T H E
J O U R N A L
O F
PITYSICAL CHEMISTRY Registered in U . S . Patent O f i c e
@ Copyright, 1964, by the A m e r i c a n Chemical Society
VOLUME 68, NUMBER 8 AUGUST 14, 1964
Microwave Absorption and Molecular Structure in Liquids.
LVIII.
The Dielectric Relaxations, Infrared Spectra, and Intramolecular Hydrogen Bonding of 2,6-Dichloro-p-nitroaniline and Four Substituted
by Arthur A. A n t ~ n y Francis ,~ K. Fong, and Charles P. Smyth Frick Chemical Laboratory,~Princeton University, Princeton, .+mu Jersey
(Receiaed April 90, 1964)
The dielectric constants and losses of dilute benzene solutions of 2,6-dibromophenol, 2,6dibromo-p-nitrophenol, 2,6-diahloro-p-nitrophenol, 2,4-dibroniophenol, and 2,G-dichlorlop-nitroaniline have been measured at 20, 40, and 60" a t several microwave frequencies and used to calculate the dielectric relaxation times. The infrared spectra in the OH stretching region have been measured for solutions of 2,6-dibronio-p-nitrophenol, 2,6-dibroinophenol, 2,4-dibromophenol, and o-bromophenol. 2,6-Dibromo-p-nitrophenol and 2,6-dichloro-pnitrophenol behave as rigid molecules, but 2,6-dibromophenol can be treated in ternis of two relaxation times, one due to molecular rotation and the other due to OH group rotation. 2,4-Dibromophenal probably has more than two relaxation times. The dipole moment of this niolecule can be interpreted in terms of the two conformations in which the OH is coplanar with the benzene ring.
Previous work4 on the dielectric relaxation of chloroform in solution indicates that hydrogen bonding increases the dielectric relaxation time. Substitution of halogens in the ortho position in phenol should result in an intr.amolecular hydrogen bond and, therefore, an increased barrier to internal rotation. In order to confirm the presence of an intramolecular hydrogen bond and in order to ascertain the relative strength of this bond in the conipounds studied, infrared spectroscopic measurements have been made in the region of the OH stretching fundamental of the phenols. The
unexpectedly large band widths observed in benzene solution prompted some additional infrared studies, which are also reported in this paper. (1) This research was supported by the Office of Naval Research, the National Science Foundation, and the U. S. Army Research Office (Durham). Reproduction, translation, use, or disposal in whole or in part by or for the United States Government is permitted. (2) This paper represents part of the work submitted by A. A. Antony to the Graduate School of Princeton University in partial fulfillment of the requirement for the degree of Doctor of Philosophy. (3) Allied Chemical Fellow, 1961-1962; Woodrow Wilson Fellow, 1959-1960.
2035
A. A. ANTONY,F. K. FOKG, AND C. P. SMYTH
2036
Experimental Purification of Materials. Five compounds obtained from Eastman Kodak Co. were treated as follows: 2,6-dibromophenol was shaken with decolorizing charcoal, filtered, and then recrystallized from ether; 2,6dibromo-p-nitrophenol, 2,6-dichloro-p-nitrophenol, and 2,6-dichloro-p-nitroaniline were recrystallized from ether ; 2,4-dibromophenol was recrystallized from chloroform. The following melting points were obtained : 2,6-dibroniophenol, 54.8-55.3’ : 2,B-dibromo-pnitrophenol, 143-144’; 2,6-dichloro-p-nitrophenol, 122.5’, decomposition accompanying melting; 2,6dichloro-p-nitroaniline, 193.5-194.4 ’; 2,4-dibromophenol, 38-40’. 0-Bromophenol from hlatheson Colenian and Bell was used without further purification. Reagent grade benzene and carbon tetrachloride from Baker and Adamson were also used without further purification. Cyclohexane from Matheson Coleman and Bell was fractionally distilled. Apparatus. Dielectric constants and losses were measured by methods previously d e ~ c r i b e d . ~ -Infra~ red spectra were measured on a Perkin-Elmer 421 grating spectrometer. Results Slopes ao, a’, aD, and a” were obtained by plotting the static dielectric constants eo, the high frequency dielectric constants e’, the refractive indices for the sodium D-line, and the losses e” of the solutions against the concentrations of the polar solutes. Values of the distribution parameter a and the most probable relaxation time T~ were obtained from Cole-Cole plots of a’ against In Table I the experimental values for aol a’, and a” are listed. I n Table I1 the values for a , uD, a,, and T O are listed. Because of the very low solubility of 2,B-dichloro-pnitroaniline in benzene, measurements were made only on the saturated solution and r 0 was obtained from the frequency at maximuni loss froin a plot of 68’’ - EB’’ against log frequency, where E S ” is the loss of the solution and eB” is the loss of benzene. Table I11 lists the OH stretching fundamental frequencies and half-band widths measured in the infrared region. Some of the values listed are taken from the literature. The frequency associated with the nonhydrogen-bonded OH is listed under the column “Free” and that associated with the hydrogen-bonded OH under the column “Bonded.” The band width at half-height is given under Avl/,. Discussion All of the compounds discussed in this paper may be expected to show intramolecular hydrogen bonding. The Journal of Physical Chemistry
Table I : Slopes for the Dependence of Dielectric Constant and Loss on Concentration in Benzene Solution Wave length, om.
