2 Physical Methods Part (ii) Nuclear Magnetic Resonance By I. H. SADLER Department of Chemistry University ofEdinburgh West Mains Road Edinburgh EH9 3JJ The widespread use of n.m.r. spectroscopy in structure and stereochemical studies makes a complete survey of the field impracticable here. For such a survey the reader is referred to Volume 1 of the Chemical Society Specialist Periodical Report entitled 'Nuclear Magnetic Resonance,' which was published earlier this year. This Report deals primarily with three areas of increasing importance to organic chemists :the use of shift reagents chemically induced dynamic nuclear polarization and carbon-1 3 magnetic resonance. Other aspects of n.m.r. including conformation and kinetic studies have received somewhat short measure under miscellaneous studies where emphasis has been placed on methods rather than results.Some areas including theoretical work solvent effects and studies of nuclei other than hydrogen and carbon have had to be omitted. The previous survey of n.m.r. in Annual Reports appeared in 1968 and therefore where appropriate significant material from 1969 and 1970 has been included in this Report. 1 Chemical Shift Reagents A principal shortcoming of proton magnetic resonance lies in its inability to distinguish clearly between similar but not equivalent protons and the resulting region of the spectrum is often complex and difficult to analyse. The addition of transition-metal ions to induce differential shifts among similar protons usually also results in considerable line broadening and a consequent loss of information owing to the reduction by the paramagnetic species of the proton spin-lattice relaxation times (Tl).Most of the lanthanides however are considerably less efficient in reducing Tl than the more common transition-metal ions,' and euro- pium(II1) particularly so. The tris-P-diketone complexes of these metals behave as weak Lewis acids and readily increase their co-ordination number using ligands possessing lone electron pairs. Initially Hinckley reported2 that the presence of the dipyridine adduct of bis(dipivalomethanato)europium(m) Eu(dpm) ,2py (I) results in substantial ' R. L. Conger and P. W. Selwood J. Chem. Phys. 1952 20 383. C. C. Hinckley J. Amer. Chem. SOC.,1969 91 5160.18 Physical Methods-Part (ii) Nuclear Magnetic Resonance 19 downfield shifts of the proton resonances of cholesterol with very little line broadening. Larger downfield shifts are observed when the pyridine-free com- plex Eu(dpm) (2) is used ;,the resonances from different protons are shifted by different amounts and the spectrum assumes a more first-order appearance. The addition of 0.5 molar equivalents of (2) to a solution of 4-t-butylcyclo- hexanone causes the proton resonances to be separated over the range of 10p.p.m. clearly revealing the proton couplings. Asimilar separation of the spectrum is obtained for quinoline. The tris(dibenzoylmethanat0)-complex,Eu(dbm) (3) (1) R = CMe,; + 2py Eu (2) R = CMe (3) R = Ph 0-c \ is too insoluble for general use.The praseodymium complex Pr(dpm) causes upfield shifts approximately three times as large as the downfield shifts of Eu(dpm) but the signals are slightly less well defined.4 Shifts of proton reso- nances have been observed for a variety of molecules including alcoholsY2-' ester~,~*~,' amines,3,6.12,13 epoxide~,'~ 5,16 ethers, ketone^,^.^.'^,'^ oximes,' sulph~xides,'~ polymers,20 carbohydrates,21 steroid^,^,^,',^^ and terpene~.~~ J. K. M. Saunders and D. H. Williams Chem. Comm. 1970,422; J. Amer. Chem. Soc. 1971 93 641. J. Briggs G. H. Frost F. A. Hart G. P. Moss and M. L. Staniforth Chem. Comm. 1970 749. P. V. Demarco T. K. Elzey R. B. Lewis and E. Wenkert J. Amer. Chem. Soc. 1970 92 5734 5737. D.R. Crump J. K. M. Saunders and D. H. Williams Tetrahedron Letters 1970 4419,4949. ' C. C. Hinckley J. Org. Chem. 1970 35 2834. G. V. Smith W. A. Boyd and C. C. Hinckley J. Amer. Chem. Soc. 1971,93 6319. D. E. Sunco L. Tomic Z. Majerski and M. Tomic Chem. Comm. 1971 719. lo G. H. Waal and M. R. Peterson Chem. Comm. 1970 1167. l1 A. F. Cockerill and D. M. Rackham Tetrahedron Letters 1970 5149. l2 D. C. Remy and W. A. Van Sam Tetrahedron Letters 1971,2463; H. Van Brederode and W. G. B. Huysmans Tetrahedron Letters 1971 1695. l3 E. Ludger Chem.-Ztg. 1971 95 325. l4 M. R. Willcott J. F. M. Oth J. Thio G. Plinkie and G. Schroder Tetrahedron Letters 1971 1579. F. I. Carroll and J. T. Blackwell Tetrahedron Letters 1970 4173. l6 M. Yoshimoto T. Hiraoker H.Kuwano and Y. Kishida Chem. and Pharm. Bull. (Japan) 1971 19 849. P. Belanger G. Freppel D. Tizane and J. C. Richer Chem. Comm. 1971 266. Z. W. Wolkowski Tetrahedron Letters 1971 825. l9 R. R. Fraser and Y. Y. Wigfield Chem. Comm. 1970 1471. 2o A. R. Katritsky and A. Smith Tetrahedron Letters 1971 1765; F. F. L. Ho J. Polymer Sci.,Part B Polvmer Letters 1971 9 491. 21 I. Armitage and L. D. Hall Canad. J. Chem. 1971,49 2770. 22 J. E. Hertz V. M. Rodriguez and P. Joseph-Nathan Tetrahedron Letters 1971 2947. 23 D. G. Buckley G. H. Green E. Ritchie and W. C. Taylor Chem. and Ind. 1971 298; 0. Achmatowicz A. Ejchart J. Jarczak L. Kozerski and J. St. Pyrek Chem. Comm. 1971 98. 20 I. H. Sadler Nitro-compounds show only very small shifts and compounds containing only halogens or carbon-carbon double bonds remain ~naffected.~ The presence of acidic or phenolic groups causes decomposition of the shift reagents.The use of dry pure materials and solvents is necessary during the preparation and use of the reagents since they preferentially complex with water and their effec- tiveness is greatly reduced if not entirely eradicated. For similar reasons polar solvents are to be avoided. The magnitude of the shift normally varies linearly over a shift-reagent-to-substrate molar ratio range of 0.14.5. This is to be expected for a rapid equilibrium between a reagent-substrate complex and substrate molecules." At molar ratios >0.5 the shift magnitude tends to a limiting value. Shifts are reported either as values extrapolated linearly to a molar ratio of 1.0 or preferably as the gradient (p.p.m./mol.M(dpm),/mol. substrate) of the linear portion of the shift us. molar ratio graph. These gradients depend upon the substrate functional group and its environment. In the absence of steric factors the shift gradients of the methylene group in RCH,X decrease3 along the sequence X = NH > OH > C0.C > OR > C0,R > CN. Shift magnitudes can be increased by lowering the temperature9* but this normally results in broader signals. The shifts are to a pseudocontact interaction between the metal atom and a substrate atom possessing one or more lone electron pairs. Since this is a through-space magnetic dipolar interaction which alters the local mag- netic field but does not affect the bonding electron densities coupling constants should not be altered.Providing therefore that the resonances are not too ill defined it is valid to obtain coupling constants directly from the spectrum if first-order conditions hold. Thus it has been possible24 to obtain long-range coupling constants for the interaction of the ring protons with the methyl protons in 2-methylpyrazine and its monomethoxy-derivatives. The magnitude (Ad) of this type of shift is expressed by the equation:25 A6 = K(3 COS~ $ -1)r-where K is a constant for a particular complex at a given temperature 4 is the proton-metal-co-ordination site internuclear angle and r is the proton-metal distance. Thus in a particular example the relative shifts depend only on the associated angles and distances.Several worker^^.^.^,'^*' have found that such shifts often seem to a fair approximation to show only a dependence on r-3. In these cases the metal atom must be so located that the variation in 4 is suffi-ciently small enough to be ignored. Reports5*' ',I3 of correlation with other orders of r can probably be attributed to an incorrect assumption of the location of the metal atom or neglect of the angular dependence or both. The importance of the angular variation is illustrated26 by the Eu(dpm) induced upjeld shifts of the methoxy-proton resonances in (4). All other proton resonances are dis- placed downfield. This behaviour is expected when (3 cos 4 -1) < 0 i.e. 24 A. F. Bramwell G. Riezebos and R.D. Wells Tetrahedron Letters 1971 2489. 25 H. M. McConnell and R. E. Robertson J. Chem. Phys. 1958,29 1361. 