A 1H NMR study of the structure and conformation of phenol oriented in nematic phases

Journal of Molecular Structure, 96 (1983) 315-324 Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands

A ‘H NMR STUDY OF THE STRUCTURE AND CONFORMATION PHENOL ORIENTED IN NEMATIC PHASES

T. BJORHOLM Department (Denmark)

OF

and J. P. JACOBSEN*

of Chemistry,

Odense

University,

Campusvej

55, Odense M, DK-5230

(Received 14 June 1982)

ABSTRACT The proton spectra of phenol in the nematic phases Merck ZLI 1167 and ZLI 1132 have been recorded and analyzed. The ‘H, W dipole--dipole couplings have been obtained from the 13C satellites and the indirect IH, W and ‘H, ‘H couplings determined from an isotropic sample. The dipoledipole couplings involving the ring protons are corrected for harmonic vibrations and used to calculate the molecular r, structure of the phenyl ring. This structure was found to deviate considerably between the two phases. In the nematic phase ZLI 1167 the hydroxyl group is found to be coplanar with the phenyl ring. The same conclusion could not be reached directly for phenol in ZLI 1132 for which three models for the internal motion are discussed. INTRODUCTION

NMR spectroscopy of rigid molecules oriented in a nematic liquid crystalline phase is a useful method for obtaining structural information in the liquid state. This technique is also suitable for studying conformational changes if spin nuclei are involved in the internal motion. During the last decade, a great many molecules with internal motion were investigated by this method. However, the conclusions drawn in these cases are often based on a small number of observables (dipole-dipole couplings) compared with the number of unknown parameters. In a few cases [ 1, 21 the i3C satellites in the proton spectrum were included in the investigations thereby increasing significantly the number of observable couplings. In an earlier study of phenol dissolved in the Merck nematic Phase IV, Diehl and Henrichs [ 31 observed an unusual ordering of the molecule, which they explained by hydrogen bonding between the solvent molecules and the -OH group. Furthermore, the geometry of the -OH group contrasted with that found by other methods. This disagreement could be explained by twisting the hydroxyl group 30-40” out of the phenyl ring plane. In a related compound, thiophenol, Lunazzi et al. [ 4 ] also found that the CSH plane was twisted with respect to the phenyl ring. In both cases only three dipole*To whom correspondence 0022-2860/83/0000-0000/$03.00

should be addressed. o 1983 Elsevier Scientific Publishing Company

316

dipole couplings were available to determine the off-diagonal ordering parameter and the three coordinates of the -OH proton. Three off-diagonal ordering parameters are required, however, if the -OH group is twisted out of the plane. Thus, some drastic assumptions were necessarily introduced. In an attempt to clarify aspects of the geometry and internal motion of phenol in the liquid state we have reinvestigated the ‘H NMR spectra of phenol dissolved in the two nematic phases, Merck Licristal ZLI 1167 and Merck Licristal ZLI 1132. The “C satellites in the ‘H spectra are included in this investigation. EXPERIMENTAL

The spectrum of phenol in ZLI 1167 was recorded on a JEOL FXGOQ NMR spectrometer. The spectral width was 10 kHz and 80 000 scans provided a sufficient signal-to-noise ratio. The accumulated FID data were transferred to a Univac 1100/62 E2 computer where they were multiplied by a Gaussian function before Fourier transformation in order to enhance the resolution. The spectrum of phenol in ZLI 1132 was recorded on a Bruker HX 270D NMR spectrometer. The spectral widths were 15 kHz and 5 000 scans were accumulated. In this case resolution enhancement was obtained by adding a fraction of the original spectrum to the derivative of the spectrum [ 51. Due to the different signs for the bulk diamagnetic susceptibility of the two liquid crystals, sample spinning was allowed for in both samples. The line width was approximately 4 Hz in all spectra. 13C-enriched phenol was used in the investigations in nematic phases. The enrichment was 9% st.atisticalIy in each position. Both nematic samples were prepared as 10 wt% mixtures in 5-mm double-walled tubes. A small amount of acid alumina was added to prevent exchanges of the -OH proton with traces of water in the liquid crystal. The samples were degassed and sealed under vacuum. For the measurement of indirect coupling constants an isotropic solution of 50 wt% phenol in acetone-d, was prepared in a lo-mm tube. To this sample was likewise added a trace of acid alumina. The isotropic spectra were recorded on a JEOL FXGOQ NMR spectrometer equipped with a 13C/lH dual probe. The SPT experiments were performed according to the method of Sdrensen et al. [6] by applying a selective n-pulse to the proton spectrum immediately before the.13C spectrum was recorded. The proton pulse was adjusted in order to satisfy various selectivity requirements. A typical pulse to a pulse width of approximately 2 Hz. time was rn = 0.5 s, corresponding RESULTS