1.25 3.22 10.0 575 m. 1.25 3 22 10.0 25.0 50.0 575 m. 3.22 10.0 25.0 50.0 575 m. 1.25 3.22 10.0 575 m.
200
60‘
40‘ a”
a’
Q’
(1’’
2,6-Dibromophenol (0-0.0195) 1.9 1.05 2.2 1.20 3.2 1.57 3.5 1.49 5.0 1.53 4.9 1.1C 5.9 5.5
2,6-Dibromo-p-nitrophenol 1.6 1.75 2.1 2.69 3 . 4 3.7 8.9 7.25 11.1 15.8 5.04 14.6 16.7 2 . 9 8 15.1 17.66 16.00
a’
Q“
2.4 3.5 4 9 5.2
(0-0,0125) 1.88 2.0 3.9 4.6 6.2 12.0 3 . 7 5 13.5 2 . 1 0 13.9 14.32
1.33 1.43 0.87
2.23 4.4 4.44 2.55 1.4
2,4-Dibromophenol (0-0,0125) 0.72 1.8 1.8 0.66 0.63 2.5 2.35 0.43 2.75 0.28 2.05 0.35 3.05 0.20 2.8 0,l8 3.1 2.85 2,6-Dichloro-p-nitrophenol 2.68 2.63 1 . 4 3.71 3.37 2.98 7.31 6.06 8.54 15.96 13.96
(0-0.0158) 1.70 2.58 3.28 4.05 5.21 9.22 11.84
1.70 3.67 3.98
2,6-Dichloro-p-nitroaniline ea”
1.25 3.22 10.0
- fB”
0,0122 0.0220 0.0327
6s‘’
- rg”
0.0164 0,0235 0.0268
68’’
- sg”
0,0192 0,0246 0.0198
If halogens are substituted in both the 2- and the 6positions, the OH or S H 2group can be intramolecularly bonded to either position and, in an alternating electric field, might acquire enough energy to break the bond to one halogen, jump over the barrier to internal rotation, and form a hydrogen bond to the other halogen. This behavior would manifest itself by a second intramolecular relaxation time, in addition to that associated with the over-all molecular rotation. It would (4) A. A. Antony and C. P. Smyth, J . A m . Chem. SOC., 86, 152 (1964). (5) W. M. Heston, Jr., A. D. Franklin, E. J. Hennelly, and C. P. Smyth, i b i d . , 72, 3443 (1950). (6) D. A. Pitt and C. P. Smyth, 1. Phys. Chem., 63, 582 (1959). (7) L. M.Kushner and C. P. Smyth, J . Am. Chem. Soc., 71, 1401 (1949).
(8) K . S. Cole and R. H. Cole, J . Chem. Phys., 9, 341 (1941). (9) C. P. Smyth, “Dielectric Behavior and Structure,” McGraw-Hill Book Co., New York, N. Y., 1955, Chapter 11.
2037
NIICROWAVEABSORPTION AND MOLECULAR STRUCTURE m LIQUIDS
Table 11: Slopes, a ~ for , the Dependence of the Square of the Refractive Index on Concentration, Infinite Frequency Intercepts am,Relaxation Times 70,Distribution Parameters CY, and Dipole h'hments p TO r
t! OC.
aD
a m
10-12 Bee.
