5 Part (i) HIGH RESOLUTION NUCLEAR MAGNETIC RESONANCE SPECTROSCOPY By J. Feeney ( Varian Research Laboratory, Klausstrasse 43, Ziirich 8, Switzerland) NUCLEAR magnetic resonance spectroscopy continues to have widespread use in analytical and structural investigations. Further support for such routine n.m.r. applications has been provided recently by extensive compila- tions of data by Brugell and Bovey.2 In this review no attempt will be made to cover the many excellent examples of the use of 1i.ni.r. for molecular structural determinations. Likewise, little consideration will be given to the well established procedures for conformational analysis and kinetic studies of inter- and intra-molecular processes. However, special attention will be paid to any advances which have been made in instrumental procedures and in theoretical aspects of the technique. It is worth noting that the develop- ment of the theory of chemical shifts and spin-spin coupling constants is not being retarded by lack of excellent experimental data of which there is an abundance available.Experimental Techniques.-The most exciting recent advance has been the use of pulse techniques to remove dipolar and quadrupolar broadening in solid samples to an extent where chemical shifts and scalar couplings can be ~bserved.~ -7 Normally to observe high resolution n.m.r. spectra it is necessary to examine samples as gases or liquids, or in solution, to provide the molecules with sufficient tumbling and rotational motion to average out the dipolar interactions. Waugh and his co-workers have shown that by pulsing repeti- tively with brief intense pulse trains, phase shifted by k90" and/or & 180", one can observe effects similar but not equivalent to the motional narrowing observed in liquids.A typical 8-pulse cycle used is (T, Px; 27, P-x; 2, Py; 22, P-y; 2, P-x; 22, Px; 2, P-y; 27, Py). The effect of a fast 90" train is analogous to that of molecular motion, and the transient n.m.r. signal observed is the Fourier transform of the high resolution spectrum revealed by the r.f. analogue of motional narrowing. When the system is subjected to a repetitive cycle of pulses, the spin operators 1 W. Briigel, 'NMR Spectra and Chemical Structure,' Academic Press, New York, 1967. F. A. Bovey, 'NMR Data Tables for Organic Compounds,' Wiley, New York, 1967.P. Mansfield and D. Ware, Phys. Letters, 1968, 27A, 159. 3 J. S. Waugh, L. M. Huber, and U . Haeberlen, Phys. Rev. Letters, 1968,20, 180. 5 J. S. Waugh, L. M. Huber, and E. D. Ostroff, Phys. Letters, 1968,26A, 211. 6 D. Gill, Phys. Letters, 1968, 26A, 544. J. S . Waugh, C . H. Wang, L. M. Huber, and R. L. Vold, J . Clieni. Phys., 1968, 48, 662. 6364 J . Feerzey (rather than the lattice operator in motional averaging) bccome time depen- dent. A series of 90" r.f. pulses of different phase has been applied to a sinall single crystal of calcium fluoride in the presence of some liquid trifluoro- methylbenzene, and the phase detected Bloch decay has been observed. The signal from the solid gave a fast decay followed by the much longer decay of the liquid signal. A compound modulation was obtained because of the difference in chemical shift between the two samples (the dipolar interactions in the solid having been removed).The Fourier transform of the compound modulation gives the slow passage spectrum with all dipolar interactions removed and all chemical shifts reduced by a factor of 4 3 . Linewidths of less than 200 Hz have been observed for solid calcium fluoride by this method and no doubt these linewidths will be further improved. This method promises to be much more convenient than the alternative method of removing dipolar broadening in solids by rapidly rotating the samples around an axis inclined at 54" 44' to the direction of the external magnetic field. It has been shown recently that when a molecule has a spin coupling tensor which has an antisymmetric component this component is not usually removed by the rotation.8 The quest for sensitivity enhancement in n.ni.r. has continued, stimulated largely by the increasing interest in nuclei other than protons where the natural abundance and sensitivity to n.m.r.detection often pose serious sensitivity problems. For example for 13C nuclei the natural abundance is only 1.1 % and the natural sensitivity to detection at constant field strength is only 1.6% that of protons. All known methods of sensitivity enhancement (for example use of large sample tubes, high fields, spectrum accumulation, Fourier transform) benefit by having a high sensitivity in the basic instru- ment: over the last seven years most commercial n.m.r.spectrometers have been able to improve their basic sensitivity specifications by a factor of five as a result of instrumental improvements. During the last year, heteronuclear noise decoupling has become an added powerful means of improving sen- sitivity. Its application to 13C spectroscopy is perhaps the most drama ti^.^ By irradiating at the proton frequency with a noise-modulated r.f. source it is possible to achieve a broad band decoupling effect which decouples simul- taneously all the chemically shifted protons (even those with large C-H coupling constants of 200 Hz). The signal-to-noise ratio of the decoupled 13C spectrum is improved not only by virtue of the fact that the intensity of the collapsed multiplet is now concentrated into a single absorption band but also because of a positive nuclear-nuclear Overhauser effect, which can give signal-to-noise enhancement of a factor of three or more: the two effects combine to improve the sensitivity by a factor of seven or eight, which corresponds to a considerable time saving when spectrum accumulation techniques must be used.By use of this method it is possible to detect all * E. R. Andrew and L. F. Farnell, Mol. Phys., 1968, 15, 157. 9 L. F. Johnson, 9th Experimental Conference, Pittsburgh, Pennsylvania, March 1968.Iiigh Resolution Nuclear Magnetic Resonarice Spectroscopy 65 the carbon atoms in cholesterol after one hundred scans.l*, l1 Broad-band decoupling can be achieved without use of noise modulation techniques, by use of a very high power coherent r.f.source.12 However, one can only use this approach with a time-sharing spectrometer where the measuring r.f., the irradiating r.f., and the receiver are switched on in turn for short non-over- lapping intervals at the rate of a few kHz. Hewittl3 has used a time-sharing spectrometer of this type to obtain beat-free decoupled spectra of bicyclo- butane in a nematic solvent. The need for small dedicated digital computers in nuclear magnetic reso- nance has long been realised, but it is only recently that such facilities have become available. Some obvious uses are data acquisition in digitised form, data reduction to a few significant parameters, data handling, spectrum accumulation spectrum smoothing, resolution enhancement, line-shape and spectral analysis and simulation of multispin systems.Computers will also be used increasingly to control n.m.r. spectrometer operation. An obvious example is the automatic control of magnetic field