- Email: [email protected]
water mixtures studied by FT-NIR spectroscopy and DFT calculations
Journal of Molecular Structure 974 (2010) 60–67
Contents lists available at ScienceDirect
Journal of Molecular Structure journal homepage: www.elsevier.com/locate/molstruc
Molecular structure and hydrogen bonding in liquid cyclohexanol and cyclohexanol/water mixtures studied by FT-NIR spectroscopy and DFT calculations Mirosław Antoni Czarnecki *, Andrzej S. Muszyn´ski, Helena Troczyn´ska Faculty of Chemistry, University of Wrocław, F. Joliot-Curie 14, 50-383 Wrocław, Poland
a r t i c l e
i n f o
Article history: Received 28 October 2009 Received in revised form 4 December 2009 Accepted 4 December 2009 Available online 8 January 2010 Keywords: Hydrogen bonding Alcohols Near-infrared Two-dimensional correlation spectroscopy
a b s t r a c t The molecular structure and hydrogen bonding in liquid cyclohexanol and cyclohexanol/water mixtures has been examined by Fourier-transform near-infrared (FT-NIR) spectroscopy. FT-NIR spectra of pure cyclohexanol and binary mixtures with water at selected water mole fractions (X H2 O ) from 30 to 80 °C and the spectra of the mixtures from X H2 O = 0–0.4 at 30 °C were measured. Besides, FT-IR and FT-NIR spectra of cyclohexanol in CCl4 and cyclohexane solutions were recorded. The experimental spectra were analyzed by two-dimensional (2D) correlation approach and chemometrics methods. Interpretation of the spectra was guided by DFT calculations. It has been shown that small to moderate water content has a negligible effect on the structure of liquid cyclohexanol at constant temperature. Water molecules predominantly act as double donors to different species of cyclohexanol and this hydrogen bonding is stronger than that in bulk water. At lower water content appears a noticeable amount of singly bonded water molecules, however, population of this species in cyclohexanol is significantly smaller as compared with that in butyl alcohols. This results from much higher viscosity of cyclohexanol that stabilizes the cyclohexanol–water interactions. Increasing water content leads to creation of small clusters of water, where the water–water interaction is much weaker than that in bulk water. The temperature-induced breaking of smaller associates of cyclohexanol occurs easier in the presence of water, while an opposite effect was observed for the higher associates. The hydrophobic interactions in the cyclohexanol/water mixtures are of minor importance. Ó 2009 Elsevier B.V. All rights reserved.
1. Introduction Cyclohexanol is an important organic solvent used in laboratories and industry. The water content is one of the most important factors determining the properties of organic solvents. It is of particular note that the solubility of water in cyclohexanol is relatively high compared to 1-hexanol and other alcohols [1,2]. In the alcohol–water mixtures different kinds of interactions are possible: alcohol–alcohol, water–water and alcohol–water. As shown in previous vibrational and dielectric studies, in pure liquid n-alcohols the molecules predominantly form linear associates, whereas the branching increases the tendency of creation of the cyclic associates [3]. In the alcohol-rich region the structure of alcohols has been postulated to be the same as that in the pure liquid phase [5–9]. The spectroscopic, dielectric and thermodynamic studies on alcohol/water binary mixtures reveal that water molecules loses its hydrogen bond network and mixes into the solution as a single molecule [7–9]. The water molecules are singly bonded to * Corresponding author. Fax: +48 71 3282348. E-mail address: [email protected] (M.A. Czarnecki). 0022-2860/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.molstruc.2009.12.012
the hydrophilic site of the alcohol molecule at the end of the chain that act preferentially as H-bond acceptors. This implies that the amount of the free OwH groups (Ow – water oxygen) should be comparable to that of the hydrogen-bonded ones. Near-infrared (NIR) studies on butyl alcohol/water mixtures clearly show that most of the water molecules are doubly bonded to the alcohols, and the population of singly bonded water to the terminal OH groups of the alcohol associates is relatively small [10]. The solubility measurements suggest that water molecules can be incorporated in a variety of ways in one or two chains of self-associated alcohols simultaneously [11]. In the second case the water molecules interconnect two chains. At higher water content the formation of hydrogen bonds between water molecules already incorporated in the alcohol chains becomes more and more important [10,11]. On the other hand, the neutron diffraction measurements on small alcohols reveal that water molecules in the mixture (X H2 O > 0:3) exist as small hydrogen-bonded clusters with the three-dimensional hydrogen-bonded network structure of bulk water [12]. Thus, the local structure of water in alcohol-rich region of the mixture is close to that found in bulk water. The anomalous
M.A. Czarnecki et al. / Journal of Molecular Structure 974 (2010) 60–67
thermodynamic properties of alcohol–water mixtures result from incomplete mixing at molecular level that leads to coexistence of the alcohol-rich and water-rich clusters in the binary mixtures. Similar conclusion was obtained from low-frequency Raman spectra [13] and mass spectra of clusters of aqueous solutions of aliphatic alcohols [14]. Most of studies on alcohol–water mixtures have been performed in the water-rich region with the aim of determination the influence of alcohol presence on the structure of liquid water. In contrast, the studies in the alcohol-rich region are limited and were performed on small alcohols [4–9]. As yet, only a few spectroscopic studies were devoted to solutions of cyclohexanol in CCl4 [15,16]. Here we present the first FT-NIR study of cyclohexanol in the pure liquid state, in CCl4 and cyclohexane solutions as well as in the binary mixtures with water. The interpretation of the experimental results was supported by DFT calculations. 2. Experimental 2.1. Materials and spectroscopic measurements Cyclohexanol, cyclohexane and CCl4 of high purity (>99%) were purchased from Aldrich Chemical Co. (Germany). The samples were dried under freshly prepared molecular sieves (4A). High-purity water (resistivity 18.2 MX-cm) was obtained by the Simplicity 185 Ultrapure Water System (Millipore Corporation). FT-NIR and FT-IR spectra were recorded at resolutions of 4 and 1 cm1, respectively, on a Nicolet Magna 860 spectrometer with DTGS or MCT detector. The spectra were measured in a variable-temperature quartz cells (Hellma) of 2 and 5 mm thickness as a function of temperature [30:5:80] °C at constant X H2 O (0, 0.01, 0.05, 0.1, 0.2, 0.3, 0.4) and as a function of water content X H2 O = [0:0.05:0.4] at 30 °C. The spectra of cyclohexanol in CCl4 and cyclohexane were recorded at concentrations of 0.002, 0.01, 0.1 and 1 M at 30 °C.
61
3. Results and discussion 3.1. DFT calculations Both the experimental and theoretical studies reveal that the cyclohexanol molecules adopt a chair conformation in all phases [24–26]. Our calculations show that this conformation is 6– 7 kcal/mol more stable than the boat ones. Therefore, the further discussion is limited to the chair conformers only. Fig. 1 shows the lowest-energy conformers of isolated molecule of cyclohexanol obtained from DFT calculations. The energy difference between all four isomers (not shown) is less than 1.5 kcal/mol and agrees well with the value obtained by Jansen et al. [25]. It is of note that the length of the OH bond in all monomer conformations is very similar (0.966 ± 0.001 Å). This means that the conformation of the hydrocarbon chain has a small effect on properties of the free OH group (OHf). An acceptation of the proton by the OH group (Fig. 2A) also does not lead to noticeable variation in the length of the O–H bond in the open dimer. In contrast, the proton donation lengthen the OH bond to 0.9755 ± 0.0005 Å and the magnitude of this effect is similar for the three most stable conformers of cyclohexanol. The fourth conformer (not shown) has an OH group directed towards the chair and hence, the hydrogen bonding creation in this conformation is less favorable. Our recent results have shown that cyclohexanol may form very stable cyclic tetramers (Fig. 2B), where an average length of the O–H bond is 0.990 ± 0.002 Å. This kind of species should be abundant in the liquid cyclohexanol at lower temperatures. In 1:1 complex with water the OH groups of cyclohexanol can act as an acceptor (OHa) (Fig. 3A) or as a donor (OHd) (Fig. 3B). Both complexes should be distinguishable by NIR spectroscopy since they differ with the number of the free and associated OH groups in water and cyclohexanol. Our calculations show that the length of the intermolecular O H bonding is shorter (0.01–0.02 Å) when
2.2. DFT calculations Density functional (DFT) calculations were carried out at the B3LYP/6-31G(d,p) level of theory using Gaussian 03W [17]. The total energy of each conformer was corrected with the zero point vibrational energy, and the harmonic wavenumbers were calculated. 2.3. 2D correlation analysis Prior to 2D correlation analysis the spectra were corrected for the density change with temperature and then the baseline fluctuations were minimized by an offset at 9000 cm1. The spectra of pure samples at 30 °C and average spectra were used as references for the concentration- and temperature-dependent data, respectively. The generalized 2D correlation spectra were calculated by approach given by Noda [18,19], using MATLAB 7.0.4 (The Math Works Inc.) based software written in our laboratory. 2.4. Chemometric analysis The number of significant species was estimated by principal component analysis (PCA) [20] and then confirmed by evolving factor analysis (EFA) [21]. The results of EFA provided an initial approximation of the concentration profiles. The real concentration profiles and the pure component spectra were resolved by multivariate curve resolution-alternating least squares (MCR-ALS) approach with constraints (non-negativity on concentrations and spectra) [22,23]. The analysis was performed by PLS-Toolbox 4.0.2 (Eigenvector Research Inc.) for use with MATLAB.
