1. Introduction
Biological methanation (BM) plays an important role in improving the energy independence of Europe by providing sustainable and decarbonized biomethane [
1]. For example, the IRENA report on “Bioenergy for energy transition” states that the enhanced use of biomass materials and energy is an important factor in meeting the 1.5 °C scenario. In addition, the combination of bioenergy with carbon capture and storage (BECCS) may further reduce global greenhouse gas emissions. It may even provide negative emissions, as previously biologically bound carbon dioxide is captured and stored [
2]. The most known example of bioenergy is biogas from anaerobic fermentation, containing around 45–65% methane and 35–55% carbon dioxide, with traces of ammonia and hydrogen sulfide. However, the injection of biogas into the existing gas grid is restricted in terms of different contaminants, which need to be separated [
3]. BM is capable of converting the remaining CO
2 and upgrading biogas to biomethane, which can be injected and distributed throughout the European gas grid. For this purpose, different methanogenic archaea are used, which were found and investigated in the last century.
Methanothermobacter marburgensis is the most widely used methanogenic archaeon for BM, which was first isolated from sewage sludge in Marburg (strain Marburg
[
4]), as well as various other locations [
5,
6]. The organism was identified as a variant of the previously found
Methanobacterium thermoautotrophicum , isolated from sewage sludge in Urbana, Illinois [
7]. Both organisms were investigated using DNA–DNA hybridization by Brandis et al. They concluded that these strains are not closely related but the differences are not evident enough to generate a new species [
8]. However, Wasserfallen et al. showed with three independent datasets (16S rRNA sequences, antigenic fingerprinting, and plasmid and phage typing) that the strains are different from each other and described the new genus
Methanothermobacter with the respective species, which was also proposed by Boone et al. [
9,
10].
The isolation and cultivation of anaerobic methanogenic archaea like
M. marburgensis is difficult. Special care should be taken to ensure that the medium used meets the nutrient requirements of the organism, which correspond to the environment prior to isolation and are based on the elemental composition of the cell. Furthermore, oxygen can react with substances in cells and inhibit growth or increase toxicity to organisms. Mylroie and Hungate verified in different experiments that a low ORP is needed for the growth of
M. formicicum [
11]. To simplify cultivation under oxygen-free conditions, Hungate developed the so-called roll tube technique in 1950 [
12], which was further improved [
13] but also adapted for different applications by various researchers. One example is the commonly used methanogenic bacteria cultivation technique by Balch and Wolfe [
14]. In order to bind remaining oxygen, the medium is then reduced by the addition of different suitable chemicals (see
Table 1).
2. State of the Art
In the past, different reducing agents were determined and compared. In 1954, Mylroie and Hungate conducted experiments with
M. formicicum using sodium sulfide as a reducing agent for the medium. They found that substitution with cysteine did not improve the results [
11]. In 1961, Bryant et al. reported that the usage of sodium sulfide instead of cysteine, which was commonly used at this time, led to a greater growth of ruminal bacteria. They also stated that the use of other reducing agents, including dithionite, was not satisfactory for the cultivation of such organisms [
25]. Later on, Hungate published an article about his roll-tube technique in 1969, for which he used several reducing agents to adjust the ORP of the media. He found that hydrogen sulfide, which is the product of sodium sulfide at pHs of 6–7, may be the best option as a reducing agent. Furthermore, he stated that the reducing agents had an inhibiting or even toxic effect at higher concentrations. This applies in particular to sodium dithionite [
12,
26]. As
M. marburgensis was further described and investigated with respect to the growth conditions and trace elements needed for metabolism by Schönheit et al. in 1979 and 1980 [
27,
28], they also used sodium sulfide as a reducing agent based on the findings mentioned above. Overall, it can be said that based on this research, sodium sulfide was established as reducing agent, whereas other reducing agents like sodium dithionite were not applied.