homogeneity by monitoring some resolution criterion such as peak height and automatically modifying the electric shim coil homogeneity controls to optimise the resolu- tion.14 Of the possible criteria for homogeneity, only the second moment and related criteria can allow one to construct a complete set of non-interacting shim adjustments : however, for a resonance line with sufficient natural linewidth the peak height can be related to the second moment and may also give non-interaxting control. The advent of field/frequency locked spectrometers has made practical a simple method of measuring the r.f.observing field, Hl.l5 One selects a sample with a single narrow absorption line and at a low non-saturating value of r.f. power the spectrometer is adjusted to be exactly on resonance: the r.f. power is then suddenly increased by 40 dB such that a transient wiggle signal is observed (o ca. 2 Hz). This frequency is a measure of yH1 to about 0.5% accuracy. By use of a steady-state method, an electric field effect has been observed on the 14N spectrum of highly purified nitrobenzene when a very high field (60 kv/cm.) is applied.16 As expected, a line-splitting was detected but this could not be described completely by use of either the Lorentz or the Onsager local field model. By use of a low field ( 1 0 ~ ) n.m.r. spectrometer with very large samples (1-25 1.) signals from whole living animals have been detected for the first time.17 lo J.D. Roberts, NATO NMR Summer School, Coimbra, 1968. l1 L. F. Johnson, personal communication. l2 E. Lippmaa and T. Pehk, Kemian Teollisuus, 1967, 24, 1001. l 3 R. C. Hewitt, Rev. Sci. Instr., 1968, 39, 1066. l4 R. R. Ernst, Rev. Sci. Instr., 1968, 39, 998. l5 J. S. Leigh jun., Rev. Sci. Instr., 1968, 39, 1594. C. W. Hilbers and C. MacLean, Chem. Phys. Letters, 1968, 2 , 445. J. A. Jackson and W. H. Langham, Rev. Sci. Instr., 1968, 39, 510.66 J. Feeney Novel Techniques.-Several workers18-22 have shown that it is possible to observe n.m.r. emission and enhanced absorption during rapid free-radical reactions. This is referred to as chemically induced dynamic nuclear polarisa- tion (CIDNP).When free radicals are generated by the breaking of chemical bonds, the unpaired electron states have equal populations initially and tend to reach a Boltzman distribution by spin-lattice relaxation. If fluctuating electron spin-nuclear spin interactions are present, cross relaxation (combined electron spin/nuclear spin) transitions may be induced which disturb the nuclear Boltzman equilibrium. This can result in enhanced nuclear polarisa- tion. When radicals react further the polarisation is transferred to the dia- magnetic reaction products, which give rise to increased n.m.r. signals. Whether or not one obtains emission or enhanced absorption depends on whether there is a dipolar or a scalar coupling involved. CIDNP proton spectra have been reported recently for methyl, ethyl, propyl, isopropyl and t-butyl radicals by Kaptein.22 This author also clarified the conditions con- cerning relaxation times and radical life-times necessary for successful obser- vation of this effect.Cocivera23 has described a method for studying optically excited molecules by n.m.r. Using a 3500 w mercury lamp operating over 3000-4000 8, he irradiated 0.005 M anthraquinone in benzene. Under these conditions one obtains a steady-state concentration of anthraquinone in the triplet state. Because of the low concentration of the triplet state molecules, the n.m.r. spectrum of only the ground state molecules is observed, but it is modified by the presence of the triplet states. Thus we see Overhauser effects which change the spectrum of the ground state dramatically, leading to inversion of most of the lines in the ground state spectrum by a process of stimulated emission induced by the optical excitation of the molecule to its lowest lying triplet state.Multiple-resonance experiments form the basis for several novel applica- tions. Sinevee and Sa1m-1~~ have developed a theory to predict the effects of triple homonuclear resonance by use of weak r.f. fields: their theory predicts not only the line shapes but a new phenomenon, namely, the appearance under certain conditions of a signal with a frequency which is a combination of the basic frequencies. This has been experimentally substantiated by By~trov,~~ who has examined the triple resonance spectra of AMX systems.Freeman and Gestblom26 have described a method of measuring frequency l8 J. Bargon, H. Fischer, and U. Johnsen, Z . Nnrurforsch., 1967, 22a, 1551. l9 J. Bargon and H. Fischer, 2. Nuturforsch., 1967, 22a, 1556. 20 H. R. Ward and R. G. Lawler, J . Amer. Chem. Soc., 1967, 89, 5518. 21 R. G. Lawler, J . Amer. Chem. SOC., 1967, 89, 5519. 22 R. Kaptein, Chem. Phys. Letters, 1968, 2, 261. 33 M. Cocivera, J . Arner. Chem. Soc., 1968, 90, 3261. 24 V. Sinevee and V. Salum, BUN. Acnd. Sci. Estonian S.S.R., Phys. Marh. Sci. Ser., 25 V. F. Bystrov, J . MoI. Spectroscopy, 1968, 28, 81. 26 R. Freeman and B. Gestblom, J . Chem. Phys., 1968, 48, 5008. 1968, 17, N1.High Resolution Nuclear Magnetic Resonance Spectroscopy 67 differences to & 1 mHz. In high resolution n.m.r.the dominant line-broaden- ing mechanism is very often the spatial inhomogeneity of the magnetic field; in such cases it is possible to carry out a double resonance experiment to impose a very weak r.f. field, H z , on a given line, Wa, in the spectrum, such that the saturation is localised to a restricted region of the sample volume (referred to as ‘burning a hole’ in the line). If one then examines a line, Wb, regressively connected to Wa, one observes a similar hole in the exactly cor- responding position on the line profile resulting from changes in the popula- tion of the common energy level. This correlation arises from the intra- molecular nature of the coupling and it forms the basis of a technique for measuring frequency separations with a precision not determined by in- homogeneous line widths.Irradiation with a weak perturbing field in a spin-tickling experiment can be used to locate precisely hidden 13C.CH satellite lines in a proton spectrum and hence to determine the magnitude and signs of CCH coupling constants. The technique is applicable mainly to two or three spin systems but by com- bination with spin decoupling in a triple resonance experiment it can be extended to more complex systems, such as found in ally1 bromide.27 By irradiating at the CHzBr frequency the molecule is effectively reduced to a three-spin system, which is sufficiently simple to carry out the required spin- tickling experiments on the 13C.CH satellite spectra. Govil and BernsteinZ8 have found the same relative signs for HgHF and JtHF in 1,1,2,2-tetrabromo- fluoroethane in the gauche and trans rotameric forms.This was possible because the nuclei in the slowly interconverting forms retain the same nuclear orientations with respect to each other during the exchange process: thus irradiation at part of the multiplet of one nucleus will cause transfer of saturation to only part of the multiplet of the interacting nucleus in the other form involved in the slow intramolecular exchange process. Hoffman and ForsenZ9 were the first to use double-resonance methods to study kinetic effects by observing the transfer of saturation from an irradiated nucleus to a nucleus at another nuclear site involved in exchange with the irradiated nucleus. Fung30 has extended the method to cases where the spin saturation can be propagated along an aliphatic chain to involve protons not participating in the direct exchange process but strongly coupled to the nuclei involved in the exchange.This has been illustrated in experiments with mixtures of 2-phenoxyethanol and t-butyl alcohol. Double quantum transi- tions (DQT) in four-spin systems such as (I) have been used to obtain the relative signs of the coupling constants.31 Strong DQT were observed for low 27 J. Feeney and P. J. S . Pauwels, Mol. Phys., 1968, 14, 209. 