Physicochemical and electrochemical properties of a new series of protic ionic liquids with N-chloroalkyl functionalized cations

Abstract

Six new protic ionic liquids (PILs) based on N-chloroalkyl functionalized morpholinium, piperidinium, pyrrolidinium and alkylammonium cations, with bis[(trifluoromethyl)sulfonyl]imide as counter-ion, were synthesized by a metathesis reaction. To understand the differences of structure, charge distribution and volume between the various investigated N-chloroalkyl functionalized cations, as well as between their non-chloro analogues, computational methods were used to generate the COSMO volume and the sigma profile of each ion. Physicochemical investigations showed lowered melting point of these PILs (−8.5 °C < T_m < 34.1 °C) as compared to their non-functionalized analogues and a high thermal stability with T_5%onset in the range 280–337 °C. The alkylammonium and pyrrolidinium-based PILs display reasonable conductivity (1.23 mS cm⁻¹ < σ < 1.71 mS cm⁻¹), although their viscosity values are relatively high (0.0665 Pa s < η < 0.1093 Pa s). The effect of temperature on the transport properties of each PIL has then investigated by fitting the experimental data with the Arrhenius law and the Vogel–Tamman–Fulcher (VTF) equations, revealing the convergence of viscosity with the former model and conductivity with the latter one. The electrochemical stability of cations towards reduction is discussed in light of the frontier orbital theory. All N-chloroalkyl functionalized PILs display a wide electrochemical stability window (4.1–5.1 V), in the same range as the representative of non-chloro functionalized analogues, [HN₂₂₂][TFSI], (4.3 V).

---

1. Introduction

Ionic Liquids (ILs), claimed to be an outstanding class of materials, are defined as compounds consisting entirely of ions (an organic cation and an organic or inorganic anion) with melting point arbitrarily limited to 100 °C. They possess a very wide liquid range and, owing to their low volatility and non-flammability, they display the characteristic utility of molecular solvents, serving as a reaction medium with enhanced safety.^1–6 ILs are classified into two broad categories: aprotic ionic liquids (AILs) and protic ionic liquids (PILs), distinguished by at least one hydrogen atom attached to the central atom of the cation,⁷ which might be ammonium,⁸ phosphonium⁹ or sulphonium.¹⁰

Although PILs are less known than AILs, they are the object of a growing interest in recent years, not only due to the simplicity of their manufacturing process and the ubiquity of the starting materials, such as amines and acids, but also because they often have lower melting point and glass transition temperature than their aprotic counterparts.¹¹ Usually, PILs are formed by a proton transfer from a Brönsted acid to a Brönsted base,¹² and the most widely reported PILs are synthesized by amine neutralization with common organic acids like formic and acetic acid,^13–15 or inorganic ones like nitric and sulphuric acid.^16–18 PILs may be also obtained by a metathesis reaction through replacing a halide anion combined with protonated cation by another one,¹⁹ such as tetrafluoroborate [BF₄]⁻, hexafluorophosphate [PF₆]⁻, bis(perfluoroethylsulfonyl)imide [BETI]⁻ or bis[(trifluoromethyl)sulfonyl]imide [TFSI]⁻.^11,18

Since PILs can be tailored like all ILs, the selection of appropriate cation and anion is intended to provide compatibility with specific requirements. Considering electrochemical applications, a wide electrochemical stability window combined with high conductivity and low viscosity is demanded. PILs with anions such as [NO₃]⁻, [HCOO]⁻, [CH₃COO]⁻, [CF₃COO]⁻, together with other anions derived from oxoacids or carboxylic acids, are the most popular owing to their low cost, easy synthesis procedure and relatively good transport properties, yet their electrochemical window is rather narrow, e.g., [HPyrr][HCOO] with EW of 2.55 V on glassy carbon, 1.82 V on gold and 1.45 V on platinum.²⁰ The [PF₆]⁻, [BF₄]⁻ and [TFSI]⁻ anions provide enhanced electrochemical stability, however, the coupling of [PF₆]⁻ or [BF₄]⁻ with common ammonium cations results mostly in solid PILs at RT or even protic molten salts (with melting temperature higher than 100 °C, according to ILs definition¹²). Beside high electrochemical stability, [TFSI]⁻ enlarges the liquid range of ILs, both for upper and lower temperatures limits,^21–23 whereas it is not susceptible to hydrolysis in presence of water traces, and is preferred to cheaper alternatives such as [BF₄]⁻ and [PF₆]⁻ anions slowly undergoing this process with concomitant release of HF.^24,25 Countless publications report on AILs with [TFSI]⁻ anion, whereas the most significant part of research on PILs with [TFSI]⁻ has been presented by Watanabe et al. for application in fuel cells, owing to the presence of a dry labile proton.^18,21,26,27 However, among a long list of common protonated amines combined with [TFSI]⁻, such as pyrrolidine, pyridine, imidazole, pyrazole etc., only the PIL with triethylammonium cation [HN₂₂₂]⁺ exhibits a melting point below RT (here 3.5 °C).²¹ PILs were also explored as alternative protic medium in electrochemical capacitors (ECs) with RuO₂ electrodes, where the transferable proton takes part in quick faradic redox processes (similarly to what is observed for aqueous solutions), enhancing capacitance through a pseudocapacitive contribution.^28–30 PILs, namely pyrrolidinium nitrate [Pyrr][NO₃], were also applied in ECs with activated carbon (AC) electrodes, where the proton is involved in faradic reactions with the surface functionality of carbon.³¹ By contrast, when utilizing [HN₂₂₂][TFSI] in presence of AC electrodes, only electrical double-layer (EDL) charging was observed at voltages up to 2 V,¹⁹ whereas the [HN₂₂₂]⁺ cation is reduced liberating nascent hydrogen which is adsorbed onto AC when the voltage is raised to 2.5 V.³² Further works confirmed the advantage of applying the [TFSI]⁻ anion for enlarging the electrochemical stability on AC, while pointing that an adequate selection of the counter cation importantly influences the capacitance of the system.³³

Notwithstanding, the number of room temperature PILs with small cations and [TFSI]⁻ anion is limited. In the case of AILs involving the [TFSI]⁻ anion, cations functionalized with hydroxyl³⁴ or methoxy^34–36 groups in their alkyl substituents (to increase asymmetry) were successfully utilized for lowering the melting point as well as for improving the transport properties. However, in case of the methoxy substituent, the electrochemical stability is unfortunately reduced.³⁷

In this work, searching for PILs liquid at least at room temperature and displaying possibly attractive characteristics for electrochemical applications, the [TFSI]⁻ anion was coupled with a series of N-chloroalkyl functionalized cations. The already mentioned beneficial characteristics of [TFSI]⁻, including its bulkiness, high flexibility and extensive charge delocalization which contribute to formation of low-temperature ILs as well as great electrochemical stability,³⁸ was combined with the positive effect of functionalized cations displaying increased asymmetry, thus also facilitating the lowering of melting temperature of the resulting PILs. Chlorine itself generates an electronegative withdrawing effect slightly affecting the charge distribution within the molecule. The examples of AILs with chloralkyl substituted cations so far reported in the literature are sporadic and focus on the (2-chloroethyl)trimethylammonium cation coupled with various biologically active anions; the corresponding PILs demonstrate phytopharmaceutical/herbicidal properties.^39–41 Apart from these features, the application of a heteroatom like chlorine is expected to change the molecules polarizability and consequently PILs organization in the electrical double-layer (EDL) on a charged surface.⁴²

Hence, we present a series of six new protic ionic liquids with N-chloroalkyl functionalized cations and [TFSI]⁻ as counteranion. The cations were selected in aim to achieve electrochemically stable PILs, liquid at least at room temperature and having reasonable transport properties. We describe their preparation, physicochemical and thermal properties as well as electrochemical stability window. These properties depend on the nature of the cation, i.e. the chloroalkyl substituent, and are discussed in light of the cation COSMO volume, sigma profile and frontier orbital theory. Overall, this systematic study allows for understanding and developing genuinely tailorable PILs.

2. Experimental

2.1. Materials and synthesis

4-(2-Chloroethyl)morpholine hydrochloride (99%), 1-(2-chloroethyl)piperidine hydrochloride (98%), 1-(2-chloroethyl)pyrrolidine hydrochloride (98%); (2-chloroethyl)diethylamine hydrochloride (99%), (2-chloroethyl)dimethylamine hydrochloride (99%), (3-chloropropyl)dimethylamine hydrochloride (96%) were purchased from Sigma Aldrich and lithium bis[(trifluoromethyl)sulfonyl]imide – Fluorad HQ-115 from 3M™, and used without further purification.

