± 0.54
Biological Oxygen Demand (mg/L)	196.50 ± 57.80
Chemical Oxygen Demand (mg/L)	497.60 ± 43.44
Total Coliforms (×105) (CFU/100 mL)	110.60 ± 80.63
Fecal Coliforms (×105) (CFU/100 mL)	60.42 ± 63.94
Total Nitrogen (mg/L)	229.74 ± 167.16
Total Phosphorus (mg/L)	10.50 ± 2.41
E. coli (×105) (CFU/100 mL)	60.42 ± 63.94
Residual chlorine (mg/L)	CELL
Color (Pt-Co.)	572.20 ± 309.13

displays the initial physicochemical characteristics of the used activated effluent.

2.2. Experimental Scheme and Operational Procedure

2.2.1. Biochemical Hydrogen Potential and Reactor Configuration

The biochemical hydrogen potential (BHP) method was implemented using a jacketed acrylic batch reactor with a capacity of 7 L under mesophilic conditions over 39 consecutive days (Figure 1). Concurrently, three control experiments (consisting solely of inoculum) were conducted to account for endogenous respiration and biogas production solely from bacterial activity. These were thoroughly mixed after proportionally adjusting the inoculum and substrates (Section 2.2.2). Subsequently, the reactor was purged with nitrogen for 5 min to eliminate any remaining oxygen.

Percolate recirculation was employed to maintain uniform microbial conditions throughout the experiment. This ensured even distribution and facilitated optimal contact between the microorganisms and substrates. Mixing was intentionally omitted to allow for a complete batch reaction. The system also featured a peristaltic pump (Masterflex, 77910-55 L/S, Barrington, IL, USA) that recirculated leachates, distributing and percolating them into the digestate via a distribution device. This pump was operated daily for ~4 h. A water bath (LabTech, LCB-11D, Sorisole, Italy) was also used to recirculate hot water, maintaining a temperature of 35 °C ± 2 °C inside the reactor. This temperature, as shown in

Table 3. Environmental requirements for optimal residue degradation.

Parameter	Hydrolysis/Acidogenesis	Methane Formation
Temperature	25–35 °C	Mesophilic: 32–42 °C/Thermophilic: 50–58 °C
pH value	5.2–6.3	6.5–7.5
C:N ratio	10–45	20–30
DM content	<40% DM	<30% DM
Redox potential	+400 to −300 mV	<−250 mV
Required C:N:P:S ratio	500:15:5:3	600:15:5:3
Trace elements	No special requirements	Essential: Ni, Co, Mo, Se

, lies within the ideal temperature range for hydrolysis and acidogenesis reactions, which promote hydrogen generation. pH and oxidation-reduction potential (ORP) were continuously monitored using two electrode sensors (Vernier, Beaverton, OR, USA). An interphase data-logging system (Vernier, LabQuest 3, Beaverton, OR, USA) was used to monitor the system throughout the fermentation.

2.2.2. Operational Conditions

Many parameters influence the H₂ potential during fermentation, and some of them were controlled as much as possible to maintain optimum conditions. Drawing from Deublein and Steinhauser, 2008 [19], these parameters are shown in Table 3.

For the pH adjustment, NaOH was supplied to maintain a pH within that range throughout the reaction. Alkalinity was determined according to standard methods APHA 2017 [20], resulting in an average 528 mg/LCaCO₃ value. This value indicates the carbonate/bicarbonate and non-protonated forms of volatile fatty acid (VFA) contents.

As for TS concentration, according to Raposo et al., 2012 [21], low or high loads of solid substrate introduced into the reactors could limit or inhibit biogas production; thus, highlighting the importance of proper TS contents. For this test, considering the initial intention to maintain high-solids contents, the TS was adjusted to 24.5%, from which 73.8% comprises volatile solids (VS). Total and volatile solids were calculated according to the regulations (NMX-AA-016-1984, 1984) [22] and the technical report (NREL/TP-510-42621, 2008) [23] and (NREL/TP-510-42622, 2005) [24] of the US National Renewable Energy Laboratory (NREL).

Additionally, establishing and maintaining an appropriate C/N ratio is a key factor in successful digestion. Codigestion was also implemented as a method widely used to primarily reach and maintain C/N ratios between the optimal 10–45 ranges mentioned by Deublein and Steinhauser, 2008 [19].

The substrate/inoculum ratio is essential in BHP tests. The substrate and inoculum fractions depend on the inoculum’s activity, biomass concentration, and biodegradability. An S/I ratio of 1.5 was considered optimal from an operational perspective, and at the same time, it allowed for the maintenance of the aforementioned TS contents of 24.5%, which also prevented the hoses from clogging. Even when lower S/I ratios have been reported as optimal for biohydrogen production, a value of 1.5 was used to represent pilot conditions, which may be replicable in larger industrial-scale reactors. Further studies to investigate and compare diverse S/I ratios may contribute to enhancing the system’s performance and increasing hydrogen production in pilot-scale projects.

