DOI: 10.1002/cphc.201500252
Articles
Measurement of Large Dipolar Couplings of a Liquid Crystal with Terminal Phenyl Rings and Estimation of the Order Parameters R. V. Sudheer Kumar[a, b] and Krishna V. Ramanathan*[a] NMR spectroscopy is a powerful means of studying liquid-crystalline systems at atomic resolutions. Of the many parameters that can provide information on the dynamics and order of the systems, 1H–13C dipolar couplings are an important means of obtaining such information. Depending on the details of the molecular structure and the magnitude of the order parameters, the dipolar couplings can vary over a wide range of values. Thus the method employed to estimate the dipolar couplings should be capable of estimating both large and small dipolar couplings at the same time. For this purpose, we
consider here a two-dimensional NMR experiment that works similar to the insensitive nuclei enhanced by polarization transfer (INEPT) experiment in solution. With the incorporation of a modification proposed earlier for experiments with low radio frequency power, the scheme is observed to enable a wide range of dipolar couplings to be estimated at the same time. We utilized this approach to obtain dipolar couplings in a liquid crystal with phenyl rings attached to either end of the molecule, and estimated its local order parameters.
1. Introduction Liquid crystals are the quintessential partially ordered systems, and find uses in a variety of applications. Of these systems, the design of thermotropic liquid crystals has undergone significant changes over the years. In the conventional approach, the molecules were either rod-like or disk-like.[1–3] In recent times, however, many systems with different shapes, such as bent core, l, H, and S, have also been proposed and exhibit interesting novel phases.[4–6] In this scenario, NMR spectroscopy has been found to be a powerful tool for investigating the structure, topology, and order that provide correlations between the molecular shape and the mesophase properties of the material. The design of liquid-crystalline molecules involves one or more core units connected to flexible chains.[1–3, 7] The study of the core units has drawn significant attention due to the fact that they give rigidity to the molecule and are also involved in intermolecular interactions. NMR spectroscopy has been highly useful in the study of such systems because it has been shown to provide information with atomistic resolution.[8, 9] Although nuclei like 1H, 2H, and 19F have been utilized for investigating such systems,[10–13] in recent years the naturalabundance 13C NMR spectroscopic technique is being increasingly used due to the fact that this approach does not require special sample preparation, such as isotope labeling, and provides high-resolution spectra in the oriented phase.[14–21] It is possible to obtain two anisotropic parameters from 13C NMR [a] R. V. S. Kumar, Prof. K. V. Ramanathan NMR Research Centre, Indian Institute of Science Bangalore 560012 (India) E-mail: [email protected] [b] R. V. S. Kumar Department of Physics, Indian Institute of Science Bangalore 560012 (India)
ChemPhysChem 2015, 16, 2199 – 2205
spectroscopy. The first is the alignment-induced chemical shift (AIS), which is the difference in the chemical shift of the carbon in the isotropic phase to that in the liquid crystalline phase, and arises from the alignment of the molecule in the magnetic field. The magnitude of the AIS depends on two parameters, namely, the order parameter and the chemical shift anisotropy tensors in a complex manner.[15, 22] The second parameter is the 1H–13C heteronuclear dipolar coupling. The magnitude of the dipolar coupling, like AIS, also depends on the magnitude and the orientation of the order parameter and dipolar tensors.[15, 21] For a rigid molecular structure, the magnitude and the direction of the principal components of the dipolar tensor are known. Thus dipolar couplings provide the local order parameters in a straightforward manner. Experimentally, however, the determination of the dipolar couplings is more involved and requires the use of the separated local field (SLF) 2D NMR method.[23] In this method, the chemical shift of the rare spin (S) is presented along the horizontal axis and the heteronuclear dipolar coupling of the S spin to another spin I, usually a proton, is depicted along the vertical axis. The experiment also involves removal of homonuclear couplings between protons while evolution under I–S coupling is carried out. There are several variations of the SLF experiment, each with relative advantages and disadvantages. These methods generally differ in the manner in which the evolution under the heteronuclear I–S interactions occurs. In the conventional SLF method,[23] the S spin evolution takes place under multiple I–S couplings, which results in broadening of the resonances due to overlapped multiplet structures. Conversely, the polarization inversion spin exchange at magic angle (PISEMA)[24] method has been extensively utilized for obtaining isolated I–S couplings, such as those that involve 15N–1H in peptides and
2199
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles proteins.[25, 26] In the presence of multiple I–S couplings, as happosed by Kharkov et al.[39] In this modification, a low r.f. field is 1 13 pens in the study of H– C dipolar couplings in liquid crysused for one BLEW-48 cycle, which enables the dwell time to tals,[27] a method that has significant advantages is the protonbe halved. We initially demonstrate this approach for the encoded local field (PELF) spectroscopy experiment.[28–31] In this known liquid crystal 4-cyano-4’-pentylbiphenyl (5CB). Subseapproach, the abundant I spins are allowed to evolve in the quently, we use it to obtain the dipolar couplings in a novel presence of their couplings to the rare nucleus S. Thus, an I liquid crystal with four phenyl groups and with an aliphatic spin has only a single S spin coupling. All the I–S dipolar couchain inserted between two of the phenyl groups, and demonplings are observed with the corresponding doublets separatstrate its utility in measuring large dipolar couplings and the ed out according to their magnitude. This results in narrow corresponding order parameters in a straightforward manner. and well-resolved spectral lines that correspond to different I–S spin pairs. The basic approaches mentioned above have been 2. Pulse Sequence further modified to address some of the problems encountered in practical applications, such as the heating effects due The basic pulse sequence is shown in Figure 1a, and utilizes to the continuous application of radio frequency (r.f.), the the BLEW-48 sequence for both homonuclear decoupling and effect of proton r.f. offsets, and so on.[32–34] Of these methods, for heteronuclear polarization transfer. As a consequence of we consider here the dipolar assisted polarization transfer the application of the BLEW-48 pulse sequence, the heteronu(DAPT) pulse sequence proposed recently,[35] which has the adclear Hamiltonian (HIS) has the form k2DISIXSZ, in which DIS is the vantage of being used either as an S-nucleus-evolved SLF-like heteronuclear dipolar coupling between the I and S spins and pulse sequence[36] or an I-spin-evolved PELF-like pulse sek is the scaling factor. For the BLEW-48 sequence, the scaling quence. This sequence, like the insensitive nuclei enhanced by factor was 0.424 and the effective field was along the transpolarization transfer (INEPT) pulse sequence[37] used in solution verse plane. During the initial BLEW-48 period P0, the IZ magNMR, uses only a simple 908 pulse on the S spin, which thus netization evolved under the scaled heteronuclear dipolar coueliminates the high r.f. power required in the case of cross-popling to produce a two-spin-order term, 2IYSZsin (kDISt1). This larization-based sequences with simultaneous RF irradiation was converted to 2IYSYsin (kDISt1) by a 908 pulse applied on the applied on the I and the S spin channels. Furthermore, usage S channel. Subsequent evolution during t2 took place under of homonuclear decoupling throughout ensures that the transthe phase-shifted BLEW-48 sequence P90, in which the phases fer of polarization between I and S spins takes place between of all the pulses were shifted by 908 degrees with respect to P0 coupled I–S spin pair only, which avoids indirect transfers to provide an effective HIS of k2DISIYSZ. This refocused the twothrough the I–I coupling pathway. In the original presentation spin-order term 2IYSY to create the observable single-spin SX of this sequence, a BLEW-12[38] pulse sequence was used for operator term. During the acquisition period, the S spin signal homonuclear proton decoupling. BLEW-12 is observed to be observed with heteronuclear dipolar decoupling after the less efficient in comparison with other sequences, such as chemical shift evolution during t2 was refocused with a 1808 BLEW-48,[38] and also has an effective field inclined at an angle pulse applied on the S spin channel. It may be noted that the (q) of 26.68 with reference to the z axis. The latter condition reobserved S spin signal was modulated by dipolar evolution quires additional pre-pulses to be applied to appropriately that took place during both t1 and t2 periods, which resulted align the proton magnetization and maximize signals. Conin a signal with the form SXsin (kDISt1) sin (kDISt2). Interestingly, versely, BLEW-48 has an effective field in the transverse plane, during the t1 period the evolution was of the I (or proton) which is an advantage for experiments that involve proton chemical shifts or heteronuclear dipolar evolution with homonuclear decoupling. Herein, the DAPT experiment, modified to include the BLEW-48 pulses, has been exploited to obtain PELF spectra. However, the BLEW-48 pulse sequence has a long cycle time, which causes the dwell time in the dipolar dimension to increase. The increased dwell time is a disadvantage when large dipolar couplings are required to be measured because Figure 1. a) Pulse sequence for homonuclear decoupling evolved local field (HDELF) spectroscopy. P0 and P90 are it will result in folding. To cirBLEW-48 homonuclear decoupling pulse sequence blocks. The phases of all r.f. pulses are shifted by 908 between cumvent this problem, we have P0 and P90. b) HDELF used as a proton-evolved pulse sequence (P-HDELF). Use of reduced r.f. power for one BLEWadopted a modification to the 48 block at the center of the t1 period and increasing its duration by 1.5 times for alternate t1 increments decreasPELF sequence recently proes the dwell time by half in the indirect dimension (see main text and Table 1 for details). ChemPhysChem 2015, 16, 2199 – 2205
