J. Chem. SOC., Faraday Trans I, 1988, 84(1), 29-36 The Interaction of Water with Non-electrolytes The System Water-Acetonitrile-1.,4-Dioxane Hero Mirti" and Vincenzo Zelano Dipartimento di Chimica Analitica, Universita' di Torino, via Giuria 5 , I10125 Torino, Italy 'H, 13C and 1 7 0 magnetic resonance has been used to investigate the degree of modification of the aqueous structure in water-acetonitrile-l,4-dioxane mixtures of different compositions. Chemical-shift data of the hydrogen and oxygen nuclei of water have enabled the calculation of a parameter cf,), which may be considered representative of the average situation experienced by one molecule of water in the presence of one molecule of non-electrolyte. Chemical-shift data of the hydrogen and carbon nuclei of acetonitrile and dioxan have been used to determine similar parameters uiN and f,"") representative of the situation experienced by one molecule of water in the presence of one molecule of acetonitrile or dioxane, respectively.The values of fiN and f,"" accord with those off,, and indicate that the degree of modification of the aqueous structure is generally independent of the composition of the non-electrolyte. It is known that solvent features can affect chemical processes in solution. The observed effects, such as a change in the reaction rate or a shift of equilibrium conditions, can frequently be correlated with the physico-chemical properties of the medium. As a change in the latter can dramatically alter the chemical behaviour of a given compound, it is often useful to search for the most suitable medium by mixing different solvents in appropriate amounts.In this way some properties can be varied, e.g. the donor capability or the dielectric constant, simply by varying the composition of the medium. Because of its role in the natural environment, water is one of the most recurrent components of mixed solvents. The addition of a cosolvent to water alters the hydrogen- bonded structure of the latter, and this may be one factor in determining the effect of the medium on the behaviour of the species in solution. The most likely modification made to the network of H,O molecules is the distortion or rupture of the existing links,'-' even though some enhancement of the aqueous structure has been observed at low temperatures in the presence of small quantities of certain non-electrolytes.2* 3, '-12 Whether water-water hydrogen bonds are distorted rather than broken in the presence of organic cosolvents can be widely discussed.The most current view seems to favour the idea of a distorted network, but deciding at what point a distorted link is turned into a broken one may be largely a matter of opinion. It seems likely, however, that an actual rupture of links should gradually take place when the amount of cosolvent becomes predominant in the system and fewer molecules of water are dispersed in the organic matrix, with water-water bonds eventually replaced by water-solvent links. One can generally speak of a weakening of the structure of water, as opposed to its strengthening, but it must be borne in mind that the former may be the result of either simple distortion or more severe breaking of the bonds involved.In this context it is useful to remember that the features of the hydrogen bond network have been argued even in pure water,13-18 where the intermolecular links may be distorted (or broken) by a change of temperature. 2930 Interact ion of Water with Non-elec troly tes In a previous paperlg some binary systems formed by water and a non-hydroxylic solvent are reported. The results obtained showed a correlation between the extent of modification of the water structure and the physico-chemical properties of the cosolvent. In this context it was surprising to note that acetonitrile (AN) and 1,4-dioxane (DO) caused very similar effects, in spite of their different properties.A subsequent study of a ternary system of water, dimethyl sulphoxide (DMSO) and DO showed that the non-electrolytes can affect each other's capability of modifying the structure of water.20 This suggested a study of water-acetonitrile4ioxan mixtures, to discover whether AN and DO can replace each other without changing the degree of distortion (or rupture) of the hydrogen-bond network of water. Among the many techniques allowing one to collect information on the structural features of mixed solvents, nuclear magnetic resonance is one of the most suitable, because the chemical shift (6) of the nuclei of the molecules involved is