All PILs were synthesized in a simple metathesis reaction¹⁹ during which the chloride anion was replaced by the [TFSI]⁻ one. The relevant chloroalkylamine hydrochloride (0.2 mol) was dissolved in 20 mL of distilled water. In parallel, an equimolar amount of lithium [TFSI] (0.2 mol) was dissolved with 30 mL of distilled water. Both solutions were mixed together and stirred for 1 hour, and at the end the desired PIL appeared as a hydrophobic phase at the bottom of the aqueous solution. The mixture was transferred into a funnel in order to separate the PIL from the by-product (lithium chloride) and residual unreacted starting materials. The PIL product was rinsed a few times with small amounts of distilled water until the test with silver nitrate did not reveal the presence of chlorides. The obtained PILs were dried at 80 °C for 24 h under reduced pressure of 100 mPa in a vacuum dryer (Büchi Glass Oven B-585). The amount of residual water in the dried PILs was determined by the Karl Fisher method using 831 KF Coulometer (Metrohm); all PILs have less than 200 ppm water. The names, abbreviations and structural formulas of the obtained PILs are presented in Table 1. The acronyms of alicyclic cations were created as follows: morpholinium referred to as “Morph”, piperidinium to as “Pip”, pyrrolidinium to as “Pyrr”, whereas the chloroethyl substituent was denoted to as “C_2-Cl-” (or for the series of non-functionalized cations ethyl substituent referred to as “C_2-”). In the case of linear alkylammonium cations, “H” stands for hydrogen, “_2-Cl-” for chloroethyl whereas “1” means methyl, “2” ethyl and “3” propyl. As the anion is the same in all the tested PILs, they are further named by the cation abbreviation in the figures.

Table 1 Name, abbreviation and structural formula of the synthesized PILs

Name	Abbreviation	Structural formula
4-(2-Chloroethyl)morpholinium bis[(trifluoromethyl)sulfonyl]imide	[C2-Cl-Morph][TFSI]
1-(2-Chloroethyl)piperidinium bis[(trifluoromethyl)sulfonyl]imide	[C2-Cl-Pip][TFSI]
1-(2-Chloroethyl)pyrrolidinium bis[(trifluoromethyl)sulfonyl]imide	[C2-Cl-Pyrr][TFSI]
(2-Chloroethyl)diethylammonium bis[(trifluoromethyl)sulfonyl]imide	[HN2-Cl-22][TFSI]
(2-Chloroethyl)dimethylammonium bis[(trifluoromethyl)sulfonyl]imide	[HN2-Cl-11][TFSI]
(3-Chloropropyl)dimethylammonium bis[(trifluoromethyl)sulfonyl]imide	[HN3-Cl-11][TFSI]

---

¹H, ¹³C and ¹⁹F NMR spectra of the prepared PILs were obtained on a Varian VNMRS spectrometer (402.64 MHz for ¹H, 378.82 MHz for ¹⁹F and 101.25 MHz for ¹³C NMR) using DMSO-d₆ as solvent. The ¹H and ¹³C chemical shifts (in ppm) were referred to TMS internal standard, and for ¹⁹F to CFCl₃. The values obtained for the various PILs were as follows (bs-broad singlet, t-triplet, s-singlet, m-multiplet):

2.1.1 [C_2-Cl-Morph][TFSI]. ¹H NMR (DMSO-d₆) δ ppm = 3.16 (bs, 2H); 3.47 (bs, 2H); 3.56 (t, J = 6.23 Hz, 2H); 3.72 (bs, 2H); 3.99 (t, J = 6.20 Hz, 4H); 9.78 (s, 1H broad peak); ¹³C NMR (DMSO-d₆) δ ppm = 124.32; 121.15; 117.97; 114.79; 63.14; 56.79; 51.43; 37.22; ¹⁹F NMR (DMSO-d₆) δ ppm = −78.35.

2.1.2 [C_2-Cl-Pip][TFSI]. ¹H NMR (DMSO-d₆) δ ppm = 1.40 (s, 2H); 1.70 (bs, 2H); 1.81 (bs, 2H); 2.95 (bs, 2H); 3.47 (t, J = 6.51 Hz, 4H); 3.97 (t, J = 6.68 Hz, 2H); 9.46 (s, 1H broad peak); ¹³C NMR (DMSO-d₆) δ ppm = 124.33; 121.14; 117.97; 114.79; 56.49; 52.50; 37.45; 22.33; 21.11; ¹⁹F NMR (DMSO-d₆) δ ppm = −78.37.

2.1.3 [C_2-Cl-Pyrr][TFSI]. ¹H NMR (DMSO-d₆) δ ppm = 1.96 (m, 4H); 3.12 (s, 2H); 3.57 (t, J = 6.17 Hz, 2H); 3.95 (t, J = 6.00 Hz, 2H); 9.57 (s, 1H broad peak); ¹³C NMR (DMSO-d₆) δ ppm = 125.99; 121.72; 117.46; 113.19; 54.99; 53.81; 39.37; 22.49; ¹⁹F NMR (DMSO-d₆) δ ppm = −78.34.

2.1.4 [HN_2-Cl-22][TFSI]. ¹H NMR (DMSO-d₆) δ ppm = 1.21 (t, J = 7.25 Hz, 6H); 3.20 (m, 4H); 3.49 (q, J = 6.37 Hz, 2H); 3.97 (t, J = 6.35 Hz, 2H); 9.22 (s, 1H broad peak); ¹³C NMR (DMSO-d₆) δ ppm = 124.33; 121.15; 117.97; 114.79; 52.13; 46.93; 37.96; 8.36; ¹⁹F NMR (DMSO-d₆) δ ppm = −78.33.

2.1.5 [HN_2-Cl-11][TFSI]. ¹H NMR (DMSO-d₆) δ ppm = 2.84 (s, 6H); 3.49 (t, J = 6.23 Hz, 2H); 3.98 (t, J = 6.11 Hz, 2H); 9.46 (s, 1H broad peak); ¹³C NMR (DMSO-d₆) δ ppm = 125.95; 121.68; 117.42; 113.15; 57.24; 42.57; 38.08; ¹⁹F NMR (d DMSO-d₆) δ ppm = −78.35.

2.1.6 [HN_3-Cl-11][TFSI]. ¹H NMR (DMSO-d₆) δ ppm = 2.11 (m, 2H); 2.81 (s, 6H); 3.17 (t, J = 8.00 Hz, 2H); 3.71 (t, J = 6.35 Hz, 2H); 9.36 (s, 1H broad peak) ¹³C NMR (DMSO-d₆) δ ppm = 124.31; 121.13; 117.95; 114.77; 54.58; 42.44; 42.00; 26.91; ¹⁹F NMR (DMSO-d₆) δ ppm = −78.35.

2.2. Physico-chemical and electrochemical properties of the PILs

The thermal stability of the PILs was measured by thermogravimetric analysis (TGA) with nitrogen as shielding gas using a TGA/DSC 1 (Mettler Toledo). The samples were heated from 30 to 450 °C, at a rate of 10 °C min⁻¹. Differential scanning calorimetry (DSC) was carried out on a Mettler Toledo DSC 1 coupled with TC 125-MT intracooler (Huber). The samples were heated from 25 to 120 °C, and then cooled down to −100 °C and finally heated to 120 °C at a rate of 10 °C min⁻¹. A densimeter DDM 2911 (Rudolph Research Analytical) was used to measure the density (ρ) with ±0.00002 g cm⁻³ accuracy in a temperature range from 20 to 90 °C ± 0.02 °C. The refractive index (n_D) was measured with accuracy of ±0.00005 at atmospheric conditions in the temperature range from 20 to 90 °C ± 0.02 °C using a J357 Automatic Refractometer (Rudolph Research Analytical). The surface tension (γ) was measured at room temperature by the platinum ring method using a Lauda tensiometer (Germany); all measurements were repeated at least twice. The dynamic viscosity (η) was determined in a temperature range from 20 to 80 °C ± 0.01 °C at an operating speed of 500 rad s⁻¹ using a rotational rheometer AR 1000 (TA Instruments) with a conical geometry coupled with an integrated thermostat; the uncertainty on viscosity did not exceed ±0.01 Pa s. The conductivity (σ) was determined at temperatures from 20 to 100 °C ± 0.5 °C on an electrode/electrolyte/electrode system assembled in a glove box under argon atmosphere, using electrochemical impedance spectroscopy (EIS, VMP3, Biologic); the frequency range was 500 kHz–1 Hz and the amplitude of ± 5 mV; a 0.1 mol L⁻¹ KCl solution was used to calibrate the cell. The resistance values (R) were determined from the intersection of the Nyquist plot with the real axis (Z′), and the conductivity was calculated according to the equation: σ = l/RA where l is the distance between the two electrodes in the cell and A is the surface area of the electrodes.

