Removal of hydrogen sulfide (H₂S) with Thiobacillus denitrificans biofilter: study of the microbial community conducted through 16S rRNA sequencing analysis

Abstract

Hydrogen sulfide (H₂S) emissions from oil and gas operations, sewage treatment facilities, and landfills are challenges to the quality of life. The main objective of this work is to study the effect of different filter packing materials on Thiobacillus denitrificans-mediated lab-scale bioreactor for H₂S removal using 16S rRNA metagenomic sequencing. In this study, the bioreactor column, which has three distinct layers of Ceramic Ball Filter Media (CBFM), Filter Media Ceramic Rings (FMCR), and Filter Bio Balls (FBB), was designed and operated for 60 days. The microbial community samples adhered to the surfaces of the filling materials were investigated using 16S rRNA metagenomic sequencing with paired-end 2 × 150 base reads (Illumina). The results showed that the H₂S gas removal efficiency reached its maximum of 99% by the end of the seventh day, followed by a steady-state pattern. Compared with polypropylene surfaces, ceramic materials successfully hosted the T. denitrificans bacteria. Changing the filter material altered the phylum species diversity of the microcosms on the filter material, as shown by alpha diversity indices (Shannon and Simpson values). The dominant Phylum across all samples, regardless of the treatment and filter material type, was Proteobacteria, followed by Firmicutes and Bacteroidetes. Comamonas thiooxydans, Comamonas testosterone, Enterococcus faecalis, Staphylococcus aureus, Staphylococcus cohnii, and Mesorhizobium terrae are the most abundant species detected on the filter materials. In addition, changing the filter material causes a substantial alteration of the dominant species in the microcosm of the bioreactor. These findings highlight the critical role of filter material in supporting H₂S-removing microorganisms.

Introduction

H₂S is a highly toxic, odorous gas commonly produced throughout many physical and biological systems. Inhalation of H₂S concentrations above ten ppm can result in headaches and lung irritation, whereas 100 ppm and above could result in loss of olfactory perception and may cause damage to the lungs and a threat to life and health. In addition to a significant odor nuisance, releasing H₂S into the atmosphere is considered a health concern because of its toxic and corrosive effects. (Valdez et al. 2012). The problems and challenges arising from urban air pollution and waste disposal have affected human health and the environment locally and globally. (Manisalidis et al. 2020). Open dumping areas or sanitary landfills, where municipal solid wastes (MSWs) are disposed of, are among the most significant emission sources of harmful air pollutants. (Siddiqua et al. 2022). During decomposition processes, various harmful and greenhouse gases (GHG) such as methane (CH₄), carbon dioxide (CO₂), hydrogen sulfide (H₂S), ammonia (NH₃), carbon monoxide (CO), volatile organic compounds (VOCs), and odor are vented to the atmosphere (González et al. 2019; Calvin et al. 2023).

Landfills produce massive amounts of methane gas (CH₄) and toxic gases such as hydrogen sulfide (H₂S). Using biogas as a renewable energy source is becoming dominant in avoiding the adverse effects of global warming and local air pollution issues (Liang and Liang 2013; Aita et al. 2016). However, H₂S in biogas has restricted its use in bioenergy production, mainly due to corrosive gases and acids. (Ghimire et al. 2021). H₂S is an odorous and highly toxic gas commonly encountered throughout many physical and biological systems. In addition to a significant odor nuisance, releasing H₂S into the atmosphere is considered a health concern because of its toxic and corrosive effects. (Valdez et al. 2012). The European Union Commission directive 2009/161/E.U. has established 8-h time-weighted average and short-term occupational exposure limit values for H₂S as five ppm and ten ppm, respectively.

