2 Physical Methods Part (ii) Nuclear Magnetic Resonance By I. H. SADLER Department of Chemistry University of Edinburgh West Mains Road Edinburgh EH9 UJ The field of n.m.r. spectroscopy is so wide that it is impossible in a Report of this size to cover all aspects of the subject. This Report concentrates therefore almost exclusively on chemically induced dynamic nuclear polarization which receives little attention in other reports and pulse Fourier-transform spectroscopy which in recent years has become the routine method for the examination of low-abundance nuclei. A limited number of papers of general interest have been covered in detail rather than making an attempt to be comprehensive. Papers concerned solely with compilation of chemical shifts or coupling constants or with the application of well established procedures have been largely excluded.A small miscellaneous section is mainly devoted to shift reagents. For all other aspects the reader is referred to the second and third volumes of the Specialist Periodical Reports on Nuclear Magnetic Resonance edited by R. K. Harris which follow the layout of the first volume. 1 Chemically Induced Dynamic Nuclear Polarization A monograph’ devoted almost entirely to CIDNP presents some theoretical material and covers a large number of applications also including several dia- grams and figures not available elsewhere. A chapter on chemically induced dynamic electron polarization (CIDEP) is also included. A review2 describes the origin of CIDNP in a qualitative manner and surveys published work.The possibility of simplifying the analysis of CIDNP spectra by the use of lanthanide shift reagents has been in~estigated.~ The fluorinated chelates [Eu(fod),] and [Pr(fod),] can successfully be used to shift the resonance lines of products possessing atoms bearing lone electron pairs although increasing concentrations of shift reagent rapidly reduce signal intensities. At reagent :sub-strate molar ratios greater than ca. 0.05 signals are quenched. Intensities of pro-ducts which complex less well are reduced less so that certain lines may be ’ ‘Chemically Induced Magnetic Polarization,’ ed. A. R. Lepley and G. L. Closs Wiley New York 1973. * D. Bethel1 and M. R. Brinkman Ado. Phys. Org. Chem. 1973,10 53. J.Bargon J. Amer. Chem. SOC.,1973 95 941. 22 Physical Methods-Part (ii) Nuclear. Magnetic Resonance selectively removed. Similarly coinciding emission and absorption lines of different products which may lead to zero signal may be separated to reveal both polarizations. During reactions line positions approach the original chemical shifts partly owing to a decrease in the reagent :substrate ratio. Systems that produce shift-reagent poisons e.g. acids as reaction products are to be avoided. High-field CIDNP continues to prove useful in the study of the mechanisms of reactions. Some examples showing interesting or unusual features are reported here. Decomposition4 of benzoyl-P-bromopropionyl peroxide in bromotri- chloromethane produces after decarboxylation of the P-bromopropionyloxyl radical a singlet radical pair which yields polarized P-bromoethyl benzoate (1) as a recombination product and unpolarized 1,2-dibromoethane (2) as a transfer product (Scheme 1).The same result is obtained for the ester (1) using chloro- benzene in place of bromotrichloromethane. These results consistent with 0 S II BrCH,CH,Cyp PhCO; eH,CH,Br -+ eH,CH,Br I I PhCO BrCCI, 0 .t + kH,CH,Br PhCO,CH,CH,Br BrCH,CH,Br + 'CCI €A unpolarized H C-CH \/ Br A Absorption; E Emission; S Singlet radical pair; T Triplet radical pair; F Radical pair formed by diffusive encounter. Scheme i Kaptein's rule,5 imply that the two pairs of methylene protons in the bromo- ethyl radical have hyperfine couplings of opposite signs and are not equivalent thus supporting the linear structure (3).A popular alternative the bridged radical (4) can be excluded as a ground state for the bromoethyl radical since this would have equivalent methylene groups which would show the same polarization in the ester (1) and also result in a polarized signal for (2). This study does not however rule out participation by bromine in the decarboxylation step. During a kinetic study of the reaction of aniline with isoamyl nitrite in carbon tetrachloride at 55-60 "Ctwo time-separated maxima were observed6 for the 'J. M. Hargis and P. B. Shevlin J.C.S. Chem. Comm. 1973 779. ' R. Kaptein Chem. Comm. 1971 732. I. P. Gragerov A. F. Levit L. A. Kiprianova and A. L.Buchachenko Org. Magn. Resonance 1973.5. 445. I. H. Sadler emission signals of benzene and also for those of chlorobenzene. This implies that these products are generated by two consecutively formed phenyl radical sources which is consistent with a previously proposed' mechanism outlined in Scheme 2. PhNH + C,H,,ONO -P PhNHNO PhN=NOH Ph-N=N-0-N=N-Ph / -Nz I 5 Ph-N=N-0' 'Ph -+ Ph' -+ PhX X = H or C1 E I Ph-N=N-OPh PhO' 'Ph -+ Ph' -+ PhX Scheme 2 Detailed investigations of the reactions occurring during Grignard-reagent formation have been reported. Polarization observed for the main product RMgX and the hydrocarbon by-products R(H) and R(-H) are shown to arise from a radical pair 2R formed by diffusive encounters of radicals formed prior to Grignard-reagent formation.A comparison of the CIDNP spectra obtained' during the ferric salt-catalysed reaction between isopropylmagnesium bromide and isopropyl bromide the reaction with both reagents deuteriated at C-2 and the reactions with only one reagent deuteriated shows that polarization arises only for the hydrocarbon products propene and propane derived from the alkyl halides. Product analysis indicates that similar quantities of these hydrocarbons are also formed from the Grignard reagent indicating that this compound must react via a non-radical pathway. On the basis of CIDNP signals obtained during the reaction of p-chlorobenzenediazoniumfluoroborate with t-butylmagnesium chloride it has been proposed" that the reduction and addition reactions proceed via the same mechanism and are competitive.Polarizations obtained' for the photo-Fries rearrangement of p-cresyl p-chlorobenzoate to 2-hydroxy-4-methyl-4'-chlorobenzophenone show that the product is formed by cage recombination of a singlet radical pair. The enhance- ment factor for the polarization of the methyl protons was in agreement with the ' I. P. Gragerov and A. F. Levit Zhur. org. Khim. 1968.4 10; ibid. 1969 5 310. H. W. H. J. Bodewitz C. Blornberg and F. Bickelhaupt Tetrahedron 1973 29 719. R. B. Allen R. G. Lawler and H. R. Ward J. Amer. Chem. Soc. 1973.95 1692. lo V. I. Savin I. D. Ternyachev and Y. P. Kitaev Org. Magn. Resonance 1973,s. 449. ' ' W. Adam J. A. de Sanabia and H. Fischer J.Org. Chem. 1973.38,257 1. Physical Methods-Part (ii) Nuclear Magnetic Resonance theoretical value for an exclusive radical-pair route. The related photo-Claisen rearrangement has provided an opportunity to examine the difference in be-haviour of singlet and triplet radical-pair recombinations resulting from the difference in sign of the spin-density distribution in the radicals involved. During the photolysis of p-cresyl P-methallyl ether (5) the polarization (A) of the p-methyl protons in the meta-product (9) was opposite to the polarizations (E) of the corresponding protons in the ortho-product (8) and the para-product (7). This initially surprising result is interpretedI2 in Scheme 3. The spin-density Scheme 3 distribution in the singlet radical pairs suggests that these are likely to undergo recombination resulting in starting material (5) and products (7) and (8).Triplet radical pairs cannot form these products in view of the unfavourable spin pairing but have the correct phase for meta coupling to give the biradical(6) which under- goes spin inversion and bond reorganization to allow the formation of product (9). The possibility of the formation of (9) by other routes was eliminated. This system is unusual in that it allows in principle at least the observation of CIDNP W. Adam H. Fischer H.-J. Hansen H. Heimgartner H. Schmid and H.