摘要:

CHEMICAL SOCIETY REVIEWS VOLUME 17,1988 0 Copyright 1989 LONDON THE ROYAL SOCIETY OF CHEMISTRY CONTENTS PAGE THE WITTIG OLEFINATION COMPOUNDSREACTIONWITH CARBONYL OTHER THAN ALDEHYDESAND KETONES. By P. J. Murphy and J. Brennan i SPECTROSCOPY REACTIONS.TILDEN LECTURE. OVERTONE AND UNIMOLECULAK By M. S. Child 31 SPECTROSCOPIC OF OPFN-SHELLPOLYATOMICSTRUCTURE CATIONS.By John P. Maier 45 MACROCYCLIC OF LANTHANIDESSCHIFF BASE COMPLEXES AND ACTINIDES. By D. E. Fenton and P. A. Vigato 69 THE DAKIN-WEST REACTION.By G. L. Buchanan 91 ROBERT ROBINSON LECTURE. RETROSYNTHETIC ANDTHINKING-ESSENTIALS EXAMPLES. By E. J. Corey 111 EDUCATIONNYHOLM LECTURE. NEW TRENDS IN CHEMICAL AND CHEMISTRY TEACHER WORLDWIDE.EDUCATION By Marjorie Gardner 135 A CONFORMATIONAL OF TRANSITION STERICANALYSIS METALq '-ACYL COMPLEXES: AND STEREOELECTRONICINTERACTIONS EFFECTS.By Brent K. Blackburn, Stephen G. Davies, Kevin H. Sutton, and Mark Whittaker 147 OF MOLTEN ACETAMIDE COMPLEXES.THE CHEMISTRY AND ACETAMIDE By D. H. Kerridge 181 REARRANGEMENTS.VINYLCYCLOPROPANE By Z. Goldschmidt and B. Crammer 229 MELDOLA MEDAL LECTURE. ORGANOTRANSITION IN-METAL COMPLEXES CORPORATING BISMUTH. By N. C. Norman 269 REACTIVITYOF SUBSTITUTED ALIPHATIC NITRO-COMPOUNDS WITH NUCLEO-PHILES. By W. R. Bowman 283 VOLTAMMETRIC STUDIES OF ION TRANSFER MEMBRANES.ACROSS MODEL BIOLOGICAL By Christopher J. Bender 317 BIOMOLECULAR By Stephen Mason 347HOMOCHIRALITY. RADICAL REACTIONS SIMONSEN LECTURE. COBALT-MEDIATED IN ORGANIC SYNTHESIS. By Gerald Pattenden 36 1 OF SECONDARY IN MICROORGANISMS.ENZYMES METABOLISM By J. A. Robinson 383 METALLOCENES REACTIONINTERMEDIATES.AS By Peter Grebenik, Roger Grinter, and Robin N. Perutz 453 1988 Indexes 49 1

ISSN:0306-0012    

DOI:10.1039/CS98817FP001

出版商:RSC    

年代:1988

数据来源: RSC

摘要:

Cheni. SOC.Rev., 1988, 17, 31-44 TILDEN LECTURE* Overtone Spectroscopy and Unimolecular Reactions By M. S. Childt DEPARTMENT OF CHEMISTRY AND BIOCHEMISTRY, UNIVERSITY OF COLORADO, BOULDER, CO 80309, USA 1 Introduction Traditional studies of intramolecular dynamics have concentrated either on the low energy (up to 4 000 cm-' or 50 kJ mol-') normal coordinate vibrational spectrum' or on the kinetics of unimolecular diss~ciation~-~ at typically 20G400 kJ mol-' or 20 000-30 000 cm-'. The former implies that the energy is permanently locked into one normal mode or the other, while the latter are known to be consistent with the RRKM3-4 picture of free internal energy flow between all degrees of freedom. The interesting intermediate energy range' where one type of behaviour goes over to the other has until recently defied investigation, due primarily to the selection rule Ac = 1 governing strong vibrational transition^.^.^ Various recent spectroscopic advances have begun to overcome this difficulty.One is to employ fourier transform or laser which have the sensitivity to detect up to Ac = 9 transitions in favourable circumstances. Another is to access specific high vibrational levels of the ground state So by energy selective laser excitation to a higher electronic state S1which couples back to So by internal conversion.' 5-1 The internal states of any decomposition fragments can then be detected by laser-induced fluorescence.16-' A third approach is to employ pico- second flash photolysis methods to follow the rate of decomposition in real I G.Herzberg, 'Infra-red and Raman Spectra'. van Nostrand, New York, 1945.'N. B. Slater, 'Theory of Unimolecular Reaction', Cornell University, Ithaca. 1959. P. J. Robinson and K. A. Holbrook. 'Unimolecular Reactions', Wiley. London. 1972. W. Forst, 'Theory of Unimolecular Reactions', Academic Press, New York. 1973. 'Energy Storage and Redistribution in Molecules', ed. J. Hinze, Plenum Press, New York, 1983. 'H.-R. Dubal and M. Quack, J. Chc~tn.Phjx. 1984, 81, 3779. ' R. B. Sanderson, in 'Molecular Spectroscopy. Modern Research', ed. K. N. Rao and C. W. Matthews. Academic Press, New York. 1972.'R. L. Swofford. M. E. Long, and A. C. Albrecht, J. Clrcn7. Pkj..s., 1976. 65, 179. 'R. L. Swofford, M.E. Long, M. S. Burberry. and A. C. Albrecht, J. Chem. Phjx, 1977, 66, 664. lo R. L. Swofford, M. S. Burberry. J. A. Morrell. and A. C. Albrecht, J. Chetn. Php., 1977. 66, 5245.' R. G. Bray and M. J. Berry, J. Chrm. P/i!,.s., 1979, 71, 4909. I J. S. Wong and C. B. Moore, J. Clicw. P/ij..s., 1982. 77, 603. l3 K. K. Lehman, G. J. Scherer. and W. Klemperer. J. CIien7. Phj..~.,1982, 77, 2853. F. F. Crim, A1717. Rrr. Phj,.s. Chrtn., 1982. 35.657. l5 R. P. Wayne. 'Photochemistry'. Elsevier, New York, 1970. I' 1. Nadler, M. Noble, H. Reisler. and C. Wittig, J. Ciiem. Phj~.,1985, 82, 2608. I' H. Reisler and C. Wittig, An/?.Rrr. Phrx Clrem., 1986, 37, 307. * Delivered at a Symposium of the Faraday Division of the Royal Society of Chemistry, Scientific Societies' Lecture Theatre, London.on 6th May, 1987. t Present address: Department of Theoretical Chemistry, University of Oxford, 1 South Parks Road, Oxford OX 13 TG Overtone Spectroscopj? and Unimolecular Reactions Corresponding information also comes from high resolution spectroscopy of electronically excited states.” This article concentrates on the first method-overtone spectroscopy, which at present implies the spectroscopy of X-H stretching motions, because these are among the few which offer the energy and intensity to access states above 10000 cm-I (120 kJ mol-’). Section 2 below describes the detailed spectroscopy of very small molecules such as H,O, CH,, and C2H2 and explains how the bond anharmonicity of the X-H vibrations leads to a transition from ‘normal mode’ to ‘local mode’ The importance of Fermi resonance as a mechanism for energy transfer between different types of is then discussed in Section 3, with particular relevance to the overtone spectra of CHFC1225 and benzene.”.24 The following section shows the sense in which rates of intramolecular vibrational energy transfer may be inferred from the widths of overtone bonds.Finally Section 5 covers experiments on the overtone induced dissociation of H20226-27 Conclusionsand its theoretical interpretati~n.’~~’~~’~ are summarized in Section 6. 2 Local X-H Stretching Vibrations Henry’ discovered many years ago that the predominant X-H overtone bands of many molecules were consistent with the single mode Morse oscillator energy expression O and Figure 1 shows the frequencies of such bands for a variety of molecules, each containing several equivalent X-H bonds.More recently Swofford et a1.l’ have found that the overtone bands in benzene are unchangedh frequency or width on partial deuteration but decrease in intensity in proportion to the number of remaining C-H bonds; Figure 2 shows the AvcH = 6 band as an example. It therefore appears that the X-H bonds oscillate independently (as local modes) in these highly excited levels and that the bands shown have degeneracies equal to the number of equivalent X-H bonds. Very detailed recent work has fully confirmed this picture and solved a number of L. R. Khundar, J. L. Knee, and A. H. Zewail, J.Chern. Pliys.. 1987, 87, 77. l9 N. F. Scherer and A. H. Zewail. J. Clreni. PIijx, 1987, 87, 97. 2o C. S. Parmeter, J. Plijx Cliem., 1982. 86, 1735. 21 B. R. Henry. Acc. Chern. Res., 1977. 10, 207. 22 M. L. Sage and J. Jortner, Ad. C/icwi. Phj..s., 1981, 47, 293. 23 M. S. Child and L. Halonen, A& Chmi. Phjx, 1984, 57, 1. 24 E. L. Sibert, J. T. Hynes, and W. P. Reinhardt. J. Chern. Phjx. 1984, 81, I! 15. 2s A. Amrein, H.-R. Dubal, and M. Quack, Mnl. Pliys., 1985, 56, 727. 26 T. R. Rizzo. C. C. Hayden, and F. F. Crim, J. Chrm. Phjx, 1984, 81, 4501.*’ L. J. Butler, T. M. Tichich, M. D. Likar, and F. F. Crim, J. Chew. Phjx, 1986, 85. 2331. 28 H.-R. Dubal and F. F. Crim, J. Clieni. Ph~x,1985. 83. 3863. 29 T. Uzer, J. T. Hynes, and W.P. Reinhardt. J. Cliem. Pli~x,1986, 85, 5791. 30 G. Herzberg, ‘Spectra of Diatomic Molecules,’ van Nostrand, New York. 1945. 32 Child rl 1 IuI 10000 v/cm-l 20000 Figure 1 Frequencies of obsertied orertone bands. The shaded area lies above the dissociation limit ,for H,O, ENERGY (ern-') Figure 2 The lick, = 6 ouertone band qf ChHh compared with the same band in C,D,H increased in intensitj by a factor of six (Reproduced by permission from J. Chem. Phys., 1977, 66, 5245) consequent problems, the major difficulty being to account for the simplicity of the spectrum. Even in water, for example, the assignment of five vibrational quanta to Overtone Spectroscopj?and Uninzolecular Recrctions two equivalent stretching modes should give rise to six symmetry allowed Au = O-+ 5 transitions, of which only a narrowly separated pair (Au = 0.4 cm-') appear at first sight to contribute to the spectrum.On closer inspection, however, three very weak Av = 5 OH stretching bands are ~bserved,~~.~~ and a fourth band is confidently predicted.33 Furthermore the relative spacings of these bands, shown in Figure 3, and similar patterns in other molecules, have led to a clear understanding of the factors responsible for the normal mode to local mode transition. Two competing factors are at work: the interbond coupling responsible for normal coordinate behaviour and the bond anharmonicity, evidenced by the Morse oscillator-like progressions given by equation 1. Classically this competition may be understood by noting that the mutual response of two coupled oscillators varies inversely with the difference between their natural frequencies.Thus the transferability of characteristic C-H stretching frequencies from one molecule to another depends as much on their large frequency separation from other modes as from the absence of coupling.' The situation is more subtle in the case of two equivalent bonds because their natural fundamental frequencies must be identical, but in this case the bond anharmonicity causes a frequency decrease as the energy rises. In quantitative terms, the semi-classical identity (equation 2) between the classical frequency oand the quantum mechanical level spacing, shows for water (with o = 3 875 cm-' and x = 84 cm-') that there is a change of frequency of -840 cm-' between the u = 5 and u = 0 states, which may be compared with a splitting of only 100 cm-' between the stretching fundamentals, due to interbond coupling.One can therefore see that the interbond coupling will appear relatively strong (near normal-mode-like) in states In, m) if n % nz, but that the two bonds will oscillate almost independently if their excitation levels are very different (M = 0 and ni = 5 for example). Note that quantum numbers n and m are used to label local excitation states to avoid confusion with the traditional normal model labels ul, u2, etc..' The quantum mechanical equivalent of the above classical argument is illustrated in Figure 3. The left-hand side shows the energies of the six possible u = n + m = 5 two-mode Morse oscillator states together with the dominant An = 1, Am = & 1 coupling scheme implied by the potential functions used to analyse the spectrum.34 The right-hand side shows the five observed energies and one predicted level, from which it is apparent that the two directly coupled 12, 3) and )3,2) levels give rise to a large first order (or normal-mode-like) splitting of 224 31 J.W. Swensson, W. S. Benedict. L. Delboiville, and G.Roland, Mmm. Soc. R. Sci. Liege, special vol. 5, 1970. 32 L. S. Rothman, Appl. Opr., 1978, 17. 3517. 33 M. S. Child, Clrmi. Phjx Lett. 1982, 87, 217. 34 M. S. Child and R. T. Lawton, Fnradciy Discussion Climni. Soc., 1981, 71, 273. 34 Child 3 2-/ -4 1st order 3 '\ \ -3 2' 41 ----3rd order 4 41' 50------I-5th order 5 50' Figure 3 The local mode coupling scheme and resulting eigenvalue pattern for the v = 5 onertone manifold of H,O cm-', whereas the two lowest levels derived from 10, 5) and-)5,0>,which give rise to the predominant absorption bands, are split only in fifth order by 0.4 cm-'.Furthermore the predominance of the 15,O) over 14,l) and 13,2) bands is readily understood in terms of a localized bond dipole model because the transition intensity from 10, 0) to In, m) then depends on terms like (Olpln) (Olm), and the second factor (Ojm) vanishes unless m = 0.22723 The overall conclusion is that the strong bond anharmonicity of the X-H stretching modes drastically disturbs the familiar normal mode picture in a way that actually simplifies the appearance of the spectrum.Moreover recognition of this simplicity has led to remarkably accurate predictions of the overtone absorption frequencies of several small molecules.22 For example the frequencies of the predicted methane bands which are compared with experiment in Table 1 were derived from a model involving only three potential parameter^:^' two to specify the Morse frequency and anharmonicity and one to fix the interbond coupling strength. 35 L. Halonen and M. S. Child, Mol. Plzp., 1982. 46, 239. 35 Overtone Spectroscopj.and Unimolecular Reactions Table 1 Ohserwd und culculured CH stretching .frequencies in merhune.23 Assignments qfhrucketed hands are uncertuin.L' 1 2 916.47 2 916.4 3 019.49 3 021.0 2 - 5 791.6 5 861 5 856.4 6 004.65 6 010.2 3 8 604 8 612.5 8 807 8 807.8 8 900 8 909.5 9 045.92 9 041.7 4 11 270 11 262.5 (I 1 620) 11 553.3 (I 1 885) I1 907.2 5 13 790 13 796.4 14 220 14 206.9 (14 640) 14 641.1 6 16 160 16 213.3 16 740 16 745.1 7 18 420 18 513.5 19 120 19 166.0 8 20 600 20 697.4 9 22 660 22 765.3 3 Fermi Resonance Coupling The above isolated local mode picture is at best limited to small molecules with high frequency motions and a consequent low density of states. The situation in larger molecules is rapidly complicated by coupling between different types of mode, due in particular to Fermi-resonance types of ~oupling.~*~~-*~ Even the spectra of H,O and CH, are slightly perturbed by Fermi resonance, but the situation becomes much more drastic in benzene, where the CH stretching character is spread over possibly hundreds of molecular eigenstates in any particular overtone band.One therefore thinks of the X-H stretching modes as 'vibrational chromophores'6 rather than labels for any particular eigenstate. The transition from H,O-like to benzene-like behaviour is well illustrated by the low resolution survey spectrum of CHC1F225 shown in Figure 4. (It should be noticed that the optical path length and absorbing gas pressure have been adjusted to make all bands appear in equal strength, although the intrinsic absorption intensity decreases by orders of magnitude over this frequency range).Three factors are important for the interpretation. First the C-H stretching frequency, v, 2 3 020 cm-' is roughly twice that of the two C-H bending modes, v, 1 350-' and v,, 2 1 310 cm-', while Amrein et d2' estimate stretch-bend coupling constants of the order of 100cm-'. Hence a strong Fermi resonance' between bands characterized by the composite quantum number (equation 3) may be expected. Child CHCI F2 N-7/2 3180 kPa.m 205 kPa.rn x-1.5x-1 .o x 0.5 0.ob Figure 4 Survey spectrum ojCHC1,F in terms of reduced absorbance .xln(Zo/I), with values of .x and of the product of pressure times path length given in the inserts of the Jigure.The stick spectra indicate bandpositions and band strengths calculated from the multiple resonance model. (Reproduced by permission from Mol. Phys., 1982, 46, 239) N = V, + + +u,, (3) Secondly the number of component bands, with given N, increases in proportion to (N + 1)2. Finally the anharmonic constant in equation 1, zc, 60 cm-', of the stretching mode causes the higher us level separations to tune towards those of the more harmonic bending modes (x, =Xb 2 7 cm-I), leading to much stronger mixing for N = 4than N = 2. The theoretical model25 based on these ideas is seen, from Figure 4,to be in remarkably close agreement with the frequencies and intensities of the experimental spectrum. It was assumed that the relative intensities in each so called polyad are all borrowed from the Nth stretching state (us, u,, ub) = IN, 0, O), if N is an integer, or from the two dominantly stretching states (In, 1,O) and In, 0, 1)) if N = n + 1/2.37 Overtone Spectroscopy and Unimolecular Reactions Thus, for example, the strong band at N = 2 corresponds to an almost pure (2,0,0) eigenstate while the remaining weaker bands can be assigned to states with us = 1, ub = 2, or v2 = 0, u, + ub = 4. By contrast, at N = 4 the mixing is so strong that it is no longer possible to identify the parent 10,0,0) -14,0,O) transition; instead there are two roughly equally intense sub-bands plus a third with roughly half the intensity-a picture that persists in the higher frequency N = 5 and N = 6 visible absorption bands.36 Although individual peaks can still be identified, the individual quantum numbers L'~, v,, and L'b have lost their meaning; only the resultant N has a useful physical significance.This loss of information, or spread of excitation, is taken to a further stage in benzene to the extent that the overtone spectrum' ' consists of a sequence of broad peaks with widths of 30-100 cm-' (see Figures 2 and 5) attributable to successive local mode C-H stretching states coupled to a high density of background states. Sibert et have followed through the implications of a Fermi resonance model, assuming a geometrically determined interaction between the six CH stretching and six CH wagging modes, with the latter coupled in turn to six C-C stretching modes; furthermore vCH2 2v, where N includes the two latter types of mode.Figure 5(a) shows that the state density now increases dramatically from one level of the polyad to another, to the extend that the authors24 could include only the first three levels INCH),I(N -l),,,, 2N), and I(N -2),,, 4N) in their calculations; but they argue plausibly that inclusion of the complete polyads would merely fill in the gaps in Figure 5(b) to give smooth looking peaks with widths in excellent agreement with those found experimentally.' ' Each line in Figure 5(b) should also be broadened by rotational structure for comparison with experiment, although this 'heterogeneous' contribution (due to a superposition of distinct vibrational spectra from different initial rotational states) to the band width is likely to be quite small in benzene in view of its very small rotational constants.The conclusion from this calculation is that benzene has no simple high excited C-H overtone states. Instead the CH excitation is at least spread over a forest of CH wagging and C-C stretching states, coupled to it by Fermi resonance. The extent to which this coupling extends to other vibrational degrees of freedom (of which there are 30 in benzene, all told) remains, however, to be established. 4 Intramolecular Vibrational Relaxation The above picture of almost complete quantum mechanical mixing can also be cast in the time domain in terms of intramolecular vibrational relaxation at a rate determined by the width of the band.Imagine the (no longer quite hypotheti~al'~~'~) 'instantaneous' excitation of a pure local mode state <Do which by our previous assumptions carries the entire oscillator strength of the band. Since Q0 is not a stationary state (i.e. not a strict eigenstate), its composition will change with time. The probability of remaining in the original state for time t may, however, be calculated as ~{@o(0)~@o(~))\2,which is shown below to decrease at a rate proportional to the width of the band. 36 J. S. Wong. Dissertation. Berkeley. 1981. 38 Child -200 t I5CH ) -I6CH) i 19CH ) -v -p0 (cm-') Figure 5 A diugrani to slin~~ (u)the increasing density of Ferwi coupled levels responsible for the I&) bunduf benzene, and (b)the resulting calculatedstructure for the )5,,,), )ti,,, und 19cH)bands (Reproduced by permission from J.Chem. Phys., 1984, 81, 1115) The mathematical argument is as follows. Let (pidenote a set ofzeroth order time- independent wavefunctions, labelled by recognizable quantum numbers, and let 'po be the pure X-H overtone state. A typical eigenstate ynmay then be represented (equation 4)as where en", which determines the absorption strength, will be assumed consistent Overtone Spectroscopj. and Uninzoleculur Reactions with a Lorrentzian profile normalized such that where tio, denotes the energy of y,. Now equation 4 may be inverted to give, in particular, 'Po = c4hyn (7)n and the orthogonal nature of the transformation' ensures that do, = c,~.Equation 7 applies, however, only at time t = 0.It is replaced at later times by37 Qo(r) = CcnoVne-'"~' (8)n The survival probability in state (Do is therefore determined Here the orthogonality relation (y,,ly,) = 6,,, has been used in the first line, and the following standard integral3* has been employed The final result, as promised, is equation 11 I<@,(o)Q0(t))l2 = e-26', (1 1) where the time constant, 2b, of the exponential decay is readily seen to correspond with the full width at half maximum of the assumed spectroscopic profile in equation 5. 37 P. W. Atkins, 'Molecular Quantum Mechanics', 2nd Edn., Oxford University Press. 38 H. B. Dwight, 'Tables of Integrals'. 4th Edn., Macrnillan, 1961.40 Child Notice that this derivation says nothing about the density of lines in the profile. All that is required is that the number with perceptible intensity should exceed say 10,in order to justify replacement of the sums in equations 6 and 9 by integrals. The line density is, however, extremely interesting in this general context because it relates directly to the number of degrees of freedom into which the initially excited state DO(O)relaxes. These considerations explain the current interest in two exciting areas of spectroscopy. On one hand there is a drive to develop femto-second laser in order coherently to excite full band profiles in a time that is short compared with their relaxation time. Secondly, one seeks frequency-locked continuous wave lasers with the resolution to detect single eigenstates in situations where the line density may reach 100 lines per cm-'.Both types of experiment have been brought to bear on the problem of intersystem crossing in pyra~ine,~~?~' which is conceptually very similar to that of overtone relaxation, but without the added technical problem of very low oscillator strengths. 5 Overtone Induced Dissociation It is evident from Figure 1 that excitation to the uOH = 6 overtone level of hydrogen peroxide provides sufficient energy to break the 0-0 bond. Recent e~perirnental~~.~ studies of this process raise a number of and the~retical~~,~~ interesting questions. The first point is that the crude width (~86cm-' of the room temperature overtone band, shown in the right hand upper panel of Figure 6, would imply a dissociation lifetime, T 2 0.05 ps, which is seriously at variance with RRKM-like estimates28 of 5-50 ps and with a detailed dynamical estimate29 of about 8 ps.The middle panels of Figure 6 show however that the apparent band width can be seriously misleading with respect to the dissociation lifetime for a molecule with as large a rotational constant as H202.(A 10 cm-' for rotation about the 0-0 axis) becauseinhomogeneous rotational broadening,estimated2' as 2(2AkT)* 80cm-', can account for the entire profile. It is, in any case, apparent from the upper panels of Figure 6 that the 4vOHband which cannot excite dissociation has much the same width as the 6~0".The lower panels show, however, that jet cooled samples,27 with only the few very lowest rotational states populated, give quite different pictures for the 4vOHand 6~0,bands.The former is ultimately resolvable into individual lines with widths ofthe order Av 0.08cm-', which aregoverned by experimental factors. The 6~0,band is also drastically simplified by jet cooling but the features are substantially broader, and the authors2' conclude, after careful tests, that this breadth is due predominantly to homogeneous lifetime broadening. They take the 1.5 _+ 0.3 cm-' of the narrowest feature near 18 942 cm-' as setting a lower limit on the dissociation lifetime (-2mAv) of 3.5 ps, in good agreement with the above theoretical estimate^.^^'^^ 39 B.J. van der Meer, H. T. Jonkman, J. Kolmmandeur, W. L. Meerts, and N. A. Majewski, Clzem. Phvs. Lrrt.. 1982, 92, 565. 'O B. J. van der Meer, H. T. Jonkman, G. ter Horst. and J. Kommandeur, J. Cliem. Phjx, 1982, 76, 2099. 41 Overtone Spectroscopj* and Unindecular Reactions .3200 13300 “OH 13340 18945 18950 A? = 008 cm-’ 1cm-’ 13358 13362 13370 13374 13392 13394 Wavenumber, 3 (cm-’ ) Figure 6 Spectra of ilibrationul oilertone transition to u bound state (4t;,,) und to a predissociatiue state (6u,,,) of hydrogen peroxide: (a) and (d) It‘ere taken at loit. resolution.from room temperature samples; (b), (c), and (e) are high resolution spectra from jet-cooled samples. Note the marked dijference in line width hettr3een (c)and (e).indicatizle ofhomogeneous like-time broadening in the 6v,, band (Reproduced by permission from J. Clwm. Plij~s,,1986. 85, 233 1) The two calculations give, however, rather different pictures of the dissociation dynamics. Dubal and Crim” use an RRKM-like statistical adiabatic channel mode141 based on the assumption that the energy from the initial OH stretching motion is equilibrated between all internal degrees of freedom before dissociation. Uzer et on the other hand, follow the classical dynamics on a spectroscopically based potential energy surface, and find that the energy flow is by no means completely statistical. A fairly close (but not exact) Fermi resonance leads to rapid (0.2 ps) partial energy exchange between the initially excited oscillator and the directly coupled H-0-0 bend.There is also some excitation transfer to the second 0-0-H bending motion, but no smooth energy build-up in the 0-0 dissociation mode, nor any significant energy transfer to the second OH stretch or to the two torsional degrees of freedom. Instead the 0-0 bond appears to break 41 J. Troe. J. C/ieni. P/IJX.1983. 79. 6017. Child when the system reaches some as yet uncharacterized (Slater-like?) critical configuration. Information about the internal states of the nascent OH fragments also bears on the question of statistical or non-statistical energy flow, because the energy of the uOH = 6 stretching state exceeds the dissociation energy by roughly 20 kJ (1 600 cm-’), which is sufficient to populate at least the J = 8 rotational state of OH, and populations up to J = 10 are reasonably accessible from a room temperature sample.Rizzo et a1.26find, by laser-induced fluorescence, that the OH and OD fragment rotational population, from the 6~0, overtone dissociation of HOOH and HOOD are consistent with a statistical distribution over product states which is in turn consistent with (but not strictly dependent on) a statistical energy distribution over the molecular modes before dissociation. On the other hand a weak band at the slightly higher frequency 600, + 385 cm-’, gives rise to a shift in the fragment rotational distribution to lower J values, contrary to the prediction of any statistical As a further test it would be very interesting to see whether the product rotational distributions from jet cooled samples are similar to those obtained at room temperature in case the ‘statistical’ product state distribution arises from statistical averaging over initial populations rather than from statistical dynamics during the dissociation process.Another interesting experiment would be to examine the relative OH(c = 1) and OD(u = 1) product yields from higher nuoH overtone bands of HOOD, because extrapolation of the dynamical results29 would predict only vibrationally excited OH. Unfortunately, however, the interpretation would be complicated by the possible influence of a second electronic state at energies above the v~~ = 7 energy level. Further information on the extent to which an initial OH excitation is distributed over the molecule is in principle contained in the line densities of the band. (Similar information from the 6~oHband would be even more interesting, but the homogenous line broadening due to dissociation will probably prohibit the necessary resolution.) Butler et al.27 estimate a mean density of 3.6 vibrational lines per cm-’ at the 4cOH energy level if the OH excitation were shared between all degrees of freedom, whereas the density of moderately strong lines in Figure 6(c) is at best one per cm-’ and some at least of these come from initially rotationally excited levels.Hence there is again an indication that the energy flow is not fully statistical. 6 Conclusions Several conclusions can be drawn from these new spectroscopic experiments.In the first place we are reminded that the normal coordinate picture of vibrational modes is dependent on the harmonic approximation, and it has been shown that the anharmonicity of X-H stretching potentials leads to a transition from the normal mode to a local mode picture. Moreover the relatively high intensity of X-H overtone transitions provides a series of windows for investigation into the nature of intramolecular dynamics. In considering the extent of intramolecular relaxation from X-H stretching modes to other degrees of freedom, Fermi resonance coupling to neighbouring Overtone Spectroscopy and Unimolecular Reactions X-H bending modes was seen to provide probably the fastest general relaxation mechanism.Alternatively, in the frequency domain, Fermi resonance provides the dominant contribution to the breadth of the band over which the X-H overtone state is spread. Conversely the density of lines within the band in principle gives a measure of the number of significantly coupled degrees of freedom, but the measurement of such densities puts a high premium on the spectroscopic resolution. Finally the overtone induced dissociation of hydrogen peroxide was shown to offer direct insight into the detailed mechanism of the decomposition process. Careful line width measurements fix the rate; laser-induced flourescence can assess the statistical or non-statistical nature of the energy disposal over fragment states; and the line density of neighbouring non-dissociative bands can give information about the number of coupled degrees of freedom.At present there is information both for and against a purely statistical picture of the dissociation. The weight of evidence favours a less than fully statistical description, but it will be interesting to see how the balance tilts in the light of future experiments. Acknowledgement. The author is glad to acknowledge the hospitality of the Department of Chemistry and Biochemistry at the University of Colorado in Boulder, where this article was written.

ISSN:0306-0012    

DOI:10.1039/CS9881700031

出版商:RSC    

年代:1988

数据来源: RSC

摘要:

Chem. SOC.Reu., 1988, 17,45-67 Spectroscopic Structure of Open-shell Polyatomic Cations By John P. Maier INSTITUT FUR PHYSIKALISCHE CHEMIE, UNIVERSITAT BASEL, KLINGELBERGSTRASSE 80. CH-4056 BASEL, SWITZERLAND 1 Introduction Ions are transient species which have attracted increasing attention in the last decade. This is because they are known to occur in comets, interstellar space, planetary atmospheres, plasmas, and flames, where their presence has led to many suggestions concerning their involvement in chemical reactions. For example, the formation of hydrocarbons in dark interstellar clouds has been modelled by ion- molecule schemes.’ In addition, technological developments have allowed the design and implementation of sensitive methods for the spectroscopic characterization of such reactive species, which are invariably produced in small concentrations.The focus of attention of this review is on the approaches to characterize open- shell cations of polyatomic molecules spectroscopically by means of their electronic transitions. Such data on diatomic species have provided the foundations of modern spectroscopic theories,2 and the advent of high resolution lasers and Doppler-free techniques3 has led to further refinements and developments. Many diatomic, and a few triatomic, ions can be readily produced in discharge sources with sufficient densities for detection by classical emission and absorption method^;^ with polyatomic ions, however, their generation and the environment for their study is more demanding and crucial.Consequently such investigations have required specially designed appro ache^.^ This last area has been the centre of our research activity over the past decade and will be considered presently and illustrated by examples showing the sort of information on ions which can be obtained. There are also several other approaches used to study the spectroscopy of ions which are not discussed here. In particular, the advances in i.r. lasers, and in the means of detection of ions by modulation methods, have led to many successful measurements of vibrational transitions of closed-shell cations,6 and recently of some open-shell polyatomic cations7 and anions.* A few simple cations have also G. Winnewisser, Top.Curl. Cliem.. 1981. 99, 39. G. Herzberg, ‘Spectra of Diatomic Molecules’, van Nostrand, New York, 1950. See for example. G. Duxbury, Chem. Soc. Rer., 1983, 12, 453. G. Herzberg, Quart. Rev., Cliem. SOL..,1971,25,201;G. Herzberg, Pro(*.Indictn Not. Sci. Acad., 1985,A51, 495. J. P. Maier, Ace. Clieni. Rex, 1982, 15, 18.‘C. S. Gudeman and R. J. Saykally. Annu. Rer. Pl7y.s. Clienz., 1984, 35, 387.’M. W. Crofton. M.-F. Jagod, B. D. Rehfuss, and T. Oka, J. Cliem. Pliys., 1987, 86, 3755. a M. Polak, M. Gruebele, and R. J. Saykally, J. Ant. Climi. Soc., 1987, 109, 2884; M. Gruebele, M. Polak, and R. J. Saykally. J. C/ien7.Plijx, 1987, 86. 6631. Spectroscopic Structure qf Open-shell Polyatornic Cations been characterized by microwave spectros~opy.~ Other methods have utilized mass-selected ion-beams in conjunction with laser excitation both in the visible" and i.r.regions." A further area of activity is in obtaining information on the structures of ions by indirect means, such as those using photodissociation,' 'ion-molecule or collisional processes on mass-selected ion beams produced in conventional' as well as laser evaporation and supersonic expansion source^.'^ Some ion geometries have been determined by the Coulomb explosion approaches. As far as polyatomic cations of stable molecules (as well as of some semi-stable species) are concerned, photoelectron spectroscopy has provided a great deal of information on the energy location of the electronic states of their ions, and some on the vibrational structure in the lowest states.I6 The limitation of the method has been the resolution, typically 160 cm-', though it has recently been shown that it is possible to reduce this to a few cm-' in favourable cases.I7 Consequently the methods which we have applied are based on the detection of photons, because their energy can be determined much more a~curately.~ A further improvement came about with the incorporation of lasers, with the associated inherent resolution and intensity advantages.' The spectral characterization of open-shell cations discussed here is the result of studies of the emission spectra from electron excitation of effusive and supersonic free jets, of the laser excitation spectra with Penning ionization, and of absorption measurements on ions isolated in 5 K neon matrices." It goes without saying that the concomitant application of such techniques has enabled information to be extracted which by one method alone would not have been possible.Thus with these methods the ions are prepared with rotational and vibrational temperatures ranging from z 5 K to 300 K, using different projectiles to generate the ions. The analysis of the respective spectra yields the term values of the electronic transitions, vibrational frequencies, and, in favourable cases, the rotational constants and hence geometries of the ions in the ground and lowest excited electronic states. Related to the above are the studies of the radiative and non-radiative relaxation pathways of state selected ions.These aspects are not considered here and have been the object of other reviews based on the techniques used for such investi- R. C. Woods, J. Mol. Srrucr., 1983, 97, 195. A. Carrington. Proc. R. Soc. London A, 1979, 367, 433. 'I A. Carrington and T. P. Softley in 'Molecular Ions: Spectroscopy, Structure and Chemistry', ed. T. A. Miller and V. E. Bondybey, North-Holland, Amsterdam, 1983, p. 49. R. C. Dunbar in 'Gas-Phase Ion Chemistry', Vol. 3, 'Ions and Light'. ed. M. T. Bowers, Academic Press, New York, 1984, p. 129. l3 'Ionic Processes in the Gas Phase', ed. M. A. Almoster-Ferreira, D. Reidel, Dordrecht, 1984. l4 R. E. Smalley, Lusrr Chem., 1983, 2, 167. E. P. Kanter, Z. Vager. G. Both. and D. Zajfman, J. Chem.PAjx. 1986,85,7487, and references therein. l6 D. W. Turner, C. Baker, A. D. Baker, and C. R. Brundle,'Molecular Photoelectron Spectroscopy', Wiley- Interscience, New York. 1970. K. Muller-Dethlefs. M. Sander, and E. W. Schlag, Z. Narurfbrsch.. Tril A, 1984, 39, 1089. T. A. Miller and V. E. Bondybey, J. Chim. Php. Phjx-Chim. Biol.,1980, 77, 695. I9 J. P. Maier, J. El. Sprctro.w. Rel. Phenoni., 1986, 40, 203; J. P. Maier, J. Clirtn. Sol,..F(irritiriy Trms. 2, 1987. 83, 49. Maier gations; photoelectron-photon,20 photoion-photon,21 or photoelectron-ion22 coincidence measurements yielding lifetimes, fluorescence quantum yields, and fragmentation branching ratios. -E hv(Laser1 -hvT Ar* EMISSION LASER EXCITATION ABSORPTION IN 5K NEON Figure 1 Essentid aspects qf the emission, laser ercitation, and matrix absorption approaihes for the iiibrational characterization of cations.The ions are prepared in the three methods bl- =200 eV electron impact, by Penning ionization (Ar, He metastables) and by photoionization [H(Ly X) or Ne(I)]. The latter two processes are follo,t*ed by collisional rela.uation (M~Jlines)prior to wailelength selected photon absorption 2 Vibrational Characterization The principal features of the three approaches used for vibrationally characterizing open-shell cations of polyatomics in their ground (8)and lowest excited (A,B.. .) electronic states are depicted in Figure 1.The three methods rely on the detection of the dipole allowed electronic transition between the ionic states; in the case of the emission and laser excitation experiments, the emitted photons are monitored, whereas in the matrix the absorption is measured directly.The ions are initially generated by different means in the three approaches. In the emission method the ionization takes place by electron impact on an effusive or supersonic beam containing the molecular precursor; in the laser excitation technique Penning ionization with He or Ar metastables and collisional relaxation is employed; and for the absorption measurements, molecules embedded in a 5 K neon matrix are photoionized with either H(Ly X) or Ne(1) radiation. The emission spectra yield vibrational information mainly on the ionic ground state (g)and, usually, to a lesser extent on the excited (2)state.On the other hand, '() J. P. Maier and F. Thommen, in 'Gas-Phase Ion Chemistry'. Vol. 3, 'Ions and Light', ed. M. T. Bowers, Academic Press. London, 1984, p. 357. " S. Leach, J. Moi. Slrucr., 1986. 141, 43. "T. Baer. Adr. Cher?~.fhj.s., 1986, 64, 11 1. Spectroscopic Structure of Open-shell Polyatomic Cations the laser excitation and absorption spectra probe the excited states because the population of the ions is concentrated in the lowest vibrational level of the ground state (8)prior to the interaction with the electromagnetic radiation. A. Emission Spectroscopy.-This was the first approach used to look for the emission spectra from electronically excited polyatomic ions.23 In the initial experiments an effusive beam of the sample was ionized/excited by a collimated electron beam (typically 20-200 eV, several mA current), in a perpendicular arrangement; the resulting fluorescence was dispersed by a monochromator and detected by single photon counting electronics.The optical resolution usually attainable for studies of the vibrational structure of the electronic transitions was ~0.01nm, and this could be improved in a few cases by a factor of two. The strengths of this approach are the sensitivity and especially the ability to scan large wavelength regions (200-900 nm) easily. The results obtained in such studies24 have laid the foundation stone for the application of higher resolution laser, and of technically more demanding approaches such as photoelectron-photon coincidence measurements.2o Though the approach has by and large outlived its usefulness, it is still occasionally used to obtain the initial information on the optical transitions of ions of, for example, semi-stable (vide infva) and, recently, of the type XF; (X = C,26 Si,27 Ge2’) using similar or related experimental arrangements. The emission spectra of over one hundred open-shell polyatomic cations have been observed.These are listed in Table 1, or their structural moiety is indicated (e.g. fluoro-benzene cations). The symmetries of the electronic transitions are also given and the references to the original and subsequent studies on these ions are to be found in several corn pi la ti on^.^^^^^^^^ In two of these the lifetimes of the ions in their excited electronic states are also li~ted.~~,~~In a further review, the fluorescence quantum yields, and radiative and non-radiative rate constants of many of the ions are given and their relaxation behaviour is discussed.20 An important improvement in the quality of the emission spectra and the amount of information which could be extracted came about with the introduction of the molecular precursors into the ionization region in the form of a supersonic free jet.31 Usually seeded helium jets were employed: the molecular precursor at a pressure of 1-10 mbar was premixed with z 1 bar of He and expanded through a 50-100 pm orifice. As a result of the expansion the internal degrees were cooled 23 J. P.Maier,in ‘Kinetics of Ion-Molecule Reactions’,ed. P. Ausloos, Plenum Press, New York, 1979, p. 437. 24 J. P. Maier. Chiniici, 1980, 34.219. 2s J. P. Maier. Pliilos. Trans. R. SOC.London A, 1988, 324, 209. ”J. F. M. Aarts, S. M. Mason and R. P. Tuckett, Ma/. PIijx, 1987, 60, 761. J. F. M. Aarts, Clieni. PIijx, 1986, 101, 105; S. M. Mason and R. P. Tuckett, Mol. Plip., 1987, 60,771.’* H. Van Lonkhuyzen and J. F. M. Aarts. Clirrw. Pliys. Lett., 1987, 140, 434. 29 J. P. Maier. 0.Marthaler, L. Misev, and F. Thommen, in ‘Molecular Ions’, ed. J. Berkowitz and K.-0. Groeneveld, Plenum Press. London 1983, p. 125. T. A. Miller and V. E. Bondybey, Appl. Spwirosc. Rev.. 1982, 18, 105; T. A. Miller, Annu. Rrt.. Plivs. C/ietii.,1982. 33, 257.31 D. Klapstein. J. P. Maier. and L. Misev, in ‘Molecular Ions: Spectroscopy, Structure and Chemistry’, ed. T. A. Miller and V. E. Bondybey. North-Holland, New York, 1983. p. 175. 48 Muier Table I Polyatomic. cations for ~~hicli emission spectra have been detected in the gas phase. (Not all the isotopic dericatives studied are listed.) A literature survey of all the studies o these cations by means qf the indicated electronic transitions is to be found in seceral revie~s'~~~-~~ Cation Transition co; cs: N20+ +ocs J2n --:TI (C,,) H,O' A2Al --X2B, (C,,) H2S+ A2A, --f2B, (C,,) XCN+ B'n -f2n (C,,) X = C1, Br, I &+ -pnXCP+ A2x+ -f2n (CXV) X = H, D, F XBS' A2c+-T2n ((2%") X = H, D, F, C1 XC=CH+ A2n -Y2n (C=J X = C1, Br, I xcxx + $17, -Z2II,, (Dxv) X = C1, Br, I ClCECBr+ A2n -X2n (C=,,) RF; b2A, -c2T2 (Td) R = C, Si, Ge X(C=C),X (Dxh) X = H, D, F, C1, Br, I+ NXCXCsN (Dzh)cis-1,2-difluoroethylene+ A2A, -X2B1 (C2JH( CC ),C-N A2n -<2n (C,,)+ CH,CP+ A2A, -<'E (C,,) CH,C=CX+ A"E -X2E (C,") X = C1, Br +H(C-C),H $n, -p2n, (Dzh) NrC(C-C),C-N + A2n, -T2nq(Drh) CH ,( C=C),X + A"E -f2E (C3,) X = H, D, F, C1, Br CF,( C=C), F + H( CX), F A2E -f2E (C3,)+ CH ,(CK) CEN+ A2n, -T2nq (Dsh) R(C=C),R A2E -f2E (C3,) R = CH,, CD,, CF, + +Et (C=C),X A2Eu --X2Eg (D3J X = H, CN, Et +fvuns-1,3,5-Hexatriene J2A" -f2A"(C,) ci.r-l,3,5-Hexatriene+ A2B, -f2A, (C,,)frans-1,3.5-Heptatriene' AZA, -PB1 (C,,) trcms-1,3,5,7-0ctatetraenei 22 Ajf +2A" (C,) Bu'( C=C),Bu' + A2A, --X2Bg (c2h)22A" -22A-(CS) Benzenoid Cations C6F6 + +1,3,5-C,H,X, X = F, C1 1,3,5-C,F,X,+ X = C1, Br 1JS-C6F,(CH,), + Fluorobenzenes+ Chlorobenzenes+ Bromofluoro benzenes Bin-') -P(x-l) (C,)+ Fluorophenols + Chloroalk ylbenzenes C,F,X+ + iX = CF,, CH,, CN, CHO, NO, NH, 49 Spectroscopic Structure of' Open-shell Polyatomic Cations I1 II I 1 I I I I I I 25000 26000 -31 Trot-10K I I I I 1 I I I I I I I1 27000 28000 29000cm-' Trot -300K ,Ill ,I,I ,111 1111 I( 20000 25000 dooo cm-1 Figure 2 Emission specrrutn.A"n-T2H of c~ilorou~,etj.l~~n~~cution e.\-citcci bj. z200 eV electron impact on aseededHe .supersonic.,frec.jet (topT,,, z 10K;0.03 nm resolution)or on an effusire beam (bottom-TT,,, z 300 K, 0.16 nm fwhm).The burs uhoiie the i:ihrurional assignnients refer to the R = 112 component: [he dots refer to utoiwii' lines (T,,, z 5-10 K) and, following electron impact ionization, one obtained emission spectra of rotationally cold (T,.,, z 5-10 K) ions. In Figure 2 is shown such an emission spectrum, the A"'n -T'II transition of chloroacetylene cation, recorded in a supersonic free jet32 (top) and with an effusive beam33 (bottom). The improvements are self-evident. As a result of the reduction in the inhomogeneous broadening of the bands at rotational temperature 5-10 K, 32 D. Klapstein. R. Kuhn. and J. P. Maier, Cheni. Ph~x.1984. 86. 285. 33 M. Allan, E. Kloster.Jensen. and J. P. Maier, J. C/IWI.SOC..Fmdui.Trtim. -7. 1977. 73, 1406. 50 Maier Table 2 Fundamental vibrational frequencies (cm-') of chloroacetylene cation inferred from the emission32 and laser excitation34 spectra. The molecular values are from ref: 36 C+ n Species State v, 02 03 04 v5 35C1-CK-H X'C' 3 340 2 110 756 604 326 35C1-CK-Hf 82113/23 146 1984.5 836.8 595 235 A2113,, 3 230.7 2 063.8 595.7 224 35C1-CK-D X'C' 2612 1980 742 472 312 35C1-CS-D+ f2r13/2 2475 1882.9 817.0 476 A"2113,2 2 561.2 1919.7 587.2 216 Uncertainty in values is f0.2 cm-' for figures quoted to one decimal place, all others f1 cm-' vibrational frequencies can be inferred to within 1 cm-' (and sometimes to even less), as opposed to 10cm-' at an ambient temperature of z300 K.Furthermore, sequence and combination bands, obscured by the broadened profiles at the higher temperatures, become apparent. Important for the vibrational assignment in the case of the spectrum of chloroacetylene cation (Fig. 2) is the resolution of the bands of the naturally occurring isotopic species 35C1-C=C-H+ and 37Cl-C=C-H (in the intensity ratio 3: l), and the identification of the bands belonging to R = 3/2 and 1/2 sub-components. Although the chloroacetylene cation in the z2nand J211states is linear,34 there is a large difference in the C-Cl distance in these two states as well as in the molecular ground state (X'C') values.35 The C-C distance changes less and the C-H distance variation is small.The result of this is that the emission spectrum is congested with transitions involving many vibrational levels in both the J211and f2nionic states; the u3(C-CI stretch) vibration of X+ symmetry is especially prominent (cJFigure 2).32The vibrational frequencies of all the five fundamentals (3*C+,2.n)in the z213state are obtained from the analysis of the emission spectra of CI--C=C-H+ and Cl-C=C-D+, as well as their u3 frequency in the J2llstate. The fundamentals are listed in Table 2 (for the 35Cl derivative) together with the molecular The detection of the emission spectra has also proved fruitful in the vibrational characterization of ions of unstable molecular species. The spectroscopic studies of the latter neutral species have been the subject of a review in this journal.37 Armed with the knowledge of the photoelectron spectra of the semi-stable molecules, and with the methods used for their preparation in the microwave studies, the emission spectra of the cations of the phospha-alkynes, XCP+ (X = H, D,38F,39CH3"') 34 M.A. King, J. P. Maier, and M. Ochsner, J. Chem. Phys., 1985, 83, 3181. 35 E. Heilbronner, K. A. Muszkat, and J. Schaublin, Helv. Chim. Acta, 1971, 54, 58. 36 G. R. Hunt and M. K. Wilson, J. Chem. Phys., 1961, 34, 1301. 37 H. W. Kroto, Chem. SOC.Rev., 1982, 11, 435. 38 M. A. King, D. Klapstein, H. W. Kroto, J. P. Maier, 0.Marthaler, and J. F. Nixon, Chem. Phys. Let[., 1981, 82, 543; M. A. King, R. Kuhn, and J. P. Maier, Mol. Phys., 1987, 60, 867, 39 M. A. King, D.Klapstein, H. W. Kroto, R. Kuhn, J. P. Maier, and J. F. Nixon, J. Chem. Phys., 1984,80, 2332. 40 J. Lecoultre, M. A. King, R. Kuhn, and J. P. Maier, Chem. Phys. Lett., 1985, 120, 524. 51 Spectroscopic Structure qjOpen-shell Polyutomic Cations HBS+ A*E+ -Z2n 000-oov 0 1 2 1 1 I I I I I Ne 001-000 n BS n 000-10v - 0 1 A 21000 20000 19000 18000 17000 cm-1 Figure 3 The J2C' -p2neniission spectrum of thioborine cation (0.033 nm fwhm)obtainedby passing H2Souer B chips at = 1 OOO OC and !z200 eV electron beam ionization. The assignment of the main pr-o~ressions-u,(B-H), u,(B-S)-is indicated and of the sulphidoborons, XBS' (X = H, D,41F, C142) were recorded. In all these cases the observed band systems are due to their J2C+-f2n electronic transitions (A2Al -p2Efor CH,CP+).The analysis of the vibronic structure has enabled their fundamental frequencies, spin-orbit constants as well as Renner- Teller and Fermi resonance parameters in the f2nstate, which complicate the pattern, to be inferred. In Figure 3 is shown the 2%' -T2rIemission spectrum of HBS' obtained by passing H,S over B at 2 1 000 "C and ionization of the products by a ~200 eV electron beam with the apparatus de~cribed.~~ The essential modification to the instrument was the incorporation of an oven near to the ionization region, in much the same way as in studies of the semi-stable molecules by photoelectron spectroscopy.,' In the spectrum (Figure 3) two sub-band systems are apparent as a result of the spin-orbit splitting in the cationic ground state (A: z -323 cm-').The rotational profiles show a strong Q-branch, and a weaker P-branch, maximum component. The vibronic pattern is complicated by the overlap of the isotopic bands of H11B32S' and H10B32S+ (the bands due to the 34S modification are not clearly discernible). The two main progressions involve the two C+ modes, I;,(B-H str.) and u,(B-S str.), whereas the weaker bands are due to the excitation of the degenerate, u2, mode associated with which are further splittings and intensity changes because of Renner-Teller and Fermi resonance interactions. The observed pattern can be reproduced by the usual parameters which mimic such effects. The emission spectrum thus yields the frequencies of the three fundamentals in 41 M.A. King, D. Klapstein. R. Kuhn, J. P. Maier, and H. W. Kroto, Mol. Phj.~.,1985. 56. 871. 42 M. A. King, R. Kuhn, and J. P. Maier. J. Pl7j.s. Cheni., 1986. 90. 6460. 52 Maier Table 3 Fundamental uibrational.freguencies (cm--')and spin-orbit splittings, AO.eff.(cm-') of the phosphaetiiyne and sulphidoboron cations obtained from their emission spectra. The calues giten are for the most abundant isotopic derivative Cution State c, Cz 63 Ao,err. Rf$ HCPf .f2113,23 125.1(4) 642( 1) 1 167.1(4) -146.97(3) 38 2111i23 124.9(4) 1 159.9(4) AzC' 2 985.6(4) 706(1) 1 275.4(4) DCP' f2113i2 2 356.5(4) 499( 1) 1 112.4(4) -146.71(1) 38 'llli22 356.6(4) 1 113.4(4) AZC' 2 274.4(4) 552( 1) 1218.1(4) FCP+ d2n 1729(2) 765( 1) -190.2(6) 39 azE+ 1866(2) 817(2) HBS" Z2113iz 2 746.8(4) 659( 1) 9 75.9(4) -322.6(4) 41 2111,z 2 747.1(4) x991 J2E' 2 214.8(4) 550( 1) 1050.9(4) DBS f2113,z+ 2 071.1(4) 9 3 7.4( 4) -322.2(4) 41 2111122 074.2(4) z 993 A2Z' 1 706.6(4) 1 01 1.1(4) FBS' Z2113,,2 1721(2) 339(2) 637(2) -339(2) 42 2111,z1718(2) 633(2) A2C+ 1718(2) 691(2) ClBS' 2'l-I 1347.8(8) 508.9(8) -383( 1) 42 AZCf 1 390.6(8) 516.0(8) the T2xstate and the two stretching modes in the J2C' state (Table 3).These values have been obtained for the four isotopic derivatives, H"B32S+, H"B3*S+, D' 1B32S', and D'oB32Sf.41 In Table 3 are summarized the vibrational frequencies, and spin-orbit constants, determined from the emission spectra of all the XCP' and XBS' species studied.It can be seen that this approach has led to a fairly extensive vibrational characterization of these ions. With such information in hand, higher resolution studies using both visible and i.r. lasers, are now feasible. B. Laser Excitation Spectroscopy.-As indicated in Figure 1, the laser excitation spectra lead primarily to a vibrational characterization of the ions in the excited electronic states. Thus it yields spectroscopic data which are complementary to those obtained from the emission spectra. The reason for this lies in the means of preparation of the ions in the two methods and is illustrated by two examples: the electronic transition of bromocyanoacetylene and of chloracetylene cations.In Figure 4are shown the spectra of the A2rI-22rI transition of Br-C&-C=N+ recorded in emission in a supersonic free jet (top trace) and as a laser excitation spectrum (bottom) at a reduced ambient temperature (10G150 K).43As is seen the emission and excitation spectra provide, almost exclusively, the vibrational characterization of the ion in the 2'l-I and J2IIstates respectively. The explanation is straightforward. In the laser excitation approach the ions are prepared by 4' R. Kuhn, J. P. Maier. L. Misev, and T. Wyttenbach, J. El. Specrrosc. Rel. Phenom., 1986, 41, 265. Spectroscopic Structure of Open-shell Polyatovzic Cutions Br-CfC-EN' A2n-4% I I I I 19000 18000 17000 cm-' A2n3,2-ji2n3,2 Trot lOOK 21000 20000 19000 cm-1 Figure 4 Tk A" n-y'n electronic transition of' hronioc~anoacet~.lenecation recorded in emission (topTT,,,z 10 K, 0.05 nm fwhm) and as laser escitation spectrum (bottom-T,,, % 100 K, 0.02 nm fwhm).The emission spectrum sholz9.7 borh R = 312 and 1/2 (denoted \i,ith bar aboce the assignment) components, the e.ucitation spectrum only Q = 3/2 Penning ionization (with Ar or He metastables) and are subsequently collisionally relaxed by the excess of the rare gas atoms, which have been cooled by a liquid nitrogen environment. Thus the population of the ions is concentrated in the lowest vibrational level of the T211state and levels lying up to =500 cm-' above may be populated to a few percent. The 'll component is also completely quenched as the spin-orbit splitting in the ground state of the ion is z -900 cm-'.The excitation spectrum is consequently relatively simple (Figure 4). The laser photons induce transitions from the lowest level (1' vibrational symmetry) of the T2113,2 state to the symmetry and Franck-Condon allowed levels of the 22113,,,state. As can be seen the v,(C-Br str.) mode is particularly strongly excited; in addition the excitations Maier of the u,(C-C str.), v,(C-C str.), and of the degenerate modes 06 and u7 in double quanta are observed. A more complicated laser excitation spectrum would be obtained using ions produced under collision-free conditions. The emission spectrum of bromocyanoacetylene cations cooled rotationally to 5-10 K (Figure 4) shows that the vibrational excitation is mainly in the f211state. Only the u4 vibration, and those of the low frequency degenerate modes 06 and v7, which appear as sequence transitions to the red of each main band, are excited by one or two quanta in the J217 state.Transitions between the 2111,2-2111/2 components (denoted with a bar in the figure) are also seen because in the electron impact ionization of the molecular species in the X'C' state both R = 3/2 and 1/2 components are populated. The J'n +-X'Z' ionization process is fairly vertical (linearity is retained and all the internuclear distances are expected to be a little elongated) and only a few excited vibrational levels in the X2nstate are significantly populated because of the Franck-Condon factors.44 In the J217-X'n emission, various levels in the X'll state are accessed, those involving the u4(C-Br str.) mode are prominent as in the excitation spectrum.This is expected because the main geometrical change on passing from the X'll state is a decrease in the C-Br distance, which is readily predicted from the molecular orbital description of the electronic structure.44 This is also reflected in the u4 fundamental frequency changes: X'C+--419, 8211,i2-438, J2n,/,-359 ~m-'.~,Thus in considering the vibrational activity in the emission spectrum, the transition probabilities involve three electronic states, X'C+ -J2n-+ X'n, whereas in the laser excitation spectrum only the two ionic states need to be considered. The analysis of the emission and excitation spectra of bromocyanoacetylene cation yields the vibrational frequencies, to within +1-2 cm-', for most of the seven fundamentals in the T'll and A2ll states.Such combined studies have been carried out for many of the ions listed in Table 1; tables of the ionic vibrational frequencies are to be found in the original references. In the case of the J211--8'17 transition of chloroacetylene cation, the emission spectrum,' (Figure 2) also provided extensive information on the J211state. This is because the large geometry change on ionization to this state (C-Cl distance increase of z 0.13 results in the population of a larger part of the upper state surface than with bromocyanoacetylene cation. The laser excitation spectrum, however, is again simple34 (Figure 5)for the same reasons as discussed above. The vibronic structure is dominated by the excitation of the u> mode in progression and combination series with the ul(C-H) and u,(C-C) Zl modes.Though the emission spectrum (Figure 2) also shows the u; progression, the 2;; and u; modes are discernible only weakly, reflecting the small Franck-Condon factors to these levels in the J'n-X'X' ionization step. The frequency values obtained by the emission and laser excitation approaches in conjunction are used for the completion of Table 2. "G. Bieri, E. Heilbronner, V. Hornung, E. Kloster-Jensen, J. P. Maier, F. Thornrnen, and W. von Niessen, C'iieiii. P/i~..c..,1979, 36, 1. '5 P. Botschwina, P. Sebald. and J.P. Maier, Clieni. Plijx Lett.. 1985, 114. 353. Spectroscopic Structure of Open-shell Poljutomic Cations CI-C=C-H+ A2n-X2n I"' ~''I'" 30600 29bO 29600 28kOO 28boO 27500 27;OO 26500 cni' Figure 5 Laser excitation spectrum qfthe 2'l-I -y213transition qf chloroacetvlene cation (0.005 nm fwhm). The ions were prepared bjs Penning ionization with argon metastables and were collisionally relaxed to T,,, = 100 K. The R = 112 component bands are identified by the bar above the uihrational assignment; the sharp lines are atomic CI-c = C-H+ n u I,0: w am-N d 1.4 1. 6 1.8 2. 0 2. 2 R(C-CI) Figure 6 Potential energy contour lines (in300 cm-' intercals)of chloroacet~~lenecation in its ground andfirst excited state obtained,froni ab initio calculations and sonie experimental data45 In Figure 6 is reproduced a potential energy plot for the X2nand A"'nstates of chloroacetylene The ion is linear in the two states-this is proved from a rotational analysis of the A"2n-%211 the C-Cl and C-C distances tran~ition~~-and are the two coordinates.The C-H bond lengths are fixed. The potential energy functions were obtained from ab initio calculations, with some experimental data (u2, u3 frequencies and the rotational constants) as optimization parameters. The relative positions of the surfaces show the reason for the extensive excitation of the u2 mode in the emission and excitation spectra. On J2n+--X'C+ ionization, the C-C1 and C-C distances increase by 0.13 A and 0.05 A, and in the A"'n-.%?'TI transition the changes are a decrease of 0.20 A (C-CI) and an increase of 0.03 A (C-C).The changes in the C-H distances are small (calculated to Maier be less than 0.01 A). The distance changes have also a marked and corresponding effect on the totally symmetric stretching vibrational frequencies (Table 2) because they are essentially localized: ul(C-H), u,(C-C), u,(C-Cl). Thus the fundamental < u2(A2n)values are in the order u2(f2n) < u2(X1Cf) and u3(A"*II) < u3(X'C+) < u3(f2n),which reflect directly the described bond length changes (cf: Figure 6). On the basis of the u1 frequencies one would expect the C-H distance to be slightly longer in the X2ncompared to the A2Il state. This is borne out by the calculation^.^^ Thus it is seen that the vibrational characterization of the ionic states leads to a detailed insight into the description of the electronic structure.The treatment could be taken further by construction of the force field based on the frequency data available for the various isotopes [35*37C1-C=C-H(D)]. C.Absorption Spectroscopy in Neon Matrices.-In order to characterize ions which do not relax in their excited electronic states by emission of photons, and which are not therefore amenable to study by the methods discussed above, other approaches are required. Two such successful ones are based on direct absorption measurements either of vibrational-vibrational or electronic transitions. The former approach utilizes i.r.lasers and modulation techniques and has been applied to a variety of (mainly closed-shell) ions which can be produced in discharge sources in the gas phase.6 The second approach relies on the measurement of the absorption spectra of ions embedded in rare gas matrices.46 Only a few studies have been carried out in the i.r.,47but the electronic transitions of many large unsaturated organic cations and anions have been observed in freon and some in argon matrices.48 In such investigations the energy positions of the transitions were of interest; information on vibrational frequencies is rather limited. The latter is, however, exactly the focus of our studies, especially for the smaller open-shell cations. The scheme for the generation of the ions in a neon matrix is depicted as part of Figure 1.The molecular precursors are deposited in a low concentration (typically in ratio 1 :5 000) in a 5 K neon matrix and the ions to be studied are produced by photoionization with either a LiF windowed H(Ly x), or an open Ne(1) photon source. Because only 1 in 105-106 of the molecules present in the matrix are ionized, it is necessary to enhance the sensitivity of the absorption measurement. This is achieved by passing the light along the matrix, through its thin side, in a waveguide manner.49 The path length is thus increased ~200fold. As the matrix is deposited on a rhodium-coated copper substrate, the differences in the refractive indices of the adjacent media ensure efficient total internal reflection of the light.The enhancement is sufficient for the absorption spectra of ions to be recorded using standard light sources and modulation techniques. The first ions to be studied by this method were those of the fluorinated 46 V. E. Bondybey, T. A. Miller, and J. H. English, J. Cliem. Phys., 1980, 72, 2193. 47 L. Andrews. Annu. Reti. P1zy.r.Ciienz., 1979, 30, 79. T. Shida, E. Haselbach. and T. Bally, Acc. Chern. Rex, 1984, 17, 180. 49 R. Rossetti and L. E. Brus, Reo. Sci. Insfrum., 1980, 51, 467. Spectroscopic Structure of Open-shell Polyatomic Cations we have used this technique to characterize the ions of substituted (poly-) acetylenes. The vibrational information obtained relates to their excited electronic states since in the 5 K matrix the population is in the lowest vibrational level of the cationic ground state (cf: Figure 1).Most of the earlier studies of the electronic spectra of ions in neon matrices, whether by direct absorption or by laser excitation spectroscopy, were concerned with comparing the inferred vibrational frequencies with the gas phase values for the ions known to fluoresce. The general conclusion is that, for polyatomic cations, the gas and neon matrix frequencies rarely differ by more than 5-10 cm-'. The frequency of the electronic transition itself is significantly shifted owing to the difference in the interactions of the ground and excited state of the ion with the neon matrix. For the polyenes and aromatic cations, where the electronic transitions are of the r'-r'type, a red shift of 10&200 cm-' in the neon matrix is common.Recently, a detailed comparison of the vibrational frequencies of hexafluorobenzene cation in the matrix and gas phase have been disc~ssed.'~ References to the matrix ~~ studies can be found in two reviews.30 I Br -CX-C:NQ/Ne 0.0 4600 4700 4800 4900 5OOO 5100 5200 5300 5400 5500 [&] Figure 7 The absorption spectrum of the A2n3,2 -f2n3 electronic transition qf hromocyanoacetylene cation isolated in % 5 K neon matri-v (spectra/ band-pass 0.1 nm). The ion is generated by photoionization tt,ith Ne(1) radiation of the matrix comprising BrCCCN: Ne % 1 :4 000. The vibrational assignments .for the main site are indicated These aspects are illustrated by comparison of Figures 4 and 7, where the gas- phase laser excitation43 and neon matrix absorption of the B'll,,, +---22113,2 transition of bromocyanoacetylene cation are to be seen.In both spectra only the il = 3/2 component is observed because of the reduced temperature of the ions initially prepared (% lob150 K and 5 K respectively). The vibrational pattern is similar in both spectra; in the matrix several sharp sites are apparent and the assignment is given for the main site. The bands are actually narrower in the matrix because sequence transitions involving the low frequency '"V. E. Bondybey and T. A. Miller in 'Molecular Ions: Spectroscopy, Structure and Chemistry', ed. T. A. Miller and V. E. Bondybey, North-Holland, Amsterdam, 1983, p.125. 5' Y.-C. Hsu, R. A. Kennedy. T. A. Miller, L. A. Heimbrook. and V. E. Bondybey, Mol. Ph!x., 1987.61.225.'' S. Leutwyler. J. P. Maier, and U. Spittel, J. CIwrn. Soc.. Fnririltrj, Trims. 2. 1985. 81. 1565. Maier ,lTTaTl ' 16.0 J 15.6 152 14.8 14.4eV , Figure 8 The excited electronic. states of cyanogen cution rerealed by He(1a) photoelectron spectroscopy ('up-z,80 cm-' resoluiiori und by the absorption spectrum in u 5 K neon mairk (bottom4 cm-optical hand-pass) (Top spectrum redrawn by permission from ref. 16) degenerate modes (u6, u7) are absent; at the higher vibrational temperature (lo& 150 K) of the gas phase excitation spectrum, they are discernible to the red of each progression and combination peak.In the latter spectrum there is also broadening and red-shading due to the rotational structure (Figure 4);lowering the rotational temperature to 5-10 K in a seeded supersonic free jet, leads to the narrow bands observed in the emission spectrum43 (Figure 4, top). In both spectra the prominent excitation is of the u,(C-Br str.) mode because of the considerable increase in C-Br distance during the b211+--B'II transition. The frequencies of the four C stretching modes, u1 to u4, can be inferred (with & 4 cm-'+ uncertainty) from the matrix spectrum; from the gas phase excitation spectrum (Figure 4),u2-u4 and v6, u7 fundamentals have been deduced. The complementary nature of the measurements is evident: when a particular mode is not easily located in the excitation spectrum, it may be discernible in the matrix absorption spectrum, and uice uersa.The matrix absorption technique has recently been used to obtain vibrational information on simple, fundamental, cations for which the only spectroscopic data available came by photoelectron spectroscopy, albeit at low resolution (100-200 cm-'). Cyanogen cation is one such species and the photoelectron spectrum (top trace, Figure 8) yielded the ionization energies of the one-electron accessible Spectroscopic Structure of Open-shell Polyutomic Cations doublet states and a few vibrational frequencies to within +80 ~m-'.~~Higher resolution data are obtained for the C211state of NCCN' by measurement of the C2T13/2-T2T13/2absorption spectrum in a 5K neon matrix (bottom trace Figure 8).54 Although the bands are not as sharp as can be the case in the matrix (cf Figure 7), owing to strong electron-phonon coupling and indistinct sites, the fundamental frequencies of this ion in the c21-13,2state can be deduced to rf: 10 cm-' for the tl; = 2020 and v; = 740 cm-' totally symmetric (El) stretching modes.The photoelectron spectrum yielded the values (k80 cm-') of 2 020 and 710 cm-l re~pectively.~~Similar studies of the absorption spectra of cyanoacetylene and methylcyanoacetylene cations in neon matrices have improved our knowledge of the vibrational frequencies of these ions in their excited electronic states.54 The observed frequency of the origin band of the electronic transition in the matrix can be used to make a good estimate for the expected location of the transition in the gas phase for more detailed studies using a method such as two- photon absorption of mass-selected ion beams5' The above mentioned studies were an intermediate step towards one of the present goals-the spectral characterization of ions known only in mass-spectroscopic measurements, such as isomers and fragment ions.To this end, the ions are first studied in absorption in neon matrices, and then armed with the absorption data, higher resolution gas-phase laser methods are subsequently applied. The first success in this direction has been the detection of the PC; t3°C; absorption spectrum of C: in a neon matrix.56 The ion Cl is invariably involved in the chemical reaction schemes in interstellar clouds, in plasmas, and in flames leading to the formation of hydrocarbons and clusters.It has also been recently detected in the comets Halley and Giacobini-Zinner by in situ mass-spectrometric sampling5 However, spectroscopic information on CT is scarce. A rotationally resolved gas-phase absorption spectrum, of 2C--211 symmetry, has been attributed to C,+,58but this assignment has been questioned in two theoretical ~tudies.'~ There also exists a low resolution measurement of electronic transitions of C: by translational energy-loss spectroscopy, where broad bands were observed6' and assigned by comparison with calculations. C: was produced by a two-step process: acetylene or chloroacetylene was embedded in low concentration in a 5 K neon matrix, C2 was generated by photolysis with a Xe (147 nm) or H(Ly a)(121.6 nm) source, and C: by subsequent ionization with Ne(1) (73.6 nm) radiation.The absorption spectrum reproduced in 53 C. Baker and D. W. Turner, Proc. R. SOL..London A, 1968, 308, 19. 54 J. Fulara, S. Leutwyler, J. P. Maier, and U. Spittel, J. Phys. Chem., 1985, 89, 3190. 55 P. 0. Danis, T. Wyttenbach, and J. P. Maier, J. Chem. Phys., 1988, 88, 3451. 56 D. Forney, H. Althaus, and J. P. Maier, J. Php Ckem., 1987. 1987, 91, 6458. 57 D. Krankowsky, P. Lammerzahl, 1. Herrwerth, J. Woweries, P. Eberhardt, U. Dolder, U. Herrmann, W. Schulte, J. J. Berthelier, J. M. Illiano, R. R. Hodges, and J. H. Hoffman, Nafure, 1986, 321, 326; M.A. Coplan, K. W. Ogilvie, M. F. A'Hearn, P. Bochsler, and J. Geiss, J. Geophys. Res., 1987, 92, 39. '* H. Meinel, Can. J. Phys., 1972, 50, 158. 59 C. Petrongolo, P. J. Bruna, S. D. Peyerimhoff, and R. J. Buenker, J. Chen?. Phys., 1981, 74, 4594; P. Rosmus, H.-J. Werner, E.-A. Reinsch, and M. Larsson, J. El. Spectrosc. Rel. Phenom.. 1986, 41, 289. 6o A. O'Keefe, R. Derai, and M. T. Bowers, Chem. Phys., 1984, 91, 161. Maier 2-0 1-0 0-0 ' c; :B4t;-ji4tg I1 I I 1 I I I 23000 22000 21000 20000 Wcm' Figure 9 The 84Zi -24Z-absorption spectrum of C: in a 5 K neon matrix. C: was produced from ClCCH deposited in the matrix by Xe(1) (147 nm) photolysis&llowed by Be(!)(73.6 nm) photoionization. C; is concomitantly ,formed and two of its B2Z: -X2C, absorption bands are labelled Figure 9 56 is identified as the 84C; c-8"Cg-transition of C; on the basis of chemical evidence (using different precursors), isotopic shifts (using 3C, D derivatives), and by comparison with the calculated term values and vibrational frequencie~.~~The agreement with the latter is excellent.Based on the matrix data this transition was found in the gas phase by the laser excitation approach, and the rotational structure proves the correctness of the assignment.61 These experiments on C: pave the way for characterization of other fragment ions in the matrix and gas phase. 3 Rotational Characterization In the case of the linear open-shell cations, predominantly those of tri- and tetra- atomics, it has proved possible to resolve the rotational structure in their electronic transition^.^^ Most of such studies have used the laser excitation approach; however, emission spectroscopy using electron excitation has proved valuable in two areas outlined in the next section.As far as triatomic cations are concerned, rotational analyses of the emission spectra, and some absorption spectra obtained by classical techniques, have been carried out in the past for CO;, CS; (B'C:,~2n,-?2n,),N,O+(J2C+-f2n), H,O+, H2S' (~2A,-~2Bl)and of some of their isotopic derivatives62 and for diacetylene cation.63 A. Emission Spectroscopy.-The spectral characterization of the ions of unstable precursors, XCP', XBS+ (X = H, D, F, Cl), has so far only been possible in emission (cf: Figure 3).However, in two of the studies, the attainable resolution (z0.005 nm) was sufficient to resolve the rotational structure in the emission band 61 M. Rosslein. M. Wyttenbach. and J. P. Maier, J. Chem. Phjs., 1987, 87, 6770. b2 See S. Leach, in 'Spectroscopy of the Excited State', ed. B. di Bartolo, Plenum Press, New York, 1976, p. 369; R. J. Saykally and R. C. Woods, Annu. Rev. Phyx Chem., 1981, 32, 403.for references therein. 63 J. H. Callomon, Cm J. Phjs., 1956, 34. 1046. Spectroscopic Structure of' Open-shell Poljwtomic Cutions IIIIII11I 1 1 1 f If1 1 591 592 593 594 nm I1 I I I I1 1 1 11 I 1 1 I I I I1 597 598 599 nm Figure 10 Rotational structure qf'the origin band in the A2C' -p2rlemission spectrum of' deuteropiiospiiuethL.ne cation (0.008nm fwhm).The .spectrum \ius obtained bj. z 40 eV electron impact on an effusiiv source of the niolc~cularprecursor Table 4 Rotational constants (cm-' ) of tire polyatoniic cations inferred recenfly ,from their laser escitafion (LE)or emission (E) dectronic The constant ,for the most abundant naturallv occurring isotope is girrn. Referenws to thc cwnstants of the earlirr studied triatoniic cationsb' arc to he ,fiiund in t1r.o reriett~P2 Bcff .o Cutiori .f2n3,2 2n32 RPf. OCS' 0.194 66(8) 0.186 89(8) (A) LE 73 ClCN+ 0.204433 (50) 0.17660(13) (B) LE 74 79BrCN+ 0.141 612 (44) 0.127 474 (45) (B) LE 75 CICCH' 0.194 647 (49) 0.170 881 (48) (A) LE 34 "BrCCH+ 0.137 794 (37) 0.121 351 (36) (A) LE 69 ICCH' 0.10959(7) 0.09669(7) (2) LE 72 Bo B'n A HCP' 0.622 4( 16) 0.669 O( 17) ('X') E 64 HBS' 0.5760(2) 0.614 X(2) ('X') E 65' H(CC),H+ 0.146 90(4) 0.140 09(4) ('n) LE 70 Values in brackets correspond to one standard deviation. * CO, *, CS2+,N,O+, H,O+, H,S +.'In the given reference, these constants were incorrectly attributed to states of BS systems.These were the A2C+---f2Iltransitions of HCP+, DCP+,64 HBS', and DBS+.4' Such a recording for DCP+ is reproduced in Figure 10 where the assignment of the rotational lines is indicated. From the analysis the rotational constants for HCP+. DCP' were derivedh4 and these are to be found in Table 4. h4 M. A. King. D. Klapstem. H. W. Kroto. J. P. Maier. and J. F. Nixon. J. Mol. Spci,Irosc..1982. 80, 23. 62 Muier The values given there for HBS’ were not determined in our studies, because the rotational constants were already available in the literature;6s they were, however, incorrectly attributed to two electronic states of BS of ’C+ and ’TI symmetry. The vibrational and isotopic (DBS’) analyses of the emission spectra showed conclusively that the reported band system and constants are of the J’C’ and X217 states of HBS+.41 Higher resolution studies of the ions XCP’, XBS’ await the application of the laser excitation technique, but are hampered by the difficulties of preparing such species in sufficient quantities. The second area where the emission technique, in combination with supersonic free jets, has proved successful is in the study of the rotational structure in the electronic transition of di- and tri-atomic ions at low temperatures.Thus the rotational pattern of vibronic bands of the J2A,-f2B, transition of H20f was investigated 66 under the temperature conditions encountered in extra-terrestrial environments, such as in comets, where this fluorescence has been detected.67 The ‘desired’ rotational temperature is tuned by choice of the expansion conditions and is determined by comparison with measurements of the rotational distribution of transitions such as 82C~-~2Z:of Ni, or A’C+--X’II of N,O+. H~O+ii2~, n+X28, 080~000 Trot-100 K I.... I..,.l.,.,,,,, 16300 16200 16100 v (cm-’) Figure 11 The rotutionul structure of the 080-400 Il sub-hand in the A’ ’A, -T2B, er~ission spectrum of H,O+ at two reduced rotational temperatures, obtained 61, z200 eV electron irtipuct on a seeded heliuni supersonic free jet (0.045 nm resolution). The chosen teniperutures are typical of coniet tails In Figure 11 is shown the rotational structure of the 84 l7 sub-band within the J2A, -W’B, emission system of H20+ at rotational temperatures of = 50 K (bottom trace) and = 100 K (top).The temperatures were actually evaluated from the pattern of the C components of the 84 transitions, for reasons which are discussed in the original study.66 The temperatures of 50 and 100 K were chosen because these correspond to comet tail measurements of these bands at particular heliocentric distances. Some such cometary spectra can be found in reference 68.‘’ J. K. McDonald and K. K. Innes. J. Mol. Spcctrosc., 1969, 29, 251. 66 S. Leutwyler, D. Klapstein, and J. P. Maier, Ciimi. Phj,.~..1983. 74. 441. 67 P. A. Wehinger. S. Wyckoff. G. H. Herbig, G. Herzberg, and H. Lew. Astropiiys. J., 1974, 190, L43. 63 Spectroscopic Structure of Open-shell Poij3atomic. Cations p79 25.5 '81 l.-"l'"~I.~~'l 20 545 20 540 20 535 V/cm-1 Figure 12 Rototionullj3 resoli~~dorigin hmd (0.0008 nm fwhm) in the A"lI 3, Iuscr e.Ycitution spectrum of deuterobroniouc.i.tj.Iene cution. The ions (ire produced by Penning ionizution and collisional relusutioii (T,,, = 100 K). The 79 arid 81 Iuhels identill, rhe tn'o equallj, ahundunr hroniine isotopic derii.ntii;c.s Prior to the experimental observations at the reduced rotational temperatures, the line intensities were determined using the known rotational constants of H20+ and calculated line strengths of the 2A1-2B, transitions.68As our study showed,66 however, there are considerable differences between the predicted and observed spectra, probably due to uncertainties in the calculated line strengths.One may now instead use the experimental data (Figure 11) in a direct comparison with the cometary spectra for temperature and other determinations. This example illustrates how the techniques can be used to provide information on physicochemical parameters of inaccessible environments. Related applications in the laboratory can be envisaged for plasma and flame-characterizations using ions which are often present in such media.B. Laser Excitation Spectroscopy.-The laser excitation approach lends itself favourably to studies of the rotational structure of electronic transitions because, unlike the situation with the emission approach where the intensity limits the useable resolution of the monochromator, narrowing the laser band-width with an intracavity etalon does not cause deterioration in the signal-to-noise ratio of the experiment. In Figure 12 is shown the rotationally resolved origin band in the J'II,,, -2'113,,excitation spectrum of deuterobromoacetylene cation recorded with 0.03 cm-' laser band pass.69 The ions were produced by Penning ionization with argon ''S. Wyckoff and P. A. Wehinger, Astrop/ijx J..1976. 204, 616. 69 M. A. King, J. P. Maier, L. Misev, and M. Ochsner, Coii. J. flr~x.1984. 62, 1437 64 Maier metastables and were collisionally relaxed and thus only the R = 3/2 sub-component bands were detected because the spin-orbit splitting in the ground cation state is z -1 000 ~m-’.~~The rotational and vibrational temperature of the ions is estimated as 100-150 K. The rotational structure is composed of two overlapping R and two P branches due to the two naturally occuring isotopic forms, 79Br-CrC-D+ and 81Br-C=C-D+. The head is in the R branch and the band is red-shaded (i.e. B < B”); the Q-branch is too weak to be detected at the temperature of the measurement but should, however, become apparent by lowering the temperature to below about 20 K (as has been demonstrated on the J2n--X211transition of diacetylene cation conducted in a supersonic free jet7’).The assignment of the lines in the respective branches is indicated in Figure 12. Such recordings have been made on a number of bands of the v;(C-Br str.) progression for both HCCBr+ and DCCBr +. The rotational analyses show that the ions are linear in the X2nand J217states. The rotational and related constants were obtained by fitting the observed line positions to the differences of the eigenvalues of lower and upper state effective Hamiltonian matrices. The actual expression used for each state is derived from the diatomic matrix elements71 and second order perturbation correction for the R = 1/2 sub-component which is not sampled in the e~periment.~~The Beff..’values are given in Table 4 for the most abundant isotopic variants of bromoacetylene cation, as well as for ~hloro-~” and iodo-a~etylene’~ cations where corresponding measurements and analyses have been completed.A complete rs structure of the halogenoacetylene ions in the ground and first excited electronic states is not yet available as only the terminal atoms were isotopically substituted (i.e. either H/D or Br, 35,37C1). However, the data 79381 obtained show, for example for HCCBr’, that the average H-Br distances are 3.94 and 4.07 8, in the 2’I-Iand J2nstates respectively, as compared with 4.05 8, for the neutral molecule.69 As has been discussed for the chloroacetylene ion earlier, the dominant reason for the difference in the H X distance in the two states is the change in the C-X distance (cJ: Figure 6).Apart from the halogenoacetylene cations, the laser excitation approach has been used to obtain the rotational structure of the origin bands in the J2113,2-X’lI312 transition of OCS+,73and of the J211RX217one of H(CC),H+ and Most recently the B2l7--?’ll band systems of ~hloro-’~ D(CC)2D+.70 and bromo- cyanide75 cations have been rotationally analysed (cf: Table 4). ’’R. Kuhn, J. P. Maier, and M. Ochsner, Mol. P/7jx, 1986, 59,441. ” R. N. Zare, A. L. Schmeltekopf. W. J. Harrop, and D. L. Albritton, J. Mol. Spectrosc., 1973, 46, 37. 72 J. P. Maier and M. Ochsner, J. Chem. Soc..Faradq Trans. 2, 1985, 81, 1587. 73 M. Ochsner, M. Tsuji, and J. P. Maier, Chern. Phys. Lett., 1985, 115, 373. 74 F. G. Celii, M. Rosslein, M. A. Hanratty, and J. P. Maier, Mol. Pkys., 1988, 62, 1435. is M. A. Hanratty, M. Rosslein. F. G. Celli, T. Wyttenbach. and J. P. Maier. Mol. P/~J:F.,1988, in press. Spectroscopic Structur-e of'Opeti-slicdl Poljvtoniic Cutions 4 Outlook Concurrent with the spectral characterization of open-shell cations by the methods described above has been the development of other approaches for probing such ions at higher resolution and of the different detection schemes required for other ions. In the former category belongs stimulated emission pumping. This technique has proved useful for the spectroscopic studies of stable molecules in highly excited vibrational levels of the ground state manifold.76 We have recently shown, with experiments on diacetylene cation, that this is also a viable and attractive approach for the study of reactive species produced in small concentrations, such as open- shell The method, which depends on the fluorescence of the ions in the excited electronic state, uses two laser wavelengths.One is used to transfer the population from the ground to the excited state, whereas the second laser stimulates the transitions to the chosen level of the ground state. Detection is by measuring changes in the fluorescence intensity from the excited state. One of the features of interest of the stimulated emission pumping method is that it opens the way for the study of vibrational levels of the ions in their ground states at higher resolution than is attainable with the emission approach using a monochromator.Furthermore, as the transitions to the ground state manifold are driven by the laser, many more vibrational levels become accessible. Thus precise vibrational frequencies of the levels, and for the smaller ions the rotational constants, can be determined. The double resonance nature of the measurement is often useful in assignment clarifications. It is envisaged that the method could also be used for the preparation of ions in selected vibrational levels for subsequent ion- molecule reactions, in a similar way to that demonstrated for molecules.7* The major effort using the matrix absorption and laser excitation techniques is at present directed towards spectral characterization of ions known so far only by mass spectrometry.The first step in this direction has been the identification of the 84C,-84Cg-transition of Ci in absorption in a 5 K neon matrix.56 Based on this observation the corresponding laser excitation spectrum of C: in the gas-phase was found and rotationally analysed.61 The use of the matrix and gas-phase techniques in tandem is clearly advantageous for the characterization of other simple fragment ions. Another approach with this aim, but also for ions which do not decay radiatively, is centred on a mass-selected ion beam in conjunction with laser excitation. Several groups have successfully employed such an approach where, with the exception of a few studies which used charge-transfer reactions to monitor the electronic tran~ition,~~the detection relied on the predissociation of the excited state.*' The approach we have been pursuing is more general: the electronic transition of a mass- 76 C.E. Hamilton, J. L. Kinsey, and R. W. Field, Aitiiu. Rw. P/ij:c.. C/i~i77.,1986. 37. 493. 77 F. G. Celii. J. P. Maier. and M. Ochsner, J. ChP171. Phjx. 1986, 85, 6230. 7'S. H. Kable and A. E. W. Knight, J. C/i~in.P/ij.s.. 1987. 86, 4709. 79 A. Carrington. D. R. J. Milverton. and P. J. Sarre, Mnl. PIijx, 1978, 35. 1505. See for example S. P. Goss, R. G. McLoughlin, and J. D. Morrison, Ini. J. Mass Spc,cfroni. loit Phjx, 1985. 64, 213.66 Maier selected ion beam is induced by one laser wavelength and a second laser colour then produces a daughter ion. The latter is detected as the first laser wavelength is scanned to record the spectrum of the chosen ion. The apparatus is a triple quadrupole system with a high pressure source to produce collisionally relaxed ions for the mass selection. The feasibility of this method has now been demonstrated on the e2Zlf--B2Zc,f t-f2n,transitions of CST-the C2Z; state leads to the production of S ions which are monitored-and on analogous transitions of + N20+ and CO, +.” This paves the way for the spectral characterization of a variety of ions which can be generated by fragmentation and ion molecule reactions. Acknowledgements. It is my pleasure to acknowledge the contributions and efforts of PhD students, M. Allan, D. Forney, R. Kuhn, 0. Marthaler, L. Misev, M. Ochsner, M. Rosslein, U. Spittel, F. Thommen, T. Wyttenbach and of postdoctoral fellows, Drs. F. Celii, P. Danis, J. Fulara, M. Hanratty, M. A. King, D. Klapstein, S. Leutwyler, in realising the research projects outlined here. The financial support has been generously provided throughout the years by the ‘Schweizerischer Nationalfonds zur Forderung der wissenschaftlichen Forschung’, and I wish to thank Professor E. Heilbronner for our long-lasting, fruitful, and harmonious collaboration within this project.