Removing H₂S involves many technologies, including bioremediation (Aita et al. 2016). Several bioreactor configurations holding biofilters, scrubbers, and trickling filters have previously been implemented to reduce H₂S emission (Quijano et al. 2018; San-Valero et al. 2019; Dupnock and Deshusses 2020; Mu et al. 2021; Flores-Cortés et al. 2021; Das et al. 2022). To reduce H₂S emissions, several bioreactor configurations containing biofilters, scrubbers, and trickling filters have previously been implemented. Quijano et al. (2018) designed a two-step desulfurization system, combining a continuously aerated absorption column with an airlift bioreactor. The H₂S removal efficiency of 100% at various inlet loading rates (28, 56, and 112 g S-H₂S m⁻³ h⁻¹) was achieved. The liquid phase primarily contained SO₄²⁻ with an average sulfur oxidation performance of ∼80%. The system showed robustness during a 7-day shutdown, recording 100% H₂S removal efficiency 10 h after restarting. Microbial communities characterized by 16S rRNA pyrosequencing demonstrated successful sulfurization of H₂S-rich biogas, maintaining high and stable removal performance. San-Valero et al. (2019) developed a fully aerobic bio scrubber for H₂S-rich biogas desulfurization, combining an absorption column and a bubble-column bioreactor. The bio scrubber achieved stable H₂S removal efficiencies above 80% for 37–100 g/m³ h loading rates. It proved robust H₂S removal performance during short- and long-term shutdowns with quick performance recovery. The aerated bubble-column bioreactor, operating at a slightly alkaline pH, exhibited sulfur oxidation performances exceeding 75%—characterization of bacterial communities identified specialized and efficient H₂S-oxidizing Proteobacteria in the bioreactor. Dupnock and Deshusses (2020) investigated co-treating H₂S and CO₂ in a biological trickling filter system using hydrogenotrophic methanogens and nitrate-reducing bacteria. Thiobacillus denitrificans were introduced to upgrade a biogas mimic to > 97% methane. Severe competition for hydrogen’s electrons between CO₂ and nitrate was observed, necessitating high N:S loading rates (16:1) for H₂S removal above 98%. However, high nitrate loading negatively impacted biogas upgrading. Mu et al. (2021) developed a bio-desulfurizing system using Thioalkalivibrio versutus D301 to treat H₂S in haloalkaliphilic conditions. Optimal performance was observed at a sulfide load of 3.3 g/l/d and an ORP of −400 mV, resulting in 86% elemental sulfur and 9% sulfate selection. Linear relationships were found between sulfide load, ORP, and aeration rate. Spatial and temporal heterogeneity in the bioreactor was confirmed. Operating conditions influenced sulfur particle morphology, with 56.3% of total S^◦ particles collected within two hours by gravity settling. Flores-Cortés et al. (2021) studied a two-stage bioreactor that effectively removed H₂S from biogas with 2500–10,000 ppm concentrations, achieving over 95% removal efficiency. H₂S oxidation ranged from 8.2% to 18.7%, the ratio of formed SO₄²⁻ per gram of H₂S absorbed. The nitrogen balance showed that 2–6% of consumed NO3 was directed to biomass growth. Bacterial community changes did not affect removal efficiency, and clogging issues were avoided by eliminating elemental sulfur accumulation. Chaghouri et al. (2022) examined the efficiency of fungi-based biofilters in removing H₂S. Trichoderma harzianum, isolated from industrial digestate, proved capable of metabolizing H₂S. A bioreactor with cardboard, perlite, woodchips, and pellets as filling was utilized. Over 60 days, H₂S removal was monitored, with 97% sulfur detected in the filling material, primarily as SO₄²⁻.

Another bacterium abundantly adhered to all membrane filters is Enterococcus faecalis. The recent studies showed the homeostatic interaction between endogenous H₂S production and the biogenesis and clearance of reactive sulfur species (RSS) in E. faecalis (Walsh and Giedroc 2020; Walsh et al. 2022). Shen et al. (2018) reported that a RSS-sensing receptor protein from E. faecalis, namely CstR (copper-sensing operon repressor (CsoR)-like sulfurtransferase), regulates a cst-like operon in response to exogenous sulfide stress. These studies point out the existence of H₂S-sensing and metabolizing mechanisms in E. faecalis.

