The microscopic molecular structure and dynamics of a new deep eutectic solvent (DES) composed of an ionic liquid (1-hexyl-3-methylimidazolium chloride) and an amide (trifluoroacetamide) at various molar ratios were investigated using linear and non-linear infrared spectroscopy with a vibrational probe. The use of the ionic liquid allows us to investigate the changes that the system undergoes with the addition of the amide or, equivalently, the changes from an ionic liquid to a DES. Our studies revealed that the vibrational probe in the DES senses a very similar local environment irrespective of the cation chemical structure. In addition, the amide also appears to perceive the same molecular environment. The concentration dependence studies also showed that the amide changes from being isolated from other amides in the ionic liquid environment to an environment where the amide–amide interactions are favored. In the case of the vibrational probe, the addition of the amide produced significant changes in the slow dynamics associated with the making and breaking of the ionic cages but did not affect the rattling-in-cage motions perceived by it. Furthermore, the concentration dependence of slow dynamics showed two regimes which are linked to the changes in the overall structure of the solution. These observations are interpreted in the context of a nanoscopic heterogeneous environment in the DES which, according to the observed dynamical regimes, appears at very large concentrations of the amide (molar ratio of greater than 1:1) since for lower amide molar ratios, the amide appears to be not segregated from the ionic liquid. This proposed molecular picture is supported by small angle x-ray scattering experiments.

Deep eutectic solvents (DESs) are a new class of solvents that have started to attract significant interest from both scientific and engineering communities.1–3 So far, many different applications of DESs have been described in the literature, such as metal plating,4,5 organic synthesis,6 carbon capture,7 oil gelation,8 biomass processing,9 etc. In addition, it is now commonly believed that these new solvents can be the next generation of designer solvents because of their unrivaled properties, such as low toxicity and their facile synthesis, as compared to commonly used molecular solvents and other designer solvents, such as ionic liquids.3 In particular, DESs are often perceived as analogs of ionic liquids because of the tunability of their physical chemistry and their many shared properties with ionic liquids, such as high dissolution power and conductivity, and low volatility and flammability.1,3 However, at the microscopic level, DESs do not share any similarity with ionic liquids, since they are binary mixtures and not a single compound.1 Moreover, it is only because of the eutectic phenomenon that DESs exist.

The eutectic phenomenon usually refers to the property of a binary mixture with a fixed proportion of compounds having a single melting temperature that is lower than the fusion point of any of the individual components.1 At the molecular level, it is now commonly accepted that the decrease in the fusion point in DESs arises from intermolecular interactions between the components.3 In particular, for type III DESs which are composed of a quaternary ammonium salt and a molecule having hydrogen bond donor groups, it has been proposed that the freezing point depression is caused by the hydrogen bond interaction between the ionic components and the hydrogen bond donor (HBD), though the definition appears to represent more interactions than single tradition hydrogen bonds.3 However, this proposed molecular mechanism for the formation of DESs does not take into consideration the presence of molecular heterogeneities in these systems even when the hydrogen bonding among components might favor specific spatial distributions among them.10–14 Hence, the molecular characterization of DESs has gained significant momentum in the last few years. However, the molecular level understanding of DESs remains incomplete. To this end, many different experimental12,15–19 and theoretical12,14,20–24 approaches have been utilized to characterize the molecular structure of these liquids. The key findings of these studies can be summarized as follows. First, DESs present extensive ionic networks with complex arrangements of the components. Second, it was established that the interaction between the HBD and the anion was key for the formation of these ionic networks.25 Third, the shape and symmetry of the cation affect the spatial and dynamical heterogeneities of the different DESs.10–12 Finally, DESs observe nanoscopic segregation of the components into different apolar and polar domains.14,26,27

An important drawback in all the studies presented so far is the relatively narrow window of concentrations studied.1,3,28 In all cases, the changes in the molecular structure of the eutectic mixture as a function of the HBD concentration were limited to a small set of concentrations at room temperature due to the phase separation of the mixture caused by the high melting points of the DES components.7 In some cases, such as the DESs formed by choline chloride and ethylene glycol or glycerol, the range of studied compositions has been extended to the pure HBDs because those particular HBDs have melting temperatures below room temperature.15,19,29 However, the study of how the addition of the HBD affects the structure of the ionic compound has been inaccessible as a result of the high melting point of quaternary ammonium salt (e.g., choline chloride melting point is 302 °C). Thus, a new DES composed of a quaternary ammonium salt and a hydrogen bond donor where the quaternary ammonium is liquid at room temperature is required. To this end, a new DES composed of 1-hexyl-3-methylimidazolium (HMIM) chloride and trifluoroacetamide (TFAm) is created and used for this work. The HMIM cation is asymmetric, so it fulfills the requirements for the ionic component typically related to a strong melting point depression of the mixture. On the other hand, the use of HMIM permits us to investigate the structure of the liquids in a wide range of mixture compositions, which were impossible to investigate in other DESs formed by choline or tetraalkylammonium salts. Note that because HMIM chloride is liquid at room temperature, these mixtures cannot be classified as true DESs. However, later in this work, it is shown that there is no difference between DESs composed of either 1-butyl-3-methylimidazolium (BMIM) or HMIM chloride and trifluoroacetamide (Scheme 1), which is liquid at room temperature (Fig. 1). In the former case, BMIM chloride (BT2) is solid at room temperature, so both should be considered as DESs. In addition, both DESs have glass transitions that are lower than the pure components (see supplementary material, Table S1).

