Interactive Effects of Sulfide Addition and Heat Pretreatment on Hydrogen Production via Dark Fermentation

Abstract

Despite being recognized as toxic in anaerobic systems, sulfide’s potential to enhance hydrogen fermentation via microbial modulation remains underexplored. This study evaluated the combined effects of sulfide concentration (0–800 mg S/L) and heat pretreatment on hydrogen production during dark fermentation (DF). Without pretreatment, hydrogen yield reached 83 ± 2 mL/g COD at 0 mg S/L but declined with increasing sulfide, becoming negligible at 800 mg S/L. In contrast, heat-pretreated inocula showed markedly improved performance: peak cumulative production (4628 ± 17 mL) and yield (231 ± 1 mL/g COD) were attained at 200 mg S/L, while the maximum production rate (1462 ± 64 mL/h) occurred at 400 mg S/L. These enhancements coincided with elevated acetic and butyric acids, indicating a metabolic shift toward hydrogen-producing pathways. The microbial analysis of heat-pretreated samples revealed an enrichment of Clostridium butyricum (from 73.1% to 87.5%) and Clostridium perfringens, which peaked at 13.5% at 400 mg S/L. This species contributed to butyric acid synthesis. At 800 mg S/L, Clostridium perfringens declined sharply to 0.6%, while non-hydrogenogenic Levilinea saccharolytica proliferated, correlating with reduced butyric acid and hydrogen output. These findings indicate that sulfide supplementation, when combined with heat pretreatment, selectively restructures microbial communities and metabolic pathways, enhancing DF performance.

1. Introduction

Hydrogen is increasingly recognized as a promising clean energy carrier owing to its high energy density (142 kJ/g) and its potential to replace fossil fuels in various industrial applications [1]. Among various production routes, biohydrogen from organic waste via dark fermentation (DF) offers environmental benefits, notably renewable energy generation and greenhouse gas reduction. DF is a strictly anaerobic process that involves the microbial breakdown of carbohydrates into hydrogen and organic acids [2]. However, its performance is strongly influenced by microbial interactions and by environmental and operational conditions, including nutrient levels and inhibitory compounds [3].

To improve DF efficiency, pretreatment strategies have been widely adopted [4]. Heat pretreatment, in particular, is an effective method for enriching spore-forming hydrogen-producing bacteria (HPB), such as Clostridium spp., while simultaneously suppressing non-spore-forming, lactic acid-producing bacteria that compete for substrates [5]. Numerous studies have demonstrated the effectiveness of heat pretreatment in enhancing hydrogen yield. For instance, Liu et al. (2009) reported that heating mixed cultures at 100 °C for 30 min led to a 105% increase in hydrogen yield, accompanied by a marked enrichment of HPB [6]. Similarly, Kim et al. (2006) observed that heat pretreatment at 90 °C for 20 min successfully inhibited lactic acid-producing bacteria while promoting Clostridium dominance [7]. This genus is known for its robust metabolic versatility, resistance to environmental stress, and capacity to utilize a broad range of substrates, making it a key player in efficient DF [8,9].

In addition to pretreatment, the presence of inhibitory compounds such as sulfide can strongly influence DF performance. Sulfate (SO₄²⁻) is frequently found in waste streams from industries like petrochemicals, pulp and paper, and food processing [10]. Under anaerobic conditions, sulfate-reducing bacteria (SRB) convert sulfate to sulfide (S²⁻), which can also originate from the breakdown of sulfur-containing amino acids [11]. Sulfide presents two major challenges: competition with HPB for reducing equivalents through SRB activity, and the direct inhibition of microbial metabolism [12]. In acidic environments typical of DF, sulfide exists primarily as hydrogen sulfide (H₂S), which can penetrate microbial membranes and interfere with intracellular enzymes and electron transport chains [13]. Importantly, the influence of sulfide is dose-dependent. While concentrations as high as 500 mg S/L have been shown to severely inhibit hydrogen generation by up to 90%, low-to-moderate levels may enhance hydrogen production or redirect metabolic pathways [14,15]. These findings highlight the need for precise sulfide control to maintain optimal DF activity. Despite this, the interactions between sulfide stress and microbial community dynamics, particularly in conjunction with pretreatment, remain poorly defined. Clarifying these interactions is crucial for improving the operational robustness of DF systems treating heterogeneous organic waste streams.

