Biohydrogen Production from Industrial Waste: The Role of Pretreatment Methods

Abstract

This study aimed to investigate the effectiveness of dark fermentation in biohydrogen production from agro-industrial wastes, including apple pomace, brewer’s grains, molasses, and potato powder, subjected to different pretreatment methods. The experiments were conducted at a laboratory scale, using 1000 cm³ anaerobic reactors at a temperature of 35 °C and anaerobic sludge as the inoculum. The highest yield of hydrogen was obtained from pre-treated apple pomace (101 cm³/g VS). Molasses, a less complex substrate compared to the other raw materials, produced 25% more hydrogen yield following pretreatment. Methanogens are sensitive to high temperatures and low-pH conditions. Nevertheless, methane constituted 1–6% of the total biogas under these conditions. The key factor was appropriate treatment of the inoculum, to limit competition from methanogens. Increasing the inoculum dose from 150 cm³/dm³ to 250 cm³/dm³ had no further effect on biogas production. The physicochemical parameters and VFA data confirmed the stability and usefulness of activated sludge as a source of microbial cultures for H₂ production via dark fermentation.

1. Introduction

Developing and deploying renewable energy sources are key steps towards reducing the environmental impact of fossil fuel emissions, improving energy security, and addressing the challenges of climate change. Renewable energy sources are thus an essential foundation for economic prosperity [1]. Among renewable energy sources, hydrogen is gaining considerable attention from both scientists and industry as an emerging and sustainable alternative to fossil fuels [2]. As a clean energy carrier, hydrogen offers a range of benefits that can contribute to decarbonizing the transport, residential, commercial, and industrial sectors, in line with the European Union’s new decarbonization goal [3,4]. Hydrogen can have a combined effect with other low-carbon alternatives, enabling a more cost-effective transition to decarbonized and cleaner energy systems. Its energy content is relatively high, at approximately 122 kJ/g, which makes it about 2.75 times more efficient than conventional hydrocarbon fuels [5]. Hydrogen thus has great potential to become a mainstream energy source for transportation and manufacturing.

Unfortunately, hydrogen is still predominantly produced from non-renewable sources, through processes such as methane steam reforming, oil reforming, coal gasification, and water electrolysis [5,6]. Sustainable methods for hydrogen generation include the direct use of solar energy (e.g., for photoelectrochemical water splitting), the use of solar and wind energy for water electrolysis, and the conversion of fuels and biomass [7]. Biohydrogen production from biomass using various microorganisms offers a promising route to producing clean and sustainable fuel, contributing towards carbon neutrality and environmental sustainability [8,9,10]. The hydrogen production processes, such as dark fermentation (DF) and photofermentation, are considered more sustainable and economical than other biological methods, including bio-photolysis. However, these methods have limitations, such as low hydrogen yield and low conversion efficiency. Improvements are still necessary, including optimizing the selection of waste biomass, microorganisms, and process conditions [6,11].

There are numerous possible sources of biomass for biorefineries; however, to avoid jeopardizing food supplies, biowaste generated throughout the food supply chain (production, post-harvest processing, distribution, and consumption) seems to be a valuable substrate. Agricultural waste is a significant global issue, with an estimated annual production of between 1.3 billion and 2 billion tons [12,13]. The cost of food waste alone amounts to 1000 billion dollars annually, a sum that could potentially cover the nutritional needs of millions of people [14]. These wastes include a variety of materials, such as crop residues, animal manure, and processing byproducts, such as husks, peels, and pulp. Effective management of these wastes is crucial, as they have the potential to cause environmental pollution if not properly handled. However, they also possess value as a resource for various applications.

The waste is usually dominated by lignocellulosic biomass, which offers several notable advantages, including widespread availability, low material cost, and a high carbohydrate content, typically ranging from 40% to 65% of the dry weight. These characteristics make waste lignocellulosic biomass a promising feedstock for sustainable biohydrogen [6,15], as well as other biofuels such as biodiesel, bioethanol, and biogas [6]. Thus, agro-industrial wastes can be considered as a low-cost raw material in the production of green energy, including via DF processes [6,16].

Lignocellulosic biomass is plant-derived material composed primarily of cellulose, hemicellulose, and lignin. Due to this complex composition, most hydrolysis technologies incorporate pretreatment processes designed to disrupt the structural matrix—particularly the interactions between lignin and cellulose—to reduce cellulose crystallinity and facilitate the hydrolysis of hemicelluloses. Conventional pretreatment strategies include physical [17,18], chemical [17,19], and biological methods [15,20]. Chemical hydrolysis and high temperatures are used to break down macromolecules into monomeric sugars, including glucose, xylose, and arabinose. The resulting soluble hydrolysate is susceptible to all types of fermentation, but the insoluble fraction after pretreatment can also be further processed. From a biorefinery perspective, the use of both hydrolysates would be desirable for plant efficiency. The fermentable sugars in liquid hydrolysates (HLs) can be used as a carbon source for hydrogen production through DF bioprocess, while the insoluble biomass, i.e., hydrolyzed solids (HSs), can be converted into bioenergy, such as methane, through anaerobic digestion (AD) [21].

Dark fermentation (DF) is an anaerobic process divided into two stages: hydrolysis and acidogenesis. The organic matter is hydrolyzed into simple compounds, which are then converted into volatile fatty acids (VFAs), hydrogen (H₂), and CO₂ [5]. To increase the economic viability of the process, several key challenges must be overcome. These include enhancing hydrogen yield, reducing production costs through the optimization of DF parameters, improving substrate conversion efficiency to limit the formation of undesirable by-products and reduce the need for downstream effluent treatment, selecting effective microbial consortia for optimal substrate utilization, and successfully scaling up the process [5,22].

