Effect of Sodium Chloride on the Profiles of the IR Spectrum Bands of Kaolinite at Moistures under Plastic Limit

Abstract

This study presents data on the IR spectra of kaolinite at a moisture of 26% and after drying. The preparation of moistened samples was made with distilled water, and the solutions of NaCl with limiting and 50% of limiting concentrations at a temperature of 21 °C. To experimentally evaluate the effect of the interaction between liquid water and sorbing basal surfaces of kaolinite on the profile of the IR spectrum bands, the spectra of distilled water and NaCl solutions were additionally studied. Comparison of the band profiles in the wavenumber range of 2750–4000 cm⁻¹ for kaolinite samples allows the conclusion that, when going from distilled water to the most saline water, the adsorption effect is inferior to the effect of solvation, and it decreases with increasing salinity. During drying of the samples, an increase in the peak intensities of the bands in the wavenumber range of 3550–3750 cm⁻¹ is observed. Observed spectral variations are qualitatively interpreted by the results of calculations performed by the DFT method with the XLYP exchange–correlation potential. The presented data can be used to test theoretical approximations and computer models for calculating the structure and properties of moist minerals and salt adsorption mechanisms.

1. Introduction

Kaolinite is one of the main soil-forming minerals, and is widely used in industry, construction, and science-intensive applications. In this regard, its properties are constantly being refined. One of the areas of research is the study of the properties of kaolinite which absorbs aqueous solutions of salts. It has been found that an increase in the concentration and/or valence of salt atoms leads to a higher value of the viscosity of the suspension based on kaolinite, while an increase in the valence of electrolyte ions leads to a lower value of the Bingham shear stress [1]. It was established in Ref. [2] that aggregates are formed and settle faster in saline water. The results of studies performed in Ref. [3] made it possible to clarify the correlation between the atomic–molecular processes underlying the mechanism of water film formation, and the plastic properties of soil-forming minerals.

The study of the properties of NaCl–water solutions shows that an increase in salt concentration increases the rate of evaporation [4]. When water moistens kaolinite particles, the subsequent drying leads to the appearance of a salt film on the surface of the mineral. Formation of a salt film can significantly modify the properties of the surface, affecting the adsorption capacity of kaolinite [5].

The study of interactions in the kaolinite–water–salt system opens up wide opportunities for solving practically significant problems. The study of NaCl dissolution processes at the atomic level, performed in Ref. [6] using low-temperature scanning tunneling microscopy (STM) methods, made it possible to understand the dissolution process and to propose approaches to the design of NaCl-based insulating films. The activation of kaolinite surface by NaCl films leads to a significant change in the orientation of the sheets. In this case, automorphic pseudohexagonal plates are replaced by worm-like ones, the pattern of which provides maximum orientation for reactions involving interlayer spaces [7].

The mechanisms of interaction in the kaolinite–water–salt system at the atomic level are carried out by experimental and theoretical methods and are widely discussed in the literature. The authors of [8] studied the mechanisms of hydration of kaolinite by means of IR spectroscopy and density functional theory (DFT). Simulations showed that the H₂O molecules are predominantly adsorbed by the hydroxyl surface of kaolinite. The most energetically favorable position of a water molecule is the active center formed by three hydrogen atoms. On the siloxane surface, which has less hydrophilicity, a chemical bond is formed between the hydrogen atoms of H₂O and two surface oxygen atoms. The calculations of the adsorption energies of water molecules on the basal surfaces of kaolinite by the methods of DFT and molecular dynamics (MD) are performed in Refs. [9,10]. Atomic simulation made it possible to establish that the hydration of the hydroxyl surface in kaolinite is energetically more favorable than that of the siloxane one. In Ref. [11], the dependence of the position of the maximum and the width of the stretching-vibrations band of water at low moistures were established. Theoretical calculations of hydrated kaolinite were performed in Ref. [12] with the Hartree–Fock approximation. The results of calculating the profiles and positions of the bands of the IR spectra in the 2750–4000 cm⁻¹ range made it possible to qualitatively interpret experimentally measured regularities, i.e., the shift in the band maximum towards lower wavenumbers, and an increase in the relative intensity of the band with an increase in the moisture of the kaolinite samples. This effect is interpreted as a manifestation of the chemical bonding of surface atoms and the atoms of water molecules, which leads to an increase in the masses of oscillators and their number with an increase in the moisture content of the mineral.