Accordingly, this study aimed to investigate the combined effects of sulfide concentration and heat pretreatment on hydrogen production, organic acid metabolism, and microbial community structure in DF using sucrose as a model substrate. The specific goals were to (1) quantify hydrogen production and analyze organic acid profiles across a sulfide concentration gradient (0–800 mg S/L) under both untreated and heat-pretreated inocula; and (2) characterize shifts in microbial community structure, with a focus on HPB enrichment and the suppression of non-hydrogenogenic populations. To our knowledge, this is the first study to systematically integrate graded sulfide dosing and thermal pretreatment to assess their combined effects on hydrogen yield, organic acid metabolism, and microbial community dynamics in DF. By addressing the interactive impact of biochemical and physical stressors, this work provides novel insights into the microbial and metabolic mechanisms that form the basis of process stability and biohydrogen production in sulfide-rich waste treatment.

2. Materials and Methods

2.1. Inoculum and Substrate Preparation with Thermal Pretreatment

DF was conducted using sucrose (99%; Daejung Chemicals Co., Ltd., Siheung-si, Republic of Korea) as a substrate to investigate the effect of sulfide concentration. The seed sludge was obtained from an anaerobic digester at a brewery wastewater treatment facility in Republic of Korea. The sludge was maintained at 25 °C prior to use and showed the following characteristics (mean ± SD): TCOD, 91,912 ± 6393 mg/L; SCOD, 22,840 ± 2390 mg/L; TS, 66,281 ± 405 mg/L; VS, 59,160 ± 448 mg/L; alkalinity, 7425 ± 15 mg CaCO₃/L; and pH 7.5. The seed sludge was divided into two groups: one subjected to heat pretreatment and one untreated. The condition of 90 °C for 20 min was selected as it has been frequently employed in dark fermentation studies using food waste, where it effectively enriched hydrogen-producing bacteria and enhanced process performance [8,9,16]. Heating beyond 90 °C was avoided due to the excessive energy required for water evaporation, which can significantly reduce energy efficiency [17]. Sodium sulfide pentahydrate (Na₂S·5H₂O, 98%; Daejung Chemicals Co., Ltd., Siheung-si, Republic of Korea) was used as the sulfide source, and sulfide concentrations in this study were expressed as mg S/L.

2.2. Experimental Setup

Eight anaerobic reactors, each with a total volume of 800 mL (1450 mm in height and an internal diameter of 838 mm), were utilized for hydrogen production through DF. Each reactor was loaded with seed sludge to 30% of the working volume, corresponding to 1.8 g of VS, and the remaining volume was filled with tap water to reach a total working volume of 500 mL. Sucrose was used as the carbon source to achieve a final substrate concentration of 40 g COD/L. Sodium sulfide pentahydrate (Na₂S·5H₂O) was added to each reactor to obtain final sulfide concentrations of 0, 200, 400, and 800 mg S/L. The inoculum and substrate mixture were manually homogenized prior to sealing. The initial pH of each reactor was adjusted to 8.0 ± 0.2. During operation, pH was maintained at 5.5 ± 0.1 using 3 N KOH and 3 N HCl as needed. To ensure anaerobic conditions, the reactors were purged with high-purity nitrogen gas (99.999%) for 5 min prior to sealing. The headspace volume was approximately 300 mL. Each reactor was incubated at 35 ± 0.1 °C and stirred continuously at 280 rpm with a magnetic stirrer. Gas samples were collected from the headspace throughout the 48 h fermentation period. All experiments were performed in duplicate to ensure reproducibility.

2.3. Analytical Methods

Gas generation was quantified using a wet gas measurement system. The hydrogen generated in the collected gas was analyzed using a gas chromatograph (SRI 310, SRI Instruments, Torrance, CA, USA) equipped with a thermal conductivity detector (TCD) and a HayeSep T column (10 ft × 1/8″). High-purity nitrogen (N₂, 99.999%) was utilized as the carrier gas at a flow rate of 10 mL/min. Retention time was compared against a hydrogen standard, and quantification was performed using an external calibration curve.