2.3. Biohydrogen Production and Composition

Biogas was measured using a continuous gas flow meter (Anaero Technology, Nautilus, Cambridge, UK) with liquid volume displacement in a “tumbler bucket,” which represented a volume of 7.33 mL per turn. However, accurate biogas rates were registered considering standard temperature and pressure (STP) conditions; the flow rate was calculated in milliliters per gram of the substrate’s volatile solids (mL/gVS). To register the biogas quantities, the information technology (IT) system consisted of an Arduino 2560 Mega microcontroller (Ivrea, Italy), which acted as the central controller for the data logger. The biogas production data is presented in Annex S1.

The biogas was collected in Tedlar gas bags attached to the outlet of the reactor and then analyzed to find the diverse gas contents by gas chromatography (GC). For that, biogas samples (100 mL) were collected with a pressure lock syringe and then injected into the Clarus 580 gas chromatograph (Perkin Elmer, Inc., Waltham, MA, USA) to calculate the percentage composition of H₂, CH₄, CO₂, and other gases in each sample. A HayeSep D column (3 m × 3.2 mm, 100/120 mesh) was utilized. The injection port, oven, and detector had temperatures of 75, 30, and 120 °C, respectively. Nitrogen served as a carrier gas flowing at 30 mL/min. The purity or exact composition of H₂ in the biogas was calculated with a previously designed calibration curve (Annex S2).

At the same time, three blank assays (1 L reactors) with approximately 650 mL of the inoculum alone were set to batch conditions in a water bath Standard BMP/RBP (Gant Instruments, SG8 6GB, Cambridgeshire, UK), set at 35 °C, and with stirring for one month to determine the biogas production of the inoculum alone.

The biogas production was also measured by the Arduino continuous gas flow meter. This was carried out to consider endogenous respiration and then subtract the biogas production of the inoculum. The total biogas and hydrogen produced during the fermentation (Hmax) and the maximum biogas and hydrogen production rate (Rmax) were recorded. The assay’s key findings center on biogas and, specifically, hydrogen production, measured against a specific weight of volatile solids. This approach provides a crucial understanding of the substrate’s potential for biohydrogen production.

Furthermore, the modified Gompertz equation was implemented to describe and clarify the biohydrogen production rates and potential, as well as the start-up delay period. The cumulative hydrogen production data were used to fit the modified Gompertz equation, thereby determining the potential and maximum production rates under the given conditions.

$$H(t)=P x \mathrm{e}\mathrm{x}\mathrm{p}\left\{- \mathrm{e}\mathrm{x}\mathrm{p}\left[{\displaystyle \frac{\mathrm{R}\mathrm{m} \mathrm{x} \mathrm{e}}{\mathrm{P}}}\left(\mathrm{L}-\mathrm{t}\right)+1\right]\right\}$$

where H(t) is cumulative hydrogen production (mL); P is hydrogen production potential (mL/gVS); Rm is maximum hydrogen production rate (mL/gVS*d); e equals 2.71828; L is the lag-phase time (d); and t is time (d).

2.4. Diversity Community Analysis

Community analysis was conducted by 16S rRNA gene sequencing for the sewage sludge (inoculum) and the resulting liquid digestate (reactor sample). Biomass samples were collected from the initial inoculum and the reactor after 39 days of operation. Metagenomic DNA was extracted using the FastDNA SPIN Kit for Soil (MP Biomedicals, Irvine, CA, USA) with 5 min of cellular lysis. Gel electrophoresis and UV-VIS spectrophotometry (NanoDrop-2000, Thermo Scientific, Waltham, MA, USA) assessed the integrity and concentration of DNA fragments. Afterward, DNA purification was performed using the Clean & Concentrator kit (Zymogen, Irvine, CA, USA). The samples were then processed for Illumina MiSeq sequencing (Macrogen, Inc., Seoul, Republic of Korea). The bacterial 16S rRNA gene was amplified using primers F341 (5′-CCTACGGGGGNGGGGCWGCAG-3′) and R805 (5′-GACTACHVGGGGGTATCTAATCC-3′), targeting the V3-V4 regions. The sequencing analysis generated more than 150,000 reads in each of the evaluated samples. As a limitation of the study, the microflora derived during the hydrogen-producing phase was not tested.