affected by the structural changes. In this work, magnetic resonance of all 'H, 13C and 170 nuclei of water, acetonitrile and dioxan has been used to study the system under investigation.Experimental Dried acetonitrile and 1,4-dioxane (Riedel-De Haen, maximum water content 0.03 and 0.0 1 YO, respectively) were mixed with demineralized, twice-distilled water to prepare samples in which the mole fraction of each component could vary from 0 to 1. For each mixture, 5 x loh3 dm3 were prepared by using an Amel 233 digital automated burette operated by an Apple IIe computer. The precision of the buret (& 1 x dm3) allowed one to obtain standard deviations of the mole fraction of each component in the range 'H n.m.r. spectra were recorded by a Varian T-60 spectrometer using tetramethylsilane (TMS) or sodium 4,4-dimethyl-4-silapentane sulphonate (DSS) as the internal reference for the determination of chemical shifts.The solubility of the reference compounds in the samples under investigation determined the choice between DSS and TMS, but the latter was used whenever possible. A Jeol GX-270/89 spectrometer was used for recording 13C and 170 n.m.r. spectra, and deuterated dimethyl sulphoxide and methanol were used as external references in the determination of the chemical shifts of the carbon and oxygen nuclei, respectively. The instrumental resolution was as much as 0.01 ppm for 'H and 13C, and 0.05 ppm for 170. This led to kO.01 error in the values of a and k0.02 in those off, (see below for definitions). Exceptions were a values obtained from proton resonance of AN and DO (error up to k0.08) and fM values in the water-rich region (error up to kO.1 for xw > 0.9).Results and Discussion (1-5) x 10-3. The chemical shift of the water protons in a system containing a network of more or less distorted hydrogen bonds is given by 6 = C X i 6 , where di is the chemical shift of the protons engaged in hydrogen bonds with a determined degree of distortion and xi the corresponding fractional population with respect to the total number of protons. In the most simple situation, when hydrogen bonds are only broken and not distorted, eqn (1) becomes 6 = x,s;,+x,s; where 8; and 6; are the chemical shifts of the protons which are not hydrogen bonded and normally hydrogen bonded (i.e. without distortion), respectively, and x, and x, areP. Mirti and V. Zelano 31 the corresponding fractional populations.In the most complex situation a continuous distribution of hydrogen bond energies is present, and the sum in eqn (1) must be replaced by an integral. A determination of the true value of 6; is not simple because even pure water can contain broken or distorted bonds, and the energy distribution of these is dependent on temperature.l3-l8 Therefore 6; is more likely to be an unknown value at lower field than that obtainable by n.m.r. measurement on pure water at a given temperature (dN). On the other hand, 6; should be referred to a standard situation where the water molecules can be considered truly free, e.g. as at infinite dilution in a non-interacting solvent or in the vapour state. When water is in the presence of an interacting cosolvent a value (6,) is obtained at lower field than Sl,, owing to the coordinating properties of the cosolvent towards water.Different values'of 6, have been obtained for water molecules diluted in different non-electrolytes, and a correlation has been found between these values and the donor numbers of the cos01vents.~~ The chemical shift of the water protons at a given temperature varies with the composition of a system containing water and one (or more) non-electrolyte(s), moving upfield from the value obtained in pure water at that temperature (6,) to the value extrapolated for water infinitely diluted in that (or those) non-electrolyte(s) (aF). In spite of the complexity of the system, it seems possible to use the experimental chemical shift obtained for a given mixture as a measure of the modifications caused in the structure of water by the presence of that (or those) non-electrolyte(s) at that temperature relative to the two boundary situations represented by 6, and 6,.Because 6, and 6, are different from 6; and 6;, respectively, and depend upon temperature and cosolvent, one can only obtain relative information; however, this can allow one to infer the extent of modification of the water network in the presence of a given non-electrolyte and to compare the modifications caused by different non-electrolytes. In order to attempt a quantitative estimation of