For the analysis of organic acids, fermentation broth samples were centrifuged using a FELTA-5 centrifuge (Hanil, Seoul, Republic of Korea) and subsequently filtered through a 0.45 μm membrane filter (Advantec, Durham, NC, USA). The filtrates were analyzed using a high-performance liquid chromatograph (Ultimate 3000, Thermo Fisher Scientific, Waltham, MA, USA) equipped with a UV–Vis detector (VWD-3400RS) and an Aminex HPX-87H column (300 mm × 7.8 mm, Bio-Rad, Cotati, CA, USA). The mobile phase consisted of 0.008 N sulfuric acid (H₂SO₄), with a flow rate of 0.6 mL/min. Physicochemical parameters including COD, TS, VS, alkalinity, and pH were determined in accordance with standard methods using the closed reflux, gravimetric, and colorimetric protocols [18].

For microbial community analysis, 50 mL of fermentation slurry was centrifuged at 3000 rpm for 10 min. The pellet (approximately 2 mL) was used for DNA extraction with the Fast DNA Spin Kit for soil (QBioGene, Carlsbad, CA, USA). Purification was performed using the UltraClean Microbial DNA Isolation Kit (Mo Bio Laboratories, Carlsbad, CA, USA). The V1–V3 regions of the 16S rRNA gene were amplified universal primers 27F (5′-GAGTTTGATCMTGGCTCAG-3′) and 518R (5′-WTTACCGCGGCTGCTGG-3′). CR amplification was conducted with 5 ng of genomic DNA in a 50 µL reaction mixture containing a 5× reaction buffer, 1 mM dNTP mix, 500 nM of each primer, and Herculase II Fusion DNA Polymerase (Agilent Technologies, Santa Clara, CA, USA). The thermal cycle included initial denaturation at 95 °C for 3 min, followed by 25 cycles of 95 °C for 30 s, 55 °C for 30 s, and 72 °C for 30 s, with a final extension at 72 °C for 5 min. PCR products were purified using AMPure XP magnetic beads (Agencourt Bioscience, Beverly, MA, USA). Library preparation was performed using Nextera XT Indexed Primers, and a second PCR was conducted under the same conditions for 10 cycles. The final library was purified and quantified using qPCR (KAPA Library Quantification Kit for Illumina platforms) and quality-checked using the TapeStation D1000 ScreenTape system (Agilent Technologies, Santa Clara, CA, USA). Sequencing was performed using the Illumina MiSeq™ platform (Illumina, San Diego, CA, USA).

Sequence processing, operational taxonomic unit (OTU) clustering (97% similarity), taxonomy assignment, and diversity analysis were performed using the MOTHUR software 1.36.1 package. Chimera sequences and low-quality reads were removed according to the standard protocol described by Schloss et al. (2009) [19].

2.4. Statistical Analysis

Hydrogen production was quantified using a mass balance approach, which incorporated measurements of headspace gas composition and the total gas volume collected at each time interval. The cumulative hydrogen data were fitted to the modified Gompertz equation (Equation (1)) to estimate the kinetic parameters such as hydrogen production potential, production rate, and lag phase, following the method of Nath and Das (2011) [20].

$$H(t)=P·exp\left\{-exp\left[{\displaystyle \frac{R_m·e}{P}}\left(\lambda -t\right)+1\right]\right\}$$

(1)

where H(t) is the cumulative hydrogen (mL) at cultivation time t (d); P is the ultimate hydrogen production (mL); R_m is the hydrogen production rate (mL/d); λ is the lag phase (d); and e is the exp (1) = 2.71828.

Model fitting and parameter estimation were performed using non-linear regression techniques implemented in Origin 2023 (OriginLab Corporation, Northampton, MA, USA), and the goodness of fit was assessed by the coefficient of determination (R²).

3. Results and Discussion

3.1. Impact of Sulfide and Heat Pretreatment on Biohydrogen Production

Figure 1 illustrates the cumulative hydrogen production under various sulfide concentrations (0–800 mg S/L) with and without heat pretreatment of the inoculum. Hydrogen production parameters, including yield, production rate, and lag phase, were fitted using the modified Gompertz equation. The regression results are summarized in

Table 1. Effect of heat pretreatment and sulfide concentration (0–800 mg S/L) on cumulative hydrogen production, yield, rate, and lag phase during dark fermentation.

