A New Compartmentalized Scale (P_N) for Measuring Polarity Applied to Novel Ether-Functionalized Amino Acid Ionic Liquids

Abstract

Functionalized and environmentally friendly ionic liquids are required in many fields, but convenient methods for measuring their polarity are lacking. Two novel ether-functionalized amino acid ionic liquids, 1-(2-methoxyethyl)-3-methylimidazolium alanine ([C₁OC₂mim][Ala]) and 1-(2-ethoxyethyl)-3-methylimidazolium alanine ([C₂OC₂mim][Ala]), were synthesized by a neutralization method and their structures confirmed by NMR spectroscopy. Density, surface tension, and refractive index were determined using the standard addition method. The strength of intermolecular interactions within these ionic liquids was examined in terms of standard entropy, lattice energy, and association enthalpy. A new polarity scale, P_N, is now proposed, which divides polarity into two compartments: the surface and the body of the liquid. Surface tension is predicted via an improved Lorentz-Lorenz equation, and molar surface entropy is used to determine the polarity of the surface. This new P_N scale is based on easily measured physicochemical parameters, is validated against alternative polarity scales, and is applicable to both ionic and molecular liquids.

1. Introduction

The green chemistry concept is a widely accepted focus of modern chemical research, including the design of new materials. Ionic liquids (ILs) have emerged as useful green reaction media due to their many unique features, such as low vapor pressure and high thermal stability [1,2]. They play important roles in many fields, including energy storage [3], catalysis [4], pharmaceuticals, and medicine [5,6]. However, the relatively high viscosity of ILs is a barrier to further practical applications. Inserting ether groups into the cations of ILs has been shown to substantially reduce their viscosity without lowering thermal stability, while also reducing their toxicity [7,8,9,10]. Ether-functionalized ILs (EFILs) have demonstrated remarkable performance in many fields. For example, they dissolve lignocellulosic biomasses [7], reduce viscosity and provide coordination sites for lithium ions in Li/Li-ion batteries [8], and enhance CO₂ selectivity during CO₂ capture [11]. However, traditional ILs containing anions such as Cl⁻, [BF₄]⁻, and [PF₆]⁻ are environmentally hazardous. The development of environmentally friendly, task-specific ILs based on renewable bioresources (such as amino acids and fatty acids) is an environmental necessity. Ohno and Fukumoto were the first to prepare amino acid ILs (AAILs), using 20 different amino acids as the anion; these AAILs demonstrated lower toxicity [12,13]. AAILs have now been utilized in enantioselective separation [14], extraction separation [15], and CO₂ capture [11]. ILs with imidazolium-based cations exhibit low toxicity. Meanwhile, the shorter the alkyl chain on the imidazole ring, the lower the toxicity [16,17]. Thus, imidazolium ILs have proved more attractive than ILs based on ammonium, phosphonium, and pyridinium cations. The present study describes the preparation of two novel ether-functionalized, imidazolium-based AAILs: 1-(2-methoxyethyl)-3-methylimidazolium alanine ([C₁OC₂mim][Ala]) and 1-(2-ethoxyethyl)-face Tension, and Refractive Inde ([C₂OC₂mim][Ala]), together abbreviated as [C_nOC₂mim][Ala](n = 1, 2).

There is scope for considerable further research into EFILs because their physiochemical properties are highly susceptible to changes in structure. More experimental data are required to elucidate their structure–property relationships. This study measures the density, surface tension, and refractive index of [C_nOC₂mim][Ala](n = 1, 2). Density correlates with packing efficiency and intermolecular interactions and is a critical design property in chemical engineering [18]. Its study provides insight into the microstructure and macroscopic properties of ILs [19]. Surface tension is a crucial property at liquid–gas interfaces [20], affecting how the phases interact [21]. The surface tension of ILs is between those of alkanes and water [22]. It can be measured directly or predicted using, for example, the parachor formula [23] or group contribution methods [24]. This study predicts surface tension using an improved Lorentz-Lorenz equation. ILs are often used as solvents, so determining their polarity is crucial. Due to their non-structured nature, polarity cannot be determined by traditional methods such as relative permittivity (ε_r) and dipole moment (δ) [25]. The most widely used experimental method for IL polarity is the E_T(30) scale, which measures the solvatochromic UV−Vis absorbance shift of a solute. However, this method is time-consuming and expensive, so attempts have been made to develop predictive models [26,27].