P X 1018
a
2,6-Dibromophenol 1.50 23 18 1.51 1.50 14
0.13 0.12 0.12
20 40 60
0.6
20 40 60
1.0
20 40
0.36
2,4-Dibromophenol 0.85 23 0.83 22
20 40 60
0.45
2,6-Dichloro-p-nitrophenol 2.40 69 0 0 2.35 49 2.30 32 0
2,6-Dibromo-p-nitrophenol 1.68 56 0.03 1.68 44 0.04 1 68 28 0.06
0.24 0.28
3.32
1.35
3.04
2,6-Dichloro-p-nitroaniline 61 40 23
20 40 60
-
associated with internal group rotation. The infrared measurements in every case indicate the presence of intramolecular hydrogen bonding. The solutions were dilute enough so that no intermolecular bonding between solute molecules was expected, and no infrared peak characteristic of such a bond was found. 2,6-Dibomophenol. It was possible to analyze the data for this compound into two relaxation times by the double arc method.lO The results of this analysis are listed in Table IV. A dipole moment of 2.26 D. is calculated for this molecule from bond moments." A value of 1.46 D. was used for the m-dibromo contribution. The OH bond moment was assumed to be 1.51 D. and the CO moment 0.99 D. A COH angle of 115" was assumed. The moment w along the direction of the CO bond is 1.82 D. and the moment p z perpendicular to this is 1.36. If Debye behavior is assumed for both the over-all molecular and the intramolecular relaxation processes, C1/Cz = ( p l / p L p2,) which gives C:! as 0.36. This is close to the observed value at 20" in Table IV. The temperature dependence of Cz may be due to a greater rate of increasek of the relaxation of the hydrogen-bonded group with rising temperature, although the actual values of Cz are higher than expected.
Table IV: Relaxation Times (10-l2sec.) by the Double Arc Method for 2,6-Dibromophenol Table I11 : OH Fundamental Stretching Frequencies and Half-Band Widths
1,
Free
Bonded Y,
AVY8,
Y,
Solvent
em.-'
.em.-1
2,6-Dibromo-pnitrophenol
CEH~Z cc1, C~HE
3496 3492 3473
16 31 67
2,6-Dibromophenol
CCl, C5HE
3514 3499
36 55
2,4-Dibromophenol
CClP CBHE
3528 3512
23 62
3597
o-Bromophenol
cc14a CeHe
3529 3515
19 50
3604
Phenol
CC12 06HBb
Solute
"c.
20 40 60
71
39.2 35.2 26.5
71
12.7 10.3 10.0
cz 0.40 0.54 0.57
cm.-'
3614 3558
a I. Brown, G. Eglinton, and M. Martin-Smith, Spectrochim. Acta, 18,1593 (1962). * G. M. Huggins and G. C. Pimentel, J . Phys. Chem., 60, 1615 (1956).
thus be feasible to analyze the most probably relaxation time T~ into a relaxation time rl due to over-all molecular rotation and another relaxation time r2
The relaxation time associated with hydroxyl group rotation is larger than the value reported for compounds in which intramolecular bonding is absent. Some values which have been reported are 3.2 X see. for 2-iiaphtho1,'O 3.5 X 10-l2 see. for 2,6-dimethylphen01,'~g~~ and 8.4 X 10-l2 see. for 2,4,6-trichlorophen01.~3 A much higher barrier to internal rotation in the 2,4,6-triehlorophenol molecule is the result of the intramolecular hydrogen bond. The OH group rotation relaxation time is slightly higher for the bromo-substituted compound (Table IV) than for the chloro-substituted compound. Conclusions drawn (10) F. K. Fong and C. P. Smyth, J . Phys. Chem., 6 7 , 226 (1963). (11) C. P. Smyth, "Dielectric Behavior and Structure," McGrawHill Book Co., New York, N. Y.,1955, Chapter X. (12) F. K. Fong and C. P. Smyth, J . Am. Chem. Soc., 8 5 , 1565 (1963). (13) A. Aihsra and M, Davies, J . Colloid Sci., 1 1 , 671 (1956).
Volume 68,Number 8
August, 1964.