	Sulfide (mg S/L)	Cumulative Hydrogen Production (mL)	Hydrogen Yield (mL/g COD)	Hydrogen Production Rate (mL/h)	Lag Phase (h)
Without heat pretreatment	0	1649 ± 34	83 ± 2	575 ± 40	23 ± 0
	200	737 ± 12	37 ± 1	347 ± 24	24 ± 0
	400	578 ± 17	29 ± 1	251 ± 28	24 ± 0
	800	2 ± 0	0 ± 0	1 ± 0	21 ± 0
With heat pretreatment	0	3199 ± 36	160 ± 2	549 ± 22	17 ± 0
	200	4628 ± 17	231 ± 1	1144 ± 22	8 ± 0
	400	4463 ± 31	223 ± 2	1462 ± 64	10 ± 0
	800	3430 ± 40	172 ± 2	303 ± 6	14 ± 0

, and the coefficient of determination (R² > 0.90) for all conditions confirms the model’s reliability and adequate goodness of fit.

When the 0 mg S/L condition was compared, heat-pretreated inoculum showed substantial improvement in performance. Cumulative hydrogen production reached 3199 ± 36 mL, which was 94% higher than the non-pretreated control (1649 ± 34 mL). The hydrogen yield (160 ± 2 mL/g COD) and production rate (549 ± 22 mL/h) also improved, while the lag phase was reduced from 23 to 17 h. These findings suggested that heat pretreatment selectively survived spore-forming HPB while eliminating thermally sensitive competitors, leading to earlier fermentation onset and increased productivity.

In the absence of pretreatment, increasing sulfide concentrations led to a consistent decline in hydrogen production. The control group (0 mg S/L) showed the highest cumulative production, whereas 200 and 400 mg S/L conditions resulted in 55% and 65% reductions, respectively. At 800 mg S/L, hydrogen production was nearly completely inhibited. Similar decreasing patterns were observed for yield and production rate, and the lag phase was progressively extended. These results align with previous studies showing that elevated sulfide levels inhibit microbial activity, likely due to protein denaturation and the disruption of key metabolic enzymes [21,22].

Under heat-pretreated conditions, however, hydrogen production responded differently to sulfide levels. At moderate concentrations (200–400 mg S/L), hydrogen production was notably enhanced. The 200 mg S/L condition resulted in the highest cumulative production (4628 ± 17 mL) and yield (231 ± 1 mL/g COD), while the highest production rate was observed at 400 mg S/L. In both cases, the lag phase was shortened by up to 53% compared to the control. However, at 800 mg S/L, the overall productivity declined, and the production rate dropped below that of the control, indicating that the inhibitory effect of sulfide prevailed at high concentrations. These findings demonstrated that the response of hydrogen production to sulfide stress is significantly influenced by inoculum pretreatment. The observed patterns likely reflect changes in microbial community composition and metabolic resilience. In particular, the selective survival of spore-forming HPB and the suppression of competing microbes under heat pretreatment may help maintain hydrogen production even under moderate sulfide stress.

3.2. Modulation of Organic Acid Pathways by Sulfide Levels and Pretreatment in Dark Fermentation

This study analyzed the impact of inoculum pretreatment and sulfide concentration on organic acid profiles during DF. As key intermediates of hydrogen-producing metabolic pathways, organic acids serve as indicators of microbial activity and fermentation mechanisms. The results are presented in Figure 2 and Figure 3. Under the baseline condition with no sulfide addition (0 mg S/L), distinct differences were observed in organic acid composition depending on pretreatment. In the pretreated group, butyric acid constituted 47% of the total organic acids and reached 9795 ± 626 mg COD/L, while in the non-pretreated group it accounted for only 28% and 2805 ± 338 mg COD/L, respectively (Figure 3). Although acetic acid’s percentages were similar in both cases (25–59%), its absolute concentration was slightly lower in the pretreated sample (5177 ± 426 mg COD/L). Lactic acid, a non-hydrogen-producing byproduct, accumulated more in the pretreated group (5695 ± 2137 mg COD/L) compared to the non-pretreated one (1201 ± 581 mg COD/L), indicating a potential shift to alternative electron sinks. These results suggest that inoculum pretreatment could modulate fermentation pathways and intermediate accumulation even under the same sulfide condition.