The key element appears to be the selection of an appropriate pretreatment strategy. Single pretreatment approaches often encounter limitations, including low efficiency, high formation of byproducts, and elevated costs [18,22]. For example, although acid pretreatment effectively removes hemicellulose, it can generate inhibitory by-products such as furfural (F) and 5-hydroxymethylfurfural (HMF), which negatively impact subsequent microbial fermentation, reducing product yield and quality [22,23]. Alkaline pretreatment, although efficient at removing lignin and enhancing cellulose accessibility without the formation of F and HMF, typically requires extreme conditions, such as high temperature or pressure [15,24].

This study investigated industrial waste feedstocks and pretreatment strategies for biohydrogen production, aiming to enhance the process and reduce costs to improve the commercial viability of biohydrogen production. The presented research results are one of the few reports comparing various food industry wastes as valuable raw materials for biohydrogen production.

2. Materials and Methods

2.1. Inoculation Material

Anaerobic sludge collected from the anaerobic mesophilic digester at the Municipal Wastewater Treatment Plant in Zduńska Wola, Poland, was used as inoculum for the experiments. The concentrations of total solids (TS) and volatile solids (VS) in the inoculum were 24.58 gTS/kg and 15.45 gVS/kg, respectively. Such a natural microbial consortium may include both hydrogen-producing bacteria and hydrogen-consuming methanogens. To prevent methane production and the associated consumption of hydrogen—particularly in batch experiments—the inoculum was pretreated. This process involved heating at 80 °C for 1.5 h and adjusting the pH to approximately 5.5. The presence of naturally occurring endogenous microorganisms in agro-industrial by-products or residues may also influence hydrogen generation during DF. The pretreatment was also designed to suppress the activity of microorganisms, which can either enhance or inhibit the DF process, depending on their metabolic activity. Experiments were conducted using apple pomace (AP), brewery spent grain (BSG), molasses (M), and potato waste (PW), which were stored at −30 °C before use.

2.2. Agro-Industrial Wastes

Agro-industrial wastes were obtained from Polish factories. Apple pomace was sourced from the SAMO JABŁKO Company in Klimczyce Kolonia, Poland (52°20′54″ N 22°49′42″ E). Spent brewer’s grain was obtained from DRINK ID Company in Piotrków Trybunalski, Poland (51°24′24″ N 19°41′44″ E). Potato waste (PW), a dry powder from puree production, was obtained from the Potato Factory SOLAN in Głowno, Poland (51°57′51″ N 19°42′42″ E). Molasses was sourced from the Sugar Factory in Dobrzelin, Poland (52°13′40″ N 19°37′13″ E).

2.3. Pretreatment and Hydrolysis of Industry Wastes

Due to the material’s unique nature, particularly its chemical composition, various pretreatment methods were employed. The brewers’ spent grain and apple pomace were treated by thermal (AP-T, BSG-T), alkaline (BSG-NaOH), acid (BSP-H₃PO₄), and enzymatic methods (AP-E, AP-ME, BSP-E) (Novonesis Viscozyme^® L and Novonesis Ultraflo^® Max, Novozymes, Bagsværd, Denmark), as well as combined hydrolysis and thermochemical treatment (BSG-TNaOH, BSP-TH₃PO₄). No pre-treatment was performed on the other wastes—i.e., molasses and potato waste (

Table 1. Hydrolysis as the pretreatment method used for individual wastes.

No.	Agro-Industrial Waste	Hydrolysis	Parameters
1	Apple pomace (AP)	thermal	125 °C, 15 min
	AP-E	enzymatic	UltraFlo® Max, 45 °C, 8 h
	AP-ME	enzymatic	Viscozyme® L, UltraFlo® Max, 45 °C, 8 h
2	Spent grain (BSG)	thermal	125 °C, 15 min
	BSG-E	enzymatic	Viscozyme® L, 45 °C, 8 h
	BSG-H3PO4	acid	10% v/v H3PO4, 4 h
	BSG-TH3PO4	thermal-acid	10% v/v H3PO4, 80 °C, 8 h
	BSG-NaOH	alkaline	10% v/v NaOH, 4 h
	BSG-TNaOH	thermal-alkaline	10% v/v NaOH, 80 °C, 8 h
3	Potato powder (PW)	not used
4	Molasses (M)	not used

).

For AP and BSG, the following treatments were used: high temperature (125 °C, 15 min), acid–temperature treatment (10% v/v H₃PO₄; 80 °C, 4 h), alkaline treatment (10% NaOH, 4 h), and alkaline–temperature treatment (10% v/v NaOH; 80 °C, 4 h). The postreaction mixtures were cooled to room temperature, then neutralized with a 20% v/v NaOH or 20% v/v H₃PO₄ solution to a pH of 7–7.5.

The saccharification process was conducted using enzymes at 45 °C and pH of 5.5 on a rotary shaker (130 rpm) for 8 h. The doses of enzymes (0.1 cm³/50 cm³ medium) were established in our previous studies [25]. The commercial enzyme preparations Viscozyme^® L and UltraFlo^® Max were suspended in plain warm water to achieve a dry matter concentration of approximately 12% (w/v). The resulting homogeneous mixtures contained enzymes essential for most saccharification processes, including cellulase, xylanase, pectinase, and invertase.

2.4. Batch Experiments

Figure 1 and Figure 2 present the concept of pretreatment and experimental design.

Batch fermentation experiments were conducted in 1 dm3 glass reactors, each with a working volume of 500 cm³. To monitor daily biogas production, each reactor was connected to a 1 dm3 gas collection vessel, employing the water displacement method. The reactors were initially loaded with 150 g of inoculum, and industrial waste substrates were subsequently added to achieve an inoculum-to-substrate (X0/S0) ratio of 1:8, based on volatile solids content. This ratio, previously identified as optimal for DF processes, contrasts markedly with the more commonly applied X0/S0 ratio of 2:1 used in biomethane potential (BMP) assays [26,27]. To inhibit the activity of hydrogenotrophic microorganisms, particularly methanogens, the inoculum–substrate mixtures were submitted to a preliminary thermal pretreatment at 80 °C for 1.5 h. The pH of the mixtures was then adjusted to 5.5. Before sealing, nitrogen gas was flushed through the headspace of each reactor for 3 min to ensure anaerobic conditions. The reactor cultures were incubated at 35 °C in a thermostatically controlled chamber to maintain temperature conditions, with manual agitation performed once daily. Each fermentation trial was continued until biogas production ceased or only trace amounts were observed (duration time 5 days).