The possibility of recording changes in the IR spectra that occur during the formation of water films of different thicknesses on the basal surfaces of minerals and the interpretation of the recorded optical effects created by the features of chemical bonding of atoms on the surfaces of minerals and water molecules make it possible to start solving a similar problem in a more complex system including salts that form solvate shells.

In this connection, in this work, the bands in the FTIR spectra of samples based on sodium chloride, water, and kaolinite in the wavenumber range 2750–4000 cm⁻¹ are measured. Unlike other studies of the properties of such or similar systems, the studies performed are aimed at elucidating the features of the formation of the spectral bands during the formation of layers of moisturizing solutions with different concentrations of sodium chloride on the surface of kaolinite particles at moisture content levels close to the plastic limit (26%) [13], as well as during the formation of sodium chloride layers after drying the sample. At a selected moisture close to the plastic limit, a film of water molecules forms near the basal surfaces, which makes it possible to experimentally study the combined effect of adsorption and solvation mechanisms on the profiles and spectral positions of the IR bands.

To specify the problems with possible theoretical studies of the mechanisms leading to spectral changes associated with moisture and concentration of sodium chloride in moisturizing solutions, a qualitative interpretation of the experimental results obtained is carried out in this work. The necessary calculations of the optimal position of atoms in the system “basal surface of kaolinite—water cluster—sodium chloride molecule” is performed using the DFT method.

2. Materials and Methods

2.1. Samples Preparation

Eleven types of samples were used for experimental studies; they are described in

Table 1. Types of samples studied.

Sample Type	Sample Content
1	Zero moisture kaolinite
2	Kaolinite with a moisture of 26% moistened with distilled water
3	Kaolinite with a moisture of 26%, moistened with a 0.178 g/g NaCl solution
4	Kaolinite with a moisture of 26%, moistened with a 0.356 g/g NaCl solution
5	Sample obtained by drying sample 3
6	Sample obtained by drying sample 4
7	Kaolinite with a moisture of 26% obtained from sample 4 by drying to constant weight and moistening with a 0.356 g/g NaCl solution
8	Sample obtained by drying sample 7
9	Distilled water
10	NaCl solution with a concentration of 0.178 g/g
11	NaCl solution with a concentration of 0.356 g/g

. Kaolinite (KGa-1b) obtained from The Clay Minerals Society [14], double-purified distilled water, and chemically pure NaCl salt were used in their preparation.

The properties and mineral composition of KGa-1b were studied in Ref. [15], where electron microscopy revealed the presence of kaolinite stacks and plates in KGa-1b. At that, kaolinite KGa-1b has slightly higher crystallinity than KGa-1. According to this work, KGa-1b is a fairly pure kaolinite. No other clay impurities were detected, although anatase, Fe oxides, quartz, and micas were observed in trace abundances. According to Ref. [16], KGa-1b samples consist of kaolinite by 96% and have trace dickite (3% anatase, 1% crandallite, quartz).

Dry kaolinite (sample 1) was prepared by grinding in a porcelain mortar for 10 min. When grinding, only a slight mechanical action was applied, which made it possible to avoid mechanochemical activation of the samples. After mechanical treatment, the samples were dried to a constant weight at a temperature of 105 °C.

To prepare samples with 26% moisture, distilled water was used (sample 2), to which sodium chloride was added at a concentration of 0.178 g per 1 g of water, g/g (sample 3) and 0.356 g/g (sample 4). Initially, moistened samples were dried to a constant weight at a temperature of 105 °C, after which they were ground in a porcelain mortar for 10 min (samples 5 and 6). To prepare sample 7, sample 4 was moistened to 26% with a solution of sodium chloride with a concentration of 0.356 g/g. Sample 9 was obtained by drying sample 7. In addition, the spectra of distilled water and used sodium chloride solutions (samples 9, 10, and 11) were measured. To reduce random errors, each IR spectroscopic measurement was performed for five independently prepared samples of each type, and the IR spectra of each sample was found by averaging the results of several individual measurements.

2.2. Instrumentation and Measurements

IR spectra were obtained by the frustrated total internal reflection (FTIR-ATR) method with the FTIR spectrometer Alpha (Bruker Optik GmbH, Ettlingen, Germany). The spectra were measured in the mid-IR range from 500 to 4000 cm⁻¹ using the OPUS program (Bruker Optics GmbH, Ettlingen, Germany). Registration of the IR spectra was carried out using the Alfa-E module. The recording mode provided a resolution of 2 cm⁻¹ when averaged over 50 scans. The natural surface of a sample placed on a ZnSe crystal with an area of 19.6 mm² was studied. The spectra were recorded using a DTGS (Deuterated tri-glycine sulphate) detector. When forming samples of a given moisture content, all samples were preliminarily dried in an HD KWGDS62IF climate chamber (Hyde Science and Technology Limited) to a constant weight at a temperature of 105 ± 2 °C, according to the technology [17]. When preparing sample 2 (Table 1), moistening to the required moisture content (W) was carried out by adding distilled water in an amount determined by the ratio [17]:

