JOURNAL
OF MAGNETIC
RESONANCE
22, 327-332 (1976)
Application of Correlation NMR Spectroscopy to CIDNP YOJI ARATA Depuriment
of Chemistry,
The University
of Tokyo,
Honpo,
Tokvo
I IS, Jrrpun
Received September 3, 1975 Correlation NMR spectroscopy has been applied to ClDNP in a weakly coupled spin system of the methyl and methylene protons of 4-ethylpyridine produced by spin-selective processes in a reaction between 4-picoline-N-oxide and acetic anhydride. Different types of spectral pattern observed under different experimental conditions have been closely examined. It has been shown how one can take advantage of the correlation NMR method to investigate a weakly coupled nonequilibrium spin system produced by CIDNP using a large flip angle, which leads to the significant gain in sensitivity as compared to the pulse FT method. INTRODUCTION
and co-workers (I, 2) have shown that in pulse FT experiments of nonequilibrium spin states, a sufficiently small flip angle is necessary for the observation of the same relative intensities as obtained in a slow-passage low-power experiment. Although the pulse FT method is potentially a very promising technique for the investigation of transient phenomena such as CIDNP, the above limitation may at least to some extent hamper the usefulness of the pulse FT method, because a small flip angle would unavoidably result in a lower signal-to-noise ratio. Correlation NMR technique is generally comparable in sensitivity with the pulse FT method (3-5). In correlation spectroscopy, the resonance condition is met sequentially, in contrast to the pulse FT method, where all spins are excited simultaneously. Therefore, it appears that the correlation technique is more suitable than the pulse FT method for the investigation of nonequilibrium spin states. However, Ferretti has shown that when a large flip angle is used to obtain the correlation NM R spectrum of a coupled spin system, an anomalous intensity pattern is observed under a specified condition (6). In a weakly coupled spin system, it is in general possible in the correlation mode to cover a part of the spin system, leaving the remaining portion of it intact. In the present work, correlation NMR spectroscopy has been applied to CI DNP in a weakly coupled case. As an illustrative example, correlation NMR spectra of 4-ethylpyridine produced by spin-selective processes in a reaction between 4-picoline-N-oxide and acetic anhydride (7) will be chosen. Different types of intensity pattern obtained for the methyl and methylene protons under different experimental conditions will be closely examined. It will be demonstrated how one can take advantage of the correlation NMR method to investigate a nonequilibrium state in CIDNP in a weakly coupled case using a large flip angle, which leads to the significant gain in sensitivity as compared to the Ernst
Copyright Q 1976 by Academic Press, Inc. All rights of reproduction in any form reserved. Printed in Great Britain
327
328
YOJI ARATA
pulse FT method. Scope and limitations discussed.
of the method presented here will also be
EXPERIMENTAL
A 5-mm o.d. standard tube containing 70 mg of finely powdered 4-picoline-N-oxide dissolved (and partly suspended) in 0.5 ml of acetic anhydride was inserted into an NMR probe preheated at 95-100°C. The rate of production of 4-ethylpyridine was controlled by the probe temperature. A JEOL PS-100 NMR spectrometer operating at 100 MHz was modified for correlation spectroscopy as described previously (5). The amplitude of the radiofrequency field was determined by the method of transient nutation (8), and used to estimate the flip angle in the correlation mode. Pulse FT NMR spectra were taken at 100 MHz using a JEOL PFT-100 spectrometer. RESULTS
AND
DISCUSSION’
A flip angle of approximately 90” was employed on the correlation mode to record NMR spectra shown in Fig. 1 of a reaction mixture of 4-picoline-N-oxide and acetic
I D-----r cI c----------l
(a)
4
I
91
B-r-----
5
(b)
Ev
3
1. Correlation NMR spectra of a reaction mixture of 4-picoline-N-oxide and acetic anhydride obtained by employing a flip angle of approximately 90”. After insertion of a sample into a probe preheated at 98”, the transient was recorded at 70 set (A), 90 set (B), 110 set (C), 130 set (D), and 150 set (E). (a) Sweep from high to low field was started at 6 = 2.10; R = 205.5 Hz/set, T, = 2.048 set, P = 1024, F= 300 Hz, W= 0.9 sec. The methylene quartet is at 6 = 2.58, 2.50, 2.42, and 2.34, and a signal at 6 = 4.98 has been assigned to the methylene protons of 4-acetoxymethylpyridine. A signal at 6 = 2.62 is a carbon-13 satellite of the methyl protons of acetic anhydride. (b) Sweep from low to high field was started at 6 = 1.35; R = 170.2 Hz/set, T, = 2.048 set, P = 1024, F= 300 Hz, W= 0.9 sec. The methyl triplet is at 6 = 1.14,1.06, and 0.98, and an inverted signal at 6 = 0.05 is due to methane. FIG.
i Abbreviations used: R, sweep rate; T,, sampling time; P, number of sampling points; F, cutoff frequency of the low-pass filter; W, time constant for the exponential window exp(-f/W).
