Comparison of different pretreatment methods for hydrogen production using environmental microbial consortia on residual glycerol from biodiesel

Daniele Misturini Rossi ^a, Janaı´na Berne da Costa ^b, Elisangela Aquino de Souza ^a, Maria do Carmo Ruaro Peralba ^b, Dimitrios Samios ^b, Marco Antoˆnio Za´chia Ayub ^a,*

Biotechnology & Biochemical Engineering Laboratory (BiotecLab), Federal University of Rio Grande do Sul, Av. Bento Gonc¸alves 9500, PO Box 15090, ZC 91501-970, Porto Alegre, RS, Brazil

Department of Organic Chemistry, Institute of Chemistry, Federal University of Rio Grande do Sul, Av. Bento Gonc¸alves 9500, PO Box 15090, ZC 91501-970, Porto Alegre, RS, Brazil

A b s t r a c t

The pretreatment of environmental microbial consortia by five methods (acid, base, heat shock, dry heat and desiccation, freezing and thawing) was conducted in order to evaluate their applicability for the selection of hydrogen-producing bacteria capable of using residual glycerol from biodiesel synthesis as substrate. Results showed that substrate degradation rates of consortia pretreated with dry heat and desiccation and heat shock were higher compared with controls during the fermentation using glycerol, with degradation rates as high as 65%. The maximal hydrogen and biomass productions were obtained by dry heat and desiccation: 34.19%mol and 4340 mg/L, respectively. Dry heat and desiccation followed by heat shock are simple pretreatments methods that can be used to improve the biotechnological production of hydrogen. DNA sequencing performed to identify the bacteria strains present in the consortium showed that they belonged to the genus Klebsiella and Pantoea.

1. Introduction

Research on alternative energy sources has gained renewed interest due to the growing awareness that the accumulated carbon dioxide in the atmosphere is a potential cause of climate change. Combustion of hydrogen produces no greenhouse gases and has a high-energy yield of 122 kJ/g, which is 2.75-fold higher than hydrocarbon fuels. Thus using hydrogen as a clean fuel seems to be a promising technology. Current production of hydrogen can be achieved by either physicochemical or biological methods. Biological production of hydrogen from complex natural and residual substrates such as sugar-rich wastewaters, cellulose, municipal solid waste, sugarcane juice, corn pulp, and paper have been reported. Biological hydrogen production may be more effective if organic wastewater or other industrial wastes are employed as raw materials in this process.

The production of alternative fuels such as biodiesel and ethanol has dramatically increased over the last few years. Due to the increasing production of biodiesel, a glut of crude glycerol has resulted and the price has plummeted over the past few years. Therefore, it is imperative to find alternative uses for glycerol. Recent research has been conducted showing the possibilities of using residual glycerol in biotechnological processes, especially for the production of 1,3-propanediol, ethanol, and organic acids. However, few investigations have focused on its conversion into hydrogen. The anaerobic conversion of substrates, such as, glucose, starch and sucrose into hydrogen is a complex biochemical process. Several fermentative bacteria produce hydrogen, which functions as an intermediate energy carrier and can be used in fuel cells. In this process, bacterial hydrogenases liberate hydrogen to dispose of excess electrons. Bacteria that use these pathways are strict anaerobes (Clostridia, methanogenic bacteria, archaea), facultative anaerobes (Escherichia coli, Enterobacter, Citrobacter) and some aerobes (Alcaligenes, Bacillus). Hydrogen production is usually carried out using mixed cultures of uncharacterized bacteria and with unsterilized substrates since these processes are more practical and robust than those using pure, well define bacteria. The pretreatment methods reported for enriching hydrogen-producing bacteria consortia are: heat shock, use of acids or bases, aeration, freezing and thawing, chloroform, sodium 2-bromoethanesulfonate or its acid, and iodopropane. These reports focus on the use of sugar sources such as glucose, sucrose, or starch for the production of hydrogen, but none of them employed residual glycerol from biodiesel as a carbon source. Therefore, the objective of this study was to investigate the influence of different pretreatment methods (acid, base, heat shock, dry and desiccation, and freezing and thawing) for selecting and conditioning an environmental-isolated consortium of bacteria for their use in producing hydrogen from residual glycerol resulting from biodiesel synthesis. Finally, DNA sequencing was performed to identify the bacteria strains present in the consortium capable of metabolizing raw glyc-erol and converting it into hydrogen.

