Mechanistic Understanding on Difluoromethane Absorption Thermodynamics on Novel Deep Eutectic Solvents by COSMO-Based Molecular Simulation

Abstract

Hydrofluorocarbons (HFC) are fluorinated compounds used globally for refrigeration. These gases have been shown to contain a greenhouse potential of up to 22,000 times that of CO₂. Thus, 1298 type-5 deep eutectic solvents (DES) were examined for the absorption and interaction mechanisms of difluoromethane (R32), due to their non-polar attributes. Of these solvents, quaternary ammonium salts mixed with various species of hydrogen bond donators (HBD) produced the most favorable interactions, with ln activity coefficients predicted to be as low as −1.39 at 1:1 compositional ratio. These DES were further studied for compositional analysis where pure tetrabutylammonium bromide showed the strongest interaction potential. The pressure study showed a linear solubility increase with a pressure increase reaching up to 86 mol/mol% in a methyltrioctylammonium bromide and polyethylene glycol mixture at 9 bar. The van der Waals interaction is the driving force of absorption with ~3x enthalpic release over hydrogen bonding. All chemicals contain strong potential for an environmentally friendly solution, as is evident through an environmental health and safety analysis.

1. Introduction

Hydrofluorocarbons (HFC) are a class of fluorinated compounds used primarily in refrigeration, foam-blowing agents, aerosol propellants, solvents and fire suppressants [1]. These compounds were encouraged after policy change through the Montreal Accord on conventional chloro-fluorinated carbons (CFCs) due to ozone depletion properties [2,3,4]. While HFCs are considered a better alternative to the previous components regarding ozone depletion, due to the hydrogen allowing for much faster breakdown periods in the atmosphere, they are still persistent in the air, lasting between 15 and 270 years compared to CFCs that can last over 1000 years [5,6]. These gases also have an extreme global warming potential (GWP), being 677 times as potent as carbon dioxide in their greenhouse potential per 100 years in the atmosphere [5,7]. In 2020, the US reportedly emitted the CO₂ equivalent of 190 million metric tons of fluorinated gases [5]. Thus, providing recycling alternatives to the current separation processes is of great and immediate importance [8].

Difluoromethane is among the most used HFC today for refrigeration and fire suppression. Difluoromethane has received attention from the scientific community for capture and separation through absorption and adsorption, recently through the use of ionic liquids [9]. Complex novel solvents are required for processing HFC refrigerants such as difluoromethane [10,11]. Ionic liquids, a group of novel solvents, are generally composed of expensive, toxic, or non-biodegradable chemicals which can be detrimental to the environment [12,13]. Contrarily, deep eutectic solvents (DES) are considered analogues of ionic liquids as they possess many similar physicochemical properties, yet offer cheap, non-toxic, and environmentally benign alternatives [12]. Rather than ionic pairs, DES are formed through combinations of hydrogen bond acceptors (HBA) and hydrogen bond donors (HBD), which develop a significantly depressed eutectic point generally causing them to be liquid at room temperature [14,15]. DES are currently separated into five types. Two of these types (3 and 5) are attributed with being green solvents, while the other classes contain potentially hazardous metal components [16]. Type five DES are also termed hydrophobic DES or HDES. Due to the relatively nonpolar nature of fluorinated compounds, HBA and HBD, which are prevalent HDES, are examined in this study as they likely ensure the best solubility.

Thus far, DES have been extensively studied in their capacity as common greenhouse gas absorbents such as CO₂ and CH₄, gaseous pollutants such as siloxanes, and sulfur contaminants, continuously outperforming conventional solvents in studies [14,15,16,17]. DES have been proven to be highly effective in the capture of greenhouse gases (GHG) [18,19]. However the application of DES for HFCs, and especially difluoromethane, has yet to be explored as best to the authors knowledge. One barrier for the experimental investigation of DES in this application is the vast number of available HBA, HBD, combinations, and mixture compositions. Thus, to aid in initial explorations of DES–difluoromethane systems, many researchers have relied upon thermodynamic property prediction methods [20,21,22]. The conductor-like screening model (COSMO) suite is a thermodynamic property prediction method which employs quantum mechanical calculations, density functional theory, and QSPR methods in order to determine ab initio predictions [23]. COSMO is especially popular among researchers studying ionic liquids and DES, as one remarkable trait of these solvents is their tunability, leading to virtually infinite possible combinations of components and compositions [24,25,26,27].