CORRELATION
SPECTROSCOPY
AND
CIDNP
329
anhydride. In Fig. la the low-field region was covered by a linear sweep starting at S = 2.10 from high to low field. Data acquisition was first made at 70 set after insertion of the sample into the NMR probe preheated at 98°C followed by four data acquisitions at intervals of 20 sec. Figure lb shows the correlation NMR spectra of the high-field region recorded in a similar way using a linear sweep from low to high field, starting at 6 = 1.35. It is confirmed that the spectral pattern shows no dependence upon the flip angle. It should be noted that the spectra including the methyl and methylene proton signals of 4-ethylpyridine in a nonequilibrium state obtained using a flip angle of approximately 90” are quite similar to those observed in the CW mode (7). This result shows that in the case of weakly coupled spin systems in nonequilibrium states the observed intensities are E
FIG. 2. Pulse FT NMR spectra of a reaction mixture of 4-picoline-N-oxide and acetic anhydride. The free induction decay was acquired at 70 set after a sample was inserted into the NMR probe preheated at 98°C. T,=4.096 set, P= 8192, F= 1250 Hz, W= 2.0 sec. Flip angles employed were 17” (A), 35” (B), 59” (C), 69” (D), and 90” (E). Large peaks between the methyl and methylene signals have not been reproduced in the figure. A signal on the left-hand side of the methylene quartet is a carbon-13 satellite of the methyl protons of acetic anhydride.
not affected by a large flip angle, if each of the spin-coupled groups is scanned separately. A somewhat similar situation in the pulse FT method, where only a selected number of nuclei in systems with heteronuclear spin-spin couplings is excited by the radiofrequency pulse, has been discussed by Ernst and co-workers (9). The above result should be an obvious advantage of the correlation method over the pulse FT method, which reproduces the correct intensity pattern only when a small flip angle is employed at a sacrifice in sensitivity. Figure 2 shows pulse FT NMR spectra of the methyl and methylene protons obtained by employing several different flip angles. When large flip angles are used, the multiplet efict disappears as predicted (1,2). For convenience in the following discussion, some symbols will be introduced. Of the methyl and methylene proton signals, the one which is swept initially will be denoted by 12
330
YOJI ARATA
I, and that covered subsequently by S. The times at which the I and S groups are scanned through will be designated as t(I) and t(S), respectively. Figure 3 shows the correlation NMR spectra of 4-ethylpyridine in a nonequilibrium state where the methyl and methylene groups are swept through altogether from low to high field, using a flip angle of approximately 90”. In contrast to the results given in Fig. 1, the intensity pattern of the methyl and methylene protons is quite different at
3
2
IO
FIG. 3. Correlation NMR spectra of the methyl and methylene protons of 4-ethylpyridine produced in a reaction mixture of 4-picoline-N-oxide and acetic anhydride. The whole range was scanned from low to high field, using a flip angle of approximately 90”. The transients were acquired at 70 set (A), 90 set (B), 110 set (C), and 130 set (D) after a sample was inserted into the NMR probe preheated at 98°C. R = 170.2 Hz/set, T, = 2.048 set, P = 1024, F= 300 Hz, W= 0.9 sec.