The present study proposes a new polarity scale, P_N, which enables polarity to be predicted from the easily measured physicochemical properties of density, surface tension, and refractive index. Following on from our previous studies [28,29,30], (i) [C_nOC₂mim][Ala](n = 1, 2) are synthesized and their structures confirmed by nuclear magnetic resonance spectroscopy (NMR); (ii) their density, surface tension, and refractive index are measured from 288.15 to 328.15 K at 5 K intervals; (iii) the strength of their molecular interactions are studied based on standard entropy, lattice energy, and association enthalpy; (iv) an improved Lorentz-Lorenz equation is used to predict the surface tension of ILs and molecular liquids; and (v) a new scale, P_N, for estimating polarity is proposed, combining molar surface entropy s (which measures the polarity of the surface of a liquid) and the polarity coefficient P² (which measures the polarity of the body of a liquid).

2. Results and Discussion

2.1. Density, Surface Tension, and Refractive Index of [C_nOC₂mim][Ala](n = 1, 2)

The density (ρ), surface tension (γ), and refractive index (n_D) of [C_nOC₂mim][Ala](n = 1, 2) with various water contents over 288.15–328.15 K (at 5 K intervals) are shown in Tables S1–S3 Supplementary Materials, with each value being an average of three measurements using the standard addition method. These parameters were plotted against water content (Figure 1), producing a series of straight lines with correlation coefficient squares (r²) consistently greater than 0.99. The y-axis intercepts of these lines give the experimental value of each parameter in anhydrous [C_nOC₂mim][Ala](n = 1, 2) (

Table 1. Density (ρ), surface tension (γ), refractive index (n_D), and thermal expansion coefficient (α) at various temperatures (T) for [C_nOC₂mim][Ala](n = 1, 2).

T (K)	ρ (g·cm−3)	γ (mJ·m−2)	nD	α (K−1 × 104)
[C1OC2mim][Ala]
288.15	1.16073	51.6	1.5112	5.8555
293.15	1.15744	51.2	1.5097	5.8722
298.15	1.15423	50.9	1.5080	5.8885
303.15	1.15082	50.5	1.5066	5.9060
308.15	1.14739	50.0	1.5048	5.9236
313.15	1.14395	49.7	1.5038	5.9414
318.15	1.14049	49.3	1.5019	5.9595
323.15	1.13703	48.8	1.5004	5.9776
328.15	1.13365	48.4	1.4991	5.9954
[C2OC2mim][Ala]
288.15	1.13849	49.6	1.4942	5.8490
293.15	1.13475	49.3	1.4927	5.8683
298.15	1.13190	48.9	1.4914	5.8830
303.15	1.12854	48.5	1.4899	5.9005
308.15	1.12514	48.0	1.4885	5.9184
313.15	1.12175	47.6	1.4869	5.9363
318.15	1.11838	47.2	1.4854	5.9541
323.15	1.11521	46.7	1.4839	5.9711
328.15	1.11166	46.3	1.4824	5.9901

Standard uncertainties (u) are u(T) = 0.02 K and u(p) = 10 kPa; expanded uncertainties (U) are U(ρ) = 0.002 g·cm⁻³, U(γ) = 0.3 mJ·m⁻², and U(n_D) = 0.003, with 95% confidence (k = 2).

).

2.2. Strength of [C_nOC₂mim][Ala](n = 1, 2) Intermolecular Interactions

The density of ILs increases gradually as temperature rises. At higher temperatures, the mobility of constituent ions improves and the unit volume increases [31]. The thermal expansion coefficient (α) is defined as

$$\alpha =\left(1/V\right){\left(\partial V/\partial Y\right)}_p=-{\left(\partial ln\rho /\partial T\right)}_p$$

(1)

where V is molar volume. Molar volume is defined as

$$V=\mathrm{M}/\rho$$

(2)

The molecular volume V_m is defined as

$$V\mathrm{m}=\mathrm{M}/\mathrm{N}\rho =V/\mathrm{N}$$

(3)

where N is the Avogadro constant, V is molar volume, and M is molar mass.

At 298.15 K, V_m is 0.3300 nm³ for [C₁OC₂mim][Ala] and 0.3571 nm³ for [C₂OC₂mim][Ala] (Table S4). The difference between these indicates that the contribution of methylene (-CH₂-) to V_m is 0.0271 nm³. This is close to the average contribution methylene makes to the V_m of several other ILs listed in Table S5 (0.0278 nm³).