The bioinformatics protocol was as follows: Raw reads were quality trimmed, filtered with Trimmomatic, and assembled using FLASH [25,26]. The VSEARCH tool was used for sequence clustering, full-length dereplication, and chimera detection [27]. The high-quality sequences were also clustered into operational taxonomic units (OTUs) with a 97% similarity threshold. Taxonomic assignment of OTUs was carried out using QIIME1 (http://www.qiime.org/) and the SILVA database version 138.2 [28]. To explore associations between dominant microbial genera and key physicochemical variables, Pearson correlation analysis was performed. Genera with relative abundances >2% in at least one condition were correlated against pH, oxidation-reduction potential (ORP), and total solids (TS). Correlation coefficients were visualized in a heatmap where significant relationships (p < 0.05) were marked with a star symbol. Statistical computations were conducted in MATLAB R2020b using the Fathom Toolbox (https://www.usf.edu/marine-science/research/matlab-resources/index.aspx/ accessed on 2 October 2023) for MATLAB R2020b (MathWorks, Inc., Natick, MA, USA). This exploratory analysis aimed to identify potential microbial-environmental interactions relevant to hydrogen production under high-solids fermentation conditions.

To assess the distribution and uniformity of bacterial communities in each biomass sample, alpha-diversity metrics, including richness, evenness, and Simpson diversity index, were evaluated using a nonparametric approach [29].

3. Results and Discussion

3.1. Biogas Production

Table 4. Total and net biogas production.

		Biogas	Biohydrogen	Units
HMAX	Maximal potential production	49.23	17.1	L
	Generated by activated sludge	2.84	-
	Total generated by OFMSW	46.39	17.1
RMAX	Production rate	110.62	38.4	(NmL/gVS OFMSW)
	Production rate by activated sludge	9.56	-
	Production rate by OFMSW	101.06	38.4

and Figure 2a show that the biogas yield during the 39-day experiment reached a total of 110.62 NmL/gVS OFMSW, with the codigestion of the OFMSW and sewage sludge. It is estimated that this biogas production was adequate but did not increase, possibly due to the high total fiber content of the substrate (43.9%). Figure 2a shows that most of the biogas was generated between days five and twelve; the major production was observed on day seven. According to Dong et al., 2009 [9], and Alibardi and Cossu, 2015 [3], the rich content of fiber and lignin is characterized by low biodegradability rates due to their recalcitrance and resistance to bacterial hydrolysis, unless suitable pretreatment methods are adopted. Thus, further studies should analyze and monitor the gradual degradation process of cellulose, hemicellulose, and lignin to gain a deeper understanding of their enzymatic hydrolysis and diverse possible pretreatment pathways, thereby optimizing the system to consider a viable and economical route.

On the other hand, the high variability of vegetables and fruits, whole and peels, has resulted in diverse biohydrogen production rates in various studies [3]. After fermentation, the volatile solids content was 3.08%, which shows a significant decrease from the initial 22.09%. This yields a proportion of 44.97 NmL of hydrogen production per gram of vs. OFMSW consumed.

Accordingly, it is assumed that other factors influenced the biogas production rates as follows: (1) The HSDF process did not contribute to larger biogas production quantities due to the slow movement of microorganisms, considering that low water contents characterize this type of digestion. (2) Larger amounts of substrate (OFMSW) than inoculum (1.5:1 measured in grams VS). (3) Lower biogas production quantities than similar experiments with smaller particle sizes. In that sense, authors have suggested particle sizes equal to or smaller than 2 mm to increase the degradation efficiency [30]. Also, Yong et al., 2015 [31], compared diverse particle sizes for lignocellulosic residues and determined optimal biodegradation rates when the particle size was less than 1 mm. Elsamadony, 2015 [15], determined that a size diameter of 0.3 mm achieved the maximum H₂ yield; nevertheless, within this research, it was decided to keep a particle size of +/−5 cm, intending to consider more realistic conditions to imitate larger reactors, regardless of the possible best particle size for optimal biogas production. André et al., 2018 [32], present a list of diverse medium and large-scale high-solids anaerobic digesters demonstrating efficient biodegradation of OFMSW and other organic wastes, maintaining their original particle size.

Figure 3 illustrates that biohydrogen was prominent during the initial 26 days of the experiment. Although small amounts of carbon dioxide and methane were detected during this period, the methane content percentage progressively increased, eventually constituting up to 63.91% of the biogas produced on day 36. Nevertheless, the quantities produced on that day (0.205 mL of CH₄/gVS OFMSW) were as low as and very similar to the amounts at the initial phase (0.210 mL of CH₄/gVS OFMSW) on day 8, and minimal compared to the H₂ production (17.26 mL of H₂/gVS OFMSW) on the same day. This suggests that the small microbial populations of methanogenic archaea were maintained and did not further reproduce or augment, possibly due to the inhibition caused by the high VFA contents during DF reactions, as suggested by Sanghvi, 2024 [33]; however, VFS concentration was not detected in this experiment. This inhibition is presumed to have occurred because the substrate quantities significantly outweighed the inoculum (at a ratio of 1.5:1), but primarily due to the low methanogenic archaea populations. Furthermore, methane production did not increase, considering the low protein and lipid contents. These findings align with the study by Alibardi and Cossu, 2015 [3], which highlights that carbohydrate-rich substrates enhance biohydrogen production, while protein- and lipid-rich substrates exhibit high methane production potentials.

