2 Physical Methods and Techniques Part (i) Mass Spectrometry By T. P. TOUBE Chemistry Department. Queen Maty College London. El 4NS The literature on organic mass spectrometry continues to proliferate. Several books in the field have appeared during the period reviewed in this Report notably ones on metastable ions,’ gas chromatography-mass spectroscopy,2 pesticides and pollutant^,^ and solvents and imp~rities.~ Tables of metastable ions’ of m/e 1-500 and a four-volume registry of mass spectral data6 have also been published. Selected areas have been the subject of reviews among them structure determinati~n,~ steroids,* indoles,’ and Hammett-type correlations. lo The relationship between mass spectrometric thermolytic and pyrolytic re- activity has been reviewed.’ ’ An extremely compact and comprehensive survey of recent literature (1647 references in 25 pages!)12 has appeared.In this Report an attempt has been made to select those topics which the Re- porter believes will be of greatest interest to the majority of organic chemists. R. G. Cooks J. H. Beynon R. M. Caprioli and G. R. Lester ‘Metastable Ions’ Elsevier Amsterdam 1973. * W. McFadden ‘Techniques of Combined Gas Chromatography-Mass Spectroscopy’ Wiley-Interscience New York 1973. S. Safe and 0.Hutzinger ‘Mass Spectrometry of Pesticides and Pollutants’ CRC Press Cleveland 1973. M. Spiteller and G. Spiteller ‘Massenspektrumsammlung von Losungsmitteln Verun- reinigungen Saulen belegmaterialen und einfache aliphatische Verbindungen’ Springer- Verlag Vienna 1973.’ R. Neeter and C. W. F. Kort ‘Metastable Precursor Ions’ Elsevier New York 1973. E. Stenhagen S. Abrahamsson and F. W. McLafferty ‘Registry of Mass Spectral Data’ Wiley-Interscience New York 1974. J. M. Wilson in ‘Structure Determination in Organic Chemistry’ ed. W. D. Ollis MTP International Review of Science Organic Chemistry Series One Vol. 1 Butter-worth London 1973. * ‘Modern Methods of Steroid Analysis’ ed. E. Heftmann Academic Press New York 1973. R. Kmel’nitskii Khim. geterotsikl. Soedinenii 1974 291. lo M. M. Bursey in ‘Advances in Linear Free Energy Relationships’ ed. N. B. Chapman and J. Shorter Plenum London 1973. I ’ R. C. Dougherty Fortschr. chem. Forsch. 1974 45 93. A. L. Burlingame R.E. Cox and P.J. Derrick Analyr. Chem. 1974 46 248R. 7 T. P.Toube 1 Metastable Ions Interest in the properties and behaviour of metastable ions' has been stimulated by the introduction of MIKES (mass analysed ion kinetic energy spectroscopy)' or DAD1 (direct analysis of daughter ions).I4 These techniques provide methods for the study of ions decomposing between the source and collector regions of a mass spectrometer which are more selective than those previously available. Studies on metastable ions for the process (2)-(3) in the mass spectrum of (1)" showed that the formation of (3)involved the transfer of hydrogen only from PhNHCO(CH,),CHJ -+ ,OH1:PhNHC -+ C,H,Nf \CH (1) (2) (3) the CH group of (2). Deuterium labelling suggested that the bulk of the trans- ferred hydrogen came from the a-carbon atom of (l) but the absence of a meta- stable peak for deuterium transfer in the second stage of the reaction implied that only ions of comparatively high internal energy transfer the relevant hydrogen atom.Allylic alcohols are known16 to give spectra similar to those of the correspond- ing ketones after electron impact (EI) at 70eV. Investigation by metastable techniques of those C4H80t ions from hept-1-en-3-01 which underwent de- composition after EI indicatedI7 that they probably had structure (4) or (5). These results confirmed conclusions' from ion cyclotron resonance (ICR)which showed that the C4H,0t ion in the mass spectrum of this compound was not derived oia ketonization of the molecular ion.Using MIKES Beynon" has been able to measure kinetic isotope effects and the effect of isotopic substitution on the release of kinetic energy for the loss of methane from the molecular ion of butane using the I3C1and 2H,isotope peaks at natural abundance. Such data can normally be derived only by examination of labelled compounds. l3 J. H. Beynon R. G. Cooks J. W. Amy W. E. Baitinger and T. Y. Ridley Analyt. Chem. 1973,45 1023A. I4 K. H. Mauer C. Brunee G. Kappus K. Habfest U. Schroder and P. Schulze 19th Conference of Mass Spectroscopy Atlanta Georgia 1971. l5 K. B. Tomer and C. Djerassi Tetrahedron 1974 30 17. M. Kraft and G. Spiteller Org. Mass spectrometry 1968 1 617. l7 B. Grant and C. Djerassi J. Amer. Chem. SOC.,1974 96,3477. J.H. Beynon D. F. Brothers and R. G. Cooks Analyt. Chem. 1974. 