3718
J. Phys. Chem. B 1998, 102, 3718-3723
Assignment and Measurement of Deuterium Quadrupolar Couplings in Liquid Crystals by Deuterium-Carbon NMR Correlation Spectroscopy Ce´ line Auger,† Anne Lesage,† Stefano Caldarelli,‡ Paul Hodgkinson,† and Lyndon Emsley*,† Laboratoire de Ste´ re´ ochimie et des Interactions Mole´ culaires, UMR-117 CNRS/ENS, Ecole Normale Supe´ rieure de Lyon, 69364 Lyon, France, and Institut de Recherches sur la Catalyse, UPR-5401 CNRS, 2, aV. Albert Einstein, 69626 Villeurbanne, France ReceiVed: December 16, 1997
The performance of deuterium-carbon correlation (DECOR) spectroscopy for the assignment and measurement of deuterium quadrupolar couplings in fully deuterated liquid crystalline phases is analyzed. The experiment relies on a broadband polarization transfer step from deuterium to carbons. The efficiency of this step is demonstrated experimentally together with a brief study of cross-polarization dynamics. We show that resolution can be improved by removing the substantial contributions from heteronuclear dipolar couplings in a refocused version of the experiment. Quantification of the quadrupolar couplings is illustrated by the application to the thermotropic liquid crystal 5CB. A complete assignment is obtained by using both shortrange and long-range correlations, allowing the individual measurement of all the aromatic couplings that are overlapping in the one-dimensional deuterium spectrum. The heteronuclear C-D couplings are shown to be measurable by direct inspection of the 2D spectra. By way of demonstration, we show that the experiment allows the measurement of quadrupolar couplings as a function of temperature.
I. Introduction Liquid crystalline compounds constitute an important class of organic materials that display an intermediate state of order between crystalline lattices and isotropic solutions.1,2 However, the molecular level mechanisms that govern the physical and chemical properties of these mesophases are still not completely understood. Nuclear magnetic resonance is a tool of choice for the study of these systems, since the measurement of the various NMR interactions constitutes a detailed probe for the study of the molecular structure and dynamics at a very local atomic level.3 This is the case for instance for the averaged dipolar couplings, and we have recently shown that techniques based on protondetected local field (PDLF) spectroscopy are well-suited to the determination of both short-range and long-range carbon-proton dipolar couplings in liquid crystalline phases.4,5 Deuterium quadrupolar couplings constitute another probe that can also in principle be measured by NMR. However, owing to the very low natural abundance of deuterium (about 0.02%), except in very favourable circumstances, the measurement of the quadrupolar couplings requires chemical isotopic substitution. There are two possible approaches to labeling: selective labeling, which requires a large investment in synthesis for the fabrication of a series of selectively labeled molecules, or uniform labeling of the whole molecule (or the part of interest), which reduces the complexity of the chemistry involved but creates problems of spectral assignment and overlap. In this article we are concerned with the problem of spectral assignment and overlap in uniformly deuterated molecules. In uniaxially oriented materials the deuterium spectrum is a superposition of doublets. Each C-D pair gives rise to one * To whom correspondence should be addressed (email:
[email protected]). † Laboratoire de Ste ´ re´ochimie et des Interactions Mole´culaires. ‡ Institut de Recherches sur la Catalyse.