ISSN:0306-0012    

DOI:10.1039/CS9881700045

出版商:RSC    

年代:1988

数据来源: RSC

摘要:

Cheni. SOC.Rev., 1988, 17, 69-90 Macrocyclic Schiff Base Complexes of Lanthanides and Actinides By D. E. Fenton DEPARTMENT OF CHEMISTRY, THE UNIVERSITY, SHEFFIELD S3 7HF P. A. Vigato ISTITUTO DI CHIMICA E TECNOLOGIA DEI RADIOELEMENTI, CNR-PADOVA, ITALY 1 Introduction It would not be a total exaggeration to state that macrocyclic complexes lie at the centre of life, particularly when regarding the roles of such systems as the iron- porphyrin core in haemoglobin, the cobalt-corrin of vitamin B,, and the magnesium-hydroporphyrin in chlorophylls. Recognition of the importance of complexes containing macrocyclic ligands, where a macrocycle is defined as a cyclic compound having nine, or more, heteroatomic members and with three, or more, ligating atoms,, has led to a considerable effort being invested in developing reliable syntheses for these compounds.Ideally the complex is formed by adding the required metal ion to a preformed macrocycle but there are potential disadvantages to this approach: the synthesis of the macrocycle often results in a low yield of the desired product with side reactions such as polymerization predominating. In order to circumvent this problem the ring-closure step in the synthesis may be carried out under conditions of ‘high dil~tion’,~ or ‘rigid groups’ may be introduced to restrict rotation and internal entropy losses in the open-chain precursor^^-^ and so facilitate cyclization. One effective method for the synthesis of macrocyclic complexes involves an in situ approach wherein the presence of a metal ion in the cyclization reaction markedly increases the yield of the cyclic product.The metal ion plays an important role in directing the steric course of the reaction and this effect has been termed the metal-template effect.’ The first example of a deliberate synthesis of a macrocycle using this procedure was described by Thompson and Busch (Scheme 1),8 although Curtis had previously demonstrated the potential of template assembly through his observations that the reaction of Ni(en),(ClO,), (en = 1,2-diaminoet hane) and acetone yields isomeric tetra-azamacrocyclic complexes of ’ M. N. Hughes, ‘The Inorganic Chemistry of Biological Processes’, 2nd edn., Wiley. N.Y., 1981.’ ‘Co-ordination Chemistry of Macrocyclic Compounds’, ed.G. A. Melson, Plenum, N.Y., 1979. K. Ziegler, H. Eberle, and H. Ohlinger, Liehigs Aiiizcden, 1933, 504, 94. W. Baker, J. W. F. McOmie. and W. D. Ollis, J. Chem Sou.. 1951, 200.’ E. L. Eliel, ‘Stereochemistry of Carbon Compounds‘, McGraw-Hill, N.Y., 1962.’B. L. Shaw, J. Aiii. C/iciii.Soc., 1975, 97, 3856. L. F. Lindoy and D. H. Busch in ‘Preparative Inorganic Reactions’, ed. W. L. Jolly, Wiley-Interscience. N.Y.. Vol. 6, 1971, p. 1.’M. C. Thompson and D. H. Busch. J. Am. Chetn. Soc., 1964. 86,3651. Macrocyclic Schlff Base Coniple.ues of Lanthanides and Actinides A H2N\ /5 /Ni\sH2NW Scheme I aHead Unit N\ Lateral N N RA 1Unit[N N Figure I Scheimiic represenrarions oftlie '1 + 1' and '2 + 2' Scli(ff base macrocycles nickel(r1) (Scheme 2),9 and metal salts had been shown to facilitate the seif- condensation of o-phthalonitrile to give metal-phthalocyanine complexes (Scheme 3).'O Much of the early work featured the use of transition metal ions in the template syntheses of tetradentate macrocycles: the directional influence of the orthogonal d-orbitals was regarded as instrumental in guiding the synthetic pathway.The last decade has seen an extension of this technique to include the use of organo-transition metal derivatives to generate tridentate cyclononane complexes' '-'' 'D. A. House and N. F. Curtis. C/wu. hi.,1961, 1708. I" R. P. Linstead and A. R. Lowe, J. C/ion. Soc., 1934, 1022. D. Sellman and L. Zapf. Aqyir..C/wm..//if. Eth. EngI.. 1984. 23. 807. l2 B. N. Diel, R. C. Haltiwanger. and A. D. Norman. J. Am. Cheni. Soc.. 1982, 104, 4700. Fenton and Vigato j N’ ‘N H2 H2 Scheme 2 M 2+aCN-CN Scheme 3 and an expansion to the s-and p-block cations which guide the synthesis of penta- and hexa-dentate Schiff base macrocycles’ 3,14 and a range of tetraimine Schiff base macr~cycles.’~.’~The smaller Schiff base macrocycles have been termed ‘1 + 1’ macrocycles and the tetraimine derivatives ‘2 + 2’ macrocycles as a consequence of the number of head and lateral units present (Figure 1).l5 The metal complexes of the ‘2 + 2’ macrocycles may be mono- or di-nuclear in nature. It has been generally found that for the larger Schiff base macrocycles the transition metal cations are ineffective as templates.’ Consequently the kinetic lability of the metals present in the generation of macrocyclic complexes derived from the use of alkaline earth and main group templating agents has enabled the generation of the corresponding transition metal complexes through trans-metallation reactions (Figure 2).’3-1 This approach has been particularly successful when applied to the generation of dinuclear copper(r1) complexes of l3 S.M. Nelson. S. G. McFall, M. G. B. Drew, and A. H. Othman, Proc. R. Ir. Acotl.. Swt. B, 1977.77, 523. (N) D. H. Cook and D. E. Fenton, J. Chem. Soc., Dril/oii Trms., 1979: 266 (h) D. H. Cook and D. E. Fenton. J. C‘lieni. Soc.. Dolton Trons., 1979. 810. Is S. M. Nelson, Pure Appl.Clteiii., 1980, 52, 2461. l6 D. E. Fenton, Pure Appl. Clieni., 1986, 58, 1437. Macrocyclic Schiff Base Conip1e.ue.s of’ Lunthanides and Actinides tetraimine Schiff base macrocycles which have then been used as speculative models for the bimetallobiosites in cupro-proteins such as haemocyanin and tyrosinase. The size of the cation used as a template has proved to be of importance in directing the synthetic pathway in the Schiff base systems (Figure 3). Of the alkaline earth cations only magnesium generates the pentadentate ‘1 + 1’ macrocycle but it is ineffective in generating the hexadentate ‘1 + 1’ macrocycle which is readily synthesized in the presence of the larger cations calcium, strontium, barium, and lead(rr). These cations, however, generate the ‘2 + 2’ macrocycle derived from the components giving the ‘I + 1’ macrocycle with magr~esium.’~“.~.~~ A further size-related effect is metal-induced ring-contraction whereby if the metal ion is too small for the macrocyclic cavity and there is a functional group (=NH or -OH) available for addition to the imine bond then this can add to produce a smaller, more accommodating cavity for the available metal (Figure 4).9*20 The similarity in ionic radii between the alkaline earth metal cations and the lanthanide(ir1) cations suggested that the latter should also be efficient templating devices (Table 1). Furthermore the steady diminution in ionic radius in the lanthanide series (the ‘lanthanide contraction’) could perhaps facilitate a fine control in the synthetic pathway. This article reviews the progress made in the development of lanthanide and actinide complexes of Schiff base macrocycles through application of the above techniques.2 Lanthanide Complexes \ \N N (1) (2) (3) A. ‘2 + 2’ Macrocycles derived from Pyridine Head Units.-(i) Aliphatic Luterui Units.The early work concerning the use of lanthanide templates in the synthesis of ‘2 + 2’ tetraimine Schiff base macrocycles suggested that the potential application was limited. The reaction of hydrated lanthanide nitrates, 1,2-diaminoethane, and 2,6-diacetylpyridine in 1 :2: 2 molar ratio in alcohol gave the macrocyclic I’ S. M. Nelson. J. Trocha-Grirnshaw, A. Lavery, K. P. McKillop. and M. G.B. Drew. in ‘Biological and Inorganic Copper Chemistry, Vol. 11.’ ed. K. D. Karlin and J. Zubieta. Adenine Press. N.Y.. 1986. p. 27. lXD. H. Cook. D. E. Fenton. M. G. B. Drew. A. Rodgers, M. McCann. and S. M. Nelson, J. Chrvii. Soc.. Dultoti Trtrtis., 1979. 4 14. l9 M. G. B. Drew. J. Nelson, and S. M. Nelson, J. Chctn. Soc., Dolrorz Trtrtis., 1981, 1678. *’ N. A. Bailey, D. E. Fenton. R. J. Good, R. Moody. and C. 0.Rodriguez de Barbarin, J. Clroii. Soc.. Dtilroti Trciti.c.. 1987, 207. Fenton and Vigato Figure 2 Trunsrnetullution reaction 0 0 H,N A0 (AJnANti2 W n = 2,3 Figure 3 Schiff base niucrocycle synthesis in the presence of non-transition metal templates Macrocyclic Schiff Base Complexes of Lanthanides and Actinides (a) (b) Figure 4 Metal-induced ring contraction.The larger barium cation gives (a) as the product from the reaction of 2,6-diacetylpyridine and 1,3-diamino-2-hydro.~ypropane,whereas the smaller lead(I1) cation gives (b) Table 1 Ionic radii of alkaline earth metals and lanthanides (A) Mg2+ 0.72 Pr3+ 1.013 Dy3+ 0.908 +Ca2 1.00 Nd3+ 0.995 Ho3+ 0.894 Sr2+ 1.16 Sm3+ 0.964 Er3+ 0.881 +Bat 1.36 Eu3+ 0.950 Tm3+ 0.869 La3+ 1.061 Gd3+ 0.938 Yb3+ 0.859 +Ce3 1.034 Tb3+ 0.923 Lu3+ 0.848 (Compiled from R. D. Shannon and C. T. Prewitt, Acta Crysrallogr.. Secf. B, 1969, 25, 925) complexes [Ln( l)](N03)3 only for lanthanum and cerium." The heavier lanthanides were not found to be effective templating agents in the formation of this 18-membered ring, perhaps due to their smaller They were effective, however, as templating agents in the synthesis of the 14-membered ring (2) in which h ydrazine provides the lateral units; the complexes [Ln( 2)( H 0)2] (C10,) ,4H 0 are recovered when Ln = Tb, Dy, Ho, Er, Tm, Yb, or Lu.Employment of the lighter lanthanides as templates gave complexes of the acyclic derivative (3) and, in some experiments, lanthanide complexes of 2,6-diacetylpyridine. The discrete nature of [Ln( l)](N03)3 was confirmed by X-ray crystallography, and conductivity studies showed that the complexes of (1) existed mainly as [Ln( 1)(H20),]3 + in aqueous solution. The macrocycles were resistant to hydrolysis, and H n.m.r. experiments showed that the complexes were remarkably stable to dissociation in D20, unlike the corresponding complexes of cyclic 21 J.D. J. Backer-Dirks, C. J. Gray, F. A. Hart, M. B. Hursthouse, and B. C. Schoop,J. Chem. Soc., Chem. Coniniun., 1979, 774. 22 W. Radecka-Paryzek. Inorg. Chim. Acta, 1980, 45, L147. 23 G. Wang and L. Miao, Guodeng. Xuesiao HuaXua Xuebao. 1984,5,28;Chem. Abstr., 1984,101,182591~. 24 W. Radecka-Paryzek, Inorg. Chim.Acta, 1981, 52, 261. Fenton and Vigato polyethers. No precipitation of the cation was detected on treatment of the complex with aqueous solutions of KF or KOH. In contrast, the complexes of (2) ring-opened on reaction with water to give the acyclic complexes [Ln(3)(H20),]- (C10,)2,2H20, and in some experiments lanthanide complexes of 2,6-diacetyl- pyridine were also recovered.More recent surveys of the template potential of the lanthanides have indicated a greater breadth to the reactions. Using 2,6-diformylpyridine as the dicarbonyl precursor, complexes of macrocycle (4)were recovered for all of the lanthanides save pr~methium.~' The i.r. spectra of the complexes of the heavier lanthanides (Nd-Lu, except Pm and Eu) were different from those for the lighter lanthanides (La-Pr and Eu) in that the former group exhibited a distinctive sharp band at ca. 3 220 cm-', assigned to a secondary amine group. Addition of a water molecule across an imine double bond leads to the formation of a carbinolamine species (5), as has previously been noted for some macrocyclic complexes of transition metals.26 This is accompanied by an increase in the flexibility of the macrocycle, making it capable of accommodating the smaller lanthanide cations.Such a process could provide fine-tuning of lanthanide complexation. The carbinolamine complex may be shown to exist in solution by 13C and 'H n.m.r. studies on the lutetium derivative. A significant change occurred on recrystallization of the samarium complex of (5). The i.r. now showed that the amine band had been replaced by a new band at 3 560 cm-' ascribable to a Sm-OH group, and the X-ray structure of the product was found to consist of the discrete complex cation [Sm(4)(N03)(OH)- (H20)]+, a nitrate anion, and clathrated methanol molecules.27 During the course of the recrystallization there has been a reversion of the macrocycle to the tetraimine form.It is possible to consider that optimal cavity- cation criteria have been met and that the samarium can actually be accommodated by either form of the macrocycle. An alternative explanation is that the bulk product isolated is a complex of a carbinolamine precursor to the tetraimine Schiff base, with the former being the kinetically-favoured product and 25 K. K. Abid and D. E. Fenton, Inorg. Chini. Actci, 1984, 95, 119.'' ((0Z. P. Haque, M. McPartlin, and P. A. Tasker, horg. Clzem., 1979, 18, 2920. (b) C. Cairns, S. G. McFall, S. M. Nelson, and M. G. B. Drew, J. Clzei??.Soc., Dalton Tmns., 1979, 446. 27 K. K. Abid. D. E. Fenton. U. Casellato, P. A. Vigato, and R. Graziani,J. Cltern.Soc..Dalton Trcms..1984, 351. 75 Macrocj tclic SchifrBase ConipIe.ues of Lanthanides urid A c'tinides the latter being the thermodynamically-favoured product. The carbinolamine complex would represent the stabilization of a reaction intermediate cia the facile co-ordination of that species to the smaller lanthanide cations. On dissolution in water and recrystallization a higher temperature is reached than in the original reaction in alcohol, and this helps the reaction move to completion. The samarium complex is also hydrolysed during the recrystallization process from Sm[(4)](N03), to [Sm(4)(N0,)(OH)(H,0)]N03. It is plausible that on prolonged exposure to water the following sequence occurs: HO Sm(L)(NO,), [Sm(H,0),J3+ + L + 3N0,-Sm(H20,J3+ e[Sm(H,O),- ,(OH)]'+ + H30f [Sm(H,0),-,(OH)]2+ e[Sm(L)(OH)]Z' + (n -l)H,O This seemingly facile hydrolysis of lanthanide complexes has been observed in related systems bearing compartmental ligands,' and is also noted in macrocyclic complexes derived from 2,6-diacet~lpyridine.,~ Tetraimine Schiff bases derived from 2,6-diformylpyridine and 1,2-diamino- propane, and from 2,6-diformylpyridine and 1,3-diaminopropane have been isolated for the lanthanides (except promethi~m).~' With the latter complexes no evidence for carbinolamine intermediates was found.More recently the template procedures have been used to extend the range of complexes available from 2,6-dia~etylpyridine.~~-~'Complexes of the general formulae Ln( l)(MeCOO),Cl,nH,O (n = 3-6) and Ln( 1)(C104),0H,nH,0 (M = Ck-2) have been recovered for all of the lanthanides except pr~methium.,~ This contrasts with the earlier results; the ease and yield of the reaction depends on the counterion present.Good oxygen-donor anionic ligands such as MeCOO -favour the reaction more than CI-or ClO,-. Once the complex is formed, then an appropriate combination of anions (e.g. MeCOO- and C1-, or C104- and OH-) favours isolation in a crystalline form. In this work it was not found possible to reproduce the preparation of La( 1)(C10,),,2H20; instead the compound La( 1)(C104),0H,nH,0 was isolated consistently. Lanthanide complexes of some of the heavier lanthanides have also been prepared by a transmetallation reaction in which the alkaline earth complex Ba( 1)(CIO4), exchanges metal with the appropriate lanthanide to give two types of complexes: [Ln( 1)(N03),(H20)]N0,,H,0 (Ln = La, Ce, or Pr) and [(Ln(l)- (N0,)(H,0),},]N0,[CI04]3,4H,0(Ln = Nd, Sm, Eu, Gd, Tb, Dy, Ho, or Er).30 The smaller lanthanides appear to have too small an ionic radius to co- ordinate effectively with the macrocycle in competition with barium.The large ionic radii of the lanthanide ions, together with the electrostatic nature of the metal-ligand bonding in their complexes, leads to the generation of high l8 K. K. Abid, Inorg. Chrtii. Acto, 1985. 109, L5.'' L. De Cola. D. L. Srnailes, and L. M. Vallarino, Itiorg. Chen?.. 1986, 25, 1729. 3" A. M. Arif, J. D. J. Backer-Dirks,C. J. Gray.F. A. Hart. and M. B. Hursthouse. J. Cheni. Soc... Ddroti T~NIIS.,1987. 1665 Fenton and Vigato co-ordination numbers., The use of macrocyclic ligands such as cyclic polyethers has been said to have brought a new dimension to lanthanide chemistry as it allows study of highly co-ordinated compound^.^ la X-Ray crystal structural information is available for the complexes [Ce( l)(N0,),(H,0)]N0,,H,0,30 [{Nd( 1)(N03)- (H20),)2]N0 [CIO,] ,,4H20,,' and [Lu( 1)( MeCOO)( OH)( ClO,)],MeOH, nH2032 (Figure 5), as well as for the Ln(1) and Sm(4) complexes previously mentioned. The structures of the complexes derived from (1) show a decreasing coordination number for the metal from La-Lu: La = 12, Ce = 11, Nd = 10, and Lu = 9. The samarium complex derived from (4) has the samarium present with co-ordination number 10.For descriptive purposes the macrocyclic plane may be viewed as generating two hemispheres, one above and one below the donor atom plane (Figure 5a), and the anions or accompanying ligands will occupy sites within these hemispheres. The lanthanum ion in [La( l)][NO,), has the uncommon co-ordination number of 12.21 Although the icosahedron. is generally regarded as the most favoured polyhedron for 12-co-0rdination~~~ the co-ordination environment here consists of the six nitrogen donors in a girdle around the metal with two bidentate nitrates on one side of the macrocycle and the other bidentate nitrate on the opposite site, as has been found in the hexaoxamacrocyclic complex [Ln( 1~-c~ow~-~)](NO~)~.~~ The ligand is folded away from the hemisphere containing the extra nitrate anion.In the cerium complex the unusual co-ordination number of 11 is noted for the metal ion which is co-ordinated as follows: to the six nitrogen atoms of the macrocycle, to one bidentate nitrate anion on one side of the ligand plane, and to the remaining nitrate and the water molecule on the other side of the plane.,' The macrocycle is folded away from the hemisphere containing the extra water molecule. With a single exception, all lanthanide complexes known to have 1 l-co- ordination contain the bidentate nitrate anion, and so a distorted polyhedral structure is the rule and not the exception. It has been remarked that it is futile to try to define such complexes in terms of a particular p~lyhedron.~~ The co-ordination numbers 9 and 10 are more commonly observed in lanthanide complexes and have previously been reported for 18-crown-6 complexes of neodymium and gad~linium.~~" In the neodymium complex the metal is 10-co- ordinated by the six macrocyclic nitrogen atoms, one bidentate nitrate anion on one side of the macrocyclic plane and to two water molecules on the other side.The ligand is folded away from the hemisphere containing the two water molecules, clearly illustrating the smaller steric requirement of the nitrate to the two monodentate ligands. " ((I) J. C. G. Bunzli and D. Wessner, Coord. Cliem. Rev.. 1984,60. 191; (h)F. A. Hart, in 'Comprehensive Co-ordination Chemistry', ed.G. Wilkinson, R. D. Gillard, and J. A. McCleverty, Pergamon Press, Oxford, 1987, Chap. 39. 32 G. Bombieri. F. Benetollo, A. Polo, L. De Cola, D. L. Smailes, and L. M. Vallarino. Inorg. Chem.. 1986. 25, 1127. 33 G. J. Palenik, in 'Systematics and the Properties of the Lanthanides'. ed. S. P. Sinha. NATO AS1 Series C N0.109. D. Reidel Publishing Co., Dordrecht, 1983, Chap. 5. 3.1 J. D. J. Backer-Dirks, J. E. Cooke. A. M. R. Galas. J. S. Ghotra. C. J. Gray. F. A. Hart, and M. B. Hursthouse. J. Cherv. Soc.. Drrlror~7rm.Y., 1980, 2191. 77 Macrocydic Schiff' Buse Coriipkses ?f Lantlianides und Actinides P Fenton and Vigato Folding of the macrocycle towards the hemisphere containing the nitrate anion has also been detected in the samarium complex,26 in which the metal is 10-co- ordinated with a chelating nitrate anion on one side of the macrocyclic plane and the hydroxyl anion and water molecule on the other.In the lutetium complex two species are shown to be present in 1:l ratio, [Lu( 1)(MeCOO)(H20)](OH)(C104)(MeOH) and [Lu( l)(MeCOO)(MeOH)]-(OH)(C104).32 The complex cations contain 9-co-ordinated lutetium which is linked to the six nitrogen atoms of the macrocycle, a bidentate acetate anion, and a molecule of solvation (water or methanol). The macrocycle is in a ‘folded butterfly’ arrangement and is folded away from the hemisphere containing the acetate anion. It is proposed that this relieves strain in the macrocycle and minimizes repulsion between the heteroatoms, and thus allows optimum co-ordination at the metal.As there are non-co-ordinated but potentially strong oxygen-donor ligands present it is also suggested that nine-co-ordination is the limit in such species. There remains the question of whether the folding of the macrocycle is indeed governed by the size of the metal or by the steric requirement of the ligands. In general the non-coplanarity of the macrocycle increases in the same sequence as does the co-ordination number and appears to be a consequence of the ligand buckling to accommodate and main- tain hexadentate contact with the smaller cations; the folding is always in the direction of the hemisphere containing the single counter-ligand. The complex Eu(l)(NCS), is a solitary example of a macrocyclic complex bearing only monodentate ligand~.~~ The europium is nine-co-ordinated by the six nitrogen donors from the macrocycle and by three nitrogen atoms from monodentate thiocyanate anions, two on one side and one on the other side of the macrocyclic plane.The above trend is followed as the macrocycle is bent towards the single thiocyanate. The lanthanide complexes of (1) have been found to undergo reversible dehydration in the crystalline state and anion loss through hydrolysis, but the macrocyclic entity is stable to 240 OC-an unusually high temperature for lanthanide-bound nitrogen-donor ligands. It must be noted that the praseodymium and europium nitrate complexes both decomposed violently at 300 0C.30 In solution the macrocycle remains intact in the presence of H20 and di-methylsulphoxide (dmso) and this is evidenced by the n.m.r.spectra recorded in these solvents. Typical precipitating agents such as F-, OH-, and C2042- do not remove the lanthanide ions readily. For example the complex [Pr( 1)-(N03)2(H20)]N03,H20 is stable for some time in KOH solution before Pr(OH), is precipitated, and with KF no ready precipitation occurs. This may be contrasted with the behaviour of Pr( 18-crown-6)(N03), which gave instant precipitation with both reagent^.^' The complexes of (1) may be compared with lanthanum cryptates, edta complexes, and complexes of substituted phthalocyanines and porphyrins which are stable in water and, for the last two ligands, stable to base.A ‘H n.m.r. study of the nitrate complexes of (1) has revealed fairly considerable contact shifts.30 However, it is likely that for the smaller lanthanides (Tb-Er) 3s G. Bombieri, hiorg. Chin!.Actn. 1987, 139, 21. Macrocyclic Schiff Buse Coyde.yes of' Luntlimiidos und Actinides structural changes are occurring in solution. It cannot be ascertained whether this is by means of an n.m.r.-rapid equilibrium between two or more discrete structures which changes as the series is traversed or whether one structure is present for each metal but progressively alters with ionic radius. The crystal structures must be treated with caution as predictors of the structures in solution; this is reinforced by the observations on the samarium complex of (4).(ii) Aromatic Lateral Units. The condensation of 1,2-diaminobenzene with 2,6-diacetylpyridine in the presence of lanthanide nitrates (La, Ce, Pr, Nd) has been reported to give [Ln(6a)(N03),],2H,0. The presence of a peak corresponding to the free macrocycle in the m.s. and the interpretation of the i.r. spectrum led to the proposal that the complexes contained 12-co-ordinated lanthanide cations; the three nitrate anions were predicted to be bidentate.23*36*37 During attempts to repeat and extend this work a yellow crystalline product, having the same molecular mass peak as (6a), was consistently obtained as the major product.38 This compound was shown to be identical to the tricyclic product (7) obtained by direct reaction of the organic precursors in the absence of metal ions.The nature of (7) had remained an enigma from its first appearance39 until the solution of the X-ray crystal structure4' and its facile recovery during attempts to obtain transition metal complexes of (6a) by template procedures led to the conclusion that steric crowding inhibited the formation of (6a). This could arise via interaction of the methyl groups with proximal aromatic hydrogen atoms. In contrast, the related condensation of 1,2-diaminobenzene with 2,6-diformylpyridine in the presence of Ca", Sr", Ba", and Pb" templates readily gives complexes of the hexa-azamacrocycle (6b),41 bearing out the predictions made for this [181-annulene "W. Racecka-Paryzek. lnorg. Chn. Acrcr, 1981, 54, L251.3' W. Radecka-Paryzek. Itiorg. Chitii. Ac,~a,1985, 109. L21. '13 L. M. Vallarino, personal communication. 39 R. W. Stotz and R. C. Stoufer, Cherii. Conitii.. 1970, 1682. do J. de 0.Cabral, M. F. Cabral, M. G. B. Drew, F. S. Esho. 0.Haas, and S. M. Nelson, J. Chcwi. Soc,.. Ciimi. Cotmiuri.. 1982, 1066. dl M. G. B. Drew, J. de 0. Cabral. M. F. Cabral. F. S. Esho, and S. M. Nelson, J C/wn. Soc.. Chrni. Conitiiun., 1979, 1033. 80 Fenton and Vigato Figure 6 (a) The generation of hemispheres bj*the macrocyclic ligand: (h)the,folding of die niacrocyde to acconir,iociatc> counter-ligand presence analogue ria theoretical calculation^.^^ In a recent preliminary comm~nication~~the structure of [Pr(6a)(N03),-MeOH)ClO,,O.5MeOH has been reported, revealing the reality of the formation of macrocyclic complexes of (6a).No experimental details have been released as yet. The praseodymium is 11-co-ordinated to the six nitrogen donors from the macrocycle, with one bidentate nitrate anion on one side of the macrocycle plane and the remaining bidentate nitrate and the methanol on the other side; the macrocycle is reported as being folded. B. ‘2 + 2’ Macrocycles derived from Furan-2,5-dialdehyde.-The metal template procedure has been used to prepare complexes of the macrocycles (8)-( For the lighter lanthanides the following complexes were reported: [Ln(8)(N03),],n- H,O (n = 0-2), Ln = La-Eu except Pm; [Ln(9)(N03),,nH,0 (n = 0-2), Ln = La, Ce, Pr, and [Ln(lO)(NO,),],nH,O (n = 0, l), Ln = La, Ce, Pr.When the heavier lanthanides (Gd--Lu) were used with ethylenediamine as the lateral unit, analyses corresponding to a ratio of three metal cations to two macrocyclic units were obtained, [Ln3(8),(N0,),],4H,O. The presence of the discrete macrocycle is indicated by m.s. and the observation of this unexpected stoicheiometry may be related to decreased cation size. In the corresponding reactions with diaminopropanes intractable materials were recovered. In contrast to the pyridine-based macrocycles, but in keeping with the lanthanide complexes of the cyclic pol yet her^,^^ the complexes of (8)-( 10) decompose in water. This may be attributed to a deviation from the ‘best-fit’ criterion, leading to reduced ion- dipole interactions and so giving a more labile system.In addition, the weakly donating furan head-group would help in labilizing the system. The ‘H n.m.r. spectrum of [La(8)(N03),], in [2H,]dmso, obtained immediately on dissolution gave two sets of signals; after ca. one hour only one set remained and it was proposed that this was due to the free macrocycle. Attempts to isolate the free macrocycle were unsuccessful. C. L. Honeybourne, Terrciiiec/roti, 1973, 29. 1549. 43 K. K. Abid and D. E. Fenton. lnorg. Chiin. Acid, 1984, 82, 223. (a)J. C. G. Bunzli and A. Wessner. Hrlt.. Chini. Actcr, 1981,64, 582; (h)A. Musurneci. 1t70rg. Chin?.Acru. 1981. 53,L249. Macrocyclic Schiff Base Complexes of' Lanthanides and Actinides The kinetic lability of the alkaline earth metal cations in the Schiff base macrocyclic complexes generated by template procedures has provided a synthetic pathway to otherwise inaccessible transition metal complexes.1s.'6 The lability of the lanthanide cation has also been explored in such transmetallation reactions4, For example, the reaction of [La( lO)(NO,),] with copper(I1) perchlorate gave the homodinuclear complex [Cu,( 10)(OH),(C104),],3H20, which differs only in the hydration number from the analogous complex derived by transmetallation from [Ba( 10)(C104),].4s The general applicability of this reaction is further indicated by the ability of the lanthanum complexes of the '2 + 2' macrocycles derived from 2,6- diformylpyridine and 1,2-diaminoethane or 1,2-diaminopropane, to transmetallate with copper(~r).~~ N / N LR,N C.'1 + 1' Macrocycles derived from Pyridine Units.-Transition metal complexes of '1 + 1' Schiff base macrocycles containing pyridinyl head-units have been well studied with particular reference being made to their capability for generating high co-ordination geometries at the metal centres.' 531 Alkaline earth metals have been used as templates to prepare Schiff base macrocycles derived from facultative diamines,' 4a3b and as precursors for transmetallation reactions to synthesize otherwise inaccessible transition metal complexes.' For the lanthanides, use of the template procedure has given the complexes [La(m/c)(NO,),],nH,O for (1 1) and (12), and [La(m/c)(NO,),OH],nH,O, for (13) and ( 14).23*46,47If La(NCS), is used as the template, mixed OH-NCS complexes are obtained throughout: " S.M. Nelson and F. S. Esho. J. Clictn. Snc... Clien?.Conimun., 1981, 388. 46 A. M. Arif. C. J. Gray, F. A. Hart, and M. B. Hursthouse. Inorg. Chim. Arirr, 1985, 109. 179. D. E. Fenton and S. J. Kitchen. unpublished results. Fenton and Vigato oC i3i Figure 7 The X-ray structure of [La(l l)](N03)3(Reproduced by permission from Znorg. Chim. Acta, 1985, 109, 179). CL~(~/~)(NCS)(OH),I, and CL~(~/C)(NCS),(OH)I, [mic = (1 21, ~4)1 C~/C= (ll), (12)].47 It was, however, necessary to use transmetallation from the barium complex of (1) in order to prepare the cerium and samarium complexes of (1 l).46 The presence of the macrocycle may be confirmed by i.r.and 'H n.m.r. studies, and for [La(ll)(NO,),] an X-ray crystal structure is available to confirm the assignment (Figure 7). The lanthanum is centred in the macrocyclic ring. It is 12-co- ordinated by the six heteroatoms from the ring and the three bidentate nitrate anions, one on one side of the macrocyclic plane and two on the other side;46 such an arrangement was also noted for [La( 18-~rown-6)(NO,),]~~ and [La(l)(N0,),].2' The lanthanum is equally disposed towards the oxygen and nitrogen donors of the ring, as has been found for the alkaline earth metal complexes [Ca( 1 l)](NCS), and [Sr( 1 l)](NCS),,H,048-in contrast to the preferential disposition of the lead cation towards the nitrogen donors in [Pb(ll)](NCS)(SCN).49 In this last case the affinity of the metal for the 'softer' ligand environment is evidenced.There appears to be a consistency in the 12-co- ordination of lanthanum in the presence of 18-membered rings bearing six donor atoms and three potentially bidentate ligands. In [La( 1 l)(NO,),] the macrocycle is quite bent, folding away from the hemisphere containing the two nitrate anions towards that containing the single nitrate, exactly as found for the '2 + 2' tetraimine Schiff base complexes. The lanthanide complexes are considerably less stable in water than are the analogous complexes of (1); they are more similar in behaviour to the complexes of (8) and of cyclic polyethers. Metal hydroxide is readily precipitated on dissolution in H20 and it is probable that this reaction begins with the breaking of the three adjacent ether-lanthanum bonds.The 'H n.m.r. spectra of the complexes [Ln( 1 1)(NO3)J have been recorded in acetonitrile solution and indicate the integrity of the macrocycle in that medium.46 48 D. E. Fenton, D. H. Cook, and I. W. Nowell, J. Chem. Soc., Chem. Commun., 1978, 279. 49 D. E. Fenton, D. H. Cook, and I. W. Nowell, J. Chem. Sac., Chem. Commun., 1977, 274. 83 z IN I C L lJ I z It X C al1T I 3- % 84 Fenton and Vigato D. '2 + 2' Macrocycles derived from Phenols.-The first defined binucleating ligands were prepared from 2,6-diformyl-4-methylphenol and 1,n-diamino-alkanes.50 The co-ordination chemistry of transition metal complexes of these ligands has been much explored and has been reviewed.51 The reaction of 2,6-diformyl-4-chlorophenol, polyamines, and the appropriate lanthanum nitrate gave the macrocyclic complexes depicted in Scheme 4.52 The application of the template procedure is erratic in result and so use of preformed ligands is preferred. The best route is not to isolate these ligands but to prepare the requisite one in situ and to follow the appropriate lanthanide addition sequence. The presence of a base (LiOH, NaOH) promotes formation of the monocationic species (1 5), whereas an absence of base leads to the tricationic species (16).In the complexes derived from X = NH, one NH is non-co-ordinated and so available for further reaction.A ring contraction is found to occur in the presence of the smaller lanthanides; this allows a diminution of the macrocyclic cavity and reduction in ligand denticity (Figure 8).5 Similar contractions generating oxazolidine- or imidazoline-bearing macrocycles have been noted in the presence of main group or alkaline earth templates for the related '2 + 2' tetraimine Schiff base macrocycles derived from pyridinyl head-units and the appropriate functionalized diamine.19.5 The X-ray structure of the terbium complex of the contracted macrocycle (and of the isostructural europium analogue) has been solved and shows the cation to be nine-co-ordinated and sitting in the lateral compartment of the macrocycle derived from the open-chain amine.The co-ordination environment is completed by two bidentate nitrate anions. The third nitrate is ionic and the imidazoline-ring containing compartment is empty.53 The synthesis of homodinuclear complexes of Schiff base macrocycles was first accomplished through the use of a lead(ii) template in the condensation reaction of 2,6-diacetylpyridine and 3,6-dioxaoctane- 1,8-diami11e.~~*~ The dinuclear nature of the product of the reaction, [Pb2( 17)](NCS),, was confirmed by X-ray crystal structure analysis." Both the di-lead and corresponding mononuclear Ba" complexes underwent facile transmetallation reactions to generate homodinuclear copper(i1) complexes. These complexes are of interest as speculative models for the dinuclear copper sites in copper-containing proteins and enzymes such as haemocyanins and tyrosinase. ' The synthesis of homodinuclear macrocyclic complexes of lanthanides has been achieved through the condensation of 2,6- diformyl-4-cresol and triethylenetetramine in the presence of Ln(NO,), or Ln(CIO,), (Ln = La-Gd).56 Complexes of the general formula [Ln2(18)-50 N.H. Pilkington and R. Robson. Ausr. J. Chi., 1970, 23, 2225. 51 ((I) S. Groh. Isroel J. Climi., 1976 -77, 15,277; (h) D. E. Fenton, Ah. Itiorg. Bioirzorg. Me&, 1983,2. 187: (c) P. Zanella. S. Tamburini, P. A. Vigato. and G. Mazzocchin. Coorri. Cliem. Rev., 1987, 77,165. 52 P. Guerriero. U. Casellato, S.Tamburini, P. A. Vigato, and R. Graziani. Inorg. Chinz. Ar,m, 1987,129,127. s3 N. A. Bailey, D. E. Fenton.I. T. Jackson, R. Moody. and C. 0.Rodriguez de Barbarin, J. Clieni. Soc.. Chrwi. Coniriiuri., 1983, 1463. 54 M. G. B. Drew. A. Rodgers. M. McCann. and S. M. Nelson, J. Clzeni. Soc.. Clieni. Commuti., 1978,415. " D. H. Cook, D. E. Fenton, M. G. B. Drew. A. Rodgers, M. McCann, and S. M. Nelson, J. Clieni. Soc.. Dtrltori Trrrris.. 1979. 4 14.'' I. A. Kahwa. J. Selbin. T. C. Y. Hasieh, and R. A. Lane, Ir?org. Chin?.Acto, 1986, 118, 179. Macrocyclic Schifr Base Cor?iple.uesof Lanthanides and Actinides r CI I+ CI 3+ ring contraction L Cl Figure 8 Ring contraction in thr lantlicinide (N0,)4-x(OH)x], (x = 0-2) were recovered on concentration of the methanolic solution; diluted solutions deposited [Ln,( 18)(N0,),],nH20. The hydroxyl- containing complexes could be converted into the latter by reflux in methanol containing the tetramine and a threefold excess of Ln(NO,),.The dinuclear nature of the complexes was established by positive-ion fast atom bombardment m.s.; the interesting, mostly new, polyatomic 0x0-clusters Ln,02 +,Ln,O, +,Ln,O, +,and Ln50,+ were dominant in the mass spectrum. The organizational role of the lanthanide cations in the assembly of the binucleating macrocycle was evidenced by the recovery of intractable materials when Ca", Sr", Ba", and Pb" were used as putative templates. The condensation of 2,6-diformyl-p-cresol with triethylenetetramine in the absence of the lanthanide templates gave only intractable materials. 3 Actinide Complexes Interest in generating Schiff base macrocycles around an actinide template was stimulated by the formation of the so-called 'superphthalocyanine', formed by the reaction of 1,2-dicyanobenzene with anhydrous dioxouranium(v1) dichloride in dimethylformamide sol~tion,~' in which the uranium is equatorially bonded to five nitrogen atoms of the ring.The intermediate size of the cation (1.00 A for eight-co- ordination) and preference for equatorially directed bonding suggest that the trans- dioxouranium(v1) ions should promote the formation of flat macrocyclic systems. In the presence of UO,' the condensation of 2,6-dicarbonylpyridines and 1,2-+ diaminoethane proceeds smoothly to yield the complexes [U02(l)]X2 and [U02(4)]X2,X = ClO,, NO,, or 1.58-59The i.r.and 'H n.m.r. of these complexes suggest that all six aza-donor atoms are co-ordinated to the uranium and hat here 57 V. W. Day. T. J. Marks. and W. A. Wachter. J. Am. Chet?i. Soc.., 1975, 97, 4519.'' K. K. Abid. MSc. Thesis (University of Sheffield) 1982. '' L. De Cola. D. L. Smailes, and L. M. Vallarino. ftiorg. Chiin. Acrtr. 1985. LI. 86 Fenton and Vigato +ior\6\,~ NFN N N--\ Me*o” N N N-/ HZ-(18) is an effective D,, site symmetry. There is an exceptional inertness towards the release of U022+in solution, even towards strong acids or strongly competing ligands, which suggests that systems of this type may be of value when efficient sequestering of actinide ions is necessitated. ‘After last returns the the reaction of 2,6-diformyl-4-chlorophenyl with diethylenetriamine gave on addition of a dioxouranium(v1) salt mononuclear acyclic complexes which could be cyclized on further addition of di- or tri-amines to give the mononuclear macrocyclic complexes (19)-(22)?’ These cyclic complexes were found to act as ligands towards transition metals and so give some of the first examples of compounds in which a heterodinuclear fragment was held within a dinucleating macrocyclic periphery (23). Recent adaptation of the synthetic procedure has given the symmetric and non- symmetric macrocycles (23)-(25) and these have also been used to prepare mononuclear dioxouranium(v1) complexes [U02(m/c)].62 An X-ray crystal structure of [U02(26)] confirms that the UO,’ is equatorially co-ordinated by + the N202Sdonor atoms from one of the two identical compartments with the second remaining empty; a ‘butterfly-fold’ occurs in the ligand.63 The site occupancy in the non-symmetric macrocyclic complex [U02(25)] is believed to be in the ‘harder’ compartment, although this has yet to be verified by crystallography. Although it has not yet been possible to add a second UO,’+ to these systems to give homodinuclear complexes, the heterodinuclear complex [U02(24)Cu]-C10,,2EtOH has been reported.64 Cyclic voltammetric studies on this compound show that two successive one-electron reactions occur; a reoxidation peak ’” Robert Browning, ‘Apparent Failure’, in ’The Poetical Works of Robert Browning, Vol.VII. Smith, Elder, and Co..London, 1902. (N) U. Casellato. M. Vidali. and P. A. Vigato, 1rior.g. Nud. Client. Lett., 1974, 10,437; (h)M. Vidali, P. A. Vigato. U. Casellato, E. Tondello. and 0.Traverso, J. Itmg. Nud. Client., 1975. 37. 1715. 62 U. Casallato. D. Fregona. S. Sitran, S. Tamburini, P. A. Vigato, and D. E. Fenton. Inorg. Ckint. Acro, 1985, 110, 181. b3 U. Casellato. S. Sitran. S. Tamburini. P. A. Vigato, and R. Graziani, Inorg. Chini. AGIO, 1986, 114. 111. 64 U. Casellato. P. Guerriero, S.Tamburini. P. A. Vigato, and R. Graziani. Itiorg. Chim.Actn, 1986.119.21 5. 87 Macrocyclic Schgf Base Comple..ces qj’Lanthanides and Actinides HN-Cl (M = Cu,Co,Ni) 2x (X = ClO,) L CI I CI Fenton and Vigato (27) attributable to free copper(1) ions produced by decomplexation of the mixed- valence Cu'Uv* species is noted, as is a peak attributed to the Uv/Uv' redox An interesting application of a macrocyclic complex of dioxouranium(v1) has recently been reported.66 The co-complexation of a neutral guest molecule by both hydrogen bonding and co-ordination with a metal ion is proposed for metalloenzymes, such as urease, where urea is believed to bind at the active site via hydrogen bonding to the peptide chain and co-ordination to a nickel ion.67 Mimicry of such an event has been achieved by co-complexing urea and the dioxouranium(v1) cation within the macrocyclic cavity in the complex (27).The X-ray structure shows the cation to be held in the Schiff base compartment and the urea in the polyether compartment.The urea is co-ordinated to the cation via a lone pair from the carbonyl oxygen, advantage being taken of the capacity of the uranium to receive a fifth equatorial donor atom. The urea is also hydrogen bonded to five oxygen atoms of the polyether chain and one phenolic oxygen. These interactions lead to a highly structured ternary complex. 4 Concluding Remarks The lanthanide(II1) cations have been found to act as template reagents, generating Schiff base macrocycles in the same manner as do the larger alkaline earth metals and lead(I1). Although 18-membered hexa-azamacrocycles are found in the presence of all the lanthanide ions, there is a noticeable cation-cavity control exercised in the formation of 14-membered macrocycles, where the larger radius lanthanides are ineffective as templates, and also in the 20-membered macrocycles, where the smaller radius lanthanides are excluded.Within the series of complexes P. Zanello, A. Cinquantini, P. Guerriero, S. Tamburini, and P. A.Vigato. 1iror.g.Chim. Acrci, 1986.117.91. "C. J. van Staveren, D. E. Fenton, D. N. Reinhoudt, J. van Eerden. and S. Harkema, J. Ani. Cheni. Soc.. 1987, 109. 3456. ('7 R. K.Andrews. R. L. Blakeley. and B. Zerner, Ah. 1uor.g. Bioihet?i.. 1984, 6. 245. Macrocyclic Schiff Base Comp1e.ut.s of‘ Lanthanides and Actinides derived from the 18-membered macrocycles there does not appear to be any significant fine control exercised in the template synthesis, even though more flexible carbinolamine macrocycles were obtained in some instances. The introduction of oxygen donors restricted complex formation to the lighter lanthanides.There is a current paucity of actinide-containing Schiff base macrocycles. It has, however, proved possible to prepare heterodinuclear complexes involving uranium; the only homodinuclear complexes prepared to date in this area are restricted to those of the lanthanides. X-Ray crystal structures of the ‘2 + 2’ hexa-azamacrocycles reveal that there is a steady decrease in co-ordination number of the metal within a given series of macrocycle as the series moves from La to Lu; a similar effect has also been noted with complexes of cyclic pol yet her^.^' Many of the structures contain bidentate nitrates and these are less strongly bound to the metal than in the corresponding cyclic pol yether complexes.Generally the accompanying ligands are bound to the metal, with two on one side of the macrocyclic plane and one on the other side with the macrocycle folding in the direction of the single ligand. This folding about the lateral-unit appears to relieve steric strain and also minimizes repulsions between the accompanying ligands. In aqueous solution there is a remarkable kinetic stability of the hexa-azamacrocyclic complexes of the lanthanides relative to the corresponding cyclic polyether complexes. Given that the bonding is predominately electrostatic, this may be compared with the enhanced stability of potassium complexes of (1) relative to potassium complexes of 18-crown-6 which has been attributed to a stronger ion- dipole interaction being made available in the hexa-azamacrocycle-potassium.68 The triazatrioxamacrocyclic complexes are less stable in water, as are the furan- derived species, and so are intermediate in character between the hexa-aza and hexa-oxa-species.The inertness of the dioxouranium complexes of (1) and (4)towards the release of U02” in solution has suggested the use of such systems as effective sequestrants for the actinide ions.59 N.m.r. studies on the water-stable lanthanide complexes have indicated a possible application of the complexes as shift reagent^.^^.^' The time frame of metal exchange and ligand exchange kinetics is important when considering the use of lanthanide ions in such a way and as probes for biological systems.29 In the dilute aqueous, or aqueous organic, solutions required for such studies even a complex that is stable towards hydrolysis may be sufficiently labile towards exchange to give unreliable results.The systematic study of macrocyclic Schiff base complexes of the lanthanides has revealed a unique inertness of the complexes which, together with their solubility in both water and organic solvents, may prove to be of potential value in any application of these complexes as shift reagents or fluorescent probes. ’’ T. W. Bell and F. Guzzo, J. At??.C/wt?i.Soc., 1984, 106, 61 11.