3.2. Biohydrogen Production

The biogas exhibited an average H₂ total percent composition of 60%, demonstrating that H₂ is the most prevalent gas, resulting in a recorded production rate of ~38.4 NmLH₂/gVS OFMSW (Figure 2b). The inoculum alone demonstrated negligible biohydrogen production, thus attributing the total output solely to its combination with OFMSW. These findings are consistent with similar studies on batch AD of OFMSW, such as those by Alibardi and Cossu, 2015 [3], who observed biohydrogen production rates ranging from 26 to 86 NmLH₂/gVS or Villanueva-Galindo and I. Moreno-Andrade, 2021 [34], who tested various bioaugmentation conditions, resulting in biohydrogen production rates from 16.3 to 84.5 NmLH₂/gVS, and Favaro et al., 2013 [35], studied the effects of inoculum and indigenous microflora on hydrogen production, resulting in a production range of 16.1 to 70.1 NmLH₂/gVS. Another recent study presenting relevant production rates is that carried out by Yeshanew et al., 2018 [16], who generated a biohydrogen yield of 41.7 NmL H₂/g VS under conditions very similar to those presented in this study.

According to the results of similar studies and considering the production rates found through the present research, it can be inferred that biohydrogen production under austere and commercial-scale viable conditions could be enhanced, but to find this, deeper experimentation is required.

Diverse research highlights the importance of substrate composition variability in determining biohydrogen production rates. OFMSW comprises carbohydrate-rich materials with high H₂ potentials, alongside low protein and lipid-rich fractions with lower H₂ potentials [3]. However, it is worth noting that biohydrogen production rates may not only be influenced by substrate composition but also by the degradation rate [3]. As highlighted by Favaro et al., 2013 [35], the indigenous microbial composition is also crucial in determining biodegradation rates. These considerable variabilities in biohydrogen production could represent significant challenges for full-scale treatment plants.

Moreover, hydrogen production was boosted following the inoculum pretreatment, which decreased levels of methanogenic archaea and metabolic competition, as further described in Section 3.3. The pretreatment, which involved aeration, served two the primary purposes as follows: firstly, to achieve optimal moisture content for a high-solid anaerobic process, and secondly, to inhibit microorganisms, particularly methanogens, that compete with H₂-producing organisms and other H₂ consumers. Thus, this is expected to provide optimal conditions for enhancing hydrogen production, which can be replicated in large-scale facilities without a significant increase in costs or labor. This pretreatment method proved to be effective in preparing the inoculum for optimum biohydrogen production rates, a fact also supported by Yang et al., 2019 [36], who demonstrated aeration as one of the most efficient pretreatment methods for enriching H₂-producing bacteria. Furthermore, various studies, such as [37,38,39,40,41] have confirmed the efficient co-digestion of OFMSW with activated sludge. This practice significantly enhances H₂ production compared to treating OFMSW alone. After day 26, hydrogen production rates significantly decreased due to the depletion of simple substrates (Figure 2b).

Another critical factor influencing biohydrogen production rates was pH. As observed in Figure 4, the test was carried out with a pH range of 4.5 to 8.5 throughout the reaction. It was successfully adjusted and stabilized within the optimal range of 5.2 to 7 for 79% of the time, recognizing that most H₂-producing organisms perform optimally within these pH conditions. A basic diluted solution (NaOH 10N) was constantly added to optimize pH conditions to elevate and stabilize pH levels, which continually tended to drop due to the formation of VFAs. Approximately 30 mL of sodium bicarbonate (NaHCO₃ O.5 N) was also added to increase the buffering capacity. This adjustment was crucial, as acidification could significantly impair H2 production, as demonstrated by [15], who experienced a drastic drop in pH from 6.88 to 3.53 in their reactor, leading to a decline in H₂ production. Furthermore, co-digestion with sewage sludge enhances the system’s buffering capacity, thereby reducing the need for costly pH control measures. Sewage sludge tends to be more alkaline, as documented by Zhu et al., 2008 [37], which further supports its role in stabilizing pH levels during biohydrogen production.