-R. Waespe Angew. Chem. Internat. Edn. 1973 12 662. I. H. Sadler effects of cage recombination with accompanying escape reactions although in practice escape products were detected.CIDNP can be extremely valuable in the detection of short-lived or unstable reaction products. The photoreduction of acetaldehyde (10) and the photo- cleavage of acetoin (12) appear to proceed through a radical pair formed by diffusive encounters or in the triplet state (Scheme 4). Polarized resonances due T.F 2MeCHO 3 MekHOH kOMe & MeCHOHCOMe J MkHOH + kOMe + AfE bisproportionation EtOH + (11) Scheme 4 to vinyl alcohol (1 1) are obtained’ although these regions are free of resonances after the reaction. Linewidths suggest that this intermediate has a lifetime not shorter than one second. The enol of acetone has similarly been detectedl4 during the photoreduction of acetone in propan-2-01 and that of acetophenone” during the photodecomposition of acetophenone in phenol.In these studies detection was possible as a consequence of the large signal enhancements. Detailed analyses of the CIDNP spectra obtained during the photolysis of pivalaldehydel6 and propionaldehyde’ have been given including the effect of solvent on the primary photochemical steps. CIDNP spectra have also been observed’* for the methyl triplet (AIE)of propyl acetate formed during the thermal decomposition of 2-hydroperoxy-2-methyltetrahydrofuran.Polarization is not normally observed during hydroperoxide decompositions owing to the small proton hyperfine coupling to an electron localized on oxygen. Decarboxylation of the initially formed radical generates an alkyl radical from which polarization may be observed.Emission signals are obtained” for the proton bound to nitrogen in the amide formed during the thermal decomposition of ethyl azido- formate in decalin that are consistent with an abstraction-recombination process. ‘H and 19Femission signals respectively have been obtained during the reaction l3 B. Blank and H. Fischer Helv. Chim. Acta 1973 56 506. I4 G. P. Laroff and H. Fischer Helv. Chim. Acta 1973 56 201 1. Is S. M. Rosenfeld R. G. Lawler and H. R. Ward J. Amer. Chem. SOC.,1973 95 946. H. E. C. Chen A. Groen and M. Cocivera Canad. J. Chem. 1973 51 3032. H. E. C. Chen S. P. Vaish and M. Cocivera Chem. Phys. Letters 1973 22 576. Is A. V. Ignatenko A. V. Kessenikh V. G. Glukhovtsev and M. A. Nadtochii Org. Magn.Resonance 1973,5 219. l9 M. R. Brinkman D. Bethell and J. Hayes Tetrahedron Letters 1973 989. Physical Methods-Part (ii) Nuclear Magnetic Resonance of benzenediazonium fluoroborate6*20 and of the p-fluoro-derivative” with alkoxide ions although the proposed rationalization in terms of one-electron transfer from the alkoxide ion to the diazonium cation must be reconsidered in view of material presented later. Evidence for the formation of the biradical(l3) during the thermal decomposition of a variety of possible precursors is furnished’ by the observation of emission signals from the biradical dimers (14) and (1 5). \ / 1 The application of CIDNP to the determination of the signs of hyperfine coupling constants (avalues) is of particular value since although the magnitudes of these parameters are directly available from e.s.r.spectra their signs are not always readily obtained by this technique. Photochemical decomposition of the corresponding diaroyl peroxides in carbon tetrachloride indicates” that the fluorine a values of the pentafluorophenyl radical are positive for the ortho and *O N. N. Bubnov B. Y. Medvedev L. A. Poljakova K. A. Bilevitch and 0. Y. Okhlobystin Org. Magn. Resonance 1973 5 437. * W. R. Roth and G. Erker Angew. Chem. Infernat. Edn. 1973,12,503,505; W. R. Roth M. Heiber and G. Erker ibid. 504. 22 H. D. Roth and M. L. Kaplan J. Amer. Chem. SOC.,1973 95 262. I. H. Sadler meta positions but negative for the para position with aFo> aFm> aFP,and that the fluorine a values for both 0-and p-fluorophenyl radicals are both positive with uFo> aFp.Such results are in accord with a radical having no appreciable n-character.A similar of the photochemical decomposition of diphenyl-diazomethane and its di-p-fluoro-derivative in 0-and in m-fluorobenzyl chlorides confirms that for a benzyl radical the fluorine a value is positive for ortho-substitution but negative for meta-substitution by fluorine. CIDNP of nuclei other than 'H and "F are being increasingly studied. The low-abundance nuclei '3C and ''N have a number of advantages. Their hyperfine coupling constants can be considerably larger the nuclear relaxation times longer and the spectra are generally simpler and easier to interpret in detail than those of protons.Polarizations of both these nuclei have been observed24 during the thermal decomposition of diazoaminobenzene (16) in cycloheptanone enhancement factors being in the region of 20000 for the quaternary C atoms in the diphenylamine and biphenyl formed no 3C polarization being observable in the starting material. The use of diazoaminobenzene enriched with ''N in the central position yielded ''N polarization in the starting material only. Diazo- aminobenzene enriched at the terminal positions showed no polarization in the starting material but large polarizations with enhancement factors of at least loo00 for aniline (A) and diphenylamine (E). Such results are consistent with the reversible decomposition of diazoaminobenzene to give a phenylamino-phenyl- diazo radical pair in which the latter must be a localized a-radical in view of the apparently much greater hyperfine coupling to the radical site.A similar study25 of the azo-coupling reaction of benzenediazonium fluoroborate with alkaline solutions of phenol in methanol indicated that the coupling process is unlikely Ph-N=N-NH-Ph Ph-NZN' HNPh Ph -dimerization Ph' +-Ph' HNPh -.) PhNHPh to be that of a combination of a radical pair formed by one-electron transfer from the phenoxide ion to the diazonium cation as had previously been proposed.26 Strong approximately equal polarizations of the quaternary carbon (' 3C) and both nitrogen atoms ("N) of the starting diazo-compound implied that the Ph-NZN' radical was not involved (in view of the triazene study referred to above) and that in the radical responsible the electron must be strongly delocalized 23 M.R.Brinkman D. Bethell and J. Hayes J. Chem. Phys. 1973 59 3431. 24 E. Lippmaa T. Saluvere T. Pehk and A. Olivson Org. Magn. Resonance 1973,5,429. '' E. Lippmaa T. Pehk T. Saluvere and M. Magi Org. Mugn. Resonance 1973,5441. 26 N. N. Bubnov K. A. Bilevitch 1. A. Poljakova and 0.Y. Okhlobystin J.C.S. Chem. Comm. 1972 1058. Physical Methods-Part (ii) Nuclear Magnetic Resonance with practically equal hyperfine coupling constants to both nitrogen atoms. A possible scheme is proposed in which an unstable diazoanhydride Ph-N=N-0-N=N-Ph forms a Ph-N=N-0' (N2)Ph' radical pair in which spin selection occurs. Reaction of the polarized phenylazoxy-radical with methanol regenerates a polarized diazonium ion which may undergo coupling.Since the reaction is very fast and the nuclear relaxation times relatively long + Ph-N2-O' + MeOH -+ PhN + 'CH,OH + OH-practically all the polarization is carried over to the coupling product even though the last stage is a heterolytic process. The suggestion that this strongly delocalized radical is a phenylazoxy-radical is however in conflict with e.s.r. data for another radical assigned2' the same formula in which the electron appears to be largely confined to the oxygen atom. It is clear that the use of 15N CIDNP is going to be of great value in the study of reactions of this type of compound and of azo- compounds in general. Comparison of the results2* of 13C and proton CIDNP studies on the de- composition of acetylbenzoyl peroxide shows that for the benzoyl radical the sign of the I3C a values for C-1 and C-2 (C-6) of the ring are positive and negative respectively.A detailed of the thermal decomposition of the diazacyclo- hexanone (17) in perchlorinated solvents in the presence and in the absence of added alkyl iodide reveals that a singlet carbene (18) is first formed which reacts to give a singlet radical pair. Intersystem crossing to the triplet state and cage and escape reactions occur at comparable rates and the reaction product (19) is polarized largely in secondary reactions involving diffusive encounters of free radicals. A free-radical pathway for the rearrangement of the oxime (20)to the thio-oxime (21) is proposed30 on the basis of polarized I3C resonances of the C-1 and C(=N) atoms in the product.It cannot however be too strongly emphasized that the observation of CIDNP effects alone does not rule out the existence of alternative non-radical routes for product formation as a recent rearrangement study3 has shown. '' J. I. G. Cadogan R. M. Paton and I. C. Thomson. Chem. Comm. 1969 614. 