ISSN:0306-0012    

DOI:10.1039/CS9881700069

出版商:RSC    

年代:1988

数据来源: RSC

摘要:

C‘lienz. SOC.Reu., 1988, 17, 91-109 The Dakin-West Reaction By G. L. Buchanan DEPARTMENT OF CHEMISTRY, UNIVERSITY OF GLASGOW, GLASGOW GI2 8QQ 1 Introduction The direct conversion of an x-amino-acid into the corresponding z-acetylamino- alkyl methyl ketone by the action of acetic anhydride in the presence of a base such as pyridine, takes place with the evolution of CO, and is generally known as the Dakin-West reaction (Scheme l).’ Its discovery by these two authors in 1928 was accompanied by a brief exploration of its nature, its scope, and its mechanism.’ A little earlier, others had observed that both tyrosine and z-phenylalanine afforded ‘abnormal’ products when acetylated under these conditions. However, they were slow to identify the products4 and the reaction is invariably credited to Dakin and West.AC20 H2NCH(RICO2H base * AcNHCH(R)COMe + CO, Scheme I Most general references ’to the reaction describe only the conversion of an X-primary amino-acid, as in the above example. But, in fact, non-amino-acids such as phenylacetic acid also yield methyl ketones under these conditions, as do secondary amino-acids, which Dakin and West believed to be constitutionally incapable of undergoing this transformation.’ More recent work has also improved the efficiency of many of these conversions and this simple ‘one-pot7 conversion of a carboxylic acid into a ketone, under mild conditions, often proceeds in good yield. It deserves to be better known. These transformations do not all involve the same reaction mechanism, but they all involve the same reaction conditions and, for the purpose of this review, they are all included as examples of the general Dakin-West reaction.This review discusses these reaction mechanisms but it deals particularly with the preparative scope and efficiency of the reactions. Experimental details and yields are summarized in the Tables (pp. OOo--OOO). ’ Merck Index 10th Edn.. ONR-22. Merck & Co. Inc.. USA, 1983; H.O. House ‘Modern Synthetic Reactions’, 2nd Edn. W. A. Benjamin Inc., 1972, pp. 770- -773; ‘Comprehensive Organic Chemistry’, ed. D. H. R. Barton and W. D. Ollis, Vol. 2, Pergamon Press, Oxford 1979. p. 828.’H. D. Dakin and R. West, J. Biol. Chmi., 1928, 78. 91. 745, and 757. P. A.Levene and R. E. Steiger, J. Riol. Chon., 1927. 74. 689. P. A. Levene and R. E. Steiger. J. Biol. Chern., 1928, 79, 95. The Dukin- West Reuction 2 Primary a-Amino-acids In the original investigations' in which acetic anhydride was allowed to react with, inter ah, phenylalanine, tyrosine, leucine, alanine, and 2-phenylglycine, in the presence of pyridine at steam-bath temperature, it was noted that CO, was evolved and that a base was essential. It was shown also that pyridine could be replaced by alkylpyridines 2.5 or sodium acetate but not by quinoline or N,N-dimethylaniline. Whilst amino-acids yield ketones under these conditions, the related amines such as benzylamine are merely N-acetylated. No reaction was observed with 13-amino- acids or with x-amino-acids which lacked one C(2)-H; e.g.x-aminohydratropic acid (1).2 At these temperatures (m.100 "C)no reaction took place with secondary amino-acids, e.g. sarcosine (2) or proline (3) or with the tertiary amino-acid (4);2 however, they do react at reflux temperature (see Section 3). Acetic anhydride may be replaced by other anhydrides affording for example ethyl- or propyl-ketones in good yields, particularly under modified reaction conditions 5-' but, in general, yields diminish with increasing chain length in the acid anhydride. Benzoic Table 1 R2 I R' NHCHC02HI . (R~CO)2~+base R2 I R'NHCHCOR3 Rcw,lioti R' R2 R3 Conditions YlCld (,I(, H H Me py; reflux; 6 h; 60R' = AC HzO PhCO H Me 3-pic; r.t.; 2 h: 77 H2O PhCO H Et 3-pic; r.t.; 2 h; 73 H 2O PhCO H Pr" 2-pic; 35 "C; 3 h; 71 H20 PhCO 1-15 "C; 17 h; 75 H2O PhCO 10-15 "C; 17 h; 34 H20 PhCO I@-15 "C; 17 h; 54 H2O H Me Me py: 100 "C; 6 h; 81-88 R' = stir PhCO Me Me Et,N; DMAP; f h; 78 r.t.J. Attenburrow. D. F. Elliot. and Ci. F. Penny. J. Ckcwr. Soc.. 1948, 310. '' G. H. Cleland and C. Nieman, J. Am. C/wrn. Soc., 1949, 71, 841.' R. H. Wiley and 0. H. Rorum. J. An7. Cl7cm. Soc.. 1948, 70. 2005: Org. LYjwth. Co//., Vol IV, 5. (1963). Buchanan Table 1 x-Amino-mids (continued) Reaction R' R2 R3 Conditions Yield % Ref: H Me Ph py; 135 "C; 2i h; 42 R'= PhCO g stir H Me2CHCH2 Me py; 100 "C; 6 h 70 R' = AC h H Ph Me py; 100 "C; 5$ h 72 R' = AC h H Ph Et py; reflux; It h 75 R'= EtCO U Me py; DMAP; 25 "C; 82 i stir 20 min H PhCH, Me py; 100 "C; 5 h 79 R' = AC g H PhCH, Et py; 135 "C; 14 h 41 R' = EtCO g H PhCH, Pr" py; 135 "C; 3 h 27 R' = PrCO I: H PhCH, MeOCH, py; 11 5 "C; 1 h; 78 R' = MeOCH,CO g stir (crude) H PhCH, Ph py; 145 "C; 2 h 44 R' = PhCO g Me Et3N; DMAP; 75 R' = AC i 20 min; stir clQc0 MeSCH,CH, Me py; DMAP; 67 i EtOAc; reflux; 2 h Ph H Me reflux; 6 h 67 .i Ph H Et reflux; 4 h 55 .I Me00 H Me reflux; 1 h 454 .I Me reflux; 1 h 224 J Ph Me Me py; reflux; I h 134 Me Me,CHCH, Me 4-pic; 130 "C; 10 h 36 Me Me,CH Me py; 140 "C; 8 h 374 PhCO CH,CH,CO,H Me py; 110 "C; 45 min 90' PhCO CH,CH,CO,H Et py; reflux; 45 min 75 PhCO CH,CH,CO,H Pr" 2-pic; 110 "C; 1 h 31 PhCO CH,CH,CO,H C,H, 2-pic; 120 "C; 2 h 30' PhCO CH,CH,CO,H Ph 2-pic; 100 "C; 1 h 26 ' H CH,CH,CO,H Me Et,N; DMAP; 76 m 60 "C; 8 h a R.H. Wiley and 0.H. Borurn, J. Am. Chem. Soc.. 1948, 70, 2005; * ref. 5; R. A. F. Bullerwell and A. Lawson, J. Chqm. Soc.. 1952, 1350: E. J. Bourne, J. Burdon, V. C. R. McLoughlin, and J. C. Tatlow, J. Cliem. So(,.. 1961, 1771; 'ref. 7; ref. 16; ref. 6; ref. 63: ref. 21; ref. 34; 'ref. 32; ' ref. 15; ref. 62; " R3COQ0; R'cOOO ; NAc derivative; NCO-Et derivative. COPh Ac 93 The Dakin-West Reuction Table 2 Other Acids R'R~CHCO,H ( R~CO)*obase R' R2CHCOR3 R' R2 R3 Reuction Conditions Yield Ref: H H H H Ph Ph Ph Ph Me Me Et Pr py; reflux; 6 h py; 120 "C; 67 h; stir py; 120 "C; 42 h; stir py; 120 "C; 31 h; stir 56 + 24" 11 28 + 33" i 50 + 16" i 38 + 18' i H Me py; 120 "C; 4 h; stir 48 I Me 6-purinyl-t hio Me reflux; 5 h 57 i Me 6-purinyl-thio Et reflux; 6 h 44 i H PhS Me 2,6-lutid; reflux; 12 h 21 i H PhO Me py; reflux; 12 h; stir 17 i H 3-pyridyl Me NaOAc; reflux; 17 h 39 k reflux; 1 h 100d.r 1 reflux; I h 72 d.e 1eCozHreflux; 1 h 67d*r 1 reflux; 2 h 84 1{F; reflux; 2 h 13' 1 n-C,H 17 glass powder; 185 "C; 7 h/N, 49 f m Me 125 "C; 36 h 42 n CO 2Na Et 125 "C;36 h 20 f t1 Pr" 125 "C; 36 h 17f n Pr' 125 "C; 36 h 9f nCOzNa C02NaA Ph 125 "C; 48 h 5f n Me reflux; 15 min/N, 58 P n Et reflux; 15 min/N, 57 Po-NCH2CH2 H Et2NCH,CH2 Me reflux; 15 min/N, 47 P H Pr',NCH,CH, Me reflux; 15 min,", 66 P 0 Me NaOAc; 125 "C;35 min 48 Y OMe Me pyr; r.t.; 4 h 45 r o&C02H Buchanan Table 2 Other Acids (continued) 01; YieldR' R2 R3 Reuction Conditions Ref: Me pyr; reflux; 35 h 13 Fm-'02H " Ketonic by-product, R3 = PhCH,; * enol-ester; 'N-acyl derivative; crude; cfi structure (48); structure (52); keto-lactone; * ref.45; ' ref. 46; 'ref. 52; 'ref. 56: ' ref. 13h; ref. 56: " ref. 55; ref. 57; ref. 65; 'ref. 53: 'ref. 48. Me I P h-C-N H2 MeNHCH2C02H I mCO2H CO2H H (1 1 (2) (3 1 Ph-CH-NM e PhCON HC H2 C02Na I C02H (4) (5) anhydride gives moderate yields of the phenylketone (6),6 but neither phthalic anhydride nor succinic anhydride produces any ketone from phenylalanine.However, the bimolecular anhydride-ester derived from succinic acid behaves normally affording keto-ester in 840/;; yield (Scheme 3).8 NH2 NHCOPh I (PhC0120 I PhCHZCHCOPhPhC H2C HCOzH PY (6) Scheme 2 NHCOPh NHCOPh I ( Me02C C H2C H2C0)20 PhC H,C I HC0C H2CH,C 0, Me (84"/0)PhCHzCHC02H py,llO "C, 3h, stir Scheme 3 The use of higher reaction temperatures (e.g. 120 "C) and mechanical stirring' also improve the yields obtained with acetic anhydride. In the case of glycine, whose conversion into a-acetylaminoacetone under original (steam-bath) conditions was inexplicably poor, these modifications raise it to ca. 60%. Even better results have been obtained by employing sodium hippurate (5) in place of gly~ine.~.~ Dakin and West also observed that 'saturated' oxazol-5-ones (7) yield the same product as G.H. Cleland and F. S. Bennett, Sjnrhesis. 1985, 681. (a)R. A. F. Bullerwell and A. Lawson, J. Chem. Soc., 1952, 1350; (b)E. J. Bourne, J. Burdon, V. C. R. McLoughlin, and J. C. Tatlow, J. Chem. Soc., 1961, 1771. The Dakin- West Reaction the parent amino-acid and concluded that they might be reaction intermediates. On the other hand, the P-hydroxyamino-acid phenylserine undergoes dehydration and the reaction stops at the ‘unsaturated’ oxazolone (S).2 ‘Unsaturated’ oxazolones are unaffected by acetic anhydride and pyridine. If the reactivity of the x-CH is enhanced by an additional electrophilic group, as in the malonic half-ester (9), the conversion into ketone (10) takes place under milder conditions uiz., overnight at room temperature lo or under reflux with acetic anhydride alone.Another example is provided by 2-phenylglycine. Higher yields are claimed’ when the amino-acid is first converted into its N-benzoyl derivative and used as its sodium salt. This technique, which was first introduced by Attenburrow, Elliot, and Penny,’ affords the acyloxazolone (11) at room temperature, and from it (12), in much higher yield than by the direct Dakin-West reaction on glycine (Scheme 4).Alternatively, ethanolysis of the intermediate leads directly to the P-ketoester ( 13).5 It is even claimed that all x-benzamido-acids will undergo (slow) decarboxylative acylation by anhydrides without the need to add a base.I OH Ac20 ;base PhCO NHCH2C02Na Ph (51 (11 1 PhC 0NHC H2COMe PhCONHCHCOMe (12) (13) Scheme 4 When the Dakin-West reaction is applied to a-aminodicarboxylic acids, special problems arise.Thus, under standard conditions aspartic acid yields only impure ketonic material.2 But its N-benzoyl derivative (14) reacts with a variety of I” N. F. Albertson, B. F. Tuller, J. A. King, B. B. Fishburn. and S. Archer, J. Am. Clicw. Sor.. 1948.70, 1150. ‘I A. Lawson, J. Chem. Soc., 1953, 1046. l2 J. A. King and F. H. McMillan. J. Am. C‘liein. Soc.. 1955, 77. 2814. I3 ((I) A. Lawson, J. Cliei~.Soc.. 1954, 3363; (h) ihid., 1957. 144. Buchanan anhydrides, even in the absence of added base, to give lactones of type (16) and/or (17), presumably tlia the ketone (1 5) (Scheme 5).' 1,1 3,62 Likewise, glutamic acid, which yields a mixture of products [( 18)-(20)] under Dakin-West condition^,^,'^ responds to the Attenburrow modification, affording (21) and thence (22)' Homologous ketoacids have also been prepared by this route from the appropriate anhydrides.' PhCONHCHCO2HI CHZCOZH (14) (15) 0 COMe (16) (17) +I NH3 CI-(10) (20)R =Ac (22) (21) R = PhCO Scheme 5 More recent developments have revealed that, when a trace of 4-dimethylamino- pyridine (DMAP) is added to the reaction mixture, the Dakin-West reaction can be carried out at or near room temperature, and more rapidly than under standard reaction conditions (Scheme 6).16,' Even so, it is wasteful to use an anhydride as the acylating agent in cases where the corresponding acid is scarce, for only half of the reagent is utilized.Steglich and Hofle have overcome this objection by using the acid chloride (despite several reported failures 2*59a 'stepwise procedure' in (Scheme 7).18*19 This uses a preformed oxazolone, e.g. (23), which is 0-acylated by the acid chloride in the presence of Et,N then rearranged by DMAP2' to the J. A. King and F. H. McMillan, J. Am. Cliem. Soc., 1952, 74, 2859. Is R. A. F. Bullerwell, A. Lawson, and H. V. Morley, J. Clieni. Soc., 1954, 3283. l6 W. Steglich and G. Hofle, Angm. Chrni., In!. Ed. Engl., 1969, 8, 981. G. Hofle. W. Steglich, and H. Vorbriiggen, Angrit,. Clirm., In!. Ed.Engl., 1978, 17, 576. W. Steglich and G. Hofle, Angrw. CIieni., In[. Ed. Engl.. 1968. 7, 61. l9 W. Steglich and G. Hofle. CIwni. Bet-., 1969, 102, 883. W. Steglich and G. Hofle. Teiruheriron Let!., 1970. 4727. The Dakin-West Reaction Ac20, Et3 N MeV AcOH 30 min,r.?. PhCONHCHCOMePhCoNHCHCo2H DMAP,30 min,r.t? NQI I Me 1 Me Ph Scheme 6 R R (23) 'co h A AcOH+ .es.y'"4PY'Mu MesqcouAcOH, PY I NHCOR R (24) (25) Scheme 7 thermodynamically preferred C-acyl derivative (24) and finally decomposed with loss of CO, to the ketone (25) (R = 4-C1Ph).21The overall yield in this case is 85%. Both routes lead to high yield conversions under mild reaction conditions and are therefore applicable to sensitive molecules.8*22 In the stepwise route, ring-opening and decarboxylation, (24) -+ (25), are usually achieved by adding acetic acid and pyridine but also occur readily on boiling with water, provided the acylated oxazolone is enolizable, as in (26)?' However, 4-substituted-4-acyloxazolones such as (27) are open to attack at both carbonyl centres and yield mixtures COMe MeTCO Me+C,O H20* MeCHC02H MeCHCOMe \ I +I NHCOPh NHCOPhNYo NYoPh Ph (26) (27) Scheme 8 *' N. Engel and W. Steglich. Liehigs Ann. C'hcm.. 1978, 1916.''J. S. McMurray and D. F. Dykes, J. Org. Clieni., 1985. 50, I1 12. Buchanan (Scheme 8).19 This undesirable side-reaction has been avoided by the use of acetic anhydride and pyridine or by reaction with anhydrous oxalic acid l9 as shown in Scheme 9.Under standard Dakin-West conditions the terminal C02H of a peptide is selectively converted into methyl ketone and, if the mild reaction conditions of the DMAP-catalysed reaction are employed, sensitive protecting groups remain un- damaged.22 This conversion has been used to identify the C-terminal amino-acid of peptides by comparing (paper chromatography) the amino-acid content of the hydrolysate before and after reaction (Scheme lo). The method is applicable to 5-10 mg samples of ~eptide.~~ R' R2 R' R2I I I I -NHCHCONHCHC02H --NHCHCONHCHCOMe Scheme 10 The mechanism which Dakin and West themselves proposed for the reaction of primary amino-acids involved N-acylation followed by a C-acylation and finally decarboxylation of the P-keto-acid.Others invoked the oxazolone as an inter- mediate with its stabilized carbani~n,~?~ and later studies have only confirmed and added detail to this framework. These and other mechanistic studies have been reviewed by Allinger 24 who has set out the accepted mechanism in detail. It can be summarized as in Scheme 11. It is consistent with this mechanism that a-amino-acids lacking a C(2)-H do not react to produce a ketone,2 that intermediate C-acylated oxazolones have been isolated from the reaction and shown to be ketone precursor^,^,^^,^^ and that alanine labelled with I4C at C( 1) gave rise to labelled CO, plus unlabelled ketone.27 In a detailed examination of the acylation step (Scheme 12), Steglich and HGfle 20.28.29 ha ve shown the acylation of the oxazolone by acetic anhydride- pyridine mixtures involves a hitherto unsuspected intermediate (28) whose initial appearance and subsequent decay, with concomitant emergence of (29), has been followed by n.m.r.On the other hand, acylation by means of an acyl halide in the l3 R. A. Turner and G. Schrnerzler. J. Ani. Chen?.Soc., 1954, 76, 949. 24 N. L. Allinger, G. L. Wang, and 9. 9. Dewhurst. J. Org. Clzem.. 1974, 39, 1730. 25 P. L. Julian. E. E. Dailey. H. C. Printy. H. L. Cohen. and S. Hamashige, J. Am. Cl~enz.So(,.,1956,78,3503.'' Y. Iwakura, F. Toda, and H. Suzuki, J. Org. Cheni., 1967, 32, 440. 27 C. S. Rondestvedt. jun., 9. Manning, and S. Tabiban. J. An?. Chem. Soc., 1950, 72. 3183. 28 W.Steglich and G. Hofle, Tetrrilzeciron Lett., 1968. 1619. 2y W. Steglich and G. Hofle, CIieni. Ber.. 1971. 104, 3644. The Dakin-West Reaction COMe RCHCOzH Ac20L R~A,CO 6I 7 Nyb NHCOPh PhI bh Ph +2 AcOH +BH+ +AcO-COMe COMe AcOHI -I ~RCHCOMe + CO, tRC-C02H RC-CO-OAC I I I NHCOPh NHCOPh NHCOPh +Ac20 Scheme 11 Ac @] Ph Ac (28) J R OAcM NYo Ph Ph (29) (30) Scheme 12 presence of Et,N gives rise to a mixture of (29) and (30) in which the acyloxy- oxazole (30) can be isomerized to (29) by pyridine.'8*'" The same authors have shown that the attack of acetic acid on the acyloxazolone (29) occurs exclusively at C(5);alternative reaction at C(2) was ruled out by "0 studies.,' 30 G. Hofle, A.Prox, and W. Steglich, C%eni. Ber.. 1972, 105. 1718 100 Buchanan OJACO*HH0-co Scheme 13 It is of interest to note (Scheme 13) that in the case of 2-pyrrolidone-5-carboxylic acid (3 l), where oxazolone formation is forbidden by Bredt's Rule, no decarboxyl- ation occurs l4 and the reaction takes a new course leading to the dimer (32); i.e. acylation uia a simple Claisen-type mechanism is not observed. 3 Secondary a-Amino-acids The initial report' that secondary amino-acids are not converted into ketones under Dakin-West conditions seemed to confirm the view that oxazolone formation was an essential first step in the mechanism of the Dakin-West reaction. However, the report was mistaken because, at slightly higher temperatures, sarcosine (2) gives a moderate yield of methyl ketone (33),3' and N-methyl valine and N-methyl leucine were likewise found to undergo decarboxylative acylation in modest yield (Scheme 14).32 (33) Scheme 14 To accommodate these developments, Cornforth and Elliott proposed 33 that in secondary amino-acids the mechanism involves an oxazolinium intermediate (34).Other postulated mechanisms have not survived criticism.24 Significantly, N-methyl-2-pyrrolidone-5-carboxylicacid, which is not capable of forming the oxazolinium intermediate (35) (Bredt's rule), is returned unchanged even under forced condition^.'^ This experiment excludes the alternative of a Claisen-type mechanism which has been given con~ideration.~~ On the other hand, no ketonic products have ever been obtained from proline (3) 2314,35 or from N-formyl-N- phenylglycine (36),34although the former at least has been shown to be capable of forming an oxazolinium ion intermediate with acetic anhydride.36 Both were R.H. Wiley and 0.H. Borurn. J. Am. Cliem. SOC..1950, 72. 1626. 32 R. Hinderling. B. Prijs, and H. Erlenmeyer, Helc. Chim. Acta, 1955, 38, 1415. 33 J. W. Cornforth and D. F. Elliott, Science, 1950, 112, 534. 34 G. L. Buchanan. S. T. Reid, R. E. S. Thomson, and E. G. Wood, J. Chem. SOC.,1957, 4227. 3s 2. H. Israili and E. E. Smissman, J. Chem. Eng. Data. 1977, 22, 357. 36 R. Huisgen, G. Gotthardt. H. 0.Bayer. and F. C. Schaefer, Angel.1.. Chem., In/. Ed. Engf., 1964.3, 136. 101 The Dakin- West Reaction recovered (80-900/;) under Dakin-West conditions.However, peptides involving secondary amino-acids react normally and the reaction was employed analytically 37 in identifying N-methylvaline as the C0,H-terminal amino-acid in actinomycinic acid (39) (Scheme 15). Me Me PhNCH2C02H PhNHCHC0,H PhNHCC0,H0-co MeMI CHO (35) (36) (37) (38) Me 1. Ac20, PY Ac MeNCHMe, -CoNCHCHMe2 2.hydrolysis’ II t02H COMe (39) Scheme 15 Although all of these reactions were carried out in the presence of pyridine, it has been shown that added base is unnecessary, at least for N-arylglycines, which yield the alkyl ketone merely by boiling with acetic or propionic anhydride^.^^ Similarly (37) gave the methyl ketone, but (38) proved to be inert.34 This is consistent with a mechanism which includes C-acylation followed by decarboxylation. Little interest has been shown in the preparative application of the reaction to secondary amino-acids, but its mechanism has received detailed attention. The oxazolinium salt (41) has been isolated under mild conditions and shown to afford the methyl ketone (43) [probably uia (42)] when boiled with acetic anhydride (Scheme 16).38-40Furthermore, Knorr and Huisgen have been able to show 38 2x1-Ac20L 2 A P h +AcO-PhCONCHCOZHMe=-AC20kI AcOH Ph Ph 0 Ph (401 (41) ,, (42) Ph , J Me PhCONCHCOMe + CO,I PhCO Ph Me (44) (43) Scheme 16 3’ E.Bullock and A. W. Johnson, J. Cfietii. Soc.. 1957, 3280 38 R. Knorr and R. Huisgen.Chem. Bey., 1970, 103. 2598. 39 G. Singh and S. Singh, Trlrahedrori Let[.. 1964, 3789. 40 G. V. Boyd. J. Cheni. Sol., Cham. Comniun., 1968, 1410. 102 Buchanan that, in the absence of pyridine, the course of the third step in this sequence depends on the amount of acetic acid present. Acetic anhydride containing 1.5M acetic acid gives rise to the expected ketone (43) in up to 70% yield. However, in low concentrations of acetic acid (or in presence of pyridine) little ketone is found. The main product is the enol acetate (44), together with two pyrrole derivatives. Two routes are proposed, involving acetate attack on (42) at both C(2) and C(5).4' The ratio of C(2)/C(5) attack appears to be controlled by the concentration of acetic acid present, and the coexistence of these two routes has been established42 by the use of N-methyl-N-[' sO]benzoylphenylglycine (40*), pyridine, and acetic anhydride containing known concentrations of acetic acid (Scheme 17).MeG5h (40$0 [(41*)1 PhA$ + ACO-(421 AcI J MeN-kPh AcOA>(;;XoPh * oco + Me Ph 0 0-Ac (43)+(44) co2+(43*) +(44*) Scheme 17 By measuring the distribution of the 180-label in the reaction products it was concluded that at high concentrations, C(5) attack predominates but at low con- centrations the C(2)/C(5) ratio is about 40: 60. Complementary work, using l80-labelled acetic anhydride 30 has supported this conclusion. Thus, unlike the primary amino-acid mechanism in which the oxazolone is attacked exclusively at C(5) the oxazolinium ion (42), being more electrophilic at C(2), reacts at both centres. 41 R.Knorr and G. K. Staudinger, Chem. Ber., 1971, 104. 3621 42 R. Knorr. Cliem. Ber., 1971, 104, 3633. The Dakin-- West Reaction MeC02Na +(PrC0)20 PrCOMe + C02 Scheme 18 0Scheme 19 4 Other Acids In their initial paper,’ Dakin and West made passing reference to the fact that some non-amino-acids, for example, chloracetic, a-bromostearic and phenylacetic acids also yield ketones when boiled with acetic anhydride and pyridine, but they reported no experimental details or yields at that time or later. Of course, the formation of ketones from aliphatic carboxylic acid salts and anhydrides has been known since Perkin’s experiments a century ago (Scheme 18),43 and the same process doubtless forms the basis of the Blanc reaction 44 of certain dicarboxylic acids (Scheme 19).However, no systematic examination of the reaction was begun until much later. King and McMillan 45 showed that it is catalysed equally well by pyridine and sodium acetate but notably by tributylamine. Surprisingly, no further investigations have been reported with other organic bases. These authors also showed that the carboxylic acid involved requires at least one C(2)-H, and although the acid can be replaced by its (s-yrn)anhydride, it cannot be replaced by its ester or nitrile.46,47 Typically, phenylacetic acid gives a moderate (56%) yield of phenylpropanone when boiled under reflux with acetic anhydride and pyridine (Scheme 20), but 1,3-diphenyl propanone is also formed as a substantial (24%)by-product 45 and, on prolonged boiling, some enol-acetylation also takes pla~e.~~.~~ The formation of ketonic by-product is an undesirable feature of the reaction which has yet to be overcome.These ketones clearly arise from an initial acid-anhydride equilibration 49 generating phenylacetic anhydride and the mixed anhydride in addition to the (abundant) acetic anhydride. Thereafter it is generally agreed that base-catalysed acylation of these anhydrides by a second anhydride molecule leads to (two) P-ketoanhydrides and so to the ketones viu decarboxylation (Scheme 21). In theory, the product ought also to include propanone but this is either lost on work-up or is disfavoured at the acylation step by preferential formation of the more stable benzylic anion.PhCH2C02H + (MeCO),O -%PhCH2COMe + PhCH2COCHzPh Scheme 20 43 W. H. Perkin, J. Chem. Soc., 1886, 49, 317. 44 G. Blanc, Compres Rendus, 1907. 1356; Bull. Soc. Chim. Brlg.. 1908, 3, 118. 45 J. A. King and F. H. McMillan, J. Am. Chem. Soc., 1951. 73, 491 1. 46 G. G. Smith, J. Am. Chem. Soc., 1953, 75, 1134. 47 G. L. Buchanan and J. McArdle, J. Chem. Soc., 1952, 2944. 48 W. Wunderlich, Arch. Phurm., 1953, 286, 512. 49 D. P. N. Satchel], Quurt. Ren, 1963, 17, 177. 104 Buchanan I RCO (R = Me or PhCH,) Scheme 21 However, opinions differ on the mechanism of the acylation step and a number of suggestions have been made 45,47350,51 including a concerted cyclic model and both inter- and intra-molecular base-catalysed C-acylations.A kinetic study has revealed first-order kinetics with respect to the arylacetic acid in the presence of pyridine and excess anhydride. It also revealed an isotope effect of 1.13 with (I-I4C)anhydride but none with (l-14C)acid. This latter evidence casts doubt on the concerted mechanism but is consistent with Scheme 22,51 in which C-0 bond breaking (step ii) is rate determining. 7-Ac20 .LPhCH=C PhCHCO-OCOMe d PhCHC0-OCOMe\ -1 (ii) I OCOMe COMe Me/C-OCO Me + MeC0; 0 + PhCHCO; + Ac20 BH PhCH2COMe +C02 + 6I COMe Scheme 22 The reaction is of practical value only with suitably ‘active’ x-CH groups, as in phenylacetic acid.Thus it is reported 45 that P-phenylpropanoic acid affords no ketonic product, and phenoxyacetic acid reacts slowly giving little ket~ne,~~.~~ but p-nitrophenylacetic acid reacts rapidly to give a 48% yield of methyl ketone,46 and the highly activated pyrone (45) reacts even at room temperature (Scheme 23).53 Surprisingly, no ketone appears to be produced from diphenylacetic perhaps because of steric effects. In the course of his strychnine synthesis5’ OMe OMe Scheme 23 5o R. B. Woodward, M. P. Cava, W. D. Ollis. A.Hunger, H. U. Daeniker, and K. Schenker, Experienria, 1955. Suppl. 2, 213; Tetrahedron, 1963, 19,247.’’ G. G. Smith and D. M. Fahey, J. Am. Chem. Soc., 1959, 81.3391. 52 E. Dyer and C. E. Minnier, J. Urg. C/ieni., 1968, 33, 880.53 S. Yarnamura, K. Kato. and Y. Hirata. J. Chem. Soc., Chem. Comniun., 1968, 1580. 105 The Dakin- West Reaction Woodward made use of this ‘exceptionally simple method’ to convert the C( 14)-carboxyl group into methyl ketone during his construction of ring VI. It seems likely that the y-aromatic pyridone ring supplies the activation to the C(14)-H (Scheme 24). OAc0AJ COMe ph Ac20, p:dPh+PhPhCO,H G AcONa .-* Me CO,H (46) Scheme 24 The dicarboxylic acid (46) reacts selectively54 at the more active centre and similar selectivity was observed by Lawson ’ in o-hydroxyphenylsuccinic acid (47) where reaction with acetic anhydride alone leads to the keto-lactone (48) (Scheme 25). In the presence of pyridine, which promotes O-acetylation, the reaction is again selective, yielding (49). COMe C0,H &co2Hdo (48) OAc HO (49) (50) Scheme 25 s4 R.Stoermer and H.Stroh, Chmi. Brr., 1935, 68, 21 12. Buchanan Strangely, neither the para-isomer (50) nor phenylsuccinic acid, itself, undergo decarboxylative acetylation in the absence of a basic catalyst. This difference has received no explanation; however, Lawson has noted '3b that uncatalysed decarboxylative acylation is common to certain groups of succinic acids. These include N-benzoylaspartic acid (14) (see Section 2) whose reaction is smoother in the absence of added catalyst." Under these reaction conditions, it has been argued, 'reaction through an intermediate oxazolone is unlikely'. * 3b A further example of 'uncatalysed' acylation is provided by tricarballylic acid (51) 13' or its trisodium salt (Scheme 26).55 Although the latter gives poorer yields, this is surely a base-catalysed reaction, for improved yields of (52) are obtained by adding powdered glass, or using glassware previously washed with a strongly alkaline detergent.56 Decarboxylative acylation can be extended to homologues of acetic anhydride,46 but there is evidence that yields decline as the size of the alkyl group increases.' 3352-55 Johnson was able to overcome this shortcoming, by using the powdered glass technique described above, in his synthesis of (52) (R = C,H 7), which was a key intermediate in the synthesis 56of avenaciolide (53).It is not difficult to accept that those carboxylic acids which incorporate a basic group, i.e.tertiary amino-acids, can be acylated even in the absence of pyridine. Thus the conversion (Scheme 27) of the purinylthiopropanoic acid (54) into ketone (55) in substantial yield by acetic anhydride alone52is presumably such a case, because phenylthioacetic acid gives a poor yield of ketone with acetic anhydride in the presence of 2,6-lutidine and none at all in its absence.52 Likewise, certain y-and 6-t-amino-acids, e.g. (56), are converted into ketones by acetic anhydride alone (Scheme 28)57 and it seems plausible that here too, the molecule carries its own catalyst. Even so, the ketonic product is formed only if R is primary. If R is secondary or benzyl the lactam (58) is formed instead, via the intermediate (57).The behaviour of the t-r-amino-acids N,N-dimethylglycine (59) and its phenyl analogue (60) under Dakin-West conditions is of no preparative value, but does raise interesting mechanistic questions. These substances react with acetic an- hydride, alone or in the presence of pyridine, with evolution of CO, but the only product to be isolated has been N,N-dimethylacetamide (Scheme 29)?3' No reaction intermediate has ever been detected and the remainder of the reaction 0 COZH COZH R (51) (52) Scheme 26 '' R. Fittig. Liehigs Ann. C/imi.,1901, 314. 1. 56W. L. Parker and F. Johnson, J. Org. Chem., 1973. 38. 2489.'' P. A. Cruickshank and J. C. Sheehan, J. Am.Chem. So(,.,1961, 83, 2891 " J.A. King and F. H. McMillan. J. Am. C'/ieni. Soc.. 1951. 73. 4451. The Dukin-West Recicrion Me Me S AC2O (54) (55) Scheme 27 0 0 Scheme 28 R IMe,NCHCO,H Ac20* [Me,;:;COMe] Me2NAc + co, (59) R = H (60) R = Ph Scheme 29 0 100 "C CHzPh (621 Scheme 30 mixture is polymer. Nonetheless, it has been proposed 5x that N,N-dimethyl-acetamide is a decomposition product of the (hypothetical) ketone intermediate (61) and indeed, a specimen of this ketone (separately prepared) has been reported to give a 97:; yield of N,N-dimethylacetamide when boiled with acetic anhydride.'* Certainly, the evolution of CO, from (59) and (60) places them alongside (54) and (56)as self-catalysed examples of acylative decarboxylation.However, it is remark- able that the mixed anhydride (62) behaves so differently (Scheme 30), undergoing fragmentation at 100"Cwith evolution of carbon moi~osidc.,'~for analogous mixed anhydrides might be expected to arise from both (59) and (60) under the reaction conditions. '' V. I. Maksimov. Teir(iiidroti, 1965. 21. 687; Bull. Acd. .%i. USSR. Dir. Chrtn. Sci..1962. 99. I08 Buchanan R ACZO A+NH CH COGH-u-+ RCHO + C02PY HO-H +RCHO + CO (64) Scheme 31 Interestingly, the arylsulphonyl derivatives of x-amino-acids and their acid chlorides, (63) and (64), show a similar divergence of behaviour on treatment with base.60 Under Dakin-West conditions the acid (63) gives rise to an aldehyde and carbon dioxide whilst the acid chloride (64) reacts with base to yield the same aldehyde accompanied by carbon monoxide (Scheme 31).Both reactions have been identified by Grob 61 as examples of heterolytic fragmentation processes. AcknoMhdgement. The author is grateful to Professor Frank Johnson (State University of New York at Stony Brook, N.Y., U.S.A.)for his comments and for much stimulating discussion. "R. H. Wiley and R. P. Davis, J. Am. Chem. Soc., 1954, 76, 3496. 61 C. A. Grob and P. W. Schiess, Angew. Chem., hi.Ed. Engl., 1967, 6, 9 and 15. 62 J. Lepschy. G. Hofle, L. Wilschowitz, and W. Steglich, Liehigs Ann. Chem.. 1974, 1753. 63 R. H. Wiiey, J. Org. Chem., 1947, 12, 43. 64 A. Burger and C. R. Walter. J. Am. Chern. So(,.,1950, 72, 1988. " B. M. Goldschmidt. B. L. Van Duuren, and C. Mercado, J. Chern. Soc., (C), 1966, 2100.