Various H₂S-oxidizing bacteria species have been reported. T. denitrificans is a facultative anaerobe bacterium well known for its capacity to oxidize H₂S forms and reduce iron compounds through denitrification (Beller et al. 2006). T. denitrificans may contribute to sulfur modification in landfills, groundwater, and wastewater treatments (Koenig and Liu 2001; Sierra-Alvarez et al. 2007; Sahinkaya et al. 2011). Therefore, in this study, T. denitrificans were added to the designed Landfill gases (LFG) bioreactor and used for H₂S removal. On the other hand, not only T. denitrificans but also other H₂S-oxidizing bacteria may cooperate in the LFG column to remove H₂S. Metagenome sequencing provides information about the microbial community in the bioreactors (Tombuloglu et al. 2023). In a recent study by our group, metagenome sequencing was successfully used to decipher the microbial community in a bioreactor system (Tombuloglu et al. 2023). An earlier metagenome study has confirmed that a biofilter inoculated with Trichoderma harzianum was crucial in H₂S elimination (Chaghouri et al. 2022). In another study, Das et al. (2022) developed a hollow fiber membrane bioreactor that effectively removed over 99% H₂S and NH₃. Metagenomic sequencing revealed some sulfide oxidizers, including aerobic and anoxic chemolithotrophs, effective for H₂S bioconversion. In our study, the metagenome analysis showed that Proteobacteria, Firmicutes, and Bacteroidetes were the dominating phyla in all samples, irrespective of the filter material type and treatment.

This study will shed light and answer the following questions: What effects might different filter materials have on the stability and resilience of the microbial community involved in H₂S removal? The main objective of this work is to study the effect of different filter media packing materials on microbiome diversity on Thiobacillus denitrificans-mediated H₂S removal in a lab-scale packed bioreactor column with diverse types and structures of filling materials, such as fiber, balls, and rings made of polypropylene or ceramic, using 16S rRNA metagenomic sequencing. In this study, the bioreactor column, which has three distinct layers of Ceramic Ball Filter Media (CBFM), Filter Media Ceramic Rings (FMCR), and Filter Bio Balls (FBB), was designed and operated for 60 days. This study’s novelty is elucidating the 7-day high efficiency (up to 99%) of H₂S removal capacity using four different column materials, and bacterial community analysis verified by the metagenome sequencing technique.

Materials and methods

Bioreactor design and setup

In this study, a biofilter using filter media packing materials is designed (Fig. 1) to test the efficiency of natural ceramic materials as filter bed mediums for removing H₂S gas with Thiobacillus denitrificans bacteria. The influent H₂S gas was introduced from the bottom of the bioreactor throughout V1 (Fig. 1. H₂S gas valve). A standard certified gas cylinder provided the inlet H₂S with a concentration of 200 ppm, which was adjusted to the required mass flow rate using a digital flow meter with varying flow rates between 20 and 60 mL min⁻¹. Initially, an 8 mg day-1 H₂S mass flow rate was tested. Then, it was increased by 4 mg day⁻¹ every 12 days using a mass flow controller to evaluate the H₂S removal rate of the bioreactor system against increasing inlet H₂S mass flow rate. The effluent H₂S was collected from the top of the column (Fig. 1. H₂S gas valve V2), and its concentration in the effluent was analyzed using the MRU™ Optima7® biogas portable gas analyzer. The required moisture content and nutrients for microbial growth are provided daily from the top of the bioreactor through the valve (Fig. 1. H₂S gas valve V2) using peristaltic pumps.

Fig. 1.

Schematic view of the lab-scale bioreactor set up for H₂S removal. The bioreactor contains different filling materials (n = 4) from top to bottom: ceramic ball filter media (CBFM), filter media ceramic rings (FMCR), biofilter fiber (BFF), and filter bio balls (FBB). The bioreactors were connected to an H₂S gas cylinder regulated by a flow controller. Bacterial inoculum and nutrients were delivered using peristaltic pumps

The column bioreactor experiments lasted 60 days, with daily sampling and measuring microbial colony counts, pH, and sulfate levels.