2a A. V. Kessenikh P. V. Petrovskii and S. V. Rykov Org. Magn. Resonance. 1973 5 227. 19 G. A. Nikiforov S. A. Markaryan I. G. Plekhanova B. D. Sviridov S. V. Rykov V. V. Ershov A. L. Buchachenko T. Pehk T. Saluvere. and E. Lippmaa Org. Magn. Resonance 1973 5 339. 'O C. Brown R. F. Hudson and A. J. Lawson J. Amer. Chem. Soc. 1973.95 6500. 3' W. D. Ollis. I. 0.Sutherland and Y. Thebtaranoth. J.C.S. Chem. Comm. 1973 654. I. H. Sadler (20) Ar = C,H or 3-FC6H (21) 2 Pulsed Fourier-transform Spectroscopy Theoretical Aspects.-A computer program has been devised3' on the basis of the Bloch equations to simulate the behaviour of systems of non-coupled species in multipulse and Fourier-transform n.m.r. experiments. The effect on the magnetization vectors may be calculated for any sequence of pulses and delays. Such a program is useful for the evaluation of errors in the choice of experimental parameters such as pulse angle phase and timing which are critical when measurements of physical quantities such as relaxation times are attempted. It is also valuable in the design of new experiments.Computer simulation may indicate possible difficulties and whether a proposed new experiment may have any real advantage over existing experiments. Amongst the examples considered the conventional repetitive pulse-delay Fourier-transform sequence the measure- ment of spin-lattice (TI)and spin-spin (T,) relaxation times and the time-shared n.m.r. experiment. In addition to obtaining good agreement with experimental results it is shown that systematic noise in the conventional Fourier-transform sequence may be eliminated without loss in sensitivity by alternating the phases of consecutive pulses by 180". The effect of using a finite pulse power is also ex- amined. The advantage of the Bloch-equation approach is that quite complex pulse sequences can be treated without using large amounts of computer time.The major drawback is that tightly coupled spins cannot be examined. A second program devised3 on the basis of the Schrodinger equation has been developed to examine these systems; however the complexity of pulse sequences capable of being examined by this approach is quite restricted. The effect of the decoupling power level on the lineshape and Overhauser enhancement on a system AX is examined and the results suggest that in systems when the dipolar contribution (TI,&to the spin-lattice relaxation time (TI)is much greater than the effective spin-spin relaxation time (Tr),as is often satisfied in practice it should be possible by suitable choice of a very low decoupling power to obtain a nuclear Overhauser enhancement without decoupling thus providing an inexpensive alternative to gated de~oupling.~~ The effect of a finite pulse power for a 90" pulse on non-first- order spin systems the time development of the Overhauser enhancement and the effect of selective and non-selective 180" pulses in a 180°-o-900 sequence on AX and AB systems are also examined.Criteria are also establi~hed~~~~~ which must be met if accurate T2 values are to be obtained using the Carr-Purcell sequence and its variations. 32 P. Meakin and J. P. Jesson J. Magn. Resonance 1973 10 290. " P. Meakin and J. P. Jesson J. Magn. Resonance 1973 11 182. 34 0.A. Gansow and W. Schittenhelm J. Amer. Chem. SOC. 1971,93,4294. '' R. L. Vold R. R. Vold and H.E. Simon J. Magn. Resonance 1973 11 283. Physical Methods-Part (ii) Nuclear Magnetic Resonance 31 The saturation behaviour of spin systems in pulsed experiments has been inve~tigated~~ theoretically and compared with slow-passage continuous-wave experiments with particular reference to intensity measurements. Uncoupled spin systems (e.g. 3C resonance with simultaneous proton decoupling) show saturation behaviour qualitatively similar to that observed in slow-passage experiments. The maximum signal per unit time may be obtained for a pulse interval of T1and pulse angle of 69" if TI = T2 and any residual transverse magnetization is destroyed before a repeat pulse. Considerable differences in relative line intensities may be expected if a high pulsing rate is employed or if large differences in Tl values exist.In strongly coupled spin systems (e.g. proton and fluorine resonance) however relative signal intensities from pulse experi- ments are considerably less sensitive to saturation than those from slow-passage experiments in which inhomogeneous saturation occurs. Provided spin-lattice relaxation is dominated by an external dipolar process such as would be caused by dissolved oxygen a pulsed experiment will cause homogeneous saturation thus giving correct relative line intensities for a coupled system and is particularly suited for accurate measurements. In such cases it is advisable not to degas the samples and also to add a paramagnetic impurity to enhance external relaxation and equalize the relaxation rate.Except in the two-spin-; system intramolecular dipolar relaxation leads to slightly inhomogeneous saturation. General Techniques.-In the conventional pulsed Fourier-transform experiment the sample resonances are excited by a series of equally spaced radiofrequency pulses of constant width and amplitude. This is equivalent to simultaneously supplying a set of closely spaced frequencies of approximately equal power intensities throughout the spectrum. Fourier analysis of the excitation sequence yields amplitudes and phases of the closely spaced frequencies. Conversely the excitation could be Fourier-synthesized from the frequency spectrum. A method has been described3' for sample resonance excitation equivalent to simultaneously providing a range of frequencies having any pre-specified power-intensity dis- tribution.This is accomplished by computing the Fourier synthesis of the desired frequency distribution and using the function so obtained to amplitude- or width-modulate a sequence of radiofrequency pulses. The n.m.r. spectrum is obtained by Fourier transformation of the response of the spin system to this excitation. This capability has a number of advantages. Strong residual solvent lines may be selectively suppressed or removed while observing all the other spectral lines by choosing a frequency distribution which has zero intensity in the appropriate region and uniform intensity elsewhere. Unlike alternative solvent resonance elimination methods phase anomalies are not introduced into the final spectrum.Simultaneous homonuclear decoupling of one or more resonances may be carried out by specifying very high ( x 300) power intensities at selected positions in the frequency distribution. This has a practical advantage over the use of a conventional decoupler which requires to be frequency-locked 36 R. R. Ernst and R. E. Morgan Mol. Phys. 1973,26,49. 37 B. L. Tomlinson and H. D. W. Hill J. Chem. Phys. 1973 59 1775. 