26 P. H. Mazzocchi H. J. Tamburin and G. R. Miller Tetrahedron Letters 1971 1819. Physical Methods-Part (ii) Nuclear Magnetic Resonance 55" < 4 < 125". The opposite effect is observed using Pr(dpm),. With both reagents a sufficiently first-order spectrum was obtained to allow the coupling constants for all protons to be measured. In the presence of Eu(dpm) no shifts are observed27 for the meta- and para-proton resonances in cis-4-t-butyl-l- phenylcyclohexanol(5). This is consistent with europium-oxygen co-ordination. The proton resonance shifts observed for pyridine the picolines pyridazine pyrimidine pyrazine quinoline and isoquinoline cannot be correlated with a realistic value for the europium-nitrogen distance unless the angular dependence is included.28 Confirmatory evidence for the pseudocontact origin of the shift has been obtained from a of the Pr(dpm) induced shifts in horneol (6).The rigid stereochemistry of this molecule provided a model in which geo- metrical factors could be accurately calculated. A set of values for the geometric factor (3cos2 4 -l)r-, was calculated for all protons using a position for the metal atom estimated from inspection of Dreiding models and a consideration of other factors. These values were refined by a reiterative procedure to make deviations from proportionality with observed shifts as small as possible a value for K being selected to allow direct comparison with observed shifts.A complete assignment of the resonances and evaluation of the coupling constants was also possible. The praseodymium atom co-ordinates derived in this study have been used to calculate30 the induced shifts expected in the carbon-I3 magnetic resonance spectrum of borneol. These agreed well with the experi- mental values and allowed for the first time the complete assignment of the carbon-13 resonances. Analogous results were also obtained using the europium complex. It is noteworthy that the proportionality constant K for the best Me Me v (6) (7) 27 N. S. Bhacca and J. D. Wander Chem. Comm. 1971 1505. W. L. F. Armarego T. J. Batterhan and J. R. Kershaw Org. Magn. Resonance 1971 3 575.29 J. Briggs F. A. Hart and G. P. Moss Chem. Comm. 1970 1506. 2o J. Briggs F. A. Hart G. P. Moss and E. W. Randall Chem. Comm. 1971 364. 22 I. H. Sadler fit of the carbon-13 data is 1.16 KH where KH is the corresponding constant for the proton data. Since the pseudocontact shift A6 is independent of the nuclear magnetic moment these constants should be identical. These results show that the shifts originate predominantly if not exclusively from a pseudocontact mechanism and providing that the metal atom is correctly located there is no need to involve a major contact contribution. The use of a computer program has been reported31 to determine the position of the lanthanide ion by maxi- mizing the correlation bet ween the experimentally measured induced shift and the geometric factor.Its application to adaman tan-2-01 trans-4-t- bu tylcyclo- hexanol and the hydroxyoxetan (7) is described. The results for (7) indicate that the europium atom co-ordinates to the ether oxygen atom. The presence of a small contact contribution cannot always be excluded especially in the presence of aromatic molecules where unpaired electron spin density can be transferred through a n-electron system e.g. in quinoline the J2.3 coupling appears to decrease3 at relatively high reagent-to-substrate molar ratios. A contact contribution has been also demon~trated~~ for 31Pshifts in phosphates and phosphonates induced by Eu(NO,) ,6D,O. In molecules possessing more than one functional group the preferred order of co-ordination of Eu(dpm) to saturated molecules has been found33 to be alcohols > amines > ethers > ketones -esters.This sequence may be altered by steric factors. Unsaturation increases co-ordination to ketones. Where an ether oxygen atom is part of an aromatic system as in furan the co-ordination is very weak. In the polyethers MeO(CH,CH,O),Me symmetrical co-ordination of Eu(dpm) or Pr(dpm) to the two extreme oxygen atoms for n = 4-8 yields relatively simple spectra.34 A graphical method has been described3' for separat- ing the relative contributions to induced shifts resulting from the metal atom co-ordinating at two sites in a molecule and has been applied to testosterone and its 17wmethyl derivative. It should be emphasized that this method is based on the assumption that the angular dependence in the above equation may be ignored and should therefore be used with caution.Although less extensively st~died,~~-~~ the ytterbium complex Yb(dpm) is of value and causes shifts approximately twice those of Eu(dpm) in the same direction. A comprehensive of amines indicates that steric factors related to the accessibility and inversion of the lone electron pair are particularly impor- tant and complexing ability increases along the sequence triethylamine < pyridine < diethylamine < quinuclidine < ethylamine. The data correlate well with r-3 assuming a Yb-N distance of 3A. Ketone and aldehyde studies37 31 S. Farid A. Ateya and M. Maggio Chem. Comm. 1971 1285. 32 J. K. M. Saunders and D.H. Williams Tetrahedron Letters 1971 2813. 33 H. Hart and G. M. Love Tetrahedron Letters 1971 625. 34 A. M. Groters T. Smid and E. de Boer Tetrahedron Letters 1971 4863. 3s C. C. Hinckley M. R. Klotz and F. Patil J. Amer. Chem. SOC.,1971 93 2412. 36 C. Beaute Z. W. Wolkowski and N. Thoai Tetrahedron Letters 1971 817. 37 Z. W. Wolkowski Tetrahedron Letters 1971 821. 38 Z. W. Wolkowski C. Beaute and N. Thoai Chem. Comm. 1971 700. 39 C. Beaute Z. W. Wolkowski J. P. Merda and D. Lelandais Tetrahedron Letters 1971 2473. Physical Methods-Part (ii) Nuclear Magnetic Resonance 23 reveal stereochemical information e.g. phorone and mesityl oxide retain pre- dominantly a sym-cis conformation. In fluorenone the ytterbium atom is collinear with the carbonyl group but in indan-1-one it is displaced on a line 60" from this axis and anti to the aromatic ring.O~imes~~ appear to co-ordinate at the oxygen atom and imines azo- and nitro-compounds remain virtually ~naffected.~~ Amides show shifts comparable with those of ketones co-ordination being via the carbonyl group3* The reagent has been used to assign39 structures to several chloro- and bromo-vinylaldehydes. A systematic quantiative study of the effect of dpm complexes of all the lan- thanides except La ce and Lu on the 'H n.m.r. spectra of 4-vinylpyridine 4-picoline N-oxide and n-hexyl alcohol has been rep~rted.~' In addition to the forementioned upfield shifts are reported for Nd Sm Tb Dy and Ho while downfield shifts are observed for Er and Tm.The study confirms the geo- metric dependence of the shift ratios and the signs of the shifts correlate quali- tatively with available magnetic anisotropy data confirming their pseudocontact nature. Spectra of Gd(dpm) systems were severely broadened suggesting a contact shift which if present in any lanthanide complex could be expected to show up here. A study of relative broadening abilities of these lanthanides shows that although Tb Dy Ho and Tm cause shifts greater in magnitude than those of Pr Eu and Yb they also give broader lines. Chemical shift reagents provide a powerful tool for solving stereochemical and conformational studies. Typical applications include the assignment4' of configurations to the phosphetans (8) by examination of the resonance of the Me Me Me Me (8) R = Ph Me OEt SEt CH,Ph or NHCH,Ph C-3 proton ;the estimation42 of syn-anti ratios in oxime mixtures obtained from alkyl methyl ketones ; the determination of the stereochemistry of cis- and tr~ns-4-t-butyl-l,2,2-trimethylcyclohexanols,~~ 3-(l-naphthyl)-l,3,5,5-tetrame- thylcyclohexanols,44 and the perhydrophenalols (9) and ( of 3-endo-substituted bicyclo[3,3,l]~onanes indicate that these adopt boat-chair conformations.Possible conformational changes induced by complexing to the europium atom were shown to be only very slight. Bicyclo[3,3,1]nonan-3-one 40 W. D. Horrocks and J. P. Sipe J. Amer. Chem. SOC.,1971 93 6800. 41 S. Trippett and J. R. Corfield Chem. Comm. 1971 721. 42 K. 0. Berlin and S. Rengaraja J.Org. Chem. 1971 36 2912. 43 P. Belander C. Freppel D. Tizane and J. C. Richer Canad. J. Chem. 1971,49 1988. 44 B. L. Shapiro J. R. Hlubucek G. R. Sullivan and L. F. Johnson J. Amer. Chem. Suc. 1971,93 3281. 45 F. A. Carey J. Org. Chem. 1971 36 2199. 