Lattice energy (U_POT) and standard entropy (S^θ₂₉₈) can be calculated according to Glasser’s theory; Equation (4) is suitable for MX(1:1) type ionic salts. The constants in Equation (5) are empirical values [32].

$$U_{\mathrm{POT}}=1981.2{(\rho /\mathrm{M})}^{1/3}+103.8$$

(4)

$${S^{\mathsf{\theta }}}_{298}=1246.5V\mathrm{m}+29.5$$

(5)

U_POT reflects the strength of intermolecular interactions and can be used to measure the stability of ILs [32,33]. U_POT is 443 kJ·mol⁻¹ for [C₁OC₂mim][Ala] and 435 kJ·mol⁻¹ for [C₂OC₂mim][Ala], implying that methylene’s average contribution to U_POT is −8 kJ·mol⁻¹. U_POT is inversely related to molar volume. Addition of a methylene group will reduce the ionic or molecular packing efficiency, decreasing the strength of the interactions.

To some extent, standard entropy reflects the degree of disorder of molecular arrangements [33]. S^θ₂₉₈ is 441 J·K⁻¹·mol⁻¹ for [C₁OC₂mim][Ala] and 475 J·K⁻¹·mol⁻¹ for [C₂OC₂mim][Ala], suggesting that methylene’s average contribution to S^θ₂₉₈ is 34 J·K⁻¹·mol⁻¹ (Table S5). This indicates that ILs with longer aliphatic chains are more disordered. Standard entropy increases as molecular volume increases. S^θ₂₉₈ of ILs is usually greater than 200 J·K⁻¹·mol⁻¹, compared with the more molecularly ordered conventional inorganic salts such as NaCl (72.1 J·K⁻¹·mol⁻¹) and KCl (82.6 J·K⁻¹·mol⁻¹) [34]. This may explain why ILs are molten below 373 K.

The association enthalpy (Δ_AH_m⁰) also reflects the strength of intermolecular interactions: the higher the Δ_AH_m⁰, the stronger the gaseous state interactions. Δ_AH_m⁰ of ILs in the gaseous phase can be calculated based on the thermodynamic cycle shown in Scheme 1 [29].

Vaporization enthalpy (Δ_l^gH_m⁰) is a key parameter for calculating Δ_AH_m⁰. It is estimated from Equation (6):

$$g_s=\mathrm{a}+\mathrm{b}({{\mathsf{\Delta }}_{\mathrm{l}}}^{\mathrm{g}}{H_{\mathrm{m}}}^0-\mathrm{R}T)$$

(6)

$$g_s=\gamma V^{2/3}{\mathrm{N}}^{1/3}$$

(7)

where g_s is the molar surface Gibbs energy; γ is surface tension; V is molar volume; N is the Avogadro constant; Δ_l^gH_m⁰ is vaporization enthalpy; R is the gas constant; T is temperature; and a and b are the empirical constants −1.519 kJ·mol⁻¹ and 0.09991, respectively [29].

The estimated vaporization enthalpy is 164.1 kJ·mol⁻¹ for [C₁OC₂mim][Ala] and 165.1 kJ·mol⁻¹ for [C₂OC₂mim][Ala]. Consequently, Δ_AH_m⁰ is −278.9 kJ·mol⁻¹ for [C₁OC₂mim][Ala] and −269.9 kJ·mol⁻¹ for [C₂OC₂mim][Ala]. The Δ_AH_m⁰ of other ILs are listed in Table S5. Again, addition of methylene reduces packing efficiency and increases the degree of molecular disorder. Thus, for ILs with the same anions, the absolute value of Δ_AH_m⁰ decreases as the length of the imidazole ring alkyl side chains increases. For ILs with the same cations, Δ_AH_m⁰ decreases with increasing anion volume.

2.3. Prediction of Surface Tension Based on Molar Surface Gibbs Energy

The parameter g_s used to estimate vaporization enthalpy was developed in our previous work by modifying Li’s model [35]. The definition of g_s is consistent with the concept presented by Myers [36]. Thus, g_s is a true thermodynamic function that integrates volumetric and surface properties.

Plotting g_s against T for [C_nOC₂mim][Ala](n = 1, 2) yields straight lines (Figure 2) such that their relationship can be expressed as

$$g_s=G_0-G_{1}T$$

(8)

G₀ and G₁ for [C_nOC₂mim][Ala](n = 1, 2) are obtained from Figure 2 and substituted into Equation (8) to give the estimated molar surface Gibbs energy, g_s(est). Values of g_s, G₀, G₁, and g_s(est) are listed in

Table 2. Molar surface Gibbs energy (g_s), G₀, G₁, and estimated molar surface Gibbs energy (g_s(est)) for [C_nOC₂mim][Ala](n = 1, 2).