In contrast, Rothe et al. developed a simpler method in 2000 to cultivate methanogenic and hyperthermophilic anaerobic archaea without the use of sodium sulfide but with sodium sulfite (Na
2SO
3). They stated that sodium sulfide causes a series of problems because of its reaction with water and weak acids to hydrogen sulfide. First, in continuous reactors, H
2S is flushed out of the reactor due to the low solubility in water. Second, the addition of sulfide anions leads to the precipitation of cationic trace elements. Third, precipitated cations hinder cell growth measurement and the distinction between cells and precipitates [
29]. At laboratory scale, the amount of hydrogen sulfide produced may be small, and when glass is used, the corrosive properties are neglected. When considering the scale-up and application of high pressures to biological methanation, it is completely different since stainless steel is used for construction. Sulfur compounds such as H
2S, S
2O
3, SO
3, HS
− and even S
2O
4 are known to corrode stainless steel. Therefore, specific manufactured stainless steels, e.g., carbon or austenitic stainless steels, are needed for the whole system. However, H
2S is the most hazardous and toxic chemical compared to the other sulfur compounds [
30]. Even small amounts of this chemical can cause several challenges due to technical, economic and obvious safety aspects.
In terms of the state of the art, it is quite common to use sodium sulfide as a reducing agent despite the reported hazardous and environmental disadvantages. Therefore, this study showed that sodium dithionite and L-cysteine-HCl can be used as a substitute reducing agent for M. marburgensis with equal or even better methane evolution rates (MER) compared to sodium sulfide. For this purpose, a method was developed to compare and determine the optimal cultivation conditions with respect to MER. The method was established for commonly used 120 mL serum bottles in anaerobic cultivation. The optimal cultivation conditions were then used to compare different reducing agents at different concentrations.
3. Materials and Methods
In this chapter, the chemicals, gases, medium and laboratory devices used are described, as well as the setup of the experiments and the applied measurement methods.
3.1. Chemicals and Gases
The chemicals used were of analytical grade and purchased from Carl Roth GmbH+ Co. KG Karlsruhe, Germany.For cultivation, a mixture (4:1) of hydrogen (99.999 vol%) and carbon dioxide (99.9 vol%) resulting in 80 vol% hydrogen and 20 vol% carbon dioxide was used. For media preparation and gas chromatography, argon was used as a carrier gas with a purity of 99.999 vol%. All gases were purchased from Westfalen AG, Muenster, Germany.
3.2. Growth Medium
For all experiments,
M. marburgensis was used (DSM 2133), bought from German Collection of Microorganisms and Cell Cultures GmbH, Braunschweig, Germany. The medium was based on Schönheit et al. [
27] and was adapted according to
Table 2. The trace element solution (TES) was prepared according to recipe 141 (Modified Wolin’s mineral solution) of DSMZ at a ten times higher concentration.
As an indicator for the redox potential in biological systems, sodium resazurin was used. In its inactive form, it has a blue color. When first introduced to a medium near neutral pH, the blue color changes to pink (resorufin) in an irreversible step, usually initiated with heat. However, if resazurin is added to an alkaline solution, such as the hereby used medium with reducing agent, the color may stay blue. In the second, reversible step, the hydroresorufin/resorufin redox couple formed will be pink above −51 mV and colorless below −110 mV [
31,
32].
3.3. Setup Preparation
The cultivation was carried out in serum bottles (120 mL, bought from FloraCura, Garmisch-Partenkirchen, Germany), sealed with butyl rubber stoppers and aluminum crimp caps (both purchased from Ochs Laborfachhandel, Bovenden, Germany). The caps have an opening for injection needles, allowing the extraction of gas or suspension samples. After sealing, the reactor volume Vr was 118 mL. For anaerobic handling, an anaerobic chamber (AC, purchased from Toepffer Lab, Adelberg, Germany) was used with an atmosphere consisting of 95 vol% nitrogen and 5 vol.-% hydrogen and a palladium catalyst to regulate the oxygen concentration. For cultivation, a multiposition magnetic stirrer (MIXdrive 6 HT with MIXcontrol eco from 2mag AG, Munich, Germany) was used in an incubator (INCU-Line Prime IL 112, VWR International GmbH, Darmstadt, Germany) at 63 °C, and for each bottle a magnetic stirrer (rod-shaped, 25 × 6 mm) was used.
3.3.1. Parameter Experiment (PE)
The main parameters influencing the MER are the agitation speed and the suspension volume. To validate the performance of the parameters, the PE experiment was divided into several sub-experiments with varying agitation speeds (600, 800, 1000 rpm) and suspension volume (60 mL, 80 mL, 100 mL). For each sub-experiment, at least three batches were conducted; see
Table 3. Each batch consisted of six agitated bottles and one non-agitated control bottle as replicates. The general experimental setup is shown in
Figure 1.