It is presumed that another significant factor influencing the effective biohydrogen production rates was the carbohydrate content of OFMSW, amounting to 71% (Table 1). Research conducted by Alibardi & Cossu, 2015 [3], underscores the dependency of biohydrogen production rates on carbohydrate contents, highlighting higher production rates in substrates rich in readily biodegradable carbohydrates, primarily starch, compared to those containing cellulose, hemicellulose, or simple sugars. Consequently, it can be inferred that H₂ production rates were constrained by OFMSW’s components, such as peelings, roots, eggshells, and others with high fibrous content. The OFMSW studied contains ~43.9% of raw fiber (Table 1). Compounds rich in cellulose and lignin polymers exhibit low biodegradability rates due to their recalcitrance and resistance to bacterial hydrolysis [9]. Moreover, while proteins and lipids are typically associated with higher methane potential production [3], at the same time, they also contribute significantly to enhancing the synergistic effect, thereby increasing H₂ production [42]. Tarazona et al.’s, 2022 [42], findings suggest an optimal increase in H₂ yields with carbohydrate contents of 56%, proteins at 22%, and lipids at 22%. Accordingly, the studied OFMSW is not far from these optimal composition rates, accounting for ~71% carbohydrates, ~17% proteins, and ~12% lipids, which indicates its potential for adequate biohydrogen production. Thus, the carbohydrate composition and the OFMSW characteristics, such as humidity, pH, and volatile solids contents (shown in Table 1) present the optimal conditions for biohydrogen production through dark fermentation processes.

Figure 5 shows the resulting modified Gompertz dynamic fitting using the daily cumulative hydrogen production throughout the experiment. These dynamics help clarify the 5-day lag period before biohydrogen production, which is attributed to the fact that, since no pretreatment was applied to the substrate, the particle size exceeded 5 cm, resulting in a limited surface area at the start, considering that a larger surface area allows for more efficient and faster degradation of microorganisms.

3.3. Microbial Community Analysis

Figure 6 presents the relative abundance of microbial communities at the phylum level in both the inoculum (sewage sludge) and the liquid digestate. In the initial inoculum, microbial diversity was high, with Firmicutes representing the dominant phylum (83.75%), followed by Bacteroidota (11.94%) and Proteobacteria (3.65%). Other phyla were present in much smaller proportions, including Acidobacteriota, Cloacimonadota, Armatimonadota, Campilobacterota, Dependentiae, Patescibacteria, and Planctomycetota, together comprising 0.66%.

Alpha diversity analysis supports this heterogeneity, with 650 species detected, an evenness index of 0.67, and a low Simpson index (0.03), indicating a relatively balanced distribution of species. In contrast, the liquid digestate community was overwhelmingly dominated by Firmicutes (99.53%), indicating an intense selection pressure that significantly reduced taxonomic diversity. Minor phyla, such as Desulfobacterota, remained present in trace amounts, as shown in the expanded stacked bar plot. The drastic reduction in phylogenetic diversity suggests a shift toward a highly specialized community structure dominated by fermentative bacteria. These microbial shifts imply a functional specialization driven by process conditions, with a loss of taxa involved in complementary roles such as methanogenesis, nitrogen cycling, or phosphorus removal—functions attributed to the diverse phyla initially present.

As a result, an analysis was performed on the dominant genera identified in the inoculum (sewage sludge) and the liquid digestate, focusing on those with a relative abundance greater than 1%.

Table 5. Functional classification of dominant microbial genera identified in the inoculum (sewage sludge) and liquid digestate, based on their relative abundance (>1%) and metabolic role in dark fermentation and anaerobic digestion.

Phylums	Genus	Sewage Sludge (%)	Liquid Digestate (%)	Metabolic Pathway *	Main Products	Main Function
Firmicutes	Anaerosalibacter	6.008	6.952	Hydrolysis and Acidogenesis (Stickland-like fermentation)	Acetate, H2, CO2, fatty acids	Performs Stickland-like fermentation; contributes to early-stage amino acid catabolism.
Firmicutes	Anaerosporobacter	1.40	1.04	Hydrolysis and Acidogenesis (Stickland-like fermentation)	Acetate, butyrate, H2, CO2	Anaerobic amino acid fermenter; contributes to hydrogen and short-chain fatty acid production.
Firmicutes	Bacillus	45.11	0.46	Hydrolysis and Acidogenesis (glucose mineralization)	Acetate, butyrate, H2	Sugar fermenter via glucose mineralization; produces acidogenic intermediates including H2.
Firmicutes	Caldicoprobacter	1.096	1.305	Hydrolysis and Acidogenesis (protein and carbohydrate breakdown)	Lactate, acetate, VFAs	Thermotolerant fermenter of protein and starch; associated with lactate accumulation.
Firmicutes	Caproiciproducens	0.48	42.19	Hydrolysis and Acidogenesis (beta-oxidation)	Caproate, acetate, H2	Chain elongator; converts lactate/ethanol into caproate under syntrophic conditions.
Firmicutes	Clostridiales	0.301	8.420	Hydrolysis and Acidogenesis (mixed fermentation)	Acetate, butyrate, H2	Diverse fermenters; involved in mixed substrate fermentation under anaerobic conditions.
Firmicutes	Clostridium	1.49	11.27	Hydrolysis and Acidogenesis (amino acid fermentation and lactate oxidation)	Acetate, H2, CO2	Key acidogenic genus; ferments amino acids and sugars into hydrogen and VFAs.
Firmicutes	Desulfotomaculum	1.513	1.330	Sulfate Reduction (butyrate/acetate to H2S)	H2S, isobutyrate	Sulfate-reducing bacteria convert VFAs to H2S, reducing methanogenic efficiency.
Firmicutes	Eubacterium	1.18	0.00	Acetogenesis and Hydrogen Metabolism (Wood–Ljungdahl)	Acetate, ethanol, H2	Performs acetogenesis via Wood–Ljungdahl; contributes to acetate and ethanol pools.
Firmicutes	Lactobacillus	0.509	3.346	Hydrolysis and Acidogenesis (lactate production)	Lactate, minor acetate	Produces lactate under anaerobic conditions; may lower pH and inhibit hydrogen producers.
Bacteroidota	Proteiniphilum	2.771	0.000	Hydrolysis and Acidogenesis (protein degradation)	Acetate, H2, CO2	Proteolytic anaerobe; participates in amino acid fermentation and syntrophic acetate oxidation.
Proteobacteria	Proteus	3.625	0.000	Hydrolysis and Acidogenesis (fermentation of amino acids/lactate)	Acetate, H2, CO2	Opportunistic fermenter of amino acids/lactate; possible participant in cross-feeding reactions.
Firmicutes	Rhabdanaerobium	0.013	5.150	Hydrolysis and Acidogenesis (mixed substrate fermentation)	Acetate, butyrate, VFAs	Strict anaerobe; ferments mixed substrates to butyrate and acetate.
Firmicutes	Sedimentibacter	1.68	0.00	Hydrolysis and Acidogenesis (lactate oxidation)	Acetate, H2, CO2	Ferments lactate into acetate and hydrogen under thermophilic conditions.
Firmicutes	Sporanaerobacter	7.670	0.115	Hydrolysis and Acidogenesis (fermentation under stress)	Caproate, butyrate, H2	Ferments ethanol/lactate under stress; involved in caproate production via chain elongation.