$$\mathrm{W}(\%)=\frac{m_{H_{2}O}}{m}·100\%$$

(1)

where $m_{H_{2}O}$ is the mass of water, and m is the mass of dry kaolinite. When preparing samples 3, 4, and 7 (Table 1), moistening to the required moisture content was carried out by adding a solution with a mass m_r determined by the ratio:

$$m_r=m_{H_{2}O}·\left(1+k\right),$$

(2)

where k is the NaCl concentration used. Mass measurements were made on an AV1200-1 analytical balance (OKB VESTA, St. Petersburg, Russia) with an accuracy of 0.001 g.

2.3. Processing and Graphical Representation of Measurements Results

All studied spectra of samples of each type from Table 1 were obtained by averaging the spectra from five independently prepared samples. When averaging the spectra or comparing the spectra of samples from Table 1 containing kaolinite, the spectra were normalized so that the intensities of the bands at 1005 cm⁻¹ arising from vibrations of the valence Si–O bond were equal (Figures 1–5). The choice of this band for the normalization of the spectra is due to the fact that there are no H₂O and NaCl bands nearby, and its intensity should only slightly depend on the concentration of the sodium chloride solution in the sample. When averaging the spectra or comparing the spectra of samples that do not contain kaolinite, a libration band of water in the range of 1900–2400 cm⁻¹ was used for normalization (Figure 7). When comparing the spectra of samples of solutions and samples containing kaolinite, the intensities of the water band in samples not containing sodium chloride were equated (point d, Figure 15).

In all the figures, the abscissa shows the wavenumbers in cm⁻¹, and the ordinate shows the intensities of the FTIR spectra in arbitrary units. Comparison of intensities of the spectral bands is carried out after additional normalization described in the discussion of the figures.

2.4. Qualitative Interpretation of Experimental Results by Atomic Modeling

For a qualitative interpretation of some features of the IR spectra, the DFT method was used to model the atomic structure of clusters of sodium chloride solutions and basal surfaces of kaolinite.

This study takes into account that the size of the water cluster used should provide the experimentally observed solubility of NaCl at the temperature of the experiment (21–24 °C), equal to 0.359 g/g. Then, to dissolve one NaCl molecule (molecular weight 58.44), the minimum water cluster (molecular weight 18.00) must contain nine H₂O molecules. Such a cluster was used for subsequent studies of this work. At the same time, it should be noted that the procedure for separating one representative cluster from liquid water to calculate the profile of the IR spectrum bands is not mathematically justified. Since the study of the effect of cluster isomerism on the spectra is not addressed in this study, the results obtained with the cluster used are of a qualitative nature and reflect only the properties of the used cluster.

When constructing clusters of kaolinite basal surfaces, the crystal was split in such a way that the siloxane surface cluster contained silicon atoms and oxygen tetrahedra surrounding them; the hydroxyl surface cluster contained aluminum atoms, fragments of oxygen octahedra, and surface hydrogen atoms [12]. The use of this cluster model for qualitative interpretation of spectra was substantiated in [12] by an example of solving a similar problem of ab initio calculation of the IR spectrum of hydrated kaolinite. The work [12] demonstrated reasonable agreement between the densities of occupied and free electronic states calculated by DFT methods in the cluster model, and the model of an infinite crystal. In addition, calculations of the profile of theoretical IR spectrum in the range of 2500–4500 cm⁻¹ made it possible to qualitatively interpret the shifts in the maximum of the experimental band towards lower wavenumbers and the increase in the relative intensity of the band with the increasing moisture of the kaolinite samples. Since the hydration of the siloxane surface in kaolinite is energetically less favorable than that of the hydroxyl surface, in this work, calculations were performed using the cluster representing only the hydroxyl surface containing 60 atoms.

For calculations of electronic states and correlations in clusters of solutions, the “triple split” McLean–Chandler (MC) basis set was used, which was supplemented by three polarization functions of p-symmetry, two polarization functions of d-symmetry, and one diffuse function on atoms heavier than hydrogen. To perform calculations using clusters of kaolinite basal surfaces, diffuse functions were not taken into account due to the loss of orthogonality of the basis set. Zero-approximation kaolinite crystal structure was set using the experimental data available at http://rruff.geo.arizona.edu/AMS/amcsd.php, accessed on 10 March 2022 (American Mineralogist Crystal Structure Database, database_code_amcsd 0012237, accessed on 10 March 2022) [18].