In non-pretreated samples, increasing sulfide concentrations had a limited impact on organic acid composition. Lactic acid remained below 2000 mg COD/L across all treatments, and butyric acid began accumulating only after 16 h, reaching a modest level of 1325–3100 mg COD/L (Figure 2). Acetic acid remained relatively stable up to 20 h and then increased under higher sulfide conditions (400 and 800 mg S/L), ultimately reaching 7384–8028 mg COD/L. Despite this, the total organic acid accumulation remained constrained, with acetic acid dominant and lactic acid comprising a larger relative proportion. This indicated suppressed microbial activity and inefficient fermentative metabolism under sulfide stress when pretreatment is not applied.

In the heat-pretreated condition, sulfide concentration had a pronounced effect on organic acid profiles (Figure 2). At 200 and 400 mg S/L, lactic acid levels were minimal (1772–1785 mg COD/L), accounting for only 7% of total organic acids (Figure 3), suggesting efficient electron flow toward hydrogen-producing pathways. In contrast, acetic acid and butyric acid reached 8891–9857 mg COD/L and 13,538–13,747 mg COD/L, respectively, dominating the organic acid pool with proportions of 36–39% and 54–56%. Acetic acid and butyric acid production were directly associated with hydrogen evolution, yielding 4 mol and 2 mol H₂ per mol glucose, respectively, making these results indicative of enhanced metabolic efficiency and hydrogen productivity.

At the highest sulfide concentration (800 mg S/L), the organic acid profile shifted unfavorably. Lactic acid sharply increased to 34%, while the butyric acid concentration decreased to 9670 ± 1260 mg COD/L. Lactic acid production not only serves as an alternative electron sink, but also competes with hydrogenogenic pathways by consuming NADH. The inhibition of lactic acid dehydrogenase activity has been shown to redirect electron flow and enhance hydrogen production [23]. Moreover, the carboxylic hydrogen of lactic acid has a stronger dissociation tendency than that of acetic or butyric acid, which can promote intracellular acidification, impair cell function, and further reduce hydrogen productivity [24].

Therefore, the results suggest that elevated sulfide concentrations disrupted cellular metabolism, thereby suppressing hydrogenogenic activity and altering the overall fermentation profile [21,22]. These organic acid distribution patterns correspond with the hydrogen production trends observed in Section 3.1, particularly the superior performance under moderate sulfide levels (200–400 mg S/L) in pretreated samples. The prevalence of acetic acid- and butyric acid-type fermentations under these conditions reflects the effective routing of electrons and energy conservation. In the absence of pretreatment, such favorable metabolic routes remained underutilized, likely due to the insufficient activation of HPB and limited electron flow toward hydrogen-producing reactions. These mechanistic differences are further elaborated in Section 3.3 through microbial community analysis.

3.3. Conditional Microbial Responses and Mechanistic Interpretation Under Sulfide Stress in Dark Fermentation

3.3.1. Microbial Community Dynamics Under Sulfide and Heat Pretreatment

Building upon the findings in Section 3.1 and Section 3.2, which demonstrated that hydrogen production and organic acid profiles were significantly influenced by sulfide concentration and pretreatment, microbial community analysis was conducted to explore the underlying biological mechanisms. Due to the consistently low hydrogen production and limited fermentation activity observed in non-pretreated samples—even at the lowest sulfide concentration—these groups were considered to exhibit only minimal metabolic responsiveness to sulfide stress. Therefore, microbial community analysis was focused exclusively on heat-pretreated conditions, where distinct metabolic shifts and enhanced performance were apparent, allowing for a clearer interpretation of biologically relevant changes in response to the presence and concentration of sulfide. The microbial composition of the heat-pretreated samples under varying sulfide concentrations is presented in Figure 4, illustrating bacterial distributions at both the genus (Figure 4a) and the species levels (Figure 4b). Minor genera and species accounting for less than 0.5% of the total community were grouped as “Others”.