2. Materials and methods

2.1 Raw (residual) glycerol

The raw (or residual) glycerol was supplied from a biodiesel producing and refining facility in the South of Brazil (Passo Fundo, RS, Brazil). The raw glycerol quality and compo-sition was strictly controlled corresponding to 80.9% glycerol, 6.4% ash, 6.6% NaCl, and 11.6% moisture with a pH of 7.26. One single batch was used throughout this research.

2.2  Environmental microbial consortium and its pretreatments

The environmental microbial consortium was collected from the bottom portion of an upflow anaerobic sludge blanket reactor (UASB) at a local soybean treatment plant (Esteio, RS, Brazil). The concentration of the volatile suspended solids (VSS) of the environmental microbial consortium was 4600 mg/L.

Five physicochemical treatments were applied to the consortium. The heat-shock pretreatment was carried out by boiling the sludge at 100 C for 15 min. The acid and base pretreatments were performed by respectively adjusting the pH of the samples to 3.0 or 10.0 using 1 M HCl or NaOH and maintaining this pH for 24 h. The dry heat and desiccation was performed by storing the sample for 2 h in a drying oven at 105 C followed by desiccation in a desiccating jar for 2 h. The freezing and thawing was performed by exposing the samples to 10 C for 24 h, followed by a thawing process for 6 h at 30 C. Control experiments were carried out with the same environmental microbial consortium without applying any further pretreatments. All experiments were run as duplicates. The results were plotted as the mean of the two experiments showing the standard deviation.

2.3 Cultures

The main substrate solution consisted of organic and inor-ganic nutrients. It had a total chemical oxygen demand (COD) of 30.7 g/L O₂, mainly from glycerol. The medium used for the pretreatment tests was composed of 30 g/L of raw glycerol added of the following nutrients (in g/L): 4 (NH₄)₂SO₄; 0.125 K₂HPO₄, 0.12 MgSO₄.7H₂O; 0.01 MnSO₄.H₂O; 0.025 FeSO₄.7H₂O; 0.005 CuSO₄.5H₂O; 0.125 CoCl₂.6H₂O. A trace-element solution was added at 0.1% (v/v) to medium and contained (in g/L): 0.1 MnCl₂.4H₂O; 0.06H₃BO_3; 0.0037 CuSO₄.5H₂O; 0.2 CoCl_2.6H₂O; 0.025 NiCl₂.6H₂O; 0.035 Na₂MoO₄. 2H₂O; 0.14 ZnSO₄ 7H₂O; 8 NaHCO₃; and 0.9 ml HCl 37%. Batch runs were conducted in 60 mL glass bottles with a working volume of 30 mL, pH adjusted to 7.0 with 1 M HCl or NaOH, inoculated with a 10% (v/v) environmental microbial consortium, corre-sponding to an initial cell concentration (measured as VSS) of 460 mg/L. Each bottle was flushed with nitrogen gas to provide oxygen-free conditions, capped with a rubber stopper and placed into a reciprocal shaker at 150 rpm, 35℃for 36 h. These same culture conditions and procedures were used to culti-vate the isolated bacteria from the consortium. In these experiments, the collected gas samples at 36 h were analyzed for hydrogen.

2.4  Isolation, sequencing of 16S rDNA genes and genetic analysis of bacteria in the consortium 

The isolation and identification of the bacteria present in the microbial consortium was carried out in order to acquire information on the most relevant genera present in it. The samples were collected at the end of cultivation, serially diluted into 0.1% sterile peptone water and 0.1 mL of these cell suspensions were spread onto nutrient agar (NA) and incu-bated at 35℃. The most probable number (MPN) method was used to enumerate the bacteria cells present in the microbial consortium. After incubation for 24 and 48 h, colony counts were determined and representative colonies were sub cultured into LB (Luria Bertani) liquid medium and stored in glycerol (1:1) at 20℃ for posterior identification. Cellular morphologies were determined by bright field microscopy of Gram-stained preparations.