Optimizations of the atomic positions and calculations of the clusters’ properties were performed using the Firefly v.8.2.0 package [19], which is partially based on the GAMESS (US) source code [20]. The XLYP exchange–correlation functional was used in the calculations. When implementing the numerical procedure, the convergence thresholds for the density and self-consistent field energy were chosen to be 0.00001.When optimizing the spatial positions, the calculations were stopped when the RMS value of the components of the energy gradient vector became less than 0.00003. With this degree of optimization, the Hessians typically have residual frequencies below 10 cm⁻¹.

Cluster structures shown in the figures of this work were visualized using the MacMolPlt v 7.7 software package [21]. The following designations of atoms are used: ●—Si, ●—Al, ●—O, ●—H, ●—Cl, and ●—Na.

3. Results

3.1. Accuracy of Spectrometric Intensity Measurements

The accuracy of measuring the intensities of the IR spectra is illustrated with the spectra of sample 1. In Figure 1, black, red, green, blue and orange color lines show regions of the IR spectra of five independently prepared samples in the range of 800–1200 cm⁻¹. After normalization at point “a” with k = 1005 cm⁻¹, there is a good agreement between the intensities of the spectra of independently prepared kaolinite samples.

Figure 2 shows regions of the spectra of five independently prepared samples of dry kaolinite. The spectra are normalized so that their intensities coincide at point “a” with a wavenumber of 1005 cm⁻¹. of the spectra of five independently prepared samples in another narrow range of 3575–3725 cm⁻¹.

In this region, the spectra have noticeably different intensities. This makes it necessary to determine the errors in measured intensities.

Figure 3 shows the overview IR spectrum of kaolinite averaged over the spectra of five samples in the range of 500–4000 cm⁻¹.

3.2. IR Spectra of Kaolinite Moistened with Sodium Chloride Solutions of Different Concentrations

Figure 4 shows regions of the IR spectra of sample 2 (black line), 3 (red line), 4 (blue line), and 7 (green line) moistened to 26%. There is a change in the position and peak intensity of the band of stretching vibrations of H₂O (2750–4000 cm⁻¹) with an increase in the concentration of sodium chloride in the solution (samples 2, 3, and 4). As the concentration of NaCl in the moistening solution grows, the bandwidth decreases. At the maximum possible salinity of the solution, as shown in Figure 4, the left part of the band width, measured at half maximum, is reduced by 41 cm⁻¹. Unfortunately, it is impossible to study the change in the right part of the band width due to the overlapping with the bands of kaolinite. Comparison of the spectral bands in Figure 4 allows one to note the dependence of the profile of the band on the salt concentration. This justifies the interest in the joint analysis of the effect of hydration and adsorption on the spectral bands of stretching vibrations of water in kaolinite.

3.3. IR Spectra of Kaolinite Dried after Moistening with Sodium Chloride Solutions at Different Concentrations

Figure 5 shows regions of the IR spectra of samples 1 (black line), 5 (red line), 6 (blue line), and 8 (green line). During the preparation of these samples, the surfaces of kaolinite particles adsorbed different amounts of NaCl ions, which is reflected in the change in the intensity of the bands in the range of 3550–3750 cm⁻¹. Figure 5 shows that with an increase in the concentration of ions on the surfaces of kaolinite particles, the intensity of the bands increases. However, with an increase in concentration, the change in intensity weakens. At the same time, the accuracy of determining the spectral position does not allow registering of the bands’ chemical shifts. The observed effects make relevant the study of the effect of NaCl adsorption on the IR spectra of kaolinite.

3.4. IR Spectra of Water with Changing NaCl Concentration

Figure 6 shows regions of the IR spectra of samples 9 (black line), 10 (red line), and 11 (blue line). The spectra in Figure 6 are normalized so that the peak intensities of the libration band are equal, as shown in Figure 7. In Figure 7, the deformation band is additionally shown and the colors of the lines are the same as in Figure 6. Comparison of the spectra in Figure 6 allows one to note an increase in the peak intensity of the water band, as well as a shift in the band’s peak to larger wavenumbers with increasing sodium chloride concentrations. Simultaneously, the band’s full width at half maximum (FWHM) decreases. This decrease at the maximum salinity of the solution reaches 56 cm⁻¹.