The potential window of the PILs was determined by cyclic voltammetry in a three-electrode cell with glassy carbon (GC) as working electrode, platinum (Pt) as counter electrode and a silver wire as pseudo-reference electrode (AgQRE); the cut-off current density was 0.1 mA cm⁻² accordingly to ref. 43. The cyclic voltammograms were obtained with a VMP3 multichannel potentiostat/galvanostat (Biologic).

2.3. Computational methods

Computational methods were utilized to expound the differences in structure, charge distribution and volume between the various chloroalkyl functionalized cations, as well as between their non-chlorinated analogues (considered here as reference compounds with a simplified structure making possible to determine and describe the impact of chlorine), as previously presented in ref. 33 and 44. Firstly, each structure was optimized using the density functional theory (DFT) Gaussian version 3.0 utilizing the B3LYP (Becke, 3-parameter, Lee–Yang–Parr) method and the DGTZVP (DGauss triple zeta valence polarization) basic set. The optimized structure of each ion was then input for the generation of the COSMO file within the Turbomole program, using the BP-DFT (Becke–Perdew-density functional theory) method and the Ahlrichs-TZVP (triple zeta valence plus polarization) basic set.⁴⁵ The COSMO volume and the sigma profile of each ion were then generated by using COSMO-RS (Conductor-like Screening Model for Real Solvent) methodology within the COSMOThermX program (version 2.1, release 01.08).

3. Results and discussion

3.1. Computational evaluation of the ions properties

Literature reports^38,46,47 that two conformers of the [TFSI]⁻ anion coexist in ionic liquids in their liquid state: the first one, when the CF₃– groups are positioned on opposite sides of the S–N–S plane (transoid – conformer 1), and the second when the CF₃– groups are situated on the same side of this plane (cisoid – conformer 2); both conformers are presented in Fig. 1. The energy difference between the two conformers is only a few kJ, which explains the high flexibility of the anion due to the easy conversion of these conformers.³⁸ Owing to the –SO₂– groups in the [TFSI]⁻ anion, the negative charge is extensively delocalized by mesomery across the backbone, nitrogen and four oxygen atoms, yellow marked on the surface representing the charge density (Fig. 1).⁴⁸ Consequently, the coordinating power of [TFSI]⁻ is lowered. Since the [TFSI]⁻ anion is also bulky and asymmetric, its ionic interactions with the cations making up ILs are lowered, leading to the decrease of ILs melting point. The molecular structure of the [TFSI]⁻ anion was already interpreted using the COSMOThermX program.⁴⁹ The polar electronic charge distribution is represented as polarization charge density of the surface (Fig. 1), which results from the distribution of polarization charges given in sigma profiles (Fig. 1). As indicated in ref. 49, the sigma profile of [TFSI]⁻ given in Fig. 1 shows two peaks, one at 0.011 e Å⁻² corresponding to the –SO₂– groups acting as weak hydrogen acceptors, and the other between −0.0082 and 0.0082 e Å⁻² in the non-polar region due to the impact of the CF₃– groups.

Fig. 1 Structure, Cosmo volume and Sigma profiles of the TFSI anion and the [C_2-Cl-Pyrr]⁺ chloro-functionalized cation and its [C₂Pyrr]⁺ non-chlorinated analogue.

Apart from the impact of anion, which is common for all the investigated PILs, the role of cation type needs to be considered, particularly taking into account the effect of the chlorine substituent. Therefore, Table S1 (in ESI†) compares the Cosmo volumes of N-chloroalkyl functionalized cations and their corresponding non-chlorinated analogues. The Cosmo volume of ions containing chlorine is 21–24 ų higher than for their non-functionalized analogues, e.g., [C_2-Cl-Pyrr]⁺ with 174.11 ų vs. [C₂Pyrr]⁺ with 150.86 ų (Fig. 1). This difference is due to the higher asymmetry of the molecule when replacing one hydrogen atom by chlorine with higher van der Waals radius and longer bond length to carbon. The strong electron withdrawing effect of chlorine provokes a charge delocalization toward the halogen atom marked yellow, contrasting with light green on hydrogen atoms and pale-blue on carbon atoms of the alkyl chains or alicyclic part of the ammonium cation ring. Accordingly, in the sigma profiles, the single peak characteristic of the non-functionalized cations (Fig. S1e, g, i, k and m in ESI†) is split by the presence of chlorine in the N-chloroalkyl functionalized cations (Fig. S1f, h, j, l and n in ESI†). The peaks of the non-functionalized cations are mostly located in the non-polar region (in the range of −0.0082 to 0.0082 e Å⁻²)⁴⁹ and partially out of it, below −0.0082 e Å⁻² and in the latter case are attributed to the hydrogen bond donor region which results from the presence of acidic hydrogen (Fig. S1c–n in ESI†). In the case of the N-chloroalkyl functionalized cations, the second smaller peak is related to the presence of the terminal chlorine, and indicates weak hydrogen acceptor sites. In contrast to the other discussed cations, the different sigma profiles of the two morpholinium cations, [C₂Morph]⁺ and [C_2-Cl-Morph]⁺ (in Fig. S1c and d, respectively, in ESI†), with negative polar segment above 0.01 e Å⁻², are attributed to the impact of oxygen as hydrogen acceptor in the morpholinium ring.

3.2. Density, refractive index and surface tension

Generally, the density of ILs is in the range 1.2–1.6 g cm⁻³ at room temperature;⁵⁰ in the case of the obtained PILs with N-chloroalkyl functionalized cations, the density values are between 1.5045 and 1.6141 g cm⁻³ at 293 K (Table 2). It mainly results from the contribution of the [TFSI]⁻ anion claimed as providing higher density in comparison to ILs comprising e.g., [BF₄]⁻ or [PF₆]⁻.¹ As already extensively reported for both AILs and PILs,^1,12 the density decreases when the length of the alkyl chains in the cation and steric hindrance increases, which is confirmed by the lowest value of 1.5045 g cm⁻³ for the PIL with [HN_2-Cl-22]⁺ cation. Accordingly, the reduction of carbon atoms number in the alkyl substituent leads to the increase of density for [HN_3-Cl-11][TFSI] and [HN_2-Cl-11][TFSI]. The highest value was measured for [C_2-Cl-Morph][TFSI], which is consistent with the general observation of higher density of PILs with heterocyclic ammonium cations than with alkylammonium ones.¹² The effect of chlorine molecular weight is well-demonstrated by the higher density of [HN_2-Cl-22][TFSI] (1.5045 g cm⁻³) as compared to its non-substituted analogue [HN₂₂₂][TFSI] (1.42 g cm⁻³ taken from ref. 22). The density of PILs plotted vs. temperature shows a linear decrease with increasing temperature (Fig. 2a), which in the tested (narrow) range can be represented by eqn (1):

ρ = b + aT (1)

where a and b are fitting parameters and T is the temperature in K. As evidenced by the high values of the obtained correlation coefficients, r², (>0.999), the relationship is highly linear in the analyzed temperature range. The similar values of slope indicate comparable sensitivity to the temperature variation.

Table 2 Density (ρ) at 293 K (20 °C), fitting parameters (a and b) and correlation coefficient of fitting for density (r²) of the investigated PILs

PILS	ρ at 293 K (g cm^−3)	b (g cm^−3)	10^4a (g cm^−3 K^−1)	r^2
[C2-Cl-Morph][TFSI]	1.6141	1.9259	−10.630	0.99994
[C2-Cl-Pip][TFSI]	1.5363	1.8381	−10.290	0.99997
[C2-Cl-Pyrr][TFSI]	1.5420	1.8526	−10.583	0.99986
[HN2-Cl-22][TFSI]	1.5045	1.8071	−10.312	0.9999
[HN2-Cl-11][TFSI]	1.5875	1.9032	−10.756	0.99994
[HN3-Cl-11][TFSI]	1.5384	1.8422	−10.357	0.99993

---

Fig. 2 (a) Density and (b) refractive index of the PILs vs. temperature.

The refractive index (n_D) is a value characterizing the deviation of a light beam when it propagates from one medium to another. For all investigated PILs, n_D at 293 K was in the range of 1.4163 to 1.4391 (Table 3), which corresponds mostly to the impact of the [TFSI]⁻ anion.⁵¹ As already pointed out, in general, the refractive index decreases with the chain length in the cation.⁵² The smallest value was measured for [HN_2-Cl-11][TFSI] (n_D = 1.4163), and it increases with the number of carbon atoms in the chloroalkyl chain, like in the case of [HN_3-Cl-11][TFSI] (n_D = 1.4190), as well as for the other chains like in the case of [HN_2-Cl-22][TFSI] (n_D = 1.4230). The highest values of refractive index were measured for PILs composed of cyclic cations. Whilst the refractive index is very close for the two PILs with the six-membered ring cations – ([C_2-Cl-Morph][TFSI] and [C_2-Cl-Pip][TFSI]), the smaller value for the PIL with five-membered ring cation ([C_2-Cl-Pyrr][TFSI]) can be noted. Fig. 2b shows the plot of the refractive index vs. temperature for the studied PILs and it can be seen, for all of them, that n_D decreases linearly while temperature increases. The temperature dependence of refractive index can be described by the following eqn (2):

n_D = b + aT (2)

where a and b are fitting parameters and T is temperature. The values of slope are in a narrow range pointing to a similar effect of temperature change on the refractive index of these PILs.

