Absorptive Desulfurization of Model Biogas Stream Using Choline Chloride-Based Deep Eutectic Solvents

: The paper presents a synthesis of deep eutectic solvents (DESs) based on choline chloride (ChCl) as hydrogen bond acceptor and phenol (Ph), glycol ethylene (EG), and levulinic acid (Lev) as hydrogen bond donors in 1:2 molar ratio. DESs were successfully used as absorption solvents for removal of dimethyl disulfide from (DMDS) from model biogas steam. Several parameters affecting the absorption capacity and absorption rate has been optimized including kind of DES, temperature, the volume of absorbent, model biogas flow rate, and initial concentration of DMDS. Furthermore, reusability and regeneration of DESs by means of adsorption and nitrogen barbotage followed by the mechanism of absorptive desulfurization by means of density functional theory (DFT) as well as FT-IR analysis were investigated. Experimental results indicate that the most promising DES for biogas purification is ChCl:Ph, due to high absorption capacity, relatively long absorption rate, and easy regeneration. The research on the absorption mechanism revealed that van der Waal interaction is the main driving force for DMDS removal from model biogas. area is around the nitrogen atom and electronegative areas are around Cl, O, and S atoms. During the absorption process, the electronegative region located around sulfur atoms in the DMDS molecule interacts with the electropositive area located around nitrogen atom in ChCl:Ph molecule. These interactions provide efficiency DMDS removal from model biogas. On the other hand, it can be concluded that HBA in the DES molecule has the greatest impact on the purification process. This phenomenon will be investigated in subsequent works. Similar results can be found in ChCl:Lev-DMDS and ChCl:EG-DMDS complex.


Introduction
Currently, more and more attention is paid to the production of alternative high-quality fuels, including bio-methane, bio-hydrogen, bio-ethanol, bio-butanol, etc. Biogas is a modern form of bioenergy, which can be obtained in the process of dark fermentation using waste products from various industries and agriculture (agri-food and animal waste) [1][2][3]. Biogas usually contains 30-60% v/v CH4, 15-30% v/v CO2, 5-20% v/v N2, 1-10% v/v O2 and about 1-2% v/v of other contaminants including H2S, NH3 and numerous of organic compounds [4][5][6]. Among the organic pollutants, volatile organosulfur compounds (VSCs) are one of the most important biogas impurities group. VSCs include sulphides, disulphides, thiophenes, and thiols, and the concentrations of them are strictly dependent on the raw material used for biogas production [7]. A typical range of VSCs concentration in biogas is shown in Table 1. compressor with a membrane dryer (Ekkom, Poland) and hydrogen (purity N 5.5) generated by a 9400 Hydrogen Generator (Packard, USA) were used for the preparation of model biogas, regeneration process, and chromatographic analysis.

Preparation of DES
DESs were synthesized by mixing ChCl (HBA) with Ph, EG, and Lev (HBD) in a 1:2 molar ratio. The mixtures were stirred magnetically at 65 °C until homogeneous liquids were obtained. The liquids were then left to cool spontaneously to room temperature. The physicochemical properties of the synthesized DES were presented in Table 2.

. Absorption process
The absorption process was prepared by means of the barbotage phenomena. Nitrogen (model biogas stream) was passed through a 20-mL vial containing 5 mL of DMDS. The created mixture (nitrogen -DMDS) was diluted with a nitrogen stream to 1.0 mg/Nm 3 (DMDS) concentration. The model biogas stream containing DMDS was passed through an absorption column (total volume 60 mL) containing 50 mL of DES. The total flow of gaseous DMDS and nitrogen was kept constant at 50 mL/min. The concentration of DMDS was monitored at the inlet and outlet of the barbotage column using GC-FID. The processes were carried out for 1200 minutes. The absorptivity (A) of DMDS in the DES was calculated using equation (1): -DMDS concentration after absorption process [ppm v/v].

Regeneration of DESs
After the absorption process, DESs were regenerated using two popular methods, including nitrogen barbotage and adsorption. Nitrogen barbotage was carried on as follows, 4 mL of DES was barbotaged using nitrogen flow 50 mL/min for 2.5 or 5 hours. Three types of adsorbents i.e. AC, SG, and AO were used in the second type of regeneration process. All adsorbents were activated in a laboratory dryer at 120 °C for 2 h. The 4 mL of DES containing DMDS was mixed with 160 mg and 420 mg adsorbents in a vial. The vials were maintained in a laboratory shaker at 25 °C for 30 minutes, subsequently centrifuged for 5 min at 7000 rpm and filtered through a 0.45 µ m cellulose filter. The concentration of DMDS in DES (before and after regeneration) was controlled using static headspace coupled to gas chromatography (SHS-GC).