Fig. 1. The lowest-energy conformers of isolated molecule of cyclohexanol obtained from DFT calculations. (A) DE = 0 and (B) DE 0.1 kcal/mol.
62
M.A. Czarnecki et al. / Journal of Molecular Structure 974 (2010) 60–67
Fig. 3. The lowest-energy conformers of 1:1 cyclohexanol/water complexes obtained from DFT calculations. In complex (A) water is a donor, while in complex (B) water is an acceptor.
Fig. 2. The lowest-energy conformers of the open dimer (A) and the cyclic tetramer (B) of cyclohexanol obtained from DFT calculations.
water act as a donor. This means that the OH groups of cyclohexanol are stronger acceptors than the OH groups of water (OwH). Hence, one can expect that addition of water to liquid cyclohexanol should not break the structure of the liquid phase. The most favorable 2:1 cyclohexanol/water complex has a cyclic structure (Fig. 4A), where molecule of water is both donor and acceptor. In this complex the hydrogen bonding between the molecules of cyclohexanol is preserved. The linear complex (Fig. 4B) is less favorable (3 kcal/mol) than the cyclic one. In this complex does not occur the hydrogen bonding between molecules of cyclohexanol.
contrast, the corresponding peak in the overtone region appears only at the highest concentration (1.0 M). This is a good example of the effect reported for the first time by Luck and Ditter [27]. Vibrations due to stronger hydrogen bonds have strong fundamentals and weak overtones. In both solvents the band due to the OHf group has two components that were resolved in the second derivative spectra. The position of these bands (Table 1) is typical as that for the secondary alcohols and they were assigned to the rotational isomerism of the OHf group [28]. The anharmonicity constants (v) is similar for both rotational isomers of the OHf group and is close to that for aliphatic alcohols [3,29]. An increase in the hydrogen bonding strength leads to significant grows in v (Table 1). Interestingly, though cyclohexane is believed to be more inert solvent as compared to CCl4, the positions of the corresponding peaks in CCl4 are only slightly red-shifted (3–4 cm1). The magnitude of this shift excludes the presence of the specific interactions between cyclohexanol and CCl4.
3.3. Effect of temperature on FT-NIR spectra of pure liquid cyclohexanol and cyclohexanol/water mixtures
3.2. FT-IR and FT-NIR spectra of cyclohexanol in CCl4 and cyclohexane The IR and NIR spectra of cyclohexanol in CCl4 (Fig. 5) were divided by the corresponding molar concentration and the cell thickness. Hence, they represent the molar absorption coefficients (e). The relative intensities and positions of the peaks in the IR spectra are close to those reported by Sandorfy [16]. As expected, increasing concentration of cyclohexanol reduces e of the OHf peak in both spectral regions as a result of the hydrogen bonding creation between the molecules of cyclohexanol. The relative intensities of the bands due to the hydrogen-bonded OH groups are quite different in both spectral regions (Fig. 5). The associated species are clearly seen in the IR spectra even at concentration of 0.1 M. In
Fig. 6A and B show the effect of temperature on FT-NIR spectra of pure cyclohexanol and cyclohexanol/water mixture, respectively. The positions of the peaks in pure cyclohexanol and in the mixture are collected in Table 2. The most significant changes between the spectra are observed in the 5000–5300 cm1 region, where is located the m2 + m3 band of water. The other band of water (m1 + m3) is heavily overlapped by absorption of the associated OH groups of cyclohexanol. Interestingly, the bands due to vibrations of the hydrocarbon chain (near 5700 and 5800 cm1) are not influenced by the presence of water or temperature suggesting a minor importance of the hydrophobic interaction in the studied system. Addition of water causes a small blue-shift of the bands due to
63
M.A. Czarnecki et al. / Journal of Molecular Structure 974 (2010) 60–67
Molar absorption coefficient
10
A
8 6 4 2 0 3100 3200 3300 3400 3500 3600 3700 3800 Wavenumbers/cm-1
Molar absorption coefficient
0.04
B 0.03
0.02
0.01
6800
6900
7000
7100
7200
7300
-1
Fig. 4. The lowest-energy conformers of 2:1 cyclohexanol/water complexes obtained from DFT calculations with the cyclic (A) and linear (B) structure.
the OHf and OHad vibrations of cyclohexanol. The large shift for the OHd band (170 cm1) results from overlapping by the m1 + m3 band of water. Like for others previously studied alcohols [10], the m2 + m3 band of associated water in the mixture is red-shifted (28 cm1) as compared with that in the spectrum of bulk water. This suggests that water molecules in the mixture are involved in stronger hydrogen bonding that those in bulk water. The synchronous (not shown) and asynchronous (Fig. 7A) contour plots of pure cyclohexanol are similar to those previously obtained for aliphatic alcohols [3,30,31]. In the asynchronous spectrum are resolved the peaks due to the rotational conformers of the OHf group at (7047, 7074) cm1. The positions of these peaks are the same as those observed in CCl4 solution (Tables 1 and 2). Also position of the peak due to the cooperative hydrogen bonding (OHad) in pure cyclohexanol and in CCl4 solution is similar. In contrast, the peak attributed to the OHd group is significantly redshifted (140 cm1) on going from pure cyclohexanol to CCl4. This means that both peaks originate from different species. The peak at 6819 cm1 is due to the open chain dimers (Fig. 2A) in pure cyclohexanol, whereas the 6679 cm1 peak can be attributed to small oligomers of cyclohexanol in CCl4. In liquid cyclohexanol the later peak is overlapped by strong absorption of OHad group. It is of note that the asynchronous spectrum of the mixture (Fig. 7B) does not develop the peaks due to the rotational isomerism of the OHf group. This effect, observed for the first time in nbutyl alcohol/water mixture [32], was explained as the lack of the conformational selectivity by water molecules upon the hydrogen bonding formation with the molecules of alcohol.
Wavenumbers/cm
Fig. 5. FT-IR (A) and FT-NIR (B) spectra of cyclohexanol in CCl4 at 30 °C. Solid line – 0.002 M, dashed line – 0.01, dotted line – 0.1 M and dash-dot line – 1 M.
Table 1 Assignments of IR (m01) and NIR (m02) bands of cyclohexanol in CCl4 and cyclohexane solutions at 30 °C together with the anharmonicity constants [v = m02/2 m01]. Position (cm1) IR (CCl4)
NIR (CCl4)
NIR (cyclohexane)
2857 2933 3622 3610 3482 3336
5707 5802 7074 7047 6679 6280
– – 7080 7050 – 6283
v
Assignment
85 86 142.5 196
CH CH OHf, gauche conformer OHf, trans conformer OHd OHad
Both PCA and EFA analysis of temperature-dependent spectra of pure cyclohexanol suggest the presence of two independent components that were resolved by MCR-ALS method (Fig. 8A). However, reasonable results one can obtain by assuming the presence of the third component. The spectral and concentration profiles of this component (Figs. 8B and 8C) have maxima near 6770 cm1 and 55 °C, respectively. The concentration profiles of two other components monotonously increase and decrease with the temperature. The third component can be assigned to the OHd group and corresponds to the asynchronous peak at 6819 cm1. The difference in the position results from the fact that 2D correlation peaks are shifted in the direction of higher intensity changes [33–35].