SCHEME 1.

Chemical structure of components from left to right: 1-hexyl-3-methylimidazolium (HMIM) chloride, 1-butyl-3-methylimidazolium (BMIM) chloride, 2,2,2-trifluoroacetamide (TFAm), and tetrabutylammonium thiocyanate.

SCHEME 1.

Chemical structure of components from left to right: 1-hexyl-3-methylimidazolium (HMIM) chloride, 1-butyl-3-methylimidazolium (BMIM) chloride, 2,2,2-trifluoroacetamide (TFAm), and tetrabutylammonium thiocyanate.

Close modal
FIG. 1.

DES (BT2) formed by 1-butyl-3-methylimidazolium (BMIM) and trifluoroacetamide (TFAm).

FIG. 1.

DES (BT2) formed by 1-butyl-3-methylimidazolium (BMIM) and trifluoroacetamide (TFAm).

Close modal

The molecular structure and dynamics of the HMIM chloride DES were studied using linear and non-linear infrared spectroscopy in combination with a vibrational probe: the thiocyanate ion.30–35 In particular, the thiocyanate ion has been shown to be an excellent probe for studying the dynamics and interactions of different systems, such as liquids and interfaces.15,30,33,36–41 Moreover, the thiocyanate ion is sufficiently small (2.2 Å ionic radius)42 compared to the chloride ion (1.8 Å ionic radius)43 to intercalate within the ionic network of the ionic liquid.44–46 In addition, the use of time-resolved non-linear IR spectroscopy allows one to probe the molecular environment observed by the thiocyanate ion (vibrational probe) as well as its interactions and, more importantly, the changes in these two metrics as a function of the hydrogen bond donor concentration in the eutectic mixture. For example, previous work from our group showed that it is possible to observe the changes in the interactions when the probe is either in the pure HBD or in the DES using the thiocyanate ion.15 In addition, the structure of the solution is investigated from the HBD perspective using the amide I band of TFAm.12 Note that the two vibrational modes (amide I and nitrile stretch) have been selected among many others because they are simple, fairly localized (i.e., they are not strongly coupled to other modes), and sensitive to their immediate molecular environment.47,48 Finally, the studies are complemented with small angle x-ray scattering experiments to correlate the changes in the nanometer scale domains with the changes in the concentration of the HBD.

1-n-hexyl-3-methylimidazolium (HMIM) chloride (Alfa Aesar, 98%), 1-n-butyl-3-methylimidazolium (BMIM) chloride (iolitec, 99%), and 2,2,2-trifluoroacetamide (Oakwood, 98%) were used without further purification. Tetrabutylammonium thiocyanate (TCI, >95%) was dried in a vacuum oven at 120 °C for 24 h before its use. All samples were prepared in a glove box under nitrogen atmosphere by mixing the components at the molar ratios (Table I). The amount of water in our samples determined via Karl Fischer titration was found to be in the range of 500–1000 ppm. To achieve the mixture of the components and the formation of the eutectic mixture, the samples were vigorously shaken using a vortex and a sonicator. The probe was added at a 100 mM concentration after the liquid preparation.

TABLE I.

Compositions, molar Ratio, and abbreviations of the studied deep eutectic solvents.

ComponentsMolar ratio (salt:HBD)Abbreviations
HMIM chloride: TFAm 1:0 HMIM 
1:0.5 HT0.5 
1:1 HT1 
1:1.5 HT1.5 
1:2 HT2 
BMIM chloride: TFAm 1:2 BT2 
ComponentsMolar ratio (salt:HBD)Abbreviations
HMIM chloride: TFAm 1:0 HMIM 
1:0.5 HT0.5 
1:1 HT1 
1:1.5 HT1.5 
1:2 HT2 
BMIM chloride: TFAm 1:2 BT2 

The linear infrared spectra of the samples were collected using a Bruker Tensor 27 with a liquid nitrogen cooled narrowband MCT detector. Each spectrum has a resolution of 0.5 cm−1 and resulted from an average of 40 scans. The DES samples were held in the standard IR cell (two CaF2 windows separated by a Teflon spacer). All the IR cells were prepared in a glove box under nitrogen atmosphere to minimize moisture contamination. In addition, due to the large absorption of the samples, the spectra of the amide I band of the DESs were measured using the FTIR in the ATR (Attenuated total reflection) mode.

The 2DIR setup used in this work has been previously detailed in the literature, so only a brief description is provided here.12,49,50 The input IR pulses were generated with a Spectra Physics Spitfire Ace Ti:sapphire amplifier at a repetition rate of 5 kHz, in combination with an OPA-800C and a difference frequency generation crystal. These input IR pulses were then split into three replicas and later focused on the sample using the well-known boxcars geometry.51 The time intervals: τ (time between the first pulse and the second pulse), Tw (time between the second and the third pulse), and t (time between the third pulse and the photon echo) were monitored by computer controlled four translational stages (PI Micos). The generated photon echo with pulses in parallel polarization (〈XXXX〉) in the phase matching direction (−k1+k2+k3) was heterodyned with a fourth pulse (local oscillator) and later dispersed by using a Triax monochromator. The resulting non-linear signal and local oscillator were detected with a liquid nitrogen cooled 64 element MCT array detector (Infrared Systems Developments). Here, 2DIR data were collected by scanning τ time from −3.5 to +3.5 ps in increments of 5 fs for each waiting time in order to collect both rephasing and non-rephasing data by switching the time ordering.52 Signals were collected for waiting times from 0 to 100.0 ps with exponentially increased steps. In all the measurements, the local oscillator always preceded the photon echo signal by ∼0.5 ps. The time domain signal, collected as a function of (τ, T, λt) via monochromator-array detection, is transformed into the 2DIR spectra (ωτ, T, ωt) by means of Fourier transforms. A detailed explanation of the Fourier analysis has been described elsewhere.53,54