Table 3 Refractive index (n_D) at 293 K, fitting parameters (a and b) and fitting correlation coefficient for the refractive index (r²) of the investigated PILs

PILS	nD at 293 K	b	10^4a (K^−1)	r^2
[C2-Cl-Morph][TFSI]	1.4391	1.5121	−2.506	0.99848
[C2-Cl-Pip][TFSI]	1.4368	1.5141	−2.640	0.99987
[C2-Cl-Pyrr][TFSI]	1.4295	1.4959	−2.285	0.99731
[HN2-Cl-22][TFSI]	1.4230	1.4961	−2.498	0.99976
[HN2-Cl-11][TFSI]	1.4163	1.4921	−2.5012	0.99974
[HN3-Cl-11][TFSI]	1.4190	1.4856	−2.375	0.99898

---

It is worth mentioning that the refractive index is related to the electronic polarizability of ions and local field inside the material. It indicates the dielectric response to an electrical field induced by electromagnetic radiations, namely light. In turn, it allows the electrical properties of materials and their response to an electronic polarization to be estimated. Hence, the evaluation of refractive index can serve as an approximate method of assessing ILs electronic properties. For this purpose, both density (Table 2) and refractive index (Table 3) were employed in the Lorenz–Lorenz equation (eqn (3)) to calculate the molar refractivity (M_R) which is a measure of the total polarizability of one mole of substance:

where M_w is the molecular weight (g mol⁻¹), ρ the density (g cm⁻³), and n_D the refractive index. The values of M_R and M_w are collected in Table 4 for the series of studied PILs with [TFSI]⁻ anion.

Table 4 Molecular weight (M_w), molar refractivity (M_R) at 293 K and surface tension (γ) at 293 K of the prepared PILs

PILS	Mw (g mol^−1)	MR at 293 K (cm^3 mol^−1)	γ at 293 K (mN m^−1)
[C2-Cl-Morph][TFSI]	430.77	70.210	41.1
[C2-Cl-Pip][TFSI]	428.80	73.091	37.7
[C2-Cl-Pyrr][TFSI]	414.77	69.410	36.9
[HN2-Cl-22][TFSI]	416.79	70.542	36.7
[HN2-Cl-11][TFSI]	388.74	61.491	37.7
[HN3-Cl-11][TFSI]	402.76	66.1161	38.4
([HN222][TFSI])	382.34	65.8165	35.9

---

The M_R values calculated according to eqn (3) are in the range of 61.491 to 73.091 cm³ mol⁻¹ at 293 K; for comparison, M_R of acetonitrile at the same temperature is 11.11 cm³ mol⁻¹. The M_R difference between chlorinated and non-chlorinated analogues, e.g., [HN_2-Cl-22][TFSI] and [HN₂₂₂][TFSI] is 3.846 cm³ mol⁻¹ and is in the same range as the contribution of a single chlorine atom, i.e. 5.967 cm³ mol⁻¹.⁵³ Hence, as suggested in ref. 12, the molar refractivity of PILs can be tailored using appropriate cation and anion couple.

The measured values of the air-liquid surface tension of the N-chloroalkyl functionalized PILs are in the very narrow range of 36.7–38.4 mN m⁻¹, except for [C_2-Cl-Morph][TFSI] with 41.1 mN m⁻¹ (Table 4). These results are consistent with literature data on AILs,⁵⁴ showing that the surface tension is significantly altered when [BF₄]⁻, [PF₆]⁻ and [TFSI]⁻ anions are coupled with the [C₂C₁Im] cation, e.g., 44.0, 40.9 and 33.0 mN m⁻¹, at 298 K, respectively, in contrast to AILs with [TFSI]⁻ and diverse cations with values ranging from 31.3 to 34.4 mN m⁻¹ (at 298 K). According to,⁵⁵ the higher values of surface tension are related with the increase of PIL cohesiveness, owing to the formation of hydrogen bonds. Such trend was also observed in,⁵⁴ where ILs consisting of [TFSI]⁻ and cations with hydroxyl group in their structure displayed increased surface tension of 34.0–38.0 mN m⁻¹ at 298 K in contrast to non-hydroxyl functionalized analogues. The significant impact of hydrogen bonding is clearly demonstrated by comparing the two PILs with six-membered ring cations, e.g., [C_2-Cl-Morph][TFSI] and [C_2-Cl-Pip][TFSI]. The higher value for the former one is attributed to the oxygen atom in the morpholinium ring serving as hydrogen bond acceptor.

3.3. Thermal properties

In general, it is worth noticing that the thermal decomposition of ionic liquids consisting of organic cations and [TFSI]⁻ anion appears at higher temperatures than with more coordinating anions like [BF₄]⁻ or [PF₆]⁻,¹ e.g., [C₄C₁Im][TFSI] with 439 °C vs. [C₄C₁Im][BF₄] with 403 °C and [C₄C₁Im][PF₆] with 349 °C.⁵⁶ The thermal stability of the tested PILs was investigated by thermogravimetric analysis and determined by the onset temperature corresponding to 5 wt% mass loss (T_5%onset). All the N-chloroalkyl functionalized PILs show high thermal stability with T_5%onset in the range 280–337 °C (Fig. 3, Table S2 in ESI†). However, [C_2-Cl-Morph][TFSI] exhibits a shift of baseline (particularly close to 100 °C) which suggests moisturizing upon contact with atmosphere, due to oxygen in the ring serving as strong hydrogen bond acceptor thus enhancing water uptake. For the PILs with alicyclic cations, the lowest thermal stability was detected for [C_2-Cl-Morph][TFSI], followed by [C_2-Cl-Pip][TFSI] and [C_2-Cl-Pyrr][TFSI] (Fig. 3a and Table S2†), which is consistent with the general trend.⁵⁶ In the case of the alkyl ammonium PILs (Fig. 3b and Table S2†), the order of stability was [HN_2-Cl-22][TFSI] < [HN_2-Cl-11][TFSI] < [HN_3-Cl-11][TFSI], and the last one was the most thermally stable among all the PILs investigated in this paper.

Fig. 3 Thermogravimetric analysis of the PILs with N-chloroalkyl functionalized cations (a) alicyclic (b) aliphatic.

Whereas the thermal decomposition of the majority of ILs is reported to occur in one step with complete mass loss and volatilization of the component fragments,¹ in case of the N-chloroalkyl functionalized PILs, it occurs in two steps (Fig. 3). The mass loss during the first step ranges from 10% to 17% (Table S2†), which seems to correspond to E2 elimination with chloroethene (vinyl chloride) release. Yet, the E2 elimination of the N-substituent from imidazolium quaternary halides has been already proved as the thermal degradation pathway and presumed as the inverse of the S_N2 substitution during their formation.¹

The DSC thermograms of the PILs with N-chloroalkyl functionalized cations are shown in Fig. 4, and the corresponding thermodynamic data are reported in Table 5. For all studied PILs, the melting temperature T_m is lower than for their homologues with non chloro-substituted cations, and close to ambient temperature, from ca. 22 °C to 34 °C, except for [HN_2-Cl-22][TFSI] with T_m = −8.5 °C. In order to evaluate the effect of chlorine in the alkyl substituent on lowering the PILs melting temperature, these data on PILs with chloroalkyl-substituted cations were compared with reported literature data on their homologues with non chloro-substituted cations. The greatest melting temperature shift is observed for [C_2-Cl-Morph][TFSI] with 22.5 °C in contrast to [Morph][TFSI] with 58.5 °C.²¹ In the case of the piperidinium PILs, no significant T_m shift is detected ([Pip][TFSI] with 37.9 °C (ref. 21) and [C_2-Cl-Pip][TFSI] with 34.1 °C). A similar effect is observed for pyrrolidinium, where the two representatives [Pyrr][TFSI] and [C_2-Cl-Pyrr][TFSI] have very close T_m of 35.0 °C and 29.4 °C, respectively. In turn, the T_m shift toward lower values was more pronounced for all the aliphatic (linear) ammonium PILs. The [HN_2-Cl-22][TFSI] displays T_m of −8.5 °C in contrast to the non-chloro [HN₂₂₂][TFSI], with melting temperature of −0.8 °C (ref. 19) or 3.5 °C.²¹ Also, when one hydrogen of the ethyl group of the dimethylethylammonium cation is substituted by chlorine, we showed that the melting temperature decreases from 57.0 °C in [HN₂₁₁][TFSI]⁵⁷ to 24.2 °C in [HN_2-Cl-11][TFSI]. The more significant shift with alkylammonium PILs in comparison to alicyclic ones can be explained by more delocalized, or precisely more alkyl-shielded charge, as well as by a higher number of rotational and vibrational degrees of freedom of linear alkyl substituents in the former ions.¹ Additionally, lower T_m and correspondingly ΔH_m values (Table 5) for [C_2-Cl-Morph][TFSI] and [HN_2-Cl-22][TFSI] indicate weaker intermolecular interactions in these PILs.⁵⁸

Fig. 4 DSC thermograms of the PILs with various N-chloroalkyl functionalized cations and [TFSI] anion: (a) [C_2-Cl-Morph]; (b) [C_2-Cl-Pip]; (c) [C_2-Cl-Pyrr]; (d) [HN_2-Cl-22]; (e) [HN_2-Cl-11]; (f) [HN_3-Cl-11].