32 I. H. Sadler to the magnetic field and can only irradiate one frequency or frequency band at medThis technique should permit greater flexibility than has previously been possible in standard n.m.r. experiments and should allow extra freedom in the design of new experiments. Resolution enhancement in a conventional pulsed Fourier-transform experi- ment is normally carried out by multiplying the accumulated free induction decay [FID) by an exponentially increasing function before Fourier transformation.This procedure increases the weighting of the FID as a function of time resulting in narrower lines at the expense of the signal to noise ratio. This technique has been succ%ssfully applied * to resolve ’B-’ ’B and B-I H couplings in the ’B n.m.r. spectra of boron hydrides which normally show unresolved broad bands. A more abrupt FID weighting procedure has been found suitable39 in the case of non-proton-decoupled ’3C decays which normally show an initial decay followed by a series of beats. If only the beat pattern is transformed i.e. giving a zero weighting to the initial decay a highly resolved spectrum is obtained.The usual baseline distortions introduced into the final spectrum by truncation of the FID are negligible as this takes place between the end of the initial decay and the development of the first beat. Alternative methods4’ of resolution enhancement involve the subtraction of a ‘broadened’ spectrum from a normal one. The broad- ening may be achieved in a non-selective manner by multiplying a normal FID by an exponentially decreasing function as in signal to noise enhancement procedures or in a selective fashion by the use of a paramagnetic ion which binds selectively to a particular site in the molecule broadening only some of the resonances. In the latter case the difference spectrum obtained identifies signals from nuclei close to the binding site.These methods have been of value in the proton n.m.r. studies of proteins. INDOR spectroscopy in which the intensity of a single line is monitored while sweeping a second irradiating field through the remainder of the spectrum is a valuable tool for the analysis of complex spectra but is unfortunately confined to continuous wave (CW)operation. Two methods have been presented by which the same information may be obtained in a pulsed Fourier-transform experiment. In each case a single resonance spectrum obtained in the usual pulse-transform manner is subtracted from the spectrum of the sample in which the populations of the pair of levels associated with a particular transition are disturbed from the equilibrium distribution but otherwise obtained under the same conditions.The methods differ in the way in which the populations are altered. In one meth~d,~’ referred to as Fourier transform pseudo INDOR spectroscopy (FT$I) a weak irradiating field is applied continuously at the transition frequency. In the other method,42 referred to as selective population inversion (SPI) the ‘observing’ 38 A. 0.Clouse D.C. Moody R. R. Reitz T. Roseberry and R. Schaeffer J. Amer. Chem. SOC.,1973 95 2496. 39 W. B. Moniz and S. A. Sojka J. Magn. Resonance 1973 12 214. 40 I. D. Campbell C. M. Dobson R. J. P. Williams and A. V. Xavier J. Magn. Resonance 1973 11 172. 41 J. Feeney and P. Partington J.C.S. Chem. Comm. 1973 61 1. 42 K. G. R.Pachler and P. L. Wessels J. Magn. Resonance 1973 12 337. Physical Methods-Part (ii) Nuclear Magnetic Resonance pulse is immediately preceded by a selective 180"pulse at the transition frequency thereby inverting the populations of the connected levels only. This pre-pulse must be short compared with the 7''values of the nuclei yet of sufficiently low power not to perturb other levels. In such experiments the frequencies of the lines and the signs of their intensities in the difference spectrum so obtained are identical with those in the normal CW INDOR spectrum in which the same transition has been monitored. This follows since the progressive or regressive nature of a pair of transitions is apparent regardless of which of the pair is irradiated.The actual intensity changes however need not be the same. Spin-lattice relaxation times (Tl) are most commonly measured using the 'inversion-recovery' pulse sequence (T-180-~-90) which requires a delay (T)of about five times the longest TI in order that the spin system attains Boltzmann equilibrium before repetition of the initial 180" pulse. A new sequence43 elimin- ates this delay and also removes the necessity of producing precise 90" pulses (Scheme 5). Initially a homogeneity-spoiling pulse (HSP) ensures that the net lHSPt 1 Scheme 5 magnetization (M)is in the z-direction i.e.no transverse magnetization is present. A 90" pulse follows which shifts the net magnetization into the x-y plane and a second HSP rapidly dephases the nuclear spins (p),resulting in zero magnetization in all directions.The spin system is then allowed to relax partially during a delay (z) resulting in a new magnetization along the z-axis which is sampled with a 43 G. G. McDonald and J. S. Leigh. J. Magti. Resonance 1973 9. 358. 34 1. H. Sadler second 90"pulse and the FID signal acquired. The sequence (HSP-~C~HSP-T-~O) may then be repeated immediately. The peak intensities (A,) are given by A = A [l -exp (-z/T1)],where A represents the corresponding intensity for the spin system at equilibrium. Excellent agreement was obtained for the Tl values for H-4 and H-7 in tryptophan determined by this sequence and by the inversion- recovery technique the latter procedure taking about four times as long for a given number of transients.This sequence should therefore be particularly valuable in the measurement of I3C TI-values. This scheme is also suitable for the elimination of residual solvent resonance when obtaining proton spectra of dilute samples in deuterium oxide. A chemical method44 has been used for the removal of oxygen from samples to be used for TI determinations where conven- tional degassing techniques proved inadequate. Carbon-13 Techniques and Results.-A comprehensive book45 on I3C magnetic resonance covering the literature up to mid-1970 is now available. All chemical shift data have been referred to TMS as is now general practice. Since much of the work was written before the commercial availability of Fourier-transform n.m.r. spectrometers continuous-wave techniques are covered more extensively than pulsed methods.This book complements the other available on 13C n.m.r. I3Cmagnetic resonance is now used continually as an aid in molecular struc- ture determination in all branches of organic chemistry particularly in areas concerning natural products with complex molecular skeletons. The use of single-frequency off-resonance decoupling allows in principle at least the identi- fication of methyl methylene methine and quaternary carbon atoms from the residual splitting it being normally assumed that these carbon resonances appear as quartets triplets doublets and singlets respectively. However caution should be exercised in the interpretation of such spectra since a triplet is only guaranteed from a methylene group where the two protons have the same chemical shift.Otherwise the multiplet pattern depends upon the position of irradiation with respect to the proton4' resonances. Two lines are obtained if the irradiation lies mid-way between the proton resonances three lines are obtained if irradiation is on one of the proton resonances and four lines are obtained if irradiation is elsewhere. In this latter instance the two centre lines are often very close together and the patterns are often somewhat indistinct for a rigid or semi-rigid -CH2CH2-fragment.48 Such effects may be expected for complex natural products and once recognized may be diagnostically useful. Unusual effects are also sometimes observed49 when there is strong coupling between two or more 44 J.Homer A. R. Dudley and W. R. McWhinnie J.C.S. Chem. Comm. 1973 893. 45 J. B. Stothers 'Carbon-13 N.M.R. Spectroscopy,' Academic Press New York 1973. 46 G. C. Levy and G. L. Nelson 'Carbon-13 N.M.R. for Organic Chemists' Wiley New York 1972. 47 E. Wenkert D. W. Cochran E. W. Hagaman F. C. Schael N. Neuss A. S. Katner P. Potier C. Kan M. Plat M. Koch M. Mehri J. Poisson N. Kunesch and Y. Pollard J. Amer. Chem. SOC.,1973 95 4990. 48 I. H. Sadler unpublished observation. 