www.chemphyschem.org
2200
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles magnetization under the dipolar coupling DIS. This allows the experiment to be used as a proton-encoded experiment, with t1 used as the t1 variable. Conversely, evolution during t2 was the evolution of the S (or carbon) magnetization under DIS, which makes this an SLF experiment when t2 is used as the variable. In view of these possibilities, we refer to this pulse sequence as homonuclear decoupling evolved local field (HDELF) spectroscopy. The proton-evolved version is referred to as PHDELF and the carbon-evolved version as S-HDELF. As mentioned above, a disadvantage of the use of the BLEW-48 sequence is the long cycle time, which results in a correspondingly long dwell-time in the t1 dimension. This can be a problem when large dipolar couplings need to be measured, which leads to folding of the spectral lines in the indirect dimension. An obvious solution to this problem is to reduce the pulse width by increasing the r.f. power. For example, if the pulse width is reduced by half, then the BLEW-48 cycle time is reduced by half and the spectral width is doubled. But reducing the pulse width by half requires a fourfold increase in the r.f. power. This puts stringent requirements on the probe and contributes to sample heating. As an alternate approach, we have modified the HDELF sequence by adopting the method proposed by Kharkov et al.[39] for using SLF spectroscopy on solution-NMR spectrometers that use low r.f. power. The corresponding P-HDELF pulse sequence is shown in Figure 1b. As shown, the 1808 pulses on the I and S channels were used to refocus proton chemical shift evolution and retain only proton–carbon dipolar couplings. All the BLEW-48 blocks were used with normal power except for one block at the center of the t1 period. This block used a lower power than the power used in the rest of the blocks for alternate t1 increments and correspondingly led to longer pulse widths. It is assumed that the use of low-power r.f. for just one BLEW-48 block will not affect the homonuclear decoupling efficiency. With this modification, the desired reduction in dwell time can be achieved as follows: For a reduction of dwell time by half, the power for the center BLEW-48 block is reduced so that the pulse width becomes 1.5 times longer. This low-power pulse is applied for alternate increments of time t1 in the indirect dimension. The dwell-time increments of t1 for the normal use of the BLEW-48 pulse sequence and in the modified pulse sequence are illustrated in Table 1. It may be noted in Table 1 that for the normal pulse sequence the dwell time (t) is equal to one BLEW-48 block, whereas in the modified sequence it is t/2. Thus, the modified pulse sequence helps to double the spectral width without increasing the r.f. power applied to the coils. It is possible to have other methods to increment the spectral width by suitably adjusting the width of the BLEW-48 block at the center. Initially we applied the P-HDELF sequence to measure dipolar couplings of a standard liquid crystal sample. Subsequently, we demonstrated the utility of the sequence for samples that may have large dipolar couplings and for which the application of the original unmodified sequence will lead to folding of the spectral lines. We utilized the results obtained for a novel liquid crystal to calculate the order parameters of this system.
ChemPhysChem 2015, 16, 2199 – 2205
www.chemphyschem.org
Table 1. BLEW-48 pulse durations and dwell times used in the indirect dimension for the pulse sequences of Figure 1b. With constant r.f. power Increment Indirect dimension period 1 2 3 4
t 2
t 2
t p t ttptt t t tt 2p2 tt tttpttt
BLEW-48 cycles
Total time
3 4 5 6
3t 4t 5t 6t 3t
With reduced r.f. power for one BLEW-48 cycle of duration t0 ¼ 2 Increment Indirect dimension period BLEW-48 cycles Total time 1 2 3 4
t0
t0
t 2p2 t ttptt t t0 t0 t t 2 2p2 2 t t t tt 2p2 tt
3 4 4 5
7
2t 4t 9 2t 5t
Experimental Section A sample of 5CB was purchased from Sigma–Aldrich and was used without further purification. A liquid crystalline sample of 10-(4-((4(benzoyloxy)benzylidene)amino)phenyl)decyl benzoate (BBIAPDB) with three phenyl rings as the core and with a terminal alkoxy chain on one side that ended in a phenyl ring was kindly provided by Dr. T. Narasimhaswamy. The nematic phase for this sample ranges from 71.5 to 107.3 8C. The NMR spectra of both samples were recorded in their nematic phase by using a Bruker AV-III 500 MHz NMR spectrometer equipped with a double-resonance probe for static samples with a 5 mm solenoid coil. The proton and carbon resonance frequencies were 500.17 and 125.77 MHz, respectively. Initially, the proton-decoupled 13C NMR spectra of both samples were recorded by using both a Hartman–Hahn cross polarization (HHCP) pulse sequence and also the HDELF sequence shown in Figure 1a. For the BLEW-48 block applied to the proton channel, an r.f. power of 62.5 kHz was used to provide a 908 pulse width of 4 ms and a duration of 192 ms for one block. To obtain the 1D spectra, t1 and t2 values were equal to 192 ms. The carbon 908 pulse had a duration of 4 ms. During acquisition of the 13C spectrum, the small phase incremental alternation (SPINAL-64) heteronuclear decoupling sequence[40] was used with a proton r.f. power of 33 kHz. With 30 ms of acquisition time, eight scans were added with a long relaxation delay of 12 s between scans to avoid r.f. heating. The 2D spectra were acquired by using the P-HDELF pulse sequence shown in Figure 1b as a proton-evolved experiment. Two versions of the experiment were carried out on both samples, one with normal r.f. power for the center BLEW-48 block and another with reduced r.f. power. In the second case, the pulse width of the 908 pulses of the center block were increased to 6 ms, which increased the duration of this block to 288 ms. As detailed in Table 1, this provided a dwell time of 192 and 96 ms for the normal and low-power P-HDELF experiments, respectively. The t2 period was maintained the same as in the 1D experiment at 192 ms. In the case of the BBIAPDB sample, the selective averaging magic sandwich with polarization inversion (SAMPI-4) pulse sequence[33] was applied and the spectra were recorded for comparison. In each case, a 2D data set was collected with 128 t1 increments and 8 scans for each increment. The data set was double Fourier transformed to obtain a 2D spectrum with proton–carbon dipolar couplings along the vertical (F1) axis and carbon chemical shifts along the horizontal (F2) axis.
2201
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles 3. Results and Discussion Figure 2a shows the structure of the 5CB liquid crystal and the corresponding 1D 13C spectra obtained by using Hartmann– Hahn cross-polarization (Figure 2b) and with HDELF pulse sequences (Figure 2c, d). To obtain the spectra shown in Figure 2c and d, BLEW-12 and BLEW-48 pulse sequences, respectively, were used. It can be seen from the spectra that the HDELF scheme also works well as a polarization transfer scheme. However, the intensities of the peaks differ from what is observed for Hartmann–Hahn cross-polarization due to dipolar oscillations during t1 and t2 that are unhindered by spin-diffusion between protons due to the use of homonuclear dipolar decoupling. It is observed that Figure 2d has a better line shape and intensity profile compared with Figure 2c, which emphasizes the advantage of using BLEW-48 over BLEW-12. BLEW-48 has the effective field along the transverse plane, so the quadrature component does not contribute to any signal, which leads to better line shape. It was also observed in the 2D spectra that BLEW-48 provided sharper contours and
Figure 2. a) Structure of the 5CB liquid crystal and its 13C NMR spectra recorded at RT in its nematic phase by using b) Hartmann–Hahn cross-polarization, c) HDELF with BLEW-12, and d) HDELF with BLEW-48 pulse sequences.