28 G. Govil and H. J. Bernstein, Mol. Phys., 1968, 14, 197. 29 R. A. Hoffmann and S . Forsen, J. Chem. Phys., 1963,39, 2892. 3O B. M. Fung, J . Amer. Chem. SOC., 1968,90,219. 31 L. Lunazzi and F. Taddei, J . Mol. Spectroscopy, 1968, 25, 113.68 J. Feeney values of Av/J: for AX3 type systems when hv/J was greater than 30 it was not possible to detect DQT.Two independent methods for determining micelle formation with greater ease have been reported. Micelle formation in solutions of sodium 4,4- dimethyl-4-silapefitane-1 -sulphonate in the presence of paramagnetic ions is accompanied by enhanced chemical shift and linewidth changes compared with experiments in the absence of paramagnetic ions.32 Haque33 has exa- mined fluorinated colloidal electrolytes, sodium perfluorocaprylate and propionate, and observed large changes in 19F chemical shifts on micelle format ion. Primary, secondary, and tertiary alcohols have been characterised by preparing the dichloroacetate esters and observing the characteristic dichloro- acetyl proton resonance.34 The meltifig point of ethane has been determined by observing the motional narrowing of its proton n.m.r.signal at the phase transition; during this investigation it was noted that n.m.r. can be used as an impurity detector by measuring the impurity premelting which takes place below the melting point of the pure substance.35 Herington and Lawren~on~~ have illustrated how to determine the purity of organic samples by n.m.r. The samples were examined from 20" below their melting point up to a temperature at which all the sample is melted. The intensity of the liquid line is measured at dif- ferent temperatures and a linear graph of temperature against reciprocal of fraction melted is obtained. From the slope of the graph the purity of the sample can be estimated, and the intercept on the temperature axis gives the melting point of the pure material.Signal integration techniques, applied to separate proton signals from water molecules within and outside the primary co-ordination sphere of Nil1 com- plexes in aqueous solution at - 30", have enabled the solvation number to be determined as six. This is the first time this technique has been used with paramagnetic ions.37, 38 32 B. R. Donaldson and J. C. P. Schwarz, J. Chem. SOC. (B), 1968, 957. 33 R. Haque, J. Phys. Chem., 1968, 72, 3056. 34 J. S. Babiec, J. R. Barrante, and G. D. Vickers, Analyt. Chem., 1968, 40, 610. 35 L. J. Burnett and B. H. Muller, Nature, 1968, 219, 59. 36 E. F. G. Herington and I. J. Lawrenson, Nature, 1968, 219, 928. 37 T.J. Swift and G. P. Weinberger, J. Amer. Chern. Soc., 1968, 90, 2023. 38 N. A. Matwiyoff and P. E. Darley, J. Phys. Cheni., 1968,72, 2659.High Resolution Nuclear Magnetic Resoriance Spectroscopy 69 The tetrameric nature of alkyl-lithium compounds in solution has been confirmed by enriching methyl-lithium with 13C and observing the 7Li-13C coupling in THF solution (14.5 Hz).~* Lippmaa and his co-workers40 have measured intermolecular Overhauser effects between much larger molecules than had previously been considered possible. Coupling Constants.-The Pople and Santry4I CNDO-2SCF (complete neglect of direct overlap) approximation for calculating coupling constants has been extended by Ditchfield and M~rre11,~~ who have allowed for con- figuration interaction between calculated excited states.The Pople and Santry method is a perturbational method for calculating coupling constants based on a MO approach but which avoids using the mean energy of excited states approximation. They considered only excited states arising from single excitations between molecular orbitals obtained from a minimum basis of valance-shell atomic orbitals. The theory is formulated on the basis of a one- electron Hamiltonian such that the excitation energies are taken as the difference between orbital energies and on the assumption of no configura- tional interaction between the excited states. This approach gave satisfactory results for hydrocarbons and simple heteroatomic molecules. By introducing configurational interaction between calculated excited states, Ditchfield and Murrell have obtained good agreement between observed and calculated directly bonded C-H and C-C coupling constants in hydrocarbons. The calculated non-bonded C-H coupling constants are also in better agreement than those obtained previously.Gil and Teixeira-Dias43 have considered the effects of substitution on directly bonded C-H coupling constants within the framework of the Pople and Santry MO theory. H i r ~ i k e ~ ~ has also considered this problem. P-P coupling constants have been explained45 satisfactorily by use of an extension of the Pople and Santry theory; JPP values can be either positive or negative, as was found with JCP values.46 A linear relationship between directly bonded C-H coupling constants and calculated (SCF method) bond orders has been found for a series of hydrocarbons.47 R o s ~ i ~ ~ has proposed a theoretical explanation of the effects of deuterium substitution of a vicinal proton on directly bonded JCH values on a tetrahedrally hybridised carbon atom.The change in the time- L. D. McKeever, R. Waack, M. A. Doran, and E. B. Baker, J . Amer. Chem. SOC., 40 E. Lippmaa, M. Appa, and A. Sugis, Eesti N.S. V . Teaduste Akudeemia, Toimetised, 41 J. A. Pople and D. P. Santry, Mol. Phys., 1965, 9, 31 I . 42 R. Ditchfield and J. N. Murrell, Mol. Phys., 1968, 14, 481. 4D V. M. S. Gil and J. T. C. Teixeira-Dias, Mol. Phys., 1968, 15,47. 44 E. Iiiroike, J. Phys. SOC. Jupan, 1968,24, 1348. 45 E. G. Finer and R. K. Harris, Chem. Comm., 1968, 1 1 0 . 46 W. McFarlane, Chem. Comm., 1967, 58.47 G . Berthier, H. Faucher, and D. Gagnaire, BUN. Sue. chim. Frnrrce, 1968, 1872. 48 M. Rossi, Chem. Phys. Letters, 1968, 2, 353. 1968,90, 3244. 