Small angle x-ray scattering (SAXS) measurements were performed at the 12-ID-B station of the Advanced Photon Source (APS), Argonne National Laboratory (ANL). Samples were loaded into thin-walled quartz x-ray capillaries for all measurements using 14 keV x rays, 1.0 s exposure time, and a sample to a SAXS detector distance of 1997.08 mm. SAXS data were collected on a Pilatus 2 M detector. Data reduction and background correction were performed on-site for further analysis. Pair distribution functional analysis was performed using the program ScÅtter from ∼0.1 to 0.65 Å−1 using Moore’s method.55 

The solvation of the vibrational probe thiocyanate ion in pure HMIM and its different eutectic mixtures with TFAm (HTn), as well as the DES formed by BMIM chloride (BT2) with TFAm, were first investigated by FTIR spectroscopy. This spectrum corresponds to the nitrile stretch region (2000–2100 cm−1). For the HMIM and HTn series, a small, but notable, shift is observed in the band maxima to higher frequency with the addition of TFAm. In contrast, the bandwidth described by its full width at half maximum (FWHM) is significantly increased with the addition of TFAm (Table II). Interestingly, the BT2 sample has an almost identical CN stretch band to HT2 as seen by the perfectly overlapping spectral features of the two samples in the CN stretch region (Fig. 2).

TABLE II.

Nitrile stretch band central frequency (ω0) and full width at half maximum (FWHM).

Sampleω0 (cm−1)FWHM (cm−1)
HMIM 2053.0 ± 0.2 19.8 ± 0.1 
HT0.5 2053.8 ± 0.2 22.8 ± 0.1 
HT1 2054.7 ± 0.2 25.8 ± 0.1 
HT1.5 2055.4 ± 0.2 33.0 ± 0.1 
HT2 2056.0 ± 0.2 35.5 ± 0.1 
BT2 2055.9 ± 0.2 35.5 ± 0.1 
Sampleω0 (cm−1)FWHM (cm−1)
HMIM 2053.0 ± 0.2 19.8 ± 0.1 
HT0.5 2053.8 ± 0.2 22.8 ± 0.1 
HT1 2054.7 ± 0.2 25.8 ± 0.1 
HT1.5 2055.4 ± 0.2 33.0 ± 0.1 
HT2 2056.0 ± 0.2 35.5 ± 0.1 
BT2 2055.9 ± 0.2 35.5 ± 0.1 
FIG. 2.

IR spectra in the CN stretch region of the different samples. Top panel: HMIM (black), HT0.5 (red), HT1 (green), HT1.5 (magenta), and HT2 (blue). Bottom panel: HT2 (blue) and BT2 (open squares).

FIG. 2.

IR spectra in the CN stretch region of the different samples. Top panel: HMIM (black), HT0.5 (red), HT1 (green), HT1.5 (magenta), and HT2 (blue). Bottom panel: HT2 (blue) and BT2 (open squares).

Close modal

The structure and interaction of the different eutectic mixtures were further investigated in the amide I region (1650–1750 cm−1). In all the samples containing the amide, the spectra show a broad and asymmetric band corresponding to the amide I mode of TFAm (Fig. 3).56 While the changes in the amide I band with the addition of the HBD are significant, the changes are less pronounced after the molar ratio of TFAm is larger than 1.5. Remarkably, the amide I band of TFAm displays the same behavior as the CN stretch when comparing the HT2 and BT2 samples; i.e., the bands are almost identical.

FIG. 3.

ATR-FTIR spectra in the amide I region for different samples. The top panel contains the spectra of HMIM (black), HT0.5 (red), HT1 (green), HT1.5 (magenta), and HT2 (blue). The bottom panel has the spectra of BT2 (open squares) and HT2 (blue line).

FIG. 3.

ATR-FTIR spectra in the amide I region for different samples. The top panel contains the spectra of HMIM (black), HT0.5 (red), HT1 (green), HT1.5 (magenta), and HT2 (blue). The bottom panel has the spectra of BT2 (open squares) and HT2 (blue line).

Close modal

The solvation structure and dynamics of the vibrational probe in pure HMIM and its mixtures with TFAm were also investigated via 2DIR spectroscopy. Figure 4 shows the 2DIR spectra in the CN stretch region for the different samples at three different waiting times (Tw). All the spectra show a pair of peaks along the diagonal (black line). In each pair of peaks, the red (positive) peak represents transitions between the ground state (ν = 0) and the first excited state (ν = 1). The blue (negative) peak, downshifted by ∼23 cm−1 from the red peak, involves transitions between the first (ν = 1) and second (ν = 2) excited vibrational states.50 The downshift of this negative peak evidences the anharmonicity of the nitrile stretch vibrational potential of the thiocyanate ion.15,44,57

FIG. 4.