The objectives of this study are to determine the feasibility of difluoromethane absorption using type 5 DES, determine what types of solvent components work through ln activity coefficient analysis, determine why these solvent components work through energetic absorption mechanism analysis with the use of sigma analysis, and to determine whether the promising candidates are indeed “green” alternatives through an environmental health and safety (EHS) analysis, utilizing chemical SDS and the VEGA model. The sigma analysis includes a sigma profile and sigma potential investigation. A sigma profile is the probability distribution of a molecular surface segment having a specific charge density as computed for a molecule [22,23]. A sigma potential plot is the graphing of the chemical potential resulting from the computed molecule being placed in an environment of a specific charge [22,23].

2. Materials and Methods

2.1. COSMO Simulation

A total of 1298 solvent combinations composed of type 5 HBA (total of 22) and HBD (total of 59) at a 1:1 ratio were used in this study (Table S1). The HBAs consist of quaternary ammonium salts, quaternary phosphine salts, terpenes, and imidazolium salts. The HBDs consist of alcohols, organic acids, and profens. These components all exhibit significant van der Waals-capable surfaces which give them the potential to form a hydrophobic DES complex. The sequence of computations was performed using the COSMOLogic platform in series: TurbomoleX (TmoleX19), COSMOConfX (COSMOConf18), and COSMOThermX. The molecules that were not available in the COSMOtherm onboard database were input to TmoleX19 in the SMILES format, as found from PubChem. The lowest geometrical energies were computed along with sigma surfaces at the bp86 functional and Karlsruhe (Ahlrichs) def2-TZVP (default-2 Valence Triple-Zeta Polarization) basis set, as recommended in literature [28]. The output files (.cosmo extension) were input to COSMOConfX for conformer analysis at the same theory levels. The results were then uploaded to COSMOThermX for property predictions of the studied systems.

Equation (1) was used to compute chemical potentials $\left({\mathsf{\mu }}_{\mathrm{s}}\left(\mathsf{\sigma }\right)\right)$ of the DES–difluoromethane complex by means of sigma surface interactions, which would be used as a basis for successive property predictions.

$${\mathsf{\mu }}_{\mathrm{s}}\left(\mathsf{\sigma }\right)=-\frac{\mathrm{RT}}{{\mathrm{a}}_{\mathrm{eff}}}\ln [\int {\mathrm{p}}_{\mathrm{s}}\left( {\mathsf{\sigma }}^{\prime }\right){\mathrm{e}}^{\left(\frac{{\mathrm{a}}_{\mathrm{eff}}}{\mathrm{RT}}\left({\mathsf{\mu }}_{\mathrm{s}}\left( {\mathsf{\sigma }}^{\prime }\right)-{\mathrm{E}}_{\mathrm{misfit}}\left(\mathsf{\sigma }, {\mathsf{\sigma }}^{\prime }\right)-{\mathrm{E}}_{\mathrm{HB}}\left(\mathsf{\sigma }, {\mathsf{\sigma }}^{\prime }\right)\right)\right)}\mathrm{d}{\mathsf{\sigma }}^{\prime }]$$

(1)

where ${\mathrm{E}}_{\mathrm{misfit}}$ is the misfit energy, hydrogen bond energy is E_HB, and van der Waals energy is E_vdw; $\mathsf{\sigma } \mathrm{and} {\mathsf{\sigma }}^{\prime }$ are two interacting surface segments between two molecules x and ${\mathsf{x}}^{\prime }$, and ${\mathrm{p}}_{\mathrm{s}}\left(\mathsf{\sigma }\right)$ is the distribution function or sigma profile. R is the ideal gas constant and T is absorption temperature. ${\mathrm{a}}_{\mathrm{eff}}$ is the area of effectiveness term.