1967, 3, 385.70 J . Feeney averaged 2s character of the hybrid carbon orbital due to coupling of the bending vibrations is calculated and good values of C-H coupling constants are predicted. has developed a VB description of the contact nuclear spin-spin contribution which does not invoke the mean excitation energy approxima- tion. By use of this approach he has calculated values for long range H-H coupling i-c-electrons which are in reasonable agreement with the experi- mental values. Some of the elegant measurements of coupling constants of fundamental importance are worth mentioning. By examining the AA'BB' lH spectrum from [l ,1,2,2,3,3,4,4-2H8]cyclohexane (II) under conditions of deuterium decoupling at -103" it was possible to extract the following coupling con- Another method of obtaining coupling constants between nuclei in fixed relative positions to each other is to examine rigid bicyclic compounds.In this way the angular dependence of vicinal lH-14N coupling could be determi~~ed,~l by examining molecules such as dibenzobicyclo [2,2,2]octa-2,5- dien-7-yltrimethylammonium bromide (111). The following results were obtained: dihedral angle 0", J N H ~ 2.7 (probably positive); 60°, J N H ~ < 0.3 Hz; 120", JNHd 0.8 Hz. For compounds with dihedral angles of 180" the J N H " ~ ~ stants :50 Jas,astrans 13.12, Je q,e qtrans 2.96, Jaz,e qcis 3-65, and Jgem - 13.05 Hz.value is predicted to be 5-6 Hz. Some coupling constant sign determinations would be useful on this system. JH-S-C-H and JH-S-C-CH have been shown52 to be both positive in sign by double-resonance experiments on the bicyclo-compound 1,2,3,4- tetr achloro-5-exo-mercapt o bicyclo [2,2,1] hept-2-ene. Many other relative sign determinations for homo- and hetero-nuclear coupling constants have 49 M. Barfield, J. Chern. Phys., 1968, 48, 4458, 4463. 51 Y. Terui, K. Aono, and K. Tori, J . Amer. Chem. SOC., 1968,90,1069. b2 V. F. Bystrov and 0. P. Yablonsky, J . Mol. Spectroscopy, 1968,26, 213. E. W. Garbisch and M. G. Griffith, J. Amer. Cliem. Soc., 1968,90, 6543.High Resolution Nidear Magnetic Resonance Spectroscopy 71 been reported, as typified by references 53 to 57.Several empirical correla- tions involving coupling constants have been pointed out. Values of JHC.OH in primary alcohols follow a Karplus relationship similar to that relating JHC. CH vicinal coupling with dihedral angle.5* Long range protoii-methyl coupling constants in propene, mesitylene, and other compounds are linearly related to the square of the mobile bond 0rder.5~ A linear correlation has been found between JCF in trifluoromethyl derivatives with carbon-substituent bond length JCF = -(106rcx + 115) where YCX is bond length in A.6o JCH in a series of methyl derivatives corre- lates linearly with the :product of the electronegativity EX and the bond distance rcx.61 The coupling constants in N-substituted pyridines show a similar dependence on the N-substituent as is found in monosubstituted benezenes.62 An additivity scheme for ortho-, meta-, andpara-proton coupling constants based on changes induced by a single substituent in model com- pounds has been applied to trisubstituted benzenes: only the ortho-coupling constants could be reproduced by this approach.63 Solvent effects on coupling constants, although small, can have important implications.Fineg01d~~ has examined the solvent dependence of vicinal coupling constants in systems which are rotationally invariant in cyclic rings; a gradation of gauche vicitzal coupling constants over a 14 % range with change of solvent points to the errors involved in using medium effects to solve rotational isomerism problems. A large solvent dependence of JPF in phosphorus trifluoride [JPF (gas) 1404 Hz; JPF (Cch solution) 1423 Hz] has been explained qualitatively in terms of electric field effects.65 JBF in silver tetrafluoroborate changes sign on changing solvent from water to aceto- This observation, together with the small magnitude OfJBF (ca.1 Hz), indicates that we have cancellation of large opposing coupling contributions. The reduced X-F coupling constants in the isoelectronic series BeF42-, BF4-, CF4, and NF4+ correlate with atomic number in the same manner as found in the series TeFs, SbFs-, and S ~ F G ~ - . ~ ~ From the analysis of the proton spectrum of formaldehyde [15N]oxime68 (IV) in methyl cyanide solu- 53 R. K. Harris and C. M. Woodman, J . Mol. Spectroscopy, 1968, 26, 432. s4 R. B. Johannesen, F.E. Brinckman, and T. D. Coyle, J . Phys. Chem., 1968,72,660. 55 T. C. Farrar, R. B. Johannesen, and T. D. Coyle, J . Chern. Phys., 1968,49,281. H. Dreeskamp and G. Pfisterer, Mol. Phys., 1968, 14, 295. 57 R . B. Johannesen, J . Chem. Phys., 1968, 48, 1414, 58 E. F. Kiefer, W. Gericke, and S. T. Amimoto, J . Amer. Chem. SOC., 1968,90, 6246. 59 D. J. Blears, S. S. Danyluk, and T. Schaefer, Canad. J . Chem., 1968,46,654. 60 I. Love, Mol. Phys., 1968, 15, 93. G2 S. Castellano and R. Kostelnik, J. Amer. Chem. Suc., 1968,90, 141. 63 T. Schaefer, G. Kotowycz, H. M. Hutton, and J. W. S. Lee, Cunad. J. Chem., 1968, C . P. Yue, Cunad. J . Chem., 1968,46, 2675. 46, 2530. H. Finegold, J. Phys. Chem., 1968, 72, 3244. G5 W. T. Raynes, T. A. Sutherley, H. J. Buttery, and C.M. Fenton, Mol. Phys., 1968, 66 R. J. Gillespie, J. S. Hartman, and M. Parekh, Canad. J . Chem., 1968, 46, 1601. 67 J. Feeney, R. Haque, L. W. Reeves, and C. P. Yue, Cunad. J . Chem., 1968,46,1389. 68 D. Crkpaux and J. M. Lehn, Mol. Phys., 1968,14, 547. 14, 599.72 J. Feeney tion values of JNHA (& 14.2) and JNHB (Zt2.3 Hz) have been obtained. Proto- nated aldehyde [15N]oximes also show similar but smaller effects.68 These results can be explained in terms of the orientation of the lone pair on the nitrogen in the formaldehyde oxime which appears to greatly increase JNHA without appreciably affecting JNHR. In the protonated species where the lone pair is replaced by an N-H bond, JNHA is much smaller than in formalde- hyde oxime. The stereospecific nature of long range H-H coupling is illustrated by the observation that protons on both sp2- and sp3-hybridised benzylic carbon atoms have maximum coupling with the ortho-protons when the benzylic proton is out of plane, while the reverse is true for the coupling with the rneta-protons.69 A large number of 2 J ~ . ~ ~ values have been measured in acyclic organo- phosphorus compounds and values between 0 and 22 Hz obtained. They could not be rationalised theoretically and even their empirical use for struc- ture determination requires caution if very close analogies cannot be found. 70 Chemical Shifts.-Relatively few papers dealing with proton chemical shift calculations have appeared in the past year. It is still not possible to carry out the necessary detailed quantum mechanical calculations to predict chemical shifts and one must resort to specific physical models (such as those based on anisotropic or electric field effects) to explain differences in chemical shifts, Fraenkel and his co-~orkers~~, 72 have postulated that there are significant paramagnetic contributions to the shielding of ortho-protons in aromatic organometallic compounds of lithium, magnesium, and calcium.Magnetic mixing of the ground and low-lying excited states gives paramagnetic con- tributions to the shielding, causing low-field shifts of the ortho-protons. Homer and Callagha~~~~ have re-examined carefully the approach of using neighbour anisotropic effects to predict chemical shifts. The anisotropy of the magnetic susceptibilities of C-C and C-H bonds was investigated and the values of Axcc and AxCH obtained from the n.m.r.measurements were shown to be incompatible with those found from Cotton-Mouton constants. A similar was undertaken for C-F and C-Cl bonds, and doubt has been cast on whether the n.m.r. estimates of AxCF and AxCC1 are meaning- ful. 69 G. P. Newsoroff and S. Sternhell, Austral. J . Chem., 1968, 21, 747. 