46 I. Fleming S. W. Hanson and J. K. M. Saunders Tetrahedron Letters 1971 3733. I. H.SadIer appears to exist as a 1 1 mixture of chair-chair and chair-boat conformation^.^^ A preferred conformation has been assigned48 to griseofulvin. Structures have been assigned49 to the diastereomeric forms of bis(phenylsulphiny1)methane (Ph-SOCH,.SO-Ph) on the basis of the resonances of the methylene protons which in the absence of shift reagent appear as singlets. In the presence of Eu(dpm) those from the meso-form being non-equivalent appear as a quartet whereas those from the racemic form being equivalent remain as a singlet.The introduction of chiral shift reagents has made possible the determination of optical purity by n.m.r. In the presence of tris-[3-(trifluoromethylhydroxy-methy1ene)-(+)-camphorato]europium (1l) the singlet resonances from the C-l protons of 2-phenylbutan-2-01 may be separated” by 0.3 p.p.m. Tris-[3-pivaloyl-(+)-camphorato]europium (12) is particularly effective for amines. (11) R = CF (12) R = But Slight disadvantages of the dpm reagents lie in their low solubility in non- alcoholic solutions and their markedly reduced effectiveness when used with weak Lewis bases e.g.ethers and esters. These problems are largely overcome52 by the use of tris-(1,1,1,2,2,3,3-heptafluoro-7,7-dimethyloctadionato)europium Eu(fod) (13) and the corresponding praseodymium complex Pr(fod) .The presence of the electronegative fluorine atoms increases the acceptor properties of the reagent. Relatively high concentrations (> 1.0 molar ratio shift reagent substrate) may be used and virtually no peak-broadening is observed. The proton resonances of di-n-butyl ether which are not significantly affected by Eu(dpm), are displaced substantially by Eu(fod),. This reagent has been of 47 M. R. Vegar and R. J. Wells Tetrahedron Letters 1971 2847. 48 S. G. Levine and R. E. Hicks Tetrahedron Letters 1971 31 1. 49 J. L. Greene and P. B. Shevlin Chem. Comm. 1971 1092. H. L. Goering J. M. Eikenberry and G.S. Koermer J. Amer. Chem. Soc. 1971 93 591 3. 51 G. M. Whiteside and D. W. Lewis J. Amer. Chem. Soc. 1970,92 6979; 1971,93,5914. 52 R. E. Rondeau and R. E. Sivers J. Amer. Chem. Soc. 1971,93 1522. Physical Methods-Part (ii) Nuclear Magnetic Resonance particular value in assigning alkyl resonances of tertiary alkylamides5 and the stereo~hemistry~~ of the tricyclic benzodioxans (14). A different type of shift reagent has been reported,55 in which the compound to be studied is held in the ring-current-induced asymmetric magnetic field of a porphyrin ring system by covalent bonding to a germanium atom located at the centre. The bonds are formed by the reactions of compounds (15) and (16) with functional groups on the compound being studied.Suitable groups are weakly and strongly acidic hydroxy-groups and metallic ligands in Grignard reagents and organolithiums. In view of the inapplicability of the lanthanide shift reagents to halides carboxylic acids and phenols the new reagents are particularly useful. The proton resonances of the methyl group in dimethyl- germaniumporphin (17) obtained from (15) and methylmagnesium iodide (15) R = H; X = C1 R (16) R= Ph; X = OH (17) R= H; X = Me (18) R = H; X = n-C,H X occur as a sharp singlet at 18 z and those from the corresponding di-n-octyl derivative (18) are spread over 9-187. Since these reagents are diamagnetic they have the advantage that they cannot show line-broadening effects which sometimes becomes significant with lanthanide shift reagents.2 Chemically Induced Dynamic Nuclear Polarization The ob~ervation~~,~’ of enhanced absorption (A) or emission (E) signals in n.m.r. spectra of reacting systems is evidence for the participation of free radicals 53 L. R. Isbrandt and M. T. Rogers Chem. Comm. 1971 1378. 54 J. F. Caputo and A. R. Martin Tetrahedron Letters 1971 4547. 55 J. E. Maskasky and M. E. Kenney J. Amer. Chem. SOC.,1971,93 2060. 56 H. Fischer and J. Bargon Z. Naturforsch. 1967 22a 1551 1556; 1968 23a 2109; H. R. Ward J. Amer. Chem. SOC.,1967 89 5517. 57 H. Fischer and J. Bargon Accounrs Chem. Res. 1969 2 110; R. G. Lawler J. Amer. Chem. SOC.,1967 89 5519. 26 I. H. Sadler in the reaction and is generally referred to as chemically induced dynamic nuclear polarization (CIDNP).The effect was originally explained57 by cross- relaxation of electron and nuclear spins resulting in a non-Boltzman distribution in the nuclear spin levels of the radical which was transferred to the products upon reaction. More evidence of the phenomenon showed this theory to be untenable. In particular it is unable to explain multiplet effect^^^,^^ (the presence of both A and E within a multiplet) and places a theoretical limit on the magnitude of the polarization which is often exceeded in practice.59 The experimental observations for combination and disproportionation reactions carried out within the spectrometer are explained rather better by the radical pair theory first developed independently by CIOSS~~~~ and by Kaptein and Oosterhoff.62 Similar treatments differing slightly in quantitative formulation have been offered63 by Fischer and Adrian.The reaction of a precursor molecule (M) yields a radical pair (RP) which may either yield cage products by combination or disproportion of the components within the cage or separate into two free radicals. The separate radicals may then undergo transfer reactions or form new radical pairs by diffusion M -+ R-P-+ R-P 1 sx diR. RX,PX + R-+ Pa -+ R-R,P-P Initially the electronic state of the radical pair is that of the precursor M. Inter-actions between the unpaired electrons and neighbouring nuclei in a particular radical pair cause mixing of the electronic singlet (S) state and the M = 0 component of the triplet (To)state to an extent that is governed by the nuclear spin states.Mixing of the singlet state with the M = +1 components of the triplet (T+)state are not significant in fields greater than a few thousand gauss. Since the rate of cage-product formation depends upon the singlet character of the mixed electronic state those nuclear spin states which give rise to a state predominantly triplet in character will be depleted in the cage product and those conferring a predominantly singlet nature on the mixed state will be enhanced in the cage product. In certain cases this leads to observable polarizations. It is possible to obtain the magnitude and sign of the polarization from the hyperfine splitting constants and g-values of the component radicals and the multiplicity of the precursor.Free-radical displacement and trapping reactions 58 H. R. Ward and R. G. Lawler J. Amer. Chem. Soc. 1967 89 5518. 59 G. L. Closs and L. E. Closs J. Amer. Chem. Soc. 1969 91 4549 4550. 6o G. L. Closs J. Amer. Chem. Soc. 1969 91 4552; G. L. Closs and A. D. Trifunac J. Amer. Chem. Soc. 1969 91 4554; 1970 92 2183; G. L. Closs C. E. Doubleday and D. R. Paulson ibid. p. 2185. 61 G. L. Closs and A. D. Trifunac J. Amer. Chem. Soc. 1970 92 2186. 62 R. Kaptein and L. J. Oosterhoff Chem. Phys. Letters 1965 4 195 214. 63 H. Fischer Chem. Phys. Letters 1970 4 611; 2. Naturforsch. 1970 25a 1957; F. J. Adrian J. Chem. Phys. 1970 53 3374; 1971 54 3912. Physical Methods-Part (ii) Nuclear Magnetic Resonance 27 can be treated by the same model when nuclear relaxation processes are in- cl~ded.