ISSN:0306-0012    

DOI:10.1039/CS9881700091

出版商:RSC    

年代:1988

数据来源: RSC

摘要:

Chem. SOC.Rev., 1988, 17, 111-133 ROBERT ROBINSON LECTURE * Retrosynthetic Thinking-Essentials and Examples By E. J. Corey DEPARTMENT OF CHEMISTRY, HARVARD UNIVERSITY, CAMBRIDGE, MASSACHUSETTS 02138, USA 1 Introduction The ability of chemists to synthesize organic compounds has evolved through a number of discernible stages over the past 160 years in a progression which is marked historically by ascendance to a new and qualitatively higher level of sophistication at intervals of roughly 20 years. One clear sign of this advance is the achievement in a particular period of syntheses which are conceptually more complex and technically well beyond those realized in the preceding stage.’ The vigorous development of organic synthesis over the past decades is due in no small part to the earlier contributions of Sir Robert Robinson, a pioneer in 20th century synthesis.His insights into the relation between chemical structure, reactivity, and reaction pathways and his application of mechanistic thinking to synthesis spearheaded the advance during his active period (191@-1950) and helped set the course of modern synthesis., Also of great importance was his introduction of elegant new synthetic processes and reaction sequences, for example the famous Robinson annulation, which provided an early model and inspired subsequent generations of chemists toward the rational discovery of useful new synthetic methods. A. J. Birch wrote of Robinson, ‘His great assets in the general area of synthesis were his “feel” for mechanism and his prodigious memory for published work.’ Robinson’s genius surely also included a penetrating intellect and the ability to associate previously disparate facts in a highly original but logical way.Robert Robinson was giant in the field of chemistry. It is a great honour and privilege to present this article in his memory. This paper outlines the fundamentals of retrosynthetic thinking and illustrates the application of the method to the synthesis of three interesting organic compounds, (1)-(3), which occur in the ginkgo tree, Ginkgo biloba. The C,, * Presented at the Annual Chemical Congress of the Royal Society of Chemistry, University of Kent at Canterbury. on 12 April 1988. One indicator of the ever-increasing sophistication of organic synthesis is the survey of I.Fleming, ‘Selected Organic Syntheses’. Wiley-Interscience, 1973. A. R. Todd and J. W. Cornforth, ‘Robert Robinson’, Biograph. Mem. Fellows R. Sot., 1976, 22, 415. A. J. Birch, ‘Sir Robert Robinson: A Contemporary Historical Assessment and a Personal Memoir’, J. Proc. R. So(,.N. S. Wales, 1976, 109. 151. Retrosyn thetic Thinking-Essen tiuls and E-uimples 0 I I I I I I I I I t (3)Bilobalide Scheme 1 Structural rehiionship between ginkgdide A und bilobalide ginkgolide, ginkgolide B (l),” is a powerful antagonist of platelet activating factor and of great interest as a therapeutic agent.’ The C, ginkgolide, bilobalide (3),6 is structurally related to ginkgolides A4 and B and may derive from the common precursor ginkgolide A (2) (Scheme 1).Ginkgolide A displays strong insect antifeedant activity. The synthesis of the ginkgolides represents a formidable challenge to the synthetic chemist, unmet for more than two decades. 2 Essentials of Retrosynthetic Analysis Retrosynthetic (or ‘antithetic’) analysis is a problem-solving technique for trans-forming the structure of a ‘synthetic target’ (TGT) molecule to a sequence of progressively simpler structures along a pathway which ultimately leads to simple or commercially available starting materials for a chemical synthesis. The trans-formation of a molecule to a synthetic precursor is accomplished by the application of a ‘transform’,the exact reverse of a synthetic ‘reaction’,to a target structure.Each structure derived antithetically from a TGT then itself becomes a TGT for further ((1) N.Sakabe, S. Takada, and K. Okabe, J. Cliem. Soc.. Chem. Commun., 1967, 259; (b) K. Okabe, S. Yamada, S. Yamamura. and S.Takada. J. Cliem. Soc. C.. 1967.2201; (c) K. Nakanishi, Pure Appl. Chem.. 1967, 14, 89.’(a)P. Braquet, Drugs of’the Future, 1987, 12, 643; (15) P. Braquet and J. J. Godfroid, Trends Pharmacol. Sci., 1986, 7,397; (c) P. J. Barnes and K. F. Chung, hid., 1987, 8, 285; (d) B. Max, ibid., 1987, 8, 290. K. Nakanishi, K. Habaguchi, Y. Nakadaira, M. C. Woods, M. Maruyama, R. T. Major, M. Allauddin, A. R. Patel, K. Weinges, and W. Bahr. J. Am. Chem. Soc., 1971, 93, 3544. Corey analysis.In order for a transform to operate on a target structure to generate a synthetic predecessor the requisite structural subunit or 'retron' for that transform must be present in the target. The 'retron' for Diels-Alder reaction, for instance, is a six-membered ring containing a .n-bond and it is this structural subunit which represents the minimal 'keying' element for transform function. The logical and systematic application of retrosynthetic analysis depends on the use of higher level strategies to guide the selection of transforms. The general strategies which are available for devising retrosynthetic pathways fall into several classes including the Transform-based strategies-for example long range search or look-ahead to apply a powerfully simplifying transform (or a tactical combination of simpli- fying transforms) to a TGT with certain appropriate keying feature^.^ Usually the retron required for application of a powerful transform is not present in a complex TGT and a number of antithetic steps (subgoals) are needed to establish it.Structure-goal strategies-for example directed at the structure of a potential intermediate or potential starting material. Such a goal (S-goal) greatly narrows a retrosynthetic search and allows the application of bidirectional search techniq~es.'.'~" ' This is the oldest and most traditional of all synthetic strategies, and long the dominant strategy. The modification of a limited substructural region of a molecule, for example to accommodate the retron for some powerfully simplifying transform, leads to a substructure goal (SS-goal) which can provide the same kind of guidance as an S-goal.Topological strategies-for example the identification of one or more individual bond disconnections or correlated bond-pair disconnections as strategic.12 Topological strategies may also lead to the recognition of a key substructure for disassembly or to the use of rearrangment transforms. Stereochemical strategies-general strategies which remove stereocentres and stereorelationships under stereo~ontrol.~ Such stereocontrol can arise from transform-mechanism control or substrate control. In the case of the former the retron for a particular transform contains critical stereochemical inform- ation (absolute or relative) on one or more stereocentres. Stereochemical strategies may also dictate the retention of certain stereocentre(s) during retro- synthetic processing or the joining of atoms in three-dimensional proximity.' E. J. Corey, A. K. Long, and S. D. Rubenstein, S(,ience, 1985. 228, 408.'These strategies have been described previously in connection with the computer-assisted analysis of synthetic problems and the interactive program, LHASA, which is designed to emulate the problem solving techniques used by chemists. In turn. the LHASA project has been of great value in the development of new and general ways of thinking about synthesis. '((1) E. J. Corey, W. J. Howe, and D. A. Pensak. J. Am. Cheni. Soc.. 1974, 96,7724; (h) E.J. Corey, A. P. Johnson. and A. K. Long. J. Org. Chent.. 1980.45, 2051; (c) E. J. Corey. A. K. Long, J. Mulzer, H. W. Orf. A. P. Johnson, and A. P. Hewett, J. Cheni. InJ Comp. Sci., 1980. 20, 221. lo E. J. Corey. Quart. Rer. Cliem. Soc., 1971, 25. 45.5.' S. Hanessian. 'Total Synthesis of Natural Products: The Chiron Approach', Pergamon Press, Oxford, 1983. l2 E. J. Corey. W. J. Howe, H. W. Orf. D. A. Pensak, and G. Petersson, J. Am. C'lrem. Soc., 1975,97, 61 16. Retrosynthetic Thinking- Essentials und E.varnples 5. Functional group based strategies.’ The retrosynthetic reduction of molecular complexity involving functional groups (FGs) as keying structural subunits encompasses many important general problem-solving tactics. Single FGs or pairs of FGs (and the interconnecting atom path) can (as retrons) key directly the disconnection of a TGT skeleton to form simpler molecules. In addition FGs can signal the application of transforms which replace functional groups by hydrogen (FG removal in the retrosynthetic direction, clearly a simplifying process) or change the reactivity of FGs (corresponding to FG protection, activation, or interchange in the synthetic direction).Functional group interchange (FGI) is a commonly used tactic for generating from a TGT retrons which allow the application of simplifying transforms.’ FGs frequently key transforms which stereoselectively remove stereocentres or break topologically strategic bonds so that in effect they play a role in the other types of retrosynthetic strategies.Functional groups also may key the stereochemically dictated joining of proximate atoms to form rings in the retrosynthetic direction.I4 6. ‘Other’ types of strategies. The recognition of substructural units within a TGT which represent major obstacles to synthesis often provides major strategic input. Certain other strategies result from the requirements of a particular problem, for example economic requirements or a requirement that several related target structures be synthesized from a common intermediate. A TGT which resists retrosynthetic simplification may require that new chemical methodology be developed for a synthesis and thus suggest a line of research leading to the invention of new chemical processes. The recognition of obstacles to synthesis provides a stimulus for the discovery of such novel processes.One important human problem-solving strategy is the application of ‘imagination’ or ‘intelligent use of a chain of hypotheses’ to guide the search for an effective line of retrosynthetic analysis. This inductive problem-solving dimension has been described in somewhat different terms previously. ‘5. l6 ‘The synthetic chemist is more than a logician and strategist; he is an explorer strongly influenced to speculate, imagine, and even to create. These added elements provide the touch of artistry which can hardly be included in a cataloguing of the basic principles of synthesis, but they are very real and extremely important. Further, it must be emphasized that intellectual pro- cesses such as the recognition and use of retrons and synthons require con- siderable ability and knowledge; here, too, genius and originality find ample opportunity for expression.The proposition can be advanced that many of the most distinguished syn- thetic studies have entailed a balance between two different research philo- ’‘ E. J. Corey, R. D. Cramer 111. and W. J. Howe. J. Am. Chem. Soc,..1972. 94. 440.’‘ E. J. Corey and W. L. Jorgensen, J. An/. C’hcwi. So(,., 1976, 98, 203. l5 E. J. Corey. Pure App/. Clicwi.. 1967, 14, 30. The italicized words have been added to this quotation to clarify the original meaning since the word ‘synthon’ has now come to be used to mean synthetic ‘building block’ rather than retrosynthetic fragmentation structures.For an entirely different discussion of ’strategy in synthesis’ see P. Deslongchamps. Alr1ric.hc~hiniic.11Actci, 1984, 17, 59. Corey sophies, one embodying the ideal of a deductive analysis based on known methodology and current theory, and the other emphasizing innovation and even speculation. The appeal of a problem in synthesis and its attractiveness can be expected to reach a level out of all proportion to practical con- siderations whenever it presents a clear challenge to the creativity, originality, and imagination of the expert in synthesis.’ * Certain strategies not mentioned above relate to the question of optimization of a synthetic design after a set of pathways has been generated antithetically. Such strategies are used to determine the optimum ordering of synthetic steps, the use of protection or activation steps (or the avoidance of same),” or the determination of alternate or bypass paths for problematic segments of the ex-TGT tree of inter-mediates.Systematic and rigorous retrosynthetic analysis is the broadprinciple of synthetic problem solving under which the various types of strategies which are delineated above take their place. Another overarching strategy of great importance is the use concurrentlj, of as many independent strategies as possible to guide the search for retrosynthetic pathways.’ In general the greater the number of strategies which are used in parallel to develop u line of analjwk, the easier the analysis and the simpler the emerging synthetic plan is likely to be.During the past 20 years retrosynthetic thinking has permeated all areas of organic synthesis and, together with new methods and processes for molecular construction, has significantly enhanced the field. It is no longer possible to teach the subject of organic synthesis effectively without the extensive use of retro-synthetic concepts and thinking, as expressed previously.’ ‘These achievements (syntheses of the 1945-1 960 period) provided the impetus for further developments, which led to a great improvement in the power and elegance of synthetic planning. In the 1960s, general problem- solving strategies and methods that could be applied to the analysis of any complex synthetic problem were explicitly formulated, and the underlying principles of synthesis were defined in a way that made synthetic planning more logical, more systematic, and easier.The insights so gained had an impact on the teaching of organic synthesis as well as its practice. Even in the 195Os, synthesis was taught by the presentation of a series of illustrative (and generally unrelated) examples of actual syntheses. Chemists who learned synthesis in this inductive manner approached each problem in an ad hoc way. The intuitive search for clues to the solution of the problem at hand was not guided by effective and consciously applied general problem-solving techniques.’ ’ ” ((I) E. J. Corey, H. W. Orf. and D. A. Pensak.J. Am. Cliem. Soc., 1976.98,210; (h)E. J. Corey. A. K. Long, T. W. Greene, and J. W. Miller. J. Org. Clietn., 1985, 50. 1920. Retrosynthetic Tliinking-EEssential.s und E.umpkeJ 3 Retrosynthetic Analysis Exemplified A. Synthesis of Ginkgolides A and B.-The basic ideas of retrosynthetic analysis become much more tangible when illustrated by specific applications. The problem of synthesis of complex polycydic molecules such as the ginkgolides provides an especially useful testing ground since these molecules combine functional, topo-logical, and stereochemical complexity at high density. Also, it is not obvious from inspection of the structures of the ginkgolides which starting materials (achiral or chiral) are appropriate for molecular construction.The concurrent use of several different types of retrosynthetic strategies is essential because antithetic processing of these TGTs in an opportunistic way would lead to an enormous number and variety of synthetic possibilities, most being of dubious value. Retrosynthetic analysis of the ginkgolide B structure (1) was carried out by the concurrent use of several different strategies to guide the antithetic search. One particular retrosynthetic path toward which there was strong strategic convergence is outlined in abbreviated form in Scheme 2. For brevity several retrosynthetic steps (transforms) are combined in each of the retrosynthetic changes shown. This line of analysis was selected for experimental study and the validity of the general plan was demonstrated by successful execution of the synthesis.Nonetheless, it is important to make clear at this point that several of the individual steps and countless procedural details of the synthesis had to be developed by experimentation. 1. From the category of transform-based strategies a number of useful guides emerged. For example, the hydroxy lactonization transform and the aldol transform can be applied directly (in that order) to the A/C ring portion of (1) with consequent reduction in the number of stereocentres (by four), rings (by one), skeletal carbons (by three), functionality, and chemical reactivity level. The application of two simplifying transforms in tucticuf combinution is directly signalled because the retron for the first is present in the TGT and transform application produces the retron for the next transform to operate.The molecular simplification which results is considerable. 2. The above notwithstanding, in a practical sense it is necessary to remove the possible interference of the x-hydroxy carbonyl FG pair of the F ring in (1) before the simplifying Tfs cited above can be applied. Thus, from the category of FG-based strategies a less reactive precursor or 'equivalent' of the a-hydroxy carbonyl FG pair must first be generated. It is not uncommon that such retrosynthetic changes in FG reactivity must be made before application of the 'goal' transform for simplification even though the retron for the latter may be present in the TGT.This situation also exemplifies the enforced con- current use of transform-based and FG-based strategies. Retrons for the direct removal of ring F are absent. 3. Another reason for removing ring (I' at the earliest retrosynthetic stage is the identification of ring B for preservation (origin ring) during retrosynthetic disconnection, based on its carbocyclic nature and its substitution by a large substituent at a stereocentre which is a strong candidate as an ab initio ==3 n-0 0 u PO0 (6) (5) COOH K-(7) (8) Scheme 2 117 Re trosjn the tic Thinking- Essen tids ctnd Esamples controller of stereochemical elaboration. The t-butyl-bearing stereocentre at C(8) can potentially direct the development of the adjacent C(9) and then the C(5) and C(6) stereocentres on the B ring.This stereochemical strategy con- verges with purely topological and FG-based (reactivity) strategies which also argue for the origin status of ring B in (1). 4. After the generation of precursor (4) from (1) it is quite clear that ring B is the best candidate as origin ring in this structure as well as in (1). The A ring in (4) lacks any qualification as a possible origin ring. 5. The C-0 bonds within ring E of (4) are strategic for disconnection. They are in the maximally fused ring of (4) (fused to rings A, B, D, and F) and also eso to three rings. They are termed esendo or more specifically hetesendo bonds (since a C-heteroatom bond is involved ''). Transforms exist for sequentially disconnecting one or both of these Iiete.umdo bonds.However, because of possible functional group reactivity interference with the valid use of these transforms, it is advisable to remove retrosynthetically the keto function on ring A of (4) prior to cleavage of ring E. Retrosynthetic precursors such as (5) or the 4-deoxy analogue result. 6. Precursor 4-deoxy-(5) is also suggested by the application to (4) or (1) of key transform-based strategies which are aimed at ring disconnections. Among the key ring-transforms whose retrons map partially on to structures (1) and (4) are some which are capable of simultaneous disconnection of a pair of e.uendo bonds vicinal to the primary ring. These transforms correspond to internal cycloaddition transforms. One such specific transform is the ketene-olefin internal cycloaddition transform ' in tactical combination with the Baeyer- Villiger transform. Another is the recently developed ' internal carbolactoniz- ation (Mn"'-mediated) transform.Antithetic multistep search to apply these transforms to either (1) or (4) leads (with disconnection rings c and E or ring E, respectively) to precursor (7). 7. Strategic bond and FG-keyed strategies suggest disconnection of (7) to form precursor (8). The steps for the synthesis of (8) and for its conversion into ginkgolide B (1) will now be considered. The initial synthetic studies which led to the synthesis of (*)-( 1) have recently been described 2o as has a modification which leads enantioselectively to the naturally occurring and chiral form of (1).21 Only the latter route will be outlined here.Scheme 3 records the enantioselective synthesis of spiro-ketone (8) from 2-(2,2-dimethoxyethyl)cyclopenten-1-one. A key step in this transformation was the highly selective borane reduction of the starting cyclopentenone catalysed by the chiral oxazaborolidine derived from (S)-2-(hydroxydiphenylmethyl)pyr-rolidine and methylboronic acid.2' For this and numerous other cases, this process is outstanding because of the high enantioselectivity which can be achieved l8 E. J. Corey, M. C. Desai, and T. A. Engler. J. Am. C'kern. Soc., 1985. 107, 4339. l9 E. J. Corey and M.-c. Kang. J. Am. Chen~.Sor.. 1984, 106, 5384. 2o E. J. Corey. M.-c.Kang, M. C. Desai, A. K. Ghosh, and I. N. Houpis, J. Am. C'lirm.Soc.. 1988. 110, 649. 21 E. J. Corey and A. V. Gavai, Terralierlron Leri.. 1988, 29, 3201. 118 @Ph 1.10 mot% OH Me (f-BuCO)pO, Et3N w 2. HCI, MeOH DMAP, CH& enantlose/ecflvlty96.5 :3.5 [a]23D+36.6' (C 2.5, MeOH) 1. BH3, THF t-BuMgCI, CuCN 2. NaOH, H202 EtzO, -2OOC 3. PDC, mol. sleves,CH2CI2 OMe 9, 1. (8 1 cl9 Scheme 3 G Retrosyn the tic Thin king- Essen tials and Examples and the predictability of the absolute configuration of the produ~t.~'-~* This methodology for enzyme-like, catalytic enantioselective reduction provides an excellent route to chiral allylic alcohols, which as a class are enormously useful as starting materials for the construction of complex chiral organic molecules.25 The all-important introduction of the t-butyl group at the primary synthetic stereo- centre was achieved by stereoselective anfi S,2' displacement 26 with t-butylmagnesium chloride and 2-3 mole "/d of cuprous cyanide as reagent.The effectiveness of the origin stereocentre and its t-butyl substituent as a controller for the stereoselective generation of new stereocentres was demonstrated by the remainder of the sequence shown in Scheme 3 which led specifically to the required diastereomeric spiro-ketone (8). Scheme 4summarizes the reaction sequence used for the attachment of two rings (A and D) to the ketone (8). The enol triflate of (8) underwent efficient palladium(0)- catalysed cross coupling 27 with the ortho-ester of 2-pentynoic acid to form (9) which was reduced and deprotected to give diene-acid (7), one of the key retrosynthetic structures shown in Scheme 2.Conversion of (7) into the corre- sponding acid chloride and slow addition of the latter to a solution of tertiary amine in toluene at reflux resulted in ketene formation and subsequent internal addition 27 to form stereospecifically tetracyclic ketone (10). Elimination of methanol from the 1 -methoxy-tetrahydrofuran subunit also occurred under the reaction conditions, evidently promoted to the trialkylammonium chloride present in the reaction mixture. Selective Baeyer-Villiger ring expansion could be accomplished using alkaline trityl hydroperoxide (but not HOO-or Bu'OO-) 2o as reagent to produce tetracyclic lactone (11) in 82% yield and 100% enantiomeric purity after recrystaIIization.2 Lactone (1 1) was readily a-oxygenated by the Davis method 28 and combined with methanol under acid catalysis to form a-hydroxy-lactone (12), a potential substrate for the generation of the E ring (Scheme 5).Reaction of (12) with lead tetraacetate and iodine proceeded rapidly and cleanly to afford not the desired pentacycle (14), but the isomeric ring-closure product (13) as shown in Scheme 5." Although the formation of (13) was not useful as a synthetic step, it did provide confirmation of the stereochemistry of the intermediates (10) and (1 1). The intro- duction of the two C-0 linkages required for intermediate (4) was accomplished successfully as indicated in Schemes 6 and 7.The need for multistep sequences to achieve the conversion (11) + (4) is an indication of a considerable gap which still exists in the methodology for the selective oxygenation of organic compounds. 22 E. J. Corey, R. K. Bakshi, and S. Shibata, J. Am. Chem. Soc., 1987, 109, 5551. 23 E. J. Corey, R. K. Bakshi, S. Shibata, C.-P. Chen, and V. K. Singh. J. Am. Chem. Soc., 1987,109, 7925. 24 E. J. Corey, S. Shibata, and R. K. Bakshi, J. Org. Chem., 1988, 53, 2861. 25 For example, total synthesis of forskolin; see E. J. Corey, P. D. S. Jardine. and J. C. Rohloff. J. Am. Ckem. Soc., 1988, 110, 3672; Tetrahedron Left., 1988, in press. 26 See, E. J. Corey and N. Boaz, Tetrahedron Lett..1984, 25, 3063 and refs. cited therein. 27 See, E. J. Corey, M. C. Desai, and T. Engler, J. Am. Chem. Soc., 1985. 107. 4339. (a)F. A. Davis and 0.D. Stringer, J. Org. Chem., 1982,47, 1774; (h)F. A. Davis. L. C. Vishwakarrna, J. M. Billrners, and J. Finn, J. Org. Chem., 1984. 49, 3241. 120 Corey E c3 3 rn c- I 0 I3 2 121 Retrosynthetic Thin king-- Essen tials and Examples Scheme 5 TICI,, CH,CI,, H S ~ 25' S # H $3+,-, 0 98% PDC, AcOH CH,CI,, O", ihr P 0 Isomers O 2 : 1 , 80% MeOH -CH,CI, H,IO,, 00, 4hr HH9;0 0 75% Scheme 6 122 Corey 0 Li ,O', 30 min 0 PhS02N'q lh @O h-00 CH,CI, -MeOH CSA, 25', 48hr 1. NBS, CCI,, hv, 5' 2. ASNO, -CH,CN, 25' 3.DBU, C,H, -H20, 25' < 4. PDC, CH,C12, 25' 4-0 h-0 60% 65% (4) Scheme 7 The completion of the synthesis of ginkgolide B is shown in Schemes 8 and 9. A very critical (and in fact experimentally difficult) part of the synthesis was the attachment of ring c. As is often the case a straightforward retrosynthetic change (1) -(4)corresponded to a highly challenging segment of the laboratory synthesis. Of the many conceivable processes for the attachment of ring c only one could be demonstrated experimentally, and then only under carefully defined conditions. Elimination of methanol and alkaline epoxidation of (4)afforded epoxy-ketone (15) stereospecifically. Reaction of (15) with lithiated t-butyl propionate (16) in tetrahydrofuran-hexamethylphosphoric triamide (HMPA) produced (1 7) stereo-selectively. The use of HMPA, the presence of a 1,2-epoxy function, the trigonal centres at C( 10) and C( 1 l), and the reaction temperature of -30 "C are all crucial to the success of this aldol process.Reaction of enone (4)with the lithium enolate of t-butyl propionate at -78 "C leads selectively to the opposite mode of attack at C(3) (bottom face, trans to ring E). Camphor 10-sulphonic acid catalysed the lactonization of (17) to give (18) which was converted into ginkgolide B in four steps as outlined in Scheme 9." The hexacyclic lactone (18) was also transformed into ginkgolide A as indicated in Scheme Transformation of (1 8) to the imidazolethiocarbonyl derivative (19) and subsequent reaction with tri-n-butyltin hydride gave (20) which was 2q E.J. Corey and A. K. Ghosh, Tetruhrciron Lert., 1988, 29, 3205. Retrosynthet ic Thin king-Essen t ials and Examples OMe H4-0 1. PPlS,PhCl 135% -2. PhSCOOH R,N* iPr0 4 CSA ,CyClz ~ Bw-0- u O O C ~ HO’ O’H d (18) (17) Scheme 8 HO 1. TBMSOTI 0 Me’ Me’ HO 0’H 0 Scheme 9 124 Corey hydroxylated and oxidized to the a-hydroxy lactone (2 1). The 9-epi configuration was inverted by an oxidation-reduction sequence to produce ginkgolide A. It is noteworthy that the reactions of silylated (18) and (20) with osmium tetroxide show opposite stereoselectivities. Ginkgolide A was also synthesized from ginkgolide B using the analogous C( 1) deoxygenation process.29 B.Synthesis of Biloba1ide.-Although the structures of the C,, ginkgolides, A and B, and bilobalide (3) are clearly related to one another, retrosynthetic analysis of structure (3) produces completely different pathways for synthesis. The presence of only one carbocyclic ring in (3) and the types of carbon substituents on that ring drastically change the nature of the appropriate strategies (transform-based, topological, stereochemical, and FG-based). Another difference is the occurrence on the t-butyl-bearing ring carbon of (3) of a hydroxyl function which indepen- dently affects the stereochemical and FG-related strategies to be considered. The abbreviated retrosynthetic plan shown in Scheme 1 1 illustrates these points.The synthetic introduction of a t-butylcarbinol subunit at C(1) by conventional chemical processes poses several non-trivial problems. For example, the reaction of t-butyllithium with sterically screened cyclopentanones or even cyclopentanone itself does not lead to efficient 1,2-carbonyl addition, but to other reaction modes stemming from x-deprotonation. Other t-butyl metalloids are either unreactive or similar in behaviour to t-butyllithium. The t-butylcarbinol grouping also is a likely source of interference with a number of possible synthetic constructions. Thus, the recognition of the obstacles to synthesis associated with the t-butylcarbinol subunit provides strategic guidance, since it suggests the use of specific approaches, for instance that the t-butyl group be introduced before the oxygen atom at C(1).The first retrosynthetic change shown in Scheme 11 (which actually corresponds to several individual transforms) includes a functional group interchange (FGI) step which replaces the hydroxyl at C(l) by a 1,2-double bond. In addition the X-hydroxy-lactone unit at C(6)/C(7)is replaced by a less reactive equivalent such that the D ring of (3) is modified to the 2-methoxytetrahydrofuran system. The deoxygenation at C(6) is also suggested by stereotopological and FG-based strategies. Topologically the presence of two functionalized two-carbon append- ages on adjacent atoms CC(4) and C(5)] of ring A suggests the application of an appendage-connection strategy 4-all the more because these appendages are cis to one another on ring A and can be made identical by retrosynthetic removal of the hydroxyl at C(6).The two one-carbon appendages at C(4) and C(5)are also cis to one another and convertible into identical groups by redox FGI. Thus, the parallel application of these well-known strategies guides the retrosynthetic conversion of (3) successively into (22) and (23). Application of the connective FG-pair transform l4 C=C oxidative cleavage to (23) produces the bicyclic keto-diester (24). Although intermediate (24) has the full retron for application of the Diels-Alder transform, one of the precursors generated from (24) by that transform is the super- reactive 2,3-dimethoxycarbonyl-4-t-butylcyclopentadienone.Although the syn- thesis of this substance might be possible, it would probably be highly reactive in o<y ox 0 = -3 II o=cn h cyY a Corey cH30032CH300 -0H000--(241 (23) Scheme 11 Diels-Alder dimerization and, hence, did not seem to be an outstanding candidate for the synthesis of (24).Fortunately, another plan for the synthesis of (24) was developed which was more interesting conceptually and which also had the potential to produce (24) enantioselectively. Another possible problem with the retrosynthetic plan shown in Scheme 11 is associated with the generation of intermediate (23) which, though possessing a symmetrical subunit for further processing and eventual disconnection, contains duplicate FGs which have to be differentiated in the course of the synthesis.While such FG redundancy potentially can invalidate a synthetic approach, in this instance it poses no real problem since the FG placement and reactivities of (23) allow a straightforward synthetic differentiation. The retrosynthetic analysis out- lined in Scheme 11, in fact, provided the basis of a successful enantioselective synthesis of (3) which will now be described briefl~.~'.~~ Reaction of the (+)-menthol diester of fumaric acid with butadiene and diiso- butylaluminum chloride in 1:1 hexane-methylene chloride produced the (R,R)-Diels-Alder adduct 32 (obtained in 85% yield in pure form after chromatography) 30 E. J. Corey and W.-g. Su, J.Am. Chem. SOC.,1987, 109, 7534 [synthesis of (i)-bilobalide]. 31 E. J. Corey and W.-g. Su, Tetrahedron Lett., 1988, 29, 3423 (enantioselective synthesis of the naturally occurring form of bilobalide). 32 K. Furuta, K. Iwanaga, and H. Yamamoto, Tetrahedron Lett., 1986, 21, 4507. 127 Retrosyn the t ic Thinking-Essentials and Examples-wOOMen(+) THF, -78' "COOMen(+) (25) [a?2D3= +25.5' (C 8.0 CHCI,) tBuCrCCOOPh 40°, 12 hours I 0.5 equlv KN(SIMe3)24 GB"THF, 40'd5"R'OOC 0 R'OOC 0 (27) (26) [ag3=49.9' (C 0.88, CHCI,) [a12D3= +50.8' (c 2.72, CHCl3) 93%80% Scheme 12 with a diastereoselectivity of 43: 1 (Scheme 12, R* = menthyl). Mono-deproton- ation of (25) and reaction with phenyl 3-t-butylpropiolate afforded the Claisen product (26) which underwent base-promoted cyclization to form (27), the di- menthyl ester corresponding to (24).The mechanism of this interesting cyclization has been disc~ssed.~' The conversion of (27) into (24) was unsuccessful and so (27) was utilized directly in the synthesis. Although reduction of the methyl ester ketone (24) with sodium borohydride produced the desired allylic alcohol (HO and COOCH, trans),30 the corresponding reaction with the bismenthyl ester followed the opposite stereochemical course (Scheme 13)., Fortunately, borane reduction catalysed by the (R)-oxazaborolidine shown in Scheme 13 led to the desired allylic alcohol (28) (as predicted) with 23 :1 diastereoselectively. The transformation of (27) into (28) is an interesting example of the enzyme-like power of chiral oxaza- borolidines to overcome a strong intrinsic stereochemical bias of a substrate in carbonyl reduction.Despite considerable experimentation with a variety of Corey (+)Ma U 0.6 eqBH, RHF (+)MaBu (+)MaBu c (+)MenOOC 0 (+)MenOOC OH (+)MenOOC OH (27) (28) . .. 45% 2?40 NaBH, /methanol rn S% NahPrOH-Toluene WO 25% (no reduction with borane-THF, Sm12, Zn(BH4)2) Scheme 13 reducing agents, no other method for the stereoselective synthesis of (28) from (27) could be found. The allylic hydroxyl function provided a key for distinguishing between like sub- stituents on the five-membered ring. Ozonolysis of (28) followed by treatment with acid afforded the bicyclic diester (29) as shown in Scheme 14.Differentiation of the two ester functions and generation of the ring system of bilobalide was effected in four steps which gave (30) as indicated in this Scheme. Exposure of (30) to potassium hydroxide resulted in an interesting cleavage of the 7-acetal function to form (31), clearly by a sequence involving y-lactone hydrolysis and temporary cleavage of ring D to a dialdehyde-carboxylate inter- mediate (not shown) (Scheme 15). Reaction of (31) with methanesulphonyl chloride-triethylamine afforded an unusually stable (and isolable) 2-chlorotetra- hydrofuran derivative which underwent base-promoted elimination to produce dihydrofuran (32). Treatment of (32) with peroxy-3,5-dinitrobenzoicacid un-expectedly proceeded more rapidly at the C( 1)-C(2) double bond (t-butyl sub- stituted) than at the 6,7-double bond (vinyl ether).Therefore, (32) was converted into the bis-epoxide (33) using an excess of peracid. The D ring epoxide unit in (33) was also unusually stable for a 2,3-epoxytetrahydrofuran, as evidenced by the fact that (33) could be chromatographed over silica gel at 23 "C.It would seem that the stabilities of the chloride of (31), the dihydrofuran unit in (32), and the 2,3-epoxytetrahydrofuran unit in (33) are all a consequence of the electron- withdrawing carboxylate substituent at C(8) which reduces strongly the electron Re t rosyn the tic Thinking- Essent ials and Examples 1. OdCH2C12 -CH30H (+)Meno, c NaHC03, -78" (+)Men02 .. 2. TSOH/CH~OH/CH(OCHJ)J (28) 60°, 8 hr (29) y3 [a]~= + 160.1" (C 0.60, CHCI3) 75% 10:l LAH/THF55" I CH2C12, -78 -40" 2. 1N HCITTHF. 0" 3. PDC/4A Sleves/CH2C12 62% [aff + 145.5' (c 1.45, CHC13) Scheme 14 density about the second oxygen at C(8). Establishment of the ring B lactone function and the ring D a-oxylactone unit was next accomplished using the selective reactions outlined in Scheme 16 which converted bis-epoxide (33) into the mono- epoxide-trilactone (34). The synthesis of bilobalide from (34) was completed by the steps shown in Scheme 17. Various attempts to cleave the 1,2-epoxide function in (34) with attachment of hydrogen at C(2) were unsuccessful. Consequently, (34) was deoxygenated to (35) in a novel way by heating with triethylsilane.Hydroxylation of the 1,2-double bond of (35) proceeded stereospecifically and the resulting diol was deoxygenated as shown 30.33 to give, after deacetylation, synthetic bilobalide. 4 Epilogue This article has outlined some of the essential ideas behind the retrosynthetic approach to the planning of organic syntheses, and their application specifically to syntheses of (l), (2), and (3). For several reasons one hopes that the subject matter 33 S. C. Dolan and J. MacMillan J. Chem. Soc,., Clirw7 Commztn., 1985, 1588. 130 Corey 1.1MKOH 95% EtOH-THF 2. H30+ 1 MSCI/Et,N 2. IPr2NEVCH,CN, Reflux T DNPBNNaHCO, (331 1OCHs I(32) OCH, [a]?: + 151.3"(c 0.3, CHCI,) [a%3"= -29.7"(~0.33, CHCI,) Scheme 15 will be of interest to a broad range of chemists and students of chemistry.First of all, it provides a snapshot of contemporary organic synthesis for comparison with the past, or with events of the future as the science makes its way toward the close of the twentieth century, a period of enormous progress. Second, the emphasis on the logic underlying synthesis is intended to reflect the conceptual nature and beauty of modern synthesis and the value of the continuing study of its intellectual found- ations. And last, although chemistry is a science of vast scope and countless specialties, it has no real boundaries and many of its activities (including synthesis) interact in a synergistic way. Some aspects of organic synthesis have not changed since the era of Sir Robert Robinson.The laboratory execution of a multistep synthetic sequence is still an arduous task requiring time, ability, and effort. Although it is now much easier to analyse complex synthetic problems under retrosynthetic and strategic guidance, Retrosyn the tic Thinking-Essentiuls and E-xumples 1.1N HCUTHF, 80" 2. A%O/DMAP/CH,CI, 92%v 6CH, (33) dCH, I 1.1N HCUTHF-HOAC. 100" 2. PDC/4A Sieves, CH2Clz .... €%% (34) Y Synthetic: [a]:= -llOo(c 0.41, CHCI,) mp 240-244°C Natural: [a]:= -107.5O(c 0.40, CHCI,) mp 242-248OC Scheme 16 and although organic chemists now have at their disposal powerful new tools for analysis and separation of molecules, determining suitable conditions and optimum procedures for synthetic processes ('making reactions work') is as much of a challenge today as it ever has been.Experienced synthetic organic chemists are keenly aware of this fact and also of the many large gaps which exist in synthetic methodology. It is only through continued vigorous effort at the frontiers of synthesis that such deficiencies will be appreciated and remedied. Because of this fact and also the obvious importance of synthesis to human well-being, for example as the primary tool in the development of new therapeutic agents, it might appear that all are agreed on the need for a high degree of activity in synthesis for the foreseeable future. Not so. There are now, as there have been for many years, savants and prophets (including more than a few chemists), who proclaim that organic synthesis is a mature field with limited possibilities for future discovery.In truth, as Sir Robert Robinson would have appreciated, the frontiers and the Corey Et3SiH/toluene 300°/seJed tube 909b 1. OsOJether-pyridine 23O, l2 hours 2. methyl oxalyl chloride iPrzNEt/CH3CN 9% ~ 1. Bu,SnH/AIBN reflux toluene,s% 2.3N HCI, reflux, 70% Scheme 17 progress of organic synthesis are limited only by the creative abilities and the vision of those who work in the field. Organic synthesis is the essence of organic chemistry, just as organic chemistry is the fundamental language of life. Acknorcledgement. It is with much pleasure and gratitude that I acknowledge my collaborators in the successful syntheses of ginkgolide B, ginkgolide A, and bilo- balide which are described herein.These individuals are Myung-chol Kang, Manoj Desai, Arun Ghosh, Ioannis Houpis, Weiguo Su, and Ashvinikumar Gavai. I am also deeply indebted to the US. National Institutes of Health, the US. National Science Foundation, and Pfizer Inc. for generous financial support over the years.