Design considerations include the Inlet H₂S concentrations, gas superficial velocity and flow rate, moisture control, nutrient supply, type and depth of filter bed medium and microorganisms, reactor size, incubation time, temperature, and pH of the filter medium, and Leachate circulation. The operational parameters of biofilters include empty-bed residence time (EBRT), volumetric loading rate (VLR), mass loading rate (MLR), elimination capacity (E.C.), and H₂S removal efficiencies. In this manner, the removal efficiencies of H₂S are calculated daily and statistically treated using SPSS software.

Filter media column preparation

Figure 2 shows the lab column setup with the bioreactor containing three different layers of Ceramic Ball Filter Media (CBFM), Filter Media Ceramic Rings (FMCR), and Filter Bio Balls (FBB), separated with 5 cm internal Biofilter Fiber (BFF).

Fig. 2.

Photograph of the assembled bioreactor, showing the connections to peristaltic pumps for nutrient delivery and sampling ports

The characteristics of filter materials are denoted in Table 1. CBFM is lightweight with high surface area pores, allowing bacteria colonies to grow and efficiently remove H₂S gas. FMCRs are highly porous and offer an ideal medium for beneficial bacteria to grow by increasing their surface area through tiny pores and a perfect surface for bacterial colonies to adhere to; the FBBs are made from 100% polypropylene plastic material with a unique design of a large surface area to enhance the bacterial colony to adhere to and multiply. BFF acts as a separation layer between the mechanical and biological filter media. Before use, all filter media were washed with deionized (DI) water and left to dry at room temperature under laminar flow. The filler materials were uniformly mixed in a vessel and positioned in the lab-scale bioreactor column. The bioreactor is frequently (bidaily) inoculated with T. denitrificans bacteria augmented with necessary nutrient combinations such as NH₄Cl and KH₂PO₄ as nitrogen (N) and phosphorus (P) sources, respectively. Beyond T. denitrificans, characterizing the diversity of the bacterial communities within each column is crucial, particularly in understanding the functional role of T. denitrificans. Although T. denitrificans is anticipated to be the primary driver of H₂S bioremediation, other microorganisms are likely to contribute to the overall process. Furthermore, it is crucial to stress that interspecies bacterial interactions, including synergistic relationships and competition for nutrients and their influence on overall community diversity, must be considered for successful remediation.

Table 1.

The characteristics (material type, size, image) of filter materials

Filter media material name	Size (mm)	Material type	Image	Sample codes
Biofilter fiber (BFF)	Amorphous	Polypropylene		1TO
Ceramic ball filter media (CBFM)	25	Ceramic		2MO
Filter media ceramic rings (FMCR)	19 × 24	Ceramic		3MO
Filter bio balls (FBB)	32	Polypropylene		4BO

Bacterial cultivation and utilization

The Thiobacillus denitrificans culture was used in this study to evaluate the removal of H₂S gas from the bioreactor. The T. denitrificans were purchased from Leibniz Institute DSMZ (Product number: DSM 807) and maintained in culture. A loopful of pure culture was inoculated into the nutrient broth of 150 ml before each batch’s start. Its concentration was measured by the spread plate technique (9215C Standard method for examination of water and wastewater) for each batch. T. denitrificans with a concentration of 1 × 10⁵ CFU/ml were inoculated in the bioreactor and distilled water of 150 ml without the test organism in the control bioreactor every two days. Samples were collected every two days after inoculation of the medium in the bioreactors and analyzed for pH, sulfate, and cfu. Throughout the study period, 27 effluent samples from the bottom of the column bioreactor (valve V2) were successfully collected and analyzed for microbial colony count. Samples for CFUs were taken in a sterile container from both the operational and control bioreactors. The spread plate technique measures it by taking 0.5 ml of the sample with a sterile pipette and spreading it over a nutrient agar plate with a sterile spreader and incubating in the incubator for 24 h at 35 °C, and calculating the CFUs by the formula:

$$\text{Cfu}/\text{ml}=\left(\text{Colonies counted}/\text{sample volume}\right)\times \text{Dilution factor}.$$