ChemPhysChem 2015, 16, 2199 – 2205
www.chemphyschem.org
higher resolution and thus only the corresponding results are presented here. The use of HDELF as a PELF sequence applied to 5CB is illustrated in Figure 3. Both the normal r.f. power and the low r.f. power versions have been used and the corresponding 2D spectra are shown in Figure 3a and b, respectively. Though Figure 3b shows double the spectral range along the dipolar (F1) dimension, the spectra are identical. This was additionally verified by extracting contours across several carbon chemical shifts along the F1 dimension and comparing them. This demonstrates that the use of one low-power BLEW-48 block at the center of the t1 period does not adversely affect the results obtained. Additionally, the modification helped us in doubling the spectral range. Such a modification is not necessary for 5CB, for which the maximum CH dipolar coupling of 4.5 kHz is very much in the spectral range of 5.2 kHz for the unmodified pulse sequence. The utility of the proposed modification to the pulse sequence is exemplified in the case of the BBIAPDB liquid crystal sample, for which a dipolar coupling greater than 10 kHz is encountered. Figure 4a shows the structure of the BBIAPDB molecule, and the 1D 13C spectrum of this sample is shown in Figure 4b. The corresponding 2D spectra obtained by using both normal and low-power P-HDELF pulse sequences are shown in Figure 5a and b, respectively. A 2D spectrum (Figure 5c) obtained by using the SAMPI-4 pulse sequence is also shown for reference. Figure 5a and b appear similar except for the contours of C1 with its resonance frequency around d = 190 ppm. This carbon, with its CH vector essentially parallel to the long axis of the liquid crystal molecule, is expected to show a large dipolar coupling.[41] For the same reason, it also has a large chemical shift, similar to the other quaternary carbons of the three phenyl rings in the core. For all of them, the direction of the least-shielded chemical shift direction, namely the para axes of the phenyl rings, are nearly parallel to the liquid crystal orienting axis. In comparison, the other methine carbons of the phenyl rings, for which the CH vector is inclined very close to the magic angle with respect to the long axis, show a small dipolar coupling. Cross-sections along the dipolar dimension for one of the aromatic carbons (C13) and for carbon C1 are depicted in Figure 6. For C13, identical dipolar coupling patterns are observed in the normal (Figure 6a, top trace) and the low-power (Figure 6a, middle trace) pulse sequences. The observed characteristic doubletlike peak patterns with dipolar couplings of 1.74 and 1.29 kHz are the result of the PELF character of the pulse sequence, with two peaks that arise from dipolar couplings of C13 to its own bonded proton and to the proton at the ortho position attached to C12, respectively. The single peak in the SAMPI-4 spectrum at 2.17 kHz, shown as the bottom trace of Figure 6a, appears as expected[16] as the square root of the sum of the squares of the peak positions observed in the top traces. Similar cross-sections for C1 are shown in Figure 6b. In this figure, the peak appearing at 10.42 kHz in the middle trace corresponds to the dipolar coupling of C1 to its attached proton. This peak appears in the top trace at 4.93 kHz as a folded peak due to insufficient spectral range in the normal P-HDELF pulse sequence. The low-frequency peaks in the top and middle
2202
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles of separate pairs of peaks for each pair of protons coupled to the carbon with different dipolar couplings, which leads to resolution of dipolar couplings of different magnitudes experienced by the same carbon. In the case of the novel liquid crystalline molecule BBIAPDB, the P-HDELF spectrum provides the order parameters of the terminal phenyl ring straight away. From the dipolar coupling of 10.42 kHz observed in the spectrum in Figure 6b for C1, the order parameter S0 zz of the para axis of the corresponding phenyl ring can be obtained as the ratio of the experimental dipolar coupling to the dipolar coupling of a rigid CH pair separated by a disFigure 3. 2D P-HDELF spectra of the 5CB liquid crystal a) with constant r.f. power and b) with low r.f. power tance of 1.1 æ and equal to for the middle BLEW-48 block (see Figure 1b). Aromatic (left) and aliphatic (right) regions are shown separate22.68 kHz. A S0 zz value of 0.46 for ly. this ring was obtained. It is possible to calculate the two local order parameters S0 zz and S0 xx ¢ S0 yy for each of the phenyl rings by considering the experimentally measured dipolar couplings. They are obtained by using the following relationship:[14] DCH ¼ D0
1 2
¦¨ ¦ 1¨ ð3 cos2 qz ¢ 1ÞS0 zz þ 2 cos2 qx ¢ cos2 qy S0 xx ¢ S0 yy ð1Þ
h m g g
Figure 4. a) Structure of the liquid crystal 10-(4-(4-((4-(benzoyloxy)benzylideneamino)phenoxy)decyl benzoate (BBIAPDB) and b) the 1D 13C NMR spectrum of this liquid crystal in its nematic phase recorded at 93 8C.
traces have nearly identical values in both spectra, equal to 1.1 kHz. These peaks appear due to coupling of C1 to its ortho protons. Again, the SAMPI-4 spectrum (bottom trace) shows a peak at 10.93 kHz, which is larger than the value of 10.42 kHz obtained from P-HDELF due to combined evolutions under different dipolar couplings of the carbon. Thus, the low r.f. power modification of the HDELF pulse sequence promises to be a convenient means of estimating both large and small dipolar couplings at the same time. The utility of the pulse sequence as a PELF sequence is also seen from the appearance ChemPhysChem 2015, 16, 2199 – 2205
www.chemphyschem.org
in which D0 ¼ 04prH 3 C , with gH and gC being the gyromagnetic CH ratios of 1H and 13C nuclei, respectively. rCH is the distance between the 13C and 1H nuclei and qx, qy, and qz are the angles the CH vector makes with the coordinate axes. For each phenyl ring, we have chosen the z axis to be the para axis of the ring, the x axis to be in the plane of the ring, and the y axis to be perpendicular to the plane. The dipolar couplings of each carbon to its attached proton and to protons in the ortho position were considered. The complete list of dipolar couplings utilized are given in Table 2. From these values, the order parameters were estimated for each of the phenyl rings and the obtained values are given in Table 3. It is seen from the table that ring III has the largest order parameters. Phenyl ring IV, connected by the flexible chain, shows the lowest order parameter due to additional mobility compared with the core. It is interesting to note that ring I, with an order parameter of 0.46 for the para axis calculated in a rigorous manner by considering all the dipolar couplings available for the ring, matches well with the value calculated by considering only the dipolar coupling of C1 to its attached proton. Thus in the context of many novel liquid crystalline systems being synthesized, the method proposed herein promises to be of significant advantage in providing microscopic information about structure and order in these systems.
2203
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles 4. Conclusion Currently, a variety of novel molecular constructs are being considered for the design of liquid crystals. In this context, 1 H–13C dipolar couplings provide a convenient means of estimating the dynamics and order of these systems. We have considered here a 2D NMR experiment for the estimation of dipolar couplings in two different liquid crystals. To address the problem of a small spectral window in the dipolar dimension that could result in the folding of peaks that correspond to large dipolar couplings, a modification to the pulse sequence has been implemented. This involves reducing the r.f. power used for homonuclear decoupling in the indirect dimension for a short duration. This approach has the possibility of increasing the spectral window to any desired value. It is noted that the reduction in r.f. power for one BLEW-48 pulse cycle, which makes the Figure 5. a, b) 2D P-HDELF spectra of BBIAPDB at 93 8C recorded with a) constant r.f. power and b) with low r.f. cycle 1.5 times longer, does not power. c) 2D- SAMPI-4 spectrum of the same sample. alter the quality of the spectrum, and at the same time enables the spectral width to be doubled. The modified pulse scheme has been applied to Table 2. 1H–13C dipolar couplings of the phenyl ring carbons of the liquid a novel liquid crystal with benzene rings at either end of the crystal BBIAPDB. molecule. Dipolar couplings for the core and terminal phenyl rings have been obtained. These values and the estimated 13 [a] Phenyl rings Carbon label C chemical shift [ppm] DCH [kHz] order parameters of the phenyl rings are found to vary over ring I 1 190.2 10.42, 1.1 a wide range due to the way the molecule has been construct2 151.7 1.52, 1.24 ed. Comparison of values estimated by other previously known 3 150.8 1.48, 1.48 4 178.8 1.22, 1.22 methods shows a good match, which indicates the utility of ring II 6 204.0 1.37, 1.37 the proposed scheme as a routine experimental tool.
ring III
ring IV
7 8 9 11 12 13 14 16 17 18 19
142.8 153.2 186.1 191.3 140.0 134.1 206.3 179.3 141.9 140.6 148.8
1.35, 1.56, 1.12, 1.37, 1.76, 1.74, 1.33, 1.20, 0.63, 0.61, 1.50,
1.35 1.56 1.12 1.37 1.31 1.29 1.33 1.20 0.63 0.61 1.50
Table 3. Order parameters of the phenyl rings of the liquid crystal BBIAPDB.
