4111
SUBSTITUENT EIFFECTSON THE TAUTOMERISM OF SCHIFF BASES
Substituent, Effects on the Tautomerism of Schiff Bases
by John W. Ledbetter, Jr. Department of Chemistry, Medical College of South Carolina, Charleston, South Carolina $9401 (Received April 29, 1968)
A large number of substituted N-o-hydroxybenzylideneanilineswere prepared for studying the effects of substituents on the enol imine-keto enamine tautomerism. A wide range of electron-withdrawingand -donating groups were used to vary the imine nitrogen basicity and the hydroxyl acidity. It was found by using ultraviolet and visible spectra that the tautomerism increases with both the imine nitrogen basicity and the hydroxyl acidity. N-2-Hydroxy-5-cyanobenzylidene-2-hydroxyanilinewas found to exist largely as the keto structure in 2,2,2-trifluoroethanol.
Introduction I n earlier evidence from ultraviolet and visible spectra was obtained for the existence of an enol imine-keto enamine tautomerism of o-hydroxybenzylidene anils in solution. The tautomerism is thought to occur via an intramolecular (or intermolecular in protonic solvents) hydrogen transfer to the imine nitrogen. The tautomerism should therefore increase with
OH R3
I11
References to the literature on this subject may be found in the earlier publications. Additional work has been done recently by Cohena and nTagyj4who cited evidence for the tautomerism and its dependence on solvent polarity and the imine nitrogen basicity.
Experimental Section
the imine nitrogen basicity. Evidence for this, obtained from a few compounds, was cited earlier. I n addition, if the tautomerism occurs with the proton transfer from the hydroxyl, then the tautomerism should also inerease with the hydroxyl acidity. The work reported herein is a thorough study of the effects of substituents on the tautomerism. Electrondonating and -withdrawing groups were placed at RI and Rz to achieve a wide range of imine basicity and hy-
I
I1
droxyl acidity, respectively. The extent of tautomerism, as measured by the molar absorptivity of a band in the visible spectrum just above 400 mp, is then plotted against the Hammett substituent constant, crP, and the pK, of the corresponding amine or phenol. The second structure above utilizes the o-hydroxy in the aniline ring. It was shown earlier that this structure exhibited stronger tautomerism due to this second hydroxyl. It was therefore desirable to place substituents at Rs in an effort to obtain compounds which tautomerize to a higher degree.
The anils were prepared by the condensation of the appropriate aldehyde and amine in 95% ethanol. Those salicyladehydes not commercially available (ie., 5-methoxy-, 5-methyl-, &phenyl-, 5-fluoro-, 5iodo-, 5-carbethoxy-, and 5-cyano-) were prepared by the Duff reaction. After several recrystallizations all carbon and hydrogen analyses were in agreement with theory. N-2-Hydroxy-5-bromobenzylidene-2-hydroxyaniline was obtained from the Aldrich Chemical Go. and was recrystallized. The spectra were recorded with a Beckman DK-2A ratio recording spectrophotometer using 1.00-cm quartz cells. The instrument was calibrated with a holmium oxide filter. The solvents were of spectral quality and and 1-00 X solution concentrations were 5.00 X M for the weak absorbances.
Results and Discussion Table I lists the spectral data for structures of type I. A wide range of electron-withdrawing and -donating groups are substituted at R1 in order to vary the basicity of the imine nitrogen. It appears from the table that the molar absorptivity of the band over 400 mH, (1) J. W. Ledbetter, J . Phijs. Chem., 70, 2245 (1966). (2) J. W.Ledbetter, ibid., 71, 2351 (1967). (3) M. D. Cohen and S. Flavian, J . Chem. SOC.,B , 321 (1967). (4) P. Nagy, Magy. R e m . Folyoirat, 72, 108 (1966). (5) J. C. Duff, J . Chem. Soc., 547 (1941).
Volume 72,Number 12 November 1068
JOHN W. LEDBETTER, JR.