49 R. A. Newmark and J. R. Hill J. Amer. Chem. Soc. 1973,95,4435. Physical Methods-Part (ii) Nuclear Magnetic Resonance 35 protons with the same chemical shift but bonded to different carbon atoms in a molecule. In such cases an effect is observed similar to ‘virtual coupling’ in proton n.m.r.spectra. The olefinic carbon atom in fumaric acid appears as a quintet whose line intensities and positions depend upon the position and intensity of the irradiating frequency. This arises because the effective 13C-’H coupling constant becomes reduced to approximately the same size as the ‘H-’H coupling constant and the carbon atom appears similarly coupled to both protons. In a complex single-frequency off-resonance-decoupled spectrum the quaternary carbons may not be readily identified owing to their reduced intensity resulting from their longer relaxation times and smaller nuclear Overhauser enhancements. By the use of low power for proton noise decoupling quaternary 13C n.m.r. signals may be increased relative to those from hydrogen-bearing carbon atoms which are considerably reduced broadened or eliminated.This may be achieved directly” by the use of a low decoupling power (ca. 0.1 W) centred on and distributed over the entire proton spectrum or indirectly5 by employing narrow- band (ca.300 Hz) proton decoupling at the more usual power levels applied well upfield (6-8p.p.m.) from tetramethytsilane as in an off-resonance experiment. In such experiments it is possible to obtain a significant signal from a methylene carbon atom if the two bonded protons have the same chemical shift ;this situation is easily recognized but is not as common as might be supposed and the technique is particularly valuable in the case of relatively rigid asymmetric molecules.Resonances from protonated carbon atoms can often be assigned by selective proton-decoupling experiments or graphically52 from a series of off-resonance- decoupled spectra. This latter method has recently been shown applicable53 to molecules such as 1-nitronaphthalene that have very complex non-first-order proton spectra providing that analysis of the proton spectrum is possible. In alkyl-substituted aromatic and heteroaromatic molecules it is possible5* to distinguish between carbon atoms two or three bonds distant from the alkyl protons and ring carbon atoms more than three bonds away since the former under selective irradiation of the aromatic protons show residual splitting due to long-range l3C-IH coupling with the alkyl protons. Thus C-1 C-2 and C-6 in toluene and the quaternary carbons in tris-(3-methylphenyl)phosphinemay be readily assigned.The assignment of signals from quaternary carbon atoms is however usually less readily accomplished. For the sesquiterpenoid lactone melampodin this has been achieved55 by observing small but definite intensity increases which occur in selectively decoupled spectra where the decoupling frequency is centred on protons separated from the quaternary atom by two or by three bonds rather than on more distant protons. In large isotropically re-50 I. H. Sadler J.C.S. Chem. Comm. 1973. 809. E. Wenkert A. 0. Clause D. W. Cochran and D. Doddreil J. Amer. Chem. Soc. 1969,91 6879; A. Allerhand R. F. Childers and E. Oldfield Biochemistry 1973 12 1335 and references therein.’* B. Birdsail and J. Feeney J.C.S. Perkin 11 1972 1643. ’’ J. W. Emsley J. C. Lindon and D. Shaw J. Mugn. Resonance 1973 10 100. ’’ S. Ssrensen M. Hansen and H. J. Jakobsen J. Mugn. Resonance 1973 12 342. ’’ N. S. Bhacca F. W. Wehrli and N. H. Fischer J. Org. Chem. 1973,38 3618. 36 I. H. Sadler orientating molecules assignment can sometimes be made from Tl values since in such molecules quaternary carbon atoms are largely or completely relaxed by the intramolecular dipole-dipole mechanism. Since the nearest protons are generally those bound to a-carbon atoms the relaxation rate will be very roughly proportional to the number of a-carbon atoms. This te~hnique,'~ in conjunction with shielding arguments has allowed unambiguous assignments in the alkaloids codeine and brucine.Tl values have also been used for the assignment of proton- ated carbon atoms (see below). In some circumstances assignments may be made by careful visual examination of undecoupled 13Cn.m.r. spectra. The carbon atoms a and P to the substituents in symmetrically ortho-disubstituted benzenes show characterization patterns which may be used as finger prints.57 In anthracene benzocyclobutane 43-benzoxepin indane and naphthalene the fine structure of thedoublet components that result from the one-bond I3C-lH coupling leads to broad multiplets for the a-carbon atom and distinct doublets sometimes sharply split into further doub- lets for the P-carbon atom. Where the chemical shift difference between the a-and b-protons is smaller than 10 Hz however the pattern for the P-carbon atom becomes indistinguishable from that of the a-carbon atom.In theory complete analysis of the region would lead to the same conclusion. However this is often tedious and difficult and not always possible; in such cases 'finger prints' are extremely valuable. The chemical shift parameters58 devised by Grant and Paul from the study of a series of predominantly linear alkenes and used for 3C resonance assignments have been reviseds9 to include also highly branched alkanes and are particularly suitable for predicting shifts in olefin polymer sequences. No additional param- eters have been introduced and most deviations between calculated and measured values are considerably less than 1 p.p.m.Further analysis60v61 of the spectra of methyldecalins has led to a new and simpler set of substituent and conformational parameters6' applicable to aliphatic hydrocarbons in general although it is not always possible specifically to assign resonances closer than 2 p.p.m. Parameters for the prediction of carbon shifts in a fragment XCH,CH,Y have beendeduced62 for a series of 15 different substituents. On the basis of a set of predictive rules,63 a Fortran IV computer program entitled AMINE has been compiled64 to deduce the structures of alicyclic amines from their empirical formulae and fully proton- decoupled 13C n.m.r. spectra. Tests on over a hundred amines show that the program is accurate and selective even for large amines with many millions of 56 F.W. Wehrli J.C.S. Chem. Comm. 1973 380. 57 H. Giinther H. Schmickler and G. Jikeli J. Magn. Resonance 1973 11 344. '* D. M. Grant and E. G. Paul J. Amer. Chem. Soc. 1964,86 2984. 59 C. J. Carman A. R. Tarpley and J. H. Goldstein Mncromolecules 1973 6 719. 6o D. K. Dalling and D. M. Grant J. Amer. Chem. Soc. 1972 94 5318. 61 D. K. Dalling D. M. Grant and E. G. Paul J. Amer. Chem. SOC.,1973,95 3718. 62 G. E. Maciel L. Simeral P. L. Elliott and K. Cribley J. Phys. Chem. 1972 76 1466; L. Simeral and G. E. Maciel J. Phys. Chem. 1973 77 1590. " H. Eggert and C. Djerassi J. Amer. Chem. SOC.,1973 95 3710. 64 R. E. Carhart and C. Djerassi J.C.S. Perkin 11 1973 1753. Physical Methods-Part (ii) Nuclear Magnetic Resonance 37 structural isomers.A list of sources of predictive rules for other classes of organic compounds is also given. Numerous papers have been concerned with the measurement of 3C chemical shifts or I3C-lH coupling constants and relatively few with 13C-13C coupling constants. The effect of deuteriation on 13C chemical shifts and 13C-1H(2D) coupling constants of a number of organic compounds has been e~amined.~’ Upfield shifts of up to 1.4p.p.m. are observed for the deuteriated carbon atom. Isotope effects on shifts correlate with hybridization and electron withdrawal at the deuteriated carbon within a comparable series of compounds. Coupling constants are within 1 Hz of the predicted value [J(C H)/J(C D) = yH/yD]. The incorporation of deuterium atoms does not appear to alter the I3C-lH coupling constants.The correlation between hybridization and directly bonded 3C-1 H and 13C-13C coupling constants is an accepted method for the estimation of s-character of a C-H or C-C bond most studies having been restricted to alicyclic hydrocarbons. The determination66 of 3C-1 H coupling constants for carbonyl compounds in the bicyclo[2,2,l]heptane series shows that compared with alicyclic and monocyclic ketones coupling constants are increased for bridgehead carbon atoms o! to the carbonyl group. It is proposed that this is due to a change in hybridization resulting from a hyperconjugative interaction between the carbonyl group and a strained a-bond of the molecular skeleton. The ring-current effect important in proton studies has received comparatively little attention in 3C n.m.r.Since the absolute value of this effect a few p.p.m. at the most from proton n.m.r. is independent of the nucleus involved its importance for nuclei with wide chemical shift ranges might be expected to be minimal. Comparison of systems where ring currents may be present with geometrically similar systems with no ring current such as [12]paracyclophane with cyclopenta- de~ane~~ or bridged annulenes with corresponding ‘no ring current’ molecules,6* shows that a diamagnetic ring current does in some cases have a small effect on I3Cchemical shifts but that in most cases it is masked by other factors such as the immediate electronic environment and paramagnetic contributions to shielding that affect the shifts more strongly.There appears to be little evidence for a paramagnetic ring-current effect on the I3C resonances of biphenylene and related additivity rules accounting satisfactorily for the resonance positions. These results contrast strongly with those from proton resonance where ring currents when present dominate shifts. Thus I3C resonance cannot be used as a probe for the magnetic properties of cyclic n-electron systems. This is clearly evident” in the bridged systems (22)-(25). Proton spectra show that 65 H. N. Colli V. Gold and J. E. Pearson J.C.S. Chem. Comm. 1973,408; E. Breitmaier G. Jung W. Voelter and L. Pohl Tetrahedron 1973 29 2485. 