[a] Dipolar couplings to both directly attached protons and to protons at the ortho positions are shown.
ChemPhysChem 2015, 16, 2199 – 2205
www.chemphyschem.org
2204
Aromatic rings
S0 zz
S0 xx ¢ S0 yy
ring I ring II ring III ring IV
0.46 0.010 0.51 0.014 0.54 0.015 0.06 0.003
0.028 0.002 0.029 0.005 0.038 0.007 0.002 0.001
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles
Figure 6. 1D 1H–13C cross-sections of BBIAPDB along the dipolar dimension of a) C13 and b) C1 resonances extracted from the 2D spectra obtained with the following pulse sequences: P-HDELF with constant r.f. power (top trace); P-HDELF with low r.f. power (middle trace); and SAMPI-4 (bottom trace).
Acknowledgements The use of Bruker AV-III-500 WB NMR spectrometers, funded by the Department of Science and Technology, New Delhi at the NMR Research Centre, Indian Institute of Science, Bangalore, India, is gratefully acknowledged. The authors would like to thank Dr. T. Narasimhaswamy, Central Leather Research Institute, Adyar, Chennai, for the BBIAPDB liquid crystal sample and also Dr. V. S. Manu for useful discussions. Keywords: heteronuclear dipolar couplings · liquid crystals · local field spectroscopy · NMR spectroscopy · order parameters [1] J. W. Goodby, I. M. Saez, S. J. Cowling, J. S. Gasowska, R. A. MacDonald, S. Sia, P. Watson, K. J. Toyne, M. Hird, R. A. Lewis, S.-E. Lee, V. Vaschenko, Liq. Cryst. 2009, 36, 567 – 605. [2] T. Kato, N. Mizoshita, K. Kishimoto, Angew. Chem. Int. Ed. 2006, 45, 38 – 68; Angew. Chem. 2006, 118, 44 – 74. [3] G. W. Gray, Advances in Liquid Crystal Materials for Applications, BDH monograph, 1978. [4] R. A. Reddy, C. Tschierske, J. Mater. Chem. 2006, 16, 907 – 961.
ChemPhysChem 2015, 16, 2199 – 2205
www.chemphyschem.org
[5] D. Demus, Mol. Cryst. Liq. Cryst. Sci. Technol. Sect. A 2001, 364, 25 – 91. [6] A. Yamaguchi, I. Nishiyama, J. Yamamoto, H. Yokoyama, A. Yoshizawa, J. Mater. Chem. 2005, 15, 280 – 288. [7] C. Tschierske, Curr. Opin. Colloid Interface Sci. 2002, 7, 69 – 80. [8] A. Naito, A. Ramamoorthy in Atomistic-Resolution Structural Studies of Liquid Crystalline Materials Using Solid-State NMR Techniques (Ed.: A. Ramamoorthy), Springer, Amsterdam, 2007, pp. 85 – 116. [9] R. Y. Dong, NMR Spectroscopy in Liquid Crystalline and Ordered Phases, Wiley, Hoboken, 2012. [10] M. F. Brown, A. A. Nevzorov, Colloids Surf. A 1999, 158, 281 – 298. [11] J. Ulmius, H. Wennerstrçm, G. Lindblom, G. Arvidson, Biochim. Biophys. Acta Biomembr. 1975, 389, 197 – 202. [12] K. Chikae, Y. Shohei, T. Yoichi, I. Ken, T. Hideo, Jpn. J. Appl. Phys. 2002, 41, 6080 – 6083. [13] A. J. Montana, B. P. Dailey, J. Chem. Phys. 1977, 66, 989 – 994. [14] B. M. Fung, Prog. Nucl. Magn. Reson. Spectrosc. 2002, 41, 171 – 186. [15] J. Courtieu, J. P. Bayle, B. M. Fung, Prog. Nucl. Magn. Reson. Spectrosc. 1994, 26, 141 – 169. [16] R. Pratima, K. V. Ramanathan, J. Magn. Reson. Ser. A 1996, 118, 7 – 10. [17] H. Zimmermann, V. Bader, R. Poupko, E. J. Wachtel, Z. Luz, J. Am. Chem. Soc. 2002, 124, 15286 – 15301. [18] K. V. Ramanathan, N. Sinha in Current Developments in Solid State NMR Spectroscopy, Vol VIII (Eds.: N. Mueller, P. K. Madhu), Springer-Verlag, Wien, 55-68, 2003. [19] S. V. Dvinskikh, H. Zimmermann, A. Maliniak, D. Sandstrçm, J. Magn. Reson. 2003, 163, 46 – 55. [20] T. Narasimhaswamy, D.-K. Lee, K. Yamamoto, N. Somanathan, A. Ramamoorthy, J. Am. Chem. Soc. 2005, 127, 6958 – 6959. [21] M. K. Reddy, E. Varathan, N. P. Lobo, B. B. Das, T. Narasimhaswamy, K. V. Ramanathan, J. Phys. Chem. C 2014, 118, 15044 – 15053. [22] R. Y. Dong, J. Phys. Chem. B 2009, 113, 1933 – 1939. [23] R. K. Hester, J. L. Ackerman, B. L. Neff, J. S. Waugh, Phys. Rev. Lett. 1976, 36, 1081 – 1083. [24] C. H. Wu, A. Ramamoorthy, S. J. Opella, J. Magn. Reson. Ser. A 1994, 109, 270 – 272. [25] D. S. Thiriot, A. A. Nevzorov, L. Zagyanskiy, C. H. Wu, S. J. Opella, J. Mol. Biol. 2004, 341, 869 – 879. [26] S. Esteban-Martn, E. Strandberg, G. Fuertes, A. S. Ulrich, J. Salgado, Biophys. J. 2009, 96, 3233 – 3241. [27] S. V. Dvinskikh, H. Zimmermann, A. Maliniak, D. Sandstrçm, J. Magn. Reson. 2003, 164, 165 – 170. [28] P. Caravatti, G. Bodenhausen, R. R. Ernst, Chem. Phys. Lett. 1982, 89, 363 – 367. [29] T. Nakai, T. Terao, Magn. Reson. Chem. 1992, 30, 42 – 44. [30] K. Schmidt-Rohr, D. Nanz, L. Emsley, A. Pines, J. Phys. Chem. 1994, 98, 6668 – 6670. [31] B. M. Fung, K. Ermolaev, Y. Yu, J. Magn. Reson. 1999, 138, 28 – 35. [32] D. K. Lee, T. Narasimhaswamy, A. Ramamoorthy, Chem. Phys. Lett. 2004, 399, 359 – 362. [33] A. A. Nevzorov, S. J. Opella, J. Magn. Reson. 2007, 185, 59 – 70. [34] S. Jayanthi, N. Sinha, K. V. Ramanathan, J. Magn. Reson. 2010, 207, 206 – 212. [35] S. Jayanthi, P. K. Madhu, N. D. Kurur, K. V. Ramanathan, Chem. Phys. Lett. 2007, 439, 407 – 411. [36] S. Jayanthi, P. K. Madhu, K. V. Ramanathan, J. Phys. Chem. A 2008, 112, 11159 – 11164. [37] G. A. Morris, R. Freeman, J. Am. Chem. Soc. 1979, 101, 760 – 762. [38] D. P. Burum, M. Linder, R. R. Ernst, J. Magn. Reson. 1981, 44, 173 – 188. [39] B. B. Kharkov, V. I. Chizhik, S. V. Dvinskikh, J. Magn. Reson. 2012, 223, 73 – 79. [40] B. M. Fung, A. K. Khitrin, K. Ermolaev, J. Magn. Reson. 2000, 142, 97 – 101. [41] M. K. Reddy, K. S. Reddy, T. Narasimhaswamy, B. B. Das, N. P. Lobo, K. V. Ramanathan, New J. Chem. 2013, 37, 3195 – 3206.