46 1299. Physical Methods and Techniques-Part (i) Mass Spectrometry 9 The 1,2-elimination of H from ethane CH2=OH+ CH,=NH,+ and similar species if it occurs uia a planar transition state constitutes a 'symmetry- forbidden' process. Measurements from metastable ion characteristics of the kinetic energy released in the known 1,Zelimination of H from these molecules s~ggested'~ that the energy penalty for violation of the Woodward-Hoffmann rules appears as excess kinetic energy (ca. 1 eV in the cases studied). This con- clusion was then extended2' to the examination of the loss of H2 in species in which scrambling precluded investigation by 2H labelling. The absence of kinetic energy release for the loss of H2 from C6H7+ C2H5+ and C2H4 ions was interpreted as indicating a possible 'symmetry-permitted' 1,l-elimination of H, while for C7H9+ kinetic energy release supported a mechanism involving 1,2-elimination.A 1,l-elimination in this case of HNO ,has also been reported" from the molecular ion of Me02C(CH2),N02. Collisionally activated (CA) metastable ions have been studied using spectro- meters in which as for MIKES the magnetic analyser precedes the electrostatic analyser. Collision with an inert gas induces decomposition in previously stable ions. Field-ionized (FI) molecules have been studied by this technique using helium at ca. lop4Torr as the activating species. As would be expected the ions behaved differently from species produced in EI sources as FI yields ions whose internal energy is low.It was observed that collisional activation itself appeared to promote a certain amount of isomerization in these cases. A CA of some C2H6N + and C,H,N + ions established that only ions possessing apparently immonium structures had lifetimes greater than 10-s. Among the isomeric C2H6N+ ions it was observed that on actiuation NH2CH2CH2+ and CH3CH2NH+ isomerized to a common CH,CH=NH,+ structure which + + differed from the CH,NH=CH species formed by rearrangement of CH,NCH . The subsequent decomposition of these immonium ions proceeded without significant hydrogen scrambling. Negative-ion CA studies have been ~ndertaken.~~ Collision of the stable negative molecular ions of carboxylic acids2 with toluene induced decomposition pathways based on the acid function which were lacking in the conventional spectra.The ejection of ethane from the [M -C1]+ ion from exo-norbornyl chloride2" was shown by isotopic labelling to proceed with statistical loss of hydrogen27 and carbon.26 The 'metastable peak'26 for the process could be resolved into a l9 D. H. Williams and G. Hvistendahl J. Amer. Chem. Soc. 1974 96,6753. 'O D. H. Williams and G. Hvistendahl J. Amer. Chem. Soc. 1974 96,6755. *l T. A. Molenaar-Langeveld and N. M. M. Nibbering Org. Mass Spectrometry 19 4. 9 257. 22 K. Levsen and H. D. Beckey Org. Mass Spectrometry 1974 9 570. 23 K. Levsen and F. W. McLafferty J. Amer. Chem. Soc. 1974 96,139. 24 J. H. Bowie and S.G. Hart Internat. J. Mass Spectrometry ion Phys. 1974 13 319 25 J. H. Bowie Org. Mass Spectrometry 1974 9 304. 2b J. L. Holmes D. McGillivray and N. S Isaacs Org. Mass Spectrometry 1974 9 5 0. " J. L. Holmes D. McGillivray and N. S. Isaacs Canad. J. Chem. 1970 48 2791. 