Chromatographic analysis
GC temperature program was 120 °C ; injection port temperature was 300 °C carrier gasnitrogen (2 mL/min); injection mode: split 20:1; detector temperature was 300 °C ; detector gases flow rates were hydrogen 40 mL/min and air 400 mL/min. DESs (2 mL) after regeneration was thermostated at 80 °C for 50 min. Then 0.1 mL of gas phase was introduced into the GC injector. In order to monitor DMDS concentrations during the absorption process, 0.5 mL of model biogas was analyzed.

Theoretical studies
The molecular structures and interactions between DES and DMDS were optimized using B3LYP/6-311++G** level of theory with dispersion -corrected computational model using Orca 4.1.1 software package. All configurations were optimized for local minima using frequency calculations. The interaction energy for the gas phase between the DES and DMDS was calculated according to the following expression (2): Where: The counterpoise method has also been implemented to estimate the effects of the basis set superposition error (BSSE) on the interaction energy, based on previous studies [43]. Electrostatic potential analysis (ESP), and reduced density gradient analysis (RGD) were performed to the visual interpret the interaction nature in the DES-DMDS complex. Both, RDG and ESP analysis were performed using Multiwfn software [44][45][46]. In order to the graphical presentation of the results, the Visual Molecular Dynamics 1.9.3. the software was used.

Optimization of absorption conditions
Optimization of absorption conditions using deep eutectic solvents was carried out for DMDS as the main representative of volatile organosulfur compounds commonly found in real biogas streams [7,[47][48][49][50]. The process was optimized in terms of DES type, temperature, the volume of DES, model biogas flow, and initial concentration of DMDS.

Kind of DES
The selection of the absorption solvent is particularly important in the absorptive desulfurization process. Three types of DES were tested, including ChCl:Ph, ChCl:Lev, ChCl:EG in 1:2 molar ratio ( Figure 1). In the experiment, the following pre-selected absorption conditions were used: 50 mL of DES; initial concentration of DMDS 1 mg/Nm 3 , 25 °C temperature; model biogas flow rate 50 mL/min. Among the investigated DESs, ChCl:Ph shows the best absorption efficiency. After 800 minutes, the absorptivity value for ChCl:Ph was below 0.3 and then increased rapidly over the next 300 minutes, which indicates DES saturation. The saturation time of the other two DES was 600 and 800 min for ChCl:EG and ChCl:Lev, respectively. It can be noticed that the best result was obtained for DES which has the lowest viscosity value. As the viscosity increased, both the absorption capacity and absorption rate decreased. In DES with higher viscosity, the mass transfer is hindered, therefore it is preferable to use solvents with the lowest viscosity.
It is worth noting that the viscosity of the DES increases when the amounts of hydroxyl groups in HBD increases. The existence of extra hydroxyl groups creates a more extensive hydrogen bond network which results in lower mobility of free species within the DES [28]. However, many other factors (except viscosity) also influence the DMDS removal efficiency, therefore in section 3.2., the mechanism of absorptive desulfurization of model biogas was explained.

Volume of DES
In the studies, three-volume of DES in the range of 15-50 mL/min were investigated (Figure 2a). In the studies, other operating parameters i.e. inlet concentration DMDS 1 mg/Nm 3 , nitrogen flow rate 50 mL/min and the temperature of process 25 °C, were constant. The results showed that with the increase in DES volume, the absorption efficiency increased significantly from 695 min to 1200 min. This is due to the fact that as the volume of the absorbent increases, the contact time between the gas phase and the liquid increases, due to which the saturation time is longer. Similar results were obtained in the work [51,52]. The results indicate that the flow rate has a large impact on the overall DMDS capture process. As the flow rate increases from 20 to 70 mL/min, the DES saturation time is reduced from 1200 to 729 min. Similar results were obtained in the research [51][52][53]. This is due to the fact that as the flow velocity increases, the contact time of the polluted gas stream with the absorbent is reduced, which adversely affects the DMDS absorption process. However, only a slight change in saturation time is observed between the flow of 20 and 50 mL/min. Therefore, a 50 mL/min model biogas flow rate was considered as optimum value.

Initial concentration of DMDS
In the studies, the initial concentration of DMDS in the range of 0.1-100 mg/Nm3 was investigated (Figure 2c). The results indicate that the DMDS absorption efficiency remains fairly stable despite an increase in DMDS inlet concentration. This valuable result shows the ability of ChCl:Ph to the removal of DMDS in varying concentrations from real biogas steam which makes it desirable from an industrial point of view. Similar results were obtained in the work [54].