Received: March 23, 2015 Published online on May 26, 2015
2205
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles
Measurement of Large Dipolar Couplings of a Liquid Crystal with Terminal Phenyl Rings and Estimation of the Order Parameters R. V. Sudheer Kumar[a, b] and Krishna V. Ramanathan*[a] NMR spectroscopy is a powerful means of studying liquid-crystalline systems at atomic resolutions. Of the many parameters that can provide information on the dynamics and order of the systems, 1H–13C dipolar couplings are an important means of obtaining such information. Depending on the details of the molecular structure and the magnitude of the order parameters, the dipolar couplings can vary over a wide range of values. Thus the method employed to estimate the dipolar couplings should be capable of estimating both large and small dipolar couplings at the same time. For this purpose, we
consider here a two-dimensional NMR experiment that works similar to the insensitive nuclei enhanced by polarization transfer (INEPT) experiment in solution. With the incorporation of a modification proposed earlier for experiments with low radio frequency power, the scheme is observed to enable a wide range of dipolar couplings to be estimated at the same time. We utilized this approach to obtain dipolar couplings in a liquid crystal with phenyl rings attached to either end of the molecule, and estimated its local order parameters.
1. Introduction Liquid crystals are the quintessential partially ordered systems, and find uses in a variety of applications. Of these systems, the design of thermotropic liquid crystals has undergone significant changes over the years. In the conventional approach, the molecules were either rod-like or disk-like.[1–3] In recent times, however, many systems with different shapes, such as bent core, l, H, and S, have also been proposed and exhibit interesting novel phases.[4–6] In this scenario, NMR spectroscopy has been found to be a powerful tool for investigating the structure, topology, and order that provide correlations between the molecular shape and the mesophase properties of the material. The design of liquid-crystalline molecules involves one or more core units connected to flexible chains.[1–3, 7] The study of the core units has drawn significant attention due to the fact that they give rigidity to the molecule and are also involved in intermolecular interactions. NMR spectroscopy has been highly useful in the study of such systems because it has been shown to provide information with atomistic resolution.[8, 9] Although nuclei like 1H, 2H, and 19F have been utilized for investigating such systems,[10–13] in recent years the naturalabundance 13C NMR spectroscopic technique is being increasingly used due to the fact that this approach does not require special sample preparation, such as isotope labeling, and provides high-resolution spectra in the oriented phase.[14–21] It is possible to obtain two anisotropic parameters from 13C NMR [a] R. V. S. Kumar, Prof. K. V. Ramanathan NMR Research Centre, Indian Institute of Science Bangalore 560012 (India) E-mail: [email protected] [b] R. V. S. Kumar Department of Physics, Indian Institute of Science Bangalore 560012 (India)
ChemPhysChem 2015, 16, 2199 – 2205
spectroscopy. The first is the alignment-induced chemical shift (AIS), which is the difference in the chemical shift of the carbon in the isotropic phase to that in the liquid crystalline phase, and arises from the alignment of the molecule in the magnetic field. The magnitude of the AIS depends on two parameters, namely, the order parameter and the chemical shift anisotropy tensors in a complex manner.[15, 22] The second parameter is the 1H–13C heteronuclear dipolar coupling. The magnitude of the dipolar coupling, like AIS, also depends on the magnitude and the orientation of the order parameter and dipolar tensors.[15, 21] For a rigid molecular structure, the magnitude and the direction of the principal components of the dipolar tensor are known. Thus dipolar couplings provide the local order parameters in a straightforward manner. Experimentally, however, the determination of the dipolar couplings is more involved and requires the use of the separated local field (SLF) 2D NMR method.[23] In this method, the chemical shift of the rare spin (S) is presented along the horizontal axis and the heteronuclear dipolar coupling of the S spin to another spin I, usually a proton, is depicted along the vertical axis. The experiment also involves removal of homonuclear couplings between protons while evolution under I–S coupling is carried out. There are several variations of the SLF experiment, each with relative advantages and disadvantages. These methods generally differ in the manner in which the evolution under the heteronuclear I–S interactions occurs. In the conventional SLF method,[23] the S spin evolution takes place under multiple I–S couplings, which results in broadening of the resonances due to overlapped multiplet structures. Conversely, the polarization inversion spin exchange at magic angle (PISEMA)[24] method has been extensively utilized for obtaining isolated I–S couplings, such as those that involve 15N–1H in peptides and
2199
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles proteins.[25, 26] In the presence of multiple I–S couplings, as happosed by Kharkov et al.[39] In this modification, a low r.f. field is 1 13 pens in the study of H– C dipolar couplings in liquid crysused for one BLEW-48 cycle, which enables the dwell time to tals,[27] a method that has significant advantages is the protonbe halved. We initially demonstrate this approach for the encoded local field (PELF) spectroscopy experiment.[28–31] In this known liquid crystal 4-cyano-4’-pentylbiphenyl (5CB). Subseapproach, the abundant I spins are allowed to evolve in the quently, we use it to obtain the dipolar couplings in a novel presence of their couplings to the rare nucleus S. Thus, an I liquid crystal with four phenyl groups and with an aliphatic spin has only a single S spin coupling. All the I–S dipolar couchain inserted between two of the phenyl groups, and demonplings are observed with the corresponding doublets separatstrate its utility in measuring large dipolar couplings and the ed out according to their magnitude. This results in narrow corresponding order parameters in a straightforward manner. and well-resolved spectral lines that correspond to different I–S spin pairs. The basic approaches mentioned above have been 2. Pulse Sequence further modified to address some of the problems encountered in practical applications, such as the heating effects due The basic pulse sequence is shown in Figure 1a, and utilizes to the continuous application of radio frequency (r.f.), the the BLEW-48 sequence for both homonuclear decoupling and effect of proton r.f. offsets, and so on.[32–34] Of these methods, for heteronuclear polarization transfer. As a consequence of we consider here the dipolar assisted polarization transfer the application of the BLEW-48 pulse sequence, the heteronu(DAPT) pulse sequence proposed recently,[35] which has the adclear Hamiltonian (HIS) has the form k2DISIXSZ, in which DIS is the vantage of being used either as an S-nucleus-evolved SLF-like heteronuclear dipolar coupling between the I and S spins and pulse sequence[36] or an I-spin-evolved PELF-like pulse sek is the scaling factor. For the BLEW-48 sequence, the scaling quence. This sequence, like the insensitive nuclei enhanced by factor was 0.424 and the effective field was along the transpolarization transfer (INEPT) pulse sequence[37] used in solution verse plane. During the initial BLEW-48 period P0, the IZ magNMR, uses only a simple 908 pulse on the S spin, which thus netization evolved under the scaled heteronuclear dipolar coueliminates the high r.f. power required in the case of cross-popling to produce a two-spin-order term, 2IYSZsin (kDISt1). This larization-based sequences with simultaneous RF irradiation was converted to 2IYSYsin (kDISt1) by a 908 pulse applied on the applied on the I and the S spin channels. Furthermore, usage S channel. Subsequent evolution during t2 took place under of homonuclear decoupling throughout ensures that the transthe phase-shifted BLEW-48 sequence P90, in which the phases fer of polarization between I and S spins takes place between of all the pulses were shifted by 908 degrees with respect to P0 coupled I–S spin pair only, which avoids indirect transfers to provide an effective HIS of k2DISIYSZ. This refocused the twothrough the I–I coupling pathway. In the original presentation spin-order term 2IYSY to create the observable single-spin SX of this sequence, a BLEW-12[38] pulse sequence was used for operator term. During the acquisition period, the S spin signal homonuclear proton decoupling. BLEW-12 is observed to be observed with heteronuclear dipolar decoupling after the less efficient in comparison with other sequences, such as chemical shift evolution during t2 was refocused with a 1808 BLEW-48,[38] and also has an effective field inclined at an angle pulse applied on the S spin channel. It may be noted that the (q) of 26.68 with reference to the z axis. The latter condition reobserved S spin signal was modulated by dipolar evolution quires additional pre-pulses to be applied to appropriately that took place during both t1 and t2 