a different stage of the reaction; the shape of the methylene signals remains unchanged, whereas the methyl signal gives a different intensity pattern at a different stage of the reaction. The same is true for the methylene protons when the sweep is from high to low field, using again a flip angle of approximately 90”. Figure 4 gives an example of the spectrum where the methylene signal has completely lost the multiplet effect. With a smaller flip angle, a spectrum with anomalous intensities is observed at a later stage of the reaction. When the flip angle is further reduced, an anomaly in the intensity pattern such as that mentioned here becomes unobservable. When the I and S spins are scanned through altogether, nonequilibrium spin polarization at t(S) is presumably a sum of two contributions; one is from a group of
CORRELATION
SPECTROSCOPY AND CIDNP
331
spins perturbed by a sweep through the I resonances, and the other is due to 4-ethylpyridine freshly produced between t(1) and t(S). The second contribution should be dominant when the reaction takes place rapidly, inducing between t(I) and t(S) a sufficient amount of spin polarization which dominates spin polarization at t(S), and/or the relaxation time of the ethyl protons is much shorter than r(S) - t(I). In this case, the S signals would actually come from 4-ethylpyridine whose polarization has never been perturbed by a sweep through the I group. In view of the results given in Fig. 1, it should be natural that the same intensity pattern as
FIG. 4. Correlation NMR spectrum of the methyl and methylene protons of 4-ethylpyridine produced in a reaction mixture of 4-picoline-N-oxide and acetic anhydride. Large peaks between the methyl and methylene signals have not been reproduced. A flip angle of approximately 90” was used to sweep the whole range of the spectrum from high to low field. The reaction temperature is slightly lower than in the experiments described in Fig. 3. The transient was recorded at 70 set after insertion of a sample into the NMR probe. A signal on the left-hand side of the methylene quartet is a carbon-l 3 satellite of the methyl protons of acetic anhydride. R = 205.5 Hz/set, T, = 2.048 set, P = 1024, F= 300 Hz, W= 0.9 sec. in a slow passage low power experiment is obtained in this case, even if a large flip angle is used. Another typical pattern of the intensity is the one in which the I group retains the multiplet effect, whereas the S group loses it completely. This type of spectrum is observed either at a late stage of the reaction, as in Fig. 3D, or from the first beginning when the reaction takes place at a lower temperature (Fig. 4). The disappearance of the multiplet effect for the S group suggests that the first contribution discussed above is dominant; at t(S), magnetization of the ethyl protons would still be under a strong influence of a sweep through the I spins with a large flip angle, and when the relaxation time of the ethyl protons is much longer than t(S) - t(I), the multiplet effect should completely be lost in the S group. The different intensity pattern of the S group observed at different stages of the reaction (Fig. 3) may be interpreted qualitatively in terms of different contributions from the two types of nonequilibrium spin states mentioned above.
observed
332
YOJI ARATA
It should be pointed out that the total intensities of the I and S signals observed when both of these groups are scanned through altogether are determined by spin polarization at t(1) and t(S), and therefore are not proportional to the number of the protons in the I and S groups. CONCLUSION
It has been shown that in a weakly coupled spin system in a nonequilibrium state, the same relative intensities as those obtained in a slow-passage low-power experiment can be obtained on the correlation mode using a large flip angle by simply sweeping each of the spin-coupled groups separately. Another advantage inherent in the correlation method is, of course, a large dynamic range, which can be gained by simply not covering unwanted large signals. However, it should be emphasized that the present method is in a strict sense applicable only to a weakly coupled spin system; in the case of strongly coupled spin systems it would be difficult to limit excitation to a specified group of nuclei, resulting inevitably in an anomalous intensity pattern when large flip angles are employed (6). When a whole range of weakly coupled groups of spins is covered sequentially by one sweep, the effect of perturbing the first group in general affects the relative intensities of subsequently covered groups. The observed results should be dependent on several factors such as the rate of production of the species under investigation, flip angle, sweep rate, relaxation time. In the present paper, the effect of relaxation has not been taken into account explicitly. A slight change with time of the intensity pattern observed in Fig. 1 may at least in part be interpreted in terms of the effect of relaxation. Since in correlation spectroscopy, groups of spins with different chemical shifts are excited sequentially, a quantitative analysis of observed intensities would in principle give additional information about transient phenomena such as CIDNP. These points will be discussed in detail in a future publication. ACKNOWLEDGMENTS The author expresseshis deep gratitude to Professor S. Fujiwara for his continuing interest and encouragement. The author wishes to thank Dr. Ferretti, whose comments have been very helpful. Thanks are due to H. Ozawa for his assistance, and to M. Kunugi for taking pulse FT NMR spectra. The author is also indebted to Professor H. Iwamura for his helpful discussion about CIDNP. REFERENCES 1. S. SCHAUBLIN, A. HQHENER, AND R. R. ERNST, J. Magn. Resonance 13, 196 (1974). 2. R. R. ERNST, W. P. AUE, E. BARTHOLDE, A. H~HENER, AND S. SCH~UBLIN, Pure Appl. Chem. 37, 47 (1974). 3. J. DADOK AND R. F. SPRECHER, J. Magn. Resonance 13, 243 (1974). 4. R. K. GUPTA, J. A. FERRETTI, AND E. D. BECKER, J. Magn. Resonance 13, 275 (1974). 5. Y. ARATA AND H. OZAWA, Chem. Lett. (Tokyo) 1974,1257; H. OZAWA AND Y. ARATA, Chem. Lett. (Tokyo) 1975, 239; Y. ARATA AND H. OZAWA, J. Magn. Resonance 21, 67 (1976). 6. J. A. FERRETTI, private communication. 7. H. IWAMURA, M. IWAMURA, T. NISHIDA, AND S. SATO, J. Amer. Chem. Sot. 92,7474 (1970). 8. H. C. TORREY, Phys. Rev. 76, 1059 (1949).
9. Ref. (I), Section IIb.