2.5. Analytical Methods

Total solids (TS), volatile solids (VS), and pH level were determined according to standard procedures outlined by the American Public Health Association (APHA), the American Water Works Association (AWWA), and the Water Environment Federation (WEF), which were used in our previous studies [26]. Measurements of chemical oxygen demand (COD; LCK 514), total ammonium nitrogen (TAN; No. 8038), orthophosphates (No. 8048), total iron (No. 8008), and total volatile fatty acids (TVFA; LCK 365) were conducted using HACH-Lange test kits, as described in previous studies [26,27].

2.5.1. Biogas Composition

The proportions of hydrogen and methane in the produced biogas were measured under ambient pressure using a GA-21plus gas analyzer (Madur, Zgierz, Poland), equipped with electrochemical sensors for O₂, CO₂, CH₄, H₂, and H₂S, in accordance with the manufacturer’s protocol. Hydrogen production was quantified based on its concentration in the biogas and the corresponding daily gas volume, with values normalized to standard temperature conditions (25 °C) and adjusted for water vapor pressure. To ensure comparability with data from other studies, hydrogen yield was expressed relative to the amount of volatile solids introduced into the reactor.

2.5.2. Volatile Fatty Acids

The liquid fraction obtained after fermentation of industrial waste was also analyzed for volatile fatty acids (VFAs) and residual sugars. Individual VFAs including formic, acetic, propionic, butyric, valeric, caproic, and heptanoic acids were quantified using a gas chromatograph (model 7890A, Agilent, Santa Clara, CA, USA) equipped with a flame ionization detector (FID) and a DB-FFAP capillary column (30 m × 0.32 mm, 0.25 µm film thickness, Agilent Technologies, Santa Clara, CA, USA) containing nitroterephthalic acid-modified polyethylene glycol. Helium served as the carrier gas. The general run parameters were as follows: injector, 220 °C, split 1:10; FID, 220 °C, H2 30 cm³/min, air 400 cm³/min, Makeup He 20 cm³/min; column, initial 100 °C initial, 1.5 min hold, 8 °C/min, 200 °C final 4.0 min hold; carrier gas He (1.2 cm³/min).

2.5.3. Carbohydrates

The monosaccharide composition of the sugar beet pulp hydrolysates was analyzed using a UV spectrophotometer (Multiskan GO, Thermo Fisher Scientific, Munich, Germany) in combination with specific enzymatic assay kits provided by Megazyme Ltd. (Bray, Ireland). The following kits were employed: K-XYLOSE for D-xylose, K-URONIC for D-glucuronic and D-galacturonic acids, K-RAFGA for raffinose, K-MANGL for D-glucose, D-mannose, and D-fructose, K-ARGA for L-arabinose and D-galactose, and K-RHAMNOSE for L-rhamnose. All assays were performed according to the manufacturer’s protocols.

2.5.4. Protein

The Kjeldahl method was employed to assess the increase in protein content within the fermented biomass. Biomass was first separated from the post-culture liquid via centrifugation. A 1 g portion of the analytical sample was transferred to a digestion flask and treated with 15 cm³ of concentrated sulfuric acid. The mixture was then digested at 550 °C using a SpeedDigester K-425 (Büchi, Postfach, Switzerland) until a clear solution was obtained. Subsequent analysis was performed using a KjelFlex K-360 system. The digest was diluted with 40 cm³ of distilled water, neutralized with 60 cm³ of 30% NaOH, and then steam-distilled into 40 cm³ of 2% boric acid solution. The distillate was titrated with 0.1 mol/dm³ hydrochloric acid using a TitroLine^® 5000 titrator (SI Analytics, Weilheim, Germany). Non-fermented and non-hydrolyzed samples served as controls. The contribution of exogenous enzyme-derived protein was excluded from the final calculations.

2.6. Calculations

All the biological and analytical tests were performed in triplicate. The means and standard deviation values, along with error bars, were calculated using Microsoft Excel 2010 R version 3.5.0.

3. Results and Discussion

3.1. Chemical Characteristics of Waste Materials and Their Susceptibility to Pretreatment

Sustainable waste treatment strategies prioritize the conversion of biomass and organic waste, offering a more efficient approach to waste management [9,19]. From the perspective of a circular economy, waste pretreatment methods should also be characterized by low harmfulness to the environment. Depending on the specific properties of waste materials, such as moisture content, volatile organic matter, and ash content, various conversion technologies can be employed. These include physicochemical methods, biochemical processes (e.g., anaerobic digestion), or thermochemical procedures (e.g., pyrolysis, hydrothermal conversion). However, no single technology can solve all the challenges related to biomass waste treatment. Therefore, an integrated approach that combines multiple methods and technologies is essential for effective and sustainable biomass waste management. The type of waste being treated plays a crucial role in selecting the appropriate pretreatment method. For lignocellulosic materials, the primary challenge is their structure. Comprising cellulose, hemicellulose, and lignin, these materials require effective pretreatment strategies to break down their complex polymer matrix and enhance sugar availability [2,12,20].

In this research, we utilized a wide range of waste materials, including substrates with relatively homogeneous carbon sources (molasses—saccharose, potato powder—starch) and materials containing various substrates, most often polymers, that required appropriate pretreatment procedures (apple pomace, brewer’s grains). The differences in chemical composition between the assessed substrates determined the need for pretreatment and the type of pretreatment necessary to make these raw materials suitable for DF and hydrogen production (

Table 2. Characteristics of waste substrates.

