2 Physical Methods and Techniques Part (iii) N.M.R. Spectroscopy By J.K. M. SANDERS University Chemical Laboratory University of Cambridge Lensfield Road Cambridge CB2 1EW 1 Introduction Last year's Report' was both long and concentrated describing major advances published during 1976-78 in a variety of areas of n.m.r. spectroscopy. This Report on the 1979 literature retains the same format as before but concentrates on the development of those ideas and techniques as tools for solving chemical problems. No attempt has been made to restrict examples to specifically organic applications each section contains illustrations from physical inorganic or biological appli- cations chosen because they are instructive and interesting. 2 Spin-Lattice Relaxation For organic molecules in solution the dominant relaxation mechanism for both 'H and 13Cis dipole-dipole relaxation induced by nearby protons.The relaxation time is given by equation (l),where r is the distance from neighbouring protons and 7,is a correlation time for molecular motion. Thus Tl gives information about mobility or about inter-nuclear distances and both effects have been put to good effect this year in assigning spectra. For example the methine protons of frangulanin (1)are expected to experience similar tumbling rates and so they can be distinguished on the basis of their different distances from neighbouring protons; the methyls which all have the same geometry (and hence the same Cr-6) can be assigned through their differing mobilities.2 Similarly C-18 in asperdiol(2) is the most mobile methylene carbon and has a significantly longer Tl than the ring methylene~.~ As expected from equation (l),the ring methine carbons of (2) relax at half the rate of the methylenes.More quantitatively equation (1)can be used to extract detailed geometrical information particularly when substitution by or selective pulse experiments6 allow the contribution of a particular proton to the relaxation of J. K.M. Sanders Ann. Reports (B) 1978,75 3. * E. Haslinger and W. Robien Monatsh. 1979,110 1011. G. E. Martin J. A. Matson and A. J. Weinheimer Tetrahedron Letters 1979 2195. L. D. Hall K. F. Wong W. E. Hull and J. D. Stevens J.C.S. Chem. Comm. 1979,953. L. M.Jackman and J. C. Trewella J.Amer. Chem. SOC.,1979,101,6428. L. D. Hall K. F. Wong and H. D. W. Hill J.C.S. Chem. Comm. 1979,951. Physical Methods and Techniques -Part (iii) N.M.R. Spectroscopy :H, Ph OR RO @OROR UPh 0 (4) R=COCD3 (3) another nucleus to be determined. In the case of (3) 13Cand "N relaxation rates enabled precise measurement of the hydrogen (deuterium) bond geometry and demonstrated its asymmetry rather c~nvincingly,~ whilst substitution by deuterium at C-5 in (4) gave values for all the inter-proton distances that were in complete agreement with those from neutron-diffraction mea~urements.~All the above examples report suitable control experiments which ensure that indeed dipole- dipole interactions are responsible for the observed effects and all are for molecules that are essentially rigid.More difficult is the determination of mobility given an essentially known geometry. For example interpretation of 13Crelaxation data is crucially dependent on choosing an appropriate value for the C-H bond length in both small organic molecules7 and proteins.* Even then detailed interpretation of relaxation rates in terms of molecular motion is fraught with problem^,^ and should be attempted only by the brave. Paramagnetics are very efficient inducers of relaxation and therefore should be removed from solutions where Tl measurements are contemplated. The most common approach for aqueous solutions is to add EDTA but when this was tried with a solution containing Fe3' and phosphate the Tl of 31P actually shortened because of the formation of ternary EDTA-Fe-Pi complexes.lo An alternative approach that is sometimes successful is extraction with 8-hydroxyquinoline," but the important point is that care must be exercised.Of course paramagnetic effects on Tl may be exploited and one fine example of this -inadvertently omitted from last year's Report [Vol. 75 p. 31 -is the determination of the bound conformation of propionyl-CoA on the metal-containing enzyme transcarboxylase.'2 'R. K. Harris and R. H. Newman Mol. Phys. 1979,38 1315. K. Dill and A. Allerhand J. Amer. Chem. SOC.,1979,101,4376. P.Stilbs and M. E. Moseley J. Magn. Reson. 1979,33,209. lo G. A. Elgavish and J. Granot J. Magn. Reson. 1979,36 147. *' D. H.Live H.R. Wyssbrod A. J. Fischman W. C. Agosta C. H. Bradley and D. Cowburn J. Amer. Chem. SOC.,1979,101,474. C. H.Fung R. J. Feldmann and A. S. Mildvan Biochemistry 1976,15,75. J. K. M. Sanders Experimental Methods.