ISSN:0306-0012    

DOI:10.1039/CS9881700111

出版商:RSC    

年代:1988

数据来源: RSC

摘要:

Clzem. SOC.Rev., 1988, 17, 135-146 NYHOLM LECTURE * New Trends in Chemical Education and Chemistry Teacher Education World wide By Marjorie Gardner LAWRENCE HALL OF SCIENCE, UNIVERSITY OF CALIFORNIA, BERKELEY, CA94720, USA 1 Introduction I have waited a long time for this day. In August of 1985, I received a letter from the President of the Royal Society of Chemistry inviting me to deliver the Nyholm lectures and accept the Nyholm Medal during the 1986/87 academic session. This very happy news was followed by a series of spinal operations that made it impossible for me to come to the United Kingdom last spring to lecture and participate in the Ceremonies and International Symposium in London. This disappointment has been replaced by the excitement and anticipation of coming here for this occasion.I am very pleased to be invited and to be able to be with you today. I met Sir Ronald Nyholm on at least two occasions; no more impressive chemist existed. He was not only a giant in research and a great supporter of chemical education but also a man of personality, style, and good humour. I first met him at an international meeting held at the American Chemical Society in Washington in 1968. Later at a dinner, I sat next to Sir Ronald and watched him match wits with another noted storyteller, Henry Teterin from the USSR. Sir Ronald was delight- ful. A few years later in an IUPAC Chemical Education Symposium, I had the opportunity to see and hear the depth of this man’s commitment to the improve- ment of chemical education.I am very honoured by receiving this award and by the opportunity to speak today, in remembrance of Sir Ronald Nyholm and his commitment to excellence in chemical education. 2 New Directions/Trends Some new directions in chemical education are evident in the world today. I am going to note three major changes that are occurring. Throughout this lecture I will, in one way or another, endorse and expand on these three as I highlight the role of the teacher: Learning to use new technological tools in classroom and laboratory Dealing with social issues and the higher order thinking required (e.g. decision making, problem solving) * Delivered at the Annual Chemical Congress of the Royal Society of Chemistry, University of Kent at Canterbury, 13 April 1988.New Trends in Chemical Education Providing chemical education for all; the need for scientific literacy. As I develop these themes, I will focus quite specifically on one component of chemical education, teacher education. Teachers hold the keys to learning success for the students; teachers are the multiplying factors. Each chemistry teacher in a year influences more than a hundred students, and in a career, thousands of students. The degree of preparedness of the teacher often determines the quality of science education a nation provides. It can be high quality and support a nation’s goals or it can be low quality and grievously interfere with manpower development and public understanding.It is the teacher who can motivate some students to seek careers in science and technology and educate all others to the point where they achieve a functional level of scientific literacy. When we identify the most significant factor promoting excellence in chemical education, it is the teacher that matters. Confidence, commitment, creativity, and compassion are well identified characteristics of a good teacher. While a tendency toward these characteristics may be innate, they are brought to the forefront by experience and education and they are refined over time. The quality of the preservice experience sets the expectations of the teacher and the continuing inservice education of practising chemistry teachers allows good teachers to become even better.Inservice education programmes for chemistry teachers have been evident in many nations around the world since the early 1960s at least. Perhaps they existed before that time, but were not well known or well documented. With the advent of curriculum reform in the early 1960s, as characterized by ChemStudy in the United States and the Nuffield Chemistry Programme in the United Kingdom, and their diffusion to many areas of the world, the need to prepare practising teachers to use new curricula in their classrooms led to the development of inservice programmes. These have continued to evolve to meet national needs. 3 International Status of Teacher Education What is happening around the world in chemistry teacher education? I have recently completed a study for UNESCO where I had the opportunity to speak with chemical educators in person at international meetings or to correspond with the leaders in many nations.’ My findings indicate that teacher education has settled into a relatively routine pattern in most areas of the world.The pattern features inservice workshops and courses designed to upgrade the teacher’s know- ledge of chemistry and teaching skills. Very few of these programmes are helping teachers to address the three trends noted earlier: high technology, social issues, and scientific literacy. In a few areas of the world there are definite stirrings and new activities. There is growing recognition of the importance of the teacher’s role in nations around the world whether they be developed or developing.Let me cite just a very few examples (Figure 1). Marjorie Gardner, ‘Chemistry Inservice Training’, New Trends in Chemistry. Volume Vl. UNESCO. Paris (in press). Gardner THE INTERNATIONAL WORLD OF TEACHER EDUCATION Western Europe Asia Latin America Middle East USSR & Eastern Europe Africa North America Figure 1 In Latin America, the Tenth International Conference on Chemical Education was held last summer in Sao Paulo, Brazil. More than six hundred participants, mostly from Latin American nations, came together to learn from one another as well as from the excellent panel of speakers. The next step in Latin America is a very ambitious conference to be held in Argentina this summer.It is the First International-Argentinian Meeting on Methodology in the Teaching of Chemistry, scheduled for June 22-25. Invitations have gone out to leading chemical educators throughout Latin America and a number of other countries inviting them to come together to examine how a teacher performs in the classroom and how to improve this performance--the approaches and methods of teaching chemistry. In Asia, the Chinese government has just published a very detailed teacher’s guide and sourcebook for chemistry teachers. It begins with sections on methods of teaching (e.g. preparing for and conducting the laboratory, leading discussions). The majority of the book is devoted to chemical reactions, chemical and physical properties of elements and compounds, common uses of chemicals, etc.This compendium of chemical knowledge is being distributed to all the provinces to assist the chemistry teachers in that vast nation. In Japan, the inservice teacher education programmes are carried out through the science education centres in each prefecture. This localized system provides opportunity to every chemistry teacher in Japan for regular updating courses on content and techniques, including the use of computers and new audio visuals. One note worthy of mention, the well prepared Japanese chemistry teacher at the school level is paid a salary and accorded status very near to that of the university professor. In Eastern Europe, the pattern is relatively consistent across the USSR and other European nations.Chemistry teachers are prepared in five year preservice pro- grammes that include at least three years of intensive discipline-specific content learning. In the last two years, supervised student teaching, projects related to teaching, extensive training in methods, use of audio visuals and demonstrations become central in the curriculum. A particularly interesting inservice programme in the USSR comes from the Novosibursk University which sends out its post- graduate students across the vast areas of Siberia that this university serves. The purpose of their travels is to identify talented students for admission to the specialized science boarding school, and also to instruct teachers in the rural areas New Trends in Chemical Education in order to enhance their knowledge of chemistry and bring it up to date.In the Middle East, Israel has been a very active nation for nearly thirty years in providing continuous curriculum and teacher education reform. Other areas of the Middle East follow rather traditional patterns with limited inservice workshops and teacher preparation courses available. In Africa, activity is limited primarily to local teachers in the area of universities. In India, the National Council for Educational Research and Training provides leadership to the various states. There is no need for me to speak about the United Kingdom and Western Europe. You are the experts and I have come here to learn more from you. The conclusion in the UNESCO study is that there is not much that is innovative going on in teacher education in large segments of the world.Much more needs to be done. 4 Chemistry Teacher Education in the United States Now I will turn attention to that which I know best. The need for improved teacher education in the United States and some new programmes that are emerging to help meet that need. The literature abounds with evidence of need and recom- mendations for change.2p11 A recent research study indicates that nearly all 77 000 teachers of science at the secondary level in the United States teach at least one course out of their field of preparati~n.~ More than half of these teachers have their primary teaching assignments in a field other than the one for which they were initially prepared.Sometimes called ‘crossover teachers’, these are people who might be fully prepared in biology teaching and credentialed in that field, for ’‘1986 87 Nationwide Survey of Secondary School Teachers’. A working paper for the Advisory Workshops, American Institute of Physics, February 17.1987. (A study of physics teachers and students in American secondary schools, 1986 87.) ‘1987 Survey of Member Science Centers’. Association of Science and Technology Centers, 1987. (Statistics about teacher education programmes offered at science and technology centres.) ‘Report of the 1985 -. 86 National Survey of Science and Mathematics Education.. Research Triangle Institute. November 1987. (Statistics showing degree of science preparedness among American teachers, K-12.) ‘Science Achievement in Seventeen Countries: A Preliminary Report’.International Association for the Evaluation of Educational Achievement (IEA), 1988. (A comparison of science achievement of American students with science achievement of students in 17 other countries.) ‘‘Survey Analysis of U.S. Public and Private High Schools: 1985 86’. draft copy. National Science Teacher’s Association, Washington, D.C. April 1987. (Statistics regarding teacher preparedness in the sciences, and science course availability.) ‘Teacher Supply and Demand in California: Is the Reserve Pool a Realistic Source of Supply’. Policy Paper No. PP86-8-4, Policy Analysis for California Education. August 1986.(Statistics to project the need for increased number of teachers in California for the next 5 and 10 year increments.) ’‘Opportunities for Strategic Investment in K-12 Science Education; Options for the National Science Foundation’, SRI International. May 1987. (Mandated by Congress, this is a study of programmes, options, and recommended initiatives and budgetary levels.) Marjorie Gardner. Testimony before United States House of Representatives Subcommittee on Science. Research, and Technology, Committee on Science. Space and Technology, Washington. D.C. March 22. 1988, pages 23 to 30. lo Carnegie Forum on Education and the Economy. A Nnlion Prc.pcirrtl: Tcwi./wrs for rhc. 21.~1Ccnrury. Carnegie Forum on Education and the Economy, 1001 Connecticut Avenue.N.W., Washington, D.C. 20005, 1986.’’ The Holmes Group. Tornorroic.’s Ti.crchc,r.v: A Rcporr of tl7c Ho1vir.v Group. The Holmes Group, Inc., 501 Erickson Hall. East Lansing. Michigan 48824-1034. 1986. Gardner CHEMISTRY IN THE COMMUNITY CHEMCOM Water Nuclear Chemistry Chemical Resources Air and Climate Petroleum Health Foods The Chemical Industry A Professional Society Effort Figure 2 example, but are assigned to teach chemistry with as little preparation as one chemistry course at university level. They are teaching from a weak background and it is difficult for them to be either as knowledgeable or as enthusiastic as necessary. A series of programmes, largely funded directly by the National Science Foundation or by the Department of Education with funds channelled through the States have been designed. Their purpose is to strengthen the backgrounds of teachers ranging from those with very inadequate preparation to the most excellent and in grade level from kindergarten to twelfth grade.As examples, I will cite four new programmes that are making a difference for chemistry teachers and that have the potential to have much more impact in the next three to five years. A. Chemistry in the Community, CHEMC0M.-The first is Chemistry in the Community, CHEMCOM (Figure 2), a new course for senior high school students developed under the sponsorship of the American Chemical Society and with funding from the National Science Foundation. This course is composed of eight modules that address socially relevant problems that are chemically based.They include food, water, health, nuclear energy, air and climate, petroleum, chemical resources, and the chemical industry. An extensive teacher’s guide and two week intensive teacher preparation workshops are the resources being utilized to implement this programme in the classroom. The CHEMCOM course requires extensive laboratory activity; it is socially relevant and it promotes decision making, problem solving, and other higher order thinking skills. Materials to help the teacher implement the programme include not only the teacher’s guide, but videotapes and special publications on safety and waste disposal, for example, are a part of the teacher package.This course is being introduced into American schools this September and intensive teacher preparation inservice courses are scheduled across the nation for the next three years. CHEMCOM represents an important new direction-a professional society, ACS, taking responsibility for curriculum development and teacher education. In another innovative move, the ACS is also New Trends in Chemical Education INSTITUTE FOR CHEMICAL EDUCATION ICE Consortiumof universities -K-12 inservice education in chemistry From underprepared to mostprepared teachers Outreach and leadership A multi-university Consortium Figure 3 developing an accredited Chemical Education degree programme to be offered in university chemistry departments.B. The Institute for Chemical Education, ICE.-The second programme is the Institute for Chemical Education, ICE (Figure 3). This relatively new programme comprises a consortium of chemistry departments in major universities and provides an array of chemical education activities designed to reach and serve teachers from kindergarten through twelfth grade level and from the least prepared to the most prepared teachers. The ICE programmes include: (1) updating/funda-mentals courses for the underprepared ‘crossover7 chemistry teacher; (2) supple-ments courses to prepare teachers to operate chemistry camps for children and to capitalize on demonstrations; (3) instrumentation courses where Opportunities in Chemistry: Today and Tomorrow, serves as the framework to help the most experienced teachers learn about the theory, research, and practical uses of instruments such as nuclear magnetic resonance, lasers, gas chromatography, research computers, and new electrochemical instrumentation; and (4)special leadership conferences and workshops that allow groups of excellent chemistry teachers to attack specific problems in chemical education.I will use my own institution to illustrate how the ICE programme works. Through the Lawrence Hall of Science and the Department of Chemistry at Berkeley, forty teachers per summer from across the nation participate in the ICE fundamentals/upgrading course, another twenty in the ICE instrumentation course, and twenty more in the ICE laboratory leadership course.The laboratory leadership group, composed of award-winning teachers, is engaged in a three-year effort directed at finding effective new ways to assess the learning that occurs in the laboratory, identifying the factors that inhibit laboratory activity, and promoting further use of laboratories by incorporating micro-experiments and computer interfacing activities into the curriculum. In all three programmes at Berkeley, teachers are prepared with ideas and materials to take home and share in outreach programmes for large numbers of colleagues in their region. The headquarters Gardner CHEMSOURCE SourceBook SourcePlan Sourceview A High Tech Approach to Teacher Education Figure 4 of the ICE network is at the University of Wisconsin at Madison, and other institutions engaged in ICE activity are the University of California, the University of Arizona, the University of Maryland, and the University of Northern Colorado.The networking, outreach, and continuous innovation in teacher education that are possible when such a powerful group joins forces and shares experience and resources, are important characteristics of this new type of teacher education. C. CHEMS0URCE.-The third example is now on the drawing board, but it has powerful potential for the future. It is a programme called CHEMSOURCE that consists of three components (Figure 4). The first is SourceBook, a very detailed data base that contains essentially all of the help and information a teacher needs, including content and methods specific to particular chemistry topics (e.g.acids and bases, equilibrium, stoicheiometry) in either hard copy or disk format. SourceBook is designed to be updated at least biennially, to access other data bases, and to be available and used in every teacher's preparation room. The second component is ChemPlan, a computerized lesson planning toolkit that can draw on SourceBook and the many other new chemical education databases such as the Reactivity Network, the ICE programme, and the CHEMCOM teacher materials. This tool can be used for daily, weekly, and long-term lesson planning in preservice methods courses as well as by individual teachers in their own prep rooms. The third component of CHEMSOURCE is Sourceview, a series of videotapes that will capture exemplars of effective handling of difficulties and opportunities in the classroom.Teams of experienced high school and college chemistry teachers from across the nation are organized into cluster groups to produce this new teacher education resource, CHEMSOURCE. The initial planning has been done; if funded by the National Science Foundation, the work will begin in August of this year. The timetable calls for development, testing, and final production within a three-year period. The power of this innovation lies in giving teachers ready access to a world of assistance through computerization, video, and laser technology. D. Industrial Initiatives in Science and Mathematics Education, 1ISME.-The fourth is a very active partnership of industry and education in support of teachers New’ Trends in Chemical Education INDUSTRY INITIATIVES FOR SCIENCE AND =IISME MATH EDUCATION Figure 5 (Figure 5).Industrial partnerships have existed for decades but many of them address superficial problems and provide the frosting on the cake but are not sufficiently embedded to have significant impact on classroom teaching and on students’ attitudes and learning. IISME is a relatively new programme that has been developed in the San Francisco Bay Area under the leadership of about forty major corporations (e.g. IBM, Chevron, Hewlett Packard, AT&T, DuPont, Dow Chemical, etc.),the school systems of a nine-county region, and the Lawrence Hall of Science, where staff serve as the researchers and educational facilitators of the programme.The programme is already being replicated in other areas of the United States and in Denmark. Such exciting and substantial results from the programme are apparent that major time, effort, and resources are being invested in IISME. INDUSTRY INITIATIVES IN MATHEMATICS AND SCIENCE EDUCATION, IISME 8 weeks in industry during summer Academic year professional growth through the academy Curriculum development Telecommunication network * Research evaluation An Industry/Education Partnership Figure 6 Some of the results from industry’s perspective are: (1) teacher performance and productivity exceeded industry’s expectations; (2) a functional partnership was established with the schools; and (3) greater respect for and understanding of teachers and teaching.Equally important are the results from education’s per- spective: (1) teachers’ knowledge, especially of real world science and technology, was expanded and updated; (2) recognition of the value of group work, inter- personal communication, and problem solving/critical thinking skills; and (3) more effective teaching and career counselling through enhanced teacher credibility (Figures 6-12). In IISME, the industries are the initiators, the Lawrence Hall staff are the facilitators, and the teachers are the implementers as they carry new knowledge, skills, and teaching units back to the classroom. The industries and the National Gardner IISME THE IISME PARTNERSHIP IISME LINKS SCHOOLS WITH INDUSTRY AND GOVERNMENT RESEARCH LABORATORIES VIA SUMMER INDUSTRY ASSIGNMENTS FOR TEACHERS.THE LAWRENCE HALL OF SCIENCE PROVIDES ONGOING SUPPORT FOR TEACHERS TO INTEGRATE THIS INDUSTRY EXPERIENCE INTO CLASSROOM ACTIVITIES. Figure 7 PROGRAM MECHANISMS IISME rn PROVIDE SUMMER INDUSTRY ASSIGNMENTS FOR TEACHERS -UPDATE TEACHERS' TECHNICAL AND COMMUNICATION SKILLS -DEVELOP APPRECIATION OF TECHNICAL CAREERS AND FAMILIARITY WITH IN DUST RY STANDARDS AN D PROCEDURES -AUGMENT SALARIES -START TO "RECOGNIZE" THE TEACHING PROFESSION rn EXPAND PARTNERSHIP BETWEEN INDUSTRY AND EDUCATION -ENCOURAGE INDUSTRY TOURS, GUEST SPEAKERS FROM INDUSTRY TO CLASSROOMS, EQUIPMENT LOAN /DONATION, ONGOING DIALOGUE rn HELP TEACHERS TRANSFER SUMMER EXPERIENCE TO THEIR STUDENTS AND C0LLEA CU ES Figure 8 New Trends in Chemical Education IIiESTATEMENT OF NEED SHORTAGE OF SCIENTIFIC AND TECHNICAL PERSONNEL THREATENS U.S.0 DEMAND OUTPACES SUPPLY 0 STUDENT INTEREST AND ACHIEVEMENT IN MATH AND SCIENCE ARE DECLINING PRECOLLEGE MATH AND SCIENCE PROGRAMS ARE INADEQUATE OUT-OF-DATE CURRICULA AND EQUIPMENT 6000 MATH AND SCIENCE TEACHERS LEAVE TEACHING ANNUALLY FOR OTHER EMPLOYMENT OVER HALF THE BAY AREA MATH AND SCIENCE TEACHERS ARE NOT CREDENTIALED IN THOSE SUBJECTS FACULTY IS UNDERPAID TWO OUT OF THREE CALIFORNIA HIGH SCHOOL TEACHERS WORK AT A SECOND JOB TO AUGMENT EARNINGS Figure 9 IImE INDUSTRY'S PERSPECTIVES 0 EXCELLENT MENTOR/MANAGER/EXECUTlVE SUPPORT FOR PROGRAM -MEANINGFUL TEACHER ASSIGNMENTS WERE LOCATED -PERFORMANCE OF TEACHERS EXCEEDED INDUSTRY'S EXPECTATIONS -MINIMAL TIME AND COST TO ADMINISTER PROGRAM 0 KEYS TO SUCCESS -APPOINT EFFECTIVE COORDINATOR TO WORK WITH IISME -IDENTIFY MENTORS WITH TECHNICAL BACKGROUND AND INTEREST IN EDUCATION -IDENTIFY JOB DESCRIPTION BEFORE TEACHER IS SELECTED AND PLACED -ALLOW TEACHER ACCESS TO COMPANY RESOURCES (PERSONNEL, LIBRARY, TRAINING, ETC.) Figure 10 Science Foundation are now supporting spin-off academic-year professional growth programmes, telecommunication networks, curriculum development, and research.The importance of IISME is that it is developing an industry/education 144 Gardner IIWE EDUCAT I 0N' S PERSPEC T IV ES * TEACHERS UNDERSTANDING OF INDUSTRY IMPROVED -INCREASED KNOWLEDGE OF MODERN APPLICATIONS OF SCIENCE AND MATH IN TODAY'S INDUSTRIES -EMPHASIZED NEED FOR INDEPENDENT PROBLEM-SOLVING AND CRITICAL-THINKING SKILLS -INCREASED FAMILIARIWWITH CAREER REQUIREMENTS IN INDUSTRY * TEACHERS RETURNED TO THE CLASSROOM "REVITALIZED" TEACHERS DEVELOPED CURRICULUM, LABS.AND INSTRUCTIONAL MATERIALS TO TRANSFER SUMMER EXPERIENCE TO THE CLASSROOM -95% OF TEACHER PARTICIPANTS FEEL THEIR INSTRUCTION HAS IMPROVED AS A RESULT OF IlSME -NONE HAVE LEFT THE CLASSROOM FOR POSITIONS IN INDUSTRY; SOME FROM INDUSTRY ARE ENTERING THE CLASSROOM. Figure 11 WHAT CHANGED IN THE CLASSROOM IIiEFOR 1987 IlSME TEACHERS? 89%-ADDED OR REVISED LECTURES (CONTENT) * ADDED MORE CONCRETE EXAMPLES FROM INDUSTRY 85% TO THE CURRICULUM * EMPHASIZED WORK-RELATED SKILLS 82% EMPHASIZED ORAL AND WRITTEN COMMUNICATION SKILLS 69% 66% -INCLUDED MORE GROUP WORK ADDED OR REVISED LABORATORY ACTIVITIES 65% * INCREASED CAREER COUNSELING 62% * INCLUDED MORE PROBLEM-SOLVING ACTIVITIES 33% * INCLUDED GUEST SPEAKERS FROM INDUSTRY 31% * EMPHASIZED INDEPENDENT LEARNING 28% ADDED EQUIPMENT DONATED BY INDUSTRY 23% Figure 12 partnership that works; commitments are being forged that are deep and long- lasting.The four programmes just described-CHEMCOM, ICE, CHEMSOURCE, New Trends in Chemical Education and IISME-represent our efforts to address the trends initially noted.These included learning to use technological tools, dealing with social issues, and making chemical education available to everyone. A new day is dawning in chemical education. The world is reawakening to the value of the teacher, the instructional programmes in the schools, quality in all we do, and the need for change. We must pursue new ways to improve the preparation of our teachers and continuously upgrade their knowledge of chemistry and of research results related to teaching. Recognition of excellent teachers for their achievements and exchanges between and among teachers within each country and across national boundaries are important directions we must pursue. I would be most happy to discuss any of these ideas in detail with individuals or with groups, in person or by correspondence.At the same time, I plan to learn much more about the preservice and inservice education of chemistry teachers in this nation. I thank you for inviting me and for the honour of giving the Nyholm Lecture. Related Literature The National Commission on Excellence in Education. A Nution at Risk: The Imperatiue,fbr Educational Reform. U.S. Government Printing Office, Stock No. 065-000-001 77-2, Washington, D.C. 20402, 1983. National Science Foundation. Educating Americans for the 2lst Century. NSF, Washington, D.C. 20550, 1983. American Chemical Society. Tomorrow: The Report ofthe Task Force,fbr the Stuciy of Chemistry Education in the US. ACS, Washington, D.C. 20036. 1984. John Goodlad, A Pluce Called School: Prospects for !he Future. McGraw Hill Publishing Co., New York, 1984. Commission on Professionals in Science and Technology. Manpower Comments. Volume 24, May 1987. National Academy of Sciences. Opportunities in Chemistry: Toclaj~ und Tomorrow. National Academy Press. Washington, D.C., 1985.

ISSN:0306-0012    

DOI:10.1039/CS9881700135

出版商:RSC    

年代:1988

数据来源: RSC

摘要:

Chem. SOC.Rev., 1988, 17, 147-179 A Conformational Analysis of Transition Metal '1'-AcylComplexes: Steric Interactions and Stereoelectronic Effects By Brent K. Blackburn, Stephen G. Davies,* Kevin H. Sutton, and Mark Whittaker THE DYSON PERRINS LABORATORY, UNIVERSITY OF OXFORD, SOUTH PARKS ROAD, OXFORD OX1 3QY 1 Introduction The introduction of conformational analysis revolutionized organic chemistry such that it is now routinely used to explain and, more importantly, to predict stereo- selective reactions.' Organotransition metal complexes form a major class of compounds of interest both intrinsically and as reagents for organic synthesis.2 Highly stereoselective reactions of these complexes are being discovered, yet surprisingly little consideration has been given to the influence of conformation on reactivity and selectivity. The very large number of highly stereoselective reactions of acyl ligands attached to the iron chiral auxiliary [(q5-C,H,)Fe(CO)(PPh3)] can all be rationalized in terms of a very simple model based on a conformational analysis approach.&" This conformational analysis, derived from models, calculations, and reactivity patterns, indicates that the preferred conformations of such complexes are governed largely by steric factors.This review first generalizes this concept to all transition metal acyl complexes, and then extends it to include stereoelectronic factors. The analysis is correlated with the extensive number of crystal structures available. Finally ligands other than acyl are considered.2 Influence of Steric Interactions on the Preferred Conformation of Metal ql-Acyls A conformational analysis for a variety of metal acyl complexes of different structural types is outlined below. This analysis is based on a qualitative study of molecular models and on computer modelling performed on a series of idealized molecular structures. Consideration is given primarily to steric factors, such that ' (a) E. L. Eliel, N. L. Allinger, J. J. Angyal, and G. A. Morrison, 'Conformational Analysis', Wiley Interscience, New York, 1965: (h) P. A. Bartlett, Tetrahedron, 1980, 36,3; (c) J. I. Seeman, Chem. Rec., 1983, 83. 83. ((I) S. G. Davies, 'Organotransition Metal Chemistry: Applications to Organic Synthesis', Pergamon Press, Oxford.1982; (b)J. P. Collman, L. S. Hegedus, J. R. Norton, and R. G. Finke, 'Principles and Applications of Organotransition Metal Chemistry', University Science Books, Mill Valley, 1987. S. G. Davies, 1. M. Dordor-Hedgecock, R. J. C. Easton, S. C. Preston, K. H. Sutton, and J. C. Walker, BUN. Chitn. Soc. Fr., 1987, 608 and references cited therein. S. G. Davies and J. I. Seeman, TetrahecfronLcvt., 1984, 25, 1845. S. G. Davies and J. I. Seeman, J. Am. Chem. Soc., 1985, 107, 6522. S. G. Davies. J. I. Seeman, and I. H. Williams, Tetrahedron Lett., 1986, 27, 619.' S. L. Brown, S. G. Davies, D. F. Foster, J. I. Seeman, and P. Warner, Tetrahedron Lett., 1986, 27, 623. S. G. Davies, I. M. Dordor-Hedgecock. K. H. Sutton, and M. Whittaker.J. Am. Chem. Soc., 1987, 109, 571 I.' B. K. Blackburn, S. G. Davies, and M. Whittaker, J. Chem. Soc., Chern. Cornmun.. 1987, 1344. '" B. K. Blackburn. S. G. Davies, and M. Whittaker. in preparation. A Conformational Analysis of' Transition Metal q -Acyl Complexes M 0" H\YPH Figure 1 illustration of idealized model used for computer si~nulalions the preferred conformation of transition metal acyl complexes is initially predicted simply by taking into account the steric interactions between the acyl and the proximate ligands. Stereoelectronic contributions to conformational preferences are also considered. This type of analysis can be most easily achieved by use of the appropriate Newman projection so as to view the molecule down the acyl carbon- metal bond.Complexes are dealt with by structural type namely square planar, octahedral, square pyramidal, trigonal bipyramidal, and pseudo-octahedral cyclo- pentadienyl complexes. In each case consideration is given first to complexes comprising a full complement of small carbon monoxide ligands followed by those obtained by sequential substitution of carbon monoxide ligands proximate to the acyl ligand by the large ligand triphenylphosphine. Computer simulations (Chem-X) '' of the conformational energy profiles were obtained using simplified models of the various parent acetyl complexes. The model complexes were all assigned idealized bond angles for the relevant structural type. l2 In order to make the calculations feasible the acetyl group was assigned an axial position with only cis ligands being taken into consideration. Triphenylphosphine was, unless otherwise stated, modelled by PH,Ph, itself held in a fixed conform- ation such that the torsional angle C,(carbon of the acetyl carbony1)-M(meta1)- " Chem-X, developed and designed by Chemical Design.Ltd., Oxford, England. I2 The calculations were performed using the Chern-X molecular modelling program using the default parameters which only take into account van der Waals interactions. The calculational method is an empirical force-field calculation that determines all non-bonded interactions between nuclei. The bond angles, lengths, and torsional angles for the structures on which the calculations were performed are as follows: M-P = 2.40 A, M-CO = 1.75 A, M-COMe = 2.15 A.M-P--C,,,, = 109". C,-M-P-CipSo = O", M-P-C,,,,-C,,,,, = 90". Blackburn, Dauies, Sutton, and Whittaker 0 C -90 0 90 8 Cl ~omplex(21 Figure 2 Calculated energ-v projle .for a square planar complex with two small proximate ligands [M(C0)3COMe] (1); X-rqj. slructure of anionic complex (2).13 All the protons have been remot.ed,for clurity P-Cjps0 was zero with the plane of the phenyl ring orthogonal to the plane defined by these atoms (Figure l).5 For each idealized complex the acetyl group was permitted to rotate about the C,-metal bond with the relative energies being calculated every five degrees. In each case a plot of relative energies [kcal mol-' us the torsional angle 0-C,-M-L (L = the largest ligand)] 0 is given.Note that when the torsional angle 0 is zero the acetyl oxygen is eclipsing the largest ligand and positive angles represent a counter- clockwise rotation. These calculations are not intended to predict the absolute energy of the barrier to rotation, but to provide insight into the inherent steric forces present in these complexes. In each case the computer simulation corresponds closely with the qualitative analysis obtained from inspection of molecular models. The calculated minimum energy conformation is represented alongside each energy diagram in the form of a Newman projection along C,-M. Whenever available a crystal structure of a representative acyl complex or, on occasion, a formyl complex is shown.A. Square Planar Complexes.-(i) Two Small Proximate Ligands. The calculated energy diagram for the model square planar complex [M(C0)3COMe] (1) is shown in Figure 2. There is a bias against those conformations which would eclipse the acetyl ligand with the two proximate carbon monoxide ligands; the calculated A Conformational Analysis of’ Transition Metal q -Aq>lComplexes 0 C -90 0 90 Figure 3 Calculated energy projile for a square planar complex M‘ith one large and one small proximate ligand [M(CO),(PH,Ph)COMe] (3); X-ray structure of complex (4).14‘ All the protons have been removed,for clarity energy minimum for complex (1) places the plane defined by the acetyl group orthogonal to the line described by M(CO),.A crystal structure for the square planar anionic complex (2) shows this complex adopts the calculated minimum energy conformation in the solid state. (ii) One Large and One Small Proximate Ligand. The calculated energy diagram for the model complex cis-[M(CO),(PH,Ph)COMe] (3) is shown in Figure 3. Again the minimum energy conformation has the acetyl plane orthogonal to the line P-M-(CO), that is approximately parallel to the plane of the phenyl ring. It is of note that in order to interconvert the two degenerate minimum energy conform- ations the path involving eclipsing the acetyl methyl with the small carbon monoxide ligand is greatly favoured over the alternative of eclipsing the methyl with the large phosphine ligand.Several crystal structures for complexes of this type are known, all of which adopt the predicted conformation in the solid state; l4 that for complex (4)is illustrated in Figure 3.140 For complex (4)the acyl ligand lies close to the expected conformation with the acyl oxygen slightly tilted towards the large ligand. l3 D. B. Dell’Arnico, F. Calderazzo, and G. Pelizzi, Inorg. Chem., 1979, 18, 1165. l4 (a)R. J. Klingler, J. C. Huffrnan, and J. K. Kochi, J. Am. Ciiem.SOC.,1982,104,2147; (h)G. K. Anderson, R. J. Cross, K. W. Muir, T. Solornaun, and L. Manojlovic-Muir. J. Organomet. Chem., 1979, 170, 385; (c) A. Sen, J.-T. Chen, W. M. Vetter, and R. R. Whittle, J. Am. Chem. Soc., 1987, 109, 148. Blackburn, Dauies, Sutton, and Whittaker Complex(6) Figure 4 Calculated energy projle for a square planar complex uith two large proximate ligands [M(CO)(PH,Ph),COMe] (5); X-ray structure of complex (6).15‘All theprotons have been removed for clarity (iii) Two Large Proximate Ligands.The calculated energy diagram for the model complex trans-[M(CO)(PH,Ph),COMe] (5) is shown in Figure 4. Two degenerate minima are again observed with the acetyl plane orthogonal to the P-M-P line. A relatively high energy barrier exists to the interconversion of these degenerate conformations which must involve eclipsing of the methyl group with one of the large phosphines. Crystal structures for many complexes of this type are available and the predicted conformation is adopted in each case.” The crystal structure for complex (6)is shown in Figure 4.’ 5a Note that the two triphenylphosphine ligands proximate to the acyl ligand orientate their rotors such that the dihedral angles C,-Pt-P-Cip,, are close to zero.Furthermore both the proximate phenyl rings are close to parallel to the acyl plane, being somewhat tilted in each case to widen the gap occupied by the large alkyl chain and consequently tighten the gap occupied by the relatively small acyl oxygen. Nonetheless this structure serves to demonstrate the validity of the model chosen for the computer simulations. l5 (a) R. Bardi, A. M. Piazzesi. G. Cavinato, P. Cavoli, and L. Toniolo, J. Organornet. Chern., 1982, 224, 407; (b)L. N. Zhir-Lebed,L. G. Kuzmina. Y. T. Struchkov, 0.N. Temkin, and V.A. Golodov, Koord. Khim., 1978,4, 1046; (c)R. Bardi, A. M. Piazzesi, A. Del Pra, G.Cavinato, and L. Toniolo,J. Organornet. Chem., 1982,234, 107; (d)W. Fengsham, F. Yuguo, and C. Qiuzi, Chem. J. Chin. Unit..,1984,5,699; (e) M. A. Bennett, K.-C. Ho, J. C.Jeffrey, G. M. McLaughlin, and G. B. Robertson, Aust. J. Chem., 1982,35, 1311; (f) P. Stoppioni, P. Dapporto, and L. Sacconi, Inorg. Chem., 1978, 17, 718; (g) E. Carmona, F. Gonzalez, M. L. Poveda, J. L. Atwood, and R. D. Rodgers, J. Chem. SOC.,Dalton Trans., 1980,2108; (h) P. L. Bellon, M. Manassero, F. Porta, and M. Sansoni, J. Organomer. Chem., 1974, 80, 139; (i) G.Del Piero and M. Cesari. Acta Crystallogr. B. 1979, 35, 241 1. A Conformational Analysis of' Transition Metal q '-Acyl Complexes E 15 10 5 0 Cornplex(8) Figure 5 Calculuted energy projle ,for an octahedral comp1e.x with four small proximate ligands [M(C0)5COMe] (7); X-ruy structure of complex (8).16a All the protons have been removed, for clarity B.Octahedral Complexes.-(i) Four Small Proximate Ligands. The calculated energy diagram for the model complex [M(C0)5COMe] (7), Figure 5, suggests four degenerate minimum energy conformations with the acetyl group staggered with respect to the proximate carbon monoxide ligands. Complex (8) adopts such a staggered conformation in the solid state.16o Other complexes of this type also adopt the predicted conformation in the solid state.' (ii) One Large and Three Small Proximate Ligands. The calculated energy diagram for the model complex cis-[M(CO),(PH,Ph)COMe] (9), Figure 6, indicates two stable conformations where the acetyl ligand is staggered with respect to the four proximate ligands.The small oxygen atom is shown to reside between the carbon monoxide and phosphine ligands with the larger methyl group between two small carbon monoxide ligands. Each of the acetyl groups in the bis-acetyl anion (10) 7a adopts the predicted minimum energy conformation in the solid state. This pre- dicted conformation is also observed in other reported crystal structure deter- minations for this type of complex.' l6 (a)C. A. Casey, C. A. Bunnell, and J. C. Calabrese, J. Am. Chem. Soc., 1976,98, 1166; (6) I. S. Astakhova, V. A. Semion. and Y. T. Struchkov, Zh.Strukt. Khim., 1969, 10, 508; (c) E. J. M. De Boer, J. De With, N. Meijboom, and A. G. Orpen, Organometallics, 1985, 4, 259. (a)P. G. Lenhert, C. M. Lukehart. P. D. Sotiropoulos, and K. Srinivasan, Inorg. Chem., 1984.23, 1807; (h)C. P. Casey and C. A. Bunnell, J. Am. Chem.Soc., 1976,98,436;(c) A. Mayanza, J.-J. Bonnet, J. Galy, P. Kalck, and R. Poilblanc, J. Chem. Rex (S),1980, 146, 2101; (d)C.-H. Cheng, B. D. Spivack, and R. Eisenberg, J. Am. Chem. Soc.. 1977, 99, 3003. Blackburn, Davies, Sutton, and Whittaker Complex (10) Figure 6 Calculated energy projile,for an octahedral complex with one large and three small proximate ligands [M(CO),(PH,Ph)COMe] (9); X-ray structure of anionic complex (lo).'All the protons haoe been removed for clarity (iii) Tkvo Large crnd Two Small Proximate Ligands.The calculated energy profiles for the model complexes fac-[M(CO),(PH,Ph),COMe] (1 1) and mer-[M(CO),(PH,Ph),COMe] (13) are shown in Figures 7 and 8 respectively. When the two large phosphine ligands are cis a single minimum energy conformation is predicted with the acetyl oxygen lying staggered between the two phosphines. The crystal structure for the complex (12) illustrates this feature,"" which is common to all other reported structures for this type of complex." When the two phosphine ligands are trans, however, two degenerate conform- ations are expected where the acetyl group eclipses the small carbon monoxide ligands. The one known example, for which the X-ray crystal structure data are available, of this type of complex, ([(PhO),P]3(CO),MnCHO) (14), adopts the expected conformation in the solid state.' (iv) Three Large and One Small Proximate Ligands.In the calculated minimum energy conformation for the model complex mer-[M(CO),(PH,Ph),COMe] (15), Figure 9, the small acetyl oxygen is close to eclipsing the central phosphine and the methyl close to eclipsing the small carbon monoxide ligand. Once again the dominant interactions are those between the methyl hydrogens and the phosphine '' ((I) J. R. Anglin, H. P. Calhoun, and W. A. G. Graham, Inorg. Chem., 1977, 16, 2281; (h) M. F. McGuiggan. D. H. Doughty, and L. H. Pignolet, J. Organomet. Chem., 1980, 185, 241. (a)H. Berke. G. Huttner, 0.Scheidster, and G.Weiler, Angew. Chem.,Inl. Ed. Engl., 1984,23,735; (h)H. Berke. G. Weiler, G. Huttner, and 0.Orama. Chem. Ber., 1987, 120, 297. A Conformational Analysis of Transition Metal 11 '-Acyl Complexes 0 C co -90 0 Complex (12) Figure 7 Calculated energy projile,for an octahedral complex j19ith MY) large cis and tM'o small proximate ligands fac-[M(CO),(PH,Ph),COMe] (1 1); X-ray structure of complex (12).180All the protons have been removed,for clarity phenyls. The acyl complex (16) adopts the predicted conformation in the solid state, as do other complexes of this type.20 (v) Four Large Proximate Ligands. This case parallels that for the four small proximate ligands except that the energy barriers for the interconversion of the four degenerate staggered conformations are very much larger. Figure 10 shows the calculated energy profile for the model complex trans-[M(CO)(PH,Ph),COMe] (17).All known structures of this class of compounds 21 adopt the predicted con- formation, which is illustrated by the cationic formyl complex (18).21a C. Square Pyramidal Complexes.-The conformational analysis for square pyramidal complexes where the acyl ligand is axial exactly parallels that for the corresponding octahedral complexes, as illustrated by the Newman projection (19). However because of the increased bond angles between C, and any of the other ligands the energy barriers can be expected to be lower for the square pyramidal 2o S. A. Chawdhury, Z. Dauter, J. Mawby. C. D. Reynolds, D.R. Saunders, and M. Stephenson. Ac,tn Crystullogr.. C (Crj~t.3ruc.r. Conimun.),1983.39.985; (h)D. Milstein, W. C. Fultz, and J. C. Calabrese, J. Am. Clieni. Soc.. 1986, 108, 1336; (c) G. W. Adamson, J. J. Daly. and D. Forster, J. Orpnomet. Cliem.. 1974. 71, C17. l1 ((I) G. Smith, D. J. Cole-Hamilton. M. Thorton-Pett, and M. B. Hursthouse, Polj'liedron. 1983. 2, 1241: (h)R. L. Harlow, J. B. Kinney, and T. Herskovitz J. Client.Soc., Chem. Cornmun., 1980,8 13; (c) G. Smith, D. J. Cole-Hamilton, M. Thorton-Pett. and M. B. Hursthouse. J. Chern.Soc., Dulton Trans.. 1983, 2501. Blackburn, Davies, Sutton, and Whittaker 8 1 L CN(19) Complex( 20) structures than for the corresponding octahedral complexes. Known examples of this type of complex adopt the structure in the solid state which is predicted for the corresponding octahedral comple~es.~~-~~ Complexes (20),22"with two large cis ligands, and the cation (21),23with three large ligands, illustrate the applicability of the analysis.For square pyramidal complexes where the acyl is basal, Newman projection (22), the conformational analysis will parallel that for the trigonal bipyramid structures outlined below. However to date there are no representative crystal structure determinations available to test this hypothesis. D. Trigonal Bipyramida1.-If the acyl ligand is equatorial the analysis parallels that described for the square planar complexes above. The analysis for axial acyl complexes is outlined below."((1) C.-H. Cheng, and R. Eisenberg, Inorg. Clieni.. 1979, 18, 1418; (b)C.-H. Cheng, D. E. Hendriksen, and R. Eisenberg, J. Organomet. Clieni., 1977. 142, C65. 23 M. A. Bennett, J. C.Jeffrey, and G. B. Robertson, Inorg. Ckem., 1981, 20, 323. 24 D. L. Egglestone, M. C. Baird, C. J. L. Lock, and G. Turner, J. Chem. Soc., Dalton Trans., 1977, 1576. The X-ray crystal structure reported in this article has come under close scrutiny and is deemed un- reliable. '' 25 B. B. Wayland, B. A. Woods. and R. J. Pierce, J. Am. Chern. Soc., 1982, 104. 302. 155 A Conformational Analysis of’ Transition Metal q ‘-Acyl Complexes 1 J -90 0 90 e Complex(l4) Figure 8 Culcufuted energ)’ prqfik .fhan octahedrul cwnp1e.v 1r.ith t,ro large trans and 1~10 small roximate ligands mer-[M(CO),(PH,Ph),COMe] (13); X-ray structure of complex(14)’ B All--the protons and the P(OPh), Iigandtransto the formjd ligund have been removed for clarit)* E 0 15 10 rn5 (15) 0 -90 0 90 0 0 PMe2Ph / Ph ~B~NC-RU <o PhMe2P/ I Ph Complex(l6) Figure 9 Calculated energy projilt>for un octahedral compler Mith three large and one small prosimute Iigands mer-[M(CO)2( PH,Ph),COMe] ( 15); X-ruj* structure of’ comp1e.r (1 6).20”All the protons Iime been removed ,fiw c.1urir.v Blackburn, Davies, Sutton, and Whittaker I I -90 0 90 e Complex(l8) Figure 10 Calculated energy projle for an octahedral complex with four large proximate ligands trans-[M(CO)( PH,Ph),COMe] (17); X-ray structure of cationic complex (18).21" All the protons have been removed, for clarity MeocTco0 .-90 o 90 e 8 co\i (Ph0)3P-Fe C6 0 ComplexQ4) Figure 11 Calculated energy proJile for a trigonal bipyramidal complex with three small pro-uimate ligands [M(CO),COMe] (23);X-ray structure of anionic complex (24).26aAN the protons have been removed,for clarity 157 A Conformational Analysis of Transition Metal q ‘-Acyl Complexes -90 0 90 6 Figure 12 Calculated energqprojle for a trigonal bipyramidal cnmples ir9ith one large and two small pro.uimate ligands cis-[M( CO),( PH, Ph)COMe] (25) (i) Three Small Proximate Ligands. The calculated energy profile for the model complex [M(C0)4COMe] (23) is shown in Figure 11.The minimum energy con- formation staggers the large methyl group between two carbon monoxide ligands, and consequently eclipses the acetyl oxygen with a carbon monoxide ligand. There are therefore three degenerate minimum energy conformations. The structure of anion (24) shows it to adopt a conformation close to the expected one in the solid state.26o It is noteworthy that the triphenylphosphite ligand trans to the acyl staggers itself relative to the three proximate carbon monoxide ligands. Other complexes of this type adopt the predicted conformation in the solid state.26 (ii) One Large and Two Small Pro-uimate Ligands. The calculated conformational energy profile for the model complex cis-[M(CO),(PH,Ph)COMe] (25) is shown in Figure 12. In the minimum energy conformation the acetyl ligand is orthogonal to the metal-phosphorus bond, thus minimizing the steric interactions between the methyl and acetyl oxygen with the large phosphine ligand.There are currently no examples of structures in this class of complex. CA VEAT: The latter analysis assumes that the phosphine orientates itself to eclipse the C,-M and P-Cips0 bonds, as in structure (25). If however these bonds are staggered then the conformational analysis applicable to structure (1 1) (octahedral with two large cis proximate ligands) pertains, i.e. the acyl oxygen will eclipse the phosphorus atom, that is staggered between the two large phenyl rings. This situation exists in complex (26) 27 where the bridge between the phosphines imposes such a conformation on the diphenylphosphine residue proximate to the acyl ligand.Figure 13 shows the calculated conformational energy profile for the related bis-(dipheny1phosphino)ethane-substitutedtrigonal bipyramidal acetyl 26 (a)C. P. Casey. M. W. Meszaros, S. M. Neumann. I. G. Cesa. and K. J. Haller, Orgcrrzorne~allics,1985,4, 143; (h)D. Milstein and J. L. Huckaby. J. Am. Chenr. Soc., 1982, 104, 6150; (6.) K. H. Dotz, U. Wenicker, G. Muller. H. Alt, and D. Seyferth, Organornefullics,1986,5.2570; (d)V.Galamb, G. Palyi, F. Unguary, L. Marko, R. Boese, and G. Schrnid, J. Am. Chem. Soc., 1986, 108, 3344; (e)J. B. Wilford and H. M. Powell, J. Chem. SOC.A, 1967,2092; (f) M. Roper and C. Kruger, J. Orgunomet. Chem., 1988,339, 159.’’ A. R. Cutler, C. C. Tso, and R.K. Kullnig, J. Am. Clrern. Soc., 1987, 109, 5844. Blackburn, Davies, Sutton, and Whittaker ocvco 4%Ph/P\Ph Complex(26) -90 0 90 e Figure 13 X-Ray structure of complex (26)2' (all the protons have been removed for clarity); energy profile calculated from X-ray structure of complex (26) 2a complex.28 The preferred conformation predicted for the acetyl complex is that where the acetyl oxygen nearly eclipses the metal-phosphorus bond, which is similar to that found in the X-ray crystal structure of complex (26).27 ''This calculation was performed using CHEM-X molecular modelling program" on the basis of the reported X-ray crystal structure determination 27 using the default parameters that only take into account van der Waals interactions. A Conformational Analysis of Transition Metal q'-Acyl Complexes -90 0 90 8 Figure 14 Calculated energyprojile.fbr a trigonal hipyramidal complex with two large and one small proximate ligands [M(CO),(PH,Ph),COMe] (27) 50 (281i:0 -90 0 90 0 Figure 15 Calculated energy projile for a trigonal hipyramidal complex Kitti three large proximate ligands trans-[ M(CO)( PH ,Ph),COMe] (28) (iii) Two Large and One Small Ligands.The calculated conformational energy profile for the model complex cis-[M(CO),(PH,Ph),COMe] (27) is shown in Figure 14. The minimum energy conformation staggers the small acetyl oxygen between the two large phosphines and consequently eclipses the methyl with the proximate carbon monoxide ligand.There are no known examples of this type of complex. (iv) Three Large Ligands. The calculated conformational energy profile for the model complex trans-[M(CO)(PH,Ph),COMe] (28) is shown in Figure 15. It parallels that for the three small ligand case (Figure 11) except that the energy barriers for interconversion of the three degenerate minimum energy conform- ations will be much larger. Again, there are no known X-ray crystal structure determinations of a complex of this type. E. Tetrahedral Complexes.-The conformational analysis for the tetrahedral com- plexes parallels that for the corresponding trigonal bipyramid complexes described Blackburn, Davies, Sutton, and Whittaker co -I 0 Complex (30) Figure 16 Calculated energy proJile for a pseudo-octahedral cyclopentadienyl com lex with two small ligands [q 5-C,H,)M(CO),COMe] (29); X-ray structure of complex (30).'" All the protons have been removed for clarity above.However, rotational energy barriers will be lower for tetrahedral acyl complexes than for the analogous trigonal bipyramid complexes, because of the increased bond angles between C, and any of the proximate ligands. F. Pseudo-octahedral Cyclopentadienyl Complexes.-By examination of the avail- able X-ray crystallographic data the cyclopentadienyl complexes [(q'-C,H,)ML,] are best described as pseudo-octahedral, where each of the three ligands L are orthogonal to each other., Thus, the three ligands L occupy adjacent coordination sites of an octahedron with the cyclopentadienyl ligand occupying the three remaining coordination sites.(i) Two Small Ligands. The calculated conformational energy profile for the model complex [(q5-C,H,)M(CO),COMe] (29) is shown in Figure 16. The two degenerate minimum energy conformations place the acetyl orthogonal to the metal-cyclopentadienyl centroid line. Complex (30) adopts this conformation in the solid as do other examples.29 (ii) One Large and One Small Ligand. The calculated conformational energy diagram for the model complex [(q5-C,H,)M(CO)(PH,Ph)COMe] (31) is shown in Figure 17. The minimum energy conformation places the small acetyl oxygen 29 (a)H. Blau, W. Malisch, S. Voran, K. Blank, and C. Kruger, J. Organomet. Chem., 1980,202, C33; (b)A.Eisenstadt, F. Frolow, and A. Efraty, J. Chern. Soc.. Dalton Trans., 1982, 1013; (c) J. B. Sheridan and G. L. Geoffroy, J. Am. Ciiem. Soc., 1987, 109, 1584. A Conformational Analysis of Transition Metal q '-Acyl Complexes Complex(321 Figure 17 Calculated energy projle ,for a pseudo-octahedral cyclopentadienyl comples M'ith one large and one small ligund [(q5-C5H,)M(CO)(PH,Ph)COMe](31); X-raq' structure qfcomplex (32).30aAN the protons haw been remoi>ed,for clarity between the cyclopentadienyl and the largest ligand, the phosphine. The iron acetyl complex (32) 30a conforms to this prediction and further consideration to its conformational properties is given below. Similarly, the predicted conformation is adopted by the other known examples.30 (iii) Two Large Ligands.The calculated conformational energy profile for the model complex [(q 5-C5H5)M(PH,Ph),COMe] (33), which possesses idealized bond angles (8 = 90") between the two large ligands, is shown in Figure 18a. The two degenerate minimum energy conformations place the acyl oxygen between the cyclopentadienyl ligand and one of the phosphine ligands. This conformation is adopted in the solid state by most known e~arnples,~'".~ however, it is not observed in the solid state for complex (34)31c where the acyl oxygen resides between the two 30 ((I) I. Bernal, H. Brunner. and M. Muschiol, Inorg. Chini. A(,tn. 1988, 142.235; (h)G.J. Baird, J. A. Bandy, S. G. Davies, and K. Prout, J. Chem. Soc.. Clieni.Comniun., 1983, 1202; (c) V. A. Semion and Y. T. Struchkov, Zh. S/ruk/.Kliini., 1969. 10.664; (d)G. M. Reisner. 1. Bernal, H. Brunner, and L. Muschliol. Inorg. Chem., 1978,17,783; (e)S. G. Davies. I. M. H. Dordor-Hedgecock, K. H. Sutton. and J. C. Walker, Te/mhedron,1986,42, 5123; (1)S. G. Davies, I. M. H. Dordor-Hedgecock, K. H. Sutton, J. C. Walker, C. Bourne. R. H. Jones, and K. Prout, J. Cliem. Soc,..Cliem. Comrnun., 1986,607: (8)P. Helquist and E. J. O'Connor, J. Am. Chem. Soc.. 1982,104, 1869; (11)G. J. Baird, S. G. Davies, R. H. Jones. K. Prout, and P. Warner, J. Cliem.Soc., Cliem. Commun., 1984.145: (i) G. 0.Nelson and C. E. Sumner. OrgnnoniatnNic~s, 1986, 5. 1983; (j) K. H. Pannell. R. N. Kapoor. M. Wells. and T. Giasolli. OrgLitioiiie/~~Ni(..s,1987, 6, 663.31 (a) H. Felkin, B. Meunier, C. Pascard, and T. Prange, J. Orgonome/. Clietn.. 1977. 135. 361: (h) H. Werner, L. Hofmann, and R. Zolk, Chmi. Bey., 1987, 120, 379; (c) H. G. Alt, M. E. Eichner, B. M. Jansen, and U. Thewalt. 2. Nufurforsdi..Td B, 1982. 37, 1109. Blackburn, Davies, Sutton, and Whittaker L I -90 0 90 e PMe3 Complex (34) Figure 18 Calculated energy projle for a pseudo-octahedral cyclopentadienyl complex with ~M'Olarge ligands [(q'-C,H,)M(PH,Ph),COMe] (33) {0(<P-M-P) = (a)90"[I. . .]; (b)92" [--. --3; (c)95" [I----];(d)100" [-I}; X-ray structure of complex (34).31cAll theprotons have been removed for clarity large phosphines. The angle between the two PMe, ligands in (34), however, is larger (0 = 99.1') than in the model complex (33).Revision of the model by widening the bond angle between the two phosphorus atoms results in new calculated energy diagrams (Figure 17b-d). It was calculated that even a small change of only two degrees results in a new energetically favoured conformation, (39, which is similar to that observed for complex (34) in the solid state. The predicted conformations for the types of pseudo-octahedral cyclopenta- dienyl complexes described above are related by isolobal analogy3* to the predictions made for the octahedral complexes (Sections B i-iii). Figure 19 shows 32 (a)A. R. Pinhas, T. A. Albright, P. Hofmann, and R. Hoffmann, Helv. Chem. Acta, 1980,63, 29; (b)R. Hoffmann, Angew'. Chem., In(.Ed. Engl., 1982, 21, 711.A Conformational Anulysis oj‘ Transition Metul q ’ -Aqd Comp1e.ue.y 0C oc+co C 0 (7) 0 C R co ---0% 19) (31) 0C a (11) (353 Figure 19 Illustration of’isolohulanalogies hetlz,een octahedral und the corresponding pseudo-octahedral conipleses this relationship between the preferred conformation for an acetyl ligand attached to octahedral and the corresponding pseudo-octahedral types of complexes. This relationship results from substituting three coordination sites in the octahedral complexes, each occupied by a small carbon monoxide ligand, by a cyclopenta-dienyl ligand, which is representative of a medium sized ligand, in the pseudo-octahedral complexes. Thus the preferred conformations for the octahedral complexes are directly analogous to those predicted for the corresponding cyclo-pentadienyl complexes, after adjustments to take into account the medium size of the cyclopentadienyl ligand.The influence of steric interactions on acyl con- Blackburn, Davies, Sutton, and Whittaker n Figure 20 Graphical representation of d, -p,. orbital overlap formation by other types of ligands not covered here can, therefore, be accomplished through consideration of their relative steric size and isolobal analogy. 3 Electronic and Electrostatic Contributions to the Preferred Conformation A. Stereoelectronic Effects.-The above analysis correctly correlates prediction of acyl geometry based on steric interactions to the wealth of X-ray crystallographic data available, thus providing a working hypothesis whereby the preferred con- formation of an 11'-acyl complex may be confidently predicted; acyl geometry is predominately dictated by steric interactions.However, this simple rule does not allow for stereoelectronic contributions arising from electron delocalization of a lone pair of electrons on the metal into the 7c* orbital of the acyl ligand. Resonance stabilization of this type, through d,-p,, orbital over la^,^^-^' is maximized when the acyl ligand lies perpendicular to a filled molecular orbital, of the correct symmetry and energy, on the metal fragment (Figure 20). Only when such a stereo- electronic contribution outweighs the steric effects will it take precedence in defining acyl geometry.Three situations may be envisioned for organotransition metal complexes, namely when there are zero, one, or two lone pairs of electrons in high lying molecular orbitals on the metal that are of the appropriate symmetry to participate in d,-p,, orbital overlap. For metal fragments with zero lone pairs no stereo- electronic effect can exist. Metal fragments with two lone pairs, which lie orthogonal to each other, will exert a stereoelectronic influence on the acyl ligand but it will not be directional. For metal fragments with one lone pair of electrons, however, a directional stereoelectronic effect on acyl conformation may exist. Figure 21 illustrates the stereochemical influence of zero (a), one (b), and two (c) pairs of electrons on the metal fragment in relation to the preferred acyl con- formation.For zero lone pairs of electrons there is no significant d,-p,, orbital overlap and thus no stereoelectronic effect. Evidence in support of this is the commonly observed shortening of the M-C, bond of approximately 0.2 8, when going from zero to one (or more) lone pairs of electrons on the metal. For example, the rhenium complex [(CO),ReCOC,H,Cl], having zero lone pairs (vide infra), 33 (a)P. M. Treichel, R. L. Shubkin, K. W. Barnett, and D. Reichard, Inorg. Chem., 1966, 5, 177; (6) M. R. Churchill and J. P. Fennessey, Inorg. Chem., 1968. 7, 953. 34 E. Hadicke and W. Hoppe, Acta Crystallogr., B, 1971, 27, 760. A Conformational Analysis of Transition Metal q -Acyl Complexes has a M-C, bond length of 2.22 whereas the rhenium acyl complex RS,SR-[(q5-C,H5)Re(NO)(PPh,)COCH(Me)CH,Ph],which has one lone pair of electrons (oide infra), is observed to have a M-C, bond length of 2.08 A.35b9cTwo orthogonal, energetically similar lone pairs of electrons (Figure 21c) are able to participate equally in orbital overlap and in consequence do not exert a con- formational preference on the acyl ligand.Therefore, when there is zero, two, or more lone pairs of electrons centred at the metal, steric interactions will dominate conformational preferences. Only in the case where there is a single, energetically distinct lone pair of electrons (or HOMO) on the metal fragment can there be a significant stereoelectronic influence on acyl conformation.Clearly, therefore, only for complexes where a single lone pair of electrons is centred at the metal is it necessary to consider the influence of stereoelectronic preferences. The number of lone pairs of electrons on the metal can be readily determined by the following method, which is in accordance with previously reported theoretical analyses of the frontier orbitals of ML, fragment~.~~-~O (i) Considering the complex of interest determine the number of valence d-electrons of the metal (Group Number) 41 and subtract (for positively charged complexes) or add (for negatively charged complexes) the total charge of the complex. Subtract from this number the number of hydrocarbon ligands with odd hapto (q) numbers and the number of other odd electron ligands (e.g.hydride, halide, nitrosyl, etc.). Finally, divide by two to give the total number of electron pairs potentially available for backbonding. (When the result is a count of four reduce by one, since the fourth lone pair is in an orbital inappropriate for backbonding according to the rules of ~ymmetry.~’) 35 (a)W. K. Wong, W. Tam, C. E. Strouse, and J. A. Gladysz, J. Chem. Soc., Chem. Commun., 1979,530 (b) D. E. Smith and J. A. Gladysz, Organomeiallics. 1985,4, 1480; (c)G. S. Bodner, A. J. Patton, D. E. Smith, S. Georgiou, W. Tam, W.-K. Wong. C. E. Strouse, and J. A. Gladysz, Organometallics, 1987, 6, 1954. 36 T. A. Albright, J. K. Burdett, M.-H. Whangbo, ‘Orbital Interactions in Chemistry’, Wiley Interscience: New York, 1985 and references cited therein.”(a)M. Elian and R. Hoffmann, Inorg. Chem., 1975.14, 1059; (h)R. Hoffmann, Science, 1981,211,995;(c) T. A. Albright, Tetrahedron, 1982, 38, 1339;(c)D. E. Sherwood and M. B. Hall, Inorg. Chem., 1983, 22, 93; (d)D. E. Sherwood and M. B. Hall, Inorg. Chem., 1980,19, 1805; (e)R. Hoffmann and P. Hofmann, J. Am. Ckem. Soc., 1981,103,4320;(f’) P. Kubacek and R. Hoffmann, J. Am. Chem. Soc., 1981,103,4320; (8)T. Ziegler and A. Rauk, Inorg. Chem., 1979, 18, 1755; (h)J. L. Templeton, P. B. Winston, and B. C. Ward, J. Am. Chem. Soc., 1981, 103,7713; (i)R. Hoffmann and H. Berke, J. Am. Chem. Soc., 1978,100, 7224 (j)R. Hoffmann and A. R. Rossi, Inorg. Chem., 1975,14,365;(k)J. K. Burdett, Inorg.Chem., 1975, 14, 375; (I) idem, J. Chem. Soc., Faraday Trans. 2, 1971, 70, 1599; (m) D. M. P. Mingos. J. Chem. Soc., Dalton Trans., 1977, 602; (n) P. Hofmann. Angew. C/zem., 1977, 15, 1. ’13 (a)R. Hoffman, B. E. R. Schilling, and D. L. Lichtenberger, J. Am. Chem. SOL..,1979,101,585;(b)B. E. R. Schilling, R. Hoffmann, and J. W. Faller, J. Am. Chem. Soc., 1979, 101, 592. 39 (a)R. F. Fenske, M. C. Milletti, and M. Arndt, Organometallics, 1986,52316;(h)W. A. Kiel, G.-Y. Lin, A. G. Constable, F. B. McCormick, C. E. Strouse, 0.Eisenstein, and J. A. Gladysz, J. Am. Chem. Soc., 1982, 104, 4865; (c) S. Georgiou and J. A. Gladysz, Teirahedron. 1986, 42, 1109. 40 (a)D. M. P. Mingos, Adu. Organomet. Chem., 1977.15, I; (h)D. M. P. Mingos, ‘Bonding of Unsaturated Organic Molecules to Transition Metals’ in Comprehensiw Organometallic Chemistry; ed.G. Wilkinson, F. G. A. Stone and E. W. Abel, Pergamon Press: New York, 1982, Vol. 3, I; (c) T. A. Albright. R. Hoffmann, J. C. Thibeault, and D. L. Thorn, J. Am. Chem. SOL..,1979,101,3801;(d)0.Eisenstein and R. Hoffmann, J. Am. Chem. Soc., 1981, 103, 4308. 41 The Group Number refers to the new periodic group notation recently adopted by the IUPAC and ACS nomenclature committees. Ligands are classified according to ref. 2a page 2. 42 By definition, only up to three orbitals may have the appropriate symmetry to participate in orbital mixing with a ligand p,. orbital. 166 Bluckhurn, Davies, Sut ton, and Whit taker L L F (a1 (b) (C) Figure 21 Ii?iplicatrons for acyl ligund conformation on metal fragments "ith (a)zero, (h)one, und ((2) t)t'o lone pairs of electrons (ii) Now, only considering the ligands on the ML, fragment, reduce the number of electron pairs (0-3) by one (1) for every pair of trans 7c acidic ligands (e.g.carbon monoxide, linear nitrosyl, isonitrile ligands, etc.) and one for all other strongly 7c acidic ligands.* A calculation of this sort will yield a count of zero to three lone pairs of electrons where a result of one lone pair is of particular interest. In order to predict acyl conformation, it is necessary to determine the orientation of the HOMO containing the single lone pair of electrons, since the acyl ligand will prefer to lie orthogonal to this orbital to achieve maximum orbital overlap. The orientation of the HOMO can be predicted to lie along the line defined by the bond between the poorest d, acceptor ligand and metal In general, the poorest d, acceptor ligand is a phosphine ligand, which commonly is also the sterically most demanding ligand.Therefore, in these types of acyl complexes stereoelectronic and steric interactions are complementary. Figure 22 shows the orientation of the HOMO for the types of metal fragments which possess a single lone pair of electrons. Again, as in Section 2, the metal fragments are dealt with by structural type with a sequential substitution of the proximate ligands, initially carbon monoxide, with a tertiary phosphine. The description of the HOMO on the metal fragments, as outlined above and depicted in Figure 22, is based principally on previously reported molecular orbital calculation^.^^ -40 When considering the preferred conformation of an organotransition metal acyl complex it is important to weigh the effects of both steric and stereoelectronic interactions and determine which predominates. Several examples are shown in Table 1 of known complexes that are categorized by the number of pairs of electrons centred at the metal.The metal fragments that are determined to have zero, two, or more pairs of electrons have the acyl geometry predicted on steric arguments. For the metal fragments calculated to have one lone pair, the stereo- * Generally. even-hydrocarbon, phosphine, and phosphite ligands are not strong n acidic ligands whereas q'-carbene.q2-vinylidene. carbon monoxide. linear nitrosyl, and isonitrile ligands are strongly n acidic ligdnds. A Conformational Analysis qf Transition Metal q -Acyl Complexes 0 0 PIIi) I C c c P Iii) R P R:CO,P I C P Figure 22 Postulated orientation of a single lone-pair of electrons on meta1,fragments of the comp1e.u types: (i) octahedral; (iia) trigonal bipyramid equatorial1.v substituted; (iib) trigonal bipyramid axially substituted; (iii) pseudo-octahedral electronic effect, depending on its magnitude relative to steric effects, is shown to dictate the acyl orientation, reinforce with the steric restrictions present in the complex, or in certain cases be completely outweighed by the steric interactions.The pseudo-octahedral cyclopentadienyl manganese anion [(q ,-C,H,)Mn-(CO),COPh]-(36)34 is determined to have three pairs of electrons available for backbonding, by adding one for a negative charge then subtracting the total number of odd hapto (q) ligands from the seven valence &electrons and finally dividing by two. Reducing by two to allow for backbonding into the carbon monoxide ligands yields one lone pair of electrons (or HOMO) which, upon inspection of Figure 22, lies parallel to the cyclopentadienyl ring. The X-ray crystal A Conformational Analysis of Transition Metal q -Acyl Complexes Table I Selected complexes categorized by the number of pairs of electrons available for d,-p,, orbital overlap Number of 0 1 2 3 lone pairs square planar ci~-[C0Pt(Cl),C0N(Pr'),] ' Cl(PPh,),PtCOPr '54 (2) (6)C 0 octahedral (CO),MnCOCOMe 16' [P(OEt),],Mn(CO),CHO lY M (8) (14)P L trigonal [(MeO),P]Fe(CO),COMe 26n (dppe)Co(CO),COR 27 E bipyramidal (24) (26)X q5-cyclopentadienyl (C,H,)Re(NO)(CO)COR (C,H,)Re(NO)(PPh,)COPh (C,H,)Fe(CO)(PPh,)COMe (C,H,)Fe(PMe,),COEt In pseudo-octahedral (30) (32) (34) (C5H,)Fe(CO),COR [C,H,)Fe(NO)(PPh,)COR] ' A Conformational Analysis of' Transition Metal q '-AcyI Complexes (36) Figure 23 X-Ray crystai structure of the anionic conip1e.v [(q5-C,H5)Mn(CO),COPh](36).34The diagram shohi5.y the Newman proption along the alpha carbon to manganese bond.The protons hare been removed for clarity structure determination for complex (36) shows that the stereoelectronically preferred conformation is adopted in the solid state (Figure 23), where maximum d,-p,, orbital overlap can be achieved.If the preferred conformation of the acyl function in (36) was controlled by steric effects then the benzoyl moiety would tend to lie parallel to the cyclopentadienyl ligand (see Section 2E, i). In this case the stereoelectronic effect is greater than the steric effect. Another example of the use of the guidelines discussed above is the comparison of the chiral iron acetyl [(q5-C5H5)Fe(CO)(PPh3)COMe](32) and its rhenium analogue [(q5-C,H,)Re(NO>(PPh,)COMe] (37). Iron has a total of three pairs of electrons available for backbonding as determined by having eight valence d-electrons, subtracting the number of odd q ligands, and dividing by two.Back- bonding into the carbon monoxide ligand leaves two lone pairs of electrons that are of similar energy and which are potentially available for resonance interaction with the acyl ligand. It is concluded then that the preferred conformation of the iron acyl complex will be dictated by steric interactions. Note that an electronic contribution from the metal to acyl ligand is operating in the chiral auxiliary [(q5-C5H5)- Fe(CO)(PPh,)] but does not exert any significant directional influence on attached acyl ligands. In a similar calculation for complex (37), rhenium having seven valence d-electrons results in a total of two electron pairs. Subtraction of one pair for backbonding to the strongly n-acidic nitrosyl ligand leaves only one lone pair of electrons on the rhenium metal.An electronic interaction with the HOMO of the rhenium fragment, which lies in the plane containing the phosphorus, rhenium, and carbonyl carbon atoms (Figure 22), and the acyl ligand will result in a strong stereoelectronic conformational preference for the acetyl ligand on the rhenium complex (37) to lie orthogonal to the rhenium-phosphorus bond. Direct comparison of the X-ray crystal structures of [(q5-CSH5)Fe(CO)(PPh,)- COMe] (32) 30a and RS,SR-[(q5-C,H,)Fe(CO)(PPh,)COCH(Me)Et](38) 30k with RS,SR-[(q5-C5H,)Re(NO)(PPh,)COCH(Me)CH,Ph](39) 35h.r shows that the methyl and iso-butyl groups for the two iron acyl complexes (32) and (38), respectively, occupy the space between the cyclopentadienyl and carbon monoxide Blackburn, Dauies, Sutton, and Whittaker 138) 139) Figure 24 X-Ray crystal structures of RS,SR-[(q5-C,H,)Fe(CO)(PPh3)COCH(Me)Et] (38)30k and RS,SR-[(q5-C,H,)Re(NO)(PPh,)COCH(Me)CH,Ph](39).35b.cThe diagram .slioic*sthe Neit5man projections along the alpha carbon to metal bond.Selectedprotons hazje been remoaed.for clarity ligands (Figure 24), which is the expected result based on a simple analysis derived solely on steric interactions (see Section 2E, ii).4-10,43 For the rhenium acyl complex (39), however, the CH(Me)CH,Ph group, a sterically more demanding moiety than methyl, nearly eclipses the nitrosyl ligand. Figure 17 indicates that the energy due to eclipsing interactions (0 = -90') is not significant, as compared to the preferred conformation (0 = -60"), and can be overcome by electronic forces.Thus, the stereoelectronic requirement present in the rhenium complex outweighs the steric forces that result from eclipsing interactions. It is of interest to note that calculations on both the iron and rhenium complexes have been reported and have shown the HOMO in each complex to lie in the plane defined by the phosphorus, metal, and C, carbon atoms with the second highest occupied molecular orbital (SHOMO) orthogonal to this plane.39.40 The energy difference between the HOMO and SHOMO for the iron auxiliary is calculated 44 to be a.1 1 kcal mol-' whereas for the rhenium auxiliary this difference is ca.48 kcal m~l-'.~'Based on PMO theory46 the preferred conformation of these metal acyl 43 B. K. Blackburn, S. G. Davies, and M. Whittaker, in 'Stereochemistry of Organometallic and Inorganic Compounds', ed. I. Bernal, Elsevier Science: Amsterdam, 1988, Vol. 3. 44 This value is taken from the EHMO calculations performed by Hoffmann and co-workers for the unsubstituted analogue [(qs-C,H,)Mo(CO)(PH,)]+ and is the energy difference between the a" (HOMO) and 2a' (SHOMO) molecular orbitals for [(q5-C,H,)Fe(CO)(PPh,)]'.35b "This value is the calculated energy difference between the HOMO and SHOMO on [(q'-C,H,)Re- (NO)(PH,)Me] 360 where there is expected to be no substantial orbital interactions from the metal to alkyl ligand and should, therefore, act as a satisfactory model for the fragment [(q'-C,H,)Re(NO)- (PH3)I-. 4h (a)K.Fukui, Bull. Chetn. SOC.Jpn., 1966,39,498; (h)K. Fukui and H. Fujimoto. ihid., 1968,41, 1989; (c) ideni. ihiri.. 1969. 42, 3399; (d)K. Fukui. ACC.Cliem. Res., 1971. 4, 57: (e) idem, Angew. Chem., In,. Ed Gig/.. 1982,21, 801; (.fl iriein. 'Theory of Orientation and Stereoselection'. Springer-Verlag: West Berlin, 1975 171 A Conformational Analysis of' Transition Metul q '-Acj,l Comp1e.ue.s LUMO I I I ISHOMO ', I I I I Figure 25 Comparison of moleculur orbiful infeructionsfor [(qs-C,H,)Fe(CO)(PPh3)CORJund [(q5-C,H,)Re(NO)(PPh,)COR] based on PMO theory complexes is where the acyl ligand lies perpendicular to the iron phosphorus bond to maximize overlap between the n* orbital of the carbonyl ligand and the HOMO on the metal.However, the orthogonal conformation will also allow dr-p,* inter-action to a similar extent for the iron auxiliary but to a very much lesser extent for the rhenium complex (Figure 25). This difference manifests itself in a strong stereo- electronic preference for the acyl to lie perpendicular to the phosphorus metal bond in rhenium but not in the iron acyl complexes. Note that steric repulsion continues to play a role in the rhenium complex by having the large CH(Me)CH,Ph group syn to the small nitrosyl ligand and away from the more sterically demanding zone between the cyclopentadienyl and triphenylphosphine ligands (Figure 24).5v43 Furthermore, the hydrogen atom and methyl group straddle the nitrosyl ligand with the small hydrogen atom residing in the sterically compressed zone between the carbon monoxide and triphenylphosphine ligand~."'~ Complexes where the steric interactions are greater than the stabilization that would be derived from resonance through d,-p,, orbital overlap are complexes (26) 27 (see Section 2D, ii) and [(q5-C,H,)Ru(C0)2CO(C,,H1,)I (40).29b Each of these complexes has a single lone pair of electrons as determined by the simple set of guidelines described above and, therefore, might be expected to adopt a con- Blackburn, Davies, Sutton, and Whittaker (40) Figure 26 X-Raj.crj~stalstructure qf”(q5-C,H,)Ru(CO),COC,oHl 5] (40).29bThe diagram sho~..vtlw Neir~man projecrion along the alpha carbon to ruthenium bond.The protons have been rcwioivd,fiw clarity formation with the acyl ligand orthogonal to the HOMO. However, the solid state conformation for each of these complexes is the one where steric interactions are minimized. It is suggested that for these complexes the steric repulsion between the acyl moiety and the proximate ligands is greater than the energy potentially gained by electron delocalization. Note that for complex (40) a certain degree of electron delocalization may occur through back donation from the metal (HOMO) to acyl ligand (p,.) but steric effects preclude maximization of this type of stabilization. B. Electrostatic Effects.-(i) DipoIar Effects.For most metal acyl complexes, a dipole exists in the acyl ligand as C( +ve)-O( -ve) due to electronic effects, and in consequence minimization of dipoledipole interactions in the relatively non-polar solid-state environment will be favoured. The dipole-dipole interactions for an organotransition metal acyl complex will be small relative to steric interactions that would be present and, therefore, will not affect conformational preferences that are predicted from steric arguments. For formyl complexes, however, dipolar inter- actions may affect formyl orientation since the formyl ligand is sterically smaller than an acyl ligand. For example, the complex { [P(OEt),],Re(CO),CHO} (41)47 is determined to have one pair of electrons that is expected to align with the plane containing the two trans phosphorus atoms, the rhenium, and C, carbon atoms.It is apparent from the X-ray crystal structure of complex (41)47 a stereochemical preference does exist, such that the formyl ligand eclipses the trans carbon monoxide and phosphine ligands (Figure 27), whereby maximum overlap can be achieved between the single lone pair of electrons centred at the metal and the 7r* orbital of the formyl ligand. Since the metal-phosphorus bond is a polarized as M( -ve)-P( +ve) the direction of the formyl ligand is opposite to that predicted by simple steric considerations and is directed towards the phosphine ligand to minimize the overall dipole in the solid state. 47 C. Sontag, 0.Orama, and H. Berke, Clrem.Ber., 1987, 120, 559. 173 A Conformational Analysis qf’ Transition Metal q -Acyl Complexes 0 (411 Figure 27 Newman projection along alpha carbon rhenium bond shohing the solid state coKformation qf { [P(OEt),],Re(CO),CHO) (41).4’ (42) Figure 28 X-Ray crysral structure of [(q5-C,H,)Re(NO)(PPh,),CHO] (42).35a.cThe diagram sh0NV.s the New?man projection along the alpha carbon to rhenium bond. Selected protons have been removed .for clariry In the rhenium formyl complex [(q5-C,H5)Re(NO)(PPh,)CHO] (42),35a.c shown in Figure 28, the formyl ligand lies perpendicular to the plane containing the phosphorus, rhenium, carbonyl carbon atoms, as determined by stereoelectronic requirements (see Section 3A), and the C-0 dipole is directed opposite to the dipole associated with the nitrosyl ligand, thus minimizing its dipole4ipole interactions in the solid-state.(ii) Hydrogen Bonding. Hydrogen bonding between an acyl oxygen atom and proximal ligands has been shown to play a role in determining acyl orientation 48 and is clearly illustrated in the complex [Re(CO),(NH,Ph)COMe] (43), shown in Figure 29.48a Cross-linking between adjacent molecules in the crystal lattice 48 (a)C. M. Lukehart and J. V. Zeile, J. Am. Cliem.Sue., 1978,100,2774;(h)A. J. Baskor, C. M. Lukehart. K. Srinivasan,J. An?.Chem. Suc., 1981,103,1461; (c)J. D. Korp and I. Bernal, J. Organumet. Chem.,1981, 220,355; (d)C. M. Lukehart and J. V. Zeile, J. Organornet. Chem., 1977,140.309;(a)C. M. Lukehart and J. V.Zeile, J. Am. Cliem.Soc., 1976,98,2365;(./) V. G. Albano, P. L. Bellon, and M. Sarson, Inorg. Chem., 1969, 8, 298; (g) F. A. Cotton, B. A. Frenz, and A. Shauer, Inorg. Cliim. Acta, 1973, 7, 161. 174 Blackburn, Davies, Sutton, and Whittaker (431 Figure 29 X-Ray crystal strucfure qf[Re(CO),(NH,Ph)COMe] (43).48aThe diagram shows the Netvman projection along the alpha carbon to rhenium bond. The protons hatle been removed for cluritj, through hydrogen bonding has been proposed for certain complexes and may also affect acyl ~rientation.~’ 4 Conformational Analysis of Ligands other than 7 ‘-Acyl The guidelines for determining acyl conformation set forth here may also be applied to other ligands, such as for example carbenes” and vinylidene~.~’ Thus, the preferred conformation of a carbene or vinylidene ligand in an organotransition metal complex will be primarily governed by steric interactions with a stereo- electronic preference existing only when there is a single lone pair of electrons on the metal fragment.52 X-Ray crystal structure determinations of a number of carbene 1.53*54 complexes have been reported; the complexes shown in Figure 30 J9 ((I) D.M. Chipman and R. A. Jacobson, Inorg. Chmm. Acta, 1967, 1, 393; (h) G. L. Breneman, D. M. Chipman, E. J. Galles, and R. A. Jacobson, lnorg. Chem. A~IA,1969, 3, 447; (c) A. J. Lindsay, S. Kim, R. A. Jacobson, and R. J. Angelici, Organometallics, 1984,3, 1523;(d) D. Messer, G. Landgraf, and H. Behren, J. Organomer. Chem., 1979, 172, 34 (e)H.Wagner, A. Jungbauer, G. Thiele, and H. Z. Behren, Nrrfurforscli., Teil B, 1979, 34, 1487. For leading references see: (a)U.Schubert, Coord. Chem. Rec., 1984,55, 272; (b)K. H. Dotz, H. Fischer, P. Hofmann, F. R. Kreisgl, U.Schubert, and K. Weiss, ‘Transition Metal Carbene Complexes’, Verlag Chemie: Weinheim, 1983; (c) F. J. Brown. Prog. Inorg. Chem., 1980, 27, 1; (d) D. J. Cardin and R. J. Norton. Organornet. Chem., 1982, 12, 213; (e) K. H. Dotz, Angew. Chem.. lnt. Ed. Engl., 1984, 23, 587; (1’)J. E. Hahn. Prog. lnorg. Chem., 1984, 31, 205; (g) D. B. Pourreau and G. L. Geoffroy, Ah. Organornet. Clrem., 1985,24. 249: (h)T. Aratani, Pure Appl. Cliem., 1985,57, 1839; (i) M. Brookhart and W. B. Studabaker, Chem. Rev., 1987. 87, 411.” For leading references see: (a)M. 1. Bruce and A. G. Swincer, Ah. Organornet. Chem., 1983,22, 59; (h) S. Abbot, D.Phi1. Thesis. Oxford University, 1984; (c) S. G. Davies S. Abbott, and P. Warner, J. Orgnnomet. Chem., 1983. 256, C65; (d)A. G. M. Barrett and M. A. Sturgess, J. Org. Chem., 1987, 52, 381 1; (e)G. Consiglio and F. Morandini, lnorg. Chim. Acta, 1987, 127, 79. 52 Note that steric interactions in vinylidene complexes are tempered due to the greater distance between the groups on the allenic ligand and proximal ligands. ‘3 (a)G. Huttner and H. Lorenz, Chem. Ber., 1975,108, 1864; (b)S.G. Davies, T. R. Maberly, R. H. Jones, M. E. C. Polywka. and G. Baird, in preparation; (c) T. R. Maberly, D.Phil. Thesis, Oxford University, 1986 (d)S. Fontana, U.Schubert, and E. 0.Fisher, J. Organomel. Chem., 1978, 146, 39. 54 P. Helquist, C. Knors, G.-H. Kuo, J. W. Lauher, and C. Eigenbrot, Organometallics, 1987, 6, 988. 175 A Conformational Analysis of Transition Metal q'-Acj>lComplexes 0 C I c PPh, c0 0 (44) (45) (46) Figure 30 Newman projections along alpha carbon metal bond showing the solid state con- formations for the carbene complexes [Cr(CO),C(OMe)Ph] (44),'3" [(q'-C,H )Fe(CO)-(PPh,)C(OMe)Et]+ (45),53b-Eand [(qs-C,H,)Mn(CO),C(OMe)menthyl] (46).53d.55 illustrate the steric and stereoelectronic influences on conformation in these types of complexes. Complexes [Cr(CO),C(OMe)Ph] (44) 53a and [(q ,-C,H,)Fe(CO)-(PPh,)C(OMe)Et] (45)53b,' are determined to have zero and two lone pairs of electrons, respectively (see Section 3A), and therefore adopt a conformation that reduces all steric interactions.Complex [(q5-C,H,)(CO),MnC(OMe)menthyl] (46) 53d*55 has a single lone pair of electrons on the metal fragment and thus adopts a conformation in which the n* orbital of the ligand (LUMO) is aligned parallel to the lone pair (HOMO). This stereoelectronic conformational preference is also seen in the homologous vinylidene complex [(q 5-C,H,)Mn(CO),CCMe,] (47),56 shown in Figure 31. It should be noted that in certain cases steric interactions may be so great as to overwhelm stereoelectronic preference^.^^ For example, complex (48) [(q ,-C,H,)-Fe(CO),C(SMe)Me] adopts a conformation that reduces steric interactions by + placing the thiomethyl and methyl groups bound to the carbene carbon atom in the least sterically congested space between the carbon monoxide and cyclopentadienyl ligands (Figure 32).54 However, the complex [(q5-C5H5)(C0)2FeCH(SPh)]+ (49),54 a closely related complex to (48), adopts the conformation that is predicted on stereoelectronic arguments (Figure 32).In the crystal structure determination of complex (49) there was identified two randomly distributed contributors to the structure that were determined to be (49a) (80%) and (49b) (20%). Despite the problems associated with disorder in the solid state, the crystal structure deter- minations for (48) and (49) are sufficient to reveal their conformational preferences. Comparison of complexes (48) and (49) shows that (49) does not contain the same 55 Several authors have noted that complex (46)adopts a conformation where the methoxy group occupies the space between the cyclopentadienyl and carbon monoxide ligands, but, as seen by its Newman projection in Figure 30 this is clearly not the case.[(a)U. Schubert, Organometalfics, 1982, I, 1085; (h) W. M. Jones, F. J. Manganiello, and M. D. Radcliffe, J. Organornet. Chem., 1982, 228, 271.14'" 56 H. Berke, G. Huttner, and J. von Seryerl, J. Organomet. Chem., 1981,218, 193. For other X-ray crystal structure determinations of vinylidene complexes see: (a)A. N. Nesmeyanov, G. G. Aleksandrov, A. E. Antonova, K. N. Anisimov, N. E. Kolobova, and Y. T. Struchkov,J. Organomer. Chem., 1976,110, C36; (b)M.I. Bruce, F. S. Wong, B. W. Skelton, and A. H. White, J. Chem. Soc., Dalton Trans., 1982, 2203; (c) R. M. Kirchner and S. Iber. Inorg. Chem., 1974, 13, 1667; (d)G. L. Geoffroy, D. B. Pourreau, A. L. Rheingold. and S. J. Geib, Organometallic.s, 1986, 5, 1337. 176 Blackburn, Davies, Sutton, and Whittaker (47) Figure 31 X-Ray crystal structure of [(q5-C,H,)Mn(CO),CCMe2] (47).56 The diagram yho\r:s the Newman projection along the alpha carbon to manganese bond. The protons have been remotled I;w clarity cC c 0 0 0 (48) (49a) (49b)Figure 32 Newman projections along alpha carbon metal bond showing the solid state con- formations,for the cationic carbene complexes [(qS-C,H,)Fe(CO),C(SMe)Me] + (4yi54 and [(q '-C,H,)Fe(CO),C(SPh)H] (49) :(49a) 80% contributor, (49b) 20% contributor+ steric demands as in (48)-there being a greater steric requirement for a methyl group relative to a hydrogen atom-and thus adopts the stereoelectronically preferred conformation.Interestingly, all the known pseudo-octahedral cyclopentadienyl manganese dicarbonyl complexes adopt the conformation where maximum orbital overlap between the p-orbital on the carbene and HOMO on the metal can be achieved, whereas the majority of the corresponding iron complexes adopt a conformation predicted on steric interactions. This result is surprising, since all of these types of complexes are isoelectronic and, therefore, are predicted to have a single lone pair of electrons. It is postulated that this difference stems from the energetic imbalance between the HOMO and an orbital of lower energy, that lies orthogonal to the HOMO and is also available for backbonding (Figure 33),38 on the iron and manganese fragments. A similar argument was presented above for the observed difference between the acyl complexes of [(q 5-C,H,)Fe(CO)(PPh,)] and its rhenium analogue (Figure 25).In order to rationalize the results observed for the iron and manganese systems it is suggested that for most of the iron complexes A Conformational Anal-psis of Transition Metal q ‘-Acyl Complexes Ll LL 111 111 I L i HOMO secondary orbital Figure 33 Illustration of’theorientation c?f‘tht.HOMO and the secondary lo).t’erenerg?.orbital steric interactions are greater than the energy gained by preferentially mixing the empty carbene p-orbital with the HOMO relative to the lower lying orbital, which was calculated to be approximately 6 kcal m01-l.~~ For the manganese complexes that are known, however, this energy difference is greater than the steric inter- actions present and they therefore adopt the conformation where the carbene p- orbital can efficiently interact with the HOMO.The difference described for these two complexes is commonly observed for similar types of complexes of the iron triad relative to the manganese triad. A single lone pair of electrons on the metal, in addition to the stereoelectronic conformational preferences, will also manifest itself in an increase in the barrier to r~tation.~~’,~’-~~For example, essentially free rotation about the iron to alpha carbon bond is observed for carbene ligands attached to the iron chiral auxiliary [(q 5-C,H,)Fe(CO)(PPh,)],57” predicted to have two lone pairs (Section 3A), whereas syn and anti rotomers can be isolated for carbene ligands attached to the rhenium chiral auxiliary [(q5-C,H,)Re(NO)(PPh,)1,58 calculated to have one lone 57 For example: ((I) M.Brookhart, J. R. Tucker, T. C. Flood, and J. Jensen, J. Am. Chem. Soc.. 1980. 102, 1203; (h) U. Guerchais and C. Lapinte, J. Chern. Soc., Chern. Commun., 1986, 894; (c) A. Davison and D. L. Reger, J. Am. Chern. Sor.., 1972,94, 9273: ((1) M. Brookhart, J. R. Tucker, and G. R. Husk. J. Am. Clrem. Soc., 1983, 105. 258.58 A. M. Crespi and D. F. Schriver. Orgnnomrtallic:c., 1985, 4, 1830. 59 ((1) W. E. Buhro, S. Georgiou, J. M. Fernandez. A. T. Patton, C. E. Strouse, and J. A. Gladysz, Orgnnometallics, 1986,5956: (h) W. A. Kiel. W. E. Buhro, and J. A. Gladysz, ihfd., 1984,3,879; (c) W. G. Hatton and J. A. Gladysz. J. Am. Chem. Soc.. 1983, 105. 6157. Blackburn, Dauies, Sutton, and Whittaker pair of electrons on the metal (Section 3A). In the former case the electronic effect is not significantly directional whereas in the latter a strong directional stereo- electronic effect is apparent. Similar results are observed for the corresponding vinylidene complexes of [(q5-C5H5)Fe(CO)(PPh3)] 6o and [(q5-C5H5)Re(NO)- (PPh3)].6' 5 Conclusions The dominant force that controls acyl conformation in transition metal q '-acyl complexes is steric interactions.Therefore, the preferred conformation of an q '-acyl complex can, in general, be rapidly, and correctly predicted by taking into account the steric interactions that would be present between the acyl ligand and its proximate ligands. We have shown this to be the case by correlating the sterically preferred conformation, as determined by using the CHEM-X molecular modelling program only calculating van der Waals interactions, to the corresponding X-ray crystal structures available. Only in exceptional cases, where a stereoelectronic conformational preference is present in the system, is the sterically preferred conformation not strictly adhered to. We have provided a set of guidelines where the electronic contribution of the metal fragment can be readily predicted, including several examples of their use on known complexes.Furthermore, this hypothesis can be successfully extrapolated to ligands other than acyl; for example, Fischer carbenes and vinylidenes. Acknowledgement. We thank BP International Limited for a Venture Research Award. 60 B. E. Boland-Lussier. M. R. Churchill. R. P. Hughes, and A. L. Rheingold, Organometallics, 1982.1.628.'' J. A. Gladysz and A. J. Wong, J. Am. Chem. Soc.. 1982. 104, 4948. 179