~~ Both multiplet and net emission and absorption effects are explained and no limit is placed on the magnitude of the polarization.Combination of identical radicals can only give rise to multiplet effects (not observable for single resonances). Cage products are expected to show opposite polarization to products resulting from radicals escaping from the pair (e.g. transfer products) and this is observed in practice.65 Polarizations resulting from a singlet precursor are opposite in sign to those from a triplet precursor. Recombination of radical pairs formed by diffusive encounters gives polarization of the same sign as a triplet precursor but with reduced intensity.66 The model provides a satisfactory explanation for a large number of observed polarizations including those of products from the singlet and triplet decompositions of benzoyl peroxides in a variety of solvent^,^' and those of 1,1,2-triarylethanes obtained6' by (a) the photochemical decomposition of diphenyldiazomethane in toluenes (b) the thermolyses of the appropriate azo-compounds and (c) the decomposition of peroxides in mixtures of toluenes and diphenylmethanes.Tomkiewicz and Cocivera have extended6* the theory to account for the different intensity distributions (as well as the difference in sign) observed for products resulting from singlet and triplet precursors of a radical pair. Kaptein has presented69 two qualitative rules from which the sign of polariza- tion observed for a product may be predicted.The CIDNP spectrum of a nucleus i originally belonging to radical a of a pair ab and possibly coupled to a nucleusj in the product is described by the sign of r for net effects (A or E) and rmfor multiplet effects (E/A low-field part of a multiplet E high-field part A ;A/E reverse of E/A) r =~EA~A~ +ve for A -ve for E r =pcA,A,o,,Ji +ve for E/A -ve for A/E where the parameters take the following signs -for ab formed from a singlet precursor +for ab formed from a triplet precursor 4+ for ab formed by diffusive encounter of two radicals +for cage recombination (or disproportionation) products &{ -for products of escaped radicals (e.g. transfer products) +for nuclei i and j belonging to the same radical gij{ -for nuclei i and j belonging to different radicals +for g-factor of a greater than g-factor of b -for g-factor of a less than g-factor of b 64 G.L. Closs and A. D. Trifunac J. Amer. Chem. SOC. 1970,92 7227; G. L. Closs and D. R. Paulson J. Amer. Chem. Soc. 1970 92 7229. 65 B. Blank and H. Fischer Helv. Chim. Acta 1971 89 905. 66 M. Lehnig and H. Fischer Z. Naturforsch. 1970 25a 1963; J. Phys. Chem. 1971 75 3410. 67 R. Kaptein J. A. den Hollander D. Antheunis and L. J. Oosterhoff Chem. Comm. 1970 1687; S. R. Farenholtz and A. M. Trozzolo J. Amer. Chem. SOC.,1971,93 253. 68 M. Tomkiewicz and M. Cocivera Chem. Phys. Letters 1971 8 595. 69 R. Kaptein Chem. Comm. 1971 732. I.H. Sadler Ai Aj,and Jijtake the signs of the appropriate hyperfine splitting and nuclear spin coupling constants. The correct polarization is obtained in straightforward cases where the signal exhibits either a net effect or a multiplet effect. Where both multiplet and net effects are observed the spectrum is usually described by the combination of both rules; however under conditions where AgPHOis very large compared with Ai or where the spectra are very strongly coupled the multi- plet effect may be reversed. The rules imply that the strong emission signal observed for p-dichlorobenzene obtained during the decomposition of di-(p-chlorobenzoyl) peroxide in hexa- chloroacetone originates from p-chlorophenyl radicals that have escaped from an aroyloxy-aryl radical pair.This conclusion was also reached from chemical arguments65 concerning the effect of trapping agents upon the CIDNP patterns and product distributions. These rules will be of obvious value in mechanistic studies with regard to precursor multiplicity and radical pair composition. Emission signals have also been observed for benzene formed via the de- compo~ition~~ of N-nitrosoacetanilide phenylazotriphenylmethane and dia- zoaminobenzene in cyclohexane toluene cumene and acetic anhydride and for methane formed by the thermal decomposition7’ of 00f-diacetyl-4-hydroxy- aminoquinoline 1-oxide (19) in dioxan. On the basis of CIDNP effects observed during the acetylation of y-picoline N-oxide with acetic anhydride the inter- mediacy of the radical pair (20) is proposed.72 The ob~ervation’~ of an E/A OAc (19) multiplet for the methine proton Ha,in the aniline derivative (23) obtained from the reaction of benzyne with NN-dimethylbenzylamine has been used as con- firmatory evidence for the formation of radical pair (22) via the steps shown from an ortho-betaine (21).The sign of the multiplet is in accordance with Kaptein’s Rules. A similar polarization effect is reported74 for an analogous reaction between benzyne and dibenzyl sulphide. Strong emission by the methine proton in the oxime ether (25) obtained75 by thermolysis of fluorenone ’O L. F. Kasukhin M. P. Ponomarchuck and V. N. Kalinin Zhur. org. Khim. 1970 6 2531. Y. Kawdzoe and M. Araki Chem. and Pharm. Bull. (Japan) 1971 19 1278.72 H. Iwamura M. Iwamura T. Nishida and S. Sato J. Amer. Chem. Soc. 1970,92,7474. 73 A. R. Lepley R. H. Becker and A. G. Giumanini J. Org. Chem. 1971 36 1222. 74 H. Iwamura M. Iwamura T. Nishida M. Yoshida and J. Nakayana Tetrahedron Letters 1971 63. ’5 D. G. Morris Chem. Comm. 1971 221. Physical Methods-Part (ii) Nuclear Magnetic Resonance 29 Me Me Me Me t) I I p$:Me Ph-A+-Me + [Ph-N?1 .Me "-;'-M] I --+ Ph-y: I + I -H-C-Ph Ph-C-Ph-C-Ph -C* Ph-C-Me I I I H H Ha N-benzhydrylnitrone (24) provides further evidence for the homolytic nature of the Martynoff rearrangement. CIDNP during the photolysis of di-isopropyl ketone has been described.76 In most solvents polarization is in accord with radical pair formation via a Norrish Type I split from a triplet state ketone.In carbon tetrachloride the reaction probably involves complex formation of the solvent with the excited singlet state of the ketone. 0-+/ X=N -P X=N-O-CHPh, \ CHPh (24) X = 9-Fluorenylidene (25) A of the decomposition of 0.2 moll -' isobutyryl peroxide in hexa- chlorobutadiene containing varying initial concentrations of bromotrichloro- methane shows a variation and sign reversal for the polarization of the chloroform signal implying formation by two competing routes (see Scheme). At concentra-tions of bromotrichloromethane greater than 0.11 mol I-' an enhanced absorp- tion signal is obtained indicative of chloroform formation mainly by dispropor- tionation of a singlet [.CCl, CHMe,] radical pair (S); this being formed by S ~ (RCO,) + 2C0 + 2R.--+ RR RH R(-H) S RBr + R. CCI --+ RCCI, CHCI, R(-H) lBrCCII (A) RBr + 2433 R. + .CCl R. CCI + RCCI, CHCI, R( -H) (E) Scheme l6 J. A. den Hollander R. Kaptein and T. A. T. M. Brand Chem. Phys. Letters 1971 10 430. l7 R. Kaptein F. W. Verheus and L. J. Oosterhoff Chem. Comm. 1971 877. I. H. Sadler reaction of the initial singlet [2*CHMe,] radical pair with bromotrichloromethane. At concentrations of bromotrichloromethane less than 0.11mol I-' an emission signal results indicating the major route is from a radical pair (F) formed by diffusive encounters of CC1 and CHMe radicals. Consideration of the likely reaction rates suggests that spin correlation effects of radical pairs in solution may be of relatively long duration (microseconds).A variation of product polarization with concentration of peroxide and transfer reagent has also been observed78 in the decomposition of phenyl acetyl peroxide in carbon tetra- chloride-bromotrichloromethane mixtures. CIDNP studies79 of the decomposition of bis(pentafluorobenzoy1) peroxide in hexachlorobutadiene containing a variety of aromatic compounds suggest that the pentafluorobenzoyloxyl radical shows marked electrophilic properties. The photolysis of diazomethane in carbon tetrachloride to yield pentaerythrityl tetrachloride is thought to proceed via a multistep chain reaction initiated by reaction of methylene with carbon tetrachloride. Roth has obtained" an emission signal from a minor side-product 1,1,1,2-tetrachloroethane,formed by cage recombination of the radical pair produced in the initiation step indicating that methylene is formed predominantly if not exclusively as a singlet species.Tri- chloromethyl radicals escaping the pair are involved in propagation of the chain process. CIDNP spectra obtained" during the singlet and triplet photo- sensitized photolysis of diazirine in deuteriotrichloromethane indicate singlet methylene preferentially abstracts chlorine atoms whereas triplet methylene abstracts hydrogen atoms ;the presence of one-step C-H insertion by methylene is not excluded by these results. Strong emission signals have been obtained8' for all dimerization products of the diradical (27) formed by thermolysis of the bicyclic azo-compound (26).Closs has rationalized8 this using an extension of the radical pair theory. A one-step reaction between two triplet trimethylene diradicals may give either multiplet or emission spectra ;if the reaction occurs in two steps by an initial singlet-triplet combination to give a triplet diradical pure emission only may be observed resulting from singlet-triplet transitions. No polarization is observable for a one-step reaction between two singlet trimethylene Me (26) 78 C. Walling and A. R. Lepley J. Amer. Chem. Soc. 1971 93 546. 79 J. Bargon J. Amer. Chem. SOC.,1971 93 4630. H. D. Roth J. Amer. Chem. Soc. 1971 93 1527. H. D. Roth J. Amer. Chem. SOC.,1971,93,4935. *' J. A. Berson R. J. Bushby J. M. McBride and M.Tremelling J. Amer. Chem. SOC. 1971,93 1545. 