-The search continues in many laboratories for an ideal way of measuring TI values in any given sample. One problem has always been the generation of a good 180” pulse to invert all the nuclear spins but Levitt and FreernanI3 have described T. composite ‘sandwich’ pulse which self-compensates for any errors and which should therefore make the null-point method more reliable. The variable-flip-angle approach to measurement of Tl has been extended into two variants one employing fixed delays between the perturbing and monitoring pulses’4 and the other allowing simultaneous nulling of an interfering HOD ~igna1.l~ A completely different way of removing unwanted resonances is to use spin echoes in TI sequences.16*17 Spin echoes are discussed in detail in Section 5 of this Report.Spin-lattice relaxation in the rotating frame a hitherto rather obscure subject has proved to have a simple application in the study of rapid internal motions and it may well become popular.’* The Nuclear Overhauser Effect.-This has experienced a revival in the guise of n.0.e. difference spectroscopy and is now proving to be perhaps the most subtle and powerful manifestation of spin-lattice relaxation phenomena in organic chemistry. The ‘traditional’ n.0.e. experiment is very powerful and it has been exploited skilfully to determine the structure conformation and absolute configuration of the polypeptide antibiotic ristocetin (mol.wt. 3000),19but it is limited in scope because the signal to be observed must be resolved and because the minimum credible effect using integration is ca. 5%. In n.0.e. difference spectroscopy a control spectrum without n.0.e. is subtracted from the spectrum with n.O.e. so that only spectral changes appear. The signal of interest need no longer be resolved in the normal spectrum and the lower limit of observable effect is determined only by instrument stability. Thus all the protons of the steroids 1-dehydrotestosterone (5) and llp-hydroxyprogesterone (6) have been resolved and assigned using n.O.e.3 in the 0.5-5% range (to ‘see’ between and across rings) in combination with a difference decoupling technique (see Section 9.’’ The same combination of techniques has been used to resolve and assign the protons of tyrocidine A a decapeptide.” l3 M.H. Levitt and R. Freeman J. Magn. Reson. 1979 33,473. l4 R. K. Gupta G. H. Weiss J. A. Ferretti and E. D. Becker J. Magn. Reson. 1979 35 301. R. K. Gupta and P. Gupta J. Magn. Reson. 1979 34 657. l6 J. Hochmann R. C. Rosanske and G. C. Levy J. Magn. Reson. 1979,33,275. D. L. Rabenstein T. Nakashima and G. Bigam J. Magn. Reson. 1979 34,669. l8 D. M. Doddrell M. R. Bendall P. F. Barron pd D. T. Begg J.C.S. Chern. Cornm. 1979,77. l9 D. H. Williams V. Rajananda G. Bojesen and M. P. Williamson J.C.S. Chem. Cornrn.,1979,906 and refs. therein. *’ L.D. Hall and J. K. M. Sanders J.C.S. Chern. Cornrn. 1980,368. *’ M. Kuo and W. A. Gibbons J. Biol. Chem. 1979,254,6278. Physical Methods and Techniques -Part (iii) N.M.R. Spectroscopy 21 Nuclear Overhauser effect difference spectroscopy has also been used22 to assign non-equivalent protons in the primary amide group-CONH2 to study the con- formation of trimethoprim bound to dihydrofolate red~ctase,~~ and to characterize the interaction of ADP with creatine kina~e.~~ Most n.0.e. experiments are carried out under steady-state conditions but in large proteins where spin diffusion eventually dilutes the useful r-6 dependence of the n.O.e. it is helpful to acquire ‘truncated’ n.O.e.’s just after the saturation has begun to build up.25 Alternatively ‘transient’ n.O.e.’s arising from the selective inversion of a single resonance can be useful in geometry measurements9 or spectral assign- ments2’ for organic molecules.Since the build-up of transient and truncated effects depends on the relaxation rates of both nuclei involved the rates of build-up can be used to measure TI values of signals that are otherwise inaccessible because they have large negative n.0.e.’s.26 Finally we come to Chemically Induced n.O.e. or deceptive CIDNP,27an elegant technique for generating enhancements in a small part of the molecule of interest that molecule must (for the present) contain an exchangeable OH or NH proton. To the solution is added an aromatic acceptor and a suitable tertiary amine. Irradiation with light in the probe leads to electron transfer [equation (2)] and the resulting amine radical cation exchanges its now acidic protons (H) with the molecule under study hv Acceptor + R2NCH2Me AT (R2NCHzMe)t (2) The exchanged protons carry into their ‘host’ CIDNP which then imparts negative enhancements to any nucleus that is J-coupled to them and a positive enhancement to protons that are dipolar-coupled to them.This trick has promise both as an assignment tool and as a sensitivity-enhancement method for nuclei such as 15N. 3 Other Nuclei Deuterium,’H.