doublet, which yields directly the averaged quadrupolar coupling. These quadrupolar splittings δυQ are probes of local order and molecular dynamics and can be expressed as
δυQ ) 2ωQ )
3 e2qQ 2
h
2 2 D0,m (θ, ψ) Dm,0 (β, R) ∑ m
where (β, R) are the polar angles of the C-2H vector in the molecular frame and (θ, ψ) the polar angles of the director in the molecular frame.6 For a rigid molecule in an uniaxial phase, 2 (θ, ψ) are directly related to the order paramthe terms D0,m eters describing the molecular orientation with respect to the phase axis. By way of an example, Figure 1 shows the deuterium NMR spectrum of the fully deuterated nematic liquid crystal 5CB (4-n-pentyl-4′-cyanobiphenyl). Despite the simplicity of the deuterium spectra, one-dimensional deuterium spectroscopy is limited, even in such relatively simple systems, since it is not possible to assign the deuterium spectra without recourse to prior knowledge or to models. This is due to the fact that there are essentially no chemical shift differences between the deuterons. Several years ago, it was shown that twodimensional deuterium autocorrelation experiments can be used to assign the deuterium spectrum of solutes dissolved in liquid crystals in the special case where the deuterium-deuterium dipolar couplings are resolved, but this is not generally the case.7 We recently introduced a two-dimensional carbon-deuterium correlation experiment, referred to as DECOR for DEuterium CORrelation, which allows the assignment of deuterium spectra for oriented materials.8 The deuterium assignment is based on prior knowledge of the carbon spectrum, which can be assigned using several existing methods.9 In this paper we present a detailed description of the experiment and its application to the measurement of the quadrupolar couplings in 5CB, a wellcharacterized model thermotropic liquid crystal.10,11 We provide extensions of the original experiment to eliminate splittings due
S1089-5647(98)00539-2 CCC: $15.00 © 1998 American Chemical Society Published on Web 04/17/1998
Deuterium Quadrupolar Couplings in Liquid Crystals
Figure 1. Molecular model of the nematic liquid crystal 5CB together with one-dimensional 61.4 MHz deuterium and 125 MHz carbon NMR spectra of a static sample of 5CB. The deuterium spectrum was obtained from a fully deuterated sample of 5CB with a simple one-pulse sequence. This spectrum is relatively simple, showing a set of doublets for each C-D pair. The weak splittings observed on the R peaks are due to the homonuclear D-D dipolar couplings. The carbon spectrum was obtained with a cross-polarization sequence and proton decoupling. The carbon spectrum shows 13 well-resolved lines.
Figure 2. DECOR pulse sequence. Deuterium single-quantum coherence is allowed to evolve during the first evolution time t1 under the deuterium quadrupolar couplings. A cross-polarization step, using a linear off-resonance ramp on deuterium, transfers the magnetization to carbon spins. The carbon resonances are then detected in t2 with continuous wave deuterium decoupling (pulse program and phase cycle available on request).
to dipolar couplings (thereby increasing resolution and sensitivity), and we show that remaining overlap problems can be overcome by the use of long-range correlations. II. The DECOR Experiment All the experiments in this article were obtained using a fully deuterated sample of 5CB and were collected (unless otherwise indicated) at room temperature on a Bruker DSX400 wide-bore spectrometer using a 7-mm triple resonance (2H-13C-1H) CPMAS probe without sample rotation. The pulse sequence used for the experiments is illustrated in Figure 2. The 2H magnetization is excited by a 90° pulse and evolves during t1 under the effect of the dominant quadrupolar interactions and the homonuclear dipolar interactions and, when a deuterium 180° pulse is not applied in t1, under the effect of the deuterium chemical shift and the heteronuclear dipolar interactions. A cross-polarization step then transfers the deuterium magnetization to carbon spins that are observed during t2. Transfer of deuterium (single-quantum) polarization can be achieved by
J. Phys. Chem. B, Vol. 102, No. 19, 1998 3719
Figure 3. One-dimensional deuterium-to-carbon cross-polarization experiments on 5CB-d19. All the experiments were recorded with 4096 scans and a 5 ms contact time for cross-polarization; the deuterium offset frequency during the cross-polarization step ΩCP was set to 18 kHz and the deuterium decoupling field during acquisition to 34.5 kHz. During the CP, for spectra a and b, a constant amplitude rf field of 15.5 kHz was used on deuterium with carbon rf fields of 14.7 and 18.0 kHz, respectively. For spectrum c a linear ramp on deuterium was applied ranging between 14.6 ( 4.9 kHz (carbon rf field of 19 kHz). (d) One-dimensional carbon spectrum of 5CB-d19 obtained with a simple one-pulse sequence and continuous wave deuterium decoupling during acquisition (decoupling field strength of 34.5 kHz).