83 G. L. Closs J. Amer. Chem. Soc. 1971 93 1546. Physical Methods-Part (ii) Nuclear Magnetic Resonance 31 diradicals. Unexpected6' observation'" of CIDNP effects in the spectrum of dec- 1 -ene obtainable from a decamethylene diradical formed during the thermol- ysis of cyclohexane diperoxide (28),is attributed to the availability of multiplicity independent competing reaction paths for the diradical. Relatively few studies have been carried out using low-strength magnetic fields. In such cases e.g. the rea~tion'~ of act-dichlorotoluene with ethyl-lithium polarization is detected by transferring the sample to the spectrometer after the reaction is complete. Under these conditions mixing of the singlet state with all three triplet states becomes irnp~rtant'~ and Kaptein's rules require m~dification.~' A simple method has been developed" for the calculation of nuclear polarizations in products of radical coupling reactions carried out at zero magnetic field and explains the polarization of 4-bromo-2,3-dideuterio- chlorobenzene obtained by the decomposition of the parent diacyl peroxide in bromotrichloromethane-hexachlorobutadieneoutside the spectrometer.The high-field spectrum is correlated with the zero-field energy levels. Low-field polarizations have also been observed*' for reaction of the naphthalene anion- radical with isopropyl chloride. No polarization is observed or expected at high fields. CIDNP is obviously of immense value in mechanistic organic chemistry ; however it cannot be stressed too strongly that the observation of such effects imply neither that the product is necessarily the major product nor that the ob- served pathway is a major one ; neither can pathways be excluded solely on the basis of no observed effect.3 Carbon-13 Resonance The low natural abundance (1.1 %) and inherent insensitivity to detection (1.6% of that for the proton for a given field strength) of the nucleus has hindered carbon-13 magnetic resonance studies by conventional continuous-wave tech- niques. Improvements over the past few years in instrument design have led to increased basic sensitivity and stability the latter allowing the effective use of spectrum accumulation. Proton noise decoupling the broad-band irradiation of the entire proton resonance frequency range while recording the 13C n.m.r.spectrum increases sensitivity by a factor of about seven by combining the com- ponents of a multiplet within a single peak and also by virtue of a positive inter- nuclear Overhauser effect. Although it gives a simpler spectrum this procedure eliminates structural information about the number of protons bonded to each carbon nucleus. This can be remedied by off-resonance deco~pling,~' where irradiation is applied some two or three hundred cycles away from the proton 84 R. Kaptein M. Frater-Schroder and L. J. Oosterhoff Chem. Phys. Letters 1971,12 16. 85 H. R. Ward R. G. Lawler H. Y. Loken and R. A. Cooper J. Amer. Chem. SOC. 1969,91,4928.86 F. J. Adrian Chem. Phys. Letters 1971 10 70. 87 J. L. Charlton and J. Bargon Chem. Phys. Letters 1971 8 442. J. F. Garst F. E. Barton and J. I. Morris J. Amer. Chem. SOC.,1971 93 4310. 89 E. Wenkert A. 0. Clause D. W. Cochrane and D. Doddrell J. Amer. Chem. SOC. 1969,91 6878. 32 I.H. Sadler resonances. This results in a partially decoupled spectrum in which the carbon resonances show relatively small splittings (10-50 Hz) from the directly bonded protons only while the Overhauser enhancement remains virtually unaffected. A major development has been the commercial availability of instruments for routine Fourier transform nuclear magnetic resonance spectrosc~py~~~~~ (FT-n.m.r.). This technique involves the irradiation of the sample with a radio frequency pulse short enough to simultaneously excite the resonances of all nuclei of a given isotope in the molecule.The resulting decay of magnetization of the sample the free induction decay is recorded as a function of time and converted by Fourier transformation into a conventional frequency spectrum. The pulse duration is a few microseconds and the free induction decay is recorded over ca. one second. A complete spectrum may therefore be obtained in a time considerably less than the spin-lattice relaxation times of the nuclei. Time averaging of many runs leads to about a ten-fold improvement in sensitivity over normal methods for the same total observation time. Measurement of spin-lattice relaxation times (TI)for individual carbon-13 nuclei is fast becoming a routine procedure in noise-decoupled FT-n.m.r.Normally an inversion-recovery technique is empl~yed,~~,~~ in which a 180" pulse is applied to invert the spin-level populations followed after a delay time T,by a 90" pulse for initiation of the free induction decay. The pulse sequence (180-7-90) may be repeated when thermal equilibrium of the spin system has been re-established usually at least three times the longest value. This method has also been termed 'partially relaxed Fourier transform n.m.r.' (PRFT).93 The peak intensities (A,) in a PRFT spectrum are giveng4 by A = A,[1 -2exp(-z/T1)] where A is the corresponding intensity for the spin system at equilibrium. Resonances appear as emission or absorption lines according to whether 7 is smaller or greater than 7''In 2.This display of the spectrum often resolves overlapping resonances arising from carbon atoms with very different relaxation times. Normally a series of PRFT spectra are obtained as a function of z. Application of this technique to different organic molecules has shown that TI values for carbon-13 nuclei are not always as great as has been formerly supposed. For small and highly symmetrical molecules Tl values are often greater than ten seconds and sometimes as long as a minute for example 3,5-dimethylcyclohex-2- enone (29)95 and 2-ethylpyridine (30),92(Tl values in seconds). For relatively large and asymmetric molecules however Tl values are often less than five seconds for example adenosine monophosphate (3 1).94 Allerhand and Doddrell and co-workers have determined carbon-13 7'' values for several widely differing types of natural product including protein ribonu~lease,~~ cholesteryl chloride 90 W.Bremser H. D. W. Hill and R. Freeman Messtechnik 1971 79 14; T. C. Farrar and E. D. Becker 'Pulse and Fourier Transform NMR,' Academic Press New York 1971. y1 E. Breitmaier G. Jung and W. Voelter Angew. Chem. Internat. Edn. 1971 10 673. 92 R. Freeman and H. D. W. Hill J. Chem. Phys. 1970,53 4103. 93 A. Allerhand D. Doddrell V. Glusko D. W. Cochran E. Wenkert P. J. Lawson and F. R. N. Gurd J. Amer. Chem. Sor. 1971,93 544. 94 A. Allerhand and D. Doddrell J. Chem. Phys. 1971 55 189. 95 R. Freeman and H. D. W. Hill J. Amer. Chem. Soc. 1971,54 3367.Physical Methods-Part (ii)Nuclear Magnetic Resonance 0 o\\P,o-6H2Q!H 0.19 HO/\OH OH and sucrose,94 stachyose and raffino~e,’~ and vitamin B, co-enzyme B1, and other corrinoids.” These workers have shown how Tl values are of immense value in assigning the carbon-13 resonances and in the study of internal molecular reorientation.” They propose three principle^'^ which should be considered when using PRFT spectra for resonance assignments in large and asymmetric molecules (a) Spin-lattice relaxation of protonated carbon atoms results almost exclusively from dipolar interactions with attached protons with Tl given by 1/T = NFt2yc2y,2rc,-6~f where N is the number of directly attached protons yc and yH are appropriate gyromagnetic ratios rCHis the C-H distance and T~ is the effective correlation time for rotational orientation ; (b) non-protonated carbon atoms invariably have greater Tl values than protonated carbon atoms ;(c) different carbon atoms within a molecule may not all have the same T,.differences arising from aniso- tropic effects or internal molecular orientation. An alternative method of measur-ing T values by a pulse method based on the progressive saturation technique has also been described” and the values obtained for 3,5-dimethylcyclohex-2-enone are in good agreement with the PRFT values. A recent review” of FT-n.m.r. is somewhat confusing when discussing multipulse techniques. A new method for measuring spin-spin relaxation times (T,) has been reported99 in which a selective 90” pulse is applied along the x-axis of the rotating frame followed by a continuous r.f.field applied along the y-axis for a period of seconds (t). During this period the magnetization is aligned along y and decays at a rate determined by the spin-lattice relaxation time (T,J in the rotating frame essen ially equal 96 A. Allerhand and D. Doddrell J. Amer. Chem. Soc. 1971 93 2777. ’’ D. Doddrell and A. Allerhand Proc. Nut. Acad. Sci. U.S.A. 1971 68 083; Chem. Comm. 1971 728. 