-Direct observation of 2H spectra can be a very powerful way of studying reaction mechanisms and one of this year’s finest examples is the deter- mination of the mechanism and of the conformation of the transition state of the vitamin-D3-previtamin-D3equilibrium.28 Direct comparison of proton deuterium and tritium chemical shifts in intramolecular hydrogen bonds gives a detailed picture of the shape of the potential well easily distinguishing a symmetrical hydrogen bond from two unsymmetrical rapidly interconverting bonds.29 This approach nicely complements Jackman’s,’ which was described in Section 2.Deuterium is an almost ideal nucleus owing to the ease of incorporation into biological molecules the 22 A. G. Redfield and S. Waelder J. Amer. Chem. SOC.,1979 101,6151. 23 P. J. Cayley J. P. Albrand J. Feeney G. C. K. Roberts E. A. Piper and A. S. V. Burgen Biochemistry 1979,18,3886. 24 M. Vasik K. Nagayama K. Wuthrich M. L. Mertens and J.H. R. Kagi Biochemistry,1979,18,5050. 25 G. Wagner and K. Wiithrich J. Magn. Reson. 1979 33 675; A. A. Bothner-By and J. H. Noggle J. Amer. Chem. SOC. 1979,101,5152. 26 S. J. Opella R. A. Friedman M. C. Jarema and P. Lu J. Magn. Reson. 1979 36 81. 27 J. Bargon and G. P. Gardini J. Amer. Chem. Soc. 1979,101,7732. M. Sheves E. Berman Y. Mazur and Z. V. 1. Zaretskii J. Amer. Chem. SOC. 1979,101 1882. 29 L. J. Altman D. Laungani G. Gunnarsson H. Wennerstrom and S. Forsen J. Amer. Chem. SOC. 1978 100,8264. J. K. M. Sanders simplicity of the ensuing spectra and the straightforward interpretation of its relaxation parameters. For example the histidine of lysozyme has been labelled with deuterium by exchange and the 2H nucleus then used to study aggregation of the enzyme;3o deuterium has been incorporated by chemical modification into proteins and glycopro teins for n .m.r. studies. 3' More important for the organic chemist are the isotope effects of deuterium on 'H and 13Cchemical shifts (relaxation effects having been discussed in Section 2). Anet and Dekmezian have very carefully distinguished between equilibrium isotope effects which are due to a change in the population of two or more equilibrating species (usually conformations) and intrinsic effects that are observed in single species.32 As an example of the latter they describe a shift of 0.0137 p.p.m. in H* of the half-cage (7) when X is changed from H to D. Isotope effects on 13Cshifts of carbohydrates in H20 us. D20provide an elegant assignment method,33 and small deuterium effects are strongly amplified at pH values near the pK, allowing extraction of all the microscopic pK values.34 A similar amplification effect has long been known with lanthanide shift reagents3' Nitrogen-15.-Last year's Report stated that this nucleus was probably after 'H and 13 C the most interesting nucleus for many chemists and biochemists its main drawback being low sensitivity.Section 5 describes several methods which may increase its effective sensitivity sufficiently to make "N a reasonably accessible nucleus; chemically induced n.0.e. may also help in this regard. Meanwhile it is possible (with patience and adequate access to spectrometers) to acquire natural- abundance spectra of e.g. reserpine and related alkaloids and to draw up all the expected correlations of chemical shift with structure and c~nformation.~~ Most "N spectra are taken with molecules that have been highly enriched synthetically [see for example (3)] or biosynthetically.Thus measurements of Tl and of the n.0.e. of labelled oxytocin yield information on its conformation," and the mechanism of action of a serine protease has been thoroughly investigated using "N-labelled histidine in the 'catalytic triad'.37 Perhaps most remarkable of all are the "N spectra of biosynthetically labelled bacterial cell walls which supply extra- ordinarily detailed pictures of the chemical and physical similarities and differences 30 J. B. Wooten and J. S. Cohen Biochemistry 1979,18,4188. 31 M.A. Bernstein L. D. Hall and W. E. Hull J. Amer. Chem. SOC. 1979,101 2744. 32 F. A. L.Anet and A. H. Dekmezian J. Amer. Chem. SOC. 1979,101,5449. 33 P. E. Pfeffer K. M. Valentine and F. W. Parrish J. Amer. Chem. SOC. 1979,101 1265,7438. 34 J. J. Led and S. B. Petersen J. Magn. Reson. 1979,33 603. 3s J. K.M.Sanders and D. H. Williams J.C.S. Chem. Comm. 1972,436. 36 S. N. Y.Fanso-Free G. T. Furst P. R. Srinivasan R. L. Lichter R. B. Nelson J. A. Panetta and G. W. Gribble J. Amer. Chem. SOC. 1979,101 1549. 37 W. W. Bachovchin and J. D. Roberts J. Amer. Chem. SOC. 1978,100 8041. Physical Methods and Techniques -Part (iii) N.M.R. Spectroscopy 23 between one species and another.38 It is worth noting that I5Nseems to be far more sensitive to the presence of paramagnetics than are 'H and l3C.'' The great potential in biosynthetic investigations of detecting 15N through its coupling with the common nuclei was discussed in last year's Report.It should be emphasized that the couplings looked for in such experiments must be reliably known or estimated from model experiments; the dangers are stressed in a study of labelled adenine which found that some 'JNcare actually very small indeed and may not be see also Section 5. 'Other' Other Nuclei.