the above, one can introduce a parameter (f), which is required as a measure of the average degree of distortion of the water structure provoked by one molecule of non-electrolyte.If different molecules give additive contributions, and if x, is the mole fraction of water in the system, fT1- x,) is a measure of the overall variation of the aqueous structure from pure water at a working temperature. If hydrogen bonds are not distorted but actually broken, f becomes the number of links broken by a single molecule of non-electrolyte. A more convenient way to cope with the matter may be the use of a correlated parameter fM =f/xw, which is related to an average situation experienced by a single molecule of water. A practical reason for usingf, instead offlies in the fact thatfmust be zero for both x, = 1 and x, -+ 0 (i.e. where no cosolvent or water is present, and therefore no structural modifications are possible). On the other hand,f, can range from zero (x, = 1, where the structure is that of pure water at the working temperature) to 2 (x, -+ 0, where all the water-water bonds must be broken and partially replaced by water-cosolvent interactions).As long as hydrogen bonds are only broken and not distorted, eqn (1) is replaced by eqn (2), and this can be rewritten as where fT1 -xw)/2x, is the fraction of water protons not engaged in water-water hydrogen bonds and [2x, -A1 -xw)]/2x, is the fraction of bonded 0nes.l' In this case fM can be obtained as (4) In a more general case, when hydrogen bonds are distorted rather than broken, it seems possible that values off, obtained from eqn (4) may equally well be used to give a picture of the situation created by the presence of a non-electrolyte in water.In fact, because the same experimental value of 6 can stem from only a few broken links f M = 2(SN - 6)/[(6N - 'F) ( l - xw)l' 2 FAR 132 Interaction of Water with Non-electrolytes xw Fig. 1. Variation of the chemical shift of the 'H (a) and I7O nuclei (b) of water as a function of the mole fraction of H,O in mixtures H,O-AN-DO containing equal mole fractions of acetonitrile and dioxane. 3 0 0.2 0.4 0.6 0.8 1 XAN/@AN + XDO) Fig. 2. Variation of the chemical shift of the 'H (a) and I7O (b) nuclei of water as a function of the composition of the non-electrolyte in H,O-AN-DO mixtures containing 0.5 mole fraction of H,O. as well as from many distorted ones, it seems possible to speak in terms of an apparent number of bonds broken by the non-electrolyte.Within the limits of the above assumptions, eqn (4) has been used to calculate values off, from the chemical shifts of both 'H and ''0 nuclei of the water molecules; the values obtained from the hydrogen and oxygen nuclei agree with each other. Fig. 1 shows the variation of the chemical shift of the lH and l 7 0 nuclei with the mole fraction of water in samples containing a fixed molar ratio AN/DO, whereas fig. 2 reports the variation of these chemical shifts as a function of the composition of the cosolvent in samples with a fixed water content. The complete set of proton chemical shifts is given in table 1. The values off, obtained are given in table 2, together with those calculated previously for the binary mixtures H,O-AN and H2O-DO.l9 It is inferred from the data of table 2 that the degree of modification of the aqueous structure is generally independent of the composition of the non-electrolyte except whenP.Mirti and V. Zelano 33 Table 1. Chemical shifts of the hydrogen nuclei of water (aw), acetonitrile (aAN) and dioxane (ano) in H,GAN-DO mixtures of different compositions 0.90 0.90 0.90 0.90 0.90 0.70 0.70 0.70 0.70 0.70 0.50 0.50 0.50 0.50 0.50 0.30 0.30 0.30 0.30 0.30 0.10 0.10 0.10 0.10 0.10 0.90 0.70 0.50 0.30 0.10 0.90 0.70 0.50 0.30 0.10 0.90 0.70 0.50 0.30 0.10 0.90 0.70 0.50 0.30 0.10 0.90 0.70 0.50 0.30 0.10 4.49 4.50 4.51 4.5 1 4.52 4.1 1 4.12 4.15 4.14 4.15 3.76 3.77 3.78 3.79 3.79 3.26 3.3 1 3.34 3.35 3.39 2.58 2.68 2.71 2.74 2.75 3.72 3.72 3.72 3.72 3.72 3.67 3.67 3.68 3.65 3.67 3.66 3.65 3.65 3.64 3.62 3.63 3.62 3.62 3.61 3.61 3.62 3.60 3.60 3.58 3.57 2.06 2.06 2.06 2.06 2.06 2.03 2.03 2.03 2.04 2.03 2.0 1 2.01 2.00 2.03 2.0 1 2.00 2.00 1.98 1.99 2.00 1.98 1.97 1.98 1.96 1.95 0.05 0.05 0.05 0.05 0.05 0.03 0.03 0.03 0.03 0.03 0.02 0.02 0.02 0.02 0.02 0.01 0.0 1 0.01 0.0 1 0.01 0.90 0.70 0.50 0.30 0.10 0.90 0.70 0.50 0.30 0.10 0.90 0.70 0.50 0.30 0.10 0.90 0.70 0.50 0.30 0.10 2.4 1 2.49 2.53 2.55 2.56 2.34 2.40 2.46 2.49 2.50 2.30 2.40 2.45 2.45 2.45 2.22 2.33 2.40 2.43 2.41 3.62 3.60 3.59 3.58 3.58 3.60 3.60 