The activity coefficient $\left({\mathsf{\gamma }}_{\mathrm{S}}^{\mathrm{i}}\right)$ was calculated through the difference in the chemical potentials of component difluoromethane in solvent (s) and pure form (p), as seen in Equation (2).

$$\ln {\mathsf{\gamma }}_{\mathrm{s}}^{\mathrm{R}32}=\frac{\left({\mathsf{\mu }}_{\mathrm{s}}^{\mathrm{R}32}-{\mathsf{\mu }}_{\mathrm{p}}^{\mathrm{R}32}\right)}{\mathrm{RT}}$$

(2)

In Equation (3), Henry’s constant (H) was used for the validation procedure and was calculated through an iterative process of varying the pressures (P) with the concentration (C).

$$\mathrm{C}=\mathrm{H}\times \mathrm{P}$$

(3)

The predicted solubility of R-32 in DES is produced through Equation (4).

$${\mathrm{x}}_{\mathrm{R}32}=\frac{{\mathrm{p}}_{\mathrm{R}32}}{{\mathrm{p}}_{\mathrm{R}32}^{\mathrm{o}}\times { \mathsf{\gamma }}_{\mathrm{R}32}}$$

(4)

where ${\mathrm{p}}_{\mathrm{j}}$ is the partial pressure of R-32 in the DES system, ${\mathrm{p}}_{\mathrm{j}}^{\mathrm{o}}$ is the partial pressure of R-32 in its pure form, ${\mathrm{x}}_{\mathrm{j}}$ is the solubility of R-32 in HDES, and ${\mathsf{\gamma }}_{\mathrm{j}}$ is the activity coefficient of R-32 in solvent.

Finally, the excess enthalpy of mixing ${\mathrm{H}}_{\mathrm{mix}}$ was determined through the addition of three enthalpic contributors available in Equation (5).

$${\mathrm{H}}_{\mathrm{mix}}={\mathrm{H}}_{\mathrm{mf}}+{\mathrm{H}}_{\mathrm{hb}}+{\mathrm{H}}_{\mathrm{vdw}}$$

(5)

here, these contributions are from enthalpic change due to the misfit of sigma charge segments ${\mathrm{H}}_{\mathrm{mf}}$, enthalpic change due to hydrogen bonding (${\mathrm{H}}_{\mathrm{hb}}$), and contribution due to van der Waals interactions (${\mathrm{H}}_{\mathrm{vdw}}$).

2.2. Environmental Health and Safety

The “Virtual models for property Evaluation of chemicals within a Global Architecture” (VEGA) [29,30] K-Nearest Neighbors (KNN) environmental health and safety (EHS) predictive model was used was used in the evaluation of persistence, bioconcentration factor (BCF), mutagenicity, carcinogenicity, and acute toxicity for selected DES components. VEGA has been relied upon for the EHS property analysis of novel solvents including DES [31,32,33]. The model requires an input of each molecule in the Simplified Molecular Input Line Entry System (SMILES) format. The structure of the input molecule is matched with stored experimental values of molecules with the most similar structures in VEGA, and a prediction is made from this along with a measure for reliability. All involved chemicals were cross evaluated with the safety data sheets from Fischer Scientific. Much data were unavailable in SDS, making the VEGA model an important inclusion to fill the gaps in the literature. The National Fire Protection Association (NFPA) value rankings and the Occupational Safety and Health Administration (OSHA) Hazard Communication Standard (OHCS) categorization are used from the literature for the validation of VEGA results when available.

3. Results and Discussion

3.1. COSMO Validation

Due to the novelty of the systems studied, little to no DES–difluoromethane system data was readily available to the authors. Thus, the approach for validating the COSMO system was to compare solubility data for difluoromethane in ionic liquid literature data, of which DES are considered analogues. This approach has been reportedly used in such novel systems where experimental data are unavailable, as is shown through the work of Abedin et al. [34]. The benchmarking data set consists of 42 molar solubility datapoints in various ionic liquids at various pressures at 298.15 k as reported by Shiflett et al. [35]. The ionic liquid components were generated using the same methods reported in Section 2.1, along with the solubility computations. These generated values were regressed with the experimental values with a results R² value of 0.86. The same data set produced an absolute average relative deviation of 25.7% average absolute relative deviation (AARD). Figure 1 contains the resulting datapoints from a y-axis of experimental values, and an x-axis of COSMO predicted values from this work. While the single point error (AARD) is relatively high, the qualitative information predicted about these systems is well preserved as is evident from the high R-squared value. Thus, for the purpose of this study into the relative effectiveness of the solvents and energetic mechanisms driving them, COSMO is deemed an appropriate method of examination.

3.2. Evaluation of Novel HDES Solvent Combinations for R-32 Absorption

Table 1. In activity coefficients for 1:1 ratio of top performing type 5 DES HBA (columns) and HBD (rows). All computed activity coefficients are available in Figure S1.