-Phosphorus-31 continues to be a powerful probe of intact biological entities. It has been used to monitor the internal pH of the chromaffin granules of adrenal gland and to determine the co-ordination state of the ATP within those granules,4o and also to monitor release of ATP in blood platelet^.^' More exciting possibilities include the simultaneous acquisition of 31P and I3C spectra from a single sample using a probe that is quadruply tuned to both these frequencies and to 'H (for decoupling) and to 2H (lock),42 and obtaining n.m.r.spectra from parts of bodies using probes which simply lie on the skin.43 Isotope effects of l80on 13C and 31Pchemical shifts are proving to be very useful mechanistic tools the former has only been used in organometallic chemistry44 as yet but it should be generally applicable. The latter has rather quickly become an established method in enzyme chemistry.'*45 This year such isotope effects have been used to show that purine nucleotide phosphorylase catalyses exchange of phosphate on a-D-ribose 1-phosphate with C-0 cleavage (rather than the expect- ed P-0 cleavage)46 and to show that in two amino-acyl tRNA synthetases the nucleotidyl-transfer step takes place with inversion at phosph~rus.~~ Oxygen-17 has much organic chemical potential this year's most intriguing example being the direct observation of H30+(quartet JOH= 106 Hz) in cold wet; acidic carbon tetra~hloride.~~ Finally we should note that it is relatively easy to observe 113Cd-113 Cd coupling in suitably labelled samples of the protein metallo- t hionein.50 4 Paramagnetics Shift Reagents and Metal Complexes.-It is a measure of the maturity of the field that there is relatively little of novelty to report on shift reagents. Reuben" has developed a system for the n.m.r.resolution of chiral molecules in aqueous solutions 38 A. Lapidot and C. S. Irving Biochemistry 1979,18,704. 39 M. Kainosho J. Amer. Chem. Soc. 1979,101 1031. 40 H. B.Pollard H. Shindo C. E. Creutz C. J. Pazoles and J. S. Cohen JBiol. Chem. 1979,254 1170. 4.' K. Ugurbil H. Holmsen and R. G. Shulman Roc. Natl. Acad. Sci. USA 1979,76,2227. 42 P.Styles C. Grathwohl and F. F. Brown J. Magn. Reson. 1979 35 329. 43 J. J. H. Ackerman T. H. Grove G. G. Wong D. G. Gadian and G. K. Radda Nature 1980,283,167. 44 D.J. Darensbourg and B. J. Baldwin J. Amer. Chem. SOC., 1979,101,6447. 4s M. C.Summers,Ann. Reports (B),1978,75 391. 46 F.Jordan J. A. Patrick and S. Salamone Jr. J. Biol. Chem. 1979 254 2384. 47 S.P. Langdon and G. Lowe Nature 1979,281,320.48 See for example M. Katoh T. Sugawara Y. Kawada and H. Iwamura Bull. Chem. SOC. Japan 1979,52 3475and refs. therein. 49 G. D. Mateescu and G. M. Benedikt J. Amer. Chem. SOC.,1979,101,3959. J. D. Otvos and I. M. Armitage J. Amer. Chem. SOC.,1979,101,7734. '' J. Reuben J.C.S. Chem. Comm. 1979,68. 24 J. K.M. Sanders a-hydroxy-carboxylic acids form 2 :1complexes with lanthanide ions so the addi- tion of a 1:1 mixture of e.g. D-mandelate [PhCH(OH)COJ and lanthanide to another optically impure a-hydroxy-acid will result in the formation of dia- stereoisomeric complexes with different stabilities an& geometries. Alternatively enantiotopic groups in a prochiral molecule (the methyls of a-hydroxybutyrate or the methylene protons of glycollate HOCH2C02-) can be rendered non-equivalent.’I The fitting of shift data to geometrical models is perhaps becoming less risky as both the models and their statistical evaluation improve.Thus for adamantan-2-one the goodness of fit improves dramatically as the model is refined from a single metal-binding site that is collinear with the carbonyl group to a chemically sensible model with two sites along the carbonyl lone pairs and improves again in a four-site model that minimizes steric interactions between reagent and s~bstrate.’~ The standard Hamilton R test for the statistical evaluation of such geometrical models has been shown to be quite inappropriate and simple alternatives have been propo~ed.’~ Indeed the apparent failure of shift reagents to distinguish diastereo- isomers which was discussed in last year’s Report (p.14 ref. 125) turns out to be only a failure of the statistical method -the use of valid statistics shows that the diastereoisomers were distinguishable after all. It is a matter of taste whether one wants chemical problem-solving to be reduced to such statistics particularly when ‘significance testing [can] only test whether one mathematical model of a real situation fits the data better than another such model. . .. [It] can never determine whether such a model is corre~t.”