J. CHEM. SOC. FARADAY TRANS., 1994, 90(6),865-868 Preferential Solvation of a p-Sensitive Dye in Binary Mixtures of a Non-protic and a Hydroxylic Solvent Marivania Scremin, Sandra Patricia Zanotto, Vanderlei Gageiro Machado and Marcos Caroli Rezende” Departamento de Quimica ,Universidade Federal de S. Catarina, Norianopolis, SC 88040-970,Brasil The prefer enti aI soIvation of the soIvat oc h rom i c dye N,N,N’,N’-tetramethyleth yIened iami noacety I-acetonatocopper(ii) perchlorate (l),[Cu(tmen)(aca)]CIO, , in binary solvent mixtures comprising one hydroxylic component (water, methanol, ethanol or propan-2-01) and a non-protic co-solvent (acetone, acetonitrile or dimethylformamide) is discussed and interpreted in terms of hydrogen-bonding effects which alter the ‘intrinsic’ electron-donating ability of the hydroxylic solvent. Solvatochromic dyes have often been used in the investiga- tion of the properties of solvents and solvent mixtures.The realization that their behaviour in solution ultimately reflects specific interactions with their microenvironment has made them ideal probes for studies of preferential solvation in binary mixtures. Research in this area is often concerned with protic-aprotic solvent mixtures, owing not only to the wide use of such media in chemical processes, but also to their complexity, with strong dipole-dipole attractions occurring side by side with equally important hydrogen-bonding inter- actions. Dyes may be classified according to their sensitivity to dif- ferent properties of the medium.Compounds which are pre- dominantly sensitive to the electron-accepting properties of a solvent may be characterized as r-sensitive, whereas p-sensitive dyes mainly reflect the donor properties of the medium. The variety of solvatochromic probes employed in such studies have in common the fact that they are, in most cases. strongly sensitive to the hydrogen-bond accepting andjor donating ability of the medium.’-’ Thus, deviations from an ideal behaviour in mixtures comprising protic solvents are generally ascribed to hydrogen bonding of the solvent to the dye. Studies utilizing dyes which are sensitive to the electron- donating ability of the medium are much more scarce. Among these dyes, the series of salts of ethylene-diaminoacetylacetonatocopper(n), first prepared by Fukuda and Sone‘ are unique in that they are not sensitive to the electron-accepting ability of the polarizability of the medium, being considered as exclusive probes for the solvent donating ability.’ Thus, Marcus and Migron have employed com-pound 1 (Scheme 1) in the determination of Kamlet and Taft’s parameter /3 for quite a few solvents and solvent mix- tures.’~~ ’CH:, ICH: 1 J Scheme 1 A study on the solvatochromism of dye 1 in dimethyl-formamide (DMF)-nitromethane mixtures has been published.” We decided in the present work to extend this study to protic-aprotic solvent mixtures for various reasons. Besides the aforementioned scarcity of such investigations, we were interested in detecting anv Deculiar behaviour in these mixtures, as frequently happens when an r-sensitive dye is In addition, and contrary to the expected behaviour of DMF-nitromethane solutions of 1, where the electron-donating abilities of the two components are well established and are very different, the donating behaviour of solvent mixtures comprising protic components is in principle unpredictable.This stems from the fact that the values for electron-donating ability of protic solvents vary widely according to the method utilized for their obtention, an effect which may reflect the indirect action of hydrogen bonding in solution. Accordingly, one may encounter values of ‘bulk’ electron-donating ability, which differ appreciably from donor numbers of ‘isolated’ molecules.Such variations should, in principle, be found in binary mixtures of variable composition, where the degree of hydrogen bonding of the protic co-solvent changes with its mole fraction in the mixture. Thus, in spite of the simplification introduced in a system with a pure @-sensitive probe (no direct hydrogen- bonding interaction between the dye and the protic co-solvent), deviations from the ‘normal’ behaviour found in non-protic solvent mixtures may still occur, an anticipation which justifies the present work. Experimental The spectra of dye 1 in all binary mixtures were recorded on a Beckman DU-65 spectrophotometer. The copper dye 1 was prepared following a reported pro- cedure.6 Redistilled water and analytically pure alcohols were employed in all solutions.In addition, pure dimethyl-formamide and redistilled acetone and acetonitrile were dried over molecular sieves prior to use. Results and Discussion The solvatochromic behaviour of dye 1 in binary mixtures comprising one hydroxylic component is shown for acetone- ROH (Fig. l), acetonitrile-ROH (Fig. 2) and DMF-ROH (Fig. 3). The hydroxylic solvents employed were water, meth- anol, ethanol and propan-2-01. The data plotted in Fig. 1-3, which give the variations of the wavenumber, V,,,, of the dye in mixtures of variable molar composition, X, were fitted to third-order polynomials by means of a least-squares method. With the exception of the water-acetone plot, the general equation Vmax = a +bX +cX2 +dX3 was found to reproduce all data rather accu- rately, as shown in the figures.The values of the coefficients u, b, c and d for each binary mixture are given in Table 1. Values of Vmax in acetone-water mixtures were constant and equal to 16.86 x loM3cm-’ for water mole fractions ~0.13. J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 X X 0.2 0.4 0.6 0.8 1 1 1 1 1 1 ' 1 1 17.4 17.4 17.2 7 17.0 5 m 16.8 t t 16.8 1 16.8 1 II 0.2 0.4 0.6 0.8 0.2 0.4 0.6 0.8 X X Fig. 1 Variations of V,,, of dye 1 in binary mixtures with mole frac- Fig. 2 Variations of Vmax of dye 1in binary mixtures with mole frac- tion of the hydroxylic component: (a) acetone-water; (b) acetone-tion of the hydroxylic component: (a) acetonitrile-water; (b) methanol; (c)acetone-ethanol; (d) acetone-propan-2-01 acetonitrile-methanol; (c) acetonitrile-ethanol; (6) acetonitrile-propan-2-01 X For binary mixtures richer in acetone, the observed G,,, values were as follows 5,,, (X)]: 17.51(0),17.12(0.03), 17.01(0.05), 16.92 (0.075), 16.89 (0.1).A first distinction may be drawn between binary mixtures 16.9 1containing the relatively weak, non-protic donor solvents acetone and acetonitrile and those containing the strong donor DMF. Whereas in mixtures of the latter, the dye is in all cases preferentially solvated by DMF, in the acetone- Table 1 Values of the coefficients a, b, c and d in the polynomial Vmrx = a + bX + cX2+ dX3for each binary mixture 3 16.9coefficients/10-binary mixture a b C d ~ acetone-MeOH 17.51 -1.37 1.64 -0.84 Iacetone-EtOH 17.51 -0.30 -0.68 0.39 16-7* acetone-Pr'OH 17.51 -0.55 0.8 1 -0.69 acetonit rile-H 2O 17.30 -1.56 2.22 -1.16 acetonitrile-MeOH 17.30 -0.24 0.01 -0.22 0.2 0.4 0.6 0.8 acetonitrile-EtOH 17.30 -0.53 0.74 -0.71 X acetonitrile-Pr'OH 17.30 -0.21 0.89 -1.04 DMF-H 20 16.58 -0.03 0.49 -0.19 Fig.3 Variations of Cmnx of dye 1 in binary mixtures with mole frac- DMF-EtOH 16.58 0.12 -0.56 0.72 tion of the hydroxylic component: (a) DMF-water; (b) DMF-ethanol J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 Table 2 Estimated molar percentage of the hydroxylic component. ROH, in the solvation shell of dye 1, for a binary mixture with a 1 1 bulk molar composition non-protic co-solvent (%) - ~~ ROH acetone acetonitrile DMF ~ H2O ca. 100 74 30 MeOH 66 33 - EtOH 46 34 4 Pr’OH 37 3 - ROH and acetonitrile-ROH systems, preferential solvation IS shifted from the hydroxylic component to the non-protic co-solvent as we change from water to propan-2-01. The degree of solvation of the dye by the hydroxylic co- solvent in each binary mixture may be more readily grasped by a comparison of the values of Table 2.These values, obtained from the curves drawn in Fig. 1-3, are estimated percentages of ROH in the solvation shell of 1 when the bulk composition of the solvent mixture is 1 : 1. By employing the third-order polynomials given in Table 1, the percentages of ROH may be obtained from the relationship ROH(%) = (ru -Vo)/(V1 -V,), where the subscripts 0.5, 0 and 1 refer to the calculated Vmax values for solutions where X = 0.5, 0 and 1, respectively. Thus, in a 1 : 1 molar acetone-water