placing the transmitter frequency at the frequency of one component of the quadrupolar doublet, ΩCP ) (ωQ. In the limit where the radio-frequency (rf) field on 2H, ω1I , ωQ, there is a sharp CP match condition of ω1S ) (2)1/2ω1l.12,13 The match condition is modified if ω1I is comparable to ωQ, which is the case in our experiment, but the match is still strongly dependent on the value of the quadrupolar couplings. In the case of 5CB, for which the quadrupolar couplings range from 5 to 30 kHz (see Figure 1), the efficiency of cross-polarization is thus strongly dependent on ωQ, as illustrated in Figure 3. When using a constant radio-frequency field amplitude on deuterium, CP conditions can be found that are effective for either the aliphatic part (Figure 3a) or the aromatic part (Figure 3b) of the carbon spectrum, but no single match condition is effective for the whole spectrum. In order to render the process broad band, we make use of variable amplitude cross-polarization,14-16 in which we sweep linearly the amplitude of the deuterium radio-frequency field, while keeping the carbon field constant. This technique enlarges the match condition with respect to the quadrupolar interaction as shown in Figure 3c, where we observe both aliphatic and aromatic carbons; theoretical details of deuterium to carbon CP under these conditions will be given elsewhere.17 Figure 3d shows the carbon spectrum obtained with a one-pulse experiment and continuous wave deuterium decoupling during acquisition. Note that, even when using a linear ramp on deuterium, the position of the 2H transmitter frequency ΩCP is important and needs to be placed as close as possible to all the deuterium resonances, i.e., approximately in the middle of one side of the deuterium spectrum. Indeed, when
3720 J. Phys. Chem. B, Vol. 102, No. 19, 1998 the offset is set to the center of the deuterium spectrum only the aromatic carbons are visible whereas placing the offset 26 kHz upfield yields only the aliphatic carbons. For 5CB under our experimental conditions, if the offset is set to about 16 kHz, all of the carbons are cross-polarized with a roughly equivalent efficiency. Although this cross-polarization technique may appear somewhat complicated, in fact the conditions for a broad band match are relatively robust and easily established. In experimental circumstances, one should simply place the transmitter frequency ΩCP roughly in the center of one side of the deuterium spectra. A ramp having a range of rf amplitudes of roughly (5 kHz should be used for samples having typical distributions of quadrupolar couplings for liquid crystalline samples (0-30 kHz). The match condition should then be optimized for deuterium rf field amplitudes comparable to the offset frequency ω1I ≈ ΩCP. Using these initial conditions, the cross-polarization parameters can be optimized (if necessary) by observation of the one-dimensional deuterium-carbon CP spectrum (like those shown in Figure 3), before starting the two-dimensional correlation experiment. After the cross-polarization step, the carbon resonances are detected in t2 under continuous wave deuterium decoupling with the 2H transmitter frequency placed in the center of the deuterium spectrum. The sequence leads to a two-dimensional deuterium-carbon correlation spectrum in which the carbon chemical shifts in ω2 are correlated with the deuterium interactions in ω1. Note that the experiment yields an asymmetric spectrum in the ω1 dimension, since only one-half of the deuterium spectrum is spin-locked during the CP and thus transfers its magnetization to carbon. Note that complete spectra can be acquired, if necessary, by addition of two spectra, one acquired with ΩCP ) +ωQ and another with ΩCP ) -ωQ, corresponding to the two complementary halves of the spectrum. III. Assignment and Measurement of the Quadrupolar Couplings in 5CB Figure 4a shows the aliphatic region of a DECOR spectrum recorded with the original experiment,8 i.e., without a 180° pulse during t1. In this case, the deuterium dimension is dominated by the quadrupolar interactions but also includes the deuterium chemical shift and dipolar interactions. Since only half of the spectrum is obtained, the presence of the chemical shift prevents a quantitative measurement of the quadrupolar couplings (although it does not prevent assignment). Additionally, the spectrum clearly shows splittings for each of the R, β, and γ deuterons that are attributed to the 13C-2H heteronuclear dipolar couplings. The deuterium dimension can be substantially simplified by the addition of a 180° deuterium pulse during t1 to refocus the deuterium chemical shift and heteronuclear dipolar interactions. In the refocused spectrum shown in Figure 4b, the deuterium dimension is pure quadrupolar. The removal of the heteronuclear couplings results in a clear narrowing of all the five lines in ω1, not only for those exhibiting the resolved couplings (R, β, and γ) but also for the δ and ω deuterons, indicating that a significant part of the unrefocused line widths is due to the heteronuclear dipolar couplings. Note that, since the evolution under the deuterium chemical shift is also removed, the distance of the peaks from the center of the deuterium dimension in the refocused DECOR experiment gives the quadrupolar couplings directly. The heteronuclear C-D couplings can be, on the other hand, extracted from DECOR spectra obtained without a 180° pulse during t1 by direct inspection of the 2D maps. Note that 2H chemical shifts can also be in principle measured by comparison with the refocused
Auger et al.