2DIR spectra of the CN stretch region for HMIM, HT0.5, HT1, HT1.5, and HT2 at three different Tws: 0, 10.7, and 68 ps. Dashed circles show the position of the cross peaks.

FIG. 4.

2DIR spectra of the CN stretch region for HMIM, HT0.5, HT1, HT1.5, and HT2 at three different Tws: 0, 10.7, and 68 ps. Dashed circles show the position of the cross peaks.

Close modal

The time dependence of the 2DIR spectra shows that the peaks are elongated along the diagonal line at short waiting times. This peak elongation is clearly different for the different samples and represents the inhomogeneous broadening of the CN stretch due to the environment. At later waiting times, 2DIR spectra of all the samples display a change in the peak shape to more rounded due to the spectral diffusion process.50 Note that here the term spectral diffusion refers to both structural and oritentational induced components.58 Notably, none of the 2DIR peaks acquire a completely upright shape within the investigated time window. Furthermore, the HT0.5 and HT1 present positive peaks with a rhombus-like shape at 68 ps, which evidences the presence of low intensity cross peaks at this waiting time. Nonetheless, the evolution of a positive peak to rhombus-like shape is not clearly evident when the molar fraction of the eutectic mixture is larger than 1 (HT1.5 and HT2). Note that BT2 shows almost identical 2DIR spectra to HT2, completely parallel to the observations made for both FTIR spectra (see supplementary material, Fig. S1).

Finally, the short range structural ordering of the samples was studied by small angle x-ray scattering (SAXS). The SAXS diffractograms (Fig. 5) of the different eutectic mixtures (HTn) reveal the presence of a prepeak at 0.37 Å−1 in similarity to the pure ionic liquid.59–62 Moreover, the samples do not show any appreciable difference in the SAXS diffractograms.

FIG. 5.

Small angle x-ray scattering (SAXS) of the HTn DESs, HT1 (black), HT1.5 (red), and HT2 (blue).

FIG. 5.

Small angle x-ray scattering (SAXS) of the HTn DESs, HT1 (black), HT1.5 (red), and HT2 (blue).

Close modal

The structure of the eutectic mixtures was first evaluated from the ionic component perspective using the thiocyanate probe. The CN stretch of the thiocyanate ion in HMIM presents a narrow and almost symmetric band centered at ∼2053 cm−1 with a FWHM of 19.8 cm−1 (Table II), which is within the range of values previously reported for the CN stretch of the thiocyanate ion in pure and mixtures of alkyl-methylimidazolium ionic liquids.30,44,57,63 In the eutectic mixtures (HTn), the CN stretch band not only shifts to higher frequencies but also becomes significantly broader. The blue shift in the absorption of the CN band has been previously assigned to the solvatochromic shift caused by an environment in which the formation of hydrogen bonds with the nitrile group of thiocyanate is possible.38,46,64,65 Hence, the blue shift of the CN stretch band in the HTn samples could be assigned to the presence of a hydrogen bond environment produced by the existence of a HBD (TFAm). However, in true hydrogen bonding environments such as methanol, glycerol, ethylene glycol, etc., the CN stretch band has a much larger bandwidth (FWHM > 40 cm−1).15,66 Moreover, a change in the hydrogen bonding state of a nitrile group is usually accompanied by a large change in the central frequency, which in the case of methanol is ∼16 cm−1.67 Thus, the small change in the bandwidth suggests that the thiocyanate does not sense a true change in the local environment but the change in the overall structure of the solution, where the local environment of the thiocyanate ion is maintained, but there are changes in the overall molecular structure as demonstrated by the nitrile stretch band FWHM (Table II). It appears unreasonable to think that the thiocyanate ion does not form hydrogen bonds with its surroundings in the presence of large amounts of HBD. However, it has been previously established that thiocyanate hydrogen bonds through its sulfur atom while positioning its nitrile group directly toward the cation.15 Hence, the arrangement of the thiocyanate ion and the cation results in the nitrile group sensing the cation (non-hydrogen bonded) environment rather than the HBD (hydrogen bonded) environment. Note that the formation of the hydrogen bond via the sulfur atom of thiocyanate cannot be observed directly from the S–C stretch of thiocyanate because this mode overlaps with many other low frequency modes (including bending, torsion, and combinational) of the other components in the sample.

The environment around the ionic components is further deduced from the 2DIR spectra. The 2DIR spectra of all the samples do not present clear indications of a two-state system formed by a hydrogen bond and a non-hydrogen bonded environment as observed in solvents such as methanol.67 Moreover, it appears that the distinction between the two states disappears as the concentration of TFAm increases, since the cross peak (seen as a rhombus shape of the positive peak of Fig. 4) becomes imperceptible at higher concentrations of the HBD (HT1.5 and HT2). Note that the presence of these two states is supported by temperature dependent FTIR of the mixture and the ionic liquid (see supplementary material, Figs. S4 and S5). The observation of two different environments can be explained by a system having two different solution structures depending on the HBD concentration. In the low HBD concentration regime, it is expected that the HBD will be well solvated by the ionic liquid continuum, whereas in the high concentration regime, the HMIM salt segregates from the HBD forming nanoscopic domains within the eutectic mixture (Scheme 2). This hypothesis is in line with previous observations in mixtures of ionic liquids and polar solvents.68 While there is now plentiful evidence of the existence of nanoscopic domains when the concentration of HBD in the sample is high,14,26,27 the presence of a regime in which the HBD is dispersed and forms an almost ideal solution has not been previously reported. While the previous discussion is centered around the HTn eutectic mixtures, it is expected that the same considerations also apply to the BTn mixtures because of the almost identical spectroscopic signatures that the BT2 sample presents in the IR spectra (see Fig. 2 and supplementary material, Fig. S1).