T (K)	gs (kJ·mol−1)	G0	G1	gs(est) (kJ·mol−1)
[C1OC2mim][Ala]
288.15	14.78	19,803	17.4	14.79
293.15	14.69	19,803	17.4	14.70
298.15	14.63	19,803	17.4	14.62
303.15	14.55	19,803	17.4	14.53
308.15	14.43	19,803	17.4	14.44
313.15	14.37	19,803	17.4	14.35
318.15	14.29	19,803	17.4	14.27
323.15	14.17	19,803	17.4	14.18
328.15	14.08	19,803	17.4	14.09
[C2OC2mim][Ala]
288.15	14.88	20,658	19.9	14.92
293.15	14.82	20,658	19.9	14.82
298.15	14.73	20,658	19.9	14.72
303.15	14.63	20,658	19.9	14.63
308.15	14.51	20,658	19.9	14.53
313.15	14.42	20,658	19.9	14.43
318.15	14.33	20,658	19.9	14.33
323.15	14.20	20,658	19.9	14.23
328.15	14.11	20,658	19.9	14.13

.

The Lorentz-Lorenz equation expresses the relationship between n_D and the mean molecular polarizability (α_p) [37]:

$$R_{\mathrm{m}}=[({n_{\mathrm{D}}}^2-1)/({n_{\mathrm{D}}}^2+2)]·V=(4\pi \mathrm{N}/3)·{\alpha }_{\mathrm{p}}$$

(9)

where R_m is molar refraction, α_p is mean molecular polarizability, and n_D is refractive index. This has been combined with g_s to give an improved Lorentz-Lorenz equation [38] that can predict surface tension, γ_(est):

$${{\gamma }_{\left(\mathrm{est}\right)}}^{3/2}=[{g_{\mathrm{s}\left(\mathrm{est}\right)}}^{3/2}/{\mathrm{N}}^{1/2}{R}_{\mathrm{m}}]({n_{\mathrm{D}}}^2-1)/({n_{\mathrm{D}}}^2+2)$$

(10)

The R_m, α_p, and γ_(est) for [C_nOC₂mim][Ala](n = 1, 2) are listed in

Table 3. Molar refraction (R_m), mean molecular polarizability (α_p), and estimated surface tension (γ_(est)) of [C_nOC₂mim][Ala](n = 1, 2).

T (K)	Rm	αp × 1024	γ(est)
[C1OC2mim][Ala]
288.15	57.74	22.91	51.6
293.15	57.77	22.92	51.2
298.15	57.76	22.92	50.8
303.15	57.80	22.93	50.4
308.15	57.82	22.94	50.0
313.15	57.84	22.95	49.6
318.15	57.89	22.97	49.2
323.15	57.90	22.97	48.8
328.15	57.89	22.97	48.4
[C2OC2mim][Ala]
288.15	62.24	24.69	49.4
293.15	62.28	24.71	49.0
298.15	62.30	24.72	48.6
303.15	62.32	24.73	48.2
308.15	62.36	24.74	47.7
313.15	62.37	24.75	47.3
318.15	62.39	24.76	46.9
323.15	62.41	24.76	46.5
328.15	62.43	24.77	46.1

. Plotting estimated surface tension values against their corresponding experimental values produces a straight line (Figure 3). A similar plot for other ionic and molecular liquids also illustrates a linear relationship (Table S6, Figure 4), showing that this method is applicable for the surface tension prediction of both types of liquid.

For ILs sharing the same anions (Table 1 and Table S6), n_D decreases as the length of the alkyl chain in the cations increases. Refractive index correlates with dipole moment [39,40], which increases with higher molecular packing density [41]. Thus, a larger dipole moment results in a higher refractive index. The n_D is larger in [C₁OC₂mim][Ala] than [C₂OC₂mim][Ala], so the packing density of the former is higher, which is consistent with the density trend of the two ILs.

As polarizability increases, Coulomb interactions are reduced and ion mobility rises [42]. The α_p of [C₁OC₂mim][Ala] is lower than [C₂OC₂mim][Ala], so its Coulomb interactions are stronger. This finding is similar to the pattern observed for U_POT and Δ_AH_m⁰.