The molecular identification of the isolates were conducted at the ACTGene Laboratory (Biotechnology Centre, UFRGS, Porto Alegre, RS, Brazil) using the automatic sequencer ABI-PRISM 3100 Genetic Analyzer equipped with 50 cm capillaries and POP6 polymer (Applied Biosystems, USA). DNA templates (30e45 ng) were labeled with 3.2 pmol of the primer 5⁰-NNNNNNNNNNNNN-3⁰ and 2 mL of BigDye Terminator v3.1 Cycle Sequencing RR-100 (Applied Biosystems, USA) to a final volume of 10 mL. Labeling reactions were performed in a GeneAmp PCR System 9700 (Applied Biosystems, USA) ther-mocycler with a initial denaturing step of 96℃ for 3 min followed by 25 cycles of 96℃ for 10 s, 55℃ for 5 s, and 60℃ for 4 min. Labeled samples were purified by isopropanol precipi-tation followed by 70% ethanol rinsing. Precipitated products were re-suspended in 10 mL formamide, denatured at 95℃ for 5 min, ice-cooled for 5 min and electroinjected in the automatic sequencer. Sequencing data were collected using the software Data Collection v1.0.1 (Applied Biosystems, USA), programmed with the following parameters: Dye Set “Z”; Mobility File “DT3100POP6 {BDv3}v1.mob”; BioLIMS Project “3100_Project1”; Run Module 1 “StdSeq50_POP6_50 cm_cfv_100”; and Analysis Module 1 “BC-3100SR_Seq_FASTA.saz”.

2.5  Analytical methods

The soluble metabolites were analyzed using a Shimadzu high performance liquid chromatography system (Shimadzu Corp., Japan) with a RID-10A refractive index detector. The stationary and mobile phases were an Aminex HPX-87H column (300 7.8 mm) (Bio-Rad, USA) and 0.005 mol/L H₂SO₄ solution at 0.8 mL/min, respectively. The column temperature was controlled at 65℃. The hydrogen production was analyzed using a gas chromatograph (Agilent 6890N, USA) equipped with a thermal conductivity detector (TCD) and a PoropaK Q column (mesh 80/100, 6ft, 1/8 in). The tempera-tures of the column and the TCD detector were 80℃ and 150℃, respectively. The COD (Chemical Oxygen Demand) was measured according to the closed reflux colorimetric method and volatile suspended solids (VSS) were analyzed according to the procedures described in APHA standard methods.

3. Results and discussions

3.1. Effect of different pretreatments on hydrogen production by environmental microbial consortium

Fig. 1 illustrates the effect of different pretreatments methods on the hydrogen production by environmental microbial consortium. Hydrogen production was higher for the dry heat and desiccation (34.2% mol), and for the heat shock (27.3% mol) treatments, while the base pretreatment did not show any hydrogen production. Biological hydrogen production by dark fermentation processes shares many common features with methanogenic anaerobic digestion. Typical anaerobic mixed cultures cannot produce hydrogen once it is rapidly consumed by methane-producing bacteria. Therefore, the most effective way to enhance hydrogen production in anaerobic microbial cultures is to restrict methanogenesis by allowing hydrogen to become an end product of the metabolic flow. Pretreatment of cultures have been used to selectively enrich specific groups of bacteria. Spore forming hydrogen producing bacteria such as Clostridium will form endospores as a result of bacterial stress when environmental conditions are harmful (high temperature, desiccation, nutrient limitation, extreme acidity and alkalinity), while methanogenic bacteria will have no such capabilities, preventing the competitive growth of these bacteria, which are hydrogen consumers.

Wang and Wan  described the use of five methods of sludge pretreatment (heat shock, acid, base, aeration and chloroform) in order to induce hydrogen production by a consortium collected from anaerobic digested sludge at Beijing Sewage treatment plant. The results showed that the higher hydrogen production (215.4 mL) was obtained using heat shock pretreatment with temperature of 35℃, initial pH ¼ 7.0 and glucose as carbon source. Cheong and Hansen have also carried out a study comparing five pretreatment methods (acid, sodium 2-bromoethanesulfonate, wet and heat shock, dry heat and desiccation, and freezing and thawing) for enriching hydrogen producing bacteria from cattle manure sludge growing on glucose. These authors found that the acid pretreatment was the more efficient.