SCHEME 2.

Cartoon depicting the molecular picture of the solution structure as a function of the concentration of the amide. Top numbers are the molar ratios of the ionic components to the solvent.

SCHEME 2.

Cartoon depicting the molecular picture of the solution structure as a function of the concentration of the amide. Top numbers are the molar ratios of the ionic components to the solvent.

Close modal

The evidence of the low concentration HBD regime in which the TFAm is fully solvated by the ionic component is consistent with the concentration dependence of the amide I band. The amide I bands of the HTn samples show a narrow band for HT0.5, which undergoes a significant broadening when the amount of amide in the solution is increased. The broadening in the amide I band is associated with the formation of vibrational excitons due to the strong amide–amide association in the samples as previously demonstrated.12 Hence, the broad amide I band of TFAm when the molar ratio for the HBD is 1 or above showcases the strong interaction among amides, which explains why the ionic component nanosegregates to form the nanoscopic domains. Moreover, the 2DIR spectra demonstrate a significant increase in the disorder of the eutectic solutions (inhomogeneous broadening) as seen by the increase in the diagonal width of the positive peak of the 2DIR spectra at TW = 0 ps (Fig. 4) when the TFAm concentration is increased. The change in the diagonal width is directly correlated with the changes observed in the linear IR spectrum of the nitrile stretch under the same conditions. This is in direct agreement with the structural model proposed for the solution, since the number of different environments (inhomogeneous broadening) must increase as the eutectic solution becomes more nanoscopically heterogeneous (Scheme 2). On the contrary, at low molar ratios of the HBD, the amide is likely to be far away from other amides and interact exclusively with the ionic components of the solution. Hence, the two different environments observed at low molar ratios of TFAm (molar ratios smaller than 1) are likely to represent those environments having thiocyanate ions with or without an amide in its first solvation shell (Scheme 2). Furthermore, the presence of chemical exchange, as seen by the cross peaks, is likely caused by the motion of TFAm from one solvation shell to another. The presence of such states is supported by the narrow amide I band of TFAm observed in the spectrum of HT0.5, which indicates that TFAm does not directly interact with other TFAm molecules but is rather isolated from other HBDs. Moreover, the shape of the amide I band appears to be independent of the structure of the cation of the quaternary ammonium salt, since the HT2 and BT2 samples have the same IR signatures (Fig. 3). The result indicates that the organization of the amides in the solution is similar irrespective of the chemical identity of the cation even when the cation has a larger alkyl chain, such as in the case of HMIM.

The changes in the liquid structure are also deduced from the interactions perceived by the thiocyanate ion through its dynamics. The dynamics of the spectral diffusion (containing both structural and orientational components58) of the ionic probe, obtained using the using the central line slope (CLS) methodology,69 shows a time evolution (Fig. 6), which is described by a bi-exponential function of the form

where A1 and A2, and τ1 and τ2 are the amplitudes and decorrelation times for the fast and slower components, respectively. The fitting of the spectral diffusion (Table III) reveals that the fast component of the dynamics has a time constant of ∼10 ps and a normalized amplitude of 0.15 in all the samples, while the slower dynamical component has a varying decay time, which gets faster as the concentration of the TFAm increases. It is important to note that the analysis performed here did not account for the two contributions of spectral diffusion arising from structurally and orientationally induced loss of correlations.58 While the detailed analysis of data required for evaluating the weight of each contribution to the total decorrelation dynamics is beyond the scope of this paper, it is likely that these components have the same molecular origin as that previously proposed.70 Following this assumption, the two dynamical components of the spectral diffusion are assigned to rattling-in-cage motions (fast) and making and breaking of the ionic cage (slow).15,44 The lack of change in the short time component indicates that the thiocyanate observes the same kind of interactions irrespective of the composition of the eutectic mixture. Moreover, the dynamics for the HT2 and BT2 samples are almost identical (see supplementary material, Fig. S2). This is only possible if the vibrational probe always senses the same molecular environment even when the concentration of the HBD is significantly changed from the pure HMIM sample to HT2. Hence, this result agrees with the molecular picture in which the probe remains interacting with the cation via its nitrogen atom in all the eutectic mixtures and the pure ionic liquid. In contrast, the slow decay time shows a speed-up of the dynamics as the concentration of TFAm increases. Since the slow dynamical component is related to the making and breaking of the ionic cages, it is expected that this motion is dependent on the whole ionic network because displacing an ion involves rearranging all the Coulombic interactions with the other nearby ions, so it requires a major reorganization of the ionic components beyond the first solvation shell.15,44

FIG. 6.

Normalized central line slope (CLS) decay curves from CLS of DESs, HMIM (black), HT0.5 (red), HT1 (green), HT1.5 (magenta), and HT2 (blue). The solid lines are two exponential decay fittings to the CLS as described in the text. The normalization indicates that the CLS at TW = 0 ps is set to one.

FIG. 6.