7O M. J. Gallagher, Austral. J . Chem., 1968,21, 1197. 71 G. Fraenkel, D. G. Adams, and R. R. Dean, J . Phys. Chem., 1968,72, 914. 72 G. Fraenkel, S. Dayagi, and S. Kobayashi, J , Phys. Chem., 1968, 72, 953. 73 J. Homer and D. Callaghan, J . Chem. Soc. (A), 1968, 439. 74 J. Homer and D. Callaghan, J . Chem. SOC. (A), 1968, 518.High Resoliltion Nuclear Magnetic Resoriance Spectroscopy 73 The contribution to proton shielding from steric compression is known to result in a deshielding effect.Cheney75 has correlated this with the com- ponent of the non-bonded H-H repulsive force along the C-H bond axis. The model used postulates that there is induced electronic charge polarisation in the C-H bond as a result of the H-H interaction. An empirical linear relationship was found : = - 105 Z cos 8i exp (-2.671r;) p.p.m. i where 8i is the angle between C-H and the H . . . H internuclear line,and ri is the separation between the sterically interacting protons. The summation is taken over all hydrogen atoms which interact significantly with the hydrogen under study. A similar relationship has been used to predict contributions to 13C chemical shifts resulting from steric crowding. These effects are to be dis- tinguished from electric-field effects resulting from dispersion forces which show no angular dependence.Using the SCF-LCAO-MO method of Pople, E m ~ l e y ~ ~ has computed the 'H, 13C, and 14N chemical shifts in pyridine and pyridinium ion from the charge densities: he found that one cannot neglect c polarisation in such calculations. Tokuhiro and his co-workers calculated the diamagnetic and paramagnetic contributions to lH and 13C shielding in pyridine and found values in good agreement with experiment. Wu78 has used the P ~ p l e ~ ~ theory to calculate 14N chemical shifts. In the absence of low-lying transitions (n --f x*) which give rise to paramagnetic shifts, the 14N shifts were found to depend primarily on (i) the electron density on the nitrogen, (ii) the mean excitation energy, and (iii) the effect of multiple bonds on the nitrogen.Emsley80 has reviewed the relationship between charge densities and 19F chemical shifts in aromatic compounds. For fluorine nuclei para to the substituent a semi-empirical SCF-LCAO-MO method was used to verify the existence of the linear correlation between 19F chemical shifts and x- electron density on the fluorine or the attached carbon atom. 19F Shielding constants calculated by the method of Karplus and Pople, by use of an aver- age energy approach (AE = 8.15 ev) are of the correct order of magnitude, but the calculated changes with substitution are too small. This cannot be explained simply by using different AE values for different substituents and it was concluded that this approach cannot reliably predict lgF chemical shifts.By studying molecules such as (V) and (VI), where the substitution is remote from the fluorine nuclei,s1, 82 the only effects on the fluorine shielding 75 B. V. Cheney, J . Amer. Chem. SOC., 1968,90, 5386. 76 J. W. Emsley, J. Chem. SOC. (A), 1958, 1387. 77 T. Tokuhiro, N. K. Wilson, and G. Fraenkel, J . Amer. Chem. SOC., 1968,90, 3622. 78 T. K. Wu, J . Chem. Phys., 1968,49, 1139. 79 J. A. Pople, J . Chem. Phys., 1962, 37, 53, 60. 8o J. W. Emsley, J. Chem. SOC. (A), 1968, 2018. 8l E. W. Della, Chem. Comm., 1968, 1558. A2 M. J. S. Dewar and T. G. Squires, J . Amer. Chem. SOC., 1968,90,210.74 J, Feeney are those observed when the substitution causes conformational distortion, which is possible in some of the decalin derivatives.82 Molecular distortions in some bicyclic fluorides83 on being substituted have also been cited as the origin of I9F chemical shift changes accompanying substitution.Several empirical correlations between chemical shifts and other molecular parameters have been pointed out. Chemical shifts of benzylic protons in para-substituted toluenes give a linear correlation with Hammett c values of para-substituents.84 Likewise the arnine proton chemical shifts in a large series of substitute-d anilines show a linear correlation with appropriate Hammett cs c0nstants.~5 Proton chemical shields in polyenylic ions (VII) have been correlated with calculated carbon chargs densities.86 There has been an increased interest in the study of isotopic effects on chemical shifts.Isotopic chemical shifts observed on replacing protons by deuterium atoms in the isoelectronic series BH4-, CH4, NH4+ have been measured.87 Although deuteriation results in higher shifts for llB and I3C, lower shifts are observed for 14N, a fact which has been explained in terms of electric-field considerations involving the nitrogen atom, which is more electronegative than carbon or boron. F The calculated shift for 19F shielding between 2HF and IHF has been shown to be in good agreement with experiment.88 83 G. L. Anderson and L. M. Stock, J. Amer. Chem. Soc., 1968,90,212. 84 R. R. Fraser, Gurudata, C. Reyes-Zamora, and R. B. Swingle, Cunud. J. Cheni., 85 B. M. Lynch, B. C. MacDonald, and J.G. K. Webb, Tetrahedron, 1968,24, 3595. 88 D . Oseen, R. B. Flewwelling, and W. G. Laidlaw, J. Amer. Chem. SOC., 1968, 90, 4209. 8? M. Shporer and A. Loewenstein, Mol. Phys., 1968, 15, 9. 88 D. K. Hindermann and C. D. Cornwell, J. Chem. Phys., 1968,48,4148. 1968,46, 1595.High Resolution Nuclear Magnetic Resonarice Spectroscopy 75 When 19F chemical shifts of alkali fluorides are measured in light and heavy water solutions there is a linear variation in shift with isotopic composition of the solvent (total shift of 3 p.p.m. observed):89 increasing the heavy water content increases the 19F shielding. This can be interpreted in terms of changes in the vibrational energy of the fluoride ion F- (H20)z; resulting from isotopic substitution which changes the average excitation energy AE in the paramagnetic shielding term.By examining the IgF shielding of HF in the presence of varying amounts of potassium fluoride and using both light and heavy water as solvents, high field shifts have been found to accompany the substitution of H by D for both the F- (2.96 p.p.m.) and HF (4-7 p.p.m.) resonance frequencies.90 19F isotopic shifts resulting from substitution of 32S by 34s (natural abundance spectra) render fluorine nuclei non-equivalent in the normally symmetrical species S205F2, S206F2, and S308F2; one can thus observe the F-F coupling constants in such systems.g1 Empirical methods of calculating proton shifts in heteroaromatic com- pounds92 and polyhalogenoben~enes~3 have been proposed. Useful charts of proton shifts in oxygenated unsaturated aliphatic compoundsg4 and for acyclic methineg5 protons have been published.Carbon-13 Resonance.