10 T. P.Toube superimposition of 'dished' and 'Gaussian' shapes suggesting that two isomeric ions differing in heat of formation by ca 0.4 eV decomposed in this fashion. It is reasonable to suppose that the kinetic energy released in the unimolecular decomposition of an ion should not exceed the sum of the reverse activation energy for the reaction and the excess energy of the ion above the threshold for the process. The reverse activation energy for the process C3H,+ -+ C3H5++ H, as estimated from the appearance potentials of C3H ions produced from + thermal C3H radicals has been compared** with the kinetic energy release for the reaction measured from the 'dished metastable peak'.Thermochemical calculations showed that nearly all the reverse activation energy appeared as a loss of kinetic energy. This result implies that maximum kinetic energy release cannot alwayhbe equated with the above sum. h 2 The C7H7+Ion Since the original proposal2' that the C,H,+ ion from toluene was the ring- expanded tropylium ion (6) rather than a benzyl cation (7) the structures of the C,H ions from a large variety of sources have been extensively studied. The + problem is of particular importance in view of the gay abandon with which ring-expanded structures for all sorts of bicyclic and heterocyclic species have appeared in the literature since Meyerson's first suggestion sometimes on no firmer basis than tenuous analogy or aesthetic preference! An extensive study by McLafferty3' of C7H7+ ions from a number of ortho- meta- para- and side- chain-substituted toluenes substituted cycloheptatrienes and substituted nor- bornadienes has used CA methods to attempt to distinguish between the structures (6)-(11).The technique has particular merit as it permits the study of the (6) (7) (8) (9) (10) (1 1) fragmentation of C,H,+ ions which would not have had enough energy to decompose were they not activated by collision and the manner in which the induced fragmentations occur is not dependent on the internal energies of the ions at formation (as collisions lead to a Boltzmann distribution of these energies) but only on their structures.The relative abundances of fragment ions indicated for example that a significant proportion of C,H,+ ions from cycloheptatriene derivatives had the benzyl structure (7). In fact the ratio (6) :(7) did not vary very greatly irrespective of whether the starting material was a cycloheptatriene or a toluene. Calculated energy barriers suggested that the equilibration between these skeleta was easier at the molecular ion stage than in the C7H,+ fragment 28 J. L. Holmes and G. M. Weese Org. Mass Spectrometry 1974 9 618. 29 P. N. Rylander S. Meyerson and H. M. Grubb J. Amer. Chem. SOC. 1957 79 842. 30 F.W. McLafferty and J. Winkler J. Amer. Chem. SOC.,1974 96,5182. Physical Methods and Techniques-Part (i) Mass Spectrometry 11 itself. The tolyl ions (8)-( 10) appeared to be distinguishable from each other and the relative ease with which they interconverted with one another and with (6) and (7) was examined. The C7H7+ ions from norbornadiene derivatives remained largely norbornadienyl (11) and showed less tendency to isomerize to (6)or (7) than did the aromatic species. Differences in the kinetic energy released in the process C7H7+ --* C5HS+ previously reported31 for the metastable decomposition of C7H ions from different sources were reinterpreted3' on the + basis of the CA data as reflecting differences in internal energy distributions in the ions rather than variations in structure.E~amination~~ of the C,H,+ problem on the different time and energy scales of ICR produced the interesting result that whereas the C7H7+ ions from benzyl methyl ether and 7-methylcycloheptatriene reacted with dimethylamine no such reaction was observed with the corresponding ion from 7-methoxycyclohepta- triene. These data would normally be regarded as evidence for a structural difference between the reacting and non-reacting species. Isotopic scrambling results33 on a 13C,-labelled C7H,+ fragment from 6-hepta-2,4-dien-6-ynyl chloride have been interpreted as indicating that a proportion of such ions from an open-chain precursor isomerize to a tropylium structure (6). In view of McLafferty's observations3' such conclusions should not be drawn without great circumspection.Similar caution is necessary in dealing with data reported34 on isotope scrambling in the C,H,+ ion from [ 3C,Jtropylium iodide. More cautious conclusions were drawn from ion kinetic energy spectroscopy (IKES) on the [M-11' ions from benzyl fluoride and the isomeric monofluorotoluenes :35 the differences in the behaviour of the C7H6F+ ions were ascribed to variations in their internal energies and/or structures. The assumption that C,H,+ ions have a tropylium structure is implicit in