The spectrum of phenol in ZLI 1167 is shown in Fig. 1. The complexity of this spectrum made the analysis very difficult. The value of the offdiagonal ordering parameter, Sxz, was especially difficult to estimate using an

317

g. 1. Experimental (A) and computed (B) 60-MHz ‘H NMR spectra of 9% W enriched enol oriented in Merck ZLI 1167.

Iproximate geometry. These problems led to the use of the automatic fitlg program STREAK, developed by Diehl et al. [ 71, to start the analysis. iis program was unable to yield the final solution but did reproduce the lrrect overall shape of the spectrum of the unenriched molecule. The results the STREAK calculations were then used as starting parameters for the ting program MIMER, described by Manscher et al. [ 81. A correct and ial solution of the analysis was thus readily achieved. The orderings of phenol in ZLI 1167 and ZLI 1132 are sufficiently similar at analysis of the spectrum of the latter phase was readily performed lowing the solution for the former phase. Approximately 150 lines in the ectrum of the parent molecule were assigned in each phase, while 300 lines :re assigned in the 13Csatellite spectra. The values of the dipoledipole Nuplingsfor both phases are given in Table 1 (for the numbering of the ems, see Fig. 2). The errors in the results of Tables 1 and 2 correspond to a i%confidence limit. The values of the indirect spin-spin coupling constants !re fixed in the calculations of the spectra in the nematic phases. The ‘H, 13Cspin-spin coupling constants were determined by analysis of e fully coupled 13CNMR spectrum of phenol in isotropic solution by ?ans of the program MIMER. A complete analysis of the second-order fects in these spectra was included and the signs of the coupling constants !re determined by extensive use of the SPT double-resonance technique ‘1. The values thus found (Table 2) are close to those reported by Ernst et [ 91. The ‘H, ‘H spin-spin coupling constants, agree within experimental

318 TABLE

1

Chemical shifts, u(i) (ppm), two nematic phases, Merck Parameter

and dipole-dipole couplings, ZLI 1167 and ZLI 1132

ZLI 1167* Exp.

u(3) u(3)u(3) D(H1, D(H1, D(H1, D(H1, D(H1, D(H1, D(H1, D(H2, D(H2, D(H2, D(H2, D(H2, D(H2, D(H2, D(H2, D(H3, D(H3, D(H3, D(H3, D(H3, D(H3, D(H4, D(H4, D(H4, D(H4, D(H5, D(H5, D(H6, D(H6,

u(2) u(4) u(1) H2) H3) H4) Cl) C2) C3) C4) H3) H4) H5) H6) Cl) C2) C3) C4) H4) H5) Cl) C2) C3) C4) Cl) C2) C3) C4) C2) C3) C2) C3)

0.3766 0.3224 -3.0469 -11.36 77.71 71.14 764.60 126.48 46.34 36.53 509.29 108.48 84.62 141.02 278.27 1878.71 209.91 57.82 667.41 140.73 59.09 213.71 1909.99 277.01 33.45 52.26 219.37 1434.90 44.64 71.38 70.72 43.81

aObtained from the spectrum corded at 270 MHz.

D(i, j) (Hz), of phenol

in the

ZLI 1132b corr.