2.4. Molar Surface Entropy: Polarity Contribution from Surface Liquid

For most liquids, surface tension declines as temperature increases, as shown by the Eötvös equation:

$$\gamma V^{2/3}=k(Tc-T)$$

(11)

where T_c is critical temperature and k is the Eötvös equation parameter, which is associated with polarity. For some organic liquids with weak polarity, k is nearly 2.2 × 10⁻⁷ J·mol^−2/3·K⁻¹, while for some with strong polarity, such as molten NaCl, it is nearly 0.4 × 10⁻⁷ J·mol^−2/3·K⁻¹ [35,43]. However, the physical significance of k is not clear. Multiplying both sides of the Eötvös equation by N^1/3 [35] gives

$$g_s=C_0-C_{1}T$$

(12)

which fits the fundamental thermodynamic concept:

$$G=H-TS$$

(13)

C₁ denotes the molar surface entropy and is given by C₁= −($\frac{\partial g_s}{\partial T}$)_p.

The relationship between molar surface entropy (defined here as s [29]) and k is expressed as

$$s={\mathrm{N}}^{1/3}k$$

(14)

Entropy is directly linked to the number of microstates [44]. Higher entropy means molecules can be arranged in more ways, while the total energy remains constant. Thus, the physical significance of s is clear—it reflects the polarity of a liquid’s surface (higher s, lower surface polarity). The value of s is 17.39 J·mol⁻¹·K⁻¹ for [C₁OC₂mim][Ala] and 19.94 J·mol⁻¹·K⁻¹ for [C₂OC₂mim][Ala]. Values for other EFILs are listed in

Table 4. Molar surface entropy (s) of various EFILs.

IL	s (J·mol−1·K−1)	V × 104 (m3·mol−1)
[C1OC2mim]Cl [38]	15.99	1.52
[C1OC2mim][Ala]	17.39	1.99
[C1OC2mim][Thr] [45]	26.60	2.18
[C1OC2mim][NTf2] [19]	17.80	2.80
[C2OC2mim]Cl [38]	17.81	1.68
[C2OC2mim][Ala]	19.94	2.15
[C2OC2mim][Thr] [45]	28.12	2.37
[C2OC2mim][NTf2] [19]	19.32	2.99
[C1OC4mim][Gly] [46]	19.56	2.19
[C1OC4mim][Ala] [46]	20.54	2.29
[C1OC4mim][Thr] [46]	22.082	2.51
[C2OC1mim][NTf2] [19]	18.20	2.81
[C1OC3mim][NTf2] [19]	18.49	2.97
[C3OC2mim][NTf2] [19]	20.57	3.16

. The overall trend is that for ILs with the same cations, such as [C₁OC₂mim]⁺, [C₂OC₂mim]⁺, and [C₁OC₄mim]⁺, s increases as the volume of anions increases. Cl⁻, [Ala]⁻, [Thr]⁻, and [Gly]⁻ clearly obey this rule. It can be speculated that larger anions cause greater disordering of surface molecules, leading to lower polarity. However, [NTf₂]⁻ does not conform to this rule. This may be due to it being more symmetrical than other anions, facilitating a more orderly arrangement of surface molecules and increasing the polarity. For ILs sharing the same anions, the general trend is that s increases as the volume of cations increases. Values of s for different other ILs (Table S7) confirm this effect.

Data for [C₁OC₂mim][NTf₂], [C₂OC₁mim][NTf₂], [C₂OC₂mim][NTf₂], and [C₁OC₃mim][NTf₂] (Table 4) imply that, for ILs with the same number of alkyl side chain carbons, the position of the ether group also affects s, presumably because it affects packing efficiency on the liquid surface.

2.5. A New Model for Predicting Polarity

Experimental methods for measuring polarity, such as E_T(30), are time consuming and laborious. Here, we present a predictive model that establishes a relationship between polarity and the easily determined physicochemical properties of density, surface tension, and refractive index.

Our previous work [28], based on Hildebrand and Scott’s theory [47], proposed δ_μ as a polarity scale. δ_μ is the solubility parameter derived from the contribution of the average permanent dipole moment:

$${{\delta }_{\mathsf{\mu }}}^2={{\Delta }^{\mathrm{g}}}_{\mathrm{l}}{H^0}_{\mathrm{m}\mathsf{\mu }}/V-(1-x)\mathrm{RT}/V$$

(15)

where V is the molar volume and Δ^g_lH⁰_mμ is the contribution of the average permanent dipole moment to Δ^g_lH⁰_m, such that

$${{\Delta }^{\mathrm{g}}}_{\mathrm{l}}{H^0}_{\mathrm{m}\mathsf{\mu }}={{\Delta }^{\mathrm{g}}}_{\mathrm{l}}{H^0}_{\mathrm{m}}-{{\Delta }^{\mathrm{g}}}_{\mathrm{l}}{H^0}_{\mathrm{mn}}$$