periods, which resulted align the proton magnetization and maximize signals. Conin a signal with the form SXsin (kDISt1) sin (kDISt2). Interestingly, versely, BLEW-48 has an effective field in the transverse plane, during the t1 period the evolution was of the I (or proton) which is an advantage for experiments that involve proton chemical shifts or heteronuclear dipolar evolution with homonuclear decoupling. Herein, the DAPT experiment, modified to include the BLEW-48 pulses, has been exploited to obtain PELF spectra. However, the BLEW-48 pulse sequence has a long cycle time, which causes the dwell time in the dipolar dimension to increase. The increased dwell time is a disadvantage when large dipolar couplings are required to be measured because Figure 1. a) Pulse sequence for homonuclear decoupling evolved local field (HDELF) spectroscopy. P0 and P90 are it will result in folding. To cirBLEW-48 homonuclear decoupling pulse sequence blocks. The phases of all r.f. pulses are shifted by 908 between cumvent this problem, we have P0 and P90. b) HDELF used as a proton-evolved pulse sequence (P-HDELF). Use of reduced r.f. power for one BLEWadopted a modification to the 48 block at the center of the t1 period and increasing its duration by 1.5 times for alternate t1 increments decreasPELF sequence recently proes the dwell time by half in the indirect dimension (see main text and Table 1 for details). ChemPhysChem 2015, 16, 2199 – 2205
www.chemphyschem.org
2200
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles magnetization under the dipolar coupling DIS. This allows the experiment to be used as a proton-encoded experiment, with t1 used as the t1 variable. Conversely, evolution during t2 was the evolution of the S (or carbon) magnetization under DIS, which makes this an SLF experiment when t2 is used as the variable. In view of these possibilities, we refer to this pulse sequence as homonuclear decoupling evolved local field (HDELF) spectroscopy. The proton-evolved version is referred to as PHDELF and the carbon-evolved version as S-HDELF. As mentioned above, a disadvantage of the use of the BLEW-48 sequence is the long cycle time, which results in a correspondingly long dwell-time in the t1 dimension. This can be a problem when large dipolar couplings need to be measured, which leads to folding of the spectral lines in the indirect dimension. An obvious solution to this problem is to reduce the pulse width by increasing the r.f. power. For example, if the pulse width is reduced by half, then the BLEW-48 cycle time is reduced by half and the spectral width is doubled. But reducing the pulse width by half requires a fourfold increase in the r.f. power. This puts stringent requirements on the probe and contributes to sample heating. As an alternate approach, we have modified the HDELF sequence by adopting the method proposed by Kharkov et al.[39] for using SLF spectroscopy on solution-NMR spectrometers that use low r.f. power. The corresponding P-HDELF pulse sequence is shown in Figure 1b. As shown, the 1808 pulses on the I and S channels were used to refocus proton chemical shift evolution and retain only proton–carbon dipolar couplings. All the BLEW-48 blocks were used with normal power except for one block at the center of the t1 period. This block used a lower power than the power used in the rest of the blocks for alternate t1 increments and correspondingly led to longer pulse widths. It is assumed that the use of low-power r.f. for just one BLEW-48 block will not affect the homonuclear decoupling efficiency. With this modification, the desired reduction in dwell time can be achieved as follows: For a reduction of dwell time by half, the power for the center BLEW-48 block is reduced so that the pulse width becomes 1.5 times longer. This low-power pulse is applied for alternate increments of time t1 in the indirect dimension. The dwell-time increments of t1 for the normal use of the BLEW-48 pulse sequence and in the modified pulse sequence are illustrated in Table 1. It may be noted in Table 1 that for the normal pulse sequence the dwell time (t) is equal to one BLEW-48 block, whereas in the modified sequence it is t/2. Thus, the modified pulse sequence helps to double the spectral width without increasing the r.f. power applied to the coils. It is possible to have other methods to increment the spectral width by suitably adjusting the width of the BLEW-48 block at the center. Initially we applied the P-HDELF sequence to measure dipolar couplings of a standard liquid crystal sample. Subsequently, we demonstrated the utility of the sequence for samples that may have large dipolar couplings and for which the application of the original unmodified sequence will lead to folding of the spectral lines. We utilized the results obtained for a novel liquid crystal to calculate the order parameters of this system.
ChemPhysChem 2015, 16, 2199 – 2205
www.chemphyschem.org
Table 1. BLEW-48 pulse durations and dwell times used in the indirect dimension for the pulse sequences of Figure 1b. With constant r.f. power Increment Indirect dimension period 1 2 3 4
t 2
t 2
t p t ttptt t t tt 2p2 tt tttpttt
BLEW-48 cycles
Total time
3 4 5 6
3t 4t 5t 6t 3t
With reduced r.f. power for one BLEW-48 cycle of duration t0 ¼ 2 Increment Indirect dimension period BLEW-48 cycles Total time 1 2 3 4
t0
t0
t 2p2 t ttptt t t0 t0 t t 2 2p2 2 t t t tt 2p2 tt
3 4 4 5
7
2t 4t 9 2t 5t
Experimental Section A sample of 5CB was purchased from Sigma–Aldrich and was used without further purification. A liquid crystalline sample of 10-(4-((4(benzoyloxy)benzylidene)amino)phenyl)decyl benzoate (BBIAPDB) with three phenyl rings as the core and with a terminal alkoxy chain on one side that ended in a phenyl ring was kindly provided by Dr. T. Narasimhaswamy. The nematic phase for this sample ranges from 71.5 to 107.3 8C. The NMR spectra of both samples were recorded in their nematic phase by using a Bruker AV-III 500 MHz NMR spectrometer equipped with a double-resonance probe for static samples with a 5 mm solenoid coil. The proton and carbon resonance frequencies were 500.17 and 125.77 MHz, respectively. Initially, the proton-decoupled 13C NMR spectra of both samples were recorded by using both a Hartman–Hahn cross polarization (HHCP) pulse sequence and also the HDELF sequence shown in Figure 1a. For the BLEW-48 block applied to the proton channel, an r.f. power of 62.5 kHz was used to provide a 908 pulse width of 4 ms and a duration of 192 ms for one block. To obtain the 1D spectra, t1 and t2 values were equal to 192 ms. The carbon 908 pulse had a duration of 4 ms. During acquisition of the 13C spectrum, the small phase incremental alternation (SPINAL-64) heteronuclear decoupling sequence[40] was used with a proton r.f. power of 33 kHz. With 30 ms of acquisition time, eight scans were added with a long relaxation delay of 12 s between scans to avoid r.f. heating. The 2D spectra were acquired by using the P-HDELF pulse sequence shown in Figure 1b as a proton-evolved experiment. Two versions of the experiment were carried out on both samples, one with normal r.f. power for the center BLEW-48 block and another with reduced r.f. power. In the second case, the pulse width of the 908 pulses of the center block were increased to 6 ms, which increased the duration of this block to 288 ms. As detailed in Table 1, this provided a dwell time of 192 and 96 ms for the normal and low-power P-HDELF experiments, respectively. The t2 period was maintained the same as in the 1D experiment at 192 ms. In the case of the BBIAPDB sample, the selective averaging magic sandwich with polarization inversion (SAMPI-4) pulse sequence[33] was applied and the spectra were recorded for comparison. In each case, a 2D data set was collected with 128 t1 increments and 8 scans for each increment. The data set was double Fourier transformed to obtain a 2D spectrum with proton–carbon dipolar couplings along the vertical (F1) axis and carbon chemical shifts along the horizontal (F2) axis.
2201
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles 3. Results and Discussion Figure 2a shows the structure of the 5CB liquid crystal and the corresponding 1D 13C spectra obtained by using Hartmann– Hahn cross-polarization (Figure 2b) and with HDELF pulse sequences (Figure 2c, d). To obtain the spectra shown in Figure 2c and d, BLEW-12 and BLEW-48 pulse sequences, respectively, were used. It can be seen from the spectra that the HDELF scheme also works well as a polarization transfer scheme. However, the intensities of the peaks differ from what is observed for Hartmann–Hahn cross-polarization due to dipolar oscillations during t1 and t2 that are unhindered by spin-diffusion between protons due to the use of homonuclear dipolar decoupling. It is observed that Figure 2d has a better line shape and intensity profile compared with Figure 2c, which emphasizes the advantage of using BLEW-48 over BLEW-12. BLEW-48 has the effective field along the transverse plane, so the quadrature component does not contribute to any signal, which leads to better line shape. It was also observed in the 2D spectra that BLEW-48 provided sharper contours and
Figure 2. a) Structure of the 5CB liquid crystal and its 13C NMR spectra recorded at RT in its nematic phase by using b) Hartmann–Hahn cross-polarization, c) HDELF with BLEW-12, and d) HDELF with BLEW-48 pulse sequences.