Error

519.52 109.79 85.07 142.03 283.36 2030.29 214.41 58.39 677.95 141.74 59.60 218.56 2066.41 281.99 33.77 52.98 225.11 1574.33 44.87 71.90 71.21 44.01

0.0017 0.0017 0.0017 0.09 0.11 0.16 0.26 0.43 0.38 1.37 0.08 0.13 0.09 0.14 0.20 0.31 0.28 0.34 0.07 0.22 0.22 0.53 0.18 0.32 0.28 0.44 0.29 0.31 0.47 0.31 0.48 0.39

recorded

Exp. 0.2687 0.3289 5.4245 -22.34 -171.73 -157.42 -1550.78 -271.59 -100.14 -80.60 -1245.87 -232.69 -142.93 -212.48 -419.60 -3175.40 -463.09 -115.50 -1126.83 -212.93 -117.41 -471.18 -3214.18 -419.55 -82.79 -119.71 -456.12 -3522.32 -75.78 -109.28 -108.12 -74.86

at 60 MHz. bObtained

Corr.

Error

-1267.50 -235.40 -143.69 -214.06 -427.87 -3436.72 -473.04 -116.65 -1145.82 -214.56 -118.45 -481.88 -3482.76 -427.64 -83.54 -121.22 -467.13 -3846.32 -76.19 -110.09 -108.89 -75.21

0.0006 0.0006 0.0006 0.13 0.13 0.17 0.45 0.59 0.43 0.65 0.10 0.17 0.12 0.21 0.45 0.43 0.37 0.39 0.12 0.24 0.37 0.73 0.29 0.39 0.60 0.57 0.40 0.43 0.72 0.41 0.77 0.47

-

from the spectrum

re-

error with those found by Schaefer et al. [lo]. Splitings due to couplings with the hydroxyl proton were not observed. These couplings were therefore set equal to zero in the calculations of the dipole--dipole couplings. Vibrational corrections to the dipole-dipole couplings were introduced by an estimated force field. Since no force field for the in-plane vibrations in phenol is reported in the literature it was therefore necessary to use the values found for toluene [ 111. The force field for the out-of-plane vibrations

319 ‘ABLE

2

ndirect spin-spin coupling constants, J(i, j) (Hz), in phenol obtained from the spectra in he isotropic phasea ‘(H2, (H2, ‘(H2, (H2, ‘(H3, (H3,

H3) H4) H5) H6) H4) H5)

8.19 1.06 0.48 2.71 7.40 1.70

f + f + t f

All couplings to the hydroxyl

0.03 0.02 0.02 0.04 0.03 0.04

J(H2, J(H2, J(H2, J(H2, J(H3, J(H3, J(H3, J(H3, J(H4, J(H4, J(H4, J(H4, J(H5, J(H5, J(H6, J(H6,

Cl) C2) C3) C4) Cl) C2) C3) C4) Cl) C2) C3) C4) C2) C3) C2) C3)

-2.67 157.86 -0.03 7.41 9.50 1.20 158.08 0.90 -1.53 8.06 1.81 160.53 -1.38 8.76 4.68 -0.69

+ 0.03 f 0.02 f 0.06 1- 0.06 + 0.03 f 0.04 2 0.05 + 0.06 f 0.04 + 0.04 f 0.06 + 0.08 f 0.04 + 0.05 t 0.04 + 0.07

proton were equal to zero.

If phenol proposed by Brand et al. [ 121 was unable to reproduce the oberved vibrational frequencies. Consequently a force field was derived by ombining the force constants for the ring system in chlorobenzene [ 131 vith those involving the -OH group according to Brand et al. [12]. It was hus possible to reproduce the experimental vibrational frequencies given by lrand and co-workers [ 141 for the three isotopic species of the molecule. ‘he harmonic vibrational corrections to the dipole-dipole couplings were alculated using the program VIBR [ 151. The values of the corrected di,ole-dipole couplings are given in Table 1; those involving the -OH proton vere not corrected for vibrational contributions.

pig. 2. Numbering of the nuclei in phenol.