3.60 3.59 3.59 3.60 3.60 3.59 3.59 3.57 3.59 3.60 3.58 3.60 3.58 1.98 1.96 1.96 1.95 1.95 1.97 1.96 1.96 1.96 1.95 1.97 1.97 1.95 1.96 1.94 1.95 1.96 1.95 1.96 1.94 Table 2.Values off, for water-acetonitrile-dioxane mixtures of different compositions X , , / ( X , ~ + X , ~ ) = 0.00 0.10 0.30 0.50 0.70 0.90 1 .oo ~ ~ ~ _ _ _ _ _ _ _ _ ~ ~ ~ ~ 0.10 1.48 1.85 1.82 1.79 1.80 1.71 1.45 0.30 1.59 1.63 1.65 1.62 1.65 1.62 1.57 0.50 1.59 1.54 1.57 1.60 1.59 1.59 1.59 0.70 1.62 1.68 1.66 1.68 I .67 1.62 I .62 0.90 1.84 1.91 1.87 1.90 1.88 1.87 1.84 water is present in a great excess, when one finds an increase infM on passing from AN or DO alone to a mixed non-electrolyte; this means that the capabilities of the two solvents to modify the energy distribution of the water-water hydrogen bonds can be mutually enhanced under certain conditions.This is a situation already in evidence for systems containing water in the presence of both DMSO and DO, and it can be explained if the non-electrolytes interact differently with water, so that a cooperative action can stem from their mixing.This may happen, for example, if the donor capability of one compound (e.g. dioxane) can gain efficacy from the larger dielectric constant of the other (e.g. acetonitrile). 2-234 Interaction of Water with Non-electrolytes B 120 - h E ( o f 118 - W I 67 - 1,,,,,,,,,1 0 0.2 0.4 0.6 0.8 1 XW Fig. 3. Variation of the chemical shift of lH (A) and 13C nuclei (B) of acetonitrile (b) and (a) dioxane as a function of the mole fraction of water in H,O-AN-DO mixtures containing equal mole fractions of DO and AN. 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 XAN/(XAN + XDO) XAN/(XAN + XDO) Fig. 4. Variation of the chemical shift of 'H (A) and I3C nuclei (B) of dioxane (a) and acetonitrile (b) as a function of the composition of the non-electrolyte in H,O-AN-DO mixtures containing 0.5 mole fraction of water.Similarly to water nuclei, the chemical shifts of the nuclei of acetonitrile and dixoane (figs. 3 and 4 and table 1) can be considered representative of the variation of the structural features of AN and DO, respectively, when going from the pure cosolvent towards infinite dilution in water through the formation of complex hydro-organic species. In line with the points discussed above, if only two boundary forms contribute to the observed chemical shift, one may calculate a fractional population for each of them. As already said, this may not be the case with water, or indeed with AN or DO, either. However, pure AN and pure DO molecules are not part of a network of strong intermolecular interactions.Therefore, a wide span of intermediate situations is less probable for AN or DO than for water, so that the picture obtained by considering only two boundary structures (that of the pure cosolvent and that of the cosolvent at infinite dilution in water) may be less approximate. In this case one can calculate the fractionalP. Mirti and V. Zelano 35 Table 3. Fractional populations of acetonitrile (a,"") and dioxane (01:~) molecules coordinated to water, calculated from 'H and 13C n.m.r. data of mixtures containing equal mole fractions of either non-electrolyte atN a,Do 'H n.m.r. 13C n.m.r. 'H n.m.r. 13C n.m.r. x,, + X*N 0.10 0.85 0.83 0.8 1 0.86 0.30 0.62 0.58 0.56 0.54 0.50 0.39 0.39 0.38 0.37 0.70 0.23 0.22 0.19 0.20 0.90 - 0.08 0.06 0.1 1 Table 4.Comparison between values off,, xW(x,, + xDo) (from n.m.r. data of water nuclei) and fiNxWxAN + fgoxwxno (from n.m.r. data of acetonitrile and dioxane nuclei) obtained for H,O-AN-DO mixtures containing equal mole fractions of AN and DO 0.10 0.16 0.30 0.34 0.50 0.40 0.70 0.35 0.90 0.17 0.17 0.34 0.38 0.29 0.17 population of acetonitrile and dioxane molecules bound to water (atN and a!*, respectively) from the experimental chemical shifts as where a, may be either a,"" or a:", and 6, and 6, are the chemical shifts of the solvent considered in its boundary situations. By doing so, one neglects possible interactions between AN and DO in the mixed non- electrolyte, but the chemical shifts of the hydrogen and carbon nuclei of AN and DO vary only slightly with the composition of the cosolvent.Such a variation is increasingly less evident as the content of water increases, and disappears completely in an excess of water. This indicates that, even though interactions are possible between AN and DO, they are of much less importance than those of water with each cosolvent. The values of at" and a:" obtained according to the above considerations from data for either protons