-Because of the vast potential of 13C studies it is not surprising that a great deal of effort is now being expended in this area. Improvements in basic instrumental sensitivity and the use of sensitivity enhancement techniques have reached the stage where they have largely overcome the enormous inherent sensitivity problem in the study of carbon-13 nuclei. It is anticipated that there will be an explosive increase in carbon-13 studies when the organic chemist solving molecular structural problems acquires access to suitable facilities for such investigations. The desirability of total heteronuclear decoupling as a method of simplifying and increasing the sensitivity of 13C spectra has already been in~licated.~, 9 6 9 97 This is nor- mally achieved by noise decoupling at the proton frequency but Lippmaa has achieved similar results by using a time-sharing spectrometer where he can use very high powered coherent r.f.decoupling fields to irradiate all protons in a molecule simultaneously.97 Lippmaa and his co-workers have published several papers illustrating the elegant use of 13C rssonance in the study of 89 C. Deverell, K. Schaumburg, and H. J. Bernstein, J . Chem. Phys., 1968, 49, 1276. K. Schaumburg and C. Deverell, J . Amer. Chem. SOC., 1968, 90, 2495. 91 R . A. Stewart, S. Fujiwara, and F. Aubke, J . Chem. Phys., 1968, 49, 965. 92 I. Nicholson, Chem. Comm. 1968, 1028. 93 B. Richardson and T. Schaefer, Canad. J .Chem., 1968,46,2195. 94 N. F. Chamberlain, Analyt. Chem., 1968,40, 1317. 95 0. Yamamoto, T. Suzuki, M. Yanagisawa, and K. Hayamizu, Analyt. Chem., 96 A. Sugis and E. Lippmaa, Eesti N.S. V . Teaduste Akadeernia Toimetised, 1967, 1, 81. 97 E. Lippmaa, T. Pehk, and J. Past, Eesti N.S.V. Teaduste Akadeemia Toimetised, 1968,40, 568. 1967, 3, 345.76 J. Feeney alkanes,98, 99 alkenes,loO alkynes,lol methoxybenzenes,lo2 and strained molecules such as adamantane.97 Other workers have chosen to use the in- direct method of measuring 13C frequencies by using a heteronuclear double- resonance technique, where one observes the effects on the proton spectra when the 13C absorption bands are irradiated.lo3 In this way, the 13C chemical shifts have been measured as a function of pH in sevcral amino-acids and peptides such as glycine, diglycine, triglycine, alanine, and alanylglycine.By calculating charge densities, the 13C chemical shifts of the a-carbon atoms could be predicted. It was also possible to rationalise the 13C shifts changes accompanying protonation (NH2 -+ NH3+) and ionisation (C02H -+ COZ-).~O~ Litchman and Grantlo4 have found that although the 13C chemical shifts in halogen-substituted methanes cannot be explained in terms of a simple additive substituent relationship, when one incorporates a pair-interaction term into a linear expression the chemical shifts can be calculated. In an investigation of methyl derivatives of Group IV and Group IIB rnetalslO5 the M-C coupling constants have been measured and discussed in terms of the contact mechanism.Buccilo6 has pointed out an interesting empirical corre- lation of 13C chemical shifts in methyl and ethyl derivatives with the Pauling electronegativity (E) and numbers of lone pairs (M) of the substituent. 8l3CH3M 8CH313CH2M == 208 - 45E + = 234 - 55E + 13M 7M Correlations between 13C shifts and charge densities have been pointed out for diazoles,l07 triazoleslo7 and some 5-membered nitrogen heterocyclic108 com- pounds. 13C shifts have been reported for tetraiodomethane,lo9 acetyl compounds,110 the triphenylcyclopropenium cation,lll azines and their proto- nated cations,l12 pyrazole and substituted pyrazoles,l13 and 2-substituted pyridines,ll* where it was found that substituent effccts are often different from those observed in monosubstituted benzenes especially at the 2-positions.98 E. Lippmaa and T. Pehk, Eesti N.S. V . Teudusre Akadeemia Toimetised, 1968,3, 210. 99 E. Lippmaa and T. Pehk, Kemian Teollisiiius, 1967, 24, 1. loo E. Lippmaa, S. Rang, 0. Eisen, and T. Pehk, Eesti N.S.V. Tedusre Akadeemia lol S. Rang, T. Pehk, E. Lippmaa, and 0. Eisen, Eesti N . S . Y . Tecrdusre Akadreinia 102 T. Pehk and E. Lippmaa, Eesti N.S. V . Teaduste Alcadeemia Toimetised, 1968,3, 195. lo3 W. J. Horsley and H. Sternlicht, J . Amer. Chem. SOC., 1968, 90, 373s. 104 W. M. Litchman and D. M. Grant, J . Amer. Chem. SOC., 1968, 90, 1400. lo5 F. J. Weigert, M. Winokur, and J. D. Roberts, J . Amer. Chem. SOC., 1965,90, 1566. 106 P. Bucci, J . Amer. Chem. SOC., 1968, 90, 252. lo7 B. M.Lynch, Chem. Comm., 1968, 1337. 108 R. J. Pugmire and D. M. Grant, J . Amer. Chem. SOC., 1968,90,4232. 109 0. W. Howarth and R. J. Lynch, Mol. Phys., 1968, 15, 431. I1O G. A. Gray, P. D. Ellis, D. D. Traficante, and G. E. Maciel, J . Magnetic Resonance, 111 G. J. Ray, A. K. Colter, and R. J. Kurland, Chern. Phys. Letters, 1965, 2 , 324. 112 R. J. Pugmire and D. M. Grant, J . Amer. Chem. SOC., 1965,90, 697. 113 R. G. Rees and M. J. Green, J . Chem. SOC. (B), 1968, 387. 114 H. L. Retcofsky and R. A. Friedel, J. Phys. Chem.. 1968, 72, 2619. Toimetised, 1967, 3, 351. Toimetised, 1967, 4, 346. 1969, 1, 41.High Resohition Nuclear Magrietic Resonance Spectroscopy 77 Carbon-13 spin-lattice relaxation times are often long and their values are of some practical importance.Lippmaa and his co-workers115 have measured 13C spin-lattice relaxation times in several organic compounds. Many publications have appeared dealing with information obtainable from 13CH satellites in the proton spectra of molecules. In the analysis of the 13CH satellite spectra of para-dihalogenobenzenes it was found that long-range isotopic shifts must be introduced to account for the asymmetry in the upper and lower field l3CH satellites.l16 Studies of Other Nuclei.-Only a few of the numerous n.m.r. studies involv- ing other nuclei can be mentioned. Witanowskill7 has found empirical correlations between 14N shifts and molecular structure for sp2-hybridised nitrogen atoms in a large number of compounds. These correlations could be explained in terms of the ground-state molecular orbitals.It was possible to account for the influence of electronegativity of substituents on the nitrogen chemical shifts without reference to excited electronic states. The hetero- nuclear double-resonance technique has been used to measure 14N shifts indirectly in ureas,l18 thioureas,l18 and formamides.llg The 14N chemical shifts in formamide differ by as much as 10 p.p.m. on going from an infinitely dilute solution in acetone to one in methanol, resulting from hydrogen bonding of the methanol to the carbonyl group of the formamides. Oxygen-17 measurements are useful for studying solvation effects of metal ions, as was illustrated in an investigation of CoII in water and methanol solutions120 and also in aqueous hydrochloric acid solutions.121 By using 170-enriched methanol, separate 170 signals for co-ordinated and bulk methanol can be distinguished.To see separate signals the electron spin relaxation time of the paramagnetic ion must be short compared with the hyperfine interaction between unpaired electrons and the relevant nucleus. Also, the ratio of exchange between bulk and co-ordinated molecules must be sufficiently slow not to average out the separate signals. Hydration of API and GaIII ions has also been studied by novel 170 resonance measurements to determine co-ordination numbers122 and the lifetimes123 of the water molecules in the hydration sphere. When cobalt(I1) ions are added to the aqueous solutions, the 1 7 0 signal of the non-co-ordinated water molecules moves to lower field than that of the co-ordinated ones.Liquid Crystal Studies.