Figure 4. Contour plots of two-dimensional DECOR experiments on 5CB-d19 in the aliphatic region (a) without and (b) with a 180° deuterium pulse during t1. A total of 128 t1 increments with 96 scans each were collected, using a 90° deuterium pulse of 7.25 µs. The contact time for cross-polarization was set to 5 ms using a rf field of 18.5 kHz for carbon and a linearly ramped rf field of 14.6 ( 4.9 kHz for deuterium. During the CP step, the deuterium frequency was moved 18 kHz off-resonance to allow single-quantum cross-polarization as discussed in the text. The deuterium decoupling field strength during t2 was 34.5 kHz (on-resonance continuous wave decoupling). Quadrature detection in ω1 was achieved using TPPI. A 4 s recycle delay was used between scans to avoid problems with sample heating.
experiments or from sum experiments in which both halves of the unrefocused spectrum are acquired. The knowledge of the carbon spectrum permits the straightforward assignment of the quadrupolar couplings in the Ω1 dimension for the aliphatic deuterons. The identification of the correlations is unambiguous since the carbon resonances are well-resolved and since only direct one-bond correlations are visible in this region of the spectra (using a cross-polarization contact time of 5 ms). The assignment of the aromatic part of the DECOR spectra is not so straightforward because some twobond couplings are involved in the cross-polarization step, even for relatively short mixing times. This is illustrated by the two DECOR spectra of Figure 5, which were collected with two different cross-polarization times: 5 and 15 ms. In this figure, the 2D maps use the following notation: each one-bond correlation peak is indicated by the name of the deuteron giving the corresponding ω1 frequency (which is also the name of the carbon giving the ω2 frequency), while the long-range correlations are indicated by an arrow from the label of the originating deuteron to that of the receiving carbon. The assignment of the quadrupolar couplings for deuterons 2′ and 3′ is ambiguous in the 5 ms contact time DECOR experiment since both are correlated to both 2′ and 3′ carbons. In this case, the particular combination of geometry and molecular motion results in the one-bond and two-bond C-D dipolar couplings (between D2′C2′ and D2′-C3′ on one hand and between D3′-C3′ and D3′C2′ on the other hand) being very similar. Thus, the polarization transfer to either carbon 2′ or 3′ may arise from deuterons 2′ and 3′. The ambiguity is removed by using longer contact times to obtain weaker, long-range correlations. Indeed with a 15 ms contact time (Figure 5b), a correlation (labeled 3′ f 4′) appears at the carbon 4′ ω2 frequency, which can only reasonably be due to cross-polarization from deuteron 3′. The
Deuterium Quadrupolar Couplings in Liquid Crystals
J. Phys. Chem. B, Vol. 102, No. 19, 1998 3721 TABLE 1: Deuterium Quadrupolar Splittings δυQ and Deuterium-Carbon Dipolar Splittings Measured from 2D DECOR Spectra in 5CB nuclei aliphatic R γ β δ ω aromaticc 2 2′ 3 3′
δυQ (kHz)a ((200 Hz)
D (kHz)b ((100 Hz)
51.8 37.9 35.4 25.3 18.3
1.09 1.02 1.46
12.3 11.4 12.0 9.0
a Values of the deuterium quadrupolar splittings obtained at 299 K from a refocused DECOR experiment (contact time of 15 ms for crosspolarization) by a direct measurement with respect to the center of the spectrum. b Values of the 2H-13C dipolar splittings obtained from an unrefocused DECOR experiment by measuring the splitting observed for each correlation peak (see Figure 4). Note that the splitting is the sum of dipolar and scalar couplings. c For the aromatic nuclei 2, 2′, 3, and 3′, the measurements were made using peaks 2 f 1, 2′ f 3′, 3, and 3′.