Normalized central line slope (CLS) decay curves from CLS of DESs, HMIM (black), HT0.5 (red), HT1 (green), HT1.5 (magenta), and HT2 (blue). The solid lines are two exponential decay fittings to the CLS as described in the text. The normalization indicates that the CLS at TW = 0 ps is set to one.

Close modal
TABLE III.

HMIM and HTn DES CLS fitting parameters obtained from a bi-exponential decay function.

CLS fit parameters
SampleA1τ1 (ps)A2τ2 (ps)
HMIM 0.14 ± 0.01 12 ± 1 0.85 ± 0.01 470 ± 40 
HT0.5 0.11 ± 0.01 12 ± 1 0.89 ± 0.01 330 ± 10 
HT1 0.15 ± 0.01 11 ± 1 0.85 ± 0.01 350 ± 20 
HT1.5 0.10 ± 0.01 8 ± 1 0.89 ± 0.01 240 ± 10 
HT2 0.15 ± 0.01 11 ± 1 0.85 ± 0.01 200 ± 10 
CLS fit parameters
SampleA1τ1 (ps)A2τ2 (ps)
HMIM 0.14 ± 0.01 12 ± 1 0.85 ± 0.01 470 ± 40 
HT0.5 0.11 ± 0.01 12 ± 1 0.89 ± 0.01 330 ± 10 
HT1 0.15 ± 0.01 11 ± 1 0.85 ± 0.01 350 ± 20 
HT1.5 0.10 ± 0.01 8 ± 1 0.89 ± 0.01 240 ± 10 
HT2 0.15 ± 0.01 11 ± 1 0.85 ± 0.01 200 ± 10 

The change in the slow dynamical component with the TFAm molar ratio shows two different regimes (Fig. 7), one at a low concentration of TFAm (molar ratio: 0.5–1) and the other at TFAm high concentration (molar ratio > 1). The abrupt change in the slow dynamics can only be explained by a sudden change in the eutectic mixture molecular structure, since this motion is directly related to the interionic potential. This finding is in line with the proposed changes in the structure solution derived from the FTIR spectra of the amide and 2DIR spectra of the probe, where the samples with molar ratios lower than 1:1 present different spectral features than those with higher molar ratios of TFAm. Overall, the observed dynamics of the environment is in agreement with the solution structure in which at low concentration, the amide is dispersed within the ionic matrix, and at high concentration, there is a nanosegregation of the HBD, which results in the formation of TFAm nanoscopic pools, but the formation of these domains does not completely disrupt the interaction between the ionic components.

FIG. 7.

Relationship between the longer CLS dynamic time constants of HMIM and HTn DESs and the molar fraction of TFAm. Dashed and dotted lines are to help guide the eye.

FIG. 7.

Relationship between the longer CLS dynamic time constants of HMIM and HTn DESs and the molar fraction of TFAm. Dashed and dotted lines are to help guide the eye.

Close modal

The proposed idea of the maintenance of the ionic structure from pure HMIM to the HTn eutectic mixture is further supported by small angle x-ray scattering (SAXS) experiments (Fig. 5). SAXS patterns of the different eutectic mixtures (HTn) reveal a peak at 0.37 Å−1, which corresponds to a real-space distance of ∼17.0 Å. Previously, it was demonstrated experimentally59–62 that imidazolium-based room-temperature ionic liquids including 1-hexyl-3-methyl imidazolium chloride have a SAXS feature at a low-Q value of ∼0.5 Å−1. This peak has been assigned to the separation distance between polar groups of the same charge.59 Here, the observed feature unequivocally shows that domains composed exclusively of HMIM chlorides exist in the different eutectic mixtures, and they are not significantly disrupted by the addition of TFAm to the solution (∼0.01 Å−1 maximum difference). To further probe the structure of the eutectic mixtures, pair distance distribution functional analysis, P(r), was performed at the high-Q features, as shown in Fig. 8.71 Again, minimal difference is observed, with a calculated radius of gyration (Rg) of 30.9 ± 0.5 Å for HT1, with HT1.5 and HT2 exhibiting slightly larger Rg of 31.8 ± 0.4 and 31.5 ± 0.8 Å, respectively. The minimal differences between eutectic mixtures can be observed in the analysis of the low-Q Porod region as well. For HT1, a slope of ∼−2.7 is calculated, whereas the other two eutectic mixtures exhibit a slightly increased slope to ∼−2.9. With a slope approaching −3, the eutectic mixtures can be approximated to have a similar collapsed coil geometry.72 Overall, the similarity in SAXS features in all eutectic mixtures provides further support to our molecular picture of a non-ideal solution between the HMIM chloride and the HDB, since the probability of finding two ionic components is not altered by the molar ratio of TFAm as demonstrated by the diffractograms.

FIG. 8.

Modeling of features in the small angle x-ray scattering for HT1 (black circles), HT1.5 (red circles), and HT2 (blue circles) as described in the text. Red and magenta lines correspond to the modeling of the denoted features. The inset showcases the Q region between 0.1 and 1.0 Å.

FIG. 8.

Modeling of features in the small angle x-ray scattering for HT1 (black circles), HT1.5 (red circles), and HT2 (blue circles) as described in the text. Red and magenta lines correspond to the modeling of the denoted features. The inset showcases the Q region between 0.1 and 1.0 Å.