4112 Table I : Spectral Data for Structure I in Ethanol
430 s (280) 435 s (180) 437 (171) 460 s (153) 433 (126) 458 s (70) 460 s (68) 455 s (66) 445 s (64) 443 s (60) 440 s (60) 475 s (48)
351 (18,900) 350 (17,700) 342 (13,700) 348 (19,100) 338 (12,300) 346 (14,900) 344 (15,300) 346 (14,300) 342 (13,900) 342 (12,800) 339 (11,200) 354 (17,000)
326 s (15,000) 326s (14,700) 319 (13,000) 330s (17,900) 316 (10,500) 322 (15,700) 323 (14,700) 320 (14,600) 319 (13,500) 319 (12,300) 315 (10,500) 323 (17,000)
which has been assigned to the quinoid tautomer,’ generally increases with the electron-donating ability of the substituent. As indices of this ability of the substituent, the Hammett substituent constant,6 up, and the pK, of the corresponding substituted aniline’ are chosen. The log E of the band is then plotted against these indices in Figures 1 and 2, respectively. The estimated uncertainty in up is included in ’the figure. Two things are immediately evident. The first is that the band intensity increases with the electrondonating ability of the substituent. If the assumption is made that the band intensity is primarily proportional to the concentration of the quinoid tautomer, it can be said that the tautomerism increases with the imine basicity. The validity of this assumption can be found on two points. First, the band originates in the quinoid portion of the molecule, which is not completely conjugated but is cross-conjugated with overlap occurring at the nitrogen. Therefore, any variation in RI will not affect the band intensity as much as in the completely conjugated system. Second, the effect on the completely conjugated system can be ascertained by observations of the second band, which is predomi-
2’6r 2.4
2.2 W
-0” 2.0
307s (10,100) 206s (10,400) 305 s (11,000) 310s (14,400) 300 (10,900) 307 (16,300) 310 s (13,100) 304 s (15,600) 306 s (12,300) 306s (11,200) 301 (9400)
294 (8900) 292 (9400)
269 (10,500) 269 (11,300) 269 (12,300) 273 (16,900) 269 (13,700) 283 (16,200) 271 (14,000) 278 (16,800) 272 (14,000) 271 (13,100) 269 (12,000) 252 (8900)
229 (21,900) 225 (24,600) 227 (20,800) 233 (15,300) 227 (21,400) 231 (19,900) 237 (15,700) 230 (20,000) 230 (19,300) 228s (17,100)
2.6-
2.4
-
2.2
-
e
W
-g 2 . 0 1.8
-
I
I
I
1
2
3
pKa
I
I
I
I
4
5
6
7
Figure 2. Correlation of the intensity of the “quinoid” band of structure I with ionization constants of the corresponding substituted anilines.
nately the T --t T* transition of the benzenoid structure.s-10 The only possible trend between the intensity of this band and R1 is that of increasing intensity with increasing both donating and withdrawing abilities. However, the halogens must be excepted. This possible trend is certainly nothing like the increasing intensity of the first band with the electrondonating ability at R1. With structure I1 it can be seen in Table I1 that there is no obvious trend in the secondband intensity with Rz. Therefore, the trend in the first-band intensity is primarily due to tautomer concentration. The second observation from Figures 1 and 2 is that the halogens give the reverse relationship, in that the band intensity decreases with the electron-donating ability of the halogens. A further check on the be-
1.8
1.6
I
-.6
I
-.4
I
-.e
I
0
I
I
I
I
I
.2
.4
.6
.8
1.0
GP
Figure 1. Correlation of the intensity of the “quinoid” band of structure I with substituent constants. The Journal of Physical Chemistry
(8) D. H. McDaniel and H. C. Brown, J . Org. Chem., 23, 420 (1958). (7) E. A. Braude and F. C. Nachod, “Determination of Organic Structures by Physical Methods,” Vol. I, Academic Press, New York, N. Y., 1955, p 567. (8) H. H. Jaff6, S. H. Yeh, and R. W. Gardner. J. MOL Spectrosc., 2, 120 (1958). (9) W. F. Smith, Tetrahedron, 19, 445 (1963). (10) V. I. Minkin, et al., Opt. Spektrosk., 18, 328 (1965).
SUBSTITUENT EFFECTS ON
THE
TAUTOMERISM OF SCHIFF BASES
4113
-
Table I1 : Spectral Data for Structure I1 in Methanol
--------
Rs
CN NO2 CzHsOOC I
Amax, - - - -mp - - - )e(
Br CeH5 c1
H CHa CHaO F HO
havior of the halogens may be found by plotting the wavelength of the first two bands against u p as shown in Figures 3 and 4. As already pointed out, the first band is the long-wavelength band of the quinoid isomer which is not completely conjugated. Therefore, any variation at R1 will only indirectly affect the spectrum through its effect on the nitrogen atom. The amine group being electron donating, as the electron-withdrawing power of R1increases, a bathochromic shift in the band will occur.11 In this plot the halogens are in agreement with other substituents; i e . , A, is in the order I > Br >. C1> F. Using the second band, a different type of plot is obtained in Figure 4. I n this molecule both electrondonating and -withdrawing groups at R1 should produce a bathochromic shift.12 Here again, the halogens behave as expected with iodine having the longer wavelength. Since the effects of the halogens on wavelength appear to be normal, the inductive effects of the halogens evidently predominate above mesomeric effects in determining tjhe imine nitrogen basicity. On this basis, the basicity would be in the order I > Br > C1> F. It might be pointed out that for para-substituted aniline the order of nitrogen basicity is F > Br > C1, while for 3-substituted pyridines the order is I > F >
470
460
a €-450 x
-
430
-
420’
228 (21,700) 228s (13,700) 230 (17,700) 238 (17,400) 239 (16,600)
356
352
-
a E.348
-
344
-
340
-
336’
-.b -.;
b
;,
.k
.6
.Ll
1.b
GP
Figure 4. Correlation of the wavelength of the T + T* transition of the benzenoid structure I with substituent constants. 4 . 0 ~
3.8
:::I
-wg3.4 3*6[
-!6
I
I
I
I
I
I
I
I
-.4
-.2
0
.2 GP
.4
.6
.8
1.0
Figure 5. Correlation of the intensity of the “quinoid” band of structure I1 with substituent constants.