A. A. ATTOSY,F. K. FOKG, AND C. P. SMYTH
2038
from infrared measurenients concerning the relative strengths of the intramolecular hydrogen bonding of the OH group to the two halogens in these compounds are ambiguous. 14--20 The dielectric studies suggest that the hydrogen bonding is stronger in the case of the bromo-substituted compound. However, the relaxation times emociated with group rotation in the two compounds were not calculated by the same method, and the difference between the two values is, perhaps, too small to permit conclusions as to the relative strengths of the two intramolecular hydrogen bonds. 2,4-Dibromophenol. Because of the small polarization of this compound, no attempts were made to analyze the data for it into more than one relaxation time. The large value of the distribution parameter indicates that more than one mechanism is contributing to dielectric relaxation. In addition to the relaxation time associated with molecular rotation, one might expect two relaxation tinies associated with group rotation, one involving breaking the intramolecular hydrogen bond and rotation of the OH group to the trans position, and the other involving rotation of the OH from the trans position to the cis position. The cistrans rotation would be expected to have a larger relaxation time than the traizs-cis rotation since the breaking of a hydrogen bond is necessary for the former.
respectively. Approximately 20% of the molecules are in the trans form. This is very close to the result obtainedz2for o-bromophenol. 2,B-Dibromo-p-nitrophenol. Attempts were made to obtain two relaxation times by the double arc methodlo used for the 2,6-dibromophenol. It was impossible to obtain two relaxation times. The expected dipole moment” along the axis of symmetry is 2.16 D., and the component perpendicular to this axis is 1.36 D. One would, therefore, expect a value of about 0.2 for Cz, assuming Debye behavior for the group relaxation. The virtual absence of a contribution to the relaxation tiine due to the OH group rotation, while a considerable contribution is observed in the absence of the nitro group, indicates that the carbon-oxygen bond is considerably stiffened by the presence of the nitro group in the para position. This is not unexpected. The nitro group is an electron-withdrawing group, and, in the para posit,ion, it aids the resonance interaction of the hydroxyl oxygen with the benzene ring, as evidenced by an increase of 0.70 in the dipole moment of the p-nitrophenol above the calculated value. The resultant double-bond character of the C-0 bond appears to be sufficient to hinder strongly rotation of the OH in the applied field, Because of the increased resonance int,eraction between the oxygen and benzene, a stronger intramolecular hydrogen bond is expected. That this is the case is indicated by the fact that the OH stretching absorption occurs a t a lower frequency for ” 0 . Q 2,6-dibromo-p-nitrophenol than for 2,6-dibromophenol. The increased hydrogen bond strength further hinders &Br the internal OH rotation. 2,6-Dichloro-p-nitrophenoE. The relaxation time of cis trans 2,6-dichloro-p-nitrophenol is larger than that of 2,6The orientation polarization, obtained from the dibromo-p-nitrophenol, in spite of the fact that the bromine atoms are larger than the chlorine atoms. This Cole-Cole arc a t 20”, is 38.15 cc. This corresponds to a dipole moment of 1.35 D., according to eq. laz1 could conceivably arise from a greater lowering of relaxation time by OH rotation in the bromo compound, although the effect is too small to permit a separation into two relaxation times. The dipole
H\
~~
Using the bond moments given above, the dipole moment is calculated to be 0.37 D. with OH in the cis configuration and 2.64 D. with it in the trans configuration. The intermediate value is an indication that both species are present. Equation 2 can be used to estimate the fraction of molecules in the trans conformation.
P
= XPnans
+ (1 -
z)Pcts
(2)
P is the measured polarization, x: the fraction of molecules in the trans conformation, and P,,,,, and Pc$sthe calculated polarizations for the trans and cis forms,
T h e Journal of Physical Chemistry
(14) 6. Rossmy, W. Luttke, and R. Mecke, J . Chem. Phys., 2 1 , 1606 (1953). (15) A. W. Baker, J. Am. Chem. Soc., 80, 3598 (1958). (16) A. W. Baker and W. W.Kaeding, ibid., 81, 5904 (1959). (17) 0. Wulf, E. Jones, and L. Deming, J . Chem. Phyls.. 8, 753 (1940). (18) L. Pauling, J . Am. Chem. Soc., 58, 94 (1936). (19) D. A. Jones and J. G. Watkinson, Chem. I n d . (London), 661 (1960). (20) I. Brown, G. Eglinton, and M. Martin-Smith, Speetrochim. Acta, 18, 1593 (1962). (21) C. P. Smyth, “Dielectric Behavior and Structure,” McGrawHill Book Co., New York, N. Y., 1955, pp. 18, 55. 22) J. H. Richards and S. Walker, Trans. Faraday SOC.,57, 412 (1961).