ISSN:0306-0012    

DOI:10.1039/CS9881700147

出版商:RSC    

年代:1988

数据来源: RSC

摘要:

Chem. SOC.Ret‘., 1988, 17, 181-227 The Chemistry of Molten Acetamide and Acetamide Complexes By D. H. Kerridge DEPARTMENT OF CHEMISTRY, UNIVERSITY OF SOUTHAMPTON, SOUTHAMPTON SO9 5NH 1 Introduction Molten acetamide is a currently under-appreciated non-aqueous liquid with unusual and useful properties. Its melting point, stability, and cost make it con-venient to use and its dipolar nature makes it a particularly good solvent; unusually it is equally good for both ionic inorganic and for covalent organic compounds. Stafford’ reported more than 50 years ago that over 400 organic compounds dissolved in molten acetamide and over 200 inorganic compounds. In fact the only substance found to be insoluble was cellulose. Unfortunately no solubility values were quoted in this report, but more recently 65 solubilities have been determined, all substantiaL2 The earlier review (following up an extensive paper “) noted the ‘water-like’ properties of molten acetamide in physical properties, acid-base reactions and in coordination, but concentrated largely on the first.The only other general review published in 1967, apart from a more recent but inaccessible work in Russian,6 was a somewhat short summary of the known information, though more specialized accounts, for example of acid-base interactions,’ have been published. This review endeavours to deal comprehensively with the now very extensive published work on the chemistry of molten acetamide and includes consideration of acetamide complexes with metals and non-metals since there is an obvious relationship between such complexes and the interactions of the appropriate ions in molten acetamide solution. Thus data on the stoicheiometry and significant properties of the former (e.g.mode of bonding, thermal stability) have been included even when such complexes have been prepared from acetamide dissolved in another solvent (often water, alcohols, ethers, benzene, or carbon tetrachloride) rather than from molten acetamide. However, the quite extensive chemistry of acetamide dissolved in such solvents will not otherwise be considered. The other boundaries of this review are that the purely organic reactions of acetamide will not be dealt with, nor in general will the chemistry of formamide, the ’ 0.F.Stafford. J. hi.Chenr. Soc., 1933, 55, 3987.’ R. C. Paul and R. Dev, Res. Bull. Panjab Unic., Sci.,1969, 20, 139. G. Winkler, ‘Chernie Nichtwasserungen Ionisierenden Losungsrnitteln’, ed. G. Jander, H. Spandau, and C. A. Addison, Interscience, 1963. Vol. 4, 203. G. Lander and G. Winkler, J. Inorg. Cizem., 1959, 9, 24.’J. W. Vaughn, ‘The Chemistry of Non-Aqueous Solvents’, Vol. 11, ed. J. J. Lagowksi, Academic Press, New York. 1967. B. Imanakunov. Vzairnodeistvie Atsetamida s Neorganicheskhirni Solyarni, Ilirn: Frunze, Kirg SSR, 1976.’S. Guiot and B. Trernillon, J. Electroanal. Chem., 1968, 18, 261. The Chemistry of Molten Acetamide and Acetamide Complexes higher homologues or the alkyl-substituted derivatives of acetamide be considered. 2 Properties of Pure Acetamide A.Crystal Structure.-Solid acetamide is known in two crystal forms. The stable modification is trigonal (rhombohedral) and is obtained by crystallization of solutions in organic solvents (e.g.ethyl acetate) and has a melting point usually quoted as 81 OC, but recently also as 829 and 80.3°C.'0 The acetamide molecules have been shown to be in the keto form with the amino group in the plane of the carbons, bond distances agreeing both from X-ray diffraction and from neutron diffraction ' measurements. There are numerous hydrogen bonds to three other molecules. The metastable orthorhombic crystal form is obtained by crystallization of molten acetamide,' 2,1 (though rhombohedral crystals grown from the melt have also been claimed 14) with a lower melting point: values of 69-73 "C have been quoted. The bond distances are similar to those in the stable form, the differences lying in the many hydrogen bonds which initially result in dimers that are further linked together in columns.B. Preparation and Purification.-Acetamide is prepared by many reactions,' including MeC0,Et + '(NH,),CO, '1 (2)MeC0,Et + NH, -NH,OOCMe A MeCONH, + H20 MeCOCl + NH, (3) (MeCO),O + NH, (4) MeCN + H,O -MeCONH, (5) CH,=C=O + NH, -MeCONH, (6) but the most efficient commercially involve distillation of ammonium acetate at 1-200 "C to give a 95%pure product from which the reagent grade is obtained by recrystallization (product >99%, 0.25% H,O, melting point 79-81 "C).The technical grade contains -0.75%water, with melting point 78-82 "C.Acetamide is odourless if pure, the mouse-like smell arising from an unidentified impurity. Since acetamide is hygroscopic, several methods for removal of water have been F. Senti and D. Harker, J. Am. Chem. Soc., 1940, 62, 2008. L. Vogel and H. Schubert, Wiss. Z. Marlin-Luther-Unit.., 1985, 34, 79. lo H. G. M. De Wit, C. G. De Krnif, and J. C. Van Miltenburg, J. Chem. Thermodyn., 1983, 15,891. 'I G. A. Jeffrey,J. R. Ruble, R. K. McMullan, D. J. De Frees, J. I. Binkley, and J. A. Pople, Acta Crysrallogr., Sect. B, 1980, 36, 2292. W. C. Hamilton, Acra Crystallogr., 1965, 18, 866. I3 W. A. Denne and R. W. H. Small, Acta Crystallogr., Sect. B, 1971, 27, 1094.l4 T. Ottersen, J. Almlof, and H. Hope, Acra Crystallogr.. Sect. B, 1980, 36, 1147. l5 'Kirk-Othner Encyclopaedia of Chemical Technology', 3rd Edn., J. Wiley, 1978. 182 Kerridge proposed. On a laboratory scale a much used method for removal of most water is distillation, with rejection of the fraction boiling below 210 "C. However a draw- back is that small quantities of acetonitrile are formed by dehydration16 (z.e. reverse of equation 5). The recommended method is therefore to dissolve reagent grade material in hot dry methanol (saturation is -125 g per 100 ml) and then treat the cooled solution with an excess of dry diethyl ether (-8 ml per g acetamide). The precipitated solid is filtered and then vacuum dried. C.Pyrolysis.-Acetamide has a boiling point of 221 0C,3 but the vapour begins to decompose at relatively low temperatures (at 220 "C forming -0.1 mole% NH,, -0.05 mole% MeCN and MeC0,H) l7 and rate measurements at higher temperatures have shown the process to be second order in acetamide, the suggested reaction being: 2MeCONH2-MeC0,H + MeCN + NH, (7) Hydrogen chloride has a marked catalytic effect (x 10 rate) which suggested that conversion of acetamide into the hydroxyimide form was required before a rapid reaction took place.' 8*1 These findings suggest that the preparative reactions (equations 1-5) involve a number of equilibria other than those stated. D.Toxicity.-Acetamide occurs in small quantities quite widely in nature. For example it has recently been reported as a minor constituent of the aroma of Tilsit cheese,20 and of the odour of poultry manure.21 Acetamide has also been widely used for many years in a variety of industries (production was -230 tons per annum in 1976), including the manufacture of cosmetics, textiles, lacquer, leather, paper, explosives, insecticides, and plastics, principally as a humectant and plasticizer, but also as a peroxide stabilizer and ingredient of soldering fluxes.New uses are also being proposed, for example as a de-icing agent.22 These widespread albeit relatively small scale uses have indicated that with normal handling and precautions there are no significant hazards to human beings, though it is a mild irritant and has a low toxicity.15 Tests on white mice indicated the maximum tolerated dose was 12 g kg-' body weight and that the minimum lethal dose was 15 g kg-'.23 In addition rats fed on a diet containing 2.51:; acetamide for 12 months developed malignant liver tumours (at 36% incidence, but 13% with 5% acetamide) 24 which were consistent with the hypothesis of chronic intracellular liberation of ammonia.No tumours were found l6 E. C. Wagner, J. Chem. Educ., 1930. 7, 1135. "D. Davidson and M. Karten, J. Am. Chem. Soc., 1956, 78, 1066. M. Hunt, J. A. Kerr. and A. F. Trotman-Dickenson, J. Chem. Soc., 1965, 5074. '' J. Aspden, A. Maccoll. and R. A. Ross, Trans. Faradaj Soc., 1968, 64, 965. 2o K. N. Ney, Feire, Seijen, Anstrichm., 1985, 87, 289. 2' A. Yasuhara. J. Chromatogr., 1987, 387, 371.22 R. Vogel, West German Patent, DE 3434953, 1986. 23 F. Bergmann and L. Haskelburg, J. Am. Ciirm. Soc.. 1941, 63, 1437. 24 B. Jackson and F. I. Desson, Lnh. Inaest.. 1961, 10, 909. The Chemistrji of Molten Acetamide and Acetamide Complc.ues in other organs.25 This report resulted in acetamide being placed prominently on lists of carcinogens, despite the relatively massive dose involved. However, more recent studies have shown acetamide to have no mutagenicity towards lower organisms (Aspergillus niduluns 26 and Salmonella typhimurium 27), nor towards Drosophilia melunogaster.28 In additon it did not cause DNA damage to rat hepatoma cells, or to primary Syrian hamster embryo cells. Interestingly, it was found not to damage liver cells in ui~o.~~ E.Physical Properties.-Extensive lists of the older values are given in Winkler,3 but more recently the density of molten acetamide has been given as: p/g~rn-~= 1.3576 -0.0012T +0.64 x T2 (8) together with values for the heat capacities of the solid and the melt.' The latent heat of fusion and the thermal conductivity, both important para- meters in the possible use of acetamide as a phase-change thermal energy storage material, have recently been remeasured. The former has been given as 249,' 225," 264,,O 265 lo kJ kg-', which are not dissimilar to the older value of 252 kJ kg-l,,l though lower than the 278 kJ kg-' determined in 191432 and still q~oted.~ Moreover, whatever the precise value proves to be, it is evident that it is satisfactorily high for its intended use, and indeed higher than that for most organics (eg.diphenyl 140, palmitic acid 165 kJ kg-') and that for a number of inorganic salts (e.g. KBr 215, NaNO, 174 kJ kg-').,, The most recent report gives the thermal conductivity as 0.287 W m-' K-' at 50 "C and 0.240 at 91 0C,34which reinforce the single values (0.252 and 0.263) given thirty years bef~re,~~.~~ and indicate only a 2% change on melting. The electrical conductivity of the solid is low (eg.3.0 x R-' cm-' at 80°C) and has been attributed to proton tran~fer,~' but no corresponding values for the melt appear to have been published. The dipole moment of acetamide in dioxane and benzene solutions have been measured several times 38-43 and the calculated values of 3.87 and 3.92D 25 J.H. Weisburger, R. S. Yamamoto, R. M. Glass, and H. H. Frenkel, Tosicol. Appl. Pliarn~ucd. 1969, 14, 163. 26 R. Crebelli, D. Bellicampi, G. Conte, L. Conte, G. Morpurgo, and A. Carere, Mutat. Res., 1986. 172. 27 E. Dybing, W. P. Garden. E. J. Soederlund. J. A. Holrne, and E. Rivedal, Ah. ESP.Med. Biol., 1986, 197. 28 R. Valencia. J. M. Mason, R. C. Woodruff, and S. Zimmering, Enoiron. Mutagen, 1985, 7, 325. 29 P. E. Arndt, J. G. Dunn. R. L. S. Willis, Therniochini. Acta. 1984, 80, 343. 30 H. H. Emons, R. Naurnann. K. Jahn, and H. J. Flamrnersheirn, Thermochim. Acta, 1986. 104. 127. 3' K. Hrynakowski and A. Smoczkiewiczowa, Roc. Chem., 1937, 17, 165. 32 A. H. R. Muller, Z. Plrysik. Client., 1914, 86, 177. 33 G.Janz. 'Molten Salts Handbook', Academic Press, New York, 1967. 34 R. Nikolic, K. Kelic. and 0.Neskovic, Appl. Phj.,s. A, 1984. 34, 199. 35 L. P. Filippov, Vestn. Mosk. Unir., Asronomi~~n, 1960, 3. 61. 36 L. P. Filippov. Vestn. Mosk. Univ.. Asronomiju, 1965, 20, 94. 37 Koichi Hirano, Bull. Ckeni. Soc. Jpn.. 1965, 38. 8425. 38 W. D. Kumler. J. Am. Chem. Soc., 1952, 74, 261. 3Q M. J. Aroney. R. J. W. Le Fevre. and A. N. Singh. J. Ckrm. Soc.. 1965, 3179. 40 C. M. Lee and W. D. Kumler, J. Am. Cliem. Soc., 1961, 83, 4586. 41 W. P. Purcell and J. A. Singer. J. Piij~.v.Clieni.. 1967, 71, 4316. 42 W. W. Bates and M. E. Hobbs. J. An]. Chem. Soc.. 1951. 73, 2151. 43 W. D. Kumler and G. W. Porter, J. Am. CIieni. Soc,., 1934.56, 2549. 184 Kerridge respectively at 25 "C are consistent with a planar, keto structure. Magnetic susceptibilities of -3.50 and -3.32 x lo-'' m3 mol-' for solid a~etamide~~,~~and refractive indices of 1.4890 for the solid and 1.4270 for the melt have been reported.46 F. Vibrational Spectroscopy.-The infrared absorption bands of acetamide-metal complexes have been much studied, since certain shifts from the values of acetamide alone have been taken as diagnostic of coordination to metal via oxygen or via: nitrogen. The band positions (and their assignment) that have been quoted for acetamide alone do vary somewhat: a compilation in chronological order is given in Table 1. However these measurements have been made on solid acetamide or its solutions, while molten acetamide has only been the subject of two ~tudies.~~.~~ The later and more q~antitative~~ showed 17 absorption bands at 95 "C which included those of Table 1 though often with quite different intensities [the relevant frequencies, intensities, and assignments are: 3 343(3.5) and 3 191(3) NH vibrations; 2 997(7) and 2 935(26) C-H vibrations; 1 664(11.5) and 1 612(6.5) C=O vibrations; 1418(2), 1 388(9.5), and 1348(8.5) C-H deformations; 1 123(9.5) NH, rock; 1001(2.5), 955(2), and 862(24) chain vibrations; 570(9) and 447(8) CON deformations].Most interestingly a new line at 984 cm-' appeared in melts at 71 "C and intensified as the temperature was lowered (acetamide can be readily supercooled). This new line was attributed to the p (orthorhombic) crystal form, because of its lower symmetry.Simultaneously with the appearance of the new line, the 570 cm-' band shifted to 593 cm-' and the intensity of the 862 cm-' band fell. The intensities of the 1664 and 1612 cm-' vibrations decreased from 90°C which indicates a progressive change in the proportions of the a and p forms. These changes were not simply the formation of the imidazole form, since they were 'eliminated by monoallylation' (presumably of the amide group) which 'does not prohibit tautomeri~ation'.~~ The earlier brief report, for which no temperature was given for the 'molten acetamide', broadly confirmed the positions of the high temperature bands.48 The broad 3 172 cm-' band has also been cited as evidence of dimerization49 and the 3 484-3 513 cm-' band of the formation of di-, tri-, and tetramers in carbon tetrachloride solution.50 However the shifts in these bands found with lithium chloride and lithium perchlorate solutions have been interpreted in terms of hydrogen bonds to the anions and Fermic resonance, rather than in terms of changes in the association of the solvent rnolecule~.~' Part of the difficulty of interpretation of the acetamide infrared spectrum lies in the fact that many absorptions are not localized between particular pairs of atoms. 44 S.K. Siddhanta and P. Ray, J. Indian Chem. Soc., 1943, 20, 359. 45 H. Francoise, Bull. Soc. Chim. Fr., 1962, 506. 46 F. Nechai. Zh. Tekli. Fi:., 1956, 26, 436. 47 S.S. Urazovskii and 0.A. Gunder, Dokl. Akad. Nauk SSSR, 1953,91, 885. 4R L. Kahovec and K. Knollmuller, Z. Plzys. Chem. B, 1941, 51, 49. 49 R. M. Badger and H. Rubakova, Proc. Nntl. Acad. Sci. USA, 1954,40, 12. 50 P. J. Kruegar and D. W. Smith, Can. J. Chem., 1967, 45, 1611.'' J. Bukowska, Pol. J. Chem.. 1984, 58, 243. 185 The Chemistry of Molten Acetamide and Acetamide Complexes U 9 99 9 9 xz ?? z z 7 z ? z F0 ?u ??uu ?u ?u hh YU d ? z ? z X z ? z \omW 3 3 33 Y h E: vi00\o g\o 3 333 -33 x z ? z 9,* F3 A 22mrn mm mmmmc‘1m- mm mm 186 Table 1 (continued) Bund frequency (cm-') and assignment Medium Re$ 1 140(~) N-H 1014(~) C-H ? C340 ('I} N-H 1 684 (s) Amide I 1 629 (s) Amide I1 1 410 (m) C-N 1045 (m) rock3 180 (s) bend 545) N-H 1675 Amide I 1 595 Amide I1 1385 C-H 1335 C-N d3 430 3 540- N-H cc1, e asym solution 3 412 N-H sym CH2C12 f3365} N-H 1670 C=O solution3 190 3 560 (s) N-H 1 710 (s) Amide I 1 605 (s) Amide I1 1 380 (s) C-N CHC1, g asym solution 3440(~) N-H 1685(~) sYm 3 445 N-H 1685 C=O 1410 C-N h i"') N-H 1684 C=O 1410 C=N3 180 3440J} N-H 1685 Amide I 1629 Amide I1 1410 C-N(3 110) 3 290 N-H 1640 C-C Nujol or k asym HFBD mull 3 138 N-H SYm 3 360 N-H 1665 C=O 1620 NH, 1395 C-N ? 1 bend I695 C-0 1410 C-N 760 N-C-0 bend The Chemistry of Molten Acetamide und Acetamide Complexes %.c CQ omss 3-?? uz 'E, QJ g 2 mm 33 09uu gz2N-mm Kerridge For example while the band labelled ‘Amide I’ (Table 1) is almost exclusively (> 80%)due to the carbonyl stretching mode the amide I1 band is stated to have contributions both from the NH in-plane bending ( -60%)and from CN stretching modes, and the amide I11 band (position not quoted for acetamide) includes CN stretching (-40%), NH in-plane bending (-30%) and CH deformation modes,52 these assignments being based on a theoretical treatment of N-methyl a~etamide.~~ Part of the variation in experimental values for these and other bands (Table 1) is due to changes in intermolecular hydrogen bonding between the solid state and solution which particularly effects the NH and CO absorptions, and even the concentration of the solution introduces considerable variation, dilute solutions showing the least hydrogen bonding.52 Such variations naturally complicate comparisons and even assignments, since solution concentration is rarely quoted, and also complicate deductions about the mode of bonding of metal cations. Coordination through oxygen reduces the CO bond order, and hence amide I frequency values of around 1 650-1 660 cm-’ are often quoted.The amide I1 band decreases somewhat (-1 590-1 610 cm-’) and the amide 111 band shows irregular variation. The N-H stretching frequencies increase from the solid state values, but decrease from those of dilute solution, so that the basis of comparison needs to be clear. The band around 1400 cm-’ variously attributed to CH or CN bonds generally increases (by -10-20 cm-’) on coordination of oxygen to metal.By contrast bonding of metal cations through nitrogen would be expected to decrease the N-H stretching frequencies still further (due to increased hydrogen bonding) and also increase the frequency of the amide I1 band. Further changes are not easy to predict, and no stable example is available for study. In contrast the acetimidate anion (MeCONH -) does bond in metal complexes via-1 nitrogen, and is observed to produce a marked lowering of the amide I band (to 55&1 600 cm-’) a further band around 1480 cm-’, a strong C-H band (-1 420 cm-’) and slightly less strong N-H bands (around 3 250 ~m-l).’~ A few Raman measurements have been reported 54 and some infrared bands of deuterated species (together with cryoscopic and ebullioscopic measurements which showed the presence of dimers and trimers in sol~tion).’~~~~ Proton 58*59 and ’5N n.m.r.60 measurements have been made in various solvents. Protonation of carbonyl oxygen in fluorosulphuric acid was indicated,61 with just enough protonation of amide nitrogen when in liquid ammonia to explain the con- ’‘W.Gerrard. M. F. Lappert. H. Pyszara, and J. W. Wallis, J. Chem. Soc., 1960, 2144. 53 T. Miyazawa, T. Shimanouchi, and S. Mizushima, J. Chem. Phys., 1958, 29, 61 1. 54 W. Kutzelnigg and R. Mecke, Spectrociiim. Acta, 1962, 18, 549. 5.5 T. Uno, K. Machida, and Y. Saito, Bull. Soc. Chim. Jpn., 1969, 42, 897. ”M. Davies and H. E. Hallam. Trans. Faraday Soc., 1951,47, 1170.” M. Davies and J. C. Evans, J. Chem. Pliys., 1952, 20. 342. D. N. Fuchs and B. M. Rade. Monatsli. Ciiem., 1981, 112, 25. 5y H. Kamei, Bull. Soc. Ciiem. Jpn.. 1965, 38, 1212. 6o H. M. Kricheldorf and E. Haupt, fnt. J. Bid. Macromol., 1983, 5, 237. “ R. J. Gillespie and T. Birchall. Canarf.J. Clieni., 1963, 41, 148. 189 The Chemistry of' Molten Acetamide and Acetamide Complexes siderable increase in N exchange rate when ammonium chloride was pre~ent.~,.~~ N.m.r. measurements also indicated a change to acid-catalysed proton exchange, via nitrogen-protonated intermediate species, as the solvent polarity was increased (the change over in mechanism occurred with 90% aqueous THF).64 3 Reactions of Acetamide A. Acidic and Basic Reactions.-The 'water-like' nature of molten acetamide has long been known,4 the auto-dissociation constant being given as 3.2 x lo-" at 94OC, and thus the possibility of it acting as a medium for acidic and basic 2MeCONH, e[MeCONH,]+ + [MeCONHl-(9) reactions together with the formation of compounds with both acids and bases.A correlation of the pH scale in water with that in acetamide has been made.7 Hydrogen chloride was noted as early as 1857 to form a 1 : 1 compound with acetamide solutions (such compounds can also be called complexes or solvates). This, under vacuum and over sodium hydroxide, lost hydrogen chloride, forming a stable 1 :2 the melting point being given as 135 and 131 "C respective- ly.48 These compounds were considered to be substituted ammonium and oxonium salts respectively (MeCONH,+CI- and Me-C //O ' O\C-CH; C1-).66'NH, .H2N' The former 1: 1 compound was also formed by acetamide in liquid hydrogen chloride and was also considered ionic.67 Structure of the latter 1 :2 compound, called the 'hemihydrochloride', has been confirmed as having strong hydrogen bonds, with hydrogen situated between two carbonyl oxygens themselves only 2.451 A apart,68 while the intense infrared absorption between 600 and 1 600 cm-', with a maximum at 800-900 cm-' 69 that changed little on deuteration (centre of gravity -920 cm-') 70 was also taken to indicate a non-centred double minimum bond." Raman absorptions have been reported.48 1 :2 Acetamide compounds have also been claimed with hydrogen bromide (m.p.142 0C),4 and with hydrogen iodide.72 With nitric acid, 1 :1 and 1:3 compounds were formed whose crystal parameters Am+-..-:--A (--rrnor',,,a 71 or-..,",,-+:..,I..\ &LA L.--,l l,,-+L'--r+L.,. .__ . -=. .~. . .~~~ -~ ~~~-~ ~ ~ ~ L. . ,._.-L,,,, -..-.--.-..-.--.__... --.-=.. ~~,,, ~~~-o----62 L. L. Strizhak, S. G. Demidenko, and A. I. Brodskii, Dokl. Akad. Nauk SSSR, 1959. 124, 1089. 63 L. I,. Gordienko and A. I. Brodskii, Dnkl. Akad. Nuuk SSSR, 1960. 134, 595. 64 C. L. Perrin and C. P. Lollo, J. Am. Chem. Soc., 1984, 106, 2754. "A. Strecker, Ann. Chew?., 1857, 103. 321. "E. Spinner, Specrrocliini. Acta, 1959, 15, 95. 67 M. E. Peach and J. C. Waddington, J. Chem. Soc., 1962. 600. 68 K. W. Muir and J.C. S. Speakman, J. Chew. Res., 1979, 277. 69 N. Albert and R. M. Badger, J. Cliem. Phys., 1958, 29. 1193. 70 E. Spinner, J. Cliem. Soc.. Perkin Tran.v. 2, 1980, 395. " J. Emsley, D. L. Jones, and J. Lucas, Reu. Inorg. Chem.. 1981. 3, 105.''G. Winkler, Dissertation. Tech. Univ. Berlin. 1957. Kerridge former clearly showing a protonated carbonyl oxygen 73,74 (rather than a proton- ated amine group). The crystal structure of the latter compound showed short, almost symmetrical, hydrogen bonds between the proton and the oxygens of the two acetamides, the oxygens of the nitrate anion each being coordinated by two amine groups.75 Infrared and heat of combustion measurements have also been made.54,75 A 1:2 compound was found4 with perchloric acid (m.p.91 "C) but in contrast a 2: 1 compound was reported with meta-iodic acid, though a structure was not determined.73 With trichloroacetic acid both 1 :1 and 2: 1 com-pounds were found; 76 solutions of the latter decarboxylated slowly at 100 0C.77 acetic acid only a 1 : 1 compound was detected (m.p. <2.5 0C).78 Titrations of per- chloric acid in molten acetamide with sodium and potassium acetimide have con- firmed the 1:2 and the formation constant of this has been deter- mined in acetic acid solution.80 The relative strength of a number of protonic acids at 100"C in molten acetamide has been given as HSO,F > H,SO, > picric acid > Cl,CCO,H > salicylic acid > Cl,HCCO,H > ClCH,CO,H > benzoic acid > CH3C02H,78'81and as HClO, > HNO, > HBr > HCl > picricacid > p-toluene sulphonic acid.4 Measurements of specific conductance, density, viscosity, and transference numbers have been made with solutions of sulphuric acid (shown to form 1: 1 and 1:2 compounds) and of phosphoric acid (only a 1:1 compound).Labelling these acids (with 35Sand 32P,and the acetamide with 14C) has shown a prototropic (or 'proton hopping') mechanism to be occurring at 0.4 to 1.0 mole fraction of acid, and almost zero ionic migration. However, in more dilute solutions ( <0.4 M) labelled carbon migrated to the cathode (as protonated acetamide) and, in the case of sulphuric acid, only labelled sulphur to the anode (as HSO,, which was thus not ~olvated).~~-~~ At higher temperatures with concentrated sulphuric acid, decomposition occurred (46.4%decomposition at 300 "C, forming 0.9 1 milli-moles SO,, 1.20 CO,, and 0.29 CO per millimole a~etamide).~~ Long chain aliphatic acids (lauric, myristic, palmitic, and stearic acids) form 1: 1 complexes.86 Ammonia has been stated to be insoluble in molten acetamide at 94 0C,77 but had earlier been presumed to form a 1 : 1 compound.78 However, compounds have not been reported with organic nitrogen bases, though measurements of con-73 M.Z. Buranbaev, A. I. Gubin, and Y. P. Gladii, Fiz. Tverd. Tela, 1982, 33. 74 A. I. Gubin, A. L. Vanovskii, Y. T. Struchkov, B. A. Beremzhanov, N. N. Nurakhmetov, and M. Z. Buranbaev, Crj-s!. Struc!. Commurt., 1980, 9, 745.75 Y. T. Struchkov. N. N. Nurakhmetov, M. Z. Buranbaev, and B. A. Beremzhanov, 1:v. Akad. Nauk Kaz. SSSR, Ser. Khim., 1986, 76. l6 I. M. Bokhovkin, Y. I. Bokhovkina, and E. D. Vitman, lzv. Vysshikh Uchehn. Zaued., Lesn. Zh., 1964,7, 137. 77 R. C. Paul, J. R. Singla, and R. Dev, Indian J. Chem., 1969, 7, 170. 78 H. H. Sisler, A. W. Davidson, R. Stoenner, and L. L. Lyon, J. Am. Chem. SOC.,1944, 66, 1888. 79 B. Gruttner, Z. Anorg. Chem., 1952, 270, 223. '"T. Higuchi and K. A. Conners, J. Phjx Chem., 1960, 64, 179. R. C. Paul and R. Dev, lndian J. Chem.. 1969, 7, 392. ''V. P. Basov, Y. A. Karapetyan, A. D. Krysenko, and Y. Y. Fialkov, Ukr. Khim. Zh., 1975, 41, 582. 83 V. P. Basov, Y. A. Karapetyan, and A. D. Krysenko, Zh. Fiz. Khim., 1974, 48, 78.84 V. P. Basov, Y. A. Karapetyan, and A. D. Krysenko, Zh. Fiz.Khim., 1973, 47, 1199. G. M. Schwarb and 0.Neuwirth, Chem. Ber., 1957, 90,567. 8b F. C. Magne and E. L. Skau, J. Am. Chem. Soc., 1952, 74, 2628. The Chemistry of Molten Acetamide and Acetamide Complc.ws ductivity and infrared absorptions have been taken to indicate attachment to acetamide via its amine group, or even protonation of the base^,^^,^' similar shifts being observed as for coordination of a Lewis acid to carboxyl oxygen. The alkali metal compounds formed by reacting the metal with acetamide in dry organic solvent (dioxan for sodium, benzene for pota~sium),~.~~ by distilling ethanol from a solution of acetamide in ethanol containing 5% sodium ethoxide, or by the reaction of sodamide with molten a~etamide,~ are certainly ionic (m.p.of NaNHCOMe 315 oC48) and an infrared study has shown the 24 bands to be very similar to those from carbo~ylates.~~*~~.~~ Lithium acetimidate has been made by ion exchange of the sodium acetimidate in molten acetamide," and both lithium and sodium compounds by solvolysis of the alkali metal hydrides with the melt." Basic solutions of lithium, sodium, or potassium acetimidates, as well as of organic bases (e.g. pyridine, quinoline, benzylamine, and piperidine) have been titrated conductiometrically against solutions of many acids with results in accord with the autodissociation (equation 9 ab~ve).~.~~.~~." *82.y1.92 Though it was early claimed that the equivalent conductances indicated that all salts [except the zinc, cadmium, and mercury(r1) halides] were completely dissociated in molten a~etamide,~ the degree of dissociation has since been shown to be smaller in most cases (varying for example from 96.4% for Et,NHBr to 64.50/;; for Li02CMe).93 B.Main Group Elements.--(i) Reactions oj' Alkali Metal Cations. In this and succeeding sections compounds of acetamide not conveniently listed as acids or bases (in Section 3A) will be considered on the basis of the Periodic Table, even though decisions about the most significant other element present are somewhat subjective. As would be expected lithium cations were shown by X-ray diffraction analysis to be coordinated to, and bridged by, acetamide molecules [eg.in solid Li(MeCONH2),,2(MeCONHCOMe)C104"1 and 0.1 M solutions of lithium per- chlorate were stable enough at 99 "C to act as the supporting electrolyte in polarographic studies.95 Lithium nitrate has been claimed to form both a 1 :1 and 1:2 complex (m.p. 137 and 105 "C) with two eute~tics~~.~~ though another later phase diagram study found no compounds and only one eutectic.'* But both sets of measurements of electrical conductivities, densities, and viscosities (over a com- position range from 25 to 225 "C) agree that a break in the property/composition "J. N. Rakshit. J. Chem. Soc., 1913, 103, 1557. P. D. Crispin and R. L. Werner. Ausr. J. Chem.. 1967, 20. 2589. 89 H. Lenormant, Compres Rendus, 1946, 222, 1293. 9o E. Blasius and F.Wolf, Z. Anal. Chem.. 1959, 171, 88. 9' R. C. Paul and R. Dev, Indian J. Chem., 1965, 3, 315. 92 V. D. Prisyazhryi, D. P. Tkalenko, N. A. Chmilenko, and S. A. Kudrya, Fiz. Khini. Elektrokhirn. Rasplud. Tcerci. Elektrolitor Tezi.~!, Dokl. Vses. Konf: Fiz. Khim. lonnrkh Rasplauoi~ Ti'ertl. Elektrolitoc 7th, 1979, 2, 58. 93 R. C. Paul and R. Dev. Rex Bull. Panjrrh Unic., 1970, 21, 517. 94 P. S. Gentile, J. G. White. and D. D. Cavalluzzo, Inorg. Chim. Acta, 1976. 20, 37. 9s M. L. Anand, J. lnrlian Chem. Soc., 1979, 56, 30. 9b M. A. Klochko and G. F. Gubskaya, Zh. Nearg. Khim., 1958, 3, 2375. 97 M. A. Klochko and G. F. Gubskaya, Zh. Neorg. Khim.. 1959. 4, 684. 98 I. M. Bokhovkin and E. D. Vitman, Zh. Ohshch. Khim., 1965. 35, 949. 192 Kerridge curves occurs at the (lower) eutectic, as is found in hydrate system^.^^.^^.^^ The phase diagrams, densities, viscosities, and electrical conductivities of lithium, sodium, potassium, and ammonium nitrates with acetamide have also been rep~rted,'~.'