~ The temperature dependence of €anthanide-induced shifts has given rise to some controversy (see last year’s Report p. 13) but McGar~ey’~ has now confirmed theoretically that the dependence should indeed be approximately and not exactly TP2.However in cases where reagent-substrate binding is bi- or ter-dentate the exchange rate becomes so low that it is directly observable in the form of line broadening and separate free/bound spectra.In such cases it is possible not only to measure temperature dependences of the shifts directly but also to determine whether the exchange mechanism is associative or dissociative.” Biologically oriented applications of paramagnetic effects include the deter- mination of conformation of enzyme-bound propionyl-CoA already referred to12 and the assignment of the five indole N-H resonances of hen egg-white ly~ozyme.’~ N.m.r. spectroscopy of haems has now developed to the point where it is detecting localized conformational flexibility in cytochromes using a methyl group attached to the haem as reporter,” and is giving exquisite detail of electronic structure co- ordination chemistry and geometry of complexes in model porphyrins and real proteins.’* 52 R. J. Abraham D. J. Chadwick L. Griffiths and F. Sancasson Tetrahedron Letters 1979,4691. s3 M. F. Richardson,S. M. Rothstein and W.-K.Li J. Magn. Reson. 1979,36 69. s4 B. R. McGarvey J. Magn. Reson. 1979 33,445. ” L. F. Lindoy and H. W. Louie J. Amer. Chem. SOC.,1979 101,841. 56 R. E. Lenkinski,J. L. Dallas and J. D. Glickson J. Amer. Chem. Soc. 1979,101 3071. ” P. D. Burns and G. N. LaMar J. Amer. Chem. SOC.,1979,101. 5844. 58 D. L. Budd G. N. LaMar K. C. Langry K. M. Smith and R. Nayyir-Mazhir J.Amer. Chem. SOC.,1979 101 6091 and references therein. Physical Methods and Techniques -Part (iii) N.M. R. Spectroscopy 25 C1DNP.-Three contributions have captured this Reporter’s imagination. Chem- ically Induced n.0.e. has been described in detail in Section 2 of this report. The detection of CIDNP with a time resolution of microseconds after a laser flash was achieved” by first saturating the equilibrium magnetization by a series of pulses then a laser flash of several ns duration is followed by a variable delay (to allow evolution of CIDNP) and an r.f. pulse to give the n.m.r. signal. Only pure CIDNP responses are observed and they can be monitored from 1gs after the flash. The potential for mechanistic chemistry is clearly very significant.Finally a series of papers from Kaptein’s laboratory developing laser photo-CIDNP as a way of simplifying protein spectra (see last year’s Report p. 15) has culminated in selective detection of the single tryptophan in bovine pancreatic phospholipase A2and the use of that tryptophan to characterize a pH-dependent conformational change.60 5 New Experimental Methods Last year’s Report highlighted two-dimensional (2-D) n.m.r. and high-resolution solid-state n.rn.r. as fundamental developments that would soon have a great impact on many areas of chemistry and indeed the application of these techniques is already moving from physics into organic chemistry Thus all the geminal and vicinal proton coupling constants in a steroid have been resolved by 2-D J spectroscopy; the sensitivity of ”N n.m.r.has been increased in some cases by one hundred fold using population-transfer tricks; solid-state chemical reactions at catalyst surfaces have been observed. This is only the beginning of a new phase in n.m.r. spectroscopy which will develop rapidly in the next few years as the implications are digested by the organic chemist. Two-dimensional (2-D) and Spin-echo Techniques.-As explained more fully in last year’s Report 2-D spectra display the response of the nuclear spins to two different sets of frequencies in contrast to l-D(traditional) spectra which are a function of only one frequency. As a bonus most 2-D techniques are based on spin-echo methods which means that all effects due to inhomogeneity of the magnetic field are removed and thus that linewidths are determined only by the spin-spin relaxation time T2.2-0 J Spectroscopy. A simple pulse sequence i.e. 9OO-delay-180°-delay-collect FID gives spin-echo spectra in which the signal phases depend on the size and number of J couplings. Repetition of the sequence with a set of different delays gives after suitable manipulation of the data,’ spectra in which chemical shifts and coupling constants are separated in different dimensions (for weakly coupled systems). This leads of course to the resolution of signals that are otherwise hopelessly overlapping. The most spectacular use of proton-proton 2-D J spectro-scopy known to the Reporter (in January 1980)is shown in Figure 1 the spectrum allowed the measurement of all the chemical shifts and the geminal and several long-range couplings of the steroid l-dehydrotestosterone (5),61 but others will 59 G.L. Closs and R. J. Miller J. Amer. Chem. SOC.,1979,101 1639. 6o E.H.J. M. Jansen G. J. M. van Scharrenburg A. F. Slotbloom G. H. de Haas and R. Kaptein J. Amer. Chem. SOC.,1979,101,7397. 61 L. D. Hall J. K. M. Sanders and S. Sukumar J.C.S. Chem. Comm. 1980,366. J. K.M. Sanders / I / Figure 1 Partial spectra (400MHz) of 1-dehydrotestosterone (5) (data are from unpublished spectra and from ref. 61). (a) Normal one-dimensional spectrum. (6) Two-dimensional J spectrum of the same region. (c) ‘Proton-decoupled’ proton spectrum obtained by summing all the two-dimensional cross-sections onto the chemical-shift axis.