Sample collection and DNA isolation

Microbial community samples attached to the surfaces of the filter materials were collected for DNA isolation. One whole piece from each layer of the filter media materials was collected. Then, the pieces were transferred into 15 mL Falcon tubes filled with 5 mL of distilled water. The samples were vortexed for 5 min at maximum speed and sonicated for 5 min in an ultrasonic water bath (Powersonic 410, Hwashin Technology, Korea). The suspension was centrifuged at 5,000 rpm for 5 min to collect the bacteria. According to the manufacturer’s recommendations, the Quick/DNA TM Fecal–Soil Microbe Miniprep Kit (Zymo et al. No: D6010) was used to extract bacterial DNA. After the extraction, the integrity and quantity of the DNA were tested using NanoDrop™ 2000/2000c (Thermo Fisher), Qubit 2.0 fluorometer (Invitrogen), and gel electrophoresis.

Amplification of 16S rRNA V3–V4 region

Forward and reverse primer sets (5′-CCTACGGGNGGCWGCAG-3′ and 5′-GACTACHVGGGTATCTAATCC-3′) were used to amplify the V3/V4 region of the 16S rRNA. The PCR reaction was carried out in a thermal cycler (MiniAmp™ Thermal Cycler, Applied Biosystems) using the following settings: 95 °C for 10 min (1x), 95 °C for 45 s, 53–50 °C for 45 s, 72 °C for 60 s (35x), and 72 °C for 3 min (1x).

Purification, library preparation, and sequencing

To explore the microbial spectrum and diversity on each filter material, total DNA was extracted and used for 16S rRNA metagenome sequencing. On the 60th day, the microbial community samples adhered to the surfaces of the filling materials (n = 4: 1TO, biofilter fiber (BFF); 2MO, ceramic ball filter media (CBFM), 3MO, filter media ceramic rings (FMCR), 4BO, filter bio balls (FBB)) were collected and investigated by metagenome sequencing. Before the library preparation, the amplified 16S rRNA was cleaned using a PCR Clean-Up kit (Applied Biological Materials, ABM). The purified DNA was quantified using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific), and the DNA concentration was adjusted to 5 ng/µL for subsequent library preparation. The sequencing libraries were prepared utilizing the Nextera XT DNA Library Prep Kit (Illumina), following the manufacturer’s protocol and making minor modifications. Accordingly, the purified PCR product (1 ng) was fragmented and tagged with adapter sequences by the Nextera XT transposome. Nextera XT Index Kit v2 Set A (Ilumina) was used to prepare samples for indexing. The PCR program consisted of an initial denaturation at 95°C for 3 min; 8 cycles of denaturation at 95°C for 30 s, annealing at 55°C for 30 s, and extension at 72°C for 30 s; and a final extension at 72°C for 5 min. The libraries were purified to remove short fragments and reaction reagents. The Illumina MiSeq instrument with MiSeq Reagent Kit v3 (2 × 300 cycles) and a MiSeq flow cell v3 were utilized for sequencing. Paired-end sequencing with 2 × 150 base reads was performed. The sequencing run parameters were set according to the manufacturer’s recommendations.

Bioinformatics and biodiversity analyses

Metagenome sequencing provides information about the microbial community in the bioreactor. (Tombuloglu et al. 2023). Using the Kraken metagenomics system, the raw sequencing reads were quality verified, trimmed, and categorized according to operational taxonomic unit (OTU) standards (Wood and Salzberg 2014). Our approach identifies the short DNA sequences with taxonomic categories with great accuracy and speed. A report on the bacterial distributions in the samples was created, and extra data, including the whole OTU distribution for each sample and different diversity studies, were acquired.

Simpson’s and Shannon’s Diversity Index were calculated for the biodiversity analyses. A measure of diversity known as Simpson’s Diversity Index provides information on the number of species found and the relative abundance of each species. The Shannon diversity index is a widely utilized measure in ecology. It is derived from Claude Shannon’s entropy formula and estimates species diversity within a given environment. This index considers both the richness, or the number of species present in a habitat, and their relative abundance, referred to as evenness. It is calculated as follows:

$$H=-\sum [\left(\text{pi}\right)\times \text{ln}\left(\text{pi}\right)]$$

n: The number of individuals of a given species.

N: The total number of unique species in a community.