Figure 5. Contour plots of two-dimensional refocused DECOR spectra of 5CB-d19 recorded with two different contact times for crosspolarization of 5 and 15 ms. The RF field during CP on deuterium was ramped linearly between 11 ( 6 kHz, with a carbon rf field of 16.5 kHz. A 180° pulse was applied to deuterium in t1. For each of the 128 t1 increments, a total of 128 and 600 scans were collected, respectively, for the 5 and 15 ms contact time experiments. The other experimental conditions are the same as in Figure 4. See the text for a discussion of the long-range correlations observed with the longer mixing time.
correlation peaks due to deuterons 2′ and 3′ are thus unambiguously identified (peaks 2′, 3′, 2′ f 3′, and 3′ f 2′). Note that the correlation peaks 2 and 3 probably correspond to both onebond and two-bond transfer from deuterons 2 and 3. In this case the two types of transfer cannot be distinguished since the two deuterons have the same quadrupolar couplings within the experimental precision. With long contact times the quaternary carbons can be polarized. This is the case for carbon 4′ as we have just mentioned, but also for carbons 1 and 4. We observe that, as may be expected, carbon 4 is correlated with the R deuterons (peak labeled R f 4) and that carbon 1 is correlated with deuteron 2 (peak 2 f 1). The peak labeled A may correspond to both a polarization transfer 3 f 4 and 2′ f 1′. Long-range correlation peaks are also observed within the aliphatic chain. The ω deuteron appears to be correlated, as expected, with carbon ω (one-bond transfer corresponding to the peak labeled ω), but also with carbons γ (three-bond transfer corresponding to the peak labeled ω f γ). On the other hand, no correlations due to a two-bond transfer ω f δ or to a complementary transfer from deuteron γ to carbon ω are visible. This is due to the dependence of the effective dipolar couplings on the orientation of the different 13C-2H vectors with respect to the magnetic field. Another four-bond correlation peak δ f R is also clearly visible. The observation of the three- and four-bond couplings (and the absence of shorter-bond couplings) is not in agreement with what we may expect from the measurement of the heteronuclear 1H-13C couplings in protonated 5CB using the 3D PDLF experiment5 (notably the heteronuclear couplings for the ω
proton range, in order of size Hω-Cω > Hω-Cδ > Hω-Cγ > Hω-Cβ). As a consequence, we conclude that the mechanism for long-range transfer depends, at least partly, on the values of the homonuclear 2H dipolar and quadrupolar couplings, possibly involving deuterium spin diffusion. However, preliminary experiments do not indicate that deuterium spin diffusion occurs during the 15 ms spin-lock, so caution should be associated with the interpretation of very long-range correlations (although they could potentially provide information about molecular geometry). As pointed out previously, the assignment of the DECOR spectra allows the measurement of the quadrupolar couplings and the heteronuclear dipolar C-D couplings (visible on the nonrefocused DECOR spectra for the R, β, and γ peaks). All the parameters we have obtained for 5CB are summarized in Table 1. This is the first time that the quadrupolar couplings have been measured unambiguously for the aromatic part of 5CB. Note in particular that the quadrupolar couplings for deuterium 2 and 2′, which are overlapping in the 1D deuterium spectrum, are now clearly resolved. IV. Temperature Dependence of Quadrupolar Couplings in 5CB The DECOR experiment can be used to study phase transitions of thermotropic liquid crystals in more detail than onedimensional deuterium spectra since the 2D maps allow us to observe the evolution as a function of temperature of quadrupolar couplings that give rise to overlaping resonances in the 1D deuterium spectrum. We used the experiment to follow the transition between the nematic and isotropic phase for 5CB, as illustrated in Figure 6. Similar observations for the aliphatic part of the molecule using one-dimensional deuterium spectroscopy have been reported previously,10 and our measurements are in good agreement with this data. The DECOR experiment has enabled us, however, to distinguish separately the evolution of the quadrupolar couplings of the aromatic 2, 2′, 3, and 3′ deuterons (using peaks 2 f 1, 2′ f 3′, 3, and 3′, respectively). The quadrupolar couplings as a function of temperature are given in Table 2. They should be important experimental constraints in the development of models for the influence of local flexibility on the global or local ordering (such models
3722 J. Phys. Chem. B, Vol. 102, No. 19, 1998
Auger et al.