64
M.A. Czarnecki et al. / Journal of Molecular Structure 974 (2010) 60–67
0.6
A Wavenumber/cm-1, ν2
0.5 Absorbance
A
0.4 0.3 0.2 0.1
7000
6500
6000
4500
5000
5500
6000
6500
7000
6000
7500
7000
Wavenumbers/cm-1, ν1
Wavenumbers/cm -1 1.5
B
B Wavenumber/cm-1, ν2
1.2 Absorbance
6500
0.9 0.6
7000
6500
0.3
6000 4500
5000
5500
6000
6500
7000
7500
Wavenumbers/cm -1 Fig. 6. FT-NIR spectra of pure cyclohexanol (A) and cyclohexanol/water mixture with X H2 O ¼ 0:4 (B) from 30 to 80 °C. The arrows indicate directions of intensity changes on the temperature increase.
Table 2 Frequencies and assignments of selected NIR bands of cyclohexanol (1) and water (2) in the pure liquid phases and in the cyclohexanol/water mixture (X H2 O ¼ 0:4). Position (cm1)
a b c d
Pure (1)
Mixture (1 + 2)
8325 7074a 7047a 6819b 6243 5802 5705 – – 5184d
8325 7075a 7050a 6990b,c 6250 5802 5705 5295 5200b 5156
Vibration
m03CH m02OH m02OH m02OH m02OH m02CH m02CH m2 + m3 m2 + m3 m2 + m3
Assignment
OHf, gauche conformer OHf, trans conformer OHd (non-cooperative) OHad (cooperative)
Water, OwHf Water, OwHd (non-cooperative) Water, OwHad (cooperative)
Peaks resolved in second derivative spectra. Peaks resolved in 2D correlation spectra. Overlapped by the m1 + m3 band of water. In bulk water.
The 2D correlation contour plots constructed from the temperature-dependent spectra of the mixtures indicate the presence of at least two different water species. Yet, the band shift observed in the temperature-dependent spectra may generate artifacts in 2D correlation contour plots [33–35]. The presence of three distinct water species was confirmed by MCR-ALS analysis of the temperature-dependent spectra of the mixture with X H2 O ¼ 0:4 (Fig. 9). The
6000
6500
7000
Wavenumber/cm-1, ν1 Fig. 7. Asynchronous spectra of pure cyclohexanol (A) and cyclohexanol/water mixture with X H2 O ¼ 0:4 (B) from 30 to 80 °C. The positive contours are shaded.
positions of the corresponding peaks obtained by 2D correlation analysis and MCR-ALS method are similar making these results more reliable. The concentration profiles of the mixture (not shown) are similar to those obtained for the pure cyclohexanol (Fig. 8C). Fig. 10 shows the power spectra constructed from the temperature-dependent spectra of pure cyclohexanol and cyclohexanol/ water mixtures. The power spectrum is a diagonal of the synchronous spectrum and it represents the overall extent of intensity changes with respect to the reference spectrum. As can be seen, the extent of spectral changes of the OHf group increases on going from the pure cyclohexanol to the mixture. Since the spectral changes at all wavenumbers are continuous one can conclude that the temperature-induced increase in the population of the OHf groups of cyclohexanol occurs faster in the presence of water. In contrast, the number of the most associated species in pure cyclohexanol decreases faster than that in the mixture. This means that the population of the OHf groups increases mainly at the expense of the intermediate species. In other words, addition of water stabilizes the higher associates of cyclohexanol, whereas the shorter ones are disrupted more easily upon the temperature rise. 3.4. Effect of water content on FT-NIR spectra of cyclohexanol/water mixtures Fig. 11A shows the difference spectra calculated by subtraction of the spectrum of pure cyclohexanol at 30 °C from the concentra-
65
M.A. Czarnecki et al. / Journal of Molecular Structure 974 (2010) 60–67
0.16
A
1.2
Absorbance
Absorbance
0.12
0.08
0.8
0.4
0.04
6000
6500
7000
5000
7500
5500
6000
6500
7000
7500
Wavenumbers/cm -1
Wavenumbers/cm -1
Fig. 9. Spectral profiles obtained from MCR-ALS of the spectra of cyclohexanol/ water mixture with X H2 O ¼ 0:4 from 30 to 80 °C.
0.16
B 1.8
x 10
-3
1.5
0.08
Arbitrary units
Absorbance
0.12
0.04
6000
6500
7000
7500
1.2 0.9 0.6 0.3
wavenumbers/cm -1 4.5
6000
C
A.U.
6400
6600
6800
7000
7200
Wavenumbers/cm -1 Fig. 10. Power spectra of pure cyclohexanol (solid line) and cyclohexanol/water mixtures with X H2 O ¼ 0:2 (dotted line) and X H2 O ¼ 0:4 (dashed line) from 30 to 80 °C.
3.5
2.5
1.5 30
6200
40
50
60
70
80
Temperature/ o C Fig. 8. Spectral (A and B) and concentration (C) profiles obtained from MCR-ALS of the spectra of pure cyclohexanol from 30 to 80 °C.
tion-dependent spectra of the mixture recorded at the same temperature. As can be seen, an increase in water content does not affect intensities of the cyclohexanol bands, but changes the relative intensities of the free and associated OH groups of water. The band at 5295 cm1 clearly appears at small water content (Fig. 11A), whereas at X H2 O > 0:3 it is resolved only in the second derivative spectra (not shown). The intensity of the hydrogen-bonded band
of water increases linearly with X H2 O , whereas the changes of the band due to the OwHf group have a non-linear character (Fig. 11B). At higher water content this increase becomes slower, evidencing the growing importance of water–water interactions. Fig. 11C shows the power spectrum obtained from the concentration-dependent data. In the spectrum occur mainly the peaks due to water: m2 + m3 at 5160 cm1 and m1 + m3 at 6863 cm1. The peak resulting from absorption of the free OwH groups (5295 cm1) is less intense as compared with the analogous peak in the power spectra of butyl alcohol/water mixtures [10]. It has been observed a clear relationship between the relative intensity of this peak and the degree of association of the alcohol. The higher extent of self-association of alcohol – the smaller intensity of this peak [10]. In the case of more associated compounds like propanediols, this peak completely disappears from the power spectra [36]. The position of the peaks due to the rotational isomerism of the OHf group (Table 2) indicates that the strength of the hydrogen bonding in cyclohexanol is typical as that of other secondary alcohols [28]. Nevertheless, the high viscosity of cyclohexanol (>10 times higher than that of butyl alcohols) enhances the stability of the intermolecular hydrogen bonds.
66
M.A. Czarnecki et al. / Journal of Molecular Structure 974 (2010) 60–67
A Wavenumbers/cm-1, ν2
Absorbance
A
5400
1.2
0.9
0.6
0.3
5000
5500
6000
6500
7000
5200
5000
4800
7500
5200
5400
Wavenumbers/cm -1, ν1
Wavenumbers/cm -1
x 10
5000
-3
1.2
B
B
3 Absorbance
Arbitrary units
0.9
2
0.6
0.3
1
5000
0
0.1
0.2
0.3
0.4
XH O
C
Arbitrary units
7
5
(x10)
3
1 5000
5500
6000
6500
Wavenumbers/cm
7000
6000
6500
7000
7500
Wavenumbers/cm -1 Fig. 12. Asynchronous spectrum (A) and the spectral profiles obtained from MCRALS (B) of the spectra of the cyclohexanol/water mixtures with X H2 O = [0:0.05:0.4] at 30 °C. The positive contours in Fig. 12A are shaded. The dotted lines in Fig. 12B represent the spectra of pure cyclohexanol and pure water at 30 °C.
2
9
5500
7500
-1
Fig. 11. The difference spectra obtained by subtraction of the spectrum of pure cyclohexanol from the spectra of the cyclohexanol/water mixtures with X H2 O = [0:0.05:0.4] at 30 °C (A) and the power spectrum obtained from these data (C). In (B) are displayed the intensities of the bands at 5150 cm1 (dashed line) and 5295 cm1 (solid line) obtained from the second derivative of the difference spectra shown in Fig. 11A.