-Molecular geometries, signs of coupling constants, and chemical shift anisotropies have been measured in several systems 115 A. Olivson, E. Lippmaa, and J. Past, Eesti N.S. V . Teaduste Akadeemia Toimetised, 116 J. M. Read, R. W. Crecely, and J. H. Goldstein, J. Mol. Spectroscopy, 1968,25, 107. 117 M. Witanowski, J . Amer. Chem. SOC., 1968, 90, 5683. 118 P. Hampson and A. Mathias, J . Chem. SOC. (B), 1968, 673. 119 M. Kamei, BUN. Chem. SOC. Japan, 1968, 41, 1030. l2* D. Fiat, 2. Luz, and B. L. Silver, J. Chem. Phys., 1968, 49, 1376. 121 A. H. Zeltmann, N. A. Matwiyoff, and L. 0. Morgan, J . Phys., Chem., 1968,72,121. m2 R. E. Connick and D. Fiat, J . Chem. Phys., 1963, 39, 1349. 123 D. Fiat and R. E.Connick, J . Amer. Chey. SOC., 1968, 90, 608. XVI Koide, 1967, 390.78 J. Feeney (e.g. 3,3,3-trifl~oropropyne,l~~ p-benz~quinone,~~~ thiophen,126 and 1,3,5- trifl~orobenzenel~~) by this elegant procedure.124 -127 When racemic 3,3,3- trichloropropylene oxide was examined in an optically active liquid crystal solvent two different spectra (d and 1) were observed.12* Diehl and his co- w o r k e r ~ ~ ~ ~ have used both direct and moment methods of n.m.r. spectral analysis for oriented molecules containing two or three spin-systems. They have developed a modified LAOCOON I1 computer programme which was able to analyse iteratively the spectra obtained for symmetrical orrho-disub- stituted benzenes dissolved in a nematic phase.130 From a study of the temperature and concentration dependence of liquid crystal spectra131 it is found that the degree of solute orientation increases gradually with decreasing temperature and increases with decreasing solute concentration : in both cases the lines sharpen when the degree of orientation increases.Whereas the concentration dependence is found to be typical of the solute, the temperature dependence depends on the liquid crystal solvent rather than the solute. The degree of orientation also depends on the spinning speed of the sample if the axis of rotation is not along the magnetic field direc- tion.131 For very high magnetic fields (50 kG) it is possible to spin samples at speeds of 250 Hz without destroying the orientation.131 However, such high fields are invariably obtained by use of superconducting magnets, where the axis of rotation is parallel to the direction of the magnetic field so that spinning cannot destroy the orientation regardless of its rate.This has been verified e~perimental1y.l~~ For lower fields, such as 14,000 G, to maintain132 orientation it is necessary to have slow spinner speeds of 3-20 Hz, depending on the liquid ~rysta1.l~~ The use of oriented molecules to measure proton chemical shift aniso- tropies is limited in accuracy because of contributions to the observed shifts from changes in solvent environment in the isotropic and nematic p h a ~ e . ~ ~ ~ ~ 13* This is illustrated by a study of molecular hydrogen in the nematic phase where the measured chemical shift anisotropies are much larger than good theoretical values.Extensions to the experimental technique have involved applying electric 124 A. D. Buckingham, E. E. Burnell, C. A. de Lange, and A. J. Rest, Mol. Phys., 125 P. Diehl and C. L. Khetrapal, Mol. Phys., 1968, 14, 327. 1% P. Diehl, C. L. Khetrapal, and U. Lienhard, Cunud. J . Chem., 1968, 46, 2645. 127 C. T. Yim and D. F. R. Gilson. Personal communication. 12* E. Sackmann, S. Meiboom, and L. C. Snyder, J . Amer. Chem. SOC., 1968,90, 2183. 1z9 P. Diehl, C. L. Khetrapal, and U. Lienhard, Mul. Phys., 1968, 14, 465. I3O P. Diehl and C. L. Khetrapal, Mul. Phys., 1968, 15, 201. 131 P. Diehl and C. L. Khetrapal, Mol. Phys., 1967, 14, 283. 132 A. D. Buckingham, E. E. Burnell, and C. A. de Lange, Mol. Phys., 1968,15,285. 133 A. D. Buckingham, E. E. Burnell, and C.A. de Lange, Chem. Comm., 1968, 1408. 134 A. D. Buckingham, E. E. Burnell, and C. A. de Lange, J . Amer. Chem. Suc., 1968, 1968,14, 105. 90, 2972.High Resolution Nuclear Magnetic Resonance Spectroscopy 79 fields135 (which can rotate the molecular axis of the liquid crystal by 90") and using cholesteric liquid crystal phases.136 There are two cases of cholesteric liquid crystals which must be distinguished. In one case if the long axis of the molecule tends to align parallel to the magnetic field then no macroscopic alignment can take place unless the magnetic field is strong enough to un- wind the helical structure, and only then will a high-resolution spectrum of the solute be obtained. For the second case (which includes cholesterol deriva- tives) the long axis of the molecule tends to align perpendicular to the field, and the axis of the helical structure aligns parallel to the'magnetic field, such that no unwinding takes place: in such cases high resolution n.m.r.spectra of solute molecules can be observed. Biological Studies.-The availability of very high magnetic fields by use of superconducting systems (220 MHz for protons) is proving particularly useful to those interested in biological problems, where one is often dealing with broad, complex, overlapping absorption bands.137 By use of results obtained from studying an extensive series of commonly occurring amino- acids dissolved in trifluoroacetic [lH]- and [2H]-acids it was possible to analyse the 220 MHz spectrum of insulin.138 Bradbury and his co-workers139 have made a study of conformational analysis in polypeptides by use of 220 MHz n.m.r.spectra. Bradbury and Crane-R~binsonl~~ have indicated some of the exciting possibilities available through n.m.r. studies to those interested in biopolymers. There has been increasing use of the technique to study conformational and binding problems in enzymes. Eventually it will be possible to obtain detailed information concerning the nature of the active sites and the types of interactions with the substrates.140 For example, Gerig141 has made quantitative estimates of the bonding of tryptophan to p-chymo- trypsin by measuring line broadening, and Cohen and Jardet~kyl~~ have studied the 1H spectrum of lysozyme. Inter- and intra-molecular interactions in solutions of adenine nucleotides and other mononucleotides have been st~died.l4~ The n.m.r.spectrum of adenosine monophosphate (AMP) shows a similar concentration dependence to that of adenosine and it is concluded that AMP also forms vertical stacks in solution. Conformational studies of other nucleotides14'* and nucleosides,l45 and determinations of helix-coil 135 E. F. Carr, E. A. Hoar, and W. T. MacDonald, J. Chern. Phys., 1968,48,2822. 136 E. Sackmann, S. Meiboom, L. C. Snyder, A. E. Meixner, and R. E. Dietz, J. Amer. 137 E. M. Bradbury and C. Crane-Robinson, Nature, 1968,220, 1079. 