* Classification is grounded in established microbial pathways that transform complex organic substrates into hydrogen, volatile fatty acids (VFAs), and other intermediate metabolites. Hydrolysis and Acidogenesis. Proteins, carbohydrates, and lipids are broken down via reactions such as glycine cleavage, Stickland-like fermentation, lactate oxidation, beta-oxidation, and glucose mineralization. These reactions yield intermediates—acetate, hydrogen, carbon dioxide, and short-chain fatty acids. Acetogenesis and Hydrogen Metabolism. Syntrophic acetate oxidation, homoacetogenesis, and related pathways convert acetate, propionate, and hydrogen into further substrates. Pathways including the Wood–Ljungdahl and methylmalonyl-CoA routes are reported. Methanogenesis. Methane production occurs predominantly via hydrogenotrophic methanogenesis, with hydrogen and carbon dioxide as substrates, in addition to acetoclastic and, less commonly, methylotrophic processes. Sulfate Reduction. Sulfate-reducing bacteria convert butyrate and acetate into isobutyrate and hydrogen sulfide.

provides a functional classification of these genera based on their main metabolic pathways as reported in the scientific literature [43,44,45,46,47,48,49,50,51,52]. Each genus was functionally classified based on literature-supported evidence of its predominant metabolic activity under anaerobic conditions relevant to hydrogen production.

Among all the diverse possible pathways for biohydrogen production by anaerobes, facultative anaerobes, aerobes, methylotrophs, and photosynthetic bacteria [53], the research results suggest that the H₂ production was due to the presence of anaerobic and aerobic H₂-producing bacteria (HPB) in the reactor, specific the Firmicutes phylum, such as the genus Bacillus (44.6%), Clostridium (1.4%) and Romboutsia in smaller quantities (0.091%); and the phylum Proteobacteria with the genus Proteus (3.5%).

It has been widely identified that the phylum Firmicutes includes the main H₂-producing microorganisms and can utilize various substrates to produce H₂ [36]. Clostridium bacteria is one of the most common and usually dominant genera in these systems [54,55,56] presenting the highest H₂ yields in pure cultures via DF [56,57,58]. The Clostridium genus is usually one of the protagonist bacteria in biohydrogen production and inoculation with anaerobic sludge, according to J. Wang and Yin 2019 [55], usually leading to the domination of Clostridium spp. Nevertheless, the results of this experiment detected the presence of only 1.48% of this genus in the reactor, which is attributed to the fact that generally, the pretreatment method (to eliminate or reduce populations of metabolic competitors) for the inoculum is heat treatment, leaving only sporulating bacteria [1], such as Clostridium. However, considering that in this case, the pretreatment was through centrifugation and aeration, the preservation of non-sporulating bacteria is also suggested.