or carbon nuclei correspond well and are given in table 3. However, values of a,"" could not be obtained from the chemical shift of the oxygen nuclei of dioxane, and the methyl carbon of acetonitrile proved to be of no use in determining a,"". Within the limits of the approximations discussed above, the values of atN and a:" can be used to calculate the mole fraction of acetonitrile and dioxane bound to water (as atNxAN and a,""xD0, respectively, if x A N and x,, are the mole fractions of AN and DO in the system).Those, in turn, can lead one to calculate the mole fraction of the molecules of water bound to each cosolvent, provided that the kind of complex species formed is known. In this respect, a prevailing opinion is that the species formed most easily are those with a 1 : 1 molar ratio of water to cosolvent, even though species with 2: 1 and 1 : 2 molar ratios can be present in a great excess of water or cosolvent, re~pectively.~* ''9 21-24 If one assumes that 1 : 1 species are formed, a,ANxAN and a~"xDO are36 Interact ion of Water with Non - elec tr oly t es also likely to give the mole fractions of water molecules linked to acetonitrile and dioxan, respectively, and these must be multiplied by 2 to represent the mole fractions of water protons.If these two terms are divided by x,, and xDO, separately, and then both of them by x,, one obtains two factors which are representative of the average situation experienced by one molecule of water in the presence of one molecule of acetonitrile and dioxane, respectively (i.e. fkN = 2a,AN/xw and f go = 2a,D0/xw). As long as the overall modification of the structure of water in the presence of both acetonitrile and dioxane stems from additive contributions from the two cosolvents (as shown previously, except for the water-rich region), one can write taking into account that all fM,fiN a n d g o relate to a single molecule of water and non- electrolyte.The agreement between the two halves of eqn (6) (table 4) is particularly remarkable if one considers that the values off,” andgo, on one hand, and those of fM, on the other, are obtained from completely independent measurements of the chemical shifts of cosolvent and water nuclei, respectively. In conclusion, the results obtained confirm that acetonitrile and dioxane lead to a similar average modification of the structure of water and that a variation of the composition of an AN-DO mixed non-electrolyte has no particular effect on that structure. This, of course, does not settle the question as to whether bonds are broken or distorted : however, since the physico-chemical properties of the two non-electrolytes are different, it shows that the same apparent number of bonds broken can stem from water-cosolvent interactions of different kind.References 1 Yu. I. Naberukhin and S. I. Shuiskii, Zh. Strukt. Khim., 1967, 8, 606. 2 C. J. Clemett, J. Chem. SOC. A, 1969, 455. 3 B. Z. Gorbunov and Yu. I. Naberukhin, Zh. Strukt. Khim., 1972, 13, 20. 4 M. F. Fox and K. P. Whittingham, J. Chem. SOC., Faraday Trans. I , 1975, 71, 1407. 5 Yu. A. Volokhov, N. G. Dovbysh, V. B. Lebedev and V. E. Mironov, Zh. Strukt. Khim., 1975, 16, 6 C. Moreau and G. Douhiret, J. Chem. Thermodyn., 1976, 8, 403. 7 A. J. Easteal, Aust. J. Chem., 1979, 32, 1379. 8 D. N. Glew, H. D. Mak and N. S. Rath, J. Chem. SOC., Chem. Commun., 1968, 264. 9 C. J. Clemett, J. Chem. SOC. A , 1969, 458. 1013. 10 T. Tokuhiro, L. Menafra and H. H. Smart, J . Chem. Phys., 1974, 61, 2275. 11 0. D. Bonner and Y. S. Choi, J. Phys. Chem., 1974, 78, 1727. 12 B. Z. Gorbunov, V. S. Kozlov and Yu. I. Naberukhin, Zh. Strukt. Khim., 1975, 16, 808. 13 K. Buijs and G. R. Choppin, J. Chem. Phys., 1963, 39, 2035. 14 D. P. Stevenson, J. Phys. Chem., 1965, 69, 2145. 15 M. R. Thomas, H. A. Scheraga and E. Schrier, J. Phys. Chem., 1965, 69, 3722. 16 0. D. Bonner and Y. S. Choi, J. Phys. Chem., 1974, 78, 1723. 17 M. C. R. Symons, J. M. Harvey and S. E. Jackson, J. Chem. SOC., Faraday Trans. I , 1980, 76, 256. 18 S. A. Rice and M. G. Sceats, J . Phys. Chem., 1981,85, 1108. 19 V. Zelano and P. Mirti, 2. Phys. Chem. (Munich), 1983, 138, 31. 20 V. Zelano and P. Mirti, 2. Phys. Chem. (Leipig), 1986, 267, 857. 21 A. Fratiello and D. C. Douglass, J . Mol. Spectrosc., 1963, 11, 465. 22 A. Le Narvor, E. Gentric and P. Saumagne, Can. J. Chem., 1971,49, 1933. 23 B. Kingston and M. C. R. Symons, J. Chem. Soc., Faraday Trans. 2, 1973, 69, 978. 24 S. 0. Paul and T. A. Ford, Spectrochim. Acta, Part A, 1981, 37, 415. Paper 6/ 1598 ; Received 4th August, 1986