4. Discussion

4.1. Bands in the IR Spectra of the Kaolinite–Water–Sodium Chloride System

Figure 3 shows an overview IR spectrum of KGa-1b kaolinite in the range of 500–4000 cm⁻¹ after drying according to the technology [17]. The nature of spectral bands of kaolinite is quite fully discussed in the literature [22,23,24,25]. Stretching bands are observed in the spectrum of kaolinite (3694 cm⁻¹, 3669 cm⁻¹, 3653 cm⁻¹, and 3618 cm⁻¹). Deformations of the hydroxyl groups of the inner surface form bands at 913 cm⁻¹ and 937 cm⁻¹. Si–O bond stretching forms bands with wavenumbers 1100 cm⁻¹, 795 cm⁻¹, 755 cm⁻¹, and 695 cm⁻¹, as well as in-plane Si–O stretching bands (1030 cm⁻¹ and 1010 cm⁻¹) and Al–O−Si deformation bands (540 cm⁻¹).

When water is adsorbed on kaolinite particles, the vibrational lines of the molecules [26,27] υ₃ (3755.79 cm⁻¹), υ₁ (3656.65 cm⁻¹), and υ₂ (1594.59 cm⁻¹) turn into broad bands associated with valence and bending vibrations (3440 cm⁻¹) and deformation vibrations (1650 cm⁻¹).

The mechanisms of broadening of the bands associated with vibrations of the OH group atoms have been quite fully and long ago discussed in the literature. It is known that hydroxyl groups that do not participate in the formation of hydrogen bonds in the liquid phase usually give narrow bands in the IR spectrum while associated groups give broad intense absorption bands at lower frequencies [28]. Due to the hydrogen bond [29] between molecules, the valence band in liquid water is shifted towards lower wavenumbers while the bending vibration band is shifted to greater wavenumbers. Both the profile and position of liquid water bands depend on temperature [30], and the temperature dependence of liquid water band characteristics is very complex [31].

In addition, the complexity of the IR spectrum in the region of OH⁻ stretching vibrations can be a manifestation of the tunneling effect [32,33], overtones, and combinations of overtones with librations [34]. An absorption peak at 2270 cm⁻¹ is assigned in Ref. [35] to the overtone of the librational mode (3ν_L) or the association band of the bending and librational mode (ν_L + ν₂).

The presence of sodium chloride near the basal surfaces of kaolinite, or in a hydrated state in adsorbed water, can be reflected in the IR spectrum [36,37,38]. However, NaCl molecules do not have vibrational states in the considered range of wavenumbers. The calculation performed in this work gives the wavenumber of stretching vibrations equal to 347 cm⁻¹. Thus, the spectral effect from the appearance of NaCl molecules near kaolinite surfaces noted in [36,37,38] can be determined by the interaction of Na⁺ and Cl^– ions with surface hydroxyl groups (adsorption effect), and with adsorbed H₂O molecules (hydration effect). Both effects will be observed in the wavenumber range of 2750–4000 cm⁻¹, and therefore further analysis of the results of spectral measurements will be mainly related to this range of wavenumbers.

4.2. The Accuracy of Determining the Bands’ Intensities

The spectra shown in Figure 2 make it possible to estimate the errors in determining the bands’ intensities. Average relative errors in spectral intensity in each spectral point m are calculated with

$${\delta }_m=\frac{{\Delta }_m}{I_m}·100\%,$$

(3)

where ${\Delta }_m$ are the average values of the deviations of the intensities of the spectra of individual samples from the intensities of the averaged spectrum (Figure 3) and $I_m$ is the intensity of the averaged spectrum (Figure 2). Certain calculated ${\delta }_m$ values amounted to 3%. At the same time, the relative errors in determining the intensity of the most prominent bands at 3619 cm⁻¹ and 3688 cm⁻¹ turned out to be smaller, at about 2%. An analysis of the spectra of samples of other types (Table 1) leads to close error values; thus, the estimate of $\delta =$2–3% in determining the intensity of the bands can be taken as the final one for the whole series of measurements.

4.3. Changes in Samples Moisture during Spectral Studies

Kaolinite KGa-1b is hydrophilic, which, in the course of sample preparation and spectral studies of the dried mineral, can lead to uncontrolled absorption of water and change in the intensity of adsorbed water bands. In this work, to control the change in the moisture content of dried samples, we used the correlation of the intensity of the libration band in the range of 2300–2400 cm⁻¹ and the moisture content determined by direct measurement. This band is chosen for monitoring moisture because it lies far away from other bands of the spectrum, which simplifies the analysis of its intensity.