In this experiment, most H₂ production is attributed to the presence of the genus Bacillus, which represents 44.63% of the total microbial community and is the genus with the highest relative abundance in the reactor. However, the activity level was not determined, so further analyses are encouraged. Diverse research has proven that various Bacillus species have a significant effect on volumetric H₂ production, even with better performance than Clostridium species in certain conditions [59]. The main Bacillus species that proliferate through DF are B. subtilis [34,60], B. paramycoides [61], B. coagulans [62,63], B. circulans [63], B. licheniformis [53,64], and other Bacillus strains [65].

The presence of metabolic competitors was also detected. Lactic acid bacteria (LAB), one of the principal microbial inhibitors, was found in the experiment. Nevertheless, it did not significantly affect biohydrogen production, considering that it was found in less than 0.5% of the total microbial community (0.47%). Even when LAB is reported as one of the most frequent causes of biohydrogen production inhibition, this genus can also positively relate to HPB [66]. As Villanueva-Galindo et al., 2023 [59], mention, LAB produces the lactic acid (HLa) metabolite, which can then be used for biohydrogen production through its consumption by HPB, such as Clostridium. This is considered a novel strategy for H₂ production.

Two of the main processes linked to the consumption of the produced H₂ in the reactors are homoacetogenesis and methanogenesis (hydrogenotrophic) [67]. Methanogenic microorganisms were detected, which is the case of methanosaeta. However, as seen in Figure 7, the content of these archaea in the inoculum is less than 2% of the total, which decreased to 0.0129% at the end of the experiment. This decrease is attributed firstly to the inoculum’s pretreatment and, secondly, to the high production of VFAs; thus, a reduction in pH occured, which partially inhibited the action of methanogenic archaea. As for homoacetogenesis, another bacterium that also contributed to the consumption and decreased the amount of H₂ produced is Eubacterium, found in 1.16% of the inoculum composition. It must also be mentioned that some Clostridium species have a versatile metabolism and can consume H₂ [56], which might have also attenuated the final H₂ production. Nevertheless, according to the Pearson correlation results depicted in Figure 8, the Clostridium genus did not show a significant correlation (p > 0.05) with the physicochemical variables or the dominant genera in the inoculum. However, a negative correlation was observed between Clostridium and other dominant genera in the resulting population of the reactor (after 39 days).

Additionally, another important method of hydrogen consumption in fermentation systems is through sulfate reduction. In this regard, the relative abundance of the genus Desulfotomaculum—a well-known sulfate-reducing bacterium (SRB) from the phylum Desulfobacterota—remained relatively low and stable throughout the experiment, increasing only slightly from ~1.0% in the inoculum to 1.5% in the final reactor sample. Desulfotomaculum is a spore-forming anaerobe capable of using hydrogen as an electron donor to reduce sulfate to hydrogen sulfide (H₂S), under strictly anaerobic conditions and within an optimal pH range of 6.5 to 8.0 [68,69]. However, the mildly acidic environment during the initial 100 h of operation (~5.4 ± 1.1), along with the potential limitation of sulfate availability, may have constrained its metabolic activity and prevented a more substantial increase in population.

The limited increase in Desulfotomaculum abundance indicates that sulfate reduction was not a major hydrogen-consuming pathway under the experimental conditions. This is consistent with the marked rise in methane production observed during the later stages, suggesting that methanogenesis and homoacetogenesis were the dominant pathways. Although sulfate reduction is thermodynamically more favorable than methanogenesis [70,71], the system conditions likely suppressed SRB activity, reducing competition for hydrogen and favoring the growth of methanogenic archaea. These dynamics are consistent with the observed stabilization of pH toward neutral values (6.8 ± 0.2) and the increase in the methane production percentages between 100 and 500 h.

On the other hand, the genera Bacillus, Sporanaerobacter, Proteus, and Proteiniphilum were positively correlated (p < 0.05). These results suggest an antagonism and metabolic competition between Clostridium and the dominant genera in the reactor sample, thus explaining the reduction in relative abundance of Clostridium from 7.1% (inoculum) to 1.4%, and the decrease in population diversity and homogeneity observed at the end of the test operation.

Additionally, the genus Anaerosalibacter showed a significant correlation (p = 0.02) with the pH of the medium, indicating that members of this genus had a competitive advantage in the average pH conditions (6.7 ± 0.3) detected in the system. Anaerosalibacter, a strictly anaerobic genus capable of growth at pH 6.0–9.0, was described as a type of acetic acid bacteria that plays a crucial role in dark fermentation. Studies reported that the metabolic activity of Anaerosalibacter and other bacteria, such as Bacillus and Clostridium, can lead to the production of hydrogen and other fermentation products [72], which may explain the increase in their relative abundance from 0.1% to 6.0% after 39 days of operation.