(d)Some representative cross-sections showing the multiplet structure of individual proton resonances surely follow. In addition as heteronuclear couplings behave like chemical shifts in the spin-echo experiment it is possible to remove (without a decoupler) either all the homo- or all the hetero-nuclear couplings from a proton at will merely by changing the angle of projection of the proton in the 2-D spectrum.62 It is possible also to achieve 13C-lH 2-D J spectros~opy,~~ in which all the JCHappear in one dimension and the 13Cchemical shifts in the other? There are several technical problems associated with 2-D J spectroscopy some of which have yet to be solved satisfactorily. The most important (apart from the difficulty of second-order spectra) is that the spectra experience a ‘phase twist’ around each peak.In early work this was dealt with by plotting absolute value (AV) mode? spectra but these have the disadvantage of inducing line broadening and generating long ‘tails’ around intense peaks. One way round the problem is to use a * Other nuclear pairs are possible of course. t AV = (real2+imaginary*)‘ 62 L. D. Hall and S. Sukumar J. Arner. Chem. Soc. 1979,101,3120; J. R. Everett D. W. Hughes A. D. Physical Methods and Techniques -Part (iii) N.M.R. Spectroscopy Lorentzian to Gaussian transformation for the AV another is to plot the power spectrum (AV’) which cuts off tails but also distorts multiplets so that a 1 :2 1 triplet becomes 1:4 :l.65Better for the ‘H-’H experiment one can phase each signal of interest although some distortions and for the I3C-lH experiment several other possibilities are also available.63 It is possible to remove interfering solvent signals that have long Ti's by a standard 180” pulse followed by a delay before the spin-echo sequence,67 or to decouple during a 2-D J acquisition; the latter turns out to be an experiment fraught with complications and it looks unlikely that it will be of general applicability in organic chemistry.68 Finally 2-D J spectra contain extra spinning side-bands but these are (fortunately) at predictable places.69 2-0 Correlation Spectroscopy.In this form of 2-D spectroscopy both frequency axes contain chemical shifts in the heteronuclear variants suitable pulsing of both the proton and another nucleus yields 2-D spectra in which just one signal appears in the map at the co-ordinate for chemical shifts corresponding to the heteronucleus and its coupled proton (Figure 2a).The advantages of this technique are that the 1 2 3 4 (a) I 23 4 ‘H d ’H Figure 2 (a) Schematic two-dimensional correlation spectrum showing a carbon (8 30) coupled to a proton at S 1.5 and a carbon (6 50) coupled to a proton at S 3.5. (b) Schematic two-dimensional exchange-correlation spectrum showing a proton (X)at S 3 exchang-ing with a proton (Y)at S 2 the latter also exchanging with a proton (2)at S 1.5. heteronucleus is observed with a sensitivity approaching that of protons (by virtue of a cross-polarization effect) that the shifts of coupled nuclei are automatically correlated abolishing many assignment problems and that only the nuclei involved in such coupling will appear in the spectrum.Thus in the 31P-1H correlation experiment only ‘local’ protons that are coupled to 31Pare seen provided that the proton spectrum is fir~t-order.~’ Examples of 13C-lH correlation experiments were given in last year’s Report (refs. 174-176) once the technique becomes established it could eventually be a routine way of acquiring spectra of rare spins. The basis of these experiments is well described by Bodenhausen and Freeman with an intuitive approach that employs a relatively simple physical picture of population levels.” 64 J. C. Lindon and A. G. Ferridge J. Magn. Reson. 1979,36,277. 65 L.D. Hall S. Sukumar and G. R. Sullivan J.C.S. Chem. Comm. 1979,292. 66 M. H. Levitt and R. Freeman J. Magn. Reson. 1979,34,675; L. D. Hall and S. Sukumar ibid. 1980 38,555. 67 L. D. Hall and S. Sukumar Carbohydrate Res. 1979 74 C1. 68 K. Nagayama J. Chem. Phys. 1979,71,4404. 69 G. Bodenhausen S. P. Kempsell R. Freeman and H. D. W. Hill J. Magn. Reson. 1979,35,337. 70 P. H. Bolton and G. Bodenhausen J. Amer. Chem. SOC.,1979,101 1080. 71 G. Bodenhausen and R. Freeman J. Magn. Reson. 1979,36,221. 