Substrate	Chemical Oxygen Demand [g O2/kg]	Carbohydrates [%]	Protein [%]	Ammonia Nitrogen [mg/g]	Total Nitrogen [mg/g]	Ortho- Phosphates [mg/g]	Total Phosphorus [mg/g]	Iron [mg/g]
AP	244.47 ± 3.25	18.00 ± 0.25	5.20 ± 0.08	9.14 ± 0.60	7.53 ± 0.25	1.78 ± 0.04	0.58 ± 0.02	24.7 ± 0.02
BSG	401.70 ± 5.21	5.02 ± 0.12	24.00 ± 0.10	48.21 ± 0.52	39.70 ± 0.34	10.15 ± 0.05	3.31 ± 0.05	0.14 ± 0.01
PW	209.58 ± 1.08	20.00 ± 0.15	8.00 ± 0.35	3.87 ± 0.21	3.19 ± 0.12	0.80 ± 0.02	0.26 ± 0.08	0.02 ± 0.00
M	181.26 ± 0.85	75.00 ± 0.22	0.00 ± 0.00	9.15 ± 0.35	7.54 ± 0.32	1.24 ± 0.04	0.40 ± 0.06	0.05 ± 0.01

).

Apple pomace (AP) has a moderate carbohydrate content (18.80% DM) and a low COD (244.47 g O₂/kg), making it a promising substrate that requires little pretreatment. However, enzymatic hydrolysis of this raw material can boost sugar availability in fermentation processes. Brewer’s grains (BSG) contain a high protein level (48.21% DM) and the highest COD (401.70 g O₂/kg), though their low carbohydrate content (5.02% DM) indicates a need for pretreatment. Potato waste (PW) exhibits high COD (209.58 g O₂/kg) and carbohydrate levels (20.00% DM), suggesting it is suitable for DF processes. Molasses (M), with a very high carbohydrate content (75%) and COD (181.26 g O₂/kg), is almost ideal for hydrogen production. Still, their viscosity and high sugar levels require dilution and pH adjustment for effective fermentation. Overall, pretreatment methods should be tailored to match the unique chemical composition of each substrate. Adding nitrogen sources, pH correction, or enzymatic hydrolysis may be necessary for optimal DF and hydrogen production with microorganisms.

Raw BSG has a low reducing sugar concentration, making it less suitable for hydrogen fermentation. This waste material exhibits a dense, compact, and irregular morphology, characterized by a heterogeneous texture with barely visible fiber separation. This structural complexity challenges microbial accessibility during bioconversion processes such as DF. Among the various pretreatment strategies, alkali and acid treatments induce the most appreciable structural disruption (Figure 3). These chemical pretreatments effectively break down the lignocellulosic matrix, enhancing porosity and fiber separation, which significantly improves substrate attainability.

BSG after acid/alkaline treatment inhibited the highest reducing sugar concentration, indicating a high potential for hydrogen production. This degradation facilitates the release of fermentable sugars, primarily glucose and xylose, from cellulose and hemicellulose fractions, thereby increasing the availability of carbon sources for hydrogen-producing microorganisms [28]. Enzymatic and thermal treatments also promote structural loosening, although to a lesser extent. Enzymatic hydrolysis selectively targets polysaccharide components, enhancing sugar liberation without extensive matrix disruption, while thermal methods primarily loosen the biomass structure and partially degrade hemicellulose. Although less aggressive, these treatments contribute to improved fermentability and nutrient accessibility, supporting hydrogen yield in DF systems. In terms of the released reducing sugars, enzymatic hydrolysis proved to be the most effective pretreatment method, yielding nearly twice the concentration of sugars in the hydrolysate compared to the untreated control sample (Table 2).

Raw apple pomace has a dense and compact structure with intact cell walls. Its texture is irregular and heterogeneous, indicating limited accessibility for microbial or enzymatic attack (Figure 4).

From a biochemical and process engineering perspective, apple pomace (AP) appears to be the most suitable substrate for DF and biohydrogen production, due to its high carbohydrate content (primary fermentable substrate), low protein and nitrogen levels (reduced risk of ammonia inhibition), and adequate micronutrient content (P, Fe). The hydrolysates (especially AP-ME) show increased reducing sugar content despite lower TS and VS, indicating enhanced fermentability. AP-ME is the most promising hydrolysate for hydrogen fermentation, due to its highest reducing sugar concentration, which is critical for microbial hydrogen production.

The mixture of enzymes resulted in the most significant structural degradation of apple pomace, followed by single-enzyme and thermal treatments. These changes are crucial for improving the efficiency of DF by increasing the bioavailability of fermentable sugars. Synergistic enzyme action leads to extensive breakdown of cell wall components, maximizing substrate availability for fermentation. Numerous studies have highlighted the critical importance of pretreatment conditions in maximizing sugar recovery from lignocellulosic biomass while minimizing the formation of microbial inhibitors such as organic acids (mainly acetic acid, formic acid, and levulinic acid), furan derivatives, and phenolic compounds (primarily coumaric acid, syringaldehyde, and vanillin) [29,30]. These compounds are typically generated during the acid or alkaline hydrolysis of lignocellulosic biomass, when pentose and hexose sugars undergo dehydration reactions under elevated temperatures and acidic or basic conditions. Furfural (F) is primarily derived from the degradation of pentoses such as xylose, while 5-hydroxymethylfurfural (HMF) originates from hexoses like glucose and fructose [31]. Both compounds are known to exert cytotoxic effects on hydrogen-producing microorganisms by disrupting cellular membranes, inhibiting key metabolic enzymes, and interfering with redox balance, ultimately reducing hydrogen yields during DF. However, as demonstrated in numerous studies [31,32], the inhibitory effects on hydrogen production and the specific concentrations of inhibitory compounds responsible for this effect can vary considerably, depending on the type of substrate and the microbial consortium employed.