A negative and significant correlation (p = 0.03) was observed between the ORP and pH variables. This means that higher pH values are associated with more reductive environments (−mV), while lower pH values are associated with more oxidative environments (+mV). This result was expected because in anaerobic conditions, microorganisms typically produce reducing equivalents, such as NADH and FADH2, as part of their metabolism [73]. These reducing equivalents can contribute to lowering oxidizing agents in the system, causing a decrease in ORP. Furthermore, biohydrogen production during dark fermentation can contribute to the reduction in oxidizing agents in the system, which also leads to a decline in ORP [74]. In our experiments, it was observed that during the first 100 h of the reactor, when H₂ and CH₄ production was minimal (Figure 4), the average ORP and pH values were −360.8 ± 50.2 mV and 5.4 ± 1.1, respectively. Between 100 and 500 h, when hydrogen production reached its highest peak, the average ORP and pH values were −539.7 ± 44.1 mV and 6.8 ± 0.2, respectively. Initial pH values can influence the composition of microbial communities during dark fermentation, with alkaline pH favoring the activity of hydrogen-producing bacteria [75]. In this sense, although pH and ORP did not significantly correlate with any bacterial genera, it is hypothesized that the highly reductive environment (−mV) of the system is responsible for the change in distribution and diversity of dominant genera present in the inoculum samples (Caldisericum, Smithella, Romboutsia, Ferritrophicum, Lactivibrio, and Denitratisoma) and at the end of the operation (Bacillus, Sporanaerobacter, Proteus, and Proteiniphilum).

Accordingly, even though organic wastes used as substrate already contain a diverse community of indigenous microorganisms that produce H₂ [76], incorporating and using sewage sludge as an inoculum source introduces a mixed culture of microorganisms that carries out diverse reactions within the reactor that contribute and simultaneously slow down or inhibit biohydrogen production. Considering this, it was decided to use sewage sludge as a mixed culture inoculum to find simple methodologies for a realistic project that can be escalated and eventually turned into commercial or industrial-scale plants.

The results suggest the positive interaction and contribution of the OFMSW’s indigenous bacteria and the sewage sludge’s microbial communities. This concurs with Favaro et al., 2013 [35], who demonstrated the positive interactions between indigenous OW bacteria and anaerobic sludge, which enhanced H₂ production. They added that their hydrolytic capacity should be reinforced to produce higher amounts of H₂, which is also inferred from the results obtained in this study.

Moreover, as Villanueva-Galindo et al., 2023 [59], pointed out, the precise role of the detected microorganisms has yet to be defined, and more studies are crucial to developing more favorable conditions to optimize the systems. Although this study focused on taxonomic profiling through 16S rRNA gene sequencing, further research should integrate functional approaches to validate the metabolic activity of the identified microbial genera and strengthen the understanding of biohydrogen production pathways.

Furthermore, there is still a knowledge gap about the interactions and behavior of some microorganisms present in the DF processes for H₂ production [59]. State-of-the-art technologies, such as synthetic biology and multi-omics, are currently aiding in a better understanding of microbial interactions, which could allow for more efficient microbial consortia to improve H₂-producing systems [77]. Thus, understanding these relationships becomes relevant to promoting beneficial biological interactions that optimize biohydrogen production in HSDF systems.

4. Conclusions

Dark fermentation technologies are currently seen as a possible scalable technology for adequately treating the organic fraction of municipal solid waste and generating benefits for bioenergy and bioproduct generation. This research facilitates a better understanding of the feasibility of dark fermentation, specifically with high-solids conditions (HSDF) for this purpose, intending to simplify conditions and develop a real case situation with minimal pretreatment. In a biochemical hydrogen production test, the biohydrogen production rates of 38.4 NmL/gVS OFMSW present an exciting approximation of the viability of treating OFMSW through HSDF for efficient biohydrogen production. Higher yields could be obtained, for which further studies and explorations are needed to reinforce the hydrolytic capacity of the systems.

The microbial analysis conducted in this study was instrumental in understanding the microbial relationships in these systems. Bacillus spp. bacteria were found to be the dominant hydrogen producers. Other bacteria, primarily from the Firmicutes phylum, such as the genus Clostridium, and Romboutsia in smaller quantities, and the phylum Proteobacteria with the genus Proteus, were also identified as biohydrogen producers. These findings underscore the positive interaction and contribution of the OFMSW’s indigenous bacteria and the sewage sludge’s mixed microbial consortium, demonstrating efficient biodegradation rates when using sewage sludge as a co-substrate and an inoculum in high-solids conditions. Also, aeration was proven to be an efficient inoculum pretreatment method.

Furthermore, to make these technologies viable on a commercial scale, more exploration is needed to improve the efficiency of the systems; however, the results present an appealing insight into this potential. Furthermore, these technologies could be complemented, for example, with a high-solids anaerobic digestion system to produce biomethane in a second phase, raising revenues while reducing the quantities of organic substances. Thus, these results are expected to serve as a reference for future research, which may intend to improve conditions to increase the amounts of hydrogen produced and, at the same time, develop an industrial high-solid dark fermentation system that is technically and economically viable.