Figure 6. Deuterium quadrupolar frequencies in 5CB-d19 measured from 2D DECOR experiments plotted as a function of temperature. In panel a the evolution of the deuterium frequencies is shown for all the nuclei, whereas panel b is an expansion for the aromatic deuterons. TNI is the temperature of the nematic to isotropic phase transition and was found to be 307 K. The cross-polarization contact time was set to 11 ms. The other experimental conditions are the same as in Figure 4.
TABLE 2: Temperature Dependence of the Quadrupolar Splittings in 5CB T - TNI (K)a δυQ (kHz)
7.0
6.0
4.0
3.0
1.5
R γ β δ ω 2 3 2′ 3′
50.5 36.9 34.4 24.7 17.9 12.0 11.8 11.2 8.8
49.1 35.7 33.2 23.7 17.2 11.5 11.3 10.8 8.3
45.7 32.8 30.4 21.8
42.8 30.8 28.3
38.0 26.5 25.4
10.4 10.4 9.8 7.5
9.5 9.4 8.9 6.8
8.2 7.5
a T is the temperature of the nematic-to-isotropic phase transition. NI The errors are estimated to be about (200 Hz.
are outside the scope of this article). Notably, this is the first direct observation that the 2 and 3 deuterons have a significantly different quadrupolar coupling, in agreement with earlier work based on an analysis of the one-dimensional spectrum.18 Note the rapid disappearance of the ω peak (at 300 K). Although this could be due to a failure of decoupling, the other peaks are still well-decoupled, and we rather relate this to a change in couplings affecting the efficiency of the CP. V. Cross-Polarization Dynamics The cross-polarization dynamics from deuterium to carbon during the DECOR experiments was studied by measuring the peak intensities in the 2D maps as a function of the contact time (Figure 7). The dipolar C-D couplings can be calculated from the measurement of the dipolar C-H couplings (which have been previously reported4) by dividing by a factor of 6.2
Figure 7. Peak intensities of 2D DECOR spectra measured as a function of the cross-polarization contact time. Other parameters were the same as for Figure 4.
(the ratio of gyromagnetic ratios). The predicted splittings range between 1100 and 380 Hz for the aliphatic part and between 340 and 180 Hz for the aromatic part. Since the crosspolarization contact time needs to be of the order of the inverse of the coupling strength, this accounts for the fact that crosspolarization times of at least 5 ms are required for effective cross-polarization. Also note that these dipolar couplings scale in parallel with the quadrupolar couplings as usual for nematic liquid crystals. Figure 7 shows that, as expected, the crosspolarization efficiency roughly increases with the strength of the dipolar coupling. The R carbon, which has the strongest dipolar coupling, reaches its maximum value after only 3 ms whereas the ω carbon, which has the smallest dipolar coupling of the aliphatic chain, is still increasing in intensity at 9 ms. The aromatic carbon polarizations, for which the C-D couplings are even smaller, are also continuing to rise at 9 ms. Note however that optimal cross-polarization conditions are not (and cannot be)17 achieved simultaneously for all the carbons, which renders the quantitative interpretation of such curves difficult. Interestingly, the DECOR experiment allows a distinct study of the direct and indirect cross-polarization dynamics. Indeed we can see that the 3′ carbon magnetization transferred from the 2′ deuteron (indirect transfer) is still increasing after 9 ms (peak 2′ f 3′) whereas the 3′ carbon magnetization transferred from the 3′ deuteron (direct transfer) has reached its maximum value (peak 3′). This is a quite interesting feature since the two dipolar couplings are of the same order of magnitude. A more detailed treatment of deuterium to carbon CP will be presented elsewhere.17 VI. Conclusions We have shown that deuterium-carbon correlation spectroscopy is suitable for both the assignment and the measurement of deuterium quadrupolar couplings in liquid crystals. The method relies on a cross-polarization step from deuterium to
Deuterium Quadrupolar Couplings in Liquid Crystals carbons that has been shown to work over a wide range of quadrupolar couplings. The DECOR experiment offers a real gain in resolution in cases where 1D spectra are overlapping and thus in the precision of the measurement of deuterium quadrupolar couplings. Even in the well-characterized model compound 5CB the experiment has allowed us to measure some couplings unambiguously for the first time. The measurement of the quadrupolar couplings, from which local order parameters can be calculated directly, is particularly informative about the dynamics of molecules in the mesophases, and we believe that DECOR experiments will be very useful for such studies. We have illustrated the approach with an example of a thermotropic liquid crystal, but the technique is equally applicable to measure deuterium quadrupolar couplings in lyotropic model membrane systems19,20 or in proteins oriented in these systems,21,22 and we are currently pursuing studies in such systems. Acknowledgment. We are grateful to Professor J. W. Emsley for loaning us the sample of deuterated 5CB and to H. Foerster and F. Engelke (Bruker) for technical support and for the generous loan of the triple resonance probe. P.H. was supported during this work by a Marie Curie Fellowship from the European Commission (Contract ERBFMBICT961185). References and Notes (1) Stegenmeyer, H. Liquid Crystals; Springer: Darmstadt, 1994. (2) Kumar, S. Liquid Crystals in the Nineties and Beyond; World Scientific Publishing: London, 1995.