Close modal

The molecular structure and dynamics of DESs composed of 1-hexyl-3-methylimidazolium and trifluoroacetamide were investigated from both the ionic and molecular component perspectives using FTIR and 2DIR spectroscopies. Overall, the IR experiments confirmed the previously established molecular picture of the liquid eutectic mixtures in which the amide nanosegregates from the ionic component. However, in this study, it was also found that the nanosegregation occurs at a relatively high concentration of the amide (molar ratios of larger than 1). In addition, it is observed that the aggregation of the hydrogen bond donor is not affected by making the cation structure more hydrophobic. The change in the structure of the solvent is deduced from the changes in the amide I band in the linear IR spectra as well as the spectral diffusion dynamics of the different samples. Overall, our results reveal that the nanosegregation does not occur gradually, but it requires the presence of large amounts of HBD.

See the supplementary material for the table with fusion and glass transition temperatures of the different samples (Table S1), 2D IR spectra of BT2 (Fig. S1), CLS of HT2 and BT2 (Fig. S2), and differential scanning calorimetry (DSC) thermograms (Fig. S3).

The current work was supported, in part, by start-up funds provided to D.G.K. This research used the 12-ID-B beamline of the Advanced Photon Source, a U.S. Department of Energy (DOE) Office of Science User Facility operated for the DOE Office of Science by the Argonne National Laboratory under Contract No. DE-AC02-06CH11357. The authors would also like to thank Dr. H. B. Gobeze for helping with the data analysis and figure preparation and Dr. Rafael Cueto for doing the DSC of the samples.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