a of stable C13H1 ions from a variety of aromatic compounds. Heats of + formation of these ions were calculated (AH for each starting material being derived from 'group equivalent' values). On the basis of their magnitudes it was concluded that diphenylmethane and p-phenyltoluene gave rise to phenyltropy- lium ions but that the other compounds examined produced C 3H + species which had partially acyclic structures.A benzyl cation figured in a different role in a study3 of the loss of a benzyl radical from the molecular ion of a,o-dibenzyloxyalkanes. The magnitude of a deuterium isotope effect and results from l80labelling were consistent with a 31 R. G.Cooks,J. H. Beynon M. Bertrand and M. K. Hoffmann Ore. Mass Spectrometry 1973. 7. 1303. 32 A. Venema and N. M. M. Nibbering Tetrahedron Letrers 1974 3013. 33 C. Koppel. H. Schwarz and F. Bohlmann Org. Mass Spectrometry 1974,9 332. 34 A. Siegel J. Amer. Chem. Soc. 1974 96 1251. 35 S. Safe W.D. Jamieson and D. J. Embree Canad. J. Chem. 1974 52 867. 36 F. Bohlmann C. Koppel B. Miiller H. Schwarz and P. Weyerstahl Tetrahedron 1974 30,1011. " A. P. Bruins and N. M. M. Nibbering Tetrahedron 1974 30,499. 12 T. P.Toube mechanism which involved an S,i-type intramolecular migration of a benzyl cation. 3 Fragmentation Processes The effects of interaction between functional groups in polyfunctional alkanes on their mass spectra have been reviewed.38 Several examples of such intra- molecular effects have recently appeared in the literature. The base peak [M-PhC,H,] ,in the spectrum of N-acetyl-N-(4-phenyl)butylamine + has been shown3' to arise with the transfer of a benzylic hydrogen atom to nitrogen. Analogous processes are absent in the spectra of N-acetyl-N-butylamine and butylbenzme.It has been suggested that interactions between functional groups may depend less on their nature than on their relative positions. This proposal was made on the basis of the observation that the fragrnentation40 of dimethyl glutamate MeO,C(CH,),CH(NH,)CO,Me and of some of its derivatives includes some processes found in compounds containing a -N(CH2)3N- unit. On the other hand measurement4' of the appearance potentials of [M -Me]+ ions from bistrimethylsilyl ethers separated by 2-7 methylene groups gave values independent of the separation of the functional groups indicating that in this series no interaction between the oxygen functions was apparent. In the chemical ionization (CI) spectra (isobutane reagent gas) of cis-and trans-cyclohexanediols steric effects have been investigated42 The trans- 1,3- and 1,Cisomers behaved like monohydric alcohols preferring to form [M -HI+ ions while the cis-compound yielded [M + HI+ species presumably by forma- tion of a hydrogen bridge.The cis-and trans-1,2-diols formed such bridged ions equally well but possibly because of the greater strain involved this process was less favoured in the cyclopentane-1,2-dioIs. Relationships derived from condensed-phase reactions involving hydrogen stereochemistry have been extended to the study43 of the EI-induced 1,Zloss of acetic acid froucetates and the 1,4-loss of water from alcohols. By deuterium labelling it was shown that dehydration was stereoselective.The expulsion of acetic acid had the same trans,antipreference and deuterium isotope effect as the corresponding pyrolysis reaction suggesting a common structure for the transi- tion states for the two processes. The effect of steric hindrance on fragmentations of the type (12)-+ (13) has been in~estigated.4~ With an electron-donating group in the para-position of the 38 H. Bosshardt and M. Hesse Angew. Chem. 1974 86 256. 39 R. Wild and M. Hesse Helv. Chim. Acta 1974 57 452. 40 E. Lerch and M. Hesse Helv. Chim. Acta 1974 57 1584. 4' C. Koppel H. Schwarz and F. Bohlmann Org. Mass Spectrometry 1974 9 567. 42 J. Winkler and F. W. McLafferty Tetrahedron 1974 30 2971. 43 M. M. Green J. M. Moldowan and J. G. McGrew J. Org. Chern. 1974 39 2166. 44 H.Kuschel and H.