mixture, the dye is exclusively surrounded by water molecules, whereas, in acetone-propan-2-01, the solvation shell of the dye comprises about 37% of alcohol molecules.A comparison of the acetone and acetonitrile systems indi- cates a greater effectiveness of the latter solvent in competing with the hydroxylic co-solvent for the solvation of the dye. Preferential solvation by acetonitrile is verified in all binarq alcoholic mixtures, a trend which is reversed only when water is the hydroxylic co-solvent (Fig.2). In the case of acetone mixtures, water and methanol solvate the dye preferentially. whereas ethanol-acetone mixtures exhibit a nearly ideal behaviour, with bulk compositions very similar to local solvent distributions around the dye (Fig. 1). This greater effectiveness of acetonitrile, as compared with acetone, in solvating the copper dye probably reflects a greater association of the former with the soft Cu” ion. because of the stronger TC interactions of the -CN group with the planar dye.’0*’2 The preferential solvation of [Cu(tmen)(aca)]ClO, bq DMF in nitromethaneedimethylformamide mixtures was ascribed to the greater electron-donating ability of the latter solvent.” However, when one of the components is a hydroxylic solvent, the picture that emerges is more complex.The value of Vmax for 1 in a pure solvent may be assumed to be a measure of the donating ability of the medium. A rea-sonably good correlation was obtained betweenV,,, and Gutmann’s donor numbers (DN) for a series of solvents.6 interactions is facilitated by the assumption of different electron-donating abilities for the hydroxylic co-3olvent. Its ‘bulk’ donating ability reflects and incorporates the effects of extensive hydrogen bonding, which alters the ‘intrinsic’ donating ability of isolated molecules. Fig. 2 shows that the copper dye 1 is better solvated by acetonitrile than by any alcohol, in spite of the larger Gmax of the former. This may be due to the soft nature of the copper cation mentioned above, which associates better with aceton-itrile than with any alcohol ROH.This, however, suggests that the ‘intrinsic’ electron-donating abilities of these alco- hols are in fact smaller than their ‘bulk’ values, obtained in pure solvents. This situation is reversed when water is the hydroxylic co- solvent. Comparison of Fig. l(a) and (d), or of Fig. 2(aj and (d), shows that water solvates 1 in binary mixtures much more effectively than propan-2-01. The reason for this is not, according to our view, that water is intrinsically a better donor solvent than propan-2-01. The opposite should, in fact, be true. The alkyl groups in an alcohol should render the hydroxy group more basic than in water. The greater degree of dye solvation by water, than by propan-2-01 rather reflects differences in hydrogen-bond donation of these two solvents.In binary mixtures a strong hydrogen-bonding solvent may act as a ‘solvent scavenger’, binding and sequestering the non-protic co-solvent from the solvation shell of the dye. It may also stabilize and reinforce, through hydrogen bonds, other hydroxylic molecules already present in the solvation shell of the dye. A hydrophilic net is woven around the dye, with the more hydrophobic co-solvent being gradually expelled from its solvation shell. This is accompanied by enhanced electron-donating ability of the axial ROH ligands, because of hydrogen bonding to other ROH molecules, as shown in Scheme 2. H H,o$”------0I F+ ‘” Scheme 2 These effects oppose the ‘intrinsic’ weak electron-donating ability of the ROH molecules vis-Ci-vis the soft copper complex, and, in the case of the strongest hydrogen-bonding donor solvent water, even surpass the greater affinity of the dye for acetonitrile [Fig.2(a)]. It is interesting, at this stage, to compare our observations with similar studies published previously involving binary aqueous mixtures. Our data agree quite well with the report- ed values of Vmax for compound 1 in pure solvent^.