320

Internal motion in a molecule influences the NMR spectrum in an oriented phase to an extent which depends on the rate of the internal motion. If it is fast relative to the reorientation of the whole molecule in the phase, only the average molecular structure determines the value of the dipole-dipole couplings. If, on the other hand, the internal motion is slow compared to the overall motion of the molecule, each conformer has its own orientation matrix and the dipole-dipole couplings are then the average of the dipole-dipole couplings of the various conformers. Diehl and Henrichs [ 31 have shown that at least three ordering parameters are necessary to describe the orientation of phenol in Phase IV. The same solution is found for phenol in ZLI 1167 and ZLI 1132, as can be seen from the values of D(H1, H2). If it is assumed that the internal motion only affects the geometry of the phenyl ring to a very limited extent, then the two diagonal elements in the ordering matrix are sufficient to determine the structure of the ring from the 22 dipoledipole couplings in which the hydroxyl proton is not involved. Such calculations were performed using the program SHAPE [ 161. The values of the bond lengths, bond angles and ordering parameters are summarized in Table 3, together with the results from a MW study [ 171. The RMS errors in these calculations with the SHAPE program were 0.16 and 0.23 Hz in ZLI 1167 and ZLI 1132 respectively. The excellent agreement between the experimental and the calculated values of the dipoledipole couplings justified the use of an average molecular structure for the phenyl ring. TABLE

3

C-H and C-C bond microwave results? Parameter r( C2-H2) r( C3-H3) r(C4-H4) r(Cl-C2) r(C2-C3) r(C3-C4) L(C~-ClX2) L(Cl--C2-C3) L(C2-C3-C4) ‘!.(C3-C4-C5) L(C~--C~-H~) L(C~-C~--H~) SZZ SXX

lengths

and C-C-C

and C-C-H

ZLI 1167 1.0838 1.0788 1.0836 1.3950 1.3915 1.3950c 119.73 120.07 120.33 119.47 119.10 119.79 -0.06636 -0.09246

bond

angles of phenol

compared

ZLI 1132 + t f f +

0.0003 0.0003 0.0011 0.0012 0.0015

?- 0.14 f 0.08 + 0.03 t 0.14 + 0.12 + 0.04 f 0.00020 + 0.00011

aDistances in A and angles in degrees. from ref. 17. CAssumed value.

bMean

MWb

1.0844 1.0792 1.0698 1.3942 1.3758 1.3950c 120.70 119.55 120.00 120.20 120.10 120.58

+ + f k ? ?

to

+ + t + +

0.0002 0.0002 0.0005 0.0009 0.0009

1.084 1.084 1.080 1.391 1.393 1.395

0.12 0.07 0.01 0.12 0.08 0.02

0.15598 0.14149 values of the bond

120.9 119.3 120.7 119.2 119.6 119.9

f 0.00021 + 0.00007 lengths

and bond

angles

321

The -OH group is usually found to be coplanar with the phenyl ring [ 17, 181. This involves one non-zero offdiagonal ordering matrix element Sxz, in addition to the diagonal elements already calculated. Due to the symmetry of the molecule, S&z = -S&, where I and II refer to the two conformers. On the basis of this assumption, the geometry of the -OH group in ZLI 1167 is as given in the second column of Table 4. The RMS error involved using a modified version of SHAPE was only 0.16 Hz which strongly supports the concept of planarity between the -OH group and the phenyl ring. Even allowing for hydroxyl group rotation in the XY plane, optimum agreement between experiment and theory was obtained with the -OH group planar to the phenyl ring. Contrary to the situation in ZLI 1167, the values of the dipole-dipole couplings of phenol in ZLI 1132 could not be directly related to the existence of two planar conformers of the molecule. Three different molecular models were therefore considered: (i) with the -OH group planar to the rigid part of the molecule, (ii) with the -OH group twisted out of the phenyl plane to form four equally populated conformers and (iii) with the internal rotation described by three ordering parameters and the potential function V(d) = V,/2(1~0~28) (V, is the potential barrier to rotation and 0 the twist angle from the ring plane). The results of the analysis of phenol in ZLI 1132 are given in Table 4. Calculations were performed using modified versions of SHAPE and the geometry and ordering parameters previously found for the phenyl ring (Table 3).