(16)

where Δ^g_lH⁰_mn is the contribution of the induced dipole moment to Δ^g_lH⁰_m and can be calculated from the Lawson–Ingham equation [48]:

$${{\Delta }^{\mathrm{g}}}_{\mathrm{l}}{H^0}_{\mathrm{mn}}=\mathrm{C}[({n_{\mathrm{D}}}^2-1)/({n_{\mathrm{D}}}^2+2)]V=\mathrm{C}R\mathrm{m}$$

(17)

where C is the empirical constant 1.297 kJ·cm⁻³. In Equation (15), x represents Δ^g_lH⁰_mn/Δ^g_lH⁰_m (at 298.15 K). The δ_μ polarity of [C₁OC₂mim][Ala] (21.03 J^1/2·cm^−3/2) is larger than [C₂OC₂mim][Ala] (19.65 J^1/2·cm^−3/2).

However, there is an obvious drawback to δ_μ: it has a dimension (J^1/2·cm^−3/2), while some polarity scales, such as the dielectric constant, are non-dimensional. Furthermore, the contribution of the induced dipole moment was neglected. Therefore, δ_μ was improved as follows and designated as P [45]:

$$P={\delta }_{\mathsf{\mu }}/{\delta }_{\mathrm{n}}={\left[\right({{\Delta }^{\mathrm{g}}}_{\mathrm{l}}{H^0}_{\mathrm{m}\mathsf{\mu }}/V-(1-x)\mathrm{R}T/V)/({{\Delta }^{\mathrm{g}}}_{\mathrm{l}}{H^0}_{\mathrm{mn}}/V-x\mathrm{R}T/V\left)\right]}^{1/2}$$

(18)

where δ_n is the solubility parameter from the contribution of the induced dipole moment. Comparison of Δ^g_lH⁰_mμ and Δ^g_lH⁰_mn shows that (1−x)RT/V and xRT/V can be omitted and Equation (18) can be expressed as

$$P={({{\Delta }^{\mathrm{g}}}_{\mathrm{l}}{H^0}_{\mathrm{m}\mathsf{\mu }}/{{\Delta }^{\mathrm{g}}}_{\mathrm{l}}{H^0}_{\mathrm{mn}})}^{1/2}$$

(19)

The effects of average permanent dipole moment and induced dipole moment are both considered within P, which is dimensionless, with a large value indicating high polarity. [C₄mim][BF₄] is hydrophilic and [C₄mim][NTf₂] is hydrophobic. According to Seddon et al. [49], P is 1.226 for [C₄mim][BF₄] and 0.401 for [C₄mim][NTf₂]. This higher polarity of [C₄mim][BF₄] fits practical experience. Thus, the parameter P is capable of measuring the polarity of ILs. P is 1.191 for [C₁OC₂mim][Ala] and 1.043 for [C₂OC₂mim][Ala].

This study divided polarity into two compartments: the contribution from the body of a liquid, and the contribution from the surface. Cohesive energy density can demonstrate the strength of intermolecular interactions within the body [48]. δ_μ² is the cohesive energy density from the average permanent dipole moment and δ_n² is from the induced dipole moment. Consequently, P² can describe the polarity of the body of a liquid:

$$P^2={{\delta }_{\mathsf{\mu }}}^2/{{\delta }_{\mathrm{n}}}^2$$

(20)

Molar surface entropy (s) was proven above to reflect the polarity of a liquid surface. Combining s and P², a new polarity scale, P_N, is now proposed:

$$P_{\mathrm{N}}=s/P^2=s/({{\delta }_{\mathsf{\mu }}}^2/{{\delta }_{\mathrm{n}}}^2)$$

(21)