28 J. K.M. Sanders Homonuclear 2-D correlation methods allow the unambiguous elucidation of chemical exchange pathways72 or assignments of coupling constants.73 As with almost all the 2-D experiments mentioned above and several 1-D experiments that are described at the end of this section these techniques were developed by Ernst in a series of definitive and brilliant papers.In the chemical-exchange experiment7* the slowly exchanging peaks are ‘labelled’ by a 90”pulse allowed to evolve and then are monitored by further pulses. The resulting spectrum (Figure 2b) has two identical axes for proton chemical shift with the main peaks appearing on a diagonal. Peaks representing exchanged magnetization are off -diagonal and instantly show (in this example) that X exchanges with Y and Y with Z. The technique seems to this Reporter to have an appealing simplicity which will make it most valuable. In its present form at least the closely related coupling-correlation-pulse sequence73 gives by contrast spectra that are highly complex and low in effective sensitivity thus limiting its general applicability.Other Spin-echo Techniques. Even without resort to 2-D methods the spin echo can be very useful as it removes effects arising from inhomogeneity of the magnetic field and allows the separation of resonances that have the same chemical shift but different multiplicity or different values of T,. A simple description of these ideas has been published by Raben~tein,~~ and examples of their use include the elimina- tion of both broad17 and sharp7’ interfering resonances (see also last year’s Report p. 17). It is also possible to arrange the spin-echo experiment so that the resulting spectra depend only on instrumentally determined More importantly heteronuclear couplings can be eliminated or scaled down by flipping the abundant spins by 180”between acquisitions of the rare spin (i.e.”C).” This can provide an alternative to decoupling -see also page 29.A spin-echo sensitivity-enhancement technique is described in the following paragraph. Sensitivity Enhancement.-A series of have described methods for greatly increasing the effective sensitivity of nuclei such as 13C 15N,and 29Si. Morris and Freeman78 employed spin echoes to give a cross-polarization effect between protons and the insensitive nucleus. This is the same effect that operates in the 2-D correlation experiments described above; it requires an estimate of the H-X coupiing constant and enhances all X with that (or similar) coupling. The other papers79 employ a selective population transfer from one proton at a time giving a time saving for the coupled ”N of lo2-to lo4-fold.Both techniques allow the pulse rate to be determined by the proton TIrather than the slower X spin TI,and neither is (in principle) very demanding of hardware or software. N.M.R. of Organic Solids.-Last year’s Report described how a combination of ‘magic-angle’ spinning high-power (dipolar) decoupling and cross-polarization ” B. H. Meier and R. R. Ernst J. Amer. Chem. SOC.,1979,101,6441;J. Jeener B. H. Meier P. Bachmann and R. R. Ernst J. Chem. Phys. 1979,71,4546. ’’ K. Nagayama K. Wuthrich and R. R. Emst Biochem. Biophys. Res. Comm. 1979,90,305. 74 D.L.Rabenstein and T. T. Nakashima Anulyt. Chem. 1979,51 1465A. ” D.L.Rabenstein and A.A. Isab J. Mugn. Reson. 1979,36 281. 76 A.Bax A. F. Mehlkoff and J. Smidt J. Magn. Reson. 1979 35 167. ’‘ R.Freeman S. P. Kempsell and M. H. Levitt J. Mugn. Reson. 1979 35,447. 78 G.A.Morris and R. Freeman J. Amer. Chem. SOC.,1979,101,762. 79 H.J. Jakobsen and W. S. Brey J. Amer. Chem. SOC.,1979,101,774;J.C.S. Chem. Comm. 1979,478. Physical Methods and Techniques -Part (iii)N.M.R. Spectroscopy 29 techniques rendered high-resolution 13C spectra accessible from organic solids. Applications published this year include the demonstration of conformational preferences in the solid diethoxycarbonium ion [HC(OEt)2]' by observation of non-equivalent methylene carbons," the observation of solid-state organometallic fluxional processes,81 and a study of the gelation of sickle-cell haemoglobin.82 Significant industrial potential is clearly indicated in observations of toluene adsor- bed on Styrene can be manufactured by alkylation of toluene with methanol over zeolite catalysts and n.m.r.experiments revealed the identity of at least some of the intermediates involved. A neat and simple modification of the standard cross-polarization experiment has allowed Opella's group to observe non-protonated carbons in all that is required is to turn off the proton decoupler very briefly before FID acquisition and the protonated carbons almost instantly broaden. The technique is not only applic- able to solid amino-acids sugars steroids and other organics but sometimes turns out to be quicker for solid proteins than for protein solutions giving linewidths of ca.20 Hz. Finally it should be noted that it is possible to carry out double cross-polarization experiments from 'H to 13C to 15N (or from 'H to "N to "C) enabling either the observation of the natural-abundance 13C nuclei that are directly attached to a 15N label or the direct measurement of the relative concentration of 1sN-13C and lSN-l2C pairs without separation or purification of the solid This has opened the way for some fine biosynthetic and metabolic as the experiment monitors the history of a bond rather than an atom in an even more direct way than the 13C-2H methods discussed in last year's Report (p. 8). Miscellaneous.