98 D. Doddrell and A. Allerhand J. Amer. Chem. Soc. 1971 93 1558. 99 R. Freemann and H. D. W. Hill J. Chem. Phys. 1971 55 1985. I. H. Sadler to T2. Fourier transformation of the signal yields a conventional spectrum in which the line intensities A are given by 4= A exp(-t/T,,) where A is the initial intensity of the line.Integrated carbon-13 resonance spectra are rarely reported probably because of the uncertainty in the extent of the relative Overhauser enhancements. accurate integrals are possible for FT-n.m.r. carbon-13 spectra and it appearsg4 that for complex molecules e.g. cholesteryl chloride the Overhauser enhance- ments and therefore the integrated intensities are the same for nearly all carbon atoms. This is not necessarily so for small molecules. An intensity ratio of 3.5 :1 instead of 2 1 is obtained from the 13C n.m.r. spectrum"' of acetone and the relative intensities of vinyl acetate (32) are as shown. The Overhauser enhancement may be removed by adding very small quantities of a paramagnetic species (<0.05 moll-').Di-t-butyl nitroxide and the perchlorates"' of chro- mium(III) manganese(II) and iron(rr1) have been used effectively. At such low concentrations line-broadening effects are barely visible. The intensity loss is fully recovered by readjusting pulse intervals to take advantage of the faster relaxation. Following an earlier suggestion,lo2 it has been shown'03 that alternately pulsing broad-band proton decoupling power (1s) and carbon-13 power (30 p) to initiate free induction decay with a short delay time (0.5 s) leads to a signal improvement over the undecoupled spectrum by a factor of ca. three in the Fourier transform spectrum and true 3C-H coupling constants are obtained. In this respect the alternate-pulsed n.m.r.technique is superior to off-resonance or single-resonance decoupling methods which frequently distort lineshapes and coupling constants. Using this technique resonance assignments have been made and coupling constants determined for butan-1-01 and for dimethyl sul- phoxide. The value of noise-decoupled 13C n.m.r. for the study of conformational problems has been e~tablished,"~ carbon-13 shifts being extremely sensitive to the conformational environments of the nuclei. Problems common in 'H n.m.r. in determining linewidths and coalescence temperatures of the complex patterns arising from homonuclear spin-spin coupling are absent in 13Cn.m.r. ; loo G. N. La Mar J. Amer. Chem. SOC.,1971 93 1040. Iol R. Freeman K. G. R. Pachler and G. N.La Mar J. Chem. Phys. 1971 55 4586; G. N. La Mar Chem. Phys. Letters 1971 10 230. Io2 J. Feeney D. Shaw and D. J. S. Pauwells Chem. Comm. 1970 554. lo3 0.A. Gansow and W. Schittenhelm J. Amer. Chem. SOC.,1971 93 4294. D. K. Dalling and D. Grant J. Amer. Chem. SOC.,1967 89 6612; G. W. Buchanan and J. B. Stothers Canad. J. Chem. 1969,47 3605. Physical Methods-Part (ii) Nuclear Magnetic Resonance the chemical shifts between the alternative sites for a carbon atom in inter- converting conformations are generally much larger than those exhibited by protons and generally provide a wider range for temperature study in addition to raising the coalescence temperatures compared with those for proton reso- nances. There is often more than one equilibrating carbon atom in a molecule which can provide corroborative evidence and there are usually also non- equilibrating carbon atoms present which may be used to determine reference linewidths.A disadvantage of 13C n.m.r. compared with 'H n.m.r. is that many compounds e.g. cyclohexane because of symmetry possess no carbon atoms which equilibrate between two different environments. The '3C n.m.r. spectrum'05 of cis-decalin shows a sharp resonance owing to the carbon atoms at the ring junction which are in an identical environment in both forms. The remaining eight carbon atoms give rise to four lines at low temperatures which collapse to two lines under rapid kinetic averaging at high temperatures. Neither the low nor the high temperature methylene signals are as sharp as the ring junction resonances owing to residual kinetic effects.Activation parameters for this molecule 9-methyl-cis-decalin and 1,l- cis-1,2- trans-1,3- and cis-1,4- dimethylcyclohexane have been determined,' ' and are in good agreement with literature values determined from 'H n.m.r. studies where available. Resonances of the axial conformer of methylcyclohexane have been observed '06 for the first time by working at -110 "Cin a 59 kG magnetic field. An A value of 1.6kcal mo1-l obtained for the methyl group is in good agreement with the accepted value (1.7 kcal mol- '). A variable-temperature studylo7 of cyclo- nonane strongly supports a twist-boat-chair conformation (33) rather than the alternative twist-chair-boat form (34). The 250 MHz 'H n.m.r.spectra also support these results. The spectra of a large number of 1,3-dioxans have been rep~rted'~~*''~ and the chemical shifts of the ring and substituent carbon atoms correlate with the corresponding cyclohexane analogues if the deshielding effect of the ring oxygen atoms is taken into account for 5-axial substituents. In some cases O9 non-chair conformations are significant. Conformational and con- figurational assignments to twenty methyl and aryl glycosides have been made"' D. K. Dalling D. M. Grant and L. F. Johnson J. Amer. Chem. SOC.,1971 93 3678. F. A. L. Anet C. H. Bradley and G. W. Buchanan J. Amer. Chem. SOC.,1971,93,258. lo' F. A. L. Anet and J. J. Wagner J. Amer. Chem. SOC.,1971 93 5266. lo' A. J. Jones E. L. Eliel D.M. Grant M. C. Knoeber and W. F. Bailey J. Amer. Chem. SOC.,1971,93,4772. G. M. Kellie and F. G. Riddell Chem. Comm. 1971 1031. lo E. Breitmaier W. Voelter G. Jung and C. Tanzer Chem. Ber. 1971 104 1147. 36 I. H. Sadler on the basis of chemical shift data anomeric glycosides being readily distin- guished. The anomeric equilibria of ketoses in water have also been studied' '' by continuous wave and Fourier-transform methods demonstrating in this instance the superiority of the latter. Applications to the determination of stereoregularity in polymers have also been described. '* The feasibility of 13C n.m.r. for the study of carbonium structure has been ~urveyed"~ by Olah and his co-workers. Unlike 'H n.m.r. it is particularly suited to distinguishing between two rapidly equilibrating classical ions and the alternative non-classical ion by virtue of the marked sensitivity of carbon-13 chemical shifts to charge densities.Of general interest is the a~signment"~ of a symmetrical o-delocalized non-classical structure to the norbornyl cation (35). The 2-methyl and 2-ethyl derivatives show' l4 partial delocalization whereas the 2-benzyl derivative is virtually classical.' l4 The 1,2-dimethylnorbornyl cation (36) appears' to be a partially a-delocalized carbonium ion which undergoes a rapid 1,2-Wagner-Meerwein shift. Symmetrically bridged non-classical structures have been demonstrated' l6 for the bromonium ion (37) and phenonium ion (38). Ally1 cations appear' l7 to show very little direct 1,3-interaction.The chemical shifts of the atom bearing the charge in triaryl carbonium ions and aryl methyl carbonium ions correlate' '* well with c' substituent constants. Br 1+\ Q Me2C-CMe2 (37) H2C-CH2 (38) 111 D. Doddrell and A. Allerhand J. Amer. Chem. Soc. 1971 93 2779. 112 J. Schaefer Macromolecules 1971 4 98 105 107 110; A. Zambelli G. Gatti G. Sacchi W. 0.Crain and J. D. Roberts Macromokcules 1971 4 475. 113 G. A. Olah and A. M. White J. Amer. Chem. Soc. 1969 91 5801. 114 G. A. Olah A. M. White J. R. De Member A. Commeyras and C. Y. Lui J. Amer. Chem. Soc. 1970,.92 4627. 115 G. A. Olah J. R. De Member C. Y. Lui and R. D. Pugmire J. Amer. Chem. Soc. 1971,93 1442. 116 G. A. Olah and R. D. Porter J. Amer. Chem. SOC.,1971,93 6877.117 G. A. Olah P. R. Clifford Y. Halpern and R. G. Johanson J. Amer. Chem. Soc. 1971,93,4219. 