This compartmentalized scale is a novel method to evaluate polarity, with a large P_N indicating weak polarity. Based on literature data [50,51,52,53,54,55], P_N is 22.03 J·mol⁻¹·K⁻¹ for [C₄mim][NTf₂] and 14.50 J·mol⁻¹·K⁻¹ for [C₄mim][BF₄]. These results fit with practical experience and demonstrate the rationality of the P_N scale. The P_N of [C₁OC₂mim][Ala] is 12.26 J·mol⁻¹·K⁻¹ and is 18.33 J·mol⁻¹·K⁻¹ for [C₂OC₂mim][Ala], which is the same polarity trend observed using δ_μ and P. As discussed above, ILs with longer alkyl chains exhibit higher standard entropy and lower molecular packing efficiency. The observed higher polarity of [C₁OC₂mim][Ala] may be due to its stronger intermolecular interactions and more ordered molecular arrangement. This fact can be explained as follows: a polar molecule has a permanent electric dipole moment, and a molecule may be polar if it has low symmetry [44]. ILs have asymmetric structures, and an orderly arrangement of ILs will maintain this structure. In this situation, their permanent electric dipole moments will not counteract each other, and polarity will be enhanced. The P_N of various ether-functionalized ILs are listed in

Table 5. Polarity of various ether-functionalized ionic liquids using the new P_N scale.

IL	PN (J·mol−1·K−1)	Reference
[COC2mim]Cl	12.86	[37]
[C2OC2mim]Cl	16.76	[37]
[C1OC2mim][Ala]	12.26	This work
[C2OC2mim][Ala]	18.33	This work
[COC4mim][Ala]	21.83	[39]
[COC2mim][Thr]	22.18	[38]
[C2OC2mim][Thr]	25.46	[38]
[COC4mim][Thr]	24.52	[39]
[COC2mim][NTf2]	31.31	[13]
[C2OCmim][NTf2]	35.40	[13]
[C1OC3mim][NTf2]	35.57	[13]
[C2OC2mim][NTf2]	42.66	[13]
[C3OC2mim][NTf2]	55.1	[13]
[COC4mim][Gly]	19.64	[39]

. For ILs with the same anion, P_N declines as the length of the imidazole ring alkyl side chain increases. This trend supports the contention that the strength of intermolecular interactions and the degree of disorder of molecular arrangements influence polarity.

The P_N scale can be validated by comparison with other polarity scales (

Table 6. Polarity of various ionic liquids estimated using the P_N scale.

IL	PN (J·mol−1·K−1)
[C6mim]OAC	35.73
[C4mim]OAC	23.24
[C4mim]NTf2 [51,53]	22.03
[C4mmim]NTf2	19.96
[C4mim]BF4	14.50
[C2mmim]NTf2	14.01
[C5mim]Lact	12.87
[C2mim]BF4	8.24
[C2mim]Lact	8.16

). FTIR spectroscopy probes [56] and E_T^N [57] show that the polarity of [C₂mim]BF₄ is larger than [C₄mim]BF₄. The P_N scale gives the same qualitative result. Wu et al. determined the order of polarity of several ILs to be [C₄mim]BF₄ > [C₄mim]NTf₂ > [C₄mim]OAc using E_T^N [58]. Again, P_N gives the same result.

Furthermore, when P_N is applied to molecular liquids, the estimated polarity is broadly, and inversely, consistent with the dielectric constant ε [59] (

Table 7. P_N and dielectric constant (ε) of various molecular liquids.

Molecular Liquid	PN	ε
Ethyl acetate	169.14	6.1
Chloroform	167.19	4.8
Tetrahydrofuran	124.29	7.5
Pyridine	60.64	12.3
Benzyl alcohol	48.13	13.0
1-Hexanol	27.06	13.0
Cyclohexanol	25.09	15.0
1-Propanol	8.48	20.3
Ethanol	5.52	25.3
Methanol	2.40	33.0

). The correlation between P_N and the inverse ε⁻¹ (Figure 5) is r² = 0.94, demonstrating that P_N is also suitable for molecular liquids.

E_T(30) is one of the most popular polarity scales for evaluating ILs, but it is laborious and costly [60]. Moreover, the results vary depending on the molecular probe used [61]. However, using P_N, polarity can be determined simply from density, surface tension, and refractive index. The dielectric constant is the traditional polarity scale for organic solvents. The comparison of dielectric constant and P_N proves the universal applicability of P_N. Thus, this new P_N scale is demonstrated to be a viable predictive method for evaluating the polarity of both ionic and molecular liquids based on easily measured physicochemical properties. It will find applications in many fields, particular those employing novel ionic liquids for which the evaluation of polarity is expensive and time consuming.

3. Materials and Methods

3.1. Materials

The sources and purity of reagents are listed in

Table 8. Source and purity of reagents.