Another widely used metric to measure the diversity of species in a community.is Simpson’s Diversity Index. It is calculated as follows:

$$D=\sum \text{ni}\left(\text{ni}-1\right)/N\left(N-1\right)$$

ni: The number of organisms that belong to species i.

N: The total number of organisms.

A KRONA map was developed to visually investigate the results of species annotation between the clusters following OTU tags. We also performed a differential abundance analysis between the samples to further explore the taxonomic differences between the groups. This analysis is based on the Bioconductor tool at the genus level. (Robinson et al. 2010). In addition, a heatmap showing the top 10 differentially abundant genera, ordered by the False Discovery Rate (FDR), was generated. (Škuta et al. 2014).

Results and discussions

H₂S removal performance

Figure 3 shows that the maximum H₂S control efficiency reached 99% within the first seven days at 8 mg day⁻¹ of H₂S mass flow rate. Slight decreases in the removal efficiency were noticed when the mass flow rate was increased (a maximum 4% decrease at 24 mg day⁻¹ of H₂S mass flow rate). Following this, the system recovered shortly and achieved a maximum efficiency of 99%.

Fig. 3.

H₂S removal performance of the bioreactor. The blue and orange lines show removal efficiency (%) and inlet H₂S mass flow rate (mg/day)

The summarized statistical data from the experimental analysis are presented in Table 2. In the first stage of experiments, at 8 mg day⁻¹ of H₂S mass flow rate, viable microbial colonies were counted as 3.25 × 10⁵, and they reached up to 1.70 × 10⁹ CFU mL⁻¹ with a progressive increase at the end of the experiments. Parallel to the enhanced biogenic activity, the sulfate concentration initially measured as 166 mg L⁻¹ reached 438 mg L⁻¹ gradually by the day the experiments were completed, with an average value of 308.7 mg L⁻¹. The formation of sulfate indicates the activity of sulfur oxidation and an enhanced H₂S gas conversion to sulfate by the T. denitrificans bacteria. The pH value varied between 7.70 and 8.60, with an average of 8.25 ± 0.23 throughout the study period. The observed pH value in this study is in good agreement with the other studies that used different bioreactor configurations to remove H₂S gas (Fernández et al. 2013; Chaiprapat et al. 2015; San-Valero et al. 2019).

Table 2.

Summary of statistical data for the bioreactor after 60 days of T. denitrificans inoculation

	Microbial colony count (CFU/mL)	Sulphate (mg/L)	pH
Average	2.60 × 108	308.7	8.25
Median	2.26 × 108	317.0	8.31
SD	3.11 × 108	60.71	0.23
Min	3.25 × 105	166.0	7.70
Max	1.70 × 109	438.0	8.60

Sequencing and microbial community analysis

Table 3 depicts the sequencing statistics such as average read length (nt), reads per sample, and classified groups. Accordingly, the raw read average was 40,134, and the read length (nt) ranged between 110.4 (1TO) and 138.5(nt) (2MO), with an average of 126 (nt).

Table 3.

The sequencing statistics over four samples

Sample*	Avg read length (nt)	Reads / classified (%)
2MO	138.5	71,963 / 99.98
3MO	123.5	57,051 / 99.97
4BO	122.5	31,412 / 99.75
1TO	110.4	17,653 / 99.89

^*1TO, biofilter fiber (BFF); 2MO, ceramic ball filter media (CBFM), 3MO, filter media ceramic rings (FMCR), 4BO, filter bio balls (FBB)