118 G. A. Olah R. D. Porter and D. P. Kelley J. Amer. Chem. Soc. 1971 93 464; G. J. Ray R. J. Kurland and A. K. Colter Tetrahedron 1971 27 735. Physical Methods-Part (ii) Nuclear Magnetic Resonance 37 The carbon-13 methyl chemical shifts of acetate and methoxy substituents of several pyranoses have been measured' using the 1H-['3C] INDOR tech-nique i.e. observation of the intensity charge of a 13C-satellite peak in the 'H n.m.r. spectrum while sweeping the irradiation frequency through the range of the I3C n.m.r. spectrum. Correlations between carbon-1 3 chemical shifts and charge densities have been pointed out for aliphatic compounds,' 2o aza-indenes,12 benzene deriva- tives,' 22 benzimidazoles and purines,12j and thymines and ~racils.'~~ An em- pirical method for predicting carbon-13 chemical shifts has been presented by Mason125 who has shown that when corrected by subtraction ofa diamagnetic shielding contribution CT~, shifts may be obtained from a series of additive parameters.The value of odfor a particular carbon atom depends only on those atoms directly bonded to it. Corrected shifts for linear and branched acyclic alkanes and simple cycloalkanes (except cyclopropane) can be calculated by counting -25.4 p.p.m. for each a-carbon -8.4 for each P-carbon +1.5 for each y-carbon and -1.4 for each &carbon with -1.4 as a constant term for shifts measured from methane.The methylcyclohexane parameters vary with conformation. Additivity is also observed for polysubstitution by halogen alkoxy and aryl groups and in linked or fused aromatic molecules and simple olefins acetylenes and carbonium ions. Where conjugative interaction with multiple bonded substituents can occur the corrected shift may be reduced. A comprehensive study'26 of carbon-13 spectra of 50 norbornane and nor- bornene derivatives has appeared. The chemical shifts are approximately additive for similar compounds and can be used for structural assignments. The difference in resonance of em-and endo-methyl groups is attributed to different 1,4-non-bonded interactions between the carbon atoms and not to magnetic anisotropic effects. Carbon-1 3 and proton shift data for numerous acetylenes have been presented'" and analysed.It appears that in phenylacetylene ethyl ethynyl sulphide and triethynylphosphine and its oxide a charge shift occurs from the triple bond to the substituent whereas in alkynyl ethers and amines the direction is reversed. A relatively high shielding of carbon and phosphorus attached to the triple bond is attributed to a reinforcement of diamagnetic aniso- tropy caused by overlap of the z-systems of the triple bond and the substituent. ExaminationI2 of several N-nitrosoanilines and N-nitrosoamines shows that 'I9 R. Burton L. D. Hall and P. R. Steiner Canad. J. Chem. 1971 49 588. ''O P. Lazzeretti and F. Taddei Org. Magn. Resonance 1971 3 113. R. J. Pugmire M. J. Robins D. M. Grant and R.K. Robins J. Amer. Chem. SOC. 1971 93. 1887. P. Lazzeretti and F. Taddei Org. Magn. Resonance 1971 3 283; G. C. Levy G. L. Nelson and J. D. Cargioli Chem. Comm. 1972 506. R. J. Pugmire and D. M. Grant J. Amer. Chem. Soc. 1971 93 1880. lZ4 A. R. Tarpley and J. H. Goldstein J. Amer. Chem. Soc. 1971 93 3573. J. Mason J. Chem. SOC.(A),1971 1038. 126 E. Lippmaa T. Pehk J. Paasivivta N. Belikova and A. Plate Org. Magn. Resonance 1970 2 58 1. 12' D. Rosenberg and W. Drenth Tetrahedron 1971 27 3893. P. S. Pregosin and E. W. Randall Chem. Comm. 1971 399. 38 I.H. Sadler the carbon shifts depend upon the orientation of the nitroso oxygen atom; evidence is presented for the coplanarity of the two n-systems in N-methyl-N- nitrosoaniline. Several substituted allenes have been examined.129 These compounds are particularly suitable for I3C n.m.r. spectroscopy although not previously studied. For a given alkyl substituent a linear relationship holds between the shift of the central (/?)allene carbon atom and the number of these substituents. Ethyl methyl and s-alkyl substituents each contribute +4.8 +3.3 and +7(+2) p.p.m. respectively above the shift of the P-carbon atom. An approximately linear relationship holds between the shifts of the /?-carbon atom in monosubstituted allenes and the a-carbon atom in the corresponding p-substituted ethylenes. Other compounds examined include mono- di- and tri-substituted benzenes,' 30 polycyclic aromatic hydrocarbons and helicenes,' ' adamantane derivatives,' 32 bicycl0[2,2,2]0ctanes,'~~ cyclopentanes,' 34 cyclo-hexanes,' 3s and organophosphonates.'36 A concise review of carbon- 13 magnetic resonance has appeared. 4 Miscellaneous Studies The theory of nuclear spin-spin coupling and the calculation of coupling constants by molecular orbital methods have been discussed. '38 Systematic analyses have been pre~ented'~' for the AA'BB' and AA'BB'MX spin systems for I = 3 nuclei and expressions obtained for the transition frequencies and intensities which have maximum accuracy consistent with practicable use. Inconsistencies in earlier treatments of the AA'BB' system have been clarified. The 'H n.m.r. spectrum of naphthalene has been completely analysed 140 as an 8-spin system ; long-range couplings exist between all inter-ring pairs of protons.Analysis of the INDOR spectra of the AMX ABX A,X and related spin systems have been given,I4' together with the illustrations of the value of the INDOR technique in structure elucidation and the detection of hidden lines in complex 'H n.m.r. spectra. A method termed the 'Dihedral Angle Estimation by the Ratio Method' has been presented'42 for assigning dihedral angles to hydrogen atoms adjacent to a 129 R. Steur J. P. C. M. Van Dongen M. J. A. De Bie W. Drenth J. W. De Haan and L. J. M. Van de Ven Tetrahedron Letters 1971 3307. I3O M. Goh Y. Sasaki and M. Suzuki Chem. and Pharm. Bull. (Japan) 1971 19 2301. I3I R. H. Martin N. Defay and D. Zimmerman Tetrahedron Letters 1971 1871. T. Pehk E. Lippmaa V. V. Sevostjanova M.M. Krayuschkin and A. J. Tarasova Org. Magn. Resonance 1971 3 783. 133 G. E. Maciel and H. C. Dorn J. Amer. Chem. Soc. 1971,93 1268. *34 J. D. Roberts M. Christl and H. J. Reich J. Arner. Chem. Soc. 1971 93 3468. 135 T. Pehk and E. Lippmaa Org. Magn. Resonance 1971 3 783. 136 G. A. Gray J. Amer. Chem. Soc. 1971 93 2132. 13' E. W. Randall Chem. in Britain 1971 7 371. 13' J. N. Murrell Progr. N. M. R. Spectroscopy 1971 6 1. 13' J. J. Batterham and R. Bramley Org. Mugn. Resonunce 1971 3 83. I4O R. W. Crecely and J. H. Goldstein Org. Magn. Resonance 1970 2 613. I4I F. W. van Deursen Org. Magn. Resonance 1971 3 221. 142 K. N. Slessor and A. S. Tracey Canad. J. Chem. 1971,49 2874. Physical Met hods-Part (ii) Nuclear Magnetic Resonance methylene group by using accurate vicinal coupling constants (J,,J2) in a modified Karplus equation J + c -k cos’$ -J2 + c k2 COS’(O -$1) where k ,k are the Karplus constants 4 ,othe angles shown in the projection (39) and c another constant.The method is based on the assumption that although the Karplus constants vary their ratio k :k remains fixed equal to 1.0 (41,$2 < 42) or 0.9 < n/2 c $2). Using the theoretical value of 0.28 H (39) for c a complete solution for 4 and $2 is possible without knowledge of k,. Application of available data from the low-temperature spectrum of octadeuterio- cyclohexane143 without prior assumption of proton identities predicted a slightly flattened chair conformation with proton assignments identical with those originally proposed and also allowed calculation of the Karplus constants.Dihedral angles computed this way appear not to be influenced by the effects of ring strain and substituent electronegativity. The conformational preferences of several four- to six-membered ring systems were investigated. Lambe~t’~~ has examined the use of the R value as a measure of distortion in six-membered rings of the types (40H43) compared with cyclohexane. The n.m.r. spectrum of a -X-CH,-CH,-Y- fragment yields two coupling constants 3Jt,,,s,defined equal to 3Ja + Jee)and 3Jcis,equal to Jae. The ratio J,,,, Jcis,known as the R value is independent of the electronegativities of the attached atoms X and Y. R Values for flattened fragments are normally less than 1.9 and for puckered fragments greater than 2.2.Appli~ation’~~ of the Karplus equation relates the R value to the internal dihedral angle $ 143 E. W. Garbisch and M. G. Griffith J. Amer. Chem. SOC.,1968 90 6543. 144 J. Lambert Accounts Chem. Res. 1971 4 87. 145 J. B. Lambert and F. R. Koeng Org. Magn. Resonance 1971 3 389. 