The hydrogen production from samples pretreated with base or acid methods were similar to those of the control, possibly because extreme pH might have suppressed both methanogenic and hydrogen producing bacteria. Although endospore-forming bacteria such as Clostridia show high pH tolerance, facultative anaerobes such as Enterobacter and Klebsiella species have shown a very restricted optimal pH range between 5.0 and 6.0 for the production of hydrogen. Studies with enterobacteria isolated from sludge showed optimal pH range for hydrogen production between 6 and 6.5.

Table 1 shows the yields of hydrogen production after 36 h of cultivation for different pretreatments. The results showed that heat shock produced the highest yield of hydrogen followed by dry heat and desiccation, while acid and base pretreatments produced lower yields than the control. The hydrogen production has been studied using glucose as usual substrate and hydrogen yields in continuous culture typically can range anywhere from 0.7 to 4 mol H₂/mol glucose depending on bacterial community, temperature, retention time and other factors. Glycerol is usually used for the production of 1,3-propanediol by Klebsiella pneumoniae and this microorganism can also convert glycerol to hydrogen at high rates and yields. Li and Fang investigated the hydrogen production using biodiesel wastes with Klebsiella pneumoniae. Their results showed hydrogen yields, hydrogen evolution rate, and production of 0.53 mol/mol glycerol, 17.8 mmol/L/h, and 117.8 mmol/L, respectively. It conducted a study on hydrogen and ethanol production from waste discharged from biodiesel manufacturing process using Enterobacter aerogenes HU-101. Their results showed that yields of hydrogen from glycerol (1.2 mol/mol) exceeded the theoretical maximal yield of H₂ (1.0 mol/mol) with 1.7 g/L glycerol. However, the yields of hydrogen, ethanol and acetate decreased when the concentration of biodiesel wastes increased. Fountoulakis and Manios studied the effects of raw glycerol on the performance of a single-stage anaerobic reactor treating different types of organic waste to produce methane and hydrogen. The 1% (v/v) addition of raw glycerol to the feed increased the methane production rate from 479 mL/d to 1210 mL/d. In relation to hydrogen production, the authors showed that raw glycerol had a significant positive effect on the anaerobic fermentation with hydrogen enhanced of 2.9 mmol H₂/g glycerol when this substrate was used.

3.2 Effect of treatments on the production of other soluble metabolites

Table 2 summarizes the effects of the different pretreatments on the production of several soluble metabolites. The major soluble metabolites formed in the cultures treated by heat shock and dry heat and desiccation were acetic and butyric acids, and 1,3-propanediol. Control, acid, base and freezing and thawing pretreatments did not show any acetic acid production. Acetic and butyric acids are the main fermenta-tion products of pyruvate from C. butyricum, while 1,3-PD is produced by enterobacteria species such as K. pneumoniae and Clostridium. Results from studies using glucose as substrate for consortia fermentation have shown similar results to the ones presented in this study. In the Wang and Wan study the main soluble metabolites were ethanol, acetic acid, butyric and propionic acids showing that two different microbial metabolisms were present: mixed and ethanol fermentation. Khanal; studied the effects of pH using a composting pile as seed source and sucrose and starch as organic substrates for hydrogen production. The authors detected propionate, acetate and butyrate as the major soluble metabolites. Liu and Fang  using biodiesel wastes as substrates to produce hydrogen K. pneumoniae DSM 2026 cultures reported a production of 6.7 g/L of 1,3-propanediol.

3.3 Effect of treatments on substrate degradation rate

Table 1 shows the effect of different pretreatment methods on the rate of substrate degradation after 36 h of cultivation. The results showed that the substrate degradation rate of the environmental microbial consortium pretreated by dry heat and desiccation, heat shock, and freezing and thawing, in this order, were higher than that of the control, while the substrate degradation rate of the pretreatments with acid and base, were negatively affected. The pretreatments by dry heat and desiccation and heat shock produced a maximal substrate degradation efficiency of 66% of an amount of 30 g/L of raw glycerol. Sakai and Yagishita., studying the hydrogen and ethanol production using glycerol-containing wastes discharged from a biodiesel fuel production plant using a bio-electrochemical reactor with thionine, showed that glycerol at a concentration of 10.12 g/L was almost completely consumed (8.54 g/L) under pH ¼ 6.5 and 7.0. Wang and Wan, comparing different pretreatment methods (acid, base, heat shock, aeration and chloroform), for enriching hydrogen producing bacteria from digested sludge using glucose as a substrate, showed a degradation efficiency of 97.2% with digested sludge pretreated by heat shock.