The metabolic response to inhibitors such as F, HMF, and phenolic compounds is highly strain-dependent. For instance, furfural concentrations as low as 0.5–1.0 g/dm³ have been shown to significantly reduce hydrogen yields in DF systems, while HMF exhibits similar inhibitory effects at concentrations above 1.0 g/dm³. Phenolic compounds such as vanillin and syringaldehyde can also impair microbial activity at concentrations exceeding 0.5 g/dm³ [33,34]. Pure bacterial cultures, such as Clostridium butyricum or Enterobacter aerogenes, often exhibit lower tolerance thresholds due to their limited metabolic flexibility and lack of synergistic interactions [35]. In contrast, mixed microbial consortia—particularly those derived from anaerobic sludge, compost, or manure—tend to display greater resilience. This enhanced tolerance is attributed to microbial diversity, functional redundancy, and the presence of detoxifying species that can transform or assimilate inhibitory compounds. The origin of the inoculum also plays a critical role, as microbial communities pre-adapted to lignocellulosic hydrolysates or industrial waste streams may exhibit increased resistance to inhibitory stress. Therefore, understanding the interplay between microbial composition, substrate characteristics, and inhibitor profiles is essential for optimizing DF processes and ensuring stable hydrogen yields under variable feedstock conditions.

Molasses, a viscous by-product derived from the sugar refining industry, is a cost-effective and widely available substrate for biohydrogen production. Its use supports waste valorization and the circular economy, contributing to sustainable and renewable energy systems [36]. Chemically, molasses is characterized by a high content of fermentable sugars, predominantly sucrose, glucose, and fructose. These monosaccharides and disaccharides are easily assimilated by hydrogen-producing microorganisms, particularly strains of Clostridium spp. and Enterobacter spp., which are known for their strong fermentative abilities. The high concentration of reducing sugars—about 30.3 g/kg—highlights molasses’s excellent biodegradability and fermentability, making it an ideal feedstock for DF. During microbial metabolism, these sugars are converted into hydrogen and organic acids through enzymatic pathways, facilitating efficient biohydrogen generation [37]. Thus, molasses can serve as a valuable carbon source, enhancing the economic and environmental viability of biological hydrogen production technologies.

Potato powder can also serve as an interesting substrate for DF. It contains almost entirely starch, which can be converted into simple sugars by suitable amylases. Many bacterial and fungal strains possess strong amylolytic properties, rendering pretreatment unnecessary and making this waste material highly valuable for sustainable technologies [38,39,40].

The initial physicochemical characteristics of the substrates are presented in Table 2, including chemical oxygen demand (COD), carbohydrate and protein content, nitrogen and phosphorus levels, and iron concentration. In turn, the efficiency of each pretreatment method in terms of releasing fermentable sugars is summarized in

Table 3. Hydrolysate characteristics.

Substrate	Total Solids [g/kg]	Volatile Solids [g/kg]	Reducing Sugars [g/kg]
AP (raw)	161.94 ± 1.11	159.57 ± 0.97	14.10 ± 0.65
AP-E	98.24 ± 1.63	94.22 ± 1.60	18.90 ± 0.75
AP-ME	97.18 ± 1.45	93.11 ± 1.08	21.20 ± 0.84
AP-T	135.73 ± 1.97	133.65 ± 1.98	16.20 ± 0.25
BSG (raw)	215.75 ± 2.57	196.56 ± 1.53	1.40 ± 0.08
BSP-E	162.47 ± 1.85	150.45 ± 1.22	2.50 ± 0.12
BSG-NaOH	160.14 ± 2.54	109.76 ± 1.67	1.40 ± 0.09
BSG-H3PO4	209.33 ± 2.42	108.85 ± 2.95	1.60 ± 0.07
BSG-T	165.42 ± 2.14	148.62 ± 1.76	1.90 ± 0.08
BSG-TH3PO4	208.02 ± 2.56	107.15 ± 1.85	2.00 ± 0.11
BSG-TNaOH	159.25 ± 2.87	108.25 ± 1.69	1.50 ± 0.09
PW *	449.33 ± 2.65	442.07 ± 3.12	9.00 ± 0.21
M *	854.31 ± 3.28	760.22 ± 2.95	30.30 ± 1.05

* substrates not hydrolyzed.

.

It is important to note that M and PW were not subjected to hydrolysis. Nonetheless, the content of reducing sugars was significantly higher in comparison to the hydrolysates of spent grain after pretreatment. In turn, AP treatment, especially enzyme treatment, resulted in almost 50% increase in sugar level compared to the raw material.

3.2. Batch Experiments

This study employed a lower inoculum-to-substrate ratio of 1:8 (based on volatile solids content) than the conventional 2:1 ratio typically used in batch digestion assays [26]. According to Toledo-Alarcon et al. [41], substantial variability in inoculum characteristics, including total solids (TS), volatile solids (VS), and microbial composition, may be observed across various sources. The species influences the efficiency of hydrogen production by microorganisms, the growth substrate used, and the conditions under which they are cultured. Published studies have identified Clostridium sp. and Bacillus spp. as key contributors [42,43]. Due to the limited metabolic capabilities of individual bacterial strains, the application of pure cultures in fermenting complex substrates remains constrained. In this study, the microbial consortia demonstrated clear advantages, with mixed cultures enriched in Clostridium and Bacillus leading to a three-fold enhancement in hydrogen production.