In the control sample (0 mg S/L), the dominant genera were Clostridium (78.6%), Romboutsia (15.3%), and Paraclostridium (5.1%), all known to be spore-forming bacteria [25]. Clostridium, in particular, is widely recognized for its central role in DF due to its ability to utilize diverse substrates and tolerate harsh conditions [26]. It plays a direct role in hydrogen production through enzymatic pathways involving ferredoxin and hydrogenase, which are essential for electron transfer during fermentation [27]. Although lactic acid-producing bacteria were not detected as dominant groups at the end of fermentation, it is possible that certain lactic acid-producing bacteria capable of forming spores remained metabolically active during the early stages and contributed to transient lactic acid accumulation [28]. Nevertheless, this metabolic shift led to poor hydrogen yields compared to the moderate sulfide conditions (200–400 mg S/L), where lactic acid suppression and the enhanced activity of HPB were more prominent. Notably, the relative abundance of Clostridium increased with moderate sulfide concentrations, reaching 88.3% at 200 mg S/L and 95.4% at 400 mg S/L. These concentrations also corresponded with the highest hydrogen production levels, suggesting a strong correlation between Clostridium dominance and biohydrogen yield. Concurrently, the relative abundance of Romboutsia decreased from 15.3% to 2.6% at 200 mg S/L, while Paraclostridium increased from 5.1% to 7.8%, indicating a moderate sulfide-induced shift in community structure.

At the species level, Clostridium butyricum emerged as the most dominant HPB, accounting for 73.1% in the control sample (0 mg S/L) and increasing to 87.5% and 76.4% in the 200 and 400 mg S/L samples, respectively. This species was widely recognized for its robust hydrogenogenic potential and versatility in substrate utilization [9]. Its prevalence aligns well with the elevated hydrogen yields observed under moderate sulfide stress. Additionally, the presence of Clostridium perfringens rose notably to 13.5% at 400 mg S/L, a species associated with butyric acid-type fermentation, consistent with the increased butyric acid concentrations discussed in Section 3.2 [29,30].

Notably, Lactobacillus helveticus, a lactic acid bacterium known for its inability to form spores, was nearly undetectable under all tested conditions. This observation suggests that heat pretreatment selectively eliminated non-spore-forming populations, thereby suppressing lactic acid-producing microbes. Such suppression likely redirected reducing equivalents toward hydrogenogenic pathways rather than toward lactic acid synthesis, resulting in lower lactic acid accumulation and enhanced hydrogen production. Meanwhile, Paraclostridium benzoelyticum, which accounted for 7.7% of the microbial community at 200 mg S/L, has previously been identified as a contributor to hydrogen production under heat-pretreated conditions, indicating a possible functional role in the favorable outcomes observed at this sulfide level [31].

At the highest sulfide concentration (800 mg S/L), the microbial structure remained dominated by Clostridium genera (89.1%), with species Clostridium butyricum accounting for 87.8% of the total. However, the relative abundance of Clostridium perfringens dropped sharply to 0.6%, correlating with the reduced butyric acid levels and hydrogen productivity observed in this sample. Although Clostridium butyricum remained dominant at 800 mg S/L, hydrogen production was not sustained under this condition. This discrepancy may be explained by two factors: first, the pronounced decline of Clostridium perfringens, a key contributor to the acetyl-CoA to butyric acid fermentation pathway, likely impaired butyrate synthesis, which is closely tied to hydrogen generation [29]. In addition, Clostridium perfringens is known to harbor hydrogenase systems and to exhibit growth inhibition at elevated sulfide concentrations, suggesting that hydrogenase inactivation may have contributed to the suppression of hydrogen production at 800 mg S/L [32]. Second, the metabolic activity of Clostridium butyricum itself may have been suppressed under elevated sulfide stress. Sulfide toxicity can differentially affect bacterial groups and potentially divert electron flow away from hydrogen-producing pathways [33]. In addition, Levilinea saccharolytica appeared at a low abundance (1.4%) under elevated sulfide stress. This species, which was not typically involved in high-yield hydrogen-producing pathways, may reflect a metabolic shift toward less efficient fermentative routes [34,35].