The following procedure was conducted:
Culture preparation
- (a)
The medium was prepared in a 1 L bottle according to
Table 2, before being sealed and flushed with argon for 10 min. The bottle was then stored inside the AC.
- (b)
Seven serum bottles were sterilized by autoclaving (121 °C, 2 bar) for 20 min and then placed in the AC.
- (c)
As a starter, an inoculation culture (IC) was previously cultured in a glass bottle, incubated until fully grown (OD600 = 0.15) and placed in the AC before use. It should be noted that the IC temperature should be room temperature.
- (d)
The medium was filled in the serum bottles. Before adding IC, a 50 g/L Na2S·9 H2O stock solution was added to each bottle (1:100) and shaken to completely reduce the medium.
- (e)
IC was added and a magnetic stirrer rod was placed in each bottle. The bottles were sealed with a rubber stopper and ejected from the AC. The pH was not adjusted. The ratio of IC and medium was 1:2. The suspension volume was altered for each sub-experiment.
- (f)
With a special gas distribution station, the bottles were vacuumed (Vacuubrand VP 100 C) to 300 mbar and pressurized with the 80/20 hydrogen/carbon dioxide mixture. This routine was repeated two more times and the final pressure in the bottles was adjusted to a value of 5 bar absolute.
Sub-experiment
- (a)
Six of the seven bottles were placed on the magnetic stirrer inside the incubator, and one bottle was placed non-agitated as a control sample. The agitation speed was adjusted for each sub-experiment.
- (b)
The pressure of one random agitated bottle was recorded.
- (c)
The bottles were incubated for 3 h at 63 °C and then the pressure of each bottle was measured directly after removal from the incubator at incubation temperature.
- (d)
For the next batch, step 1f was repeated.
3.3.2. Reducing Agent Experiment (RE)
The culture preparation (step 1) for RE remained nearly the same, but the reducing agent and its concentration were changed. The reducing agent was freshly prepared for each concentration before addition to the medium. The reducing agent solution was not sterilized. Each variation was carried out in seven consecutive batches to gather a chronological progression. For RE, optimal cultivation conditions were chosen according to the PE results. The overall experimental setup is shown in
Figure 1.
3.4. Measurement Methods
The serum bottle pressure (absolute) was measured using a pressure sensor combined with a luer-lock adapter and a canula. The measured pressure data were transferred to a desktop PC and stored in a database. The sensor (MSD 6 BAE), device (GMH 5130) and software (EBS 20M 1.6) were obtained from GHM Messtechnik GmbH, Remscheid, Germany. For the RE experiment, the pH (HD 2156.1 with the electrode GE 100 BNC from GHM Messtechnik GmbH, Germany) and ORP (HD 2156.1 with the electrode GR 105 BNC from GHM Messtechnik GmbH, Germany) were measured in the AC at room temperature. ORP measurements during experiments often cannot exclude some possible poisoning of the electrodes. However, the electrode measurement was checked before and after the experiment at room temperature with a calibration solution (GRP 100, bought from GHM Messtechnik GmbH, Germany) to exclude any large deviations.
3.5. Calculation of the Methane Evolution Rate
The pressure of one random bottle was recorded during incubation and the pressure before (
) and after (
) incubation in the bottles was measured. The pressure difference was calculated as well as the incubation duration (
). The partial pressure for methane
can be calculated according to Equation (
1) as a result of stochiometric conversion. This method was adapted according to Taubner et al [
33]. According to the ideal gas law, the mole number of methane
can be calculated (see Equation (
2)) with
,
. The volume of the gas phase
is the difference between the known volume of the 118 mL bottle and the suspension volume, being 60, 80 or 100 mL. Finally, the methane evolution rate (MER) was calculated by dividing the amount of methane substance by the batch time and the reactor volume (see Equation (
3)).