ChemPhysChem 2015, 16, 2199 – 2205
www.chemphyschem.org
higher resolution and thus only the corresponding results are presented here. The use of HDELF as a PELF sequence applied to 5CB is illustrated in Figure 3. Both the normal r.f. power and the low r.f. power versions have been used and the corresponding 2D spectra are shown in Figure 3a and b, respectively. Though Figure 3b shows double the spectral range along the dipolar (F1) dimension, the spectra are identical. This was additionally verified by extracting contours across several carbon chemical shifts along the F1 dimension and comparing them. This demonstrates that the use of one low-power BLEW-48 block at the center of the t1 period does not adversely affect the results obtained. Additionally, the modification helped us in doubling the spectral range. Such a modification is not necessary for 5CB, for which the maximum CH dipolar coupling of 4.5 kHz is very much in the spectral range of 5.2 kHz for the unmodified pulse sequence. The utility of the proposed modification to the pulse sequence is exemplified in the case of the BBIAPDB liquid crystal sample, for which a dipolar coupling greater than 10 kHz is encountered. Figure 4a shows the structure of the BBIAPDB molecule, and the 1D 13C spectrum of this sample is shown in Figure 4b. The corresponding 2D spectra obtained by using both normal and low-power P-HDELF pulse sequences are shown in Figure 5a and b, respectively. A 2D spectrum (Figure 5c) obtained by using the SAMPI-4 pulse sequence is also shown for reference. Figure 5a and b appear similar except for the contours of C1 with its resonance frequency around d = 190 ppm. This carbon, with its CH vector essentially parallel to the long axis of the liquid crystal molecule, is expected to show a large dipolar coupling.[41] For the same reason, it also has a large chemical shift, similar to the other quaternary carbons of the three phenyl rings in the core. For all of them, the direction of the least-shielded chemical shift direction, namely the para axes of the phenyl rings, are nearly parallel to the liquid crystal orienting axis. In comparison, the other methine carbons of the phenyl rings, for which the CH vector is inclined very close to the magic angle with respect to the long axis, show a small dipolar coupling. Cross-sections along the dipolar dimension for one of the aromatic carbons (C13) and for carbon C1 are depicted in Figure 6. For C13, identical dipolar coupling patterns are observed in the normal (Figure 6a, top trace) and the low-power (Figure 6a, middle trace) pulse sequences. The observed characteristic doubletlike peak patterns with dipolar couplings of 1.74 and 1.29 kHz are the result of the PELF character of the pulse sequence, with two peaks that arise from dipolar couplings of C13 to its own bonded proton and to the proton at the ortho position attached to C12, respectively. The single peak in the SAMPI-4 spectrum at 2.17 kHz, shown as the bottom trace of Figure 6a, appears as expected[16] as the square root of the sum of the squares of the peak positions observed in the top traces. Similar cross-sections for C1 are shown in Figure 6b. In this figure, the peak appearing at 10.42 kHz in the middle trace corresponds to the dipolar coupling of C1 to its attached proton. This peak appears in the top trace at 4.93 kHz as a folded peak due to insufficient spectral range in the normal P-HDELF pulse sequence. The low-frequency peaks in the top and middle
2202
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles of separate pairs of peaks for each pair of protons coupled to the carbon with different dipolar couplings, which leads to resolution of dipolar couplings of different magnitudes experienced by the same carbon. In the case of the novel liquid crystalline molecule BBIAPDB, the P-HDELF spectrum provides the order parameters of the terminal phenyl ring straight away. From the dipolar coupling of 10.42 kHz observed in the spectrum in Figure 6b for C1, the order parameter S0 zz of the para axis of the corresponding phenyl ring can be obtained as the ratio of the experimental dipolar coupling to the dipolar coupling of a rigid CH pair separated by a disFigure 3. 2D P-HDELF spectra of the 5CB liquid crystal a) with constant r.f. power and b) with low r.f. power tance of 1.1 æ and equal to for the middle BLEW-48 block (see Figure 1b). Aromatic (left) and aliphatic (right) regions are shown separate22.68 kHz. A S0 zz value of 0.46 for ly. this ring was obtained. It is possible to calculate the two local order parameters S0 zz and S0 xx ¢ S0 yy for each of the phenyl rings by considering the experimentally measured dipolar couplings. They are obtained by using the following relationship:[14] DCH ¼ D0
1 2
¦¨ ¦ 1¨ ð3 cos2 qz ¢ 1ÞS0 zz þ 2 cos2 qx ¢ cos2 qy S0 xx ¢ S0 yy ð1Þ
h m g g
Figure 4. a) Structure of the liquid crystal 10-(4-(4-((4-(benzoyloxy)benzylideneamino)phenoxy)decyl benzoate (BBIAPDB) and b) the 1D 13C NMR spectrum of this liquid crystal in its nematic phase recorded at 93 8C.
traces have nearly identical values in both spectra, equal to 1.1 kHz. These peaks appear due to coupling of C1 to its ortho protons. Again, the SAMPI-4 spectrum (bottom trace) shows a peak at 10.93 kHz, which is larger than the value of 10.42 kHz obtained from P-HDELF due to combined evolutions under different dipolar couplings of the carbon. Thus, the low r.f. power modification of the HDELF pulse sequence promises to be a convenient means of estimating both large and small dipolar couplings at the same time. The utility of the pulse sequence as a PELF sequence is also seen from the appearance ChemPhysChem 2015, 16, 2199 – 2205
www.chemphyschem.org
in which D0 ¼ 04prH 3 C , with gH and gC being the gyromagnetic CH ratios of 1H and 13C nuclei, respectively. rCH is the distance between the 13C and 1H nuclei and qx, qy, and qz are the angles the CH vector makes with the coordinate axes. For each phenyl ring, we have chosen the z axis to be the para axis of the ring, the x axis to be in the plane of the ring, and the y axis to be perpendicular to the plane. The dipolar couplings of each carbon to its attached proton and to protons in the ortho position were considered. The complete list of dipolar couplings utilized are given in Table 2. From these values, the order parameters were estimated for each of the phenyl rings and the obtained values are given in Table 3. It is seen from the table that ring III has the largest order parameters. Phenyl ring IV, connected by the flexible chain, shows the lowest order parameter due to additional mobility compared with the core. It is interesting to note that ring I, with an order parameter of 0.46 for the para axis calculated in a rigorous manner by considering all the dipolar couplings available for the ring, matches well with the value calculated by considering only the dipolar coupling of C1 to its attached proton. Thus in the context of many novel liquid crystalline systems being synthesized, the method proposed herein promises to be of significant advantage in providing microscopic information about structure and order in these systems.
2203
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles 4. Conclusion Currently, a variety of novel molecular constructs are being considered for the design of liquid crystals. In this context, 1 H–13C dipolar couplings provide a convenient means of estimating the dynamics and order of these systems. We have considered here a 2D NMR experiment for the estimation of dipolar couplings in two different liquid crystals. To address the problem of a small spectral window in the dipolar dimension that could result in the folding of peaks that correspond to large dipolar couplings, a modification to the pulse sequence has been implemented. This involves reducing the r.f. power used for homonuclear decoupling in the indirect dimension for a short duration. This approach has the possibility of increasing the spectral window to any desired value. It is noted that the reduction in r.f. power for one BLEW-48 pulse cycle, which makes the Figure 5. a, b) 2D P-HDELF spectra of BBIAPDB at 93 8C recorded with a) constant r.f. power and b) with low r.f. cycle 1.5 times longer, does not power. c) 2D- SAMPI-4 spectrum of the same sample. alter the quality of the spectrum, and at the same time enables the spectral width to be doubled. The modified pulse scheme has been applied to Table 2. 1H–13C dipolar couplings of the phenyl ring carbons of the liquid a novel liquid crystal with benzene rings at either end of the crystal BBIAPDB. molecule. Dipolar couplings for the core and terminal phenyl rings have been obtained. These values and the estimated 13 [a] Phenyl rings Carbon label C chemical shift [ppm] DCH [kHz] order parameters of the phenyl rings are found to vary over ring I 1 190.2 10.42, 1.1 a wide range due to the way the molecule has been construct2 151.7 1.52, 1.24 ed. Comparison of values estimated by other previously known 3 150.8 1.48, 1.48 4 178.8 1.22, 1.22 methods shows a good match, which indicates the utility of ring II 6 204.0 1.37, 1.37 the proposed scheme as a routine experimental tool.
ring III
ring IV
7 8 9 11 12 13 14 16 17 18 19
142.8 153.2 186.1 191.3 140.0 134.1 206.3 179.3 141.9 140.6 148.8
1.35, 1.56, 1.12, 1.37, 1.76, 1.74, 1.33, 1.20, 0.63, 0.61, 1.50,
1.35 1.56 1.12 1.37 1.31 1.29 1.33 1.20 0.63 0.61 1.50
Table 3. Order parameters of the phenyl rings of the liquid crystal BBIAPDB.