TABLE 4 Geometry of the COH fragment in phenol obtained from the dipole-dipole nematic phasesa Parameter ZL11167

@-Cl) r(O-H)

LT0t.C LCOH L'rwistd Sxz SYZ SXY RMS

Mwb

ZL11132 Model (i)

--

Model

(ii)

Model@)

1.375”

1.375”

l.zmb

1.375”

0.956"

o.956b

0.966"

o.956b

1.8 111.7 -o.1484t0.0002 0.0 0.0 0.16

couplings in

1.9 109.7 0.2672 f 0.0002 0.0 0.0 1.13

1.3 108.3 19.3 0.2441 f 0.064 0.0217 it0.22 0.15 kO.13 0.22

1.9 106.3

1.375 0.956 2.5 108.8

0.2594i c ,003 0.0 0.0 0.91

aThe distances are given in A and the angles in degrees. bAssumed values from the MW investigation [ 171. CAngle between the C-O bond axis and the symmetry axis of the phenyl ring. dAngle between the phenyl plane and the O-H bond axis.

322 DISCUSSION

The phenyl ring geometry in ZLI 1167 and ZLI 1132 was found to be slightly different. The results obtained in ZLI 1167 are close to those of the MW study, unlike the ZLI 1132 data. Only minor deviations from the hexagonal form were found in ZLI 1167 and in the gas phase, while a quinoid-like shortening of the bonds was observed in ZLI 1132. Diehl et al. [ 191 investigated the deformation of the structure of benzene in different liquid crystalline phases. They found that distortion of this molecule is negligible in ZLI 1167 and ZLI 1132. Thus, it seems reasonable to suppose that the observed distortion of phenol in ZLI 1132 is induced by the hydroxyl group. The different behaviour of the hydroxyl group in the two phases is also evident from the considerable discrepancy (8.5 ppm) between the chemical shifts of the -OH proton. Since ZLI 1167 is a mixture of cyclohexylcyclohex. anes and ZLI 1132 a mixture of phenylcyclohexanes and one biphenylcyclohexane, all containing a cyan0 group as the only functional group, there is no reason to suppose that hydrogen-bonding effects should be important in either of the two phases. It therefore seems logical to explain the difference in the chemical shift as being due to chemical-shift anisotropy. In ZLI 1167 and in the isotropic phase of ZLI 1132, the chemical shifts of the aromatic protons and the hydroxyl proton were found to be almost identical. The anomalously low value of the chemical shift for the hydroxyl proton was only found in the anisotropic phase of ZLI 1132. This indicated that the favoured orientation of phenol in ZLI 1132 is such that the hydroxyl group is near the phenyl ring or the cyan0 group of the solvent molecules. The n electrons in these systems would then be able to induce a considerable chemical-shift anisotropy. Furthermore, this intermolecular interaction would account for the quinoid structure of the phenyl ring in phenol. The preferred conformation of phenol in ZLI 1167 i; undoubtedly that with the -OH group planar to the phenyl ring. The C-O bond is reportedly tilted from the symmetry axis of the ring system in agreement with our deductions from the NMR spectra in ZLI 1167 (Table 4). The spectral data in the case of ZLI 1132 give raise to a more complex situation as mentioned above. The RMS error resulting from the assumption of two planar conformers [model (i)] is unacceptably high, although the geometry seems perfectly reasonable. The error associated with model (ii) is much smaller, however, the number of dipole-dipole couplings used in this case is only one more than the number of unknown parameters, since five ordering parameters are necessary to describe the orientation in this model. Thus, it is obvious that extreme care should be taken when using a small RMS error as the argument for a correct solution. Despite the agreement between the geometry of model (ii) and that reported from a MW study, it is difficult to believe that the stable conformers should change from being planar in the gas phase and in ZLI 1167 to being twisted in ZLI li32. It is much more probable that the rate of overall rotational motion has changed from ZLI