-F. Grutzmacher Org. Mass Spectrometry 1974 9 403. Physical Methods and Techniques-Part (i) Mass Spectrometry 13 benzene ring this process was suppressed. However if the electron-donating substituent was an aryl group chosen to produce sufficient steric congestion for conjugation between the two aromatic rings to be prevented the process involving loss of the ortho hydrogen atom was once more observed. Steric effects have also been adduced45 to support the mechanism usually proposed for the losses of toluene and xylene from the molecular ions of carote- noids uia intermediates such as (14).46 The bulky end-groups of the unusual R2 C, carotenoids would crowd the cyclobutane ring in (14). The [M-92]/ [M -1061ratios in such Cs0 compounds were measured and found to have been altered in the sense which would have been predicted.The contribution of the stability of the radical lost in the formation of species which can be written as five- and six-membered cyclic bromonium ions from a variety of bromides has been assessed.47 For the loss of PhCH ,thermochemical criteria prevail. With most other radicals the reaction proceeds under kinetic control. An attempt to correlate the relative abundances of fragment ions from aryl benzoates with Hammett Q values failed.48 The lack of linearity in the plot was ascribed to some form of oxygen bridging in the excited state. The expulsion of a carbene unit from an ion in the mass spectrometer has commonly been regarded on energetic grounds as an improbable process.However several examples of this fragmentation have recently been reported. The molecular ion of perfluoropropene decomposes49 to give a C,F4t ion a formal loss of :CF . A variety of substituted trifluoromethylbenzenes have been + shown to lose :CF .” The iminosulphurane Me,SNHCO,Et C1-gives an [A4-C1]+ ion which then expels :CH .51 In the fragmentation of the corre- sponding diethyl compound it was proposed that the ejection of :CHCH occurred. Such loss of a carbene from a sulphonium ylide has previously been 45 W. Francis S. NorgBrd and S. Liaaen-Jensen Acta Chem. Scand. 1974 B28,244. “ E.g. U. Schweiter G. Englert N. Rigassi and W. Vetter Pure Appl. Chem. 1969 20 365. 47 H. Schwarz F. Bohlmann G. Hillebrand and G.Altnau Org. Mass Spectrometry 1974 9 707. 48 S. A. Shamshurina 0. S. Chizhov and B. M. Zolotarev Zhur. org. Khim. 1974 10 913. 49 B. S. Freiser and J. L. Beauchamp J. Amer. Chem. SOC.,1974 96 6260. 50 G. P. Vdovin Y. P. Egorov A. P. Krasnoshchek and L. M. Yagupol’skii Zhur. org. Khim. 1974 10 1355. 51 G. F. Whitfield H. S. Beilan D. Saika and D. Swern J. Org. Chem. 1974 39 2148. T. P.Toube po~tulated.~’Examination of the CA spectra of PhCH,CH2Br at low energies (11-15 eV) suggested53 that the C8H9+ ion varied in structure with energy. Carbene loss from this ion was observed. Energetic considerations would also be expected to discriminate against processes in which an even-electron species loses a radical to yield an odd- electron ion.Such fragmentations are not unknown in mass spectrometry. A particularly amusing example of this reaction however is provided by octaethyl- porphorin (15),54 which in the mass spectrometer exhibited the successive expulsion of all eight methyl radicals. Labelling with 13C revealed that the methyl radical lost from the molecular ion ofp-methyl~tyrene~’ arose from the positions shown in (16).No loss ofcarbon from the ring was observed. 20.1 % f/32.2 % 47.7 % (16) (17) Labelling results and metastable data for isobutenes6 led to a postulated mechanism for scrambling involving the interconversion of methylpropene but-1-ene and but-Zene structures via opening and closing of a methylcyclo- propane intermediate. Selective deuteriation of indolo[2,3-~]quinolizidine (17) revealed” that the [M -13 ion arose by expulsion of H-12b (37 %) H-7 (32 ”/,) H-6 (19”/,),and H-4 (13%).The ejection of ethane from the molecular ions of such compounds is well known. From the spectra of the labelled derivatives it was concluded that loss of 28a.m.u. occurred by a retro-Diels-Alder type of 52 P. Robson P. R. H. Speakman and D. G. Stewart J. Chem. SOC.(C) 1968 2180. 53 N. M. M. Nibbering T. Nishishita C. C. Van de Sande and F. W. McLafferty J. Amer. Chem. SOC.,1974 % 5668. 54 P. S. Clezy C. L. Lim and J. S. Shannon Austral. J. Chem. 1974 27 1103. 55 C. Koppel H. Schwarz and F. Bohlmann Org. Mass Spectrometry 1974 9 343. 56 M. S.