Upon H₂S treatment, the read counts of the T. denitrificans in each environment are given in Table 3. So, the bacteria are likely to grow well respectively in 2MO > 3MO > 4BO > 1TO. This result suggested that ceramic materials are more likely to be beneficial for T. denitrificans growth than polypropylene surfaces. In addition to T. denitrificans, the microcosm existing on different filter materials was explored. Table 3 shows the top 10 species with the highest species reading counts of the filter media materials placed in the column. The distribution of species was found to be diverse. In the 1TO microcosm, Mesorhizobium terrae (14.27%) and Rufibacter sp. DG15C (8.46%), Priestia megaterium (7.57%), and Enterococcus faecalis (7.35%) are the most abundant bacterial species. While in the 2MO microcosm, Staphylococcus cohnii (14.85%) is the most abundant colony of the total microbiome, followed by Enterococcus faecalis (9.66%), Comamonas testosteroni (7.0%), and Comamonas thiooxydans (6.84%). In the 3MO microcosm, Staphylococcus aureus (28.04%), Comamonas thiooxydans (8.77%), Enterococcus faecalis (7.34%), and Roseateles depolymerans (2.75%) are identified as the fastest-growing bacterial species. In the 4BO microcosm, Comamonas thiooxydans is the most abundant species (14.8%), followed by Enterococcus faecalis (12.11%), Comamonas testosteroni (4.82%), and Escherichia coli (4.2%).

As shown in Fig. 4, the read counts for T. denitrificans are lower than those of the most abundant species. They are not among the top 10 interspecies interactions that could influence their lower abundance, particularly competition for nutrients. However, while competition is a plausible explanation, other factors could also contribute. For instance, the differential lysis performance of T. denitrificans cells during DNA extraction may lead to an underrepresentation of its DNA in the sequencing library. Another possible reason is the primer bias. The 16S rRNA gene amplification primers may have a lower affinity for T. denitrificans, resulting in fewer amplicons and lower read counts. In addition, T. denitrificans may occupy a niche within the community where its absolute abundance is lower, even if it plays a crucial functional role in H₂S removal. Further investigation, such as using alternative DNA extraction methods, different primer sets, or metagenomic sequencing, would be helpful to definitively determine the reasons for the lower read counts of T. denitrificans.

Fig. 4.

Relative abundance of the top 10 most abundant species on T. denitrificans-inoculated filter media: ceramic ball filter media (CBFM, 2MO), ceramic rings (FMCR, 3MO), bio balls (FBB, 4BO), and biofilter fiber (BFF, 1TO)

The diversity of phyla, classes, and genera

Figure 5 depicts the phylum composition of different filter materials. The results revealed that the dominant Phylum across all samples, regardless of the treatment and filter material type, was Proteobacteria, followed by Firmicutes and Bacteroidetes. On the other hand, changing the filter material altered the phylum diversity of the microcosms. For instance, the 1TO group resides more in Firmicutes, Actinobacteria, and Bacteroidetes than in the other filter materials. This result revealed the importance of the filter material used in bioreactor columns, diverting the composition of microorganisms.

Fig. 5.

Taxonomic classification (phylum level) of dominant bacterial species in each filter material: biofilter fiber (BFF, 1TO), ceramic ball filter media (CBFM, 2MO), ceramic rings (FMCR, 3MO), and bio balls (FBB, 4BO)

Taxonomic community composition and bacterial diversity analysis

To study the taxonomic community compositions across the different operational groups, we used the Kraken tool via the OmicsBox program. Table 4 depicts the Shannon Index and Simpson index scores calculated in the microcosm of varying filter materials.

Table 4.

Estimation of the α-diversity index at the species level

Sample***	Shannon index (H) / (H / LN(N))*	Simpson index (1−D) **
1TO	4.452 / 0.7211	0.9605
2MO	4.333 / 0.7084	0.9635
3MO	4.059 / 0.6521	0.9137
4BO	4.532 / 0.7586	0.9699

^*Shannon index ranges between 0 and has no upper limit. The higher the index, the more diverse the species are in the community

^**D indicates Simpson’s diversity index, which shows species diversity in samples. Simpson’s index is between 0 and 1. 1 indicates diversity, while 0 indicates no diversity

^***1TO, biofilter fiber (BFF); 2MO, ceramic ball filter media (CBFM), 3MO, filter media ceramic rings (FMCR), 4BO, filter bio balls (FBB)

Furthermore, Fig. 6a displays all samples’ taxonomic classification on the genus level. Interestingly, the main Genus in most samples is Comamonas, followed by Cupriavidus. However, in the case of the 1TO sample, we see that Cupriavidus surpasses Comamonas as the dominant Genus (9% to 5%, respectively). In addition to Comamonas and Cupriavidus, Paraburkholderia, Ralstonia, and Thiobacillus are visibly detected in all samples with differing percentages. We also performed a differential abundance analysis between the H₂S-treated samples to investigate the taxonomic differences between the groups further. This analysis is based on the Bioconductor tool at the genus level. Figure 6c displays a heatmap showing the top 10 differentially abundant genera, ordered by the False Discovery Rate (FDR). In addition, we used Krona charts to group the samples hierarchically in a radial cladogram. Figure 6b illustrates the composition of the groups of samples.