	N4Br	BTACl	N8Br	N81Br
PEG	−1.39	−1.28	−1.11	−1.04
Camphor	−1.34	−1.20	−1.07	−1.01
Lidocaine	−1.27	−1.16	−1.06	−1.00
Eucolyptol	−1.27	−1.11	−1.02	−0.94

contains ln activity coefficients for the top preforming DES combinations and R32 systems. The entire data from the study may be found as a heatmap with values in Figure S1 of the Supplementary Files. Interestingly, all DES combinations provide ln activity coefficients below 1, suggesting non-ideal attractive forces. All compositions are kept at 1:1 molar ratio with a temperature of 25 °C and a pressure of 1 bar. The top four HBA in rank of lowest ln activity coefficients to highest were found to be tetrabutylammonium chloride (N4Br) (−1.39), tetraoctylammonium chloride (N8Br) (−1.39), benzyltriethylammonium (BTACl) (−1.27), and methyltrioctylammonium bromide (MTOA) (−1.03). These HBA outperformed the terpenes, shorter chain quaternary ammonium salts, imidazoles, and phosphonium-based HBAs. These four HBA share a common trait of being the largest molecules. The results are logical as they contain significant amounts of carbon, which offers non-polar interaction sites. The results are in line with the sigma analysis in Section 3.3. Other works have shown that longer chain quaternary ammonium salts have higher absorption capacities for non-polar gases vs. their shorter chain counterparts [12].

The accompanied HBDs that produced the lowest ln activity coefficient with HBA are polyethylene glycol (PEG), camphor, eucalyptol, and lidocaine. Three of these are terpenes and one is a glycol. These four HBD share the property of lacking polarity, which is favorable for a relatively non-polar molecule such as difluoromethane. The results are reasonable due to the electronic signatures of difluoromethane and the DES components, discussed in detail in Section 3.3. These types of DES components are utilized in non-polar absorption scenarios and aqueous extractions. The rest of the studies are performed with the 16 HBA–HBD combinations of these solvents.

3.3. Absorption Mechanism through Enthalpy and Sigma Analysis

Figure 2 contains the sigma profiles for the studied DES components following the primary y-axis. The x-axis is the sigma segment charge while the primary y-axis is the frequency of the segment on the molecule. The secondary Y-axis is the chemical potential of difluoromethane when interacting with specific charged surfaces represented on the x-axis, which are represented as a dotted line. Negative chemical potential values correspond to energetically favorable interactions between difluoromethane and a given surface charge. It is evident that the difluoromethane prefers to interact with van der Waals and hydrogen bond accepting surfaces. The parabolic nature indicates repulsion from sufficiently negative or positively charged surfaces with a hydrophobic attribute [23]. The sigma profiles of the HBA and HBD all show large peaks in the non-polar region, with varying peak tail areas in the polarized segments (<−0.0078 e/A², >0.0078 e/A²) [36]. This is expected due to the hydrophobic nature of type 5 DES which these components comprise. This compliments the difluoromethane interaction preferences, which explains the relatively low ln activity coefficient values witnessed for all solvent combinations as discussed earlier. However, the defining property for why N4Br, N8Br, BTACl, and MTOA outperformed the other HBA lies in the surface charge distributions. The four studied HBD (eucalyptol, polyethylene glycol, camphor, and lidocaine) are also included in Figure 2 and are represented as having remarkably similar sigma profiles to each other.

Table S2 contains surface area charge distribution as a percentage of the total surface area in terms of hydrogen bond donating, accepting, and van der Waals interacting surfaces. This was achieved through the integration of the sigma profile for each respective molecule with limits representing the charge regions of <−0.0078 e/Ų (hydrogen bond donating region), >0.0078 e/Ų (hydrogen bond accepting region), and −0.0078 to 0.0078 e/Ų (van der Waals interactions region). Following the discovery from the sigma potential plot of difluoromethane, [TETA]Cl and TOPO were the worst performing of the HBA and contained the highest surface area percentage in the hydrogen bond donating region, which supplies a repulsive charge to difluoromethane. However, N4Br, N8Br, BTACl, MTOA, camphor, PEG, lidocaine, and eucalyptol all lack significant area in the hydrogen bond donating region (<6%), have the majority of their area in the non-polar region (>80%), and some area in the favorable hydrogen bond accepting region (6% < x < 15%).

Table 2. Enthalpy of mixing analysis where H_int is the excess enthalpy of interaction, H_mf is the enthalpic contribution due to misfitting of sigma segments, H_hb is the enthalpic contribution from hydrogen bonding, and H_vdw is the enthalpic contribution from van der Waals interactions. All the enthalpies are presented as kJ/mol.