The peak positions in the spectra of the mixtures do not depend on water content at constant temperature. Hence, one can assign a real physical meaning for the peaks developed in the asynchronous
spectrum (Fig. 12A). As can be seen, the m2 + m3 peak is resolved into two components that are defined by four frequencies: 5133, 5204, 5225 and 5295 cm1. Taking into account the fact that the peaks located close to the diagonal are more shifted [33–35], one can assume that the peaks at 5204 and 5225 cm1 originate from the same species. Accordingly, in the asynchronous spectrum are resolved three non-equivalent OwH groups. The low frequency band (5133 cm1) can be assigned to the cooperative HB, where both OwH groups are donors to oxygens of cyclohexanol, while the high frequency component (5295 cm1) is due to absorption of the free OwH groups in one-bonded water. The presence of the third peak at intermediate position (5204 cm1) results from weakly hydrogen bonded water to cyclohexanol, e.g., from OwH donors in onebonded water. The asynchronous spectrum reveals that the changes at 5204 cm1 occur at different rate than those for two remaining bands. It is very likely that the 5204 cm1 peak includes a contribution from the water–water interactions taking place at higher water content. Due to small size of these clusters the water–water interactions are weaker than those in bulk water. Fig. 12B displays the spectral profiles obtained from MCR-ALS of the concentration-dependent data. The profile of cyclohexanol is identical with the spectrum of pure cyclohexanol. On the other hand, the peaks in the spectral profile of water are red-shifted with respect to those in the spectrum of bulk water. This means that the molecules of cyclohexanol in the mixture are in the same environ-
M.A. Czarnecki et al. / Journal of Molecular Structure 974 (2010) 60–67
ment as those in pure cyclohexanol, whereas the water molecules are involved in stronger hydrogen bonding than those in bulk water. 4. Conclusions Addition of small to moderate amount of water has a minor effect on the structure of liquid cyclohexanol at constant temperature. Molecules of cyclohexanol in the mixture are in the same environment as those in the pure cyclohexanol. DFT calculations suggest a significant population of the cyclic tetramers in liquid cyclohexanol. Most of water molecules in the mixture act as a double donors to oxygens of cyclohexanol. This cooperative hydrogen bonding appears to be stronger than that in bulk water. At lower water content exists in the mixture a certain amount of singly bonded water molecules. Yet, the number of the free OwH groups in the cyclohexanol/water mixture is much smaller as compared with that in butyl alcohol/water mixtures. It results from high viscosity of cyclohexanol that stabilize the cyclohexanol–water interaction. Strongly non-linear relationship between the population of the free OwH and X H2 O , particularly at higher water content, suggests growing importance of the water–water interactions. These interactions are responsible for creation of small clusters of water, where the hydrogen bonds are significantly weaker than those in bulk water. The temperature-induced breaking of smaller associates of cyclohexanol occurs easier in the presence of water. In contrast, addition of water leads to stabilization of the higher associates. The hydrophobic interactions in the pure liquid cyclohexanol and cyclohexanol/water mixtures do not play an important role.
[8] [9] [10] [11] [12] [13] [14] [15] [16] [17]
[18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30]
References [31] [1] A.F.M. Barton, Alcohols with Water, Solubility Data Series, vol. 15, Pergamon Press, Oxford, 1984. [2] N. Šegatin, C. Klofutar, Monatsh. Chem. 135 (2004) 241–248. [3] M.A. Czarnecki, K. Orzechowski, J. Phys. Chem. A 107 (2003) 1119–1126. [4] S. Dixit, W.K.C. Poon, J. Crain, J. Phys. Condens. Matter 12 (2000) L323–L328. [5] T. Sato, A. Chiba, R. Nozaki, J. Chem. Phys. 112 (2000) 2924–2932. [6] T. Sato, A. Chiba, R. Nozaki, J. Mol. Liq. 101 (2002) 99–111. [7] M. D’Angelo, G. Onori, A. Santucci, J. Chem. Phys. 100 (1994) 3107–3113.
[32] [33] [34] [35] [36]
67
G. Onori, A. Santucci, J. Mol. Liq. 69 (1996) 161–181. Y. Koga, Can. J. Chem. 66 (1988) 3171–3175. M.A. Czarnecki, D. Wojtków, J. Mol. Struct. 883–884 (2008) 203–208. P.L. Huyskens, M.Cl. Haulait-Pirson, I. Hanssens, J. Mullens, J. Phys. Chem. 84 (1980) 28–33. S. Dixit, J. Crain, W.C.K. Poon, J.L. Finney, A.K. Soper, Nature 416 (2002) 829– 832. K. Yoshida, T. Yamaguchi, Z. Naturforsch. 56a (2001) 529–536. A. Wakisaka, K. Matsuura, J. Mol. Liq. 129 (2006) 25–32. U. John, K.P.R. Nair, Spectrochim. Acta A 61 (2005) 2555–2559. C. Sandorfy, Bull. Pol. Acad. Sci. Chem. 43 (1995) 7–24. M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman, J.A. Montgomery Jr., T. Vreven, K.N. Kudin, J.C. Burant, J.M. Millam, S.S. Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J.E. Knox, H.P. Hratchian, J.B. Cross, C. Adamo, J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev, A.J. Austin, R. Cammi, C. Pomelli, J.W. Ochterski, P.Y. Ayala, K. Morokuma, G.A. Voth, P. Salvador, J.J. Dannenberg, V.G. Zakrzewski, S. Dapprich, A.D. Daniels, M.C. Strain, O. Farkas, D.K. Malick, A.D. Rabuck, K. Raghavachari, J.B. Foresman, J.V. Ortiz, Q. Cui, A.G. Baboul, S. Clifford, J. Cioslowski, B.B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R.L. Martin, D.J. Fox, T. Keith, M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong, C. Gonzalez, J.A. Pople, Gaussian 03, Revision C.02, Gaussian Inc., Wallingford, CT, 2004. I. Noda, Appl. Spectrosc. 47 (1993) 1329–1336. I. Noda, Appl. Spectrosc. 54 (2000) 994–999. S. Wold, K. Esbensen, P. Geladi, Chemom. Intell. Lab. Syst. 2 (1987) 37–52. H.R. Keller, D.L. Massart, Chemom. Intell. Lab. Syst. 12 (1992) 209–224. R. Tauler, B. Kowalski, S. Fleming, Anal. Chem. 65 (1993) 2040–2047. R. Tauler, A. Izquiedoro-Ridorsa, E. Casassas, Chemom. Intell. Lab. Syst. 18 (1993) 293–300. R.M. Ibberson, S. Parsons, D.R. Allen, A.N.T. Bell, Acta Crystallogr. B 64 (2008) 573–582. C. Jansen, an Mey Dieter, G. Raabe, J. Fleischauser, J. Mol. Struct. (Theochem) 398-399 (1997) 395-404. R.A. Pickering, C.C. Price, J. Am. Chem. Soc. 80 (1958) 4931–4933. W.A.P. Luck, W. Ditter, J. Mol. Struct. 1 (1967–1968) 261–282. M.A. Czarnecki, D. Wojtków, K. Haufa, Chem. Phys. Lett. 431 (2006) 294–299. D. Wojtków, M.A. Czarnecki, Appl. Spectrosc. 61 (2007) 928–934. M.A. Czarnecki, H. Maeda, Y. Ozaki, M. Suzuki, M. Iwahashi, Appl. Spectrosc. 52 (1998) 994–1000. M.A. Czarnecki, M. Maeda, Y. Ozaki, M. Suzuki, M. Iwahashi, J. Phys. Chem. A 102 (1998) 9117–9123. M.A. Czarnecki, D. Wojtków, J. Phys. Chem. A 108 (2004) 2411–2417. A. Gericke, S.J. Gadaleta, J.W. Brauner, R. Mendelsohn, Biospectroscopy 2 (1996) 341–351. M.A. Czarnecki, Appl. Spectrosc. 52 (1998) 1583–1590. M.A. Czarnecki, Appl. Spectrosc. 54 (2000) 986–993. K.Z. Haufa, M.A. Czarnecki, Vib. Spectrosc. 51 (2009) 80–85.