1.
E. L.
Smith
,
A. P.
Abbott
, and
K. S.
Ryder
,
Chem. Rev.
114
,
11060
(
2014
).
2.
Q.
Zhang
 et al.,
Chem. Soc. Rev.
41
,
7108
(
2012
).
3.
B. B.
Hansen
 et al.,
Chem. Rev.
121
,
1232
(
2021
).
4.
R.
Bernasconi
 et al., “
Electrodeposition from deep eutectics solvents
,” in (
Intech
,
2017
), p.
235
.
5.
A. P.
Abbott
,
K. J.
McKenzie
, and
K. S.
Ryder
,
Ionic Liquids IV
(
ACS Publications
,
2007
), p.
186
.
6.
D. A.
Alonso
 et al.,
Eur. J. Org. Chem.
2016
,
612
.
7.
G.
Garcia
 et al.,
Energy Fuels
29
,
2616
(
2015
).
8.
Y.
Cui
 et al.,
ACS Appl. Mater. Interfaces
9
,
33549
(
2017
).
9.
K. H.
Kim
 et al.,
Proc. Natl. Acad. Sci. U. S. A.
116
,
13816
(
2019
).
10.
S. S.
Hossain
and
A.
Samanta
,
Phys. Chem. Chem. Phys.
20
,
24613
(
2018
).
11.
S. S.
Hossain
,
S.
Paul
, and
A.
Samanta
,
J. Phys. Chem. B
123
,
6842
(
2019
).
12.
Y.
Cui
and
D. G.
Kuroda
,
J. Phys. Chem. A
122
,
1185
(
2018
).
13.
S.
McDonald
 et al.,
J. Phys. Chem. Lett.
9
,
3922
(
2018
).
14.
V.
Alizadeh
 et al.,
ChemPhysChem
20
,
1786
(
2019
).
15.
Y.
Cui
 et al.,
Phys. Chem. Chem. Phys.
18
,
31471
(
2016
).
16.
S. L.
Perkins
,
P.
Painter
, and
C. M.
Colina
,
J. Phys. Chem. B
117
,
10250
(
2013
).
17.
S.
Chatterjee
 et al.,
J. Phys. Chem. B
123
,
9355
(
2019
).
18.
N.
Subba
 et al.,
J. Phys. Chem. B
123
,
9212
(
2019
).
19.
L.
Percevault
 et al.,
J. Phys. Chem. B
124
,
9126
(
2020
).
20.
S. L.
Perkins
,
P.
Painter
, and
C. M.
Colina
,
J. Chem. Eng. Data
59
,
3652
(
2014
).
21.
O. S.
Hammond
,
D. T.
Bowron
, and
K. J.
Edler
,
Green Chem.
18
,
2736
(
2016
).
22.
Y.
Zhang
 et al.,
J. Phys. Chem. B
124
,
5251
(
2020
).
23.
S.
Kaur
,
A.
Malik
, and
H. K.
Kashyap
,
J. Phys. Chem. B
123
,
8291
(
2019
).
24.
S.
Kaur
,
A.
Gupta
, and
H. K.
Kashyap
,
J. Phys. Chem. B
124
,
2230
(
2020
).
25.
R.
Stefanovic
 et al.,
Phys. Chem. Chem. Phys.
19
,
3297
(
2017
).
26.
Y.
Cui
 et al.,
J. Phys. Chem. B
123
,
3984
(
2019
).
27.
S.
Kaur
,
A.
Gupta
, and
H. K.
Kashyap
,
J. Phys. Chem. B
120
,
6712
(
2016
).
28.
L. J. B. M.
Kollau
 et al.,
Chem. Commun.
54
,
13351
(
2018
).
29.
C. F.
Araujo
 et al.,
Phys. Chem. Chem. Phys.
19
,
17998
(
2017
).
30.
Z.
Ren
 et al.,
J. Phys. Chem. B
119
,
4699
(
2015
).
31.
D.
Czurlok
 et al.,
Phys. Chem. Chem. Phys.
17
,
29776
(
2015
).
32.
S.
Park
,
M.
Ji
, and
K. J.
Gaffney
,
J. Phys. Chem. B
114
,
6693
(
2010
).
33.
V. P.
Roy
and
K. J.
Kubarych
,
J. Phys. Chem. B
121
,
9621
(
2017
).
34.
S. R. G.
Kankanamge
and
D. G.
Kuroda
,
Phys. Chem. Chem. Phys.
21
,
833
(
2019
).
35.
M.
Maj
,
K.
Kwak
, and
M.
Cho
,
ChemPhysChem
17
,
921
(
2016
).
36.
D.
Czurlok
 et al.,
Phys. Chem. Chem. Phys.
16
,
19643
(
2014
).
37.
M.
Ji
,
S.
Park
, and
K. J.
Gaffney
,
J. Phys. Chem. Lett.
1
,
1771
(
2010
).
38.
K.
Ohta
,
J.
Tayama
, and
K.
Tominaga
,
Phys. Chem. Chem. Phys.
14
,
10455
(
2012
).
39.
M. G.
Maienschein-Cline
and
C. H.
Londergan
,
J. Phys. Chem. A
111
,
10020
(
2007
).
40.
M.
Maj
 et al.,
J. Chem. Phys.
140
,
235104
(
2014
).
41.
X.
Chen
 et al.,
J. Phys. Chem. B
124
,
4762
(
2020
).
42.
Y.
Iwadate
 et al.,
J. Phys. Chem.
86
,
5205
(
1982
).
43.
E. R.
Nightingale
,
J. Phys. Chem.
63
,
1381
(
1959
).
44.
Z.
Ren
 et al.,
J. Phys. Chem. Lett.
5
,
1541
(
2014
).
45.
T.
Brinzer
and
S.
Garrett-Roe
,
J. Chem. Phys.
147
,
194501
(
2017
).
46.
K.
Dahl
 et al.,
J. Chem. Phys.
123
,
084504
(
2005
).
47.
Z.
Getahun
 et al.,
J. Am. Chem. Soc.
125
,
405
(
2003
).
48.
49.
M. C.
Asplund
,
M. T.
Zanni
, and
R. M.
Hochstrasser
,
Proc. Natl. Acad. Sci. U. S. A.
97
,
8219
(
2000
).
50.
P.
Hamm
and
M. T.
Zanni
,
Concepts and Methods of 2D Infrared Spectroscopy
(
Cambridge University Press
,
Cambridge, New York
,
2011
).
51.
A. C.
Eckbreth
,
Appl. Phys. Lett.
32
,
421
(
1978
).
52.
M.
Khalil
,
N.
Demirdöven
, and
A.
Tokmakoff
,
J. Phys. Chem. A
107
,
5258
(
2003
).
53.
J. B.
Asbury
,
T.
Steinel
, and
M. D.
Fayer
,
J. Lumin.
107
,
271
(
2004
).
54.
Y. S.
Kim
,
J.
Wang
, and
R. M.
Hochstrasser
,
J. Phys. Chem. B
109
,
7511
(
2005
).
55.
P. B.
Moore
,
J. Appl. Crystallogr.
13
,
168
(
1980
).
56.
D.
Troitiño
,
E. S.
De La Blanca
, and
M. V.
Garcia
,
Spectrochim. Acta, Part A
46
,
1281
(
1990
).
57.
A.
Tamimi
,
H. E.
Bailey
, and
M. D.
Fayer
,
J. Phys. Chem. B
120
,
7488
(
2016
).
58.
P. L.
Kramer
 et al.,
J. Chem. Phys.
142
,
184505
(
2015
).
59.
H. V. R.
Annapureddy
 et al.,
J. Phys. Chem. B
114
,
16838
(
2010
).
60.
A.
Triolo
 et al.,
J. Phys. Chem. B
111
,
4641
(
2007
).
61.
O.
Russina
 et al.,
J. Phys. Chem. Lett.
3
,
27
(
2011
).
62.
A. E.
Bradley
 et al.,
Chem. Mater.
14
,
629
(
2002
).
63.
A.
Tamimi
and
M. D.
Fayer
,
J. Phys. Chem. B
120
,
5842
(
2016
).
64.
R.
Yuan
 et al.,
J. Phys. Chem. B
119
,
13407
(
2015
).
65.
P. W.
Schultz
,
G. E.
Leroi
, and
A. I.
Popov
,
J. Am. Chem. Soc.
118
,
10617
(
1996
).
66.
K.-I.
Oh
 et al.,
J. Chem. Phys.
128
,
154504
(
2008
).
67.
A.
Ghosh
 et al.,
Chem. Phys. Lett.
469
,
325
(
2009
).
68.
J.
Rajbangshi
 et al.,
J. Mol. Liq.
317
,
113746
(
2020
).
69.
K.
Kwak
 et al.,
J. Chem. Phys.
127
,
124503
(
2007
).
70.
B.
Wu
,
J. P.
Breen
, and
M. D.
Fayer
,
J. Phys. Chem. C
124
,
4179
(
2020
).
71.
X.
Hong
and
Q.
Hao
,
Appl. Phys. Lett.
94
,
083903
(
2009
).
72.
K.
Schmidt
 et al.,
Adv. Mater.
26
,
300
(
2014
).

Supplementary Material