Table 5 Thermal properties determined from DSC on the PILs with various N-chloroalkyl functionalized cations: glass transition (T_g), crystallization (T_c), melting (T_m), cold crystallization (T_cc) temperatures in °C; crystallization (ΔH_c), melting (ΔH_m) and cold crystallization (ΔH_cc) enthalpies in kJ mol⁻¹

PILs	Tg	Tc	Tm	Tcc	ΔHc	ΔHm	ΔHcc
[C2-Cl-Morph][TFSI]	−45.6	—	22.5	11.9	—	2.2	2.4
[C2-Cl-Pip][TFSI]	—	−8.2	34.1	—	26.4	32.8	—
[C2-Cl-Pyrr][TFSI]	—	−8.1	29.4	—	41.6	45.4	—
[HN2-Cl-22][TFSI]	−81.3	—	−8.5	−18.6	—	2.9	2.5
[HN2-Cl-11][TFSI]	—	−25.5	24.2	—	39.3	47.8	—
[HN3-Cl-11][TFSI]	—	−10.8	32.4	—	37.2	40.9	—

---

The studied PILs displayed two types of thermal behavior reflected in the DSC curves (Fig. 4). In the first group, including [C_2-Cl-Pip][TFSI], [C_2-Cl-Pyrr][TFSI], [HN_2-Cl-11][TFSI] and [HN_3-Cl-11][TFSI], first-order transition peaks corresponding to crystallization (T_c) and melting (T_m), with supercooling in the range of 37–50 °C, are well-visible. For the second group, including [C_2-Cl-Morph][TFSI] and [HN_2-Cl-22][TFSI], the supercooling domain is much larger, and a crystallization peak could not be observed down to −100 °C upon cooling. During heating, the two supercooled liquids demonstrated a glass transition (T_g) followed by cold crystallization (T_cc) before melting (Fig. 4a and d). Such cold crystallization is typically observed in glass-forming liquids and polymers and associated with viscous ILs incorporating bulky ions.⁵⁸ As previously reported,⁵⁰ due to the increasing ILs viscosity upon decreasing temperature, the formation of crystal nuclei required for crystallization is hindered, facilitating the occurrence of supercooled liquid below its melting point. In the case of [C_2-Cl-Morph][TFSI] and [HN_2-Cl-22][TFSI], the lack of such nuclei is maintained under the applied scan rate down to −100 °C. Yet, during the reverse scan, and going up to higher temperatures, these liquid-like systems with significant viscosity exhibit a glass transition (T_g). Such glass transitions have a kinetic nature, and glasses or supercooled liquids are considered as metastable in contrast to the energetically favorable formation of crystals within the liquid–crystallization transition having a thermodynamic nature. To make crystallization possible, the molecules in the melt must adapt their conformation (change of thermal energy), which is reflected by the (cold) crystallization upon heating.^50,58

3.4. Transport properties

3.4.1 Viscosity. Most ILs are classified as Newtonian fluids^59,60 demonstrating constant shear stress/shear rate ratio, unlike non-Newtonian fluids which do not exhibit this constant ratio. In Fig. 5a which represents the shear stress vs. shear rate for the PILs with N-chloroalkyl functionalized cations, one can observe linearity (Newtonian behavior) until a shear rate of 500 s⁻¹ and a negative deviation corresponding to non-Newtonian shear thinning behavior for higher values of shear rate. Examples of non-Newtonian ILs, both aprotic and protic, have been already published.^61–63 Separovic et al.⁶² have correlated the non-Newtonian viscous shear thinning occurring in ILs with the presence of a considerable amount of hydrogen bonds, especially when donor functional groups are present on the cation. These authors also indicated liquid-phase ordering or aggregation and pointed out the possible need for an active proton capable of creating such hydrogen bonding to participate in the aggregates formation. Accordingly, owing to the presence of chlorine, N-chloroalkyl functionalized PILs are susceptible to form an enhanced amount of hydrogen bonds, and consequently undergo the tendency in non-Newtonian shear thinning behavior.

Fig. 5 (a) Shear stress vs. shear rate at 293 K and (b) ln viscosity vs. 1/T of the PILs with N-chloroalkyl functionalized cations. The experimental data are fitted by straight solid lines.

In the PILs with N-chloroalkyl functionalized cations, the highest viscosity, determined at shear rate of 500 s⁻¹ and 20 °C, was measured for [C_2-Cl-Morph][TFSI] with 0.5953 Pa s, which was 3 times larger than for [C_2-Cl-Pip][TFSI] with 0.2097 Pa s and 7–9 times larger than for aliphatic and pyrrolidinium based PILs (Table 6). The viscosity values of ILs are commonly in the range from 0.01 to 0.500 Pa s,¹ which places the N-chloroalkyl functionalized PILs in the upper values. Generally, viscosity is dependent on the ion–ion interactions, including van der Waals forces and hydrogen bonds enhancing these interactions, and as a result leading to higher values of viscosity.¹² It is reported that viscosity increases when the length of alkyl chains increases and in presence of a hydroxyl group, which has the ability to form hydrogen bonds.³ It explains the difference in viscosity of [C_2-Cl-Morph][TFSI] and [C_2-Cl-Pip][TFSI], where the structures vary only by oxygen present in the ring of the former one, participating in hydrogen bond formation which enhances viscosity. The higher viscosity of the PIL with six-membered ring [C_2-Cl-Pip][TFSI] than five-membered ring [C_2-Cl-Pyrr][TFSI] is consistent with literature reports and attributed to the more stable chair-like conformation of piperidine in contrast to the pyrrolidine with lower energy difference between conformers.⁶⁴ In the case of alkylammonium-based PILs, [HN_2-Cl-22][TFSI] displays the lowest viscosity (0.0665 Pa s) among all the N-chloroalkyl functionalized PILs, yet significantly higher in comparison to literature values for [HN₂₂₂][TFSI], 0.039 Pa s (ref. 19) or 0.048 Pa s.²² This suggests that the strength of interactions between molecules, and thus viscosity, is intensified by the presence of the terminal chlorine, with a partial negative charge facilitating the occurrence of van der Waals forces, which overcompensate the negative charge delocalization of the [TFSI]⁻ anion, even more when chlorine is present, and promotes the formation of induced dipoles.^3,65

Table 6 Viscosity (η at 293 K) of the PILs with N-chloroalkyl functionalized cations and their Arrhenius equation parameters: η₀ (Pa s) pre-exponential factor, B_η fitting parameter (K), T temperature (K). For [HN_3-Cl-11][TFSI], the VTF equation was applied, replacing B_η by B′_η; T₀ stands either for temperature at which the viscosity goes to zero ideal or for glass transition temperature, r² is the correlation coefficient

PILs	η (Pa s)	η0 (Pa s)	Bη (K)	r^2
[C2-Cl-Morph][TFSI]	0.5953	3.300 × 10^−5	2873.5	0.9996
[C2-Cl-Pip][TFSI]	0.2097	1.004 × 10^−5	2935.2	0.9993
[C2-Cl-Pyrr][TFSI]	0.0801	1.619 × 10^−5	2485.2	0.9996
[HN2-Cl-22][TFSI]	0.0665	1.271 × 10^−5	2512.2	0.9997
[HN2-Cl-11][TFSI]	0.1093	5.769 × 10^−5	2908.8	0.9989

---

	η (Pa s)	η0 (Pa s)	B′η (K)	r^2	T0 (K)
[HN3-Cl-11][TFSI]	0.0945	1.018 × 10^−5	1006.1	0.9998	126

---

Fig. 5b shows the evolution of the logarithm of viscosity vs. 1/T, in the temperature range from 293 to 353 K, at a shear rate of 500 s⁻¹ where the PILs still exhibit a Newtonian behavior. Apart from [HN_3-Cl-11][TFSI] which slightly deviates from linearity, suggesting the influence of the distance between chlorine and the positively charged nitrogen, all the other PILs exhibit a typical Arrhenius dependence of viscosity vs. temperature and follow the eqn (4):

For [HN_3-Cl-11][TFSI], the VTF eqn (5):

was utilized to establish the temperature dependence of viscosity. The fitted parameters for the Arrhenius and VTF models are presented in Table 6.