-
-
221 s (14,400) 221 (17,400)
248 s (26,400) 234 (25,700) 230 s (22,100)
2.8
-
440
244 (27,500)
255 s (24,000) 268 s (7800) 257 (27,700) 267s (10,100) 265 (9880) 261 (34,500) 267 (9820) 269 (8600) 271 (10,800) 271 (9280) 266 (10,400) 270 (9100)
353 (9500) 294 (12,400) 348 (10,700) 360 (10,600) 358 (11,700) 352 (10,400) 358 (11,900) 349 (9800) 354 (12,000) 366 (10,400) 358 (12,500) 370 (9600)
436 (7300) 430 (6960) 435 (6900) 458 (2100) 455 (2100) 465 (1960) 455 (1920) 444 (1600) 459 (1460) 478 (1250) 455 (1140) 481 (695)
7
Br = CI.’ The order obtained here is I > Br > C1 = F, which best agrees with the conjugated imine in pyridine rather than aniline.
-.b
-:2
l !L
.h
.b
.6 .‘8
GP
Figure 3. Correlation of the wavelength of the “quinoid” band of structure I with substituent constants.
1.b
(11) C. N. R. Rao, et al., Proc. CoZZog. Spectrosc. Int., loth, Univ. Maryland, 1969,505 (1963). (12) H. H. Jaff6 and M. Orchin, “Theory and Applications of Ultraviolet Spectroscopy,” John Wiley & Sons, Inc., New York, N. Y., 1964, p 259. Volume 79, Number 19 November 1968
JOHNW. LEDBETTER, JR.
4114 Table 111: Spectral Data for Structure I11 in Methanol
425 s (4400) 427 (3140) 448 (2000) 444 (1600)
N0 2 C1 CaH6 H
348 (12,700) 355 (11,500) 357 (11,300) 349 (9800)
273 (14,500) 268 (12,700) 269 (32,000) 269 (8600)
221 (18,100) 230 (20,800) 221 s (27,800) 228s (13,700)
~~
Table IV : Spectral Data for N-2-Hydroxy-5-c~~anobenaylidene-2-hydroxyaniline in Solution
Cy clohexane-p-dioxane (50:50 vol %) Methanol 2,2,2-Trifluoroethanol
355 (12,600) 436 (7300) 428 (14,000)
241 (38,800)
353 (9800) 338 (7800)
255 s (24,000) 258 (27,000)
308 (7700)
244 (27,500) 243 (28,000)
4.0r
2.0-
I
7.0
I
I
I
8.0
9.0
10.0
I
11.0
4
I
300
h,mp
400
I
SO0
Figure 7. Spectra of N-2-hydroxy-5-cyanobenzylidene-2hydroxyaniline in: A, 2,2,2-trifluoroethanoI; B, methanol; C, 50:50 vol % cyclohexane-p-dioxane.
pKo
Figure 6. Correlation of the intensity of the “quinoid” band of structure I1 with ionization constants of the corresponding substituted phenols.