HBD	Property	N4Br (kJ/mol)	BTACl (kJ/mol)	N8Br (kJ/mol)	MTOA (kJ/mol)
PEG	Hint	−1.15158	−1.12681	−1.15158	−1.12756
PEG	Hmf	1.4377	1.44851	1.4377	1.42068
PEG	Hhb	−0.40207	−0.43892	−0.40207	−0.37627
PEG	Hvdw	−3.13471	−3.08391	−3.13471	−3.11947
Camphor	Hint	−1.12736	−1.10647	−1.15158	−1.11496
Camphor	Hmf	1.45191	1.45861	1.4377	1.40033
Camphor	Hhb	−0.39266	−0.42837	−0.40207	−0.33878
Camphor	Hvdw	−3.13412	−3.0842	−3.13471	−3.12401
Eucalyptol	Hint	−1.14948	−1.13061	−1.12736	−1.09226
Eucalyptol	Hmf	1.46866	1.47271	1.45191	1.41422
Eucalyptol	Hhb	−0.41278	−0.4479	−0.39266	−0.33038
Eucalyptol	Hvdw	−3.15286	−3.10292	−3.13412	−3.1236
Lidocaine	Hint	−1.12756	−1.1175	−1.14948	−1.11091
Lidocaine	Hmf	1.42068	1.4228	1.46866	1.42843
Lidocaine	Hhb	−0.37627	−0.40817	−0.41278	−0.3468
Lidocaine	Hvdw	−3.11947	−3.07964	−3.15286	−3.14005

contains the excess enthalpy of interaction computations for 16 combinations of solvents reflecting the results of Section 3.2. This table contains H_int, H_mf, H_hb, and H_vdw values. As expected from the sigma analysis, the most energy is released through van der Waals interactions, with up to ten times the energetic release from favorable hydrogen bonding. The H_vdw values are all approximately −3 and the H_HB values are all between −0.3 and −0.5. All values across components represent a variance range of 15% between components in the same computation values, thus suggesting similar performance capacity between these solvents and difluoromethane despite the variation in DES components. The H_mf values are all 1.4 kJ/mol. This suggests that little steric hindrance and charge variance occurs between these DES–difluoromethane systems, overall contributing to negative H_int values. However, the H_mf values are more significant than the H_HB, overshadowing their contributions due to mismatched surface charges. This finding implies that there exists room for improvement in DES component selection, which may reduce the misfitting of molecular interaction sites. Overall, the PEG HBD paired with N4Br HBA produces the lowest H_int, even though the H_mf is moderate due to the significant H_HB and H_vdw interactions.

3.4. Effects of Change in Composition of DES

Table 3. Compositions study measured as ln activity coefficients for difluoromethane in varying molar DES HBA and HBD compositions ranging from pure components to 3:X.

HBD	Ratio	Pure	N4Br	N8Br	BTACl	MTOA
-	-	-	−1.61	−1.43	−1.61	−1.16
PEG	1:1	−0.71	−1.39	−1.39	−1.27	−1.03
PEG	1:2	-	−1.27	−1.27	−1.2	−0.97
PEG	1:3	-	−1.19	−1.19	−1.14	−0.93
PEG	2:1	-	−1.48	−1.33	−1.48	−1.08
PEG	3:1	-	−1.52	−1.36	−1.52	−1.10
Camphor	1:1	−0.44	−1.34	−1.34	−1.19	−0.99
Camphor	1:2	-	−1.18	−1.18	−1.08	−0.89
Camphor	1:3	-	−1.07	−1.07	−1.00	−0.83
Camphor	2:1	-	−1.45	−1.28	−1.45	−1.06
Camphor	3:1	-	−1.50	−1.31	−1.50	−1.09
Eucalyptol	1:1	−0.03	−1.27	−1.27	−1.11	−0.92
Eucalyptol	1:2	-	−1.07	−1.07	−0.95	−0.78
Eucalyptol	1:3	-	−0.93	−0.93	−0.85	−0.68
Eucalyptol	2:1	-	−1.41	−1.22	−1.41	−1.02
Eucalyptol	3:1	-	−1.47	−1.27	−1.47	−1.06
Lidocaine	1:1	−0.58	−1.27	−1.27	−1.16	−0.98
Lidocaine	1:2	-	−1.10	−1.10	−1.04	−0.89
Lidocaine	1:3	-	−1.00	−1.00	−0.96	−0.83
Lidocaine	2:1	-	−1.41	−1.25	−1.41	−1.05
Lidocaine	3:1	-	−1.46	−1.29	−1.46	−1.08