Iron plays a crucial role as a cofactor in hydrogenase and other enzymes essential for biohydrogen production. Additionally, it serves as an efficient heavy metal for sulfide mitigation in anaerobic digestion through precipitation mechanisms. Supplementation of the substrate with Fe2+ has been shown to stimulate the activity of hydrogenase, an enzyme responsible for oxidizing reduced ferredoxin and subsequently generating molecular hydrogen [44]. The positive effect of iron on increasing the efficiency of DF was confirmed in our previous studies [26,27], where three compounds containing this element were tested. A significant increase in hydrogen content in biogas was observed when iron (III) oxide was used. Due to its relatively high iron concentration (24.7 mg/g), apple pomace may significantly enhance hydrogen production efficiency when utilized as a substrate. Furthermore, its relatively elevated carbohydrate levels and low nitrogen content present additional advantages for biohydrogen generation processes.

3.2.1. Hydrogen and Methane Production

The heat map shown in Figure 5 presents hydrogen and methane production levels (in dm³/ton) across various experimental runs, each corresponding to different treatment conditions. The color gradient ranges from black (indicating low production) to yellow (indicating the highest observed production, up to 900 dm³/t). Hydrogen production is visibly higher and more variable across the experimental conditions than methane, showing minimal production in all runs. The most significant hydrogen yields are observed in the AP and AP-ME treatments, indicated by the bright yellow coloration. Among the tested substrates, the maximum hydrogen yield over 135 cm³/g VS was reported for molasses. Slightly lower results were obtained for apple pomace: AP-ME (112.69 ± 6.97 Ncm³/g VS), and AP (101.06 ± 0.00 Ncm³/g VS). Furthermore, the hydrogen content in biogas from molasses exceeded 63%, which was the highest among all reported values in these experiments. In contrast, methane yields remained negligible across all substrates, yielding 0.00 ± 0.00 N cm³/g VS for molasses and raw and enzymatically treated apple pomace (AP, AP-ME). These data underscore the importance of substrate selection and process conditions in enhancing hydrogen production while minimizing methane formation, thereby improving the overall efficiency of biohydrogen generation via DF.

3.2.2. Biogas Composition and Biogas Yield

The results of the fermentation experiments, including total and peak hydrogen production, gas composition, and overall biogas yield, are presented in

Table 4. Operating parameters and performance of the batch tests.

Substrate	SGP [Ncm3/g VS]	SHP [Ncm3/g VS]	SMP [Ncm3/g VS]	Average H2 [%]	Average CH4 [%]	Average CO2 [%]	pH Initial	pH Final	Total VFA [g/dm3]
BSG	1916.67 ± 20.45	59.16 ± 1.32	0.96 ± 0.08	59.36 ± 1.48	0.92 ± 0.15	38.73 ± 1.38	5.62	5.30	1.94
BSG-E	2208.33 ± 25.67	60.09 ± 1.58	6.33 ± 0.15	51.43 ± 1.08	5.21 ± 0.24	36.56 ± 1.21	5.49	5.08	1.40
BSG-H3PO4	621.56 ± 1.93	18.56 ± 0.48	0.85 ± 0.04	50.87 ± 1.02	2.32 ± 0.12	25.60 ± 1.12	6.10	5.81	2.12
BSG-TH3PO4	791.67 ± 2.02	22.38 ± 1.45	1.55 ± 0.25	53.82 ± 0.98	3.94 ± 0.28	29.54 ± 1.31	5.80	5.41	1.40
BSG-T	1062.50 ± 15.67	29.36 ± 1.62	2.16 ± 0.35	52.76 ± 0.95	3.87 ± 0.32	33.46 ± 1.43	5.61	5.33	1.88
BSG-NaOH	312.89 ± 0.93	2.54 ± 0.15	0.54 ± 0.06	21.41 ± 0.25	0.54 ± 0.02	18.69 ± 0.16	5.98	5.48	0.42
BSG-TNaOH	468.33 ± 1.05	1.25 ± 0.28	0.00 ± 0.00	5.21 ± 0.45	0.00 ± 0.00	14.98 ± 0.45	5.85	5.51	1.05
M	4083.33 ± 24.94	135.31 ± 2.34	0.00 ± 0.00	63.09 ± 1.12	0.00 ± 0.00	31.58 ± 1.08	5.64	4.46	1.41
PW	1208.33 ± 18.65	62.95 ± 1.82	0.00 ± 0.00	47.76 ± 0.89	0.00 ± 0.00	30.56 ± 1.13	5.87	5.46	0.71
AP	3333.33 ± 21.85	101.06 ± 2.05	0.00 ± 0.00	57.73 ± 1.43	0.00 ± 0.00	29.85 ± 1.56	5.56	5.28	1.32
AP-ME	3593.33 ± 17.39	112.69 ± 2.06	0.00 ± 0.00	59.87 ± 1.23	0.00 ± 0.00	30.14 ± 1.56	5.68	5.22	1.02
AP-E	3083.33 ± 19.30	74.99 ± 1.54	0.61 ± 0.01	53.54 ± 1.18	0.00 ± 0.00	38.01 ± 1.28	5.52	5.04	1.63
AP-T	3000.00 ± 15.87	39.35 ± 1.04	0.00 ± 0.00	56.41 ± 1.45	0.00 ± 0.00	25.46 ± 1.37	5.56	5.11	4.96

, enabling a comparative evaluation of process performance across the different substrate pretreatment systems.

Table 4 also shows the effect of the chemical and thermochemical pretreatments on hydrogen production in batch experiments using BSG. Generally, all chemicals added to BSG decreased hydrogen production. Among the tested conditions, hydrolysis in the presence of a sodium base resulted in the lowest hydrogen yield, indicating it has the most detrimental effect on the overall reaction efficiency. In the control experiment, hydrogen production reached 59 cm³, serving as the reference point for evaluating the impact of different treatment conditions. In contrast, the sample subjected to alkaline hydrolysis yielded only 2.45 cm³ of hydrogen. This represents a reduction of approximately 95.8%, indicating a severe inhibitory effect of the sodium-based alkaline treatment on hydrogen generation. The reduced yield may be attributed to the inhibition of furfural, HMF, or aromatic compounds such as vanillin [45]. Lee et al. [46] noticed that high sodium concentrations caused Clostridium butyricum to allocate a larger portion of its fermentation-derived ATP toward cellular maintenance rather than biomass synthesis. Moreover, the presence of high Na+ levels redirected the metabolic pathway, favoring acetate production over butyrate formation, which resulted in a decrease in hydrogen production. Enzymatic hydrolysis using a cellulolytic enzyme resulted in only a modest increase in hydrogen production (600 mL H₂/dm³) compared to the control (590 mL H₂/dm³). However, reports in the literature [47] suggest that enzymatic pretreatment can enhance hydrogen yields from BSG by up to 4160 mL H₂/dm³ of the working volume.