^ Values Following this argument, the electron-donating abilities of ,!3. for the electron-pair donation tendency of the medium, water and the small aliphatic alcohols methanol, ethanol and propan-2-01 do not differ appreciably and are all greater than that of acetone or acetonitrile.This is apparent from the bathochromic shifts observed for the longest wavelength band of 1 in acetone or acetonitrile as a hydroxylic solvent is added. Nevertheless, preferential solvation of 1 by the better donor co-solvent is not always observed, as seen in Fig. l(cl, (4,Fig. 2(b), (c) and (4.Since direct solute-solvent hydrogen- bonding interactions may be excluded from our systems, the observed deviations and the progressive ‘take-over’ of the solvation shell by the hydroxylic component, as one changes from propan-2-01 to ethanol, methanol and water, must reflect indirect solvent interactions. The analysis of these have been reported previously for various aqueous mix- ture~.~.’.’ Migron and Marcus employed compound 1 as a p indica-tor and their data for water-acetonitrile mixtures’ agree quite well with our results, if the relationship Vmax = 18.76 -2.7938, derived by the same authors,’ is used to convert 3,,, into p values.Thus, we obtained a value of 16.86 x cm ‘ for the absorption wavenumber of 1 in water, a value very similar to the corresponding values in methanol (16.89 x lW3), ethanol (16.87 x and propan-2-01 (17.00 x lop3 cm-I). This corresponds to @ = 0.68, essen- tially the same as that reported by the authors (0.67) in their study of water-acetonitrile mixtures. Krygowski et al. studied various aqueous binary mixtures, using as solvatochromic probes 4-nitroaniline and N,N-diethyl-4-nitroaniline,' and arrived at results which are at variance with our observations.This may be due to the use of indicators which are not purely p-sen~itive.~ In fact, their results yield a very low electron-donating ability value for water (BKTor #I = 0.19), much smaller than that for methanol (0.62), ethanol (0.77) or propan-2-01 (0.88). These values were essentially the same as those reported previously by Taft and co-workers,14 who listed them between brackets as 'less certain'. Accordingly, in all aqueous mixtures studied by the authors, the #I indicators were prefentially solvated by the organic co-solvent, an observation which departs from our results with dye 1. In conclusion, the present study brings to light the ambi- guity of the electron-donating ability concept, when applied to hydroxylic solvents.Spectroscopic measurements of the pure &sensitive dye 1 yield similar values of electron-donating ability for water and small aliphatic alcohols. This observation cannot explain their different behaviour in sol-vating compound 1 in binary mixtures. Indirect hydrogen- bonding effects explain the relative order of effectiveness of the hydroxylic co-solvent in the preferential solvation of the dye. This effectiveness increases in the order Pr'OH < EtOH < MeOH < HzO, and parallels the strength of these compounds as hydrogen-bonding donor solvents. Grants by the Brazilian Conselho Nacional de Pesquisa Cientifica e Tecnologica (CNPq) are gratefully acknowledged.J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 References 1 J. G. Dawber, J. Ward and R. A. Williams, J. Chem. SOC., Faraday Trans. I, 1988,84,713. 2 (a) P. Chatterjee and S. Bagchi, J. Chem. SOC., Faraday Trans., 1991,87,587; (b)1992,88,1675. 3 E. Bosch and M. Roses, J. Chem. SOC., Faraday Trans., 1992,88, 3541. 4 C. Lerfand P. Suppan, J. Chem. SOC., Faraday Trans., 1992,88, 963. 5 E. Dutkiewicz, A. Jakubowska and M. Dutkiewicz, Spectrochim. Acta, Part A, 1992,48,1409. Y. Fukuda and K. Sone, Bull. Chem. SOC. Jpn., 1972,45,465. Y. Migron and Y. Marcus, J. Phys. Org. Chem., 1991,4,310. Y. Migron and Y. Marcus, J. Phys. Chem., 1991,95,400. Y. Migron and Y. Marcus, J. Chem. SOC., Faraday Trans., 1991, 87, 1339. 10 D. Bourdin, D. Lavabre, J. P. Beteille, G. Levy and J. C. Micheau, Bull. Chem. SOC.Jpn., 1990,63,2985. 11 Y. Marcus, J. Solution Chem., 1984,13,599. 12 M. C. Rezende, C. Machado, S. P. Zanotto and M. Scremin, J. Phys. Org. Chem., 1993,6,637. 13 T. M. Krygowski, P. W. Krona, U. Zielkowska and C. Rei- chardt, Tetrahedron, 1985,41,4519. 14 M. J. Kamlet, J. L. M. Abboud, M. H. Abraham and R.W. Taft, J. Org. Chem., 1983,48,2877. Paper 3/06092F; Received 12th October, 1993