-H. Liu and A. G. Harrison Canad. J. Chem. 1974,52 1813.57 G. W. Gribble and R. B. Nelson J. Org. Chem. 1974 39 1845. Physical Methods and Techniques-Part (i) Mass Spectrometry 15 pathway (loss of C-2 + C-3) to the extent of 207$ while loss of C-1 + C-2 yielded 80 % of the [M-281 ions. 4 Ion Cyclotron Resonance Vapour-phase reactions have been studied by ICR yielding results related to reactions in solution. On electron impact followed by collision of the ions with their parent molecules esters produced [M + H]+ ions.58 Reaction of these species with molecules of added alcohols led to an ion formed with the ejection of water by a process analogous to esterification. The trifluoroacetoxyl anion produced by EI was allowed to react with trifluoroacetic anhydride molecule^.^^ On the time-scale 1O-j s this yielded a 1 1 adduct whose structure was deduced to resemble (18) a species typical of solution chemistry.On the other 0-I FC,-C-OCOCF, I OCOCF (18) hand in solution esters tend to protonate at the carbonyl group; deuterium labelling in conjunction with ICR has shown6' that in the CI spectrum of esters CH5+ protonates the ether oxygen not the carbonyl. The basicities of mono- di- and tri-methylphosphines have been measured' by ICR by observation of their proton affinities. From these data P-H bond strengths have been estimated. From kinetic studies62 rates for some vapour-phase ion-molecule reactions of t-butyl alcohol have been determined. 5 Miscellaneous Applications An unusual correlation of properties was reported in a study of some aromatic dyes,63 PhN=NAr in which it was shown that the ratio [Ph+]/[PhN,+] in their EI mass spectra increased in the same way as the light-fastness of their 1% dyeings on cellulose acetate or polyester fabric.The volatile products of pyrolysis of some 70 compounds have been examined by mass spe~trometry.~~ The pyrolysis patterns could be used to distinguish the parent compounds. 58 J. K. Pau J. K. Kim and M. J. Caserio J.C.S. Chem. Comm. 1974 120. 59 J. H. Bowie and B. D. Williams Austral. J. Chem. 1974 27 1923. 6o C. V. Pesheck and S. E. Buttrill J. Amer. Chem. SOC. 1974 96,6027. 61 R. H. Staley and J. L. Beauchamp J. Amer. Chem. SOC. 1974 96,6252. 62 J. L. Beauchamp M. J. Caserio and T. B. McMahon J. Amer. Chem. SOC.1974 % 6243. 63 H. P. Mehta and A. T. Peters Appl. Spectroscopy 1974 23 241. 64 R. Belcher G. Ingram and J. R. Majer Microchem. J. 1974 19 191. T. P.Toube Substrates with two or more oxygen atoms or with nitrogen functions re- acted6' with added ethanolamine or ethylenediamine introduced via the liquid inlet system of a CI mass spectrometer. This procedure would provide a means of imparting additional reagent gas selectivity in CI mass spectrometry without the need to change gas cylinders! The kinetics of some unimolecular reactions of field-ionized species have been Field desorption (FD)techniques have been used to obtain molecular weights for compounds which fail to give molecular ions under other conditions. A representative recent example of this class is the production of a significant molecular ion of an unstable ~yloside~~ by FD.Benoit6* has re-examined some ionization potentials by the authors of last year's Report7' and compared the results with values from his own earlier Not surprisingly he comes to the conclusion that his own figures were the more reliable! The area of computer-aided interpretation remains a popular one. A Japanese computer program has been devised73 which is claimed to generate the structure most likely to give rise to a given mass spectrum even for compounds not stored as references. 65 D. V. Bowen and F. H. Field Org. Mass Spectrometry 1974 9 195. 66 P. J. Derrick and A. L. Burlingarne Accounts Cliem. Res. 1974 7 328. 67 J. J. Karchesy M.L. Lauer D. F. Barofsky and E. Barofsky J.C.S. Chem. Comm. 1974 649. 68 F. Benoit Org. Mass Spectrometry 1974 9 626. " T. W. Bentley and R. A. W. Johnstone J. Chem. SOC. (B),1971 263. '' R. A. W. Johnstone and F. A. Mellon J.C.S. Faraday II 1973,69 36. 71 R. A. W. Johnstone and F. A. Mellon Ann. Reports (B),1973 70 7. l2 F. Benoit Org. Mass Spectrometry 1972 6 1289. 73 F. Erni J. T. Clerc and S. Hishida Kuguku No Ryoiki 1974 71 101 1.