Contents lists available at ScienceDirect
Journal of Molecular Structure journal homepage: www.elsevier.com/locate/molstruc
Molecular structure and hydrogen bonding in liquid cyclohexanol and cyclohexanol/water mixtures studied by FT-NIR spectroscopy and DFT calculations Mirosław Antoni Czarnecki *, Andrzej S. Muszyn´ski, Helena Troczyn´ska Faculty of Chemistry, University of Wrocław, F. Joliot-Curie 14, 50-383 Wrocław, Poland
a r t i c l e
i n f o
Article history: Received 28 October 2009 Received in revised form 4 December 2009 Accepted 4 December 2009 Available online 8 January 2010 Keywords: Hydrogen bonding Alcohols Near-infrared Two-dimensional correlation spectroscopy
a b s t r a c t The molecular structure and hydrogen bonding in liquid cyclohexanol and cyclohexanol/water mixtures has been examined by Fourier-transform near-infrared (FT-NIR) spectroscopy. FT-NIR spectra of pure cyclohexanol and binary mixtures with water at selected water mole fractions (X H2 O ) from 30 to 80 °C and the spectra of the mixtures from X H2 O = 0–0.4 at 30 °C were measured. Besides, FT-IR and FT-NIR spectra of cyclohexanol in CCl4 and cyclohexane solutions were recorded. The experimental spectra were analyzed by two-dimensional (2D) correlation approach and chemometrics methods. Interpretation of the spectra was guided by DFT calculations. It has been shown that small to moderate water content has a negligible effect on the structure of liquid cyclohexanol at constant temperature. Water molecules predominantly act as double donors to different species of cyclohexanol and this hydrogen bonding is stronger than that in bulk water. At lower water content appears a noticeable amount of singly bonded water molecules, however, population of this species in cyclohexanol is significantly smaller as compared with that in butyl alcohols. This results from much higher viscosity of cyclohexanol that stabilizes the cyclohexanol–water interactions. Increasing water content leads to creation of small clusters of water, where the water–water interaction is much weaker than that in bulk water. The temperature-induced breaking of smaller associates of cyclohexanol occurs easier in the presence of water, while an opposite effect was observed for the higher associates. The hydrophobic interactions in the cyclohexanol/water mixtures are of minor importance. Ó 2009 Elsevier B.V. All rights reserved.
1. Introduction Cyclohexanol is an important organic solvent used in laboratories and industry. The water content is one of the most important factors determining the properties of organic solvents. It is of particular note that the solubility of water in cyclohexanol is relatively high compared to 1-hexanol and other alcohols [1,2]. In the alcohol–water mixtures different kinds of interactions are possible: alcohol–alcohol, water–water and alcohol–water. As shown in previous vibrational and dielectric studies, in pure liquid n-alcohols the molecules predominantly form linear associates, whereas the branching increases the tendency of creation of the cyclic associates [3]. In the alcohol-rich region the structure of alcohols has been postulated to be the same as that in the pure liquid phase [5–9]. The spectroscopic, dielectric and thermodynamic studies on alcohol/water binary mixtures reveal that water molecules loses its hydrogen bond network and mixes into the solution as a single molecule [7–9]. The water molecules are singly bonded to * Corresponding author. Fax: +48 71 3282348. E-mail address: [email protected] (M.A. Czarnecki). 0022-2860/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.molstruc.2009.12.012
the hydrophilic site of the alcohol molecule at the end of the chain that act preferentially as H-bond acceptors. This implies that the amount of the free OwH groups (Ow – water oxygen) should be comparable to that of the hydrogen-bonded ones. Near-infrared (NIR) studies on butyl alcohol/water mixtures clearly show that most of the water molecules are doubly bonded to the alcohols, and the population of singly bonded water to the terminal OH groups of the alcohol associates is relatively small [10]. The solubility measurements suggest that water molecules can be incorporated in a variety of ways in one or two chains of self-associated alcohols simultaneously [11]. In the second case the water molecules interconnect two chains. At higher water content the formation of hydrogen bonds between water molecules already incorporated in the alcohol chains becomes more and more important [10,11]. On the other hand, the neutron diffraction measurements on small alcohols reveal that water molecules in the mixture (X H2 O > 0:3) exist as small hydrogen-bonded clusters with the three-dimensional hydrogen-bonded network structure of bulk water [12]. Thus, the local structure of water in alcohol-rich region of the mixture is close to that found in bulk water. The anomalous
M.A. Czarnecki et al. / Journal of Molecular Structure 974 (2010) 60–67
thermodynamic properties of alcohol–water mixtures result from incomplete mixing at molecular level that leads to coexistence of the alcohol-rich and water-rich clusters in the binary mixtures. Similar conclusion was obtained from low-frequency Raman spectra [13] and mass spectra of clusters of aqueous solutions of aliphatic alcohols [14]. Most of studies on alcohol–water mixtures have been performed in the water-rich region with the aim of determination the influence of alcohol presence on the structure of liquid water. In contrast, the studies in the alcohol-rich region are limited and were performed on small alcohols [4–9]. As yet, only a few spectroscopic studies were devoted to solutions of cyclohexanol in CCl4 [15,16]. Here we present the first FT-NIR study of cyclohexanol in the pure liquid state, in CCl4 and cyclohexane solutions as well as in the binary mixtures with water. The interpretation of the experimental results was supported by DFT calculations. 2. Experimental 2.1. Materials and spectroscopic measurements Cyclohexanol, cyclohexane and CCl4 of high purity (>99%) were purchased from Aldrich Chemical Co. (Germany). The samples were dried under freshly prepared molecular sieves (4A). High-purity water (resistivity 18.2 MX-cm) was obtained by the Simplicity 185 Ultrapure Water System (Millipore Corporation). FT-NIR and FT-IR spectra were recorded at resolutions of 4 and 1 cm1, respectively, on a Nicolet Magna 860 spectrometer with DTGS or MCT detector. The spectra were measured in a variable-temperature quartz cells (Hellma) of 2 and 5 mm thickness as a function of temperature [30:5:80] °C at constant X H2 O (0, 0.01, 0.05, 0.1, 0.2, 0.3, 0.4) and as a function of water content X H2 O = [0:0.05:0.4] at 30 °C. The spectra of cyclohexanol in CCl4 and cyclohexane were recorded at concentrations of 0.002, 0.01, 0.1 and 1 M at 30 °C.
61
3. Results and discussion 3.1. DFT calculations Both the experimental and theoretical studies reveal that the cyclohexanol molecules adopt a chair conformation in all phases [24–26]. Our calculations show that this conformation is 6– 7 kcal/mol more stable than the boat ones. Therefore, the further discussion is limited to the chair conformers only. Fig. 1 shows the lowest-energy conformers of isolated molecule of cyclohexanol obtained from DFT calculations. The energy difference between all four isomers (not shown) is less than 1.5 kcal/mol and agrees well with the value obtained by Jansen et al. [25]. It is of note that the length of the OH bond in all monomer conformations is very similar (0.966 ± 0.001 Å). This means that the conformation of the hydrocarbon chain has a small effect on properties of the free OH group (OHf). An acceptation of the proton by the OH group (Fig. 2A) also does not lead to noticeable variation in the length of the O–H bond in the open dimer. In contrast, the proton donation lengthen the OH bond to 0.9755 ± 0.0005 Å and the magnitude of this effect is similar for the three most stable conformers of cyclohexanol. The fourth conformer (not shown) has an OH group directed towards the chair and hence, the hydrogen bonding creation in this conformation is less favorable. Our recent results have shown that cyclohexanol may form very stable cyclic tetramers (Fig. 2B), where an average length of the O–H bond is 0.990 ± 0.002 Å. This kind of species should be abundant in the liquid cyclohexanol at lower temperatures. In 1:1 complex with water the OH groups of cyclohexanol can act as an acceptor (OHa) (Fig. 3A) or as a donor (OHd) (Fig. 3B). Both complexes should be distinguishable by NIR spectroscopy since they differ with the number of the free and associated OH groups in water and cyclohexanol. Our calculations show that the length of the intermolecular O H bonding is shorter (0.01–0.02 Å) when
2.2. DFT calculations Density functional (DFT) calculations were carried out at the B3LYP/6-31G(d,p) level of theory using Gaussian 03W [17]. The total energy of each conformer was corrected with the zero point vibrational energy, and the harmonic wavenumbers were calculated. 2.3. 2D correlation analysis Prior to 2D correlation analysis the spectra were corrected for the density change with temperature and then the baseline fluctuations were minimized by an offset at 9000 cm1. The spectra of pure samples at 30 °C and average spectra were used as references for the concentration- and temperature-dependent data, respectively. The generalized 2D correlation spectra were calculated by approach given by Noda [18,19], using MATLAB 7.0.4 (The Math Works Inc.) based software written in our laboratory. 2.4. Chemometric analysis The number of significant species was estimated by principal component analysis (PCA) [20] and then confirmed by evolving factor analysis (EFA) [21]. The results of EFA provided an initial approximation of the concentration profiles. The real concentration profiles and the pure component spectra were resolved by multivariate curve resolution-alternating least squares (MCR-ALS) approach with constraints (non-negativity on concentrations and spectra) [22,23]. The analysis was performed by PLS-Toolbox 4.0.2 (Eigenvector Research Inc.) for use with MATLAB.