3.3.2. Mechanistic Insights into Sulfide Toxicity and Conditional Microbial Responses

To refine the understanding of the conditional impact of sulfide on hydrogen-producing activity, this section offers a mechanistic interpretation of the observed performance dynamics, emphasizing heat pretreatment as a pivotal determinant. While Section 3.3.1 outlined microbial community alterations in response to varying sulfide levels, the current discussion focused on how these transitions are connected to known sulfide toxicity pathways and their interaction with community structures shaped by pretreatment.

Sulfide has been extensively associated with inhibitory effects on anaerobic microbial processes. The principal mechanisms include alterations in protein structure, the deactivation of thiol-dependent enzymes, the disruption of electron transport systems, and the sequestration of essential metal cofactors required for redox metabolism [36,37,38]. These biochemical interferences can substantially reduce energy yield and substrate conversion efficiency, particularly in non-spore-forming or physiologically vulnerable microorganisms, groups typically more abundant in unconditioned microbial consortia. In contrast, spore-forming hydrogen producers such as Clostridium spp. may possess greater tolerance to sulfide stress. Although empirical data on sulfide’s direct influence on spore-formers remain limited, the present findings suggest that microbial assemblages enriched through heat pretreatment, favoring endospore-forming taxa, were less impacted by inhibitory pressure and responded positively at intermediate sulfide concentrations. This observation implies that pretreatment-induced microbial robustness is instrumental in defining system resilience to sulfide exposure.

The sulfide concentrations utilized in this study (200, 400, and 800 mg S/L) corresponded with thresholds previously reported to hinder fermentative performance, particularly at concentrations exceeding 500 mg S/L [14]. While 200 and 400 mg S/L likely fell within a functional tolerance range under heat-pretreated conditions, 800 mg S/L was associated with a shift toward non-hydrogenogenic genera such as Levilinea, and a marked reduction in Clostridium perfringens, suggesting a deviation from optimal ecological balance.

Although heat pretreatment itself induces significant shifts in microbial community composition, our findings indicate that sulfide exerts additional, concentration-dependent effects even within the selectively enriched populations. For instance, while Clostridium spp. dominated across all pretreated conditions, distinct shifts in the relative abundance of Clostridium perfringens and Clostridium butyricum were observed across sulfide concentrations. This suggests that sulfide selectively influences metabolic dynamics beyond the baseline effects of pretreatment. Therefore, the microbial responses presented in this study reflect the combined and context-specific influence of both pretreatment and sulfide stress, rather than being driven solely by one factor. These findings suggest that sulfide’s role in DF is neither inherently beneficial nor uniformly inhibitory, but instead is shaped by the specific microbial community context and the applied concentration thresholds. While moderate dosing may favor hydrogenogenic taxa in selectively enriched environments, its applicability cannot be broadly assumed. Moreover, the limited body of knowledge concerning sulfide interactions with spore-forming organisms necessitates dedicated investigation. Clarifying these dynamics is essential for advancing biologically responsive sulfide regulation strategies aimed at enhancing the stability and efficiency of DF systems.

4. Conclusions

The present investigation explored how sulfide concentration and thermal pretreatment collectively influenced hydrogen productivity and microbial community responses in DF. In heat-pretreated systems, moderate sulfide dosing (200–400 mg S/L) facilitated hydrogen enhancement through a metabolic redirection toward acetate and butyrate pathways while concurrently suppressing lactate synthesis. These shifts were linked to the selective proliferation of spore-forming hydrogenogenic taxa, notably Clostridium butyricum and Clostridium perfringens, alongside a reduction in non-hydrogenogenic and lactate-producing populations. At 800 mg S/L, however, hydrogen production was markedly inhibited, suggesting that this concentration exceeds the microbial tolerance threshold for sulfide toxicity. Despite being performed under controlled conditions, the integrated microbial and metabolic findings delineate the biological basis underpinning sulfide-responsive hydrogen production. These outcomes offer a practical foundation for designing sulfide dosing frameworks and inoculum preconditioning approaches to enhance hydrogen yields in DF.