To build a correlation between the intensity of the libration band and moisture content, the sample was dried to a constant weight according to the procedure [17], after which 10 spectra of the sample were recorded with an interval of 5 min. At the end of the experiment, the moisture content of the sample was determined, which amounted to 1.16%. Figure 8 shows the band shapes at different measurement times after normalizing the intensity at point “a” at 1005 cm⁻¹ to the value used when plotting the spectra in Figure 1.

Figure 8 shows that the band profile is essentially determined by resonance lines. In this regard, in this work, we studied the averaged integrated band intensity in the 2250–2400 cm⁻¹ range as a function of times at which the measurements were made. The results are presented in Figure 9. It can be seen that the change in the average integrated intensity is approximated quite well by linear dependence, shown by the red line. Assuming that the sample moisture is proportional to the integral intensity of water libration band, sample moistures were calculated based on measured libration band intensities. They are marked on the right vertical axis in Figure 9.

To estimate the moistures of dried samples in the process of spectral measurements, the integral intensities of the 2250–2400 cm⁻¹ libration band in the spectra shown in Figure 2 were measured. They are shown in Figure 9 with horizontal lines, the colors of which are similar to those used in Figure 1 (black, red, green, blue, and orange). It can be seen that the moistures of individual samples vary within 0.1–0.2%. Thus, the error in determining the moisture of dried samples (Table 1) can be estimated at 0.2%.

Moist mineral KGa-1b in natural conditions quickly loses excess moisture. Figure 10 shows diffuse bands of water in kaolinite measured in the range of 2750–3750 cm⁻¹ with a time interval of 15 min. In Figure 11, the peak band intensities measured during drying are compared with the results of direct moisture measurements. The initial moisture content of the sample was 26% and is marked with a black horizontal line. Figure 10 and Figure 11 show that the stabilization of sample moisture when performing spectral measurements takes about 60–70 min. It can be seen that the moisture of the naturally dry state of the kaolinite sample placed in the experimental setup is in the range of 1–2% (red horizontal line in Figure 11). A detailed analysis of the obtained dependences of the peak intensity of the water diffuse band on time at the beginning of drying allows us to conclude that to provide the moisture content of a moistened sample with an accuracy of 0.3%, the experiment time should not exceed 30 s.

4.4. Effect of Salinity on the Profile of the Valence Band of Water Adsorbed by Kaolinite

Figure 4 shows a region of the spectrum, including a broad band associated with stretching vibrations of H₂O and resonant stretching bands of hydroxyl groups. It can be seen that an increase in the salt concentration in the moistening solution leads to the suppression of the intensity of the water band components with lower wavenumbers near 3250 cm⁻¹. Simultaneously, the spectral components located near 3390 cm⁻¹ increase their intensity and shift towards larger wavenumbers, so that the maximum of the sixth-type sample band is at 3409 cm⁻¹. Measured increases in the peak intensities at 50% and 100% water salinity turned out to be 6% and 10%, which exceeds the measurement error by more than a factor of two. Variation in the band profile upon the change in the salinity of the used solution manifests itself in a change in the band’s width, which is normally measured at half maximum (FWHM). In our case, as can be seen in Figure 4, the presence of resonant stretching bands of the hydroxyl groups of kaolinite makes it difficult to measure the FWHM. In this regard, only the left parts of the widths were measured. It turned out that the use of solutions with the maximum concentration of sodium chloride leads to a decrease in the left part of the FWHM by 41 cm⁻¹.

When comparing the spectra of samples 4 and 7 (Table 1), it is necessary to take into account the peculiarity of their preparation, which consists of additional modification of the properties of the surfaces of mineral particles from sample 7 with NaCl molecules. When preparing the samples, solutions with the highest possible salt concentration were used. Then, the previously noted similarity in the profiles of the bands of samples 4 and 7 can be associated with a greater influence of the hydration effect on spectral characteristics compared to the differences in the adsorption of an aqueous solution of sodium chloride on modified and unmodified surfaces of kaolinite clay particles.

4.5. Effect of Salinity on the Profile and Position of Kaolinite Bands in the Range of 3550–3750 cm⁻¹

Drying of samples 3 and 4 causes the deposition of NaCl from solvation shells onto the surface of kaolinite particles. This leads to an increase in the interaction of hydroxyl groups of kaolinite and NaCl molecules, which manifests itself in an increase in the peak intensities of the hydroxyl groups stretching bands (3695 cm⁻¹, 3668 cm⁻¹, 3652 cm⁻¹, and 3620 cm⁻¹), see Figure 5. Measured increases in the bands’ peak intensities at 50% and 100% water salinity were 15% and 20%, and significantly exceed the measurement errors.