Fig. 6.

Taxonomic analysis of microbial communities in H₂S-treated samples. a Bar chart showing the relative abundance of bacterial genera. b Krona chart comparing the taxonomic composition of samples. c Heatmap visualizing the differences in taxonomic profiles between the H₂S-treated samples

Conclusions

The current work revealed that the H₂S was converted to sulfate with a removal efficiency of 99% in seven days, revealing the high capacity of T. denitrificans in the remediation of H₂S gas in the bioreactor systems. The sulfate formation in the collected solution shows the activity of sulfur oxidation and an enhanced H₂S gas conversion to sulfate by the T. denitrificans bacteria. This study’s slightly alkaline pH level is consistent with other studies reported in the literature that focus on the biological treatment of H₂S gas (Tilahun et al. 2018; Khanongnuch et al. 2019). These results proved the successful removal of H₂S gas with a laboratory-scale bioreactor column using T. denitrificans bacteria. When compared with the T. denitrificans read counts on each filter material, it was revealed that ceramic materials are more likely to benefit T. denitrificans growth than polypropylene surfaces. In addition to T. denitrificans, other sulfide-oxidizing bacterial species have immense potential to metabolize H₂S effectively, such as Thiobacillus, Xanthomonas, and Pseudomonas (Doungprasopsuk and Suwanvitaya 2017; Sonthiphand et al. 2022; Das et al. 2022). In addition to T. denitrificans, C. thiooxydans, C. testosteroni, E. faecalis, S. aureus, S.cohnii, and M. terrae could be a promising bioremediation strategy for removing H₂S. Alternating the filter material diversifies the species and phyla found on the filter material. Compared with polypropylene surfaces, ceramic materials are more efficient at harboring T. denitrificans bacteria, which can be preferred in industrial-level reactor applications for H₂S remediation. Conversely, altering the filter material changed the microcosms’ phylum diversity. For example, in comparison to the other filter materials, Firmicutes, Actinobacteria, and Bacteroidetes were abundant in the 1TO group. This result proved how crucial the filter material in bioreactor columns is for rerouting the microorganism composition. At the species level, Comamonas thiooxydans, Comamonas testosteroni, Enterococcus faecalis, Staphylococcus aureus, Staphylococcus cohnii, and Mesorhizobium terrae are the most abundant species detected on the filter materials.

The results of this work can be applied on a large scale to oil and gas facilities, landfills, or any wastewater treatment plant, converting the fugitive H₂S gas to sulfate utilizing Thiobacillus denitrificans bacteria.

Author contributions

OA, IA, and CY conceptualized the study, analyzed the data, wrote the original draft, and revised the manuscript. AQ carried out microbiological lab experiments. HT conducted metagenome sequencing and bioinformatics analyses. IE carried out lab and data analysis. All authors read and commented on the manuscript.

Funding

This research was funded by Imam Abdulrahman bin Faisal University (IAU) project No. (Eng-044-2018) through the Deanship of Scientific Research (DSR).

Data availability

The metagenome sequence data are uploaded to the NCBI Sequence Read Archive (SRA) database under the submission number SUB14993326, BioProject ID PRJNA1212586, and BioSample accessions SAMN46304422 – SAMN46304426.

Declarations

Conflict of interest

The authors declare that there is no conflict of interest regarding the submission of this manuscript and that all co-authors have approved the manuscript for publication. On behalf of all co-authors, I confirm that the work presented is original research, has not been published previously, and is not currently under consideration for publication elsewhere, either in whole or in part. All authors have reviewed and approved the final version of the manuscript.