The product B_ηR or B′_ηR (where R stands for the molar gas constant) is referred to as pseudo-activation energy corresponding to the activation energy (E_aη) for a viscous flow. Literally, it means that a smaller activation energy value indicates a lower sensitivity to a temperature variation, whereas a larger activation energy value shows that the compound viscosity is more responsive to a temperature change. The highest B_η values were determined for [C_2-Cl-Pip][TFSI] with 2935.2, [HN_2-Cl-11][TFSI] with 2908.8 and [C_2-Cl-Morph][TFSI] with 2873.5. The elevated B_η value of [C_2-Cl-Morph][TFSI] results from a stronger domination of hydrogen bonds, which in turn are susceptible to become considerably weaker with the increase of temperature, whereas in the case of [C_2-Cl-Pip][TFSI] the temperature impact overcomes the energetic barrier of the ring conformers transition. The latter is higher than for the five-membered ring conformers of [C_2-Cl-Pyrr][TFSI] with B_η value of 2485.2, suggesting that the viscosity is less sensitive to temperature. Notwithstanding, the B_η values are comparable for all the PILs with 2-chloroethyl substituent due to the dominant interactions caused by the chlorine presence, namely the induced dipoles formation in contrast to the PIL with 3-chloropropyl substituent in the cation displaying lower B′_η value. It suggests that the viscosity of PILs is more sensitive to temperature when chlorine is on the carbon in β position of nitrogen as compared to the γ one. In the latter case, the range of chlorine withdrawing effect does not affect the molecule as much as in the β position, hence the charge is less diffused, leading to weaker formation of induced dipoles and ion pairing, thus causing a weaker temperature effect on PIL viscosity.

3.4.2 Conductivity. ILs with [TFSI]⁻ anion are usually characterized by lower values of conductivity in comparison to both inorganic ([PF₆]⁻, [NO₃]⁻) and organic ([HCOO]⁻, [CH₃COO]⁻) anions.⁶⁶ Susan et al.²¹ reported conductivity values for a wide range of PILs with non-functionalized ammonium cations and the [TFSI]⁻ anion which can help to estimate the trend of cation-conductivity dependence. The conductivity of PILs with N-chloroalkyl functionalized cations follows the same evolution as for their non-chlorinated analogues. The conductivity values at 293 K increased from 0.18 mS cm⁻¹ for [C_2-Cl-Morph][TFSI], to 0.73 mS cm⁻¹ for [C_2-Cl-Pip][TFSI], 1.23 mS cm⁻¹ for [HN_2-Cl-22][TFSI], 1.35 mS cm⁻¹ for [HN_3-Cl-11][TFSI], 1.38 mS cm⁻¹ for [C_2-Cl-Pyrr][TFSI] and 1.71 mS cm⁻¹ for [HN_2-Cl-11][TFSI]. These values are very low when compared to AILs with the [TFSI]⁻ anion, e.g., [C₂C₁Im][TFSI] with 8.7 mS cm⁻¹, whereas the difference is not so significant as compared to [C₁Pyrr][TFSI] with 2.2 mS cm⁻¹.⁶⁷ The direct impact of the chlorine atom is clearly visible when comparing the conductivity of [HN_2-Cl-22][TFSI] (1.2 mS cm⁻¹) with [HN₂₂₂][TFSI] (4.4–5.0 mS cm⁻¹).^19,21 The presence of chlorine in the alkyl substituent of the discussed family of PILs alters their charge distribution in comparison to the non-functionalized analogues, in turn causing strengthened interactions between cation and anion, their association and ion pair formation, resulting in decreased ions mobility and consequently lowering of conductivity.

The evolution of conductivity with temperature was first represented in an Arrhenius plot:

where [HN_2-Cl-22][TFSI], [HN_2-Cl-11][TFSI], [HN_3-Cl-11][TFSI] and [C_2-Cl-Pyrr][TFSI] displayed negligible negative deviation, whereas the deviation became significant in the case of [C_2-Cl-Pip][TFSI] and [C_2-Cl-Morph][TFSI]. Therefore, the VTF equation:was harnessed to describe the temperature dependence of conductivity, and in this case the data could be fitted by a linear dependence for the six PILs (Fig. 6). The fitting parameters for the VTF model are presented in Table 7.

Fig. 6 Logarithm of ionic conductivity vs. (1/T − T₀) of the PILs with N-chloroalkyl functionalized cations. The experimental data are fitted by solid lines.

Table 7 Conductivity (σ at 293 K), molar conductivity (Λ, at 293 K) and VTF equation parameters of the PILs with N-chloroalkyl functionalized cations: σ₀ (Pa s) pre-exponential factor, B′_σ fitting parameter (K), T₀ stands either for the temperature at which the conductivity goes to zero ideal or for glass transition temperature. r² is the correlation coefficient

PILs	σ (mS cm^−1)	Λ (S cm^2 mol^−1)	σ0 (S cm^−1)	T0 (K)	B′σ (K)	r^2
[C2-Cl-Morph]	0.18	0.0467	1.834	172	1126.8	0.9982
[C2-Cl-Pip]	0.73	0.2051	3.857	133	1377.4	0.9989
[C2-Cl-Pyrr]	1.38	0.3699	2.630	113	1364.4	0.9991
[HN2-Cl-22]	1.23	0.3419	1.849	136	1166.8	0.9989
[HN2-Cl-11]	1.71	0.4197	1.132	143	940.2	0.9991
[HN3-Cl-11]	1.35	0.3543	2.860	126	1283.5	0.9989

---

The product B′_σR corresponds to the activation energy (E_Aσ) for the ionic conduction. Similarly to B′_ηR, higher values of B′_σR indicate a higher sensitivity to temperature changes. All N-chloroalkyl functionalized PILs display comparable B′_σ values of slope, except [HN_2-Cl-11][TFSI] with much smaller B′_σ of 940.2 demonstrating less steep temperature variation.

The different behavior of the studied PILs in the investigated temperature range, Arrhenius for viscosity (except [HN_3-Cl-11][TFSI]), and non-Arrhenius for conductivity, suggests that these two parameters are dominated by different interactions. Whilst viscosity is strongly related to van der Waals interactions which predominate in these PILs, conductivity is dominated by coulombic forces.

3.5. PILs ionicity

The ionicity of the PILs with N-chloroalkyl functionalized cations was assessed by the Walden plot (Fig. 7) showing log (molar conductivity) vs. log fluidity (η⁻¹) of the medium; values were plotted every 10 °C in the temperature range from 20 to 90 °C.^34,68,69 The molar conductivity Λ is calculated using the relation (8):(values at 293 K in Table 7). The so-called ideal line (solid line from one corner to the other in Fig. 7, established by using a 0.01 mol L⁻¹ solution) indicates that the mobility of ions is determined only by the medium viscosity, and that the number of ions present in the equivalent volume is the one indicated by the salt composition, i.e. all available ions contribute equally.⁵⁵ As clearly visible, the points representing the tested PILs are located below the “ideal line” and according to IL classification proposed by Angell⁶⁸ should be categorized as “poor ILs”. In terms of Walden rule, the PILs represented by points located below the “ideal line” have strongly interacting ions which partially associated together. [HN_2-Cl-22][TFSI], [HN_2-Cl-11][TFSI], [C_2-Cl-Pyrr][TFSI] and [C_2-Cl-Pip][TFSI] display almost parallel traces (similar slope), approaching the region of “good ILs” (here, above the dashed line), characterized in ref. 68 as ILs where each ion is surrounded by a relatively uniform shell of ions of opposite charge. A dramatic influence of temperature is observed for [C_2-Cl-Morph][TFSI], where hydrogen bonding caused by oxygen presence in the cation predominates. At elevated temperature (above 70–80 °C), hydrogen bonding to the fluorinated anion is reduced, thus viscosity is reduced and fluidity enhanced, and at the same time conductivity is improved, allowing this PIL to be rated as “good” one. As established before, the transport properties also significantly depend on the chain length between nitrogen and terminal chlorine. This is well-illustrated in Fig. 7 by the different temperature dependence of [HN_3-Cl-11][TFSI], which diverges from the trend observed for other N-chloroalkyl functionalized PILs; this particular behavior is ascribed to T-dependent ion-pairing.⁶⁹

Fig. 7 Walden plot representing log (molar conductivity) vs. log fluidity of the PILs with N-chloroalkyl functionalized cations. The points are plotted every 10 °C in the temperature range 20–90 °C. The solid line is the “ideal” Walden product line obtained with 0.01 mol L⁻¹ aqueous KCl.