J. Phys. Chem. B, Vol. 102, No. 19, 1998 3723 (3) Emsley, J. W. Nuclear Magnetic Resonance of Liquid Crystals; Emsley, J. W., Ed.; Reidel: Dordrecht, 1985. (4) Caldarelli, S.; Hong, M.; Emsley, L.; Pines, A. J. Phys. Chem. 1996, 100, 18696. (5) Caldarelli, S.; Lesage, A.; Emsley, L. J. Am. Chem. Soc. 1996, 118, 12224. (6) Dong, R. Y. Nuclear Magnetic Resonance of Liquid Crystals (Partially Ordered Systems); Springer-Verlag: Berlin, 1994. (7) Emsley, J. W.; Turner, D. L. Chem. Phys. Lett. 1981, 82, 447. (8) Auger, C.; Lesage, A.; Caldarelli, S.; Hodgkinson, P.; Emsley, L. J. Am. Chem. Soc. 1997, 119, 12000. Note that there is a typographical error in this article and that the appropriate condition is ω1S ) x2ω1I (and not ω1S ) x(2ω1I)). (9) Sandstro¨m, D.; Summanen, K. T.; Levitt, M. H. J. Am. Chem. Soc. 1995, 116, 9357. (10) Emsley, J. W.; Luckhurst, G. R.; Stockley, C. P. Mol. Phys. 1981, 44, 565. (11) Chandrakumar, T.; Zimmerman, D. S.; Bates, G. S.; Burnell, E. E. Liq. Cryst. 1994, 17, 457. (12) Brunner, P.; Reinhold, M.; Ernst, R. R. J. Phys. Chem. 1980, 73, 1086. (13) Vega, S.; Shattuck, T. W.; Pines, A. Phys. ReV. A 1980, 22, 638. (14) Takegoshi, K.; Ito, M.; Terao, T. Chem. Phys. Lett. 1996, 260, 159. (15) Metz, G.; Wu, X.; Smith, S. O. J. Magn. Reson. A 1994, 110, 219. (16) Hediger, S.; Meier, B. H.; Ernst, R. R. Chem. Phys. Lett. 1995, 240, 449. (17) Hodgkinson, P.; Auger, C.; Emsley, L. Submitted. (18) Hoatson, G. L.; Bailey, A. L.; Est, A. J. V. D.; Bates, G. S.; Burnell, E. E. Liq. Cryst. 1988, 3, 683. (19) Sanders, C. R.; Hare, B. J.; Howard, K. P.; Prestegard, J. H. Prog. Nucl. Magn. Reson. Spectrosc. 1994, 26, 421. (20) Prosser, R. S.; Hunt, S. A.; DiNatale, J. A.; Vold, R. R. J. Am. Chem. Soc. 1996, 118, 269. (21) Tjandra, N.; Bax, A. Science 1997, 278, 1111. (22) Howard, K. P.; Opella, S. J. J. Magn. Reson. B 1996, 112, 91.