NORMAN KULEVSKY AND LYMAN LEWIS
3582
Thermodynamics of Hydrogen Bond Formation between Phenol iazine N-Oxides by Norman Kulevsky" and Lyman Lewis Chemistry Department, University o j North Dakota, Grand Forks, North Dakota
68201
(Received May 8,1972)
Equilibrium constants and A H o values for the interactions of phenol with azine N-oxides in CHClz indicate that the relative order of basicity is pyridine N-oxide > pyrimidine N-oxide > pyridazine N-oxide > pyrazine Ai-oxide. With the exception of pyrazine N-oxide, the order toward phenol parallels that found toward It. Those data are interpreted on the basis of oxygen being the donor site and discussed in terms of the forces involved in forming the hydrogen bond. Pyridine N-oxide is found to be a stronger base than pyridine and this is expIained in terms of the solvent effect of CHzClz on the equiIibrium.
The relative electxon-donor abilities of the mono N oxides of some heterocyclic diazines toward molecular iodine have rocent1:y been studied.'V2 For these compounds, arguments based upon frequency shifts in the infrared and the trend of equilibrium constants for formation of the iodine complexes lead to the conclusion that it is oxygen and not the pyridine like nitrogen that acts as the donor site toward iodine. I n this paper we shall discuss an investigation of the stability of the Hbonded cornpiexes of phenol with these same bases which was undertaken in order to see what changes would occur in the relative basic strengths when an acid is used whose electronic and steric requirements are different from iodrne. The effect of a hydrogenbonding solvcnt~on the relative complexing ability of pyridine anti its oxide is also discussed.
Experimental ~ e ~ ~ i ~ r i The donor species were prepared and purified by procedures given in a previous paper.2 Reagent grade phenol was fractionally crystallized and vacuum dis1,illed Spectral grade solvents were stored over molecular sieve before use. All solutions were prepared in a drybox. The concentrations were obtained by weight. The assoc;iaf,ion constants were obtained by spectroscopic techniques previously used by other worker^.^ The optical densities were obtained at the maximum of the first overtcme of tile free 8-33 stretching mode in the near-infrared nl; 1.4 m. A Cary NIodel 14 spectrophotometer eciuipped. with a thermostated cell holder was used to 3btain these measurements. The cells used mere matched silica with IO-cm path length. solrption coefficient was obtained by measuring optdcal densities of from six to ten solutions at six different temperatures ranging from 11.0 to 31.9". The concentration of phenol in CHzCIz ranged from X M . Plots of optical density against concentration at each temperature were linear with eero intercept, indicating that there is no selfThe Journal of Plyaicol Chemistry, Vol. 76,N o . 25, 1973
association over this concentration range. The slopes of the plots were taken as the absorption coefficient. To obtain the equilibrium constants, solutions of phenol within this range were prepared with several different concentrations of base added. For each of the bases, the concentration ranges were 16 X low3to 42 X M . Using the previously determined absorption coefficient, the concentration of free phenol in the equilibrium mixture was calculated. Then using the stoichiometric relations the concentrations of free base, complex, and the equilibrium constant were calculated.
Results and Discussion Equilibrium constants as a function of temperature for the formation of the H-bonded complexes of phenol with pyridine and the N-oxides are given in Table I. Each of the values is the average of values obtained from measurements on from four to seven different solutions in which the concentration of base mixed with a constant concent,ration of phenol was varied by a factor of 2.5. The error limits, which are standard deviations, are ca. 2y0,although a few are higher. No discernable trends in K values could be observed over the range of base concentrations used, indicating that the stoichiometry for all the complexes is I :1 even in those cases where, as for the diazine N-oxides, the base has two potential donor sites. The values of AH" obtained from the plot of In R us. 1/1' are given in Table 11. The values obtained for the pyridine-phenol interaction are also included in these tables and the agreement between them and the value of K25 given by Rubin and Panson4is quite reasonable. (1) N.Kulevsky and R . G. Severson, Jr., Spectrochim. Acta, Part A , 26, 2227 (1970).
(2) N. Kulevsky and R. G. Severson, Jr., J . Phys. Chem., 75, 2504 (1971). (3) D. L. Powell and R. West, Spectrochim. Acta, 20, 983 (1964). (4) J. Rubin and G. S. Panaon, J . Phys. Chem., 69, 3089 (1969.