[a] Dipolar couplings to both directly attached protons and to protons at the ortho positions are shown.
ChemPhysChem 2015, 16, 2199 – 2205
www.chemphyschem.org
2204
Aromatic rings
S0 zz
S0 xx ¢ S0 yy
ring I ring II ring III ring IV
0.46 0.010 0.51 0.014 0.54 0.015 0.06 0.003
0.028 0.002 0.029 0.005 0.038 0.007 0.002 0.001
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles
Figure 6. 1D 1H–13C cross-sections of BBIAPDB along the dipolar dimension of a) C13 and b) C1 resonances extracted from the 2D spectra obtained with the following pulse sequences: P-HDELF with constant r.f. power (top trace); P-HDELF with low r.f. power (middle trace); and SAMPI-4 (bottom trace).
Acknowledgements The use of Bruker AV-III-500 WB NMR spectrometers, funded by the Department of Science and Technology, New Delhi at the NMR Research Centre, Indian Institute of Science, Bangalore, India, is gratefully acknowledged. The authors would like to thank Dr. T. Narasimhaswamy, Central Leather Research Institute, Adyar, Chennai, for the BBIAPDB liquid crystal sample and also Dr. V. S. Manu for useful discussions. Keywords: heteronuclear dipolar couplings · liquid crystals · local field spectroscopy · NMR spectroscopy · order parameters [1] J. W. Goodby, I. M. Saez, S. J. Cowling, J. S. Gasowska, R. A. MacDonald, S. Sia, P. Watson, K. J. Toyne, M. Hird, R. A. Lewis, S.-E. Lee, V. Vaschenko, Liq. Cryst. 2009, 36, 567 – 605. [2] T. Kato, N. Mizoshita, K. Kishimoto, Angew. Chem. Int. Ed. 2006, 45, 38 – 68; Angew. Chem. 2006, 118, 44 – 74. [3] G. W. Gray, Advances in Liquid Crystal Materials for Applications, BDH monograph, 1978. [4] R. A. Reddy, C. Tschierske, J. Mater. Chem. 2006, 16, 907 – 961.
ChemPhysChem 2015, 16, 2199 – 2205
www.chemphyschem.org
[5] D. Demus, Mol. Cryst. Liq. Cryst. Sci. Technol. Sect. A 2001, 364, 25 – 91. [6] A. Yamaguchi, I. Nishiyama, J. Yamamoto, H. Yokoyama, A. Yoshizawa, J. Mater. Chem. 2005, 15, 280 – 288. [7] C. Tschierske, Curr. Opin. Colloid Interface Sci. 2002, 7, 69 – 80. [8] A. Naito, A. Ramamoorthy in Atomistic-Resolution Structural Studies of Liquid Crystalline Materials Using Solid-State NMR Techniques (Ed.: A. Ramamoorthy), Springer, Amsterdam, 2007, pp. 85 – 116. [9] R. Y. Dong, NMR Spectroscopy in Liquid Crystalline and Ordered Phases, Wiley, Hoboken, 2012. [10] M. F. Brown, A. A. Nevzorov, Colloids Surf. A 1999, 158, 281 – 298. [11] J. Ulmius, H. Wennerstrçm, G. Lindblom, G. Arvidson, Biochim. Biophys. Acta Biomembr. 1975, 389, 197 – 202. [12] K. Chikae, Y. Shohei, T. Yoichi, I. Ken, T. Hideo, Jpn. J. Appl. Phys. 2002, 41, 6080 – 6083. [13] A. J. Montana, B. P. Dailey, J. Chem. Phys. 1977, 66, 989 – 994. [14] B. M. Fung, Prog. Nucl. Magn. Reson. Spectrosc. 2002, 41, 171 – 186. [15] J. Courtieu, J. P. Bayle, B. M. Fung, Prog. Nucl. Magn. Reson. Spectrosc. 1994, 26, 141 – 169. [16] R. Pratima, K. V. Ramanathan, J. Magn. Reson. Ser. A 1996, 118, 7 – 10. [17] H. Zimmermann, V. Bader, R. Poupko, E. J. Wachtel, Z. Luz, J. Am. Chem. Soc. 2002, 124, 15286 – 15301. [18] K. V. Ramanathan, N. Sinha in Current Developments in Solid State NMR Spectroscopy, Vol VIII (Eds.: N. Mueller, P. K. Madhu), Springer-Verlag, Wien, 55-68, 2003. [19] S. V. Dvinskikh, H. Zimmermann, A. Maliniak, D. Sandstrçm, J. Magn. Reson. 2003, 163, 46 – 55. [20] T. Narasimhaswamy, D.-K. Lee, K. Yamamoto, N. Somanathan, A. Ramamoorthy, J. Am. Chem. Soc. 2005, 127, 6958 – 6959. [21] M. K. Reddy, E. Varathan, N. P. Lobo, B. B. Das, T. Narasimhaswamy, K. V. Ramanathan, J. Phys. Chem. C 2014, 118, 15044 – 15053. [22] R. Y. Dong, J. Phys. Chem. B 2009, 113, 1933 – 1939. [23] R. K. Hester, J. L. Ackerman, B. L. Neff, J. S. Waugh, Phys. Rev. Lett. 1976, 36, 1081 – 1083. [24] C. H. Wu, A. Ramamoorthy, S. J. Opella, J. Magn. Reson. Ser. A 1994, 109, 270 – 272. [25] D. S. Thiriot, A. A. Nevzorov, L. Zagyanskiy, C. H. Wu, S. J. Opella, J. Mol. Biol. 2004, 341, 869 – 879. [26] S. Esteban-Martn, E. Strandberg, G. Fuertes, A. S. Ulrich, J. Salgado, Biophys. J. 2009, 96, 3233 – 3241. [27] S. V. Dvinskikh, H. Zimmermann, A. Maliniak, D. Sandstrçm, J. Magn. Reson. 2003, 164, 165 – 170. [28] P. Caravatti, G. Bodenhausen, R. R. Ernst, Chem. Phys. Lett. 1982, 89, 363 – 367. [29] T. Nakai, T. Terao, Magn. Reson. Chem. 1992, 30, 42 – 44. [30] K. Schmidt-Rohr, D. Nanz, L. Emsley, A. Pines, J. Phys. Chem. 1994, 98, 6668 – 6670. [31] B. M. Fung, K. Ermolaev, Y. Yu, J. Magn. Reson. 1999, 138, 28 – 35. [32] D. K. Lee, T. Narasimhaswamy, A. Ramamoorthy, Chem. Phys. Lett. 2004, 399, 359 – 362. [33] A. A. Nevzorov, S. J. Opella, J. Magn. Reson. 2007, 185, 59 – 70. [34] S. Jayanthi, N. Sinha, K. V. Ramanathan, J. Magn. Reson. 2010, 207, 206 – 212. [35] S. Jayanthi, P. K. Madhu, N. D. Kurur, K. V. Ramanathan, Chem. Phys. Lett. 2007, 439, 407 – 411. [36] S. Jayanthi, P. K. Madhu, K. V. Ramanathan, J. Phys. Chem. A 2008, 112, 11159 – 11164. [37] G. A. Morris, R. Freeman, J. Am. Chem. Soc. 1979, 101, 760 – 762. [38] D. P. Burum, M. Linder, R. R. Ernst, J. Magn. Reson. 1981, 44, 173 – 188. [39] B. B. Kharkov, V. I. Chizhik, S. V. Dvinskikh, J. Magn. Reson. 2012, 223, 73 – 79. [40] B. M. Fung, A. K. Khitrin, K. Ermolaev, J. Magn. Reson. 2000, 142, 97 – 101. [41] M. K. Reddy, K. S. Reddy, T. Narasimhaswamy, B. B. Das, N. P. Lobo, K. V. Ramanathan, New J. Chem. 2013, 37, 3195 – 3206.
Received: March 23, 2015 Published online on May 26, 2015
2205
Ó 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