It is essential to note that these results were obtained on a larger scale and under optimized conditions, particularly at pH 7.5, which was identified by Soares et al. [47] as the most favorable for their process. This comparison underscores the crucial importance of tailoring both the hydrolysis and DF conditions to the specific substrate being used.

Our previous studies involving beet pulp, fruit and vegetable waste, and corn silage showed that optimal pH values vary significantly depending on the substrate [48]. Therefore, applying a universal pH or process configuration may not be effective across different feedstocks. The relatively low hydrogen yields in the present study may be attributed to optimal pH conditions or enzyme-substrate incompatibility. Recent studies have highlighted that the optimal pH for DF of BSG varies depending on microbial consortia and pretreatment methods. Acidic conditions with a pH of around 5.0 seem to be favorable for hydrogen-producing bacteria, particularly Clostridium species [49]. However, excessive protein degradation during fermentation can lead to elevated ammonia concentrations (>300 mg N–NH₃/dm³), which significantly inhibit hydrogen production [50]. Future work should focus on fine-tuning enzymatic hydrolysis parameters, such as enzyme concentration, incubation time, and pH, as well as fermentation conditions, to maximize biohydrogen production for each specific substrate. Additionally, scaling up the process under optimized conditions may reveal more pronounced effects of enzymatic pretreatment. The initial pH value for the hydrogen digesters ranged from 5.49 to 6.10. The average pH value decreased in each run during the experimental period, regardless of the type of treated hydrolysates. The pH values decreased below 5.5, which could be linked to the release of volatile fatty acids (Table 4). A pH below 5.5 during DF can significantly impact the efficiency of hydrogen production. Under anaerobic conditions, fermenting microorganisms degrade organic substrates, leading to the formation of VFA such as acetic acid, propionic acid, and butyric acid. Their accumulation causes acidification, which explains the observed decrease in pH. The optimal pH range for hydrogen-producing microorganisms, such as Clostridium bacteria, is typically 5.5–6.5. When the pH drops below 5.0, enzymatic activity responsible for hydrogen production (e.g., hydrogenase) can be inhibited, reducing process efficiency. Furthermore, lower pH favors the growth of lactic acid bacteria, which compete for substrate but do not produce hydrogen, negatively impacting the final gas balance.

Previous studies have identified archaea in the digestate as belonging to the genus Methanosphaera, which is known for its ability to tolerate acidic environments and remain metabolically active at a pH as low as 5 [26,51]. The detection of methanogenic activity suggests that the thermal pretreatment applied to the inoculum may not have been entirely effective in deactivating these microorganisms. This observation aligns with findings reported by Venkata Mohan et al. [52], who also noted the persistence of methanogens following thermal treatment. The survival of these archaea could be attributed to their inherent resistance mechanisms or protective microenvironments within the inoculum matrix. Moreover, the continued activity of methanogens under suboptimal conditions may have implications for the stability and efficiency of anaerobic digestion processes, particularly in systems where acidic conditions prevail.

3.2.3. Fermentation Dynamics

To assess the effectiveness of the proposed pretreatment method for biohydrogen production, batch fermentation trials were conducted, during which both the volume and composition of the generated biogas were monitored, with particular emphasis on hydrogen yield. Hydrogen concentrations were quantified using a gas analyzer. The results were normalized to standard temperature and pressure conditions and expressed per gram of volatile solids (VS) introduced into the system. The performance of each substrate and pretreatment variant was assessed by comparing hydrogen yields, residual sugar concentrations, and volatile fatty acid (VFA) profiles in the post-fermentation biomass. These parameters provided a comprehensive assessment of substrate biodegradability, microbial conversion efficiency, and the overall optimization potential of the fermentation process. During the experiments, the biogas volumes and composition were measured daily, and the daily values were plotted to obtain cumulative curves. The highest biogas production was observed during the first two days of fermentation, reaching a maximum after 24 h, and then practically ceased after 3 days for the hydrogen experiments (Figure 6). During fermentative hydrogen production, a diverse array of liquid metabolites is generated, with volatile fatty acids (VFAs) being among the most prominent.

3.2.4. Total Volatile Acids

Figure 7 illustrates the concentrations of individual VFAs detected in digestates at the point of peak hydrogen yield, identifying eight organic acids, including acetic, propionic, butyric, and caproic acids. Acetic and butyric acids were the dominant products in the DF of apple pomace, while caproic acid was notably abundant in the brewer’s spent grain (BSG) digestate. These results align with the metabolic pathways of hydrogen-producing bacteria, particularly those favoring butyrate-type fermentation, which is known to enhance hydrogen yields [53]. The VFA profile reflects microbial activity and substrate composition, influencing both pH and process stability [54]. The highest total VFA concentration was recorded for the AP-T variant (4.96 g/dm³), accompanied by a significant decrease in pH to 5.11 from an initial value of 5.56. Although VFAs are essential intermediates, their accumulation can inhibit hydrogen production by lowering pH below the optimal range (5.5–6.0), thereby favoring hydrogen-consuming organisms.