3.6. Electrochemical window of the PILs

As already suggested by Wilkes et al., the electrochemical oxidation potential of the anions and cations of ILs is correlated to the energy (in eV) of the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO), respectively.⁷⁰ A higher LUMO energy of the cations characterizes a higher stability toward reduction (lower reduction potential). Recently, computational methods including HOMO/LUMO values calculation showed fairly good agreement with existing experimental data.⁷¹ Hence, by comparing the level of frontier orbital energy of N-chloroalkyl or non-chloroalkyl functionalized PILs, their oxidation/reduction stability toward the same electrode material, here glassy carbon (GC), could be predicted. As seen in Table 8, all the N-chloroalkyl functionalized cations display lower LUMO energy than their alkylammonium counterparts, e.g., [C_2-Cl-Pip]⁺ with −4.07 eV vs. [C₂Pip]⁺ with −3.24 eV and [HN_2-Cl-22]⁺ with −3.96 eV vs. [HN₂₂₂]⁺ with −3.27 eV. Accordingly, the presence of chlorine provokes an increase of the reduction potentials of N-chloroalkyl functionalized cations and thereby restricts the electrochemical window. It is consistent with the findings of Ceder et al. classifying electron-donating functional groups (e.g., alkyl or hydroxyalkyl) as cation stabilizing groups, in opposition to electro withdrawing groups such as halogen, cyanide and trifluoromethane.^71,72 The LUMO energy values of N-chloroalkyl functionalized cations are comparable to those of aprotic imidazolium cations, with reported values in the range from −4.92 to −4.51 eV,⁷³ whereas non-functionalized PILs have values similar to typical alkylammonium cations (−3.22 for [N₁₁₁₃]⁺ and −3.13 eV for [N₁₁₁₆]⁺).⁷³

Table 8 Energies of HOMO and LUMO orbitals of N-chloroalkyl functionalized cations and their N-alkyl functionalized analogues

Chloro-functionalized PILs	HOMO (eV)	LUMO (eV)	Non-functionalized PILs	HOMO (eV)	LUMO (eV)
[C2-Cl-Morph][TFSI]	−11.92	−4.32	[C2Morph][TFSI]	−11.79	−3.63
[C2-Cl-Pip][TFSI]	−11.88	−4.07	[C2Pip][TFSI]	−13.35	−3.24
[C2-Cl-Pyrr][TFSI]	−11.93	−4.21	[C2Pyrr][TFSI]	−13.76	−3.70
[HN2-Cl-22][TFSI]	−11.93	−3.96	[HN222][TFSI]	−14.27	−3.27
[HN2-Cl-11][TFSI]	−12.10	−4.55	[HN211][TFSI]	−14.68	−3.98
[HN3-Cl-11][TFSI]	−11.30	−4.13	[HN311][TFSI]	−13.52	−3.90

---

The experimental stability limits of the PILs with N-chloroalkyl functionalized cations were detected using GC as working electrode and the values of oxidation and reduction potentials are shown in Fig. 8. It can be seen that there is no straightforward correlation between the trend of increasing LUMO energy values, e.g., [HN_2-Cl-11]⁺ < [C_2-Cl-Morph]⁺ < [C_2-Cl-Pyrr]⁺ < [HN_3-Cl-11]⁺ < [C_2-Cl-Pip]⁺ < [HN_2-Cl-22]⁺ and the practical reduction potential values. It is hypothesized that the PILs viscosity has a significant influence on the detected cathodic stability values and/or that this discrepancy results from the synergetic impact of whole PIL, where cations and anions are proved to be more associated in N-chloro-functionalized PILs and to enhance stability. Nevertheless, the significance of the HOMO/LUMO calculations is warranted when comparing the cathodic/anodic stability limits of [HN_2-Cl-22][TFSI] and [HN₂₂₂][TFSI], with potential values of −1.69/2.43 V vs. AgQRE and −1.88/2.45 V vs. AgQRE, respectively (Fig. S2 in ESI†). The widest electrochemical window (EW) of 5.07 V was measured for [C_2-Cl-Morph][TFSI] – the most viscous PIL, whereas the other five PILs displayed EW in the range of 4.12–4.42 V. The cathodic limits slightly vary in the band of −1.69 to −2.05 V vs. AgQRE, except for [C_2-Cl-Morph][TFSI] with −2.39 V vs. AgQRE. As anticipated, owing to the impact of [TFSI]⁻, the oxidation limit of all PILs is situated in the narrow range of 2.37–2.68 V vs. AgQRE, making it possible to significantly enlarge the stability window in contrast to values reported for [HPyrr][HCOO] with EW of 2.55 V or [HPyrr][NO₃] with EW of 2.6 V on GC.²⁰

Fig. 8 Electrochemical window (EW) determination of the PILs by cyclic voltammetry (v = 10 mV s⁻¹) on glassy carbon. The cut-off current density is 0.1 mA cm⁻².⁴³

4. Conclusions

In summary, motivated by the demand for designing new ILs for electrochemical application, a new series of PILs composed of N-chloroalkyl functionalized ammonium cations and [TFSI]⁻ anion were synthesized and characterized by physico-chemical, computational and electrochemical approaches. For a number of properties, they were compared to their non-chlorinated analogues in order to determine the impact of chlorine on the various parameters. The application of chloro-functionality was proven to be an effective strategy for decreasing the PILs melting point. Owing to the characteristics of the [TFSI]⁻ anion, the obtained series of PILs displays an excellent thermal stability, even up to ca. 300 °C, hence a wide range of liquid state. The chlorine atom increases the values of density and refractive index, and consequently of calculated molar refractivity, suggesting that these PILs are more polarizable than non-functionalized ones, in good agreement with computed charge distribution. The investigated PILs are classified as sheer-thinning non-Newtonian fluids typical of ILs with functional atoms and groups on the cation. The temperature dependence of viscosity and conductivity revealed a strong domination of hydrogen bonds hampering the transport properties at RT; however, at high temperature, the impact of hydrogen bonds becomes negligible and the PILs approach the region of so-called “good ionic liquids”. The PILs with N-chloroalkyl functionalized cations are featured by a high electrochemical stability (>4.1 V on GC), as expected for ILs with the [TFSI]⁻ anion. The described characteristics suggest that the N-chloroalkyl functionalized PILs are suitable for electrochemical devices operating at elevated temperatures. Hence, this work demonstrates that the functionalization of cations is a good strategy to design PILs with optimized properties.

Acknowledgements

The authors are grateful to the Foundation for Polish Science for supporting the ECOLCAP project realized within the WELCOME program, co-financed from the European Union Regional Development Fund. Thanks to Dr J. Jacquemin from Queen's University, Belfast, for his help in computations.