n!hCROWAVE
ABSORPTIONA N D MOLECULAR STRUCTURE I N LIQUIDS
moment of 2,6-dichloro-p-nitrophenol,as calculated by eq. 1 is 3.04 D., while thatfor 2,6-dibromo-p-nitrophenol icr 3.32. Bond moment calculationsll lead one to expect! a dipole moment of 2.53 D. for 2,6-dichloro-pnitrophenol, as compared to 3.04 observed (Table IT), and 2.55 for 2,6-dibromo-p-nitrophenol, as compared t o 3.32 observed. The C-NO, bond moment was taken as 3.98 D,, and the moment due to the m-dichloro substituents was taken as 1.48, the moment of m-dichlorobenzene. The other values used have been given previously jn this paper. The measured dipole moment 3.04 is greater than 1hat expected from bond moments alone because of the large mesomeric moment, a situation similar to that in 2,6-dibroino-p-nitrophenoll. The difference between the calculated and the observed moments (D.) of p-nitrophenol is 0.70, while, for 2,6-djchloro-p-nitrophenol, it is only 0.51, and, for 2,6-djbromo-p-nitrophenol, it is 0.77. d,6-Dichloro-.p-nitroaniline. The relaxation times for this compound are calculated from the loss us. log frequency curves. The values a t 40 and 60" are more accurate than that a t 20" since the experimental points lie on both sides of the maxima. These relaxa,tion times are not to be considered as reliable as those determined from the Cole-Cole arcs for the other compounds. One would expect 2,6-dichloro-p-nitrophenol and 2,6-tlichloro-p-nitroanilineto have about the same relaxation time. The fact that the aniline has a smaller relaxation time may indicate a contribution from KHz group rotation, but this should be small in view of the large mesomeric moments of p-nitroaniline and N,Ndimethyl-p-nitroaniline, l1 which indicate a considergf double-bond character in the C-N bond. Several mo ecules containing the NH2 group have, however, been found to have somewhat lower relaxation times than the corresponding molecules with OH groups. Infrared Investigation. The OH stretching frequency of the o-halophenols is shifted to considerably lower frequencies than that of phenol. For phenol in benzene, it is shifted to a frequency 56 em. -l lower than that for phenol in carbon tetrachloride. On the other hand, in the o-halophenols, the difference between the OH frequencies in benzene and in carbon tetrachloride is 19 cm.-I or less. Thus, the position of the OH frequency indicates intramolecular bonding between the hydroxyl and halogen groups in both benzene and car-
a'nou14-
10112123
2039
bon tetrachloride solution. The OH band for 2,6dibromophenol occurs a t a lower frequency than that for 2,4-dibromophenol, indicating that the intramolecular hydrogen bond is stronger in 2,6-dibromophenol than in 2,4-dibromophenol or o-bromophenol. For o-bromophenol in benzene, a small shoulder is observed on the high frequency side of the band a t 351 5 cm.-'. This is probably due to the "free" OH band, which is shifted into the region of the intramolecularly bonded OH band because of hydrogen bonds to the benzene solvent molecules. S o t even a shoulder is observed for 2,4-dibromophenol. Because of the inductive effect of the additional bromine group and because of the fairly large breadth of both the "free" and the bonded peak, the ('free'' peak is probably cornpletely masked by the bonded peak. All of the intramolecularly bonded OH bands are significantly broader in benzene than in carbon tetrachloride. I n general, when intermolecular hydrogen bonding causes a shift to a lower frequency, it also causes band broadening24; and the stronger the bond, the broader the band. It is unlikely that the intramolecular hydrogen bond is stronger in benzene solution than in carbon tetrachloride solution. The band broadening could result from a perturbation on the intramolecular hydrogen bond exerted by the n-electrons of the benzene molecule. Thus, the benzene molecules might increase the number of possible intramolecularly bonded configurations. Similarly, the band for 2,8-dibromo-p-nitrophenol is broader in carbon tetrachloride than in cyclohexane because of the forces exerted by the C-Cl dipoles. The solute-solvent interaction due to the ability to form hydrogen bonds might also explain why the relaxation times are slightly larger than expected in comparison with other molecules of similar size and shape. For example, the relaxation time of 2,6-dibromophenol for molecular rotation is 39.2 X 10-l2 sec. a t 20" in benzene, which is considerably larger than the value 18.0 X 10-l2 sec. reportedz5for a-bromonaphthalene a t 20" in benzene, although the latter is a larger molecule.
(23) E. L. Grubb and C. P. Smyth, J . A m . Chem. SOC.,8 3 , 48'79 (1961). (24) C. M. Huggins and G. C. Pimentel, J . Phzls. Chem., 60, 1616 (1956). (25) H. Hufnagel, 2. Naturforsch., 15a, 723 (1960).
Volume 68,Number 8 August, 1964