shows the results of the compositional analysis. Due to the nature of the ln activity coefficient procedure being performed at a composition of 1:1, the optimal ratio for affinitive interactions will be studied. The ln activity coefficient of difluoromethane was computed for each solvent combination in six different compositions: 1:0, 3:1, 2:1, 1:1, 1:2, and 1:3 being HBA:HBD. Nearly all DES eutectic compositions are found in the range from 1:1 to 1:3 HBA top HBD [29,37]. Thus, the studied compositional ranges reflect this commonality. It is evident that the pure form of HBA has the lowest ln activity coefficients, which indicates a most favorable interaction with difluoromethane, and N4Br (−1.61) having the best of the four. However, the HBA have relatively high melting points compared to their DES form. To use the quaternary ammonium salts alone would require temperatures of approximately 100 °C for them to reach their melting points, which is counter conducive to gas solubility compared to room temperature absorption. Thus, the lowest possible concentration of HBD is recommended to create a room temperature solvent for difluoromethane uptake. Furthermore, the ranking of HBD follows PEG (−1.39), camphor (−1.34), lidocaine (−1.27), and eucalyptol (−1.27) for producing the lowest ln activity coefficients when paired with N4Br. This trend holds for pairings with all four HBA. This trend follows the findings of the sigma analysis where the ranked components follow the favorable charge distributions that induce favorable difluoromethane interaction.

3.5. Effects of Varying System Pressure

Table 4. Pressure study of the top 16 performingsolvents. The solubility of difluoromethane in DES is represented as mol/mol%. The system pressure is in bar.

HBD	Pressure (Bar)	N4Br	BTACl	N8Br	MTOA
Camphor	1	24.2	22.3	24.2	19.2
Camphor	3	49.8	48.3	49.8	43.9
Camphor	6	70.7	70.3	70.7	66.4
Camphor	9	85.8	85.6	84.7	86.2
Eucalyptol	1	22.9	20.9	22.9	18.0
Eucalyptol	3	47.8	46.2	47.8	41.8
Eucalyptol	6	68.9	68.6	68.9	64.2
Eucalyptol	9	86.0	85.6	84.5	86.2
Lidocaine	1	23.8	22.2	23.8	19.4
Lidocaine	3	50.2	48.8	50.2	44.8
Lidocaine	6	71.6	71.2	71.6	67.7
Lidocaine	9	83.7	83.1	81.4	84.0

shows the results of the pressure analysis. Vapor liquid systems are typically operated at pressures higher than atmospheric to maximize gaseous solubility. As pressure increases, non-ideal interactions tend to have less of an overall effect on the studied systems as has been seen in other vapor liquid studies [12]. The significance of this pressure analysis is to determine how well the difluoromethane and DES system will respond to increases in pressure, and if the studied solvent combinations will behave differently. The solubility of difluoromethane was computed for each solvent combination at four different pressures; 1, 3, 6, and 9 bar. The baseline solubility is promising as the lowest is 18% mol solvent per mol difluoromethane at 25 °C and 1 bar for MTOA: Eucalyptol at a 1:1 molar ratio. With increase in pressure, solubility increases with values reaching up to 86% mol/mol solubility at 9 bar. The solubilities are consistent among the solvents with minor deviations due to the similarity of the surface charges. The trends are relatively linear between solubility and pressure. The solvents containing MTOA are shown to increase at a higher rate starting with the lowest solubilities and ending with the highest or equal. The lowest solubility for MTOA is with eucalyptol at 1 bar with 18.03 mol/mol. The highest for MTOA and the highest of the solvents is 86.2 mol/mol which is achieved at 9 bar for the three HBD combinations of PEG, camphor, and eucalyptol. This result is likely described through hole theory, as MTOA is geometrically dissimilar from the others by being asymmetrical. This generally offers lower viscosities and pressure-assisted pore filling [30,38]. The HBA combinations for N4Br, N8Br, and BTACl all have similar solubilities, exhibiting deviations of no more than 10%. A 10% deviation is only witnessed at 1 bar, with decreasing discrepancy with an increase in pressure. The surface charge distribution is highest in the van der Waals and lowest in the hydrogen bond accepting region of the four studied HBA; this is likely the cause of this slight anomaly, as at higher pressures the polar interactions increase in significance, and the higher amount of van der Waals surfaces becomes a less dominating factor.