Fig. 1. The lowest-energy conformers of isolated molecule of cyclohexanol obtained from DFT calculations. (A) DE = 0 and (B) DE 0.1 kcal/mol.
62
M.A. Czarnecki et al. / Journal of Molecular Structure 974 (2010) 60–67
Fig. 3. The lowest-energy conformers of 1:1 cyclohexanol/water complexes obtained from DFT calculations. In complex (A) water is a donor, while in complex (B) water is an acceptor.
Fig. 2. The lowest-energy conformers of the open dimer (A) and the cyclic tetramer (B) of cyclohexanol obtained from DFT calculations.
water act as a donor. This means that the OH groups of cyclohexanol are stronger acceptors than the OH groups of water (OwH). Hence, one can expect that addition of water to liquid cyclohexanol should not break the structure of the liquid phase. The most favorable 2:1 cyclohexanol/water complex has a cyclic structure (Fig. 4A), where molecule of water is both donor and acceptor. In this complex the hydrogen bonding between the molecules of cyclohexanol is preserved. The linear complex (Fig. 4B) is less favorable (3 kcal/mol) than the cyclic one. In this complex does not occur the hydrogen bonding between molecules of cyclohexanol.
contrast, the corresponding peak in the overtone region appears only at the highest concentration (1.0 M). This is a good example of the effect reported for the first time by Luck and Ditter [27]. Vibrations due to stronger hydrogen bonds have strong fundamentals and weak overtones. In both solvents the band due to the OHf group has two components that were resolved in the second derivative spectra. The position of these bands (Table 1) is typical as that for the secondary alcohols and they were assigned to the rotational isomerism of the OHf group [28]. The anharmonicity constants (v) is similar for both rotational isomers of the OHf group and is close to that for aliphatic alcohols [3,29]. An increase in the hydrogen bonding strength leads to significant grows in v (Table 1). Interestingly, though cyclohexane is believed to be more inert solvent as compared to CCl4, the positions of the corresponding peaks in CCl4 are only slightly red-shifted (3–4 cm1). The magnitude of this shift excludes the presence of the specific interactions between cyclohexanol and CCl4.
3.3. Effect of temperature on FT-NIR spectra of pure liquid cyclohexanol and cyclohexanol/water mixtures
3.2. FT-IR and FT-NIR spectra of cyclohexanol in CCl4 and cyclohexane The IR and NIR spectra of cyclohexanol in CCl4 (Fig. 5) were divided by the corresponding molar concentration and the cell thickness. Hence, they represent the molar absorption coefficients (e). The relative intensities and positions of the peaks in the IR spectra are close to those reported by Sandorfy [16]. As expected, increasing concentration of cyclohexanol reduces e of the OHf peak in both spectral regions as a result of the hydrogen bonding creation between the molecules of cyclohexanol. The relative intensities of the bands due to the hydrogen-bonded OH groups are quite different in both spectral regions (Fig. 5). The associated species are clearly seen in the IR spectra even at concentration of 0.1 M. In
Fig. 6A and B show the effect of temperature on FT-NIR spectra of pure cyclohexanol and cyclohexanol/water mixture, respectively. The positions of the peaks in pure cyclohexanol and in the mixture are collected in Table 2. The most significant changes between the spectra are observed in the 5000–5300 cm1 region, where is located the m2 + m3 band of water. The other band of water (m1 + m3) is heavily overlapped by absorption of the associated OH groups of cyclohexanol. Interestingly, the bands due to vibrations of the hydrocarbon chain (near 5700 and 5800 cm1) are not influenced by the presence of water or temperature suggesting a minor importance of the hydrophobic interaction in the studied system. Addition of water causes a small blue-shift of the bands due to
63
M.A. Czarnecki et al. / Journal of Molecular Structure 974 (2010) 60–67
Molar absorption coefficient
10
A
8 6 4 2 0 3100 3200 3300 3400 3500 3600 3700 3800 Wavenumbers/cm-1
Molar absorption coefficient
0.04
B 0.03
0.02
0.01
6800
6900
7000
7100
7200
7300
-1
Fig. 4. The lowest-energy conformers of 2:1 cyclohexanol/water complexes obtained from DFT calculations with the cyclic (A) and linear (B) structure.
the OHf and OHad vibrations of cyclohexanol. The large shift for the OHd band (170 cm1) results from overlapping by the m1 + m3 band of water. Like for others previously studied alcohols [10], the m2 + m3 band of associated water in the mixture is red-shifted (28 cm1) as compared with that in the spectrum of bulk water. This suggests that water molecules in the mixture are involved in stronger hydrogen bonding that those in bulk water. The synchronous (not shown) and asynchronous (Fig. 7A) contour plots of pure cyclohexanol are similar to those previously obtained for aliphatic alcohols [3,30,31]. In the asynchronous spectrum are resolved the peaks due to the rotational conformers of the OHf group at (7047, 7074) cm1. The positions of these peaks are the same as those observed in CCl4 solution (Tables 1 and 2). Also position of the peak due to the cooperative hydrogen bonding (OHad) in pure cyclohexanol and in CCl4 solution is similar. In contrast, the peak attributed to the OHd group is significantly redshifted (140 cm1) on going from pure cyclohexanol to CCl4. This means that both peaks originate from different species. The peak at 6819 cm1 is due to the open chain dimers (Fig. 2A) in pure cyclohexanol, whereas the 6679 cm1 peak can be attributed to small oligomers of cyclohexanol in CCl4. In liquid cyclohexanol the later peak is overlapped by strong absorption of OHad group. It is of note that the asynchronous spectrum of the mixture (Fig. 7B) does not develop the peaks due to the rotational isomerism of the OHf group. This effect, observed for the first time in nbutyl alcohol/water mixture [32], was explained as the lack of the conformational selectivity by water molecules upon the hydrogen bonding formation with the molecules of alcohol.
Wavenumbers/cm
Fig. 5. FT-IR (A) and FT-NIR (B) spectra of cyclohexanol in CCl4 at 30 °C. Solid line – 0.002 M, dashed line – 0.01, dotted line – 0.1 M and dash-dot line – 1 M.
Table 1 Assignments of IR (m01) and NIR (m02) bands of cyclohexanol in CCl4 and cyclohexane solutions at 30 °C together with the anharmonicity constants [v = m02/2 m01]. Position (cm1) IR (CCl4)
NIR (CCl4)
NIR (cyclohexane)
2857 2933 3622 3610 3482 3336
5707 5802 7074 7047 6679 6280
– – 7080 7050 – 6283
v
Assignment
85 86 142.5 196
CH CH OHf, gauche conformer OHf, trans conformer OHd OHad
Both PCA and EFA analysis of temperature-dependent spectra of pure cyclohexanol suggest the presence of two independent components that were resolved by MCR-ALS method (Fig. 8A). However, reasonable results one can obtain by assuming the presence of the third component. The spectral and concentration profiles of this component (Figs. 8B and 8C) have maxima near 6770 cm1 and 55 °C, respectively. The concentration profiles of two other components monotonously increase and decrease with the temperature. The third component can be assigned to the OHd group and corresponds to the asynchronous peak at 6819 cm1. The difference in the position results from the fact that 2D correlation peaks are shifted in the direction of higher intensity changes [33–35].