3.4 Effect of treatments on the microbial growth and final pH of cultivation

The VSS assay was used in this study in order to estimate biomass. Table 1 shows the effects of the different pretreatment methods on biomass determined at 36 h of cultivation. The results showed that the higher biomass corresponding of 4340 mg/L was obtained with the pretreatment by dry heat and desiccation. The biomass of base pretreated sludge was lower than the control indicating that this treatment inhibited the growth of the microbial consortia contributing thereby to a lower hydrogen production and yields.

One of the key process parameter in the production of hydrogen is the system pH because it may directly affect the hydrogenase activity as well as the metabolic pathway. Table 1 shows the effects of different pretreatment methods on the final pH of cultures at 36 h of cultivation and, except for the base treatment that finished at pH ¼ 5.9, all others produced a significant drop on pH to 4.5, reflecting the organic acid production shown before. These results were expected and similar to studies with glucose. The influence of pH has been recognized as a key factor in determining the outcome of hydrogen fermentation. The pH is related to three important facts: growth limitation of methanogens, hydrogen production performance, and regulation of shift to solventogenesis. Maintaining pH in the acidophilic range of 5.5e6.0 is ideal for effective hydrogen production due to repression over methanogenic bacteria, indirectly promoting hydrogen producers within the system. The optimal pH for hydrogen production is in the range of 5e7. Van Ginkel and Sung  studied the biohydrogen production as a function of pH and substrate concentration using compost from potato and soybean soil as natural inocula and applying heat shock pretreatment. The authors showed that the high-est rate of hydrogen production occurred with pH of 5.5 with a conversion efficiency of 46.6 mL H₂/(g COD/L).

3.5. Analysis of the bacterial community present in the consortium

Only four different colonies were isolated from the cultivations, as assumed from the microscopic observation after Gram dying and cell morphology in the plates. They were all Gram-negative. Based on 16S rDNA analysis, three isolates were identified as Klebsiella pneumoniae (96% of confidence), while the fourth was identified as Pantoea agglomerans (93% of confidence). The Klebsiella isolates could not be considered as being the same strain, since they always produced very distinctive colony morphologies when plated. Therefore, strains were labeled as K. pneumoniae BLh-1, BLh-2, and BLh-3, and as P. agglomerans BL1. Some microorganisms have been isolated and evaluated for hydrogen production, including photosynthetic bacteria and anaerobic dark fermentative bacteria such as Clostridium sp., Enterobacter sp., Bacillus sp., and Klebsiella sp. Wu et al. (2010) isolated a Klebsiella sp. HE1 from sewage sludge, which produced 0.92 mol H₂/mol of sucrose, used as the carbon source for cultivation. However, few studies have been reported for the use of residual glycerol as a carbon source to produce hydrogen instead of defined media using common sugars as substrates. The isolated strains were individually cultivated under the same conditions described above in a simple experiment to check their ability to metabolize residual glycerol and the results shown in Table 3 clearly demonstrate their high potential to use glycerol as the sole carbon source and to produce hydrogen.

4. Conclusions

This research shows that residual glycerol can be used as substrate for the production of hydrogen in substitution of other, more expensive carbon sources as glucose, sucrose or starch. The pretreatments methods (acid, base, heat shock, dry heat and desiccation, freezing and thawing), used for selective enrichment of hydrogen production using an environmental anaerobic consortium showed considerable influence on hydrogen production and substrate degradation. The physical pretreatments using heat showed the best results on hydrogen production and maximal substrate degradation rate. DNA techniques used to identify bacterial strains present in the consortium showed that they belonged to the genera Klebsiella and Pantoea, the last one shown for the first time as a hydrogen producer from glycerol. The results showed that glycerol could be efficiently used in the production of hydrogen by replacing the traditional sources of carbon such as sugars. The use of this waste could then reduce production costs of hydrogen using an environmental friendly process.