66 N. H. Werstuik R. Taillefer R. A. Bell and B. Sayer Canad. J. Chem. 1973 51 3010.67 R. H. Levin and J. D. Roberts Tetrahedron Letters 1973 135. 68 H. Gunther H. Schmickler H. Konigshofen K. Recker and E. Vogel Angew. Chem. Internat. Edn. 1973 12 243. 69 A. J. Jones P. J. Garrett and K. P. C. Vollhardt Angew. Chem. Internat. Edn. 1973 12. 241. ’’ H. Gunther H. Schmickler U. H. Brinker K. Nachtkamp J. Wassen and E. Vogel Angew. Chem. Internat. Edn. 1973 12. 760. I. H. Sadler (22)and (24)are clearly olefinic whereas (23) and (25) are delocalized [14]annulenes. Maximum agreement of "C resonances however emphasizes the stereochemical similarities in pairing (22) with (23) and (24) with (25). Thus '3C resonance is here more strongly influenced by ring size strain and conformation. An erroneous suggestion7'that (24) is homoaromatic arises from incorrect ''C assignments.Increasing use of 13C n.m.r. is being made in the elucidation of biosynthetic pathways by the incorporation of 'C-enriched precursors. Comparison of line intensities in the spectra of products obtained with and without labelled precur- sors indicates the sites of incorporation. Primary and secondary pathways have been deduced for the formation of prodigiosin from Serratia marcescens using labelled amino-acid~,~' and similarly for streptovaricin from Streptornyces spectabilis using carboxy-labelled pr~pionate.'~ A new pathway for the bio- synthesis of mollisin from Mollisia caesia has been deduced from 13C-'3C coupling of resonances where doubly labelled acetate was used.74 Since the pre- paration of '3C-enriched materials leaves considerably larger quantities of '2C-enriched materials it has been proposed7' that "C labelling may become a cheaper method than 13C labelling.In such cases identification would be by missing lines rather than by enhancement. Variable-temperature ''C studies have provided (a)evidence76 for the presence of crown conformations in cyclo-octane and its derivatives (b)evidence77 for the valence-tautomeric equilibrium in the 1,6-methano[ lolannulene system (26)*(27) although observation of the individual tautomers was not possible and (c) an estimation78 of the free energy of activation (10.8kcal mol- ') of the 71 E. Wenkert E. W. Hagaman L. A. Paquette R. E. Wingard and R. K. Russel J.C.S. Chem. Comm. 1973 135. 12 H. H. Wasserman R.J.. Sykes P. Peverada C. K. Shaw R. J. Cushley and S. R. Lipsky J. Amer. Chem. SOC.,1973 95 6874. 73 B. Milavetz K. L. Rinehart J. P. Rolls and W. J. Haats J. Amer. Chem. SOC.,1973 95 5793. 74 H. Seto L. W. Cary and M. Tanabe J.C.S. Chem. Comm. 1973 867. 75 S. B. W. Roeder J. Magn. Resonance 1973 12 343. 76 F. A. L. Anet and J. T. Basus J. Amer. Chem. SOC.,1973 95 4424. 77 H. Gunther H. Schmickler W. Bremser F. A. Straube and E. Vogel Angew. Chem. Internat. Edn. 1973 12 570. 78 S. Masamume A. V. Kemp-Jones J. Green D. L. Rabenstein M. Yasunami K. Tabase and T. Nozoe J.C.S. Chem. Comm. 1973 283. Physical Methods-Part (ii) Nuclear Magnetic Resonance (27) (26) R = Me or CN degenerate rearrangement of tropolone derivatives (28).The use7' of a probe de- signed to take 20 mm diameter spinning sample tubes in a field of 14.2 kG shows that a given signal to noise may be achieved in about one twentieth of the time required on commercial instruments employing 12 mm tubes operating above 20 kG.This is of particular value in the study of concentration-limited materials such as non-degraded natural biopolymers. If a full Overhauser enhancement is obtained the lowest acceptable concentration is about 2 mM if an acceptable spectrum is to be obtained in 10h of accumulation. Numerous single-carbon resonances were observed for the first time in the aromatic carbon region of hen egg-white lysozyme. Further details have been presented" concerning the method for obtaining enhanced high-resoltion n.m.r.spectra of low-abundance nuclei (e.g. 'jC "N) in the presence of high-abundance nuclei (e.g. 'H) in solids. Carbon-13 Relaxation Studies.-A well presented review has appeared concerning I3C spin-lattice relaxation studies and their value in organic chemistry.8' The concepts of the rotating-frame and spin-lattice relaxation mechanisms are dis- cussed in a non-mathematical fashion. Methods for the measurement of TI values and for the differentiation of relaxation mechanisms are examined along with applications to the determination of molecular structure molecular tum- bling and rotational and segmental motion. The review is strongly recommended. A comprehensive study8* of spin-lattice relaxation in benzene and its deriva- tives has been reported.Some of this work appeared as preliminary communi- cations and was described in last year's Report. New results are given here. The general assumption that intermolecular '3C-'H dipole4ipole contributions to 13C relaxation are negligible was shown to be true even for non-protonated carbon atoms in toluene by studying dilute solutions of toluene in perdeuterio- toluene. These yielded the same values as were obtained with neat toluene. '' A. Allerhand R. F. Childers and E. Oldfield J. Magn. Resonance 1973 11 278; Biochemistry 1973 12 1335. 8o A. Pines M. G. Gibby and J. S. Waugh J. Chem. Phys. 1973,59 569. G. C. Levy Accounts Chem. Res. 1973 6 161. 82 G. C. Levy J. D. Cargioli and F. A. L. Anet J. Amer. Chem. Soc. 1973.95 1527. I.H. Sadler Shorter T values for carbon atoms para to essentially symmetrical substituents indicate preferred rotation of the molecule about an axis bisecting the substituent and the ring. This sometimes allows the assignment of resonances of protonated carbons where chemical shift arguments are inadequate. In 3-bromobiphenyl (29) for example rotation is anticipated to occur largely about the long axis and the C-3-C-6 axis with the former slightly preferred. On this basis assignments were made for the protonated carbon atoms (TIvalues in seconds). This behavi- our was paralleled in 3-nitrobiphenyl (30) where assignments are possible on (29) (30) chemical shift arguments alone. The self-association of phenol and aniline was also studied shorter Tl values obtained in more concentrated solutions resulting from the slower tumbling rate of the aggregated molecules.A similar study has been reported for ni~otinamide.~~ I3C relaxation for all the protonated carbons in camphor whether part of the rigid bicyclic skeleton or of a rapidly rotating methyl group occursE4 by both dipole-dipole and spin-rotation mechanisms. The former process is the major contributor below 42 “C and the latter process above that temperature molecular motion being essentially isotropic. TI values have been usedE5 to probe the ease of internal rotation of methyl groups in a series of 2,2,3,4,4-pentamethylphosphetans. In some cases pseudoequatorial methyl carbon atoms have relaxation times shorter than the more hindered corresponding pseudoaxial methyl carbon atoms indicating that shorter TI values should not automatically always be taken as implying steric crowding.3C spin-lattice relaxation times have also been used to investigate internal motions in n-alkanes,86 free and complexed polyether antibiotic^,^' gramicidin-S,88 helix and random-coiI states of homo polypeptide^,^' membrane^,^' micel-les,’ phyt01,’~ synthetic polymers,’ and histidine side-chains in protein^.'^ 83 B. Birdsall J. Feeney and P. Partington J.C.S. Perkin 11 1973 2145. 84 J. Grandjean P. Laszlo and R. Price Mol. Phys. 1973 25 695. 85 G. A. Gray and S. E. Cremer J. Magn. Resonance 1973 12 5. 86 N. J. M. Birdsall A. G. Lee Y. K. Levine J. C. Metcalf P. Partington and G. C. K. Roberts J.C.S. Chem. Comm.1973 757; Y. K. Levine J. Magn. Resonance 1973 11 421. 87 M. C. Fedarko J. Magn. Resonance 1973 12 30. 88 A. Allerhand and R. A. Komoroski J. Amer. Chem. Soc. 1973 95 8228. 89 A. Allerhand and E. Oldfield Biochemistry 1973 12. 3428. 90 Y. K. Levine N. J. M. Birdsall A. G. Lee and J. C. Metcalf Biochemistry 1972 11 1416; J. D. Robinson N. J. M. Birdsall A. G. Lee and J. C. Metcalf ibid.. p. 2907. 91 E. Williams B. Sears A. Allerhand and E. H. Cordes J. Amer. Chem. SOC. 1973 95 487 1. 92 R. A. Goodman E. Oldfield and A. Allerhand J. Amer. Chem. Sac. 1973 95 7553. 93 J. Schaefer Macromolecules 1973,6 882; D. D. Davis and W. P. Slichter ibid. p. 728; G. C. Levy J. Amer. Chem. Sac. 1973,956117. 94 D. T. Browne G. L. Kenyon E. L. Packer H.Sternlicht and D. M. Wilson J. Amer. Chem. SOC. 1973 95. 1316. Physical Methods-Part (ii ) Nuclear Magnetic Resonance 41 Other Nuclei.-The observation of deuterium n.m.r. at natural abundance is about 100times more difficult than that of '3Cn.m.r. owing mainly to the lower natural abundance. However the shorter T values are advantageous for spectra accumulation and it has been shown possibleg5 to obtain natural-abundance 2H n.m.r. spectra while employing proton noise decoupling. As expected the spectra show single resonances for the different deuterium atoms. Very little nuclear Overhauser enhancement is obtained consistent with a dominant quadrupolar relaxation although only a small line broadening is obtained. As with '3C n.m.r.deuterium coupling is negligible. Spectra of acetaldehyde benzaldehyde n-butyl alcohol ethanol ethylbenzene and pyridine have been obtained chemical shifts comparing very closely with the corresponding proton chemical shifts. Natural-abundance deuterium n.m,r. spectra can be expected to be particularly advantageous in molecules such as steroids or polycyclic aromatic molecules where the corresponding proton spectra are very complex. ''N chemical shifts for glycine alanine valine ornithine phenylalanine aspartic acid leucine tyrosine and glutamic acid have been measuredg6 using enriched materials. In no case was any significant difference found between the chemical shifts of the fully protonated form and the zwitterion. These results also agreed well in all but two instances with previously published resultsg7 for amino- acid methyl ester hydrochlorides indicating that nitrogen shifts were relatively insensitive to the state of the carboxy-group.29Si spin-lattice relaxation has been examined9* in tetramethylsilane (19 s) diphenylsilane (26 s) dimethylphenylsilane (46 s) hexamethyldisiloxane (39.5 s) tetraethyl orthosilicate (135 s) and a series of polydimethylsiloxanes. For tetra- methylsilane and hexamethyldisiloxane relaxation occurs largely by the spin- rotation mechanism whereas diphenylsilane and tetraethyl orthosilicate relax almost entirely by a dipole-dipole process. Both mechanisms are significant for the other compounds. Dipole-dipole interactions between 29Si nuclei and directly bonded protons are ca.10 times less effective than analogous 13C-'H interactions resulting in the long relaxation times. In contrast to 13Crelaxation intermolecular 29Si-' H dipole-dipole interactions can contribute up to 20 % of the relaxation of a 29Si nucleus. Localized motions along segments of the chain of linear polydimethylsiloxanes result in 29Si relaxation rapidly becoming in- dependent of chain length and viscosity for molecular weights above ca. 500. Relaxation times can be significantly reduced sometimes to as low as 5 s by the presence of paramagnetic species such as dissolved oxygen or by the addition of relaxation reagents such as tris(acety1acetonato)chromium or the corresponding iron complex. Such treatment also suppresses the negative nuclear Overhauser effect (NOE).The observationg9 of a substantial NOE for triphenylsilane and a small NOE for phenylsilane is attributed to a lower barrier to internal rotation 95 J. M. Briggs L. F. Farnell and E. W. Randall J.C.S. Chem. Comm. 1973 70. 96 J. A. Sogn W. A. Gibbons. and E. W. Randall Biochemistry 1973 12 2100. 97 P. S. Pregosin E. W. Randall and A. I. White Chem. Comm. 1971 1602. 98 G. C. Levy J. D. Cargioli P. C. Juliano and T. D. Mitchell J. Amer. Chem. Soc. 1973 95 3445. 99 R. K. Harris and B. J. Kimber J.C.S. Chem. Comm. 1973 255. I. H. Sadler about the silicon-carbon bond in the latter. In view of its resonance frequency lying at the extreme low-frequency end of the 29Si chemical shift range tetraethyl orthosilicate has been proposed'" as a standard to which shifts should be referenced.In practice tetramethylsilane [82.6 p.p.m. relative to (EtO),Si] will be used experimentally as its relaxation time is much shorter. Prior to this year only three studies of 77Se resonance all continuous wave had appeared. The sensitivity of the nucleus in natural abundance is only about three times that of 13C; however the 1H-77Se Overhauser effect will be positive. A recent Fourier-transform study'" of the 77Se resonances of 2-substituted selenophens have been measured and span a range of ca. 100p.p.m. either side of the parent heterocycle. The substituents included the four halogens methyl formyl acetyl carboxy and acetoxy. Two-bond couplings to hydrogen are 4548 Hz three- and four-bond couplings being less than 10 Hz.Two modifications have been described for XL-100spectrometers which allow the observation of nuclei other than those for which an appropriate r.f. unit is provided. The firstlo2 uses the decoupler frequency synthesizer to provide appropriate frequencies for the 'H field-frequency lock signal in order that the field may be varied from the standard value. In this way nuclei such as 23Na 27Al,"V 63Cu and 79Br all of natural abundance greater than SO% may be observed with the r.f. unit provided for I3C observation. A drawback is that decoupling facilities are lost. A second modificati~n'~~ uses the decoupler to provide the observing radio-frequency and allows the observation of about forty different nuclei whose resonant frequencies lie in the range 9.6-32 MHz at 23.5 kG.3 MiscellaneousStudies Two reviews'04 and a book' O5 have appeared concerning lanthanide shift reagents. These continue to be used widely in the simplification of spectra and for the assignment of resonances particularly for mixtures of diastereoisomers.'O6 Chiral shift reagents have been used to estimate the optical purity'07 of 2-methylpiperidine and to distinguish' O8 between enantiomers of sterically hindered asymmetrically ring-substituted cis-p-alkylstyrenes. Optical purity of a-deuteriated primary alcohols was examined' O9 using [Eu(dpm),] (dpm = Me,C.CO-CH-CO-CMe,) after conversion into their camphanic esters. The loo G. C. Levy J. D. Cargioli G. E. Maciel J. J. Natterstad E. B.Whipple and M. Ruta J. Magn. Resonance 1973 11 352. lo' S. Gronowitz I. Johnson and A.-B. Hornfeldt Chem. Scripra 1973 3 94. lo* P. D. Ellis H. C. Walsh and C. S. Peters J. Magn. Resonance 1973 11,426. Io3 C. S. Peters R. Codrington H. C. Walsh and P. D. Ellis J. Magn. Resonance 1973 11 431. '04 B. C. Mayo Chem. SOC.Rev. 1973 2. 49; A. F. Cockerill. G. L. 0. Davies. R. C. Harden and D. M. Rackham Chem. Rev. 1973 73 553. '05 'Nuclear Magnetic Resonance Shift Reagents'. ed. R. E. Sievers Academic Press New York 1973. lo' C. Kruk A. A. M. Roof A. van Wageningen and H. Cerfontain Rec. Trav. chim. 1973 92 1015. lo' R. R. Fraser J. B. Stothers and C. T. Tan J. Magn. Resonance 1973 10 95. lo* C. S. C. Yang and R. S. H. Liu Tetrahedron Letters 1973 481 1.'09 H. Gerlach and B. Zaglalak J.C.S. Chem. Comm. 1973 274. Physical Methods-Part (ii) Nuclear Magnetic Resonance 43 observation' lo that [Eu(fod),] (fod = Me,C.COCHCO-C,F,) causes shifts in the proton spectra of cis-azo-compounds but leaves the trans-isomers virtually unaffected provides a convenient distinguishing test. In aqueous solution praseodymium chloride has been used to simplify proton spectra of mono-saccharides"' and the perchlorate has been used for sequence analysis"2 of hexa- and hepta-peptides at 300 MHz. A method for studying exchanging systems has been described.' ' Increasing the proportion of shift reagent and hence the chemical shift separation causes the signals from the N-methyl protons of trimethyl carbamate Me,NCO,Me to change from a single resonance through coalescence to separate resonances at constant temperature.From the variation in lineshape and separation with mole fraction of shift reagent a value for the free energy of activation (15.5 kcal mol-') for rotation about the N-C(0) bond was obtained in good agreement with the literature value. Lanthanide reagents are also used to infer detailed substrate geometries usually on the assumption that the induced shift A6 is given by an expression of the type A6 = K(3 cos2 4 -l)r-3. This expression should however be used with caution since it is only valid for axially symmetric 1 1 shift reagent (L)-substrate (S) complexes where the shift is due exclusively to a pseudocontact (dipolar) interaction. However the magnetic axis is not necessarily along the metal-substrate axis1l4 and not all 1 1 complexes are axially symmetric.''5 Effective axial symmetry may sometimes result from rapid rotation about the metal-substrate axis in which case it is necessary to average calculated shifts over all molecular conformations before comparisons are made with observed values.''6 The use of static models can lead to errors in spectral assignments.1 :2 Complexes which have no axial symmetry are frequently present,' "*''* and even higher complexes have been observed.' Appreciable contact contri- butions to shifts have been demonstrated''9 in some cases particularly for europium reagents and also in 13C studies. It is therefore surprising that good agreement with the above expression is obtained so frequently.This is however usually achieved using the position of the lanthanide atom as a variable parameter that is adjusted to give the best possible fit of the data which may cover up 'lo S. R. Wilson and R. B. Turner J.C.S. Chem. Comm. 1973 557. I I S. J. Angyal Carbohydrate Res. 1973 26 27 1. M. Anteunis and J. Gelan J. Amer. Chem. SOC. 1973.95 6502. I I3 S. R. Tanny M. Pickering and I. S. Springer J. Amer. Chem. SOC.,1973 95. 6227. R. E. Cramer and R. Dubois J. Amer. Chem. Soc. 1973 95 3801. 'Is J. J. Uebel and R. M. Wing J. Amrr. Chem. Soc. 1972 94 8910. I l6 R. M. Wing J. J. Uebel and K. K. Anderson J. Amer. Chem. SOC.,1973 95 6046; I. M. Armitage L. D. Hall A. G. Marshal and L. G. Werbelow ibid. p. 1437. D. F. Evans and M.Wyatt J.C.S. Chem. Comm. 1973 399; J. W. ApSimon. H. Beierbeck and A. Fruchier J. Amer. Chem. SOC., 1973 95 939. I" R. Porter T. J. Marks and D. F. Shriver J. Amer. Chem. SOC. 1973 95 3548. I19 J. W. ApSimon H. Beierbeck and J. K. Saunders Canad. J. Chem. 1973 51 3874; J. Reuben J. Magn. Resonance 1973 11 103; K. Tori Y. Yoshima M. Kainosho and K. Ajisaka Tetrahedron Letters 1973 1573; 0.A. Gansow P. A. Loeffler R. E. Davis M. R. Willcott and R. E. Lenskinski J. Amer. Chem. SOC.,1973.95 3389 3390; G. E. Hawkes C. Marzin S. R. Johns and J. D. Roberts ibid. p. 1661. 44 I. H. Sadler deficiencies in the method particularly as the lanthanide-complexing site distance is longer than X-ray studies of crystalline complexes suggest. For a further discussion of the basis of the use of these reagents the reader is referred elsewhere.'2o It is noteworthy that positions of equilibria between substrate conformations have been altered"' by the presence of lanthanide shift reagents.As an alternative to the use of the above equation for structural analysis an approach has been proposed'22 based on the r-6 dependence of relative line- widths obtainable by the use of relaxation reagents such as [Gd(fod),] which rests on a firmer theoretical basis.123 The suggested procedure involves the addition of a shift reagent e.g.[Eu(fod),] to induce large shifts between resonances followed by a relaxation reagent [Gd(fod),] to broaden the lines. In this way broadening of up to several hundred Hz may be obtained which would be impossible without the expanded scale provided by the shift reagent.Minor linewidth corrections may be necessary to allow for slight broadening due to the shift reagent and to incompletely collapsed multiplets. Linewidth ratios are expected to be much less sensitive to the presence of multiple species than shift ratios and although shifts may be measured more accurately than linewidths the r-6 dependence allows significant error in the measured linewidth without sacrificing acceptable accuracy in the determination of the relative values for r. A detailed account of the preparation and applications of the macrocyclic shift reagents has been given.' 24 Their advantages and disadvantages together with the way in which they complement the lanthanide reagents are discussed.The observation of intramolecular nuclear Overhauser effects continues to be used in stereochemical problems. l2 It is particularly important to ensure that enhancements observed are valid where the different resonances to be saturated are no more than a few Hz apart.126 In strongly coupled spin systems it has been shownI2' that selective NOE experiments in which a single transition is saturated provide a better method for investigating relaxation mechanisms and molecular structure than the more general experiment where an entire multiplet is saturated. Analyses for ABX and A,X spin systems are given and compared with experiment. The intermolecular nuclear Overhauser effect (INOE) is rather less frequently encountered.In this experiment the solvent spins are saturated and the solute spins observed. Equations have been devised and solutions obtained for the intensity increases expected for the AX spin system with intramolecular dipolar relaxation and different intermolecular dipolar interactions at the sites of the A and X spins.l2* Using tetramethylsilane as solvent intensity increases of 37 and 16% are obtained for A and X spins in the AX system of 1,1,2-trichloroethane. 120 M. I. Foreman in 'Nuclear Magnetic Resonance' ed. R. K. Harris (Specialist Periodical Reports) The Chemical Society London 1973 Vol. 2 p. 378; 1974 Vol. 3 Ch. 10. l2 T. Sat0 and K. Goto J.C.S. Chem. Comm. 1973 494. 122 G. N. La Mar and J. W. Faller J. Amer. Chem. SOC.,1973 95 3817. 1. Solomon Phys.Rev. 1955 99 559. 124 J. E. Maskasky and M. E. Kenney J. Amer. Chem. SOC.,1973 95 1443. 125 R. A. Bell and J. K. Saunders Topics Stereochem. 1973 7 I. 12' A. J. Bellamy and W. Crilly J.C.S. Perkin If 1973 122. 12' N. R. Krishna P. P. Yang and S. L. Gordon J. Chem. Phys. 1973 58 2906. lz8 N. R. Krishna and S. L. Gordon J. Chem. Phys. 1973 58. 5687. Physical Me (hods- Part (ii ) Nuclear Magnetic Resonance 45 Comparison with theory allowed estimation of the fraction of internal relaxation. This approach offers some advantage over the intramolecular experiment in that all the lines of the coupled spin system can be observed and complex lineshape variations due to inhomogeneity effects are absent. It is likely that such studies applied to biological systems would lead to information about intermolecular interactions and molecular conformation of the molecules.A new technique for the observation of high-resolution n.m.r. spectra of rapidly flowing chemical systems has been developed'29 in which the problems posed by the relatively long relaxation times ofnuclei are overcome. It has been used to examine unstable Meisenheimer complexes formed immediately after mixing of reagents and should find application in the observation of other short- I ived intermediates. The following aspects and applications of n.m.r. have been reviewed the solvent dependence of coupling constants :'30 dynamic magnetic resonance :' ' the determination of mechanistic information from lineshapes for intramolecular exchange ;'32 n.m.r.in carbohydrate chemistry ;* 33 carbocations ;'34 phosphorus ~tudies:'~' and rate processes in boron compounds.'36 A book devoted to techniques,' 37 one on general aspects of n.m.r. spectr~scopy,'~~ and one con- cerned with the use of n.m.r. for quantitative analysis'39 have appeared. 129 C. A. Fyfe M. Cocivera and S. W. H. Damji J.C.S. Chem. Comm. 1973 743. I3O M. Barfield and M. D. Johnston Chem. Rev. 1973 73 53. 13' N.M. Sergeev Russ. Chem. Rev. 1973 42 339. 13' J. P. Jesson and P. Meakin Accounts Chem. Res. 1973 6 269. 133 G. Kotowycz and R. U. Lemieux Chem. Rev. 1973 73,669. G. A. Olah Angew. Chem. Internat. Edn. 1973 12 173. 135 G. Mavel Ann. Reports N.M.R. Spectroscopy 1973 5B 1. '36 H. Beall and C. H. Bushweller Chem.Rev. 1973 73,465. 13' W. McFarlane and R. F. M. White 'Techniques of High Resolution Nuclear Magnetic Resonance Spectroscopy' Butterworths London 1973. 13' J. W. Akitt 'N.M.R. and Chemistry' Chapman and Hall London 1973. I39 F. Kasler. 'Quantitative Analysis by N.M.R. Spectroscopy' Academic Press London 1973.