3.6. EHS Analysis

In general, type 5 DES are considered as comprising “green” components that are environmentally benign. The four HBA and four HBD selected from Section 3.2 are evaluated for this merit through a search of the available safety data sheet (SDS) data from Fischer Scientific [31,39]. The values that were unable to be determined were run through the VEGA KNN-based environmental health and safety (EHS) software, for which a further description can be found in Section 2.2.

Table 5. VEGA model for selected solvent components. Green indicates lower than moderate score of NFPA and/or OSHA rating; blue indicates unreliable predictions due to missing data in VEGA and no available data found in SDS database. Red indicates moderate or higher NFPA and/or OSHA rating for the property. No red markers are present, marking the potential for safe and environmentally benign substances.

	Persistence Air	Persistence Water	Persistence Soil	Mutagenicity	Acute Toxicity	BCF	Carcinogenicity
N4Br	g	g	b	g	g	b	b
BTACl	g	g	g	g	g	b	b
N8Br	b	b	b	g	g	b	b
MTOA	g	g	g	g	r	b	b
PEG	g	g	g	g	g	g	g
Camphor	b	g	g	g	g	g	g
Eucalyptol	b	g	g	g	g	g	g
Lidocaine	b	b	g	g	g	g	g

contains the results of this study where the green (g) indicates a lower than moderate score of NFPA and/or OSHA rating, blue (b) indicates unreliable predictions due to missing data in VEGA and no available data found in the SDS database. Red (r) indicates moderate or higher NFPA and/or OSCHA rating for the property. The solvent components were studied rather than their combination, as DES readily dissociates in the presence of water in the environment. Of the values in the study, only one is red, being from MTOA as it possesses moderately acute toxicity. MTOA also contains two unavailable/unpredictable properties in regard to BCF and carcinogenicity. N4Br is indicated as safe in all areas except for soil persistence, BCF, and carcinogenicity for which there was no available data and thus it was unable to be reliably predicted. N8Br contained the least amount of information with all properties being unavailable, except mutagenicity and acute toxicity which were benign. Camphor and eucalyptol both show the absence of negative properties except for air persistence, for which no data were available and unreliable predictions were made. PEG contained none of the adverse properties. BTACl contained none of the studied properties with the exception of BCF and carcinogenicity, which were undetermined. Due to the lack of significant harmful properties, the components studied have strong potential for use as environmentally safe alternatives to conventional systems. Thus, MTOA is not recommended as an HBA but the other 12 DES are candidates for this application. Difluoromethane is considered extremely flammable with an NFPA rating of 4, but is relatively non-toxic [32,40]. The inclusion of difluoromethane into the working fluid may attribute this characteristic to the solution, and precautions should be undertaken accordingly.

4. Conclusions

Overall, 1298 Type 5 DES were inspected in a 1:1 ratio of HBA and HBD for favorable interactions with the HFC difluoromethane (R32). The results of this project were limited to the top 16 DES and analyzed further for pressure, composition, enthalpic, and EHS properties. The results indicate that the main mode of absorption within these DES with regard to difluoromethane is through van der Waals interactions. The solvents containing significant van der Waals interaction surfaces and some hydrogen bond accepting areas are deemed the most effective traits for the task. It was discovered through the compositional analysis that the HBA in pure form had the highest solubility percentage for difluoromethane. However, because these chemicals have high melting points, it is recommended for the lowest possible ratio of HBD to be added to create a low temperature melting point for the absorption of difluoromethane. Through an increase in pressure, the solubility of difluoromethane increases. MTOA was shown to be the most pressure sensitive HBA of the four studied as it contained the lowest solubility at 1 bar (18 mol/mol%) and the highest at 9 bar (86.2 mol/mol%), likely due to the size of the quaternary salt being the least symmetrical of the four, following hole theory logic. By utilizing SDS and VEGA predictive software, an EHS analysis revealed a strong potential for the 16 DES to be environmentally safe, with MTOA being the only acutely toxic substance. These insights are provided to the scientific community as evidence of the strong efficiency and safety potential of adopting DES into industrial and commercial usage for the absorption of difluoromethane.