40 I. H. Sadler A quantitative assessment has been made'46 of the severe degree of flattening in various benzo-substituted five- and six-membered ring compounds. A study 147 of some specifically deuteriated t-butylcyclohexanes shows that an equatorial t-butyl group distorts the chair conformation of the ring and the chemical shifts of the ring protons differ from those in cyclohexane. In consequence the use of a t-butyl group as a holding group in conformational analysis is not always justified and previous doubts on the validity of much published work on con- formational preferences are well founded.The validity of the approximate equations (a) k = (n/$)Av and (b) k = (n/fi)d-) for calculation of rates of exchange in dynamic n.m.r. has been examined'48 by comparison of rates with those obtained by complete lineshape analysis. Although previously criticized 149 as unreliable equation (u) normally yields reliable estimates of rates at coalescence when nuclei are not spin-coupled. Equation (b) may be used for spin-coupled nuclei (AB system) but only if the chemical shift difference (Av) exceeds the coupling constant (J). It is important however to realize that errors of 10,25 and 100 in rate constants at 300 K produce errors in AG* of only 0.06,0.1,and 0.4 kcal mol-' respectively.A linewidth method has been reported' 50 for determining chemical exchange rates between two equally populated sites. The Gutowsky-Holm equation simplified by the assumption of a large spin-spin relaxation time was used to produce a family of curves for various Av values relating the rate of exchange to the linewidth at one-half the peak height corrected for both field inhomogeneity effects and couplings. For NN-dimethylacetamide and NN-dimethylcarbamoyl chloride this method yielded activation parameters in excellent agreement with those from total lineshape analyses. Examination of structurally related amides and thioamides indicated that their rotational barriers were not greatly different.Among the vast number of kinetic studies reported are the following confor- mational mobility in 18-annulene,' 4H-1,2-diazepines,'52 and cyclodecapen- taenes ; 53 nitrogen inversion in dibenzylmethylamine' 54 and N-chlorodibenzyl- amine ;'55 a degenerate valency isomerism' 56 in 7-acetyl-3-methylanthranil involving a 1,9-shift (44); tautomeric equilibria of 1-methyl-5-methylamino- tetrazoles'57 which exist predominantly in the amino-form (45); and the hin- dered rotation of t-butyl groups in 2-benzyl-2-chloro-3,3-dimethylbutane and 146 H. R. Buys Rec. Trav. chim. 1969 88 1003. 147 J. D. Menijnse H. Van Bekkum and B. M. Wepster Rec. Trav. chim. 1971 90 779. 148 D. Kost E. H. Carlson and M.Raban Chem. Comm. 1971 657. 14') G. Binsch Topics Stereochem. 1968 3 97. I5O K. C. Ramey D. J. Louick P. W. Whitehurst W. B. Wise R. Mukherjee and R. M. Moriaty Org. Magn. Resonance 1971 3 201. 151 J. M. Gilles J. F. M. Oth F. Sondheimer and E. P. Woo J. Chern. Soc. (B),1971,2177. U. Svanholm Acta Chem. Scand. 1971 25 640. S. Masamune K. Hojo K. Hojo G. Bigam and D. L. Rabenstein J. Amer. Chenr. Soc. 197 I 93 4966. '54 M. J. S. Dewar and W. B. Jennings J. Amer. Chem. Soc. 1971 93,401. W. B. Jennings and R. Spratt Chem. Cornrn. 1971 54. 156 K. Parry and C. W. Rees Chem. Comm. 1971 833. 15' G. Bianchi A. J. Boulton I. J. Fletcher and A. R. Katritzky J. Chem. Soc. (B) 1971. 2355. Physical Methods-Part (ii) Nuclear Magnetic Resonance 2-chloro-2,3,3-trimethylbutane,' 58 and t-butyldimethylamine.' 59 Many papers have been concerned with rotation about single bonds in halogenated alkanes and conformational equilibria in dioxans and saturated nitrogen heterocycles.The 'H n.m.r. spectra of carbonium ions continue to be studied widely. Observation of the spectrum of the deuteriated bicyclo[3,l,0]hex-3-en-2-yl cation (46) reveals16' a slow sigmatropic rearrangement not observable in the Me Me I L 7 To 'd Me Me Me 40 N-N il' /r( )" '+ HN N II Me Me D (45) (46) unlabelled species. Olah and Halpern have found lei' conditions for protonation of olefins in super-acids to give carbonium ions without simultaneous polymeri- zations. Unfortunately there appears to be no single general procedure applicable to all olefins ;various procedures are described.A method has been developed' 62 for the quantitative determination of the relative stabilities of aryl carbonium ions in which the equilibrium between two carbonium ions and their covalent precursors is observed directly by 'H n.m.r. Good agreement with e.m.f. methods was obtained for rneta- and para-substituted tritylcarbonium ions. Systematic differences obtained for 9-arylxanthyl derivatives where the leaving group is varied are attributed to steric interactions in the covalent precursors. Several groups of workers '63,164 have examined Meisenheimer complexes formed by the attack of alkoxide or hydroxide ion on aromatic molecules heavily substituted with electron-wit hdra wing groups.2,4,6-Trisu bsti tu ted anisoles (47)-(49) appear'63 to give short-lived 1,3- and 1,5-complexes which rearrange to the thermodynamically more stable 1,l-complexes. The n.m.r. spectra of cumyl benzyl and a-methylbenzyl anions' 65 and of the dianions of tetraphenyl-158 J. E. Anderson and H. Pearson. J. Clwm. SOC.(B),1971 1209. 159 C. H. Bushwaller J. W. O'Neil and H. S. Bilofsky Tetrahedron 1971 27 5761. I60 P. Vogel M. Saunders N. M. Hasty and J. A. Berson J. Amer. Chem. SOC.,1971 93 1551. 161 G. A. Olah and Y. Halpern J. Org. Chem. 1971 36 2354. 162 J. V. McKinley J. W. Rakshys A. E. Young and H. H. Freedman J. Amer. Chern. SOL.. 1971 93 4715. lh3 M. R. Crampton. M. A. F1 Gharian and H. A. Khan Chem. Comm. 1971 834; F.Terrier J. C. Hallc M. P. Sirnounin and M. J. Lecourt Org. Magrz. Resonunce 1971 3. 361. I.H. Sadler OMe Me0 OMe L. "rxr"", XQm2 MeO-) x@02 OMe '..* .J \ * H Y Y Y (47) X = Y = NO (48) X = C1 CF or CN; Y = NO (49) X = NO,; Y = C1 or CO,Me ethylene'66and [12]annulene' 67 have been observed with alkali-metal counter- ions in THF. A useful modification has been described'68 for an HAlOO spectrometer to allow wide sweep ranges (up to 20 khz) to be used and homonuclear INDOR experiments to be carried out in the field-frequency locked mode. A method of correcting base-line distortion and a phase shift network are given. Modifica- tions for heteronuclear spin decoupling' 69 and homonuclear 'H-[' HI and hetero- nuclear 'H-['3C] INDOR experiments have also been described.' 70 5 Reviews and Books The following aspects of n.m.r.have been reviewed nuclear magnetic double resonance including INDOR ;' ' use of the intramolecular nuclear Overhauser effect in the assignment of organic structures studies in liquid crystals ;173 nitrogen magnetic resonance,' 74 the study of hindered rotation and inversion in and molecular relaxation mechanisms in polymers ;'76 and application to the study of polymer config~rations.'~~ Reviews have also ap- peared concerning n.m.r. studies of amides' 78 and paramagnetic complexes.' 79 Two books devoted to spectral analysis,18' and single volumes on fluorine chemical shifts' 81 and n.m.r. studies of polymers' 82 have appeared.lh4 E. J. Fendler W. Ernsberger and J. H. Fendler J. Org. Chem. 1971 36 2533. Ib5 K. Takakashi M. Takaki and R. Asami Org. Magn. Resonance 1971 3 539. K. Takakashi Y. Inoue and R. Asami Org. Magn. Resonance 1971 3 349. 16' J. F. M. 0th and G. Schroeder J. Chem. Soc. (B) 1971,904. 168 P. N. Jenkins and L. Phillips J. Phys. (E) 1971 4 530. R. Burton and L. D. Hall Canad. J. Chem. 1970,48 59. R. Burton L. D. Hall and P. R. Steiner Canad. J. Chem. 1970 48 2679. 171 W. von Philipsborn Angew. Chem. Internat. Edn. 1971 10 472. 17* G. E. Bachers and T. Schaefer Chem. Rev. 1971 71 617. 173 S. Meiboom and L. C. Snyder Accounts Chem. Res. 1971 4 81. 174 E. W. Randall and D. G. Gilles Progr. N. M. R. Spectroscopy 1971 6 119. 175 H. Kessler Angew.Chem. Internat. Edn. 1970 9 219. L76 D. M. McCall Accounts Chem. Res. 1971 4 223. 17' H. Cheradame Bull. Soc. chim. France 1971 2023. 78 W. E. Stewart and T. H. Siddall Chem. Rev. 1970 70 5 17. 179 K. Schwarzhaus Angew. Chem. Internat. Edn. 1970 9 946. R. A. Hoffman S. Forsen and B. Gestbloom 'NMR Basic Principles and Progress,' Springer-Verlag Berlin 1971 vol. 5; A. J. Abraham 'Analysis of High Resolution NMR,' Elsevier Amsterdam 1971. J. W. Emsley and L. Phillips 'Progress in N.M.R. Spectroscopy,' Pergarnon Oxford 1971 vol. 7. P. Diehl E. Fluck and R. Kosfeld 'NMR Basic Principles and Progress,' Springer- Verlag Berlin 1971 vol. 4.