64
M.A. Czarnecki et al. / Journal of Molecular Structure 974 (2010) 60–67
0.6
A Wavenumber/cm-1, ν2
0.5 Absorbance
A
0.4 0.3 0.2 0.1
7000
6500
6000
4500
5000
5500
6000
6500
7000
6000
7500
7000
Wavenumbers/cm-1, ν1
Wavenumbers/cm -1 1.5
B
B Wavenumber/cm-1, ν2
1.2 Absorbance
6500
0.9 0.6
7000
6500
0.3
6000 4500
5000
5500
6000
6500
7000
7500
Wavenumbers/cm -1 Fig. 6. FT-NIR spectra of pure cyclohexanol (A) and cyclohexanol/water mixture with X H2 O ¼ 0:4 (B) from 30 to 80 °C. The arrows indicate directions of intensity changes on the temperature increase.
Table 2 Frequencies and assignments of selected NIR bands of cyclohexanol (1) and water (2) in the pure liquid phases and in the cyclohexanol/water mixture (X H2 O ¼ 0:4). Position (cm1)
a b c d
Pure (1)
Mixture (1 + 2)
8325 7074a 7047a 6819b 6243 5802 5705 – – 5184d
8325 7075a 7050a 6990b,c 6250 5802 5705 5295 5200b 5156
Vibration
m03CH m02OH m02OH m02OH m02OH m02CH m02CH m2 + m3 m2 + m3 m2 + m3
Assignment
OHf, gauche conformer OHf, trans conformer OHd (non-cooperative) OHad (cooperative)
Water, OwHf Water, OwHd (non-cooperative) Water, OwHad (cooperative)
Peaks resolved in second derivative spectra. Peaks resolved in 2D correlation spectra. Overlapped by the m1 + m3 band of water. In bulk water.
The 2D correlation contour plots constructed from the temperature-dependent spectra of the mixtures indicate the presence of at least two different water species. Yet, the band shift observed in the temperature-dependent spectra may generate artifacts in 2D correlation contour plots [33–35]. The presence of three distinct water species was confirmed by MCR-ALS analysis of the temperature-dependent spectra of the mixture with X H2 O ¼ 0:4 (Fig. 9). The
6000
6500
7000
Wavenumber/cm-1, ν1 Fig. 7. Asynchronous spectra of pure cyclohexanol (A) and cyclohexanol/water mixture with X H2 O ¼ 0:4 (B) from 30 to 80 °C. The positive contours are shaded.
positions of the corresponding peaks obtained by 2D correlation analysis and MCR-ALS method are similar making these results more reliable. The concentration profiles of the mixture (not shown) are similar to those obtained for the pure cyclohexanol (Fig. 8C). Fig. 10 shows the power spectra constructed from the temperature-dependent spectra of pure cyclohexanol and cyclohexanol/ water mixtures. The power spectrum is a diagonal of the synchronous spectrum and it represents the overall extent of intensity changes with respect to the reference spectrum. As can be seen, the extent of spectral changes of the OHf group increases on going from the pure cyclohexanol to the mixture. Since the spectral changes at all wavenumbers are continuous one can conclude that the temperature-induced increase in the population of the OHf groups of cyclohexanol occurs faster in the presence of water. In contrast, the number of the most associated species in pure cyclohexanol decreases faster than that in the mixture. This means that the population of the OHf groups increases mainly at the expense of the intermediate species. In other words, addition of water stabilizes the higher associates of cyclohexanol, whereas the shorter ones are disrupted more easily upon the temperature rise. 3.4. Effect of water content on FT-NIR spectra of cyclohexanol/water mixtures Fig. 11A shows the difference spectra calculated by subtraction of the spectrum of pure cyclohexanol at 30 °C from the concentra-
65
M.A. Czarnecki et al. / Journal of Molecular Structure 974 (2010) 60–67
0.16
A
1.2
Absorbance
Absorbance
0.12
0.08
0.8
0.4
0.04
6000
6500
7000
5000
7500
5500
6000
6500
7000
7500
Wavenumbers/cm -1
Wavenumbers/cm -1
Fig. 9. Spectral profiles obtained from MCR-ALS of the spectra of cyclohexanol/ water mixture with X H2 O ¼ 0:4 from 30 to 80 °C.
0.16
B 1.8
x 10
-3
1.5
0.08
Arbitrary units
Absorbance
0.12
0.04
6000
6500
7000
7500
1.2 0.9 0.6 0.3
wavenumbers/cm -1 4.5
6000
C
A.U.
6400
6600
6800
7000
7200
Wavenumbers/cm -1 Fig. 10. Power spectra of pure cyclohexanol (solid line) and cyclohexanol/water mixtures with X H2 O ¼ 0:2 (dotted line) and X H2 O ¼ 0:4 (dashed line) from 30 to 80 °C.
3.5
2.5
1.5 30
6200
40
50
60
70
80
Temperature/ o C Fig. 8. Spectral (A and B) and concentration (C) profiles obtained from MCR-ALS of the spectra of pure cyclohexanol from 30 to 80 °C.
tion-dependent spectra of the mixture recorded at the same temperature. As can be seen, an increase in water content does not affect intensities of the cyclohexanol bands, but changes the relative intensities of the free and associated OH groups of water. The band at 5295 cm1 clearly appears at small water content (Fig. 11A), whereas at X H2 O > 0:3 it is resolved only in the second derivative spectra (not shown). The intensity of the hydrogen-bonded band
of water increases linearly with X H2 O , whereas the changes of the band due to the OwHf group have a non-linear character (Fig. 11B). At higher water content this increase becomes slower, evidencing the growing importance of water–water interactions. Fig. 11C shows the power spectrum obtained from the concentration-dependent data. In the spectrum occur mainly the peaks due to water: m2 + m3 at 5160 cm1 and m1 + m3 at 6863 cm1. The peak resulting from absorption of the free OwH groups (5295 cm1) is less intense as compared with the analogous peak in the power spectra of butyl alcohol/water mixtures [10]. It has been observed a clear relationship between the relative intensity of this peak and the degree of association of the alcohol. The higher extent of self-association of alcohol – the smaller intensity of this peak [10]. In the case of more associated compounds like propanediols, this peak completely disappears from the power spectra [36]. The position of the peaks due to the rotational isomerism of the OHf group (Table 2) indicates that the strength of the hydrogen bonding in cyclohexanol is typical as that of other secondary alcohols [28]. Nevertheless, the high viscosity of cyclohexanol (>10 times higher than that of butyl alcohols) enhances the stability of the intermolecular hydrogen bonds.
66
M.A. Czarnecki et al. / Journal of Molecular Structure 974 (2010) 60–67
A Wavenumbers/cm-1, ν2
Absorbance
A
5400
1.2
0.9
0.6
0.3
5000
5500
6000
6500
7000
5200
5000
4800
7500
5200
5400
Wavenumbers/cm -1, ν1
Wavenumbers/cm -1
x 10
5000
-3
1.2
B
B
3 Absorbance
Arbitrary units
0.9
2
0.6
0.3
1
5000
0
0.1
0.2
0.3
0.4
XH O
C
Arbitrary units
7
5
(x10)
3
1 5000
5500
6000
6500
Wavenumbers/cm
7000
6000
6500
7000
7500
Wavenumbers/cm -1 Fig. 12. Asynchronous spectrum (A) and the spectral profiles obtained from MCRALS (B) of the spectra of the cyclohexanol/water mixtures with X H2 O = [0:0.05:0.4] at 30 °C. The positive contours in Fig. 12A are shaded. The dotted lines in Fig. 12B represent the spectra of pure cyclohexanol and pure water at 30 °C.
2
9
5500
7500
-1
Fig. 11. The difference spectra obtained by subtraction of the spectrum of pure cyclohexanol from the spectra of the cyclohexanol/water mixtures with X H2 O = [0:0.05:0.4] at 30 °C (A) and the power spectrum obtained from these data (C). In (B) are displayed the intensities of the bands at 5150 cm1 (dashed line) and 5295 cm1 (solid line) obtained from the second derivative of the difference spectra shown in Fig. 11A.