___y_e______l___r_l-_I
~~~~~
I: Association Constants (M-1) as a Function of Temperature for Phenol Complexes in CH2c112 "C
Pyridine
Pyridine N-oxide
11.0 15.0
28.1 k 0 . 2 23.1 f 0 . 6 22.3 "& 0 . 6 19.4 i 0 . 6 15.1 =k 0.4 13.4 4 0.2
99.1 0.8 85.9 f 1 . 5 '92.1 3z 2 . 2 6 3 . 3 =k 3 . 3 54.9 f 1.1 49.3 f 0.6
T,
59.7 24.0 '28.0 31.9
*
Table 11: Enthalpy of Association for Phenol. Complexes in CI-l&lz
-AHQ, kcal/rnol
Pyridine Pyridine 1 ~ T - 5 X l d e Pyridnzine N-oxide Pyriniid:ne N-oxide Pyraeinc: Ai-oxide
5.9 k 0.2 5 . 8 f0.1 5.0 zk 0 . 3 4.9 f 0.3 4.2 f 0.3
From the data given in the two tables, the relative base strength i s pyridine N-oxide > pyrimidine N oxide > pyridazine N-oxide > pyrazine N-oxide. From the values of M this order is unequivocal, although the AH" vdues are, in some cases, very close. 'This order is similar LO the one found when iodine is the reference a d , except for the position of pyridazine Noxide which forms the weakest iodine complex. If the values of AG" for the iodine and phenol complexes, exclusive of pyridazirre N-oxide, are plotted against one :inother a linear relationship caD bc observed. Since %hereare only threc points in this correlation a statistical analysis of this relationship does not seem valuable to report hme. This relationship would indicate that for pyrimidine and pyrazine N-oxides it is oxygen which 1s the donor site tom ard phenol as well as toward iodine. Since the equilibrium constants for iodine interactions correlate wrth Hammet r constants, the failure of this correlation fox the phenol pyridaxine N-oxide indicates that ~f oxygen 16 the donor site, the r value for the ortho position gilicn by Katritzky and Swinbourne5 is not valid for che phenol interactions. The reason for this observation could be the differences in the types of Interaction the base undergoes with the two acids. Iodine, a nonpoiar species, forms a charge-transfer complex whose st:bbiiity is to a large extent determined by the amount ol overlap between donor and acceptor orbitals and the electron density in the donor orbital. 'Thus, for the iodine complexes the trend in stability
Pyridssine N-oxide
Pyrimidine N-oxide
Pyrazine M-mide
13.8 f 0 . 1 12.4 f 0 . 3 11.4 Sk 0.2 10.0 I Q . 2 8.2 0.3 7 . 7 rk 0 . 3
19.1 f 0 . 3 16.9 f 0 . 3 14.'7 f 0.4 13.0 : k 0.2 11.6 i 0 . 5 10.4 f 0 . 2
10.7 A 0.3 9 . 6 rk 0.2 9.0 0 1
*
-*
~
8 . 0 & 0.4 6.9 f 0.5 6 . 6 rk 0 . 2
follows the ?r electron density at the oxygen atom previously calculated by MO theory.6 On the other hand, the hydrogen-bonding strength is determined to a large extent by dipole-dipole forces. If dipole-dipole forces are the determining Factor then for and pyrazine N-oxide where the d 5.21 19 and 1.66 D,6 respectively, plridaaine AT-oxide should have the stronger interaction with phenol. This would also imply that for pyridazine N-oxide the hydrogen of phenol is interacting with both the oxygenand the pyridine-like nitrogen, which waLs not stpparent for the iodine interaction. One anomaly is apparent if the data for pyridine and its oxide are compared. According to the values of K and AH" given in the tables, pyridine is less basic than its oxide. Romever, the reverse order of basicity is exhibited toward both aqueous acid7 and iodine* in CCI,. One possible reason Cor this reversal could be solvent effects. Rubin, et al.,4 found that association constants for phenol-pyridine varied agreat deai as the solvent was changed although there was no correlation of thesc constants with the dielectric constant of the solvent. Since the values of K mere the same in both CBdX and CHC13 and lower than the value in CCl, (by a Factor of 0.5) it is possible that the hydrogen bonding of c or CHCl, to the free base causes the large change in K values observed by them. Because in most cases pyridine N-oxide is a weaker base than pyridine it is possibic that hydrogen bonding of the solvent to 1 he free bases would not affect the equilibrium constant for the oxide as much as it does for pyridine. Therefore, the association with phenol in CH9612could appear larger for the oxide than for pyridine.. (5) A. R. Katritzky and F. J. Swinbourne, J . Chem.
&e.,
6707
(1965). (6) T. Kuhota and H. W-atanabs, Bull. Chem. 8oc. Jap., 36, 1093
(1963). (7) (a) A. Albert, R. Goldacre, and J. Phillips, J . Chem. Soc., 2240 (1948); (E) H. H. Jaffe and 6. 0. Doa.lrs, J . Amer. Chem. Soc., 79, 4441 (1955). (8) (a) W. McKiiiney and A. I. Popov, ibid., 91, 5215 (1969), and references therein; (b) T. Kubota, ibid., 87, 458 (1965).
The Journal of Physical Chemistra, Vol. 76, 'j70. 23, 1978