Reagent Name	CAS No.	Source	Purification	Mass Fraction Purity	Analysis
Anion exchange resin 717	122560-63-8	SRC	None	Granularity > 0.950	GC
N-Methylimidazole	616-47-7	ACROS	Distillation	>0.990	FM
2-Chloroethyl methyl ether	627-42-9	SRC	None	>0.995	FM
2-Chloroethyl ethyl ether	628-34-2	SRC	None	>0.995	FM
DL-Alanine	302-72-7	SRC	Recrystallization	>0.990	FM
Acetonitrile	75-05-8	SRC	None	>0.995	FM
Ethyl acetate	141-78-6	SRC	None	>0.995	FM
Anhydrous ethanol	64-17-5	SRC	None	>0.995	FM
Sodium hydroxide	1310-73-2	SRC	None	>0.960	FM
[C1OC2mim][Ala]	-	Synthesis	Solvent extraction, vacuum drying	>0.990	1H, 13C NMR
[C2OC2mim][Ala]	-	Synthesis	Solvent extraction, vacuum drying	>0.990	1H, 13C NMR

FM—Fractional melting; SRC—Shanghai Reagent Co., Ltd.

. N-Methylimidazole (AR grade) was purified by distillation, while 2-chloroethyl methyl ether and 2-chloroethyl ethyl ether (both AR grade) were used as purchased. DL-Alanine was recrystallized from water and dried in a vacuum oven [62].

3.2. Preparation of ILs [C_nOC₂mim][Ala](n = 1, 2)

[C_nOC₂mim][Ala](n = 1, 2) were prepared by Fukumoto’s neutralization method [12], and [C_nOC₂mim]Cl(n = 1, 2) by Sheldon’s method [63]. An equal molar amount of 2-chloroethyl methyl ether or 2-chloroethyl ethyl ether was added dropwise to N-methylimidazole under nitrogen in a three-necked round-bottom flask while stirring at 298.15 K. The reaction temperature reached 353.15 K with 2-chloroethyl methyl ether and 373.15 K with 2-chloroethyl ethyl ether. The reactions lasted for 48 h and produced light yellow liquids, which were then washed three times with ethyl acetate, producing [C_nOC₂mim]Cl(n = 1, 2). These were transformed into [C_nOC₂mim]OH(n = 1, 2) using basic anion exchange resin conditioned with sodium hydroxide [29]. The aqueous [C_nOC₂mim]OH(n = 1, 2) were then added dropwise (at slight excess) to aqueous DL-alanine and reacted for 72 h, yielding [C_nOC₂mim][Ala](n = 1, 2). Water was removed by rotary evaporation and excess DL-alanine by ethanol:acetonitrile (9:1). The solvents were evaporated under reduced pressure and [C_nOC₂mim][Ala](n = 1, 2) were dried in a vacuum oven for 60 h at 353.15 K. The chemical structures of [C_nOC₂mim][Ala](n = 1, 2) are shown in Figure 6.

3.3. Analytical Methods

The structures of [C_nOC₂mim][Ala](n = 1, 2) were characterized by NMR (Varian XL-300), as shown in the Supplementary Materials. The final water contents (w₂) of [C_nOC₂mim][Ala](n = 1, 2), as measured using a ZSD-2 Karl Fischer moisture titrator, were 0.00472 and 0.00640 ± 0.0001 (mass fraction), respectively.

Since [C_nOC₂mim][Ala](n = 1, 2) form strong hydrogen bonds with water, it is difficult to remove all traces of water from these ILs, which affects their density, surface tension, and refractive index. Therefore, the standard addition method was used to determine these properties. Each parameter was measured in [C_nOC₂mim][Ala](n = 1, 2) at different water contents following heating at graduated temperatures. Values were then plotted against water content, and the intercept of the regression lines yielded the parameter values in the anhydrous ILs at a given temperature.

4. Conclusions

[C_nOC₂mim][Ala](n = 1, 2) were prepared and their structures confirmed by NMR. Density, surface tension, and refractive index were determined by the standard addition method. Adding methylene to the aliphatic chain of an IL increased its standard entropy. Lattice energy and association enthalpy measurements showed that molecules of [C₁OC₂mim][Ala] were more compacted, and their intermolecular interactions stronger, than [C₂OC₂mim][Ala]. An improved Lorentz-Lorenz equation predicted the surface tension of both ionic and molecular liquids. A new compartmentalized polarity scale (P_N) based on molar surface Gibbs energy and dipole moments is presented. It encompasses the polarity of both the surface and body of a liquid. [C₁OC₂mim][Ala] is shown to have higher polarity than [C₂OC₂mim][Ala] based on P_N. The validity of P_N is demonstrated by comparison with alternative polarity scales and published polarities of both ionic and molecular liquids.