5. Discussion
5.1. Cultivation Conditions
This study showed that the proposed method is suitable for determining the optimal cultivation conditions by varying the agitation speed and suspension volume for anaerobic archaea. The standard deviation of the data was quite low and the Welch’s t-test showed that the varied cultivation conditions had a significant impact on the MER. The highest MER was achieved with 800 rpm and 60 mL of suspension in 118 mL of serum bottles for M. marburgensis. However, if the suspension volume is low (e.g., for 60 mL), faster agitation induces shear forces on the microorganisms resulting in a lower MER.
The duration of cultivation is another important factor, as a high suspension volume (e.g., 100 mL) results in a fast conversion of a low amount of gas in a shorter time due to the high suspension to gas volume ratio (85:15). With a shorter cultivation duration, the MER for 100 mL is similar compared to the MER of 60 and 80 mL. However, the deviation of the values increases with a shorter cultivation time. Therefore, 3 h cultivation is preferred as the measured results are more consistent and less errors can occur.
Furthermore, it is possible to compare different organisms without the use of gas chromatography because the calculated MER provides a suitable comparison factor. Only a pressure sensor and the ideal gas law were used to calculate the MER values. Therefore, the method is very simple and inexpensive to perform and can be used in field experiments or less equipped laboratories.
5.2. Usage of Reducing Agents
The use of alternative reducing agents instead of sodium sulfide for M. marburgensis is promising, as several advantages can be drawn:
A lower reducing agent concentration performed better than a higher concentration. Surprisingly, 0.5 g/L sodium sulfide performed poorly, although it is the state of the art to use this reducing agent and concentration.
The MER remained stable in several batches with the use of sodium dithionite, compared to sodium sulfide, which decreases over time. This could also be related to the high redox potential of sodium dithonite and the chemical stability in an alkaline or neutral solution for several weeks, as stated by Telfeyan et al. [
35].
pH and redox potential showed that the medium used is well buffered to maintain a longer incubation time. However, a medium change is needed due to the formation and overload of surface-active metabolites, such as proteins that lead to foam formation, especially for bubble-column bioreactors [
36].
For sodium sulfide and sodium dithionite, a precipitate was observed, which sedimented in the bottles. This could be related to the high redox potential of both reducing agents. This high potential may lead to the binding of oxygen from functional groups or salts. However, a correlation between MER change and precipitation could not be drawn, but it may be that the high redox potential has influence on the organisms.
Overall, sodium dithionite and L-Cysteine-HCl performed similarly with respect to MER, which was also observed by Mylroie et al. [
11]. However, the redox potential of sodium dithionite is higher and does not influence the pH level as much as L-Cys-HCl. Furthermore, the reducing time of L-Cys-HCl is very high compared to sodium dithionite, which leads to a waiting time before the inoculum can be added to the medium.
The use of resazurin can be omitted because the redox potential stays at the same stage after several batches. As the medium is frequently changed, it is ensured that the redox potential remains at this level.
Sodium dithionite is a much more sustainable and non-toxic chemical (H251, H302) than sodium sulfide (H302, H311, H314, H400). Furthermore, sodium dithionite reduces much faster than sodium sulfide and therefore allows faster process setup and a shorter delay period, which was also observed by Widdel et al. [
19].
For further research, other reducing agents can be investigated and applied for biological methanation. It is important to note that an external source of sulfur, such as sulfite or hydrogen sulfide, is needed. This is provided by adding a reducing agent, as the use of sulfate ions (
) is not suitable for the cultivation of
M. marburgensis [
29,
37]. Furthermore, the (long-term) use of sodium dithionite and lower concentrations to prevent precipitation will be further investigated and reported.
5.3. Economic Considerations
Besides the MER performance of the reducing agents, economics should also be considered. With respect to the scale-up of biological methanation reactors, larger amounts of chemicals need to be supplied. For example, with a concentration of 0.25 g/L and a reactor of 10 m³, an amount of 2.5 kg of reducing agent is needed. In
Table 6, the costs of the three used reducing agents for different suppliers are shown. The prices are changing due to quality, producer, amount and energy prices. However, it can be seen that L-Cys-HCl is a highly priced chemical, whereas sodium dithionite has the lowest cost, around EUR 60 to 70 per kilogram, and sodium sulfide is twice the price. In the future, a life cycle assessment of the chemicals used for biological methanation should be performed to provide further information on the production and treatment of waste water for such chemicals.