TabIe I1 tabulates the spectral data of structures of type 11. In these compounds the 400-mp band is much more intense. The log E of this band is plotted against up, Figure 5, and the pK, (as measured in water) of the corresponding substituted phenol,’ Figure 6. Both plots show that the tautomerism increases with the hydroxyl acidity. Also, it is observed that the halogens conform with other substituents. The hydroxyl acidity is in the order I > Br > C1 > F, which is the same for the substituted phenols. It is seen, therefore, that increasing both the imine basicity and the hydroxyl acidity increases the tautomerism. This would be expected, considering that a more basic imine nitrogen would favor enamine formation and the more acidic hydroxyl would favor keto formation. It was pointed out earlier and seen here that structure I1 exhibits more tautomerism than structure I. This is possibly due to additional hydrogen bonding by the second hydroxyl. It was expected than an increase in The Journal of Physical Chemistry
this hydroxyl acidity would enhance the tautomerism. This is confirmed in Table 111. The electron-withdrawing groups at R3enhance the tautomerism. In an effort to increase the tautomerism further, a compound was prepared having the combined structures of I1 and I11 with a nitro group a t RPand R3. The result was a bathochromic shift with the longwavelength-band definition being lost. However, the molar absorptivity at 42.5 mp is 5400, which is between those of the singly substituted nitro compounds. Therefore, this procedure does not increase the tautomerism. This type of leveling effect by like groups opposing each other has been observed earlier.13 A simple solvent study was done to achieve the highest degree of tautomerism. X-2-Hydroxy-5-cyanobenzylidene-2-hydroxyaniline was chosen from those studied, for it is the structure exhibiting the highest degree of tautomerism in methanol. Spectra of this compound were obtained in cyclohexane-p-dioxane and 2,2,2-trifluoroethano1. This alcohol was chosen because it produces the highest degree of tautomerism of (13) R. J. Argauer and C. E. White, Anal. Chnn., 36, 2141 (1964).
EPRSTUDIES 01F COMPLEX I O N FORMATION
4115
compound is present largely as the quinoid isomer in 2,2,2-trifluoroethanol. Acknowledgments. This research was supported in part by Kational Institutes of Health General Research Support Grant FR 05420.
those solvents studied, except the acids. The acids are not advantageous because of small amounts of water present which can cause hydrolysis. The results are listed in Table IV and the spectra are shown in Figure 7. It is concluded from the spectra that the
Electron Paramagnetic Resonance Studies of Complex Ion Formation between Mm2+ and F-, C1-, Br-, I-, or SO,"-' by Douglas C. McCain and Rollie J. Myers Department of Chemistry of the University of California and the Inorganic Materials Research Division of the Lawrence Radiation Laboratory, Berkeley, California 94720 (Received April 20, 1968)
While Mn2+in aqueous solution forms only rather weak complexes with simple anions, its epr line width is affected by the presence of such anions. The line width is a measure of the electron spin relaxation time, and this time is influenced by the rates of reaction and the equilibrium of complex ions. Hayes and Xyers proposed a theory which assumes a two-step process for complex ion formation, and they related the epr line width to the rate constants for each step. We have measured the line width for Mn2+in aqueous solutions containing either F-, C1-, Br-, I-, or S042-, and our results are in good agreement with this two-step theory. These data are used to evaluate certain rate and equilibrium constants at 22 and 160". Comparison is also made to similar results that have been obtained by other methods.
I. Introduction Ion-pair forination is to be expected in ionic solutions. In aqueous solutions the cations are heavily solvated so that the anions have to displace water molecules in order l o form ion pairs. One possible ion pair is a complex in which the cation largely retains its full solvation sphere, although perhaps distorted by the anion. This kind of an ion pair has a large interionic distance and only a small entropy increase due to solvation changes. At the other extreme, ion pairs can be formed in whilch the solvation sphere is breached and large entropy changes mill occur. ;\lost spectroscopic methods are not sensitive t o outer-sphere coordination, but the epr spectra of highspin, S-state transition metal ions are affected by such coordination. The line widths of these ions are due to a time-dependent zero-field splitting which is averaged in solution. While the zero-field splitting may be only slightly increased by the formation of an outer-sphere complex, its correlation or averaging time is greatly increased over that of the simple hydrated ion. In addition, the inner-sphere coordination can cause a large increase in the apparent zero-field splitting and quite broad epr lines. In a previous publication2 it was shown that X4n2+in aqueous solution had an epr spectrum which was broad-
ened by C1- and Soh2-. A two-step coordination model was proposed to explain this broadening, and sets of equations were established which depended upon the rates of formation and dissociation of an outer- and inner-sphere complex. While all these rates could not be evaluated at that time, a consistent picture was established and a few specific rates were determined from the epr data. In this paper we have obtained data for the broadening of ?\In2+by C1-, Br-, I-, F-, and S042-. For C1and Sod2- greater attention was paid to ionic strength than was previously donee2 We were also able to justify further the pattern of rate constants previously assumed and to establish firmly certain rate and equilibrium constants for Mn2+in aqueous solution.
11. Experimental Section As previously done,2 we measured the firstJderivahyperfine tive epr line width ( A H ) for the M I = component of X W f in aqueous solution. This was done both near room temperature and in sealed tubes to above 200". Samples were prepared from ordinary, laboratory distilled water and weighed amounts of (1) Presented in part at the American Chemical Society Summer Symposium, Buffalo, N. Y., June 1965. (2) R. G. Hayes and R. J. Myers, J. Chem. Phys., 40, 877 (1964).
Volume 72. Number 12 November 1968