Substrate composition played a critical role in shaping the VFA profile. The sugar-rich substrates, including molasses and apple pomace, yielded simpler acid spectra dominated by short-chain VFAs. In contrast, for structurally complex substrates, such as BSG, a broader range of acids was produced, including medium-chain VFAs, such as caproic and heptanoic acids. Chemically pretreated BSG variants (e.g., NaOH and H₃PO₄ hydrolysis) showed enhanced acid production. These findings highlight the significance of these factors in determining fermentation outcomes. This has implications for optimizing VFA production for downstream applications, including bioenergy, bioplastics, and microbial chain elongation. These findings align with previous studies that emphasize the importance of substrate accessibility and pretreatment in optimizing volatile fatty acid (VFA) production. For instance, Motte et al. [55] demonstrated that fine milling of lignocellulosic substrates significantly increased acid production due to improved bioavailability of soluble compounds. However, it also raised the risk of acidification. Similarly, Bi et al. [56] highlighted that substrate composition and enzymatic degradation pathways strongly influence microbial community dynamics and metabolite profiles in anaerobic digestion systems. Recent studies support these findings, demonstrating that microwave-assisted pretreatment enhances the predictability and efficiency of hydrogen and volatile fatty acid production, particularly butyric acid, which reached concentrations up to 24.01 g/L under optimized conditions [57]. Moreover, hydrothermal pretreatment has been shown to significantly improve organic solubilization and VFA yields from lignocellulosic substrates, with acetic acid remaining the dominant product [58]. Our results reinforce the notion that targeted pretreatment strategies can enhance hydrogen fermentation efficiency and tailor the VFA profile for specific biotechnological applications.

3.2.5. Energy Yield

Table 5. Energy yield and hydrogen/methane production.

Substrate	SHP [Ncm3/gVS]	Energy [kJ/kgVS]	Energy [kWh/kgVS]	SMP [Ncm3/gVS]	Energy [kJ/kgVS]	Energy [kWh/kgVS]	Σ Energy
BSG	59.76	763.01	0.212	0.96	38.14	0.01	0.22
BSG-E	59.09	754.45	0.210	6.33	251.5	0.07	0.28
BSG-H3PO4	18.56	236.97	0.066	0.85	33.77	0.01	0.08
BSG-TH3PO4	22.38	285.74	0.079	1.55	61.58	1.71	1.79
BSG-T	29.36	374.86	0.104	2.16	85.82	0.02	0.13
BSG-NaOH	2.54	32.43	0.009	0.54	21.46	0.01	0.01
BSG-TNaOH	1.25	15.96	0.004	0.00	0.00	0.00	0.00
M	135.31	1727.62	0.480	0.00	0.00	0.00	0.48
PW	62.95	803.74	0.223	0.00	0.00	0.00	0.22
AP	101.06	1290.32	0.358	0.00	0.00	0.00	0.36
AP-ME	112.69	1438.81	0.400	0.00	0.00	0.00	0.40
AP-E	74.99	957.46	0.266	0.61	24.24	0.01	0.27
AP-T	39.35	502.42	0.140	0.00	0.00	0.00	0.14

presents the specific hydrogen and methane production (SHP and SMP) and corresponding energy yields for a range of organic substrates, including brewer’s spent grain (BSG), apple pomace (AP) in various pretreated forms, molasses (M), and potato waste (PW). The energy yield from hydrogen production, expressed in kJ/kg VS, ranges from approximately 16 to over 1700 kJ/kg VS, depending on the substrate and pretreatment method, while the methane energy yield, expressed in kJ/kg VS, varies between 0.0 and 250. In terms of hydrogen production, the best substrates were molasses and apple pomace treated with a mixture of enzymes. However, the same raw material, after thermal and enzymatic treatment, can be characterized as a poorer source of biohydrogen. Interesting results were obtained for brewer’s grains. It turned out that improper treatment of this waste material can significantly reduce hydrogen production but slightly improve methane yield.

Notably, most substrates subjected to thermal or chemical pretreatment (e.g., BSG-TNaOH, and AP-T) tended to exhibit weakened energy recovery. These results differ from those reported in the literature, where microwave and acid pretreatments have been shown to increase biomethane yields by 4–7 times [59]. The theoretical energy content of methane is approximately 890 kJ/mol, and for hydrogen, around 286 kJ/mol, which corresponds well with the observed values in the table when normalized per unit of volatile solids. According to Obiora et al. [60], the actual energy recovery depends heavily on substrate composition, retention time, and digester configuration.

4. Conclusions

The purpose of this study was to investigate the impact of various pretreatment methods on the effectiveness of dark fermentation for the production of biohydrogen from industrial waste, including apple pomace, brewery grains, molasses, and potato powder. Dark fermentation is a promising approach for the efficient production of hydrogen from agricultural waste. It also strengthens waste management by diverting organic waste from landfills, reducing methane emissions, and lowering disposal costs. Biohydrogen production from waste offers a sustainable solution for waste management and the generation of clean energy. Integrating biohydrogen into a circular economy maximizes resource utilization, reduces environmental impact, and promotes sustainable energy. Biohydrogen production and efficient waste management could significantly support the world’s transition to a sustainable, circular energy economy.

This research emphasizes the importance of selecting the most suitable pretreatment methods for producing biohydrogen from agricultural waste. A well-designed strategy enables significant improvements in biohydrogen production, whereas choosing poorly designed methods and parameters can lead to substantial decreases in biohydrogen or biomethane yields. The results confirmed the stability and usefulness of activated sludge as a source of microorganisms for hydrogen production under dark fermentation conditions. Among the substrates subjected to pretreatment, the highest yield of hydrogen was obtained from apple pomace after hydrolysis by mixed enzymes. Molasses, a less complex substrate compared to other raw materials, produced a higher hydrogen yield. The key factor for success, apart from the proper method of waste material pretreatment, was the appropriate preparation of the inoculum, which limited competition from methanogens.

Future research should focus on improving pretreatment techniques and even developing hybrid strategies that integrate biological, thermochemical, and electrochemical processes to optimize biohydrogen technology.