138 B. Bak, C. Dambmann, F. Nicolaisen, E. J. Pedersen, and N. S. Bhacca, J. Mol. 139 E. M. Bradbury, B. G. Carpenter, C. Crane-Robinson, and H. W. E. Rattle, Nature, l40 Article 'Enzymes and N.M.R.', Nature, 1968, 218, 1107.141 J. T. Gerig, J. Amer. Chem. SOC., 1968, 90, 2681. 142 J. S. Cohen and 0. Jardetsky, Proc. Nat. Acad. Sci. U.S.A., 1968, 60, 92. 143 M. P. Schweizer, A. D. Broom, P. 0. P. Ts'o, and D. P. Hollis, J. Amer. Chern. Suc., 144 M. Smith and C. D. Jardetzky, J. Mol. Spectroscopy, 1968, 28, 70. 145 R. J. Cushley, J. F. Codington, and J. J. Fox, Cunad. J . Chem., 1968,46, 1131. Chern. SOC., 1968, 90, 3567. Spectroscopy, 1968, 26, 78. 1968, 220, 69. 1968,90, 1042.80 J . Feeney transition temperature~l~~ for different molecular weight polypeptides have been reported. Conformation Studies.-Considerable doubt has been cast on much of the published work on conformation preferences ( A values) of substituents in cyclohexyl systems in which it has been assumed that the rigid model 4-t-butylcycloliexyl derivatives have the same chemical shifts for the axiaI and equatorial methine protons as exist in the simple cyclohexyl derivative.147 By cooling cyclohexyl derivatives down to -80” it was possible in some cases to observe directly the axial and equatorial methine shifts and when these were compared with the 4-t-butyl-derivative values the assumption is found to be invalid. Eliel and Martin148 had earlier arrived at the opposite conclu- sion for cyclohexyl derivatives, but the weight of evidence is against their findings.However, Eliel and Martin14* found that 4-t-butyl derivatives are not good models for fluorine nuclei or for proton shifts in heterocyclic compounds. Of the large selection of conformational studies reported only a few can be mentioned.The barriers to chair-twist interconversion in 4-membered rings containing a trigonal sp2-carbon atom (cyclohe~anone~~~ and methylene- cyclohe~ane~~~) have been reported. Conformational studies of 1,3-dioxans,151 N-methylhydra~ones,~~~ and N N’N”N’”-tetramethylhe~ahydrotetrazine~5~ have also been made. Restricted rotation about C-C bonds in cis- and trans- but-2-ene154 and about P-N bonds155 in dimethylamino(pheny1) phosphine chloride has been observed. Kinetic Studies.-There has been continued investigation into the question of the validity of spin-echo measurements for measuring chemical exchange rates. It was found that this method gives values of rate constants which are too high in the slow-exchange region and too low in the fast-exchange region compared with the rates obtained by the method of complete line- shape ana1y~is.l~~ Delp~echl~~ has considered the coalescence with tem- perature of a complete spectrum arising from exchange between two identical configurations of one molecule in terms of coalescence of simpler partial spectra and he was able to build up a general computer programme for handling such problems.146 E. M. Bradbury, C. Crane-Robinson, H. Goldman, and H. W. E. Rattle, Nature, 147 F. R. Jensen and B. H. Beck, J . Amer. Chem. Sac., 1968,90, 3251. 149 F. R. Jensen and B. H. Beck, J . Amer. Chem. SOC., 1968,90, 1066. 150 J. T. Gerig, J . Amer. Chenz. SOC., 1968, 90, 1065. 151 E. L. Eliel and Sr. M. Carmeline Knoeber, J . Amer. Chem. SOC., 1968, 90, 3444.lS2 C. J. Karabatsos and R. A. Taller, Tetrahedron, 1968, 24, 3557. 153 J. E. Anderson and J. D. Roberts, J . Amer. Chem. SOC., 1968, 90, 4186. 154 H. G. Hecht and B. L. Victor, J . Amer. Chem. SOC., 1968,90, 3333. lS5 A. H. Cowley, M. J. S. Dewar, and W. R. Jackson, J . Amer. Chem. SOC., 1968,90, 156 P. T. Inglefield, €3. Krakower, L. W. Reeves, and R. Stewart, Mol. Phys., 1968, 15, 157 J. J. Delpuech, Mol. Phys., 1968, 14, 567. 1968,217, 812. E. L. Eliel and R. J. L. Martin, J . Amer. Chem. SOC., 1968, 90, 682. 4185. 65.Irish Rmolutiori Nircleas Magnetic Resonance Spectroscopy 81 Miscellaneous Studies.-La~zlo~~~ has written a comprehensive review on solvent effects in n.m.r., and many papers, too numerous to mention, have appeared illustrating the important analytical usage of such effects.The reaction-field model has been used to explain medium effects on frozen-out rotamers of 1,1,2,2-tetrabromofluoroethane at low temperat~es.l5~ Bec- consalPO has derived expressions for the effects of magnetically anisotropic disc- and rod-shaped solvent molecules on the nuclear shielding of a spherical solute molecule. The binary collision gas model used for interpreting pressure dependence of chemical shifts in gases has been extended to liquids consisting of non- polar solutes in magnetically isotropic solvent molecules.161 By observing the temperature and pressure dependence of the proton signal of hydrogen chloride in the gaseous state in the presence of dimethyl ether, hydrogen- bonding parameters for the gas phase interaction could be worked out.lG2 The origin of the line broadening of the n.m.r. signals from protons in the a-position to nitrogen in nitrogen heterocyclic molecules has been considered by Kintzinger and Lehn :163 the broadening is attributed to incomplete removal of N-Ha spin coupling by 14N quadrupolar relaxation. A study has been made of cross-relaxation contributions to relaxation times in systems containing protons in two different environments;164 these contributions arise from interactions between the different protons and are additional to the contributions from interactions between identical protons. Studies of nuclei attached to quadrupolar nuclei continue to provide interest. The observed temperature changes in the lgF spectrum of the hexafluoro- niobate ion have been explained in terms of chemical exchange of the fluorine atoms and also the effects of the quadrupolar induced transitions between the spin states of the g3Nb n~c1eus.l~~ It has been observed166 that the extent of collapse of multiplet structures for protons attached to loB is greater than for the analogous llB compounds, because of the larger quadrupole moment of 1°B. The extent of collapse also increases with field gradients at the boron nuclei in different compounds. When the temperature is increased, the extent of collapse decreases, as a consequence of the decrease in correlation time of the molecular motion.166 Gore and Gutowsky167 have used a density-matrix formulation to study the transient n.m.r. effects occurring when the r.f. field HI is changed dis- continuously for an AB system. undergoing rapid intramolecular exchange. 158 P. Laszlo, Progr. N.M.R. Spectroscopy, 1967, 3, 231. 159 G. Govil and H. J. Bernstein, J. Chem. Phys., 1968, 48, 285. l 6 O J. K. Becconsall, Mol. Phys., 1968, 15, 129. 161 F. H. A. Rummens, W. T. Raynes, and H. J. Bernstein, J. Phys. Chem., 1968, 72, 162 G. Govil, A. D. H. Clague, and H. J. Bernstein, J. Chem. Phys., 1968, 49, 2821. 163 J. P. Kintzinger and J. M. Lehn, Mol. Phys., 1968, 14, 133. 164 A. A. Brooks, J. D. Cutnell, E. 0. Stejskal, and V. W. Weiss, J. Chem. Phys., 1968, 165 D. W. Aksnes, S. M. Hutchison, and K. J. Packer, MoI. Phys., 1968, 14, 307. 166 H. Watanabe, T. Totani, M. Ohtsuru, and M. Kubo, MoI. Phvs., 1968, 14, 367. 167 E. S. Gore and H. S. Gutowsky, J. Chem. Phys., 1968,48, 3260. 2111. 49, 1571.