323

1167 to ZLI 1132 to an extent that implies that in the latter phase the internal motion can no longer be considered as being slower than the reorientational motion. Strictly speaking, then, the geometry is not determinable from the observed dipole-dipole couplings, since a correlation between the internal motion and the ordering of the molecule is expected. An approximate solution may possibly be obtained by using model (iii). In model (iii) a potential barrier to internal rotation of the hydroxyl group is introduced and three ordering parameters are incorporated to indicate that the internal motion is slower than the molecular reorientation. The value found for the potential barrier was 13.8 kJ mol-‘, which is close to the value of 14.1 kJ mol-’ found by MW spectroscopy [17]. The geometry also agrees with that from MW spectroscopic data, but contrasts with the results for ZLI 1167. Nevertheless, the RMS error is still unacceptably high. The use of model (iii) for the ZLI 1167 results led to an infinite potential barrier and the same RMS error and geometry as those reported in Tables 2 and 3. Hence, in ZLI 1167 phenol is concluded to exist in two planar conformations. CONCLUSION

The ‘H NMR spectrum in Merck nematic Phase ZLI 1167 shows that, in this phase, phenol exists as two stable planar conformers. The geometry of the phenyl ring is only slightly distorted from the hexagonal shape in agreement with MW results. Intermolecular forces between phenol and Merck nematic Phase ZLI 1132 give rise to a quinoid-like shortening of the bonds in the phenyl ring of phenol. Furthermore, the internal motion in this phase cannot simply be explained by the existence of two stable conformers. A more detailed description of the internal motion is thus required. ACKNOWLEDGEMENTS

We thank the Danish National Research Council (SNF) for the use of the Bruker HX 270D spectrometer and Dr. K. Schaumburg, Chemical Labora;ory V, University of Copenhagen, for recording the spectra on this instrunent. We are also very grateful to Niels Wessel Larsen, Chemical Laboratory V, University of Copenhagen, for the sample of 13C-enriched phenol. XEFERENCES 1 2 3 4

P. Diehl, J. Jokisaari and J. Amrein, Org. Magn. Reson., 13 (1981) 451. P. Diehl and F. Moia, Org. Magn. Reson., 15 (1981) 326. P. Diehl and P. M. Henrichs, Org. Magn. Reson., 3 (1971) 791. L. Lunazzi, P. Bellomo, C. A. Veracini and A. Amanzi, J. Chem. Sot., Perkin Trans. 2, (1979) 559. 5 H. Twilfer, K. Gersonde and M. Christahl, J. Magn. Reson., 44 (1981) 470. 6 R. Sdrensen, R. S. Hansen and H. J. Jacobsen, J. Magn. Reson., 14 (1974) 243.

324 7 P. Diehl, S. Sykora and J. Vogt, J. Magn. Reson., 19 (1975) 67. 8 0. Manscher, K. Schaumburg and J. P. Jacobsen, Acta Chem. Stand., Ser. A, 35 (1981) 13. 9 L. Ernst, V. Wray, V. A. Chertkov and N. M. Sergeyev, J. Magn. Reson., 25 (1977) 123. 10 T. Schaefer, J. B. Rowbotham and K. Chum, Can. J. Chem., 54 (1976) 3666. 11 C. Lau and R. G. Snyder, Spectrochim. Acta, Part A, 27 (1971) 2073. 12 J. C. D. Brand, S. Califano and D. R. Williams, J. Mol. Spectrosc., 26 (1968) 398. 13 J. R. Scherer, Spectrochim. Acta, Part A, 23 (1967) 1489. 14 H. D. Bist, J. C. D. Brand and D. R. Williams, J. Mol. Spectrosc., 24 (1967) 402. 15 S. Sykora, J. Vogt, H. Bijsiger and P. Diehl, J. Magn. Reson., 36 (1979) 53. 16 P. Diehl, P. M. Henrichs and W. Niederberger, Mol. Phys., 20 (1971) 139. 17 N. Wessel Larsen, Thesis, University of Copenhagen, 1974. 18 H. Forrest and B. P. Dailey, J. Chem. Phys., 45 (1966) 1736. 19 P. Diehl, H. Bijsiger and H. Zimmermann, J. Magn. Reson., 33 (1979) 113.