For a qualitative interpretation of this spectroscopic effect, we calculated optimal positions of the NaCl molecule near the hydroxyl basal surface of kaolinite. The resulting spatial arrangement of atoms is shown in Figure 12a,b. The change in the optimal spatial arrangement of ions near the surface was provided by changing zero-approximation coordinates when using a numerical optimization routine. Both variants of optimal positions are characterized by similar distances between Na⁺ and Cl⁻ ions, 2.52 Å (case a) and 2.48 Å (case b). Both cases demonstrate the occurrence of interaction between Cl⁻ ions of the NaCl molecule and H⁺ of the hydroxyl surface, as a result of which the intensity of the stretching vibrations of the O–H group can change. At the same time, the noted spectral effects observed with an increase in the salinity of kaolinite may also be associated with the modification of the shape of kaolinite grains, in particular, with the change of automorphic pseudohexagonal plates to worm-like ones [7]. This mechanism was not investigated in this work.

4.6. Effect of Salinity on the Profile of the IR Spectrum of Water with a Change in the Concentration of Sodium Chloride

Comparison of the spectra shown in Figure 6 allows one to note a decrease in the width and an increase in the peak intensity of the water band in the range of 2750–4000 cm⁻¹. The increases in peak intensities of the bands at 50% and 100% water salinity were 2% and 9% and show a tendency to increase with increasing salinity. In addition, there is a shift in the position of the band peak from 3340 cm⁻¹ to 3380 cm⁻¹ and a decrease in the band width by approximately 56 cm⁻¹. This may be associated with the formation of solvate shells. These effects are accompanied by the fact that the bands’ components with lower wavenumbers are suppressed, while the intensities of the components with higher wavenumbers increase. The interaction of water molecules and sodium chloride ions can cause these spectral changes.

Comparison of the data shown in Figure 7 confirms a small dependence of the position and profile of the deformation band of water (1645 cm⁻¹) on the concentration of salts in solutions, as is known from the literature, while the dissolution of salts significantly affects all other bands. Comparison of the spectra shown in Figure 6 and Figure 7 allows one to note the shifts of the bands’ peaks in opposite directions, which may be due to the different effect of sodium chloride ions on ions and groups of atoms in the liquid phase of water.

For a qualitative interpretation of the directions of shifts of the band peaks, we calculated the intensity of the libration, deformation, and stretching bands of clusters of water and sodium chloride solution with the maximum concentration, each containing nine H₂O molecules. When performing calculations, the positions of atoms were optimized. The resulting spatial arrangement of atoms is shown in Figure 13 and used to calculate the components of the IR spectra.

Since the sets of optimal coordinates are not unique, the calculations were repeated many times with different initial coordinates until a dense packing of the cluster atoms simulating liquid water was achieved.

Figure 13 shows that in cluster (b) the Cl⁻ ion is surrounded by four H₂O molecules, and the Na⁺ ion by five H₂O molecules. The optimized distance between Cl⁻ and Na⁺ ions turned out to be 3.035 Å, while for the NaCl molecule it was determined to be 2.395 Å. An increase in the distance between the Cl⁻ and Na⁺ ions upon going from a molecule to a cluster allows one to conclude that the cluster (b) simulates solvate shells interacting with each other.

In calculating the profiles of the IR spectra, individual spectral components are presented as Gaussian curves with an FWHM = 130 cm⁻¹ estimated from the width of the stretching band. Theoretical spectra are calculated as sums of the Gaussian components; they are shown in Figure 14. In Figure 14, as well as Figure 7, the spectra are normalized to equal intensities of the deformation bands.

Figure 14 shows that the addition of the NaCl molecule to a water cluster leads to an increase in the peak intensity and a shift in the peak of the valence band toward higher wavenumbers, which agrees with the experiment (Figure 6).

The inclusion of NaCl in the water cluster also causes (Figure 14) a shift in the peak of the theoretical libration band (~500 cm⁻¹) towards lower wavenumbers, while the intensity and spectral position of the theoretical deformation band (~1600 cm⁻¹) practically do not change. This result qualitatively explains the observed libration band peak shift in experimental spectra at 1900–2400 cm⁻¹ with increasing sodium chloride concentration (Figure 7) since it is interpreted by the manifestation of overtones of the libration mode (3ν_L), or bending–libration mode (ν_L + ν₂) [35].