Temperature
Three temperature values i.e. 25, 40 and 60 °C were chosen to assess the influence of temperature on DMDS absorption behavior (Figure 2d). Theoretically, an increase in temperature 7 of 18 affect the decreases in DES viscosity and hence gas transfer rate are improved. The absorption curve reveals that increasing DES temperature, that solubility of DMDS decrease. The decrease in solubility can be explained by the fact that the gas absorption process is normally exothermic. Therefore, the preferred temperature using DES is 25°C, because in using this temperature the longest effective purification time for the biogas stream is achieved, lasting up to 1200 min. When the temperature is increased to 60 °C, the absorption time is reduced to 1010 min. Consequently, the absorption process can be performed at room temperature with minimal energy consumption. Similar results were obtained at work [55].

FT-IR analysis
The experimental research on the mechanism of the absorption process was performed by FT-IR analysis. The spectra of pure DESs were compared with pure DMDS, and DES-DMDS complex spectra (Figure 3a,b,c). All characteristic bands that can be attributed to DMDS (2909.32 cm -1 -δs (CH) stretch, 1411.73 cm -1 -δas (CH3) def., 1302.68 cm -1 -δs (CH3) def., 692.82 cm -1 -C-S stretch, 540.96 cm -1 -S-S stretch) are visible in the ChCl:Ph-DMDS spectrum [56], which indicates the creation of the DES-DMDS complex (Figure 3a). Theoretically, in the absorptive DMDS removal process, both sulfur atoms can act as a donor in S-H···π and as an acceptor in O-H···S and C-H···S interactions [57][58][59]. However, on the FT-IR spectra, there are no shifts corresponding to this type of interaction. Therefore, other interactions must play a key role in the DMDS absorption process. Similar results were also obtained for ChCl:Lev ( Figure 3b

Molecular modeling
In order to better understand the mechanisms of DMDS removal from the gas phase using DES, the density functional theory (DFT) was applied. For this purpose, the most probable and stable configurations in the gas phase of ChCl:Ph-DMDS, ChCl:Lev-DMDS, ChCl:EG-DMDS was Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 18 February 2020 doi:10.20944/preprints202002.0266.v1 geometry optimized at the B3LYP/6-311++G** level of theory ( Figure 4). The results indicate that in all complex, nonbonded interaction exists between choline and chloride atom (O-H···Cl), which can be identified as strong hydrogen bond because of short distance (below 2.5 Å ) [60]. Hydrogen bonds also occur between Cl atom and two phenol molecules such as  Electrostatic potential analysis (ESP) was used for visualization of total charge distribution and relative polarity of the studied DMDS structure and DESs-DMDS complexes. The ESP of DMDS, ChCl:Ph-DMDS, ChCl:Lev-DMDS, and ChCl:EG-DMDS are mapped onto their electron densities in Figure 5. The results indicate that the electropositive areas are around the hydrogen atoms whereas the electronegative area is around sulfur atoms, in the DMDS structure. In the ChCl:Ph-DMDS complex large electropositive area is around the nitrogen atom and electronegative areas are around Cl, O, and S atoms. During the absorption process, the electronegative region located around sulfur atoms in the DMDS molecule interacts with the electropositive area located around nitrogen atom in ChCl:Ph molecule. These interactions provide efficiency DMDS removal from model biogas. On the other hand, it can be concluded that HBA in the DES molecule has the greatest impact on the purification process. This phenomenon will be investigated in subsequent works. Similar results can be found in ChCl:Lev-DMDS and ChCl:EG-DMDS complex. The Reduced density gradient (RDG) is a beneficial approach to distinguish and visualize various types of noncovalent interactions in real space. In the studies, the RDG analysis was used to visualized weak noncovalent interactions (i.e. hydrogen bond, van der Waals interaction, and repulsive effect), by plotting the RDG versus the electron density multiplied by the sign of the second Hessian eigenvalue, based on previous studies [44]. In Figure 6 b,d,f the green surfaces indicate Van der Waals interaction, red surfaces indicate strong repulsion, and blue surfaces indicate H-bond. The obtained data show that the three hydrogen bonds, as well as van der Waals interaction, were formed between HBA and HBDs in all studied DES, which corresponding to large, negative sign(λ2)ρ value (from -0.04 to -0.02 au) and 0.01 au < sign(λ2)ρ < 0.01 au, respectively in 2D diagrams (Figure 6 a,c,e). Furthermore, in ChCl:Ph-DMDS, strong repulsive bonds occur (sign(λ2)ρ = 0.02 au), due to the presence of an aromatic ring in the phenol molecule. In all studied complex, between DES and DMDS only van der Waal interaction occur. The surfaces of van der Waal interaction increase follow the order of ChCl:EG-DMDS < ChCl:Lev-DMDS < ChCl:Ph-DMDS. This indicates that the main driving force affecting DMDS removal from model biogas is van der Waals interactions. The calculated interaction energy in the gas phase between DES and DMDS were -10.8, -3.7, and -9.6 kcal/mol for ChCl:Ph-DMDS, ChCl:Lev-DMDS, and ChCl:EG-DMDS, respectively. The lower interaction energy values stand for stronger interaction between DES and DMDS. The obtained data followed a similar trend to an experimental data ChCl:Ph-DMDS < ChCl:EG-DMDS < ChCl:Lev-DMDS.