The peak positions in the spectra of the mixtures do not depend on water content at constant temperature. Hence, one can assign a real physical meaning for the peaks developed in the asynchronous
spectrum (Fig. 12A). As can be seen, the m2 + m3 peak is resolved into two components that are defined by four frequencies: 5133, 5204, 5225 and 5295 cm1. Taking into account the fact that the peaks located close to the diagonal are more shifted [33–35], one can assume that the peaks at 5204 and 5225 cm1 originate from the same species. Accordingly, in the asynchronous spectrum are resolved three non-equivalent OwH groups. The low frequency band (5133 cm1) can be assigned to the cooperative HB, where both OwH groups are donors to oxygens of cyclohexanol, while the high frequency component (5295 cm1) is due to absorption of the free OwH groups in one-bonded water. The presence of the third peak at intermediate position (5204 cm1) results from weakly hydrogen bonded water to cyclohexanol, e.g., from OwH donors in onebonded water. The asynchronous spectrum reveals that the changes at 5204 cm1 occur at different rate than those for two remaining bands. It is very likely that the 5204 cm1 peak includes a contribution from the water–water interactions taking place at higher water content. Due to small size of these clusters the water–water interactions are weaker than those in bulk water. Fig. 12B displays the spectral profiles obtained from MCR-ALS of the concentration-dependent data. The profile of cyclohexanol is identical with the spectrum of pure cyclohexanol. On the other hand, the peaks in the spectral profile of water are red-shifted with respect to those in the spectrum of bulk water. This means that the molecules of cyclohexanol in the mixture are in the same environ-
M.A. Czarnecki et al. / Journal of Molecular Structure 974 (2010) 60–67
ment as those in pure cyclohexanol, whereas the water molecules are involved in stronger hydrogen bonding than those in bulk water. 4. Conclusions Addition of small to moderate amount of water has a minor effect on the structure of liquid cyclohexanol at constant temperature. Molecules of cyclohexanol in the mixture are in the same environment as those in the pure cyclohexanol. DFT calculations suggest a significant population of the cyclic tetramers in liquid cyclohexanol. Most of water molecules in the mixture act as a double donors to oxygens of cyclohexanol. This cooperative hydrogen bonding appears to be stronger than that in bulk water. At lower water content exists in the mixture a certain amount of singly bonded water molecules. Yet, the number of the free OwH groups in the cyclohexanol/water mixture is much smaller as compared with that in butyl alcohol/water mixtures. It results from high viscosity of cyclohexanol that stabilize the cyclohexanol–water interaction. Strongly non-linear relationship between the population of the free OwH and X H2 O , particularly at higher water content, suggests growing importance of the water–water interactions. These interactions are responsible for creation of small clusters of water, where the hydrogen bonds are significantly weaker than those in bulk water. The temperature-induced breaking of smaller associates of cyclohexanol occurs easier in the presence of water. In contrast, addition of water leads to stabilization of the higher associates. The hydrophobic interactions in the pure liquid cyclohexanol and cyclohexanol/water mixtures do not play an important role.
[8] [9] [10] [11] [12] [13] [14] [15] [16] [17]
[18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30]
References [31] [1] A.F.M. Barton, Alcohols with Water, Solubility Data Series, vol. 15, Pergamon Press, Oxford, 1984. [2] N. Šegatin, C. Klofutar, Monatsh. Chem. 135 (2004) 241–248. [3] M.A. Czarnecki, K. Orzechowski, J. Phys. Chem. A 107 (2003) 1119–1126. [4] S. Dixit, W.K.C. Poon, J. Crain, J. Phys. Condens. Matter 12 (2000) L323–L328. [5] T. Sato, A. Chiba, R. Nozaki, J. Chem. Phys. 112 (2000) 2924–2932. [6] T. Sato, A. Chiba, R. Nozaki, J. Mol. Liq. 101 (2002) 99–111. [7] M. D’Angelo, G. Onori, A. Santucci, J. Chem. Phys. 100 (1994) 3107–3113.
[32] [33] [34] [35] [36]
67
G. Onori, A. Santucci, J. Mol. Liq. 69 (1996) 161–181. Y. Koga, Can. J. Chem. 66 (1988) 3171–3175. M.A. Czarnecki, D. Wojtków, J. Mol. Struct. 883–884 (2008) 203–208. P.L. Huyskens, M.Cl. Haulait-Pirson, I. Hanssens, J. Mullens, J. Phys. Chem. 84 (1980) 28–33. S. Dixit, J. Crain, W.C.K. Poon, J.L. Finney, A.K. Soper, Nature 416 (2002) 829– 832. K. Yoshida, T. Yamaguchi, Z. Naturforsch. 56a (2001) 529–536. A. Wakisaka, K. Matsuura, J. Mol. Liq. 129 (2006) 25–32. U. John, K.P.R. Nair, Spectrochim. Acta A 61 (2005) 2555–2559. C. Sandorfy, Bull. Pol. Acad. Sci. Chem. 43 (1995) 7–24. M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman, J.A. Montgomery Jr., T. Vreven, K.N. Kudin, J.C. Burant, J.M. Millam, S.S. Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J.E. Knox, H.P. Hratchian, J.B. Cross, C. Adamo, J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev, A.J. Austin, R. Cammi, C. Pomelli, J.W. Ochterski, P.Y. Ayala, K. Morokuma, G.A. Voth, P. Salvador, J.J. Dannenberg, V.G. Zakrzewski, S. Dapprich, A.D. Daniels, M.C. Strain, O. Farkas, D.K. Malick, A.D. Rabuck, K. Raghavachari, J.B. Foresman, J.V. Ortiz, Q. Cui, A.G. Baboul, S. Clifford, J. Cioslowski, B.B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R.L. Martin, D.J. Fox, T. Keith, M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong, C. Gonzalez, J.A. Pople, Gaussian 03, Revision C.02, Gaussian Inc., Wallingford, CT, 2004. I. Noda, Appl. Spectrosc. 47 (1993) 1329–1336. I. Noda, Appl. Spectrosc. 54 (2000) 994–999. S. Wold, K. Esbensen, P. Geladi, Chemom. Intell. Lab. Syst. 2 (1987) 37–52. H.R. Keller, D.L. Massart, Chemom. Intell. Lab. Syst. 12 (1992) 209–224. R. Tauler, B. Kowalski, S. Fleming, Anal. Chem. 65 (1993) 2040–2047. R. Tauler, A. Izquiedoro-Ridorsa, E. Casassas, Chemom. Intell. Lab. Syst. 18 (1993) 293–300. R.M. Ibberson, S. Parsons, D.R. Allen, A.N.T. Bell, Acta Crystallogr. B 64 (2008) 573–582. C. Jansen, an Mey Dieter, G. Raabe, J. Fleischauser, J. Mol. Struct. (Theochem) 398-399 (1997) 395-404. R.A. Pickering, C.C. Price, J. Am. Chem. Soc. 80 (1958) 4931–4933. W.A.P. Luck, W. Ditter, J. Mol. Struct. 1 (1967–1968) 261–282. M.A. Czarnecki, D. Wojtków, K. Haufa, Chem. Phys. Lett. 431 (2006) 294–299. D. Wojtków, M.A. Czarnecki, Appl. Spectrosc. 61 (2007) 928–934. M.A. Czarnecki, H. Maeda, Y. Ozaki, M. Suzuki, M. Iwahashi, Appl. Spectrosc. 52 (1998) 994–1000. M.A. Czarnecki, M. Maeda, Y. Ozaki, M. Suzuki, M. Iwahashi, J. Phys. Chem. A 102 (1998) 9117–9123. M.A. Czarnecki, D. Wojtków, J. Phys. Chem. A 108 (2004) 2411–2417. A. Gericke, S.J. Gadaleta, J.W. Brauner, R. Mendelsohn, Biospectroscopy 2 (1996) 341–351. M.A. Czarnecki, Appl. Spectrosc. 52 (1998) 1583–1590. M.A. Czarnecki, Appl. Spectrosc. 54 (2000) 986–993. K.Z. Haufa, M.A. Czarnecki, Vib. Spectrosc. 51 (2009) 80–85.