A rigorous theoretical study of the observed spectral effect should include an investigation of the influence of cluster isomerism on the profile and position of the IR spectrum bands.

4.7. Effect of Salinization and Adsorption Mechanisms on the Profile and Position of the Water Band at 2750–4000cm⁻¹

Figure 15 shows regions of the IR spectra of samples 2–4 (moistened samples), as well as the spectra of samples 9–11 (solutions used in moistening), previously shown in Figure 4 and Figure 6. In Figure 15, the data from Figure 4 and Figure 6 are normalized so that the intensities of the bands of water-moistened kaolinite and water coincide at point “d”, and the spectra of distilled water and NaCl solutions (samples 7–9) are shifted by 40 cm⁻¹. The line colors of the spectra in Figure 15 match those in Figure 4 and Figure 6, but the spectra of liquid-phase samples are represented by dotted lines. The spectra shown in Figure 15 make it possible to evaluate the change in the shape and position of the bands of solutions with different salt concentrations upon transition from the liquid phase to the state of sorption on kaolinite. Comparison of the above results allows one to conclude that when going from distilled water to the most saline one, the adsorption effect is inferior to the effect of solvation OB > OA (dissection along the line a-a, Figure 15). In addition, the effect of adsorption decreases with increasing salinity OA > BC; this can be explained by the retention of H₂O molecules from spreading over the basal kaolinite surfaces by the ions of solvate shells.

For a qualitative interpretation of the manifestation of solvation and adsorption mechanisms, the calculations of the optimal positions of atoms of water clusters and sodium chloride solution (Figure 13) near the hydroxyl surface of kaolinite were performed. The resulting spatial arrangement of atoms is shown in Figure 16a,b. Figure 16a illustrates the adsorption effect due to the formation of a chemical bond between O⁻ ions of water molecules and H⁺ ions of the hydroxyl surface of mineral particles. The placement of a NaCl solution cluster near this surface leads to the fact that the Cl⁻ ion replaces oxygen ions in these bonds (Figure 16b). The resulting interaction leads to the fact that the Cl⁻ ion leaves the solution cluster, and the water molecules of iAts solvate shell move to the Na⁺ ion screening the Cl⁻-Na⁺ interaction. As a result, the distance between these ions increases to 6.74 Å. In this case, the solvation effect prevails over the adsorption effect.

5. Conclusions

In this work, the profiles and spectral positions of the bands of stretching vibrations of the O–H group of objects listed in Table 1 were studied experimentally.

The main goal of the study was to experimentally determine the spectral features in the wavenumber range of 2750–4000 cm⁻¹ associated with moistening of kaolinite up to 26% with sodium chloride solutions at different concentrations, as well as with the modification of the properties of kaolinite surfaces due to the deposition of sodium chloride molecules from solutions. Some of the experimental results are qualitatively interpreted by the results of DFT calculations. Comparison of the profiles and positions of the bands at 2750–4000 cm⁻¹ made it possible to reach the following conclusions.

• An increase in the salt concentration in solutions used for moistening kaolinite leads to an increase in the peak intensity of the valence band of water and its shift to greater wavenumbers.
• The deposition of NaCl on the surface of kaolinite particles leads to an increase in the interaction of hydroxyl groups of kaolinite and NaCl molecules, which manifests itself in an increase in the peak intensities of stretching bands of hydroxyl groups. In this case, the band shift is not observed.
• The formation of solvate shells around sodium chloride ions during the salinization of water leads to a decrease in the width of the stretching vibration band, an increase in the peak intensity, and a shift in the band to greater wavenumbers.
• Use of the “triple split” McLean–Chandler basis set supplemented by three polarization functions of p-, two functions of d- symmetries, and a diffuse function on oxygen, as well as the exchange–correlation potential XLYP for the calculation of the IR spectrum of the liquid water makes it possible to qualitatively interpret the shift in the peak of the valence band toward higher wavenumbers and the shift in the peak of the libration band toward lower wavenumbers with the increasing salinity of the solution.
• At a moisture close to the plastic limit of kaolinite (26%), the transition from moistening with distilled water to moistening with maximally saline water leads to the fact that the adsorption effect is inferior to the solvation effect. As the salinity of the moisturizing solution increases, water molecules adsorbed on the basal surfaces of kaolinite tend to form solvate shells of sodium chloride ions.

The results obtained in the work show that the IR spectroscopy method provides an opportunity to study the features of the process of adsorption of hydrated NaCl salts on the basal surfaces of kaolinite at different moistures and ion concentrations in the liquid phase.