Regeneration and reusability of DES
For the industrial point of view, the regeneration of DES is an essential and significant factor because it has a great impact on the operating cost. Therefore, regeneration of DES was carried on through one of the best-known regenerative methods i.e. was nitrogen barbotage (NG) (which was carried for 2.5 and 5 hours) and adsorption process (with different types and amount of adsorbents). In the adsorption process, three types of adsorbents were tested including SG, AC, and AO, in the amount of 160 and 420 mg which were added to 4 mL of each DES and shaken for 30 min. The results indicate, that the nitrogen barboage is the most efficiency desorption method. Desorption efficiency of DMDS from all DES after 5 hours is higher than 99.999%. From ChCl:Ph, DMDS can be completely removed after 3 hours (Figure 7). For the remaining DES, this time should be extended, which would affect the cost of the process. In addition, the results show, that slightly lower regeneration efficiencies for all DES were obtained using the adsorption process. The regeneration efficiency were in the range of 87.8 -84.5 % for ChCl:Ph, 87.1 -63.4 % for ChCl:EG, and 96.6 -91.3 % ChCl:Lev. The highest absorption efficiency was obtained for 420 mg of SG. In most cases, the increase in adsorbent mass relative to the amount of DES containing DMDS resulted in increased adsorption efficiency, due to the increased adsorption surface. However, from an economical industrial point of view, the adsorbent amount should be as low as possible. Desorption is the most effective method to remove DMDS from ChCl:Lev. This is probably due to the low sorption capacity of ChCl:Lev and the absorption relatively small amount of DMDS.
The absorption−desorption results indicate, that DMDS could be completely removed from DES and absorbent could be re-use minimum five times without significant loss of absorption capacity DMDS (Figure 8a). In order to examine whether there were any structural changes in DES after the regeneration process (nitrogen barbotage), FT-IR analysis was used (Figure 8b,c,d). No additional peaks or shifts are observed in the DES spectra before and after regeneration, which indicates DES stability during the regeneration process.

Conclusions
The choline chloride-based deep eutectic solvents have been synthesized and tested as alternative eco-friendly and green solvents for absorptive desulfurization of model biogas. Effect of selected absorption parameters including kind of DES, temperature, absorbent volume, model biogas flow rate and initial concentration of DMDS were studied. It was found that the optimum absorption parameters for DMDS removal were absorption solvent ChCl:Ph in 1:2 molar ratio, 50 mL of DES, temperature 25 °C, and 50 mL/min flow rate. The influence of DMDS concentration indicates that the initial amount of DMDS in model biogas has only a minor effect on the absorption capacity and rate. In the optimum conditions, DMDS was removed with high efficiency for 800 minutes. After this time, the gradual saturation of DES occurred. After the absorption process, Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 18 February 2020 doi:10.20944/preprints202002.0266.v1 ChCl:Ph in 1:2 molar ratio could be regenerated by means of nitrogen barbotage and reuse without loss absorption capacity. The studies on the absorptive desulfurization mechanism indicate that van der Waal interaction is the main driving force for the efficient removal of DMDS from model biogas.
The developed absorption process with the choline chloride-based deep eutectic solvents provides a promising alternative method for the removal of volatile organosulfur compounds from the real biogas stream.
The paper presents preliminary results of research on the removal of DMDS from a model biogas stream. However, due to the rich composition of real biogas samples, other volatile organic compound groups should also be included in future studies. In addition, in the future, to verify the suitability of the developed method, the studies using real samples of heavily contaminated biogas streams from sewage treatment plants and landfills will be carried out.