Notes and references

1. P. Wasserscheid and T. Welton, Ionic liquids in synthesis, Wiley-VCH Verlag, Weinheim, 2007.
2. T. Welton, Chem. Rev., 1999, 99, 2071.
3. P. Bonhôte, A.-P. Dias, N. Papageorgiou, K. Kalyanasundaram and M. Grätzel, Inorg. Chem., 1996, 35, 1168.
4. N. V. Plechkova and K. R. Seddon, Chem. Soc. Rev., 2008, 37, 123.
5. R. D. Rogers and K. R. Seddon, Science, 2003, 302, 792.
6. M. J. Earle and K. R. Seddon, Pure Appl. Chem., 2000, 72, 391.
7. C. A. Angell, N. Byrne and J.-P. Belieres, Acc. Chem. Res., 2007, 40, 1228.
8. D. R. MacFarlane, S. A. Forsyth, J. Golding and G. B. Deacon, Green Chem., 2002, 4, 444.
9. E. Frackowiak, G. Lota and J. Pernak, Appl. Phys. Lett., 2005, 86, 164104.
10. M. Anouti, L. Timperman, M. El hilali, A. Boisset and H. Galiano, J. Phys. Chem. C, 2012, 116, 9412.
11. J.-P. Belieres and C. A. Angell, J. Phys. Chem. B, 2007, 111, 4926.
12. T. L. Greaves and C. J. Drummond, Chem. Rev., 2008, 108, 206.
13. C. Zhao, G. Burrell, A. A. J. Torriero, F. Separovic, N. F. Dunlop, D. R. MacFarlane and A. M. Bond, J. Phys. Chem. B, 2008, 112, 6923.
14. J. Stoimenovski, E. I. Izgorodina and D. R. MacFarlane, Phys. Chem. Chem. Phys., 2010, 12, 10341.
15. K. M. Johansson, E. I. Izgorodina, M. Forsyth, D. R. MacFarlane and K. R. Seddon, Phys. Chem. Chem. Phys., 2008, 10, 2972.
16. J. Pires, L. Timperman, J. Jacquemin, A. Balducci and M. Anouti, J. Chem. Thermodyn., 2013, 59, 10.
17. M. Anouti, J. Jacquemin and P. Porion, J. Phys. Chem. B, 2012, 116, 4228.
18. H. Nakamoto and M. Watanabe, Chem. Commun., 2007, 2539.
19. L. Timperman, P. Skowron, A. Boisset, H. Galiano, D. Lemordant, E. Frackowiak, F. Béguin and M. Anouti, Phys. Chem. Chem. Phys., 2012, 14, 8199.
20. A. Brandt, J. Pires, M. Anouti and A. Balducci, Electrochim. Acta, 2013, 108, 226.
21. M. A. B. H. Susan, A. Noda, S. Mitsushima and M. Watanabe, Chem. Commun., 2003, 938.
22. H. Matsumoto, H. Sakaebe and K. Tatsumi, J. Power Sources, 2005, 146, 45.
23. A. Balducci, R. Dugas, P. L. Taberna, P. Simon, D. Plée, M. Mastragostino and S. Passerini, J. Power Sources, 2007, 165, 922.
24. J. G. Huddleston, A. E. Visser, W. M. Reichert, H. D. Willauer, G. A. Broker and R. D. Rogers, Green Chem., 2001, 3, 156.
25. M. G. Freire, C. M. S. S. Neves, I. M. Marrucho, J. A. P. Coutinho and A. M. Fernandes, J. Phys. Chem. A, 2010, 114, 3744.
26. A. Noda, M. A. B. H. Susan, K. Kudo, S. Mitsushima, K. Hayamizu and M. Watanabe, J. Phys. Chem. B, 2003, 107, 4024.
27. S. Y. Lee, A. Ogawa, M. Kanno, H. Nakamoto, T. Yasuda and M. Watanabe, J. Am. Chem. Soc., 2010, 132, 9764.
28. D. Rochefort and A.-L. Pont, Electrochem. Commun., 2006, 8, 1539.
29. L. Mayrand-Provencher and D. Rochefort, J. Phys. Chem. C, 2009, 113, 1632.
30. L. Mayrand-Provencher, S. Lin, D. Lazzerini and D. Rochefort, J. Power Sources, 2010, 195, 5114.
31. R. Mysyk, E. Raymundo-Piñero, M. Anouti, D. Lemordant and F. Béguin, Electrochem. Commun., 2010, 12, 414.
32. L. Demarconnay, E. G. Calvo, L. Timperman, M. Anouti, D. Lemordant, E. Raymundo-Piñero, A. Arenillas, J. A. Menéndez and F. Béguin, Electrochim. Acta, 2013, 108, 361.
33. L. Timperman, F. Béguin, E. Frackowiak and M. Anouti, J. Electrochem. Soc., 2014, 161, A228.
34. T.-Y. Wu, S.-G. Su, K.-F. Lin, Y.-C. Lin, H. P. Wang, M.-W. Lin, S.-T. Gung and I.-W. Sun, Electrochim. Acta, 2011, 56, 7278.
35. T. Sato, G. Masuda and K. Takagi, Electrochim. Acta, 2004, 49, 3603.
36. J. Reiter, E. Paillard, L. Grande, M. Winter and S. Passerini, Electrochim. Acta, 2013, 91, 101.
37. C. Arbizzani, M. Biso, D. Cericola, M. Lazzari, F. Soavi and M. Mastragostino, J. Power Sources, 2008, 185, 1575.
38. A. Martinelli, A. Matic, P. Johansson, P. Jacobsson, L. Borjesson, A. Fernicola, S. Panero, B. Scrosati and H. Ohno, J. Raman Spectrosc., 2011, 42, 522.
39. J. Pernak, A. Syguda, D. Janiszewska, K. Materna and T. Praczyk, Tetrahedron, 2011, 67, 4838.
40. J. Pernak, M. Niemczak, K. Materna, K. Marcinkowska and T. Praczyk, Tetrahedron, 2013, 69, 4665.
41. J. Pernak, M. Niemczak, K. Zakrocka and T. Praczyk, Tetrahedron, 2013, 69, 8132.
42. Y. Lauw, M. D. Horne, T. Rodopoulos and F. A. M. Leermakers, Phys. Rev. Lett., 2009, 103, 117801.
43. H. Matsumoto, in Electrochemical Aspects of Ionic Liquids, ed. H. Ohno, Hoboken, New Jersey, 2005, p. 35.
44. N. A. Manan, C. Hardacre, J. Jacquemin, D. W. Rooney and T. G. A. Youngs, J. Chem. Eng. Data, 2009, 54, 2005.
45. A. Schäfer, C. Huber and R. Ahlrichs, J. Chem. Phys., 1994, 100, 5829.
46. K. Fujii, T. Fujimori, T. Takamuku, R. Kanzaki, Y. Umebayashi and S.-I. Ishiguro, J. Phys. Chem. B, 2006, 110, 8179.
47. F. M. Vitucci, F. Trequattrini, O. Palumbo, J.-B. Brubach, P. Roy, M. A. Navarra, S. Panero and A. Paolone, J. Phys. Chem. A, 2014, 118, 8758.
48. K. Xu, Chem. Rev., 2004, 104, 4303.
49. J. Palomar, J. S. Torrecilla, J. Lemus, V. R. Ferro and F. Rodríguez, Phys. Chem. Chem. Phys., 2010, 12, 1991.
50. M. Galiński, A. Lewandowski and I. Stepniak, Electrochim. Acta, 2006, 51, 5567.
51. N. Papaiconomou, J. Estager, Y. Traore, P. Bauduin, C. Bas, S. Legeai, S. Viboud and M. Draye, J. Chem. Eng. Data, 2010, 55, 1971.
52. T. L. Greaves, A. Weerawardena, C. Fong, I. Krodkiewska and C. J. Drummond, J. Phys. Chem. B, 2006, 110, 22479.
53. J. A. Dean, Lange's Handbook of Chemistry, McGraw-Hill, 15th edn, 1998.
54. H. Jin, B. O'Hare, J. Dong, S. Arzhantsev, G. A. Baker, J. F. Wishart, A. J. Benesi and M. Maroncelli, J. Phys. Chem. B, 2008, 112, 81.
55. D. R. MacFarlane, M. Forsyth, E. I. Izgorodina, A. P. Abbott, G. Annat and K. Fraser, Phys. Chem. Chem. Phys., 2009, 11, 4962.
56. C. Maton, N. De Vos and C. V. Stevens, Chem. Soc. Rev., 2013, 42, 5963.
57. B. Gorska et al., unpublished results.
58. J. L. Lebga-Nebane, S. E. Rock, J. Franclemont, D. Roy and S. Krishnan, Ind. Eng. Chem. Res., 2012, 51, 14084.
59. J. Jacquemin, P. Husson, A. A. H. Padua and V. Majer, Green Chem., 2006, 8, 172.
60. K. R. Seddon, A. Stark and M. J. Torres, in Viscosity and density of 1-alkyl-3-methyl imidazolium ionic liquids, ed. M. Abraham and L. Moens, ACS Symposium Series, 2002, vol. 819, p. 34.
61. P. S. Kulkarni, L. C. Branco, J. G. Crespo, M. C. Nunes, A. Raymundo and C. A. M. Afonso, Chem.–Eur. J., 2007, 13, 8478.
62. G. L. Burrell, N. F. Dunlop and F. Separovic, Soft Matter, 2010, 6, 2080.
63. J. Jacquemin, M. Anouti and D. Lemordant, J. Chem. Eng. Data, 2011, 56, 556.
64. T. Yim, H. Y. Lee, H.-J. Kim, J. Mun, S. Kim, S. M. Oh and Y. G. Kim, Bull. Korean Chem. Soc., 2007, 28, 1567.
65. A. Noda, K. Hayamizu and M. Watanabe, J. Phys. Chem. B, 2001, 105, 4603.
66. H. Ohtani, S. Ishimura and M. Kumai, Anal. Sci., 2008, 24, 1335.
67. S. Fang, Z. Zhang, Y. Jin, L. Yang, S.-I. Hirano, K. Tachibana and S. Katayama, J. Power Sources, 2011, 196, 5637.
68. W. Xu, E. I. Cooper and C. A. Angell, J. Phys. Chem. B, 2003, 107, 6170.
69. M. Yoshizawa, W. Xu and C. A. Angell, J. Am. Chem. Soc., 2003, 125, 15411.
70. J. S. Wilkes, Green Chem., 2002, 4, 73.
71. S. P. Ong, O. Andreussi, Y. Wu, N. Marzari and G. Ceder, Chem. Mater., 2011, 23, 2979.
72. S. P. Ong and G. Ceder, Electrochim. Acta, 2010, 55, 3804.
73. W. Buijs, G.-J. Witkamp and M. C. Kroon, Int. J. Electrochem. Sci., 2012, 589050.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra12152g

---