-Selective excitation of a single resonance using DANTE was described in last year's Report on p.19. Extensions this year include its use to generate transient n.0.e.'s2' and a slight modification to yield off -resonance de- coupled ~pectra.~ An alternative named CASS" has been described which does not require such modern computer control but instead needs a rather good decoupler system.86 Other new data-acquisition methods include coils for body surfaces,43 the simultaneous observation of 13C and 31P,42and improved ways of noise-decoupling which do not cause overheating of the ample.'^*^^ The flow of data-processing tricks also continues. A simple resolution-enhance- ment routine which requires no programming has been proposed,88 and Marshall has shown that a plot of absorption us. dispersion for a broad line will give valuable information on the broadening mechanism Two groups have used * A clever allusion to the senior author's history.J. R. Lyerla C. S. Yannoni D. Bruck and C. A. Fyfe J. Amer. Chem. SOC. 1979,101,4770. J. R. Lyerla C. A. Fyfe and C. S. Yannoni J. Amer. Chem. SOC. 1979,101 1351. J. W. H. Sutherland W. Egan A. N. Schecter and D. A. Torchia Biochemistry 1979,18 1797. M. D. Sefcik J. Amer. Chem. SOC. 1979,101,2164. 84 S. J. Opella M. H. Frey and T. A. Cross J. Amer. Chem. SOC.,1979,101 5856. 85 (a)J. Schaefer R. A. McKay and E. 0.Stejskal J. Magn. Reson. 1979,34,443; (6)J. Schaefer E. 0. Stejskal and R. A. McKay Biochem. Biophys. Res. Comm. 1979 88 274. 86 G. T. Andrews I. J. Colquhoun B. R. Doggett W. McFarlane B.E. Stacey and M. R. Taylor J.C.S. Chem. Comm. 1979,89. " V. J. Basus P. D Ellis H. D. W. Hill and J. S. Waugh J. Mugn. Reson. 1979 35 19. 88 B. Clin J. de Bony P. Lalanne J. Blais and B. Lemanceau J. Magn. Reson. 1979,33,457. A. G. Marshall and D. C. Roe J. Mugn.Reson. 1979,33 551. 30 J. K.M. Sanders decoupling difference spectroscopy to simplify and assign spectra 20,21 a spectrum with decoupling is subtracted from one without leaving responses only from the decoupled nuclei. This is particularly useful when the resonances are not otherwise resolved. ‘Conventional’ flow n.m.r. has proved to be useful for studying reactions that are too fast for standard n.m.r. techniques but which are slow on the time-scale of FID collection for example it is possible at -4O”C to see intermediates in the nucleophilic substitution of 2,4,6-trinitroanisole by b~tylamine.’~ Ernst has now developed a theory and tested it experimentally for stopped-flow n.m.r.when chemical evolution is occurring on the time-scale of one FID collectton i.e. rates of around 50-100 1mol-’ s-’.~’ Characteristic line shapes quite different from those seen in exchange equilibria were predicted theoretically and then observed experimentally. The beautifully designed apparatus achieved a dead time of only ca. 2 ms by abandoning the idea of a sample tube the receiver coil was simply wound around the mixing chamber.” The determination of molecular geometry by dissolving a compound in a nematic solvent to restore some dipolar coupling interactions is a well-established technique.However in a definitive paper from Diehl et al. the structure of [l-13C]benzene has been determined with a precision of better than 0.001Angstroms for all the inter- atomic ?;his paper discusses lucidly and in great detail all the problems which may arise from for example vibrational motion. Ernst has shown that geometrical information can be re-introduced into nematic-phase proton-decoupled 13 C spectra by the selective introduction of 2H or 15N. These added nuclei induce splittings in the 13C signals that can be directly converted into distance inf~rmation.~~ In the same paper a 2-D J technique is also described that was used to give geometrical inf~rmation.~~ The principles of spin imaging (also called spin mapping or Zeugmatography) were outlined in last year’s Report.The technique is potentially of medical importance if it can be made sufficiently sensitive that it can map portions of the body in a short time. To this end Ernst has carried out a detailed theoretical study of the potential sensitivity characteristics of the various different imaging methodsg4 An imaging method has been described which is based on 2-D n.m.r. and n.m.r. images have been obtained both from water flowing in the pores of ceramics96 and from solids.97 90 C. A. Fyfe S. W. H. Darnji and A. Koll J. Amer. Chem. Soc. 1979,101,951. 91 R. 0.Kuhne T. Schaffhauser,A. Wokaun and R. R. Ernst J. Magn. Reson. 1979,35,39. 92 P. Diehl H. Bosiger and H. Zirnrnerrnann J.Magn. Reson. 1979,33 113. 93 A. Hohener L. Muller and R. R. Ernst Mol. Phys. 1979,38,909. 94 P. Brunner and R. R. Emst J. Magn. Reson. 1979 33 83. 95 D. I. Hoult J. Magn. Reson. 1979 33 183. 96 R. J. Gumrnerson C. Hall W. D. Hoff R. Hawkes G. N. Holland and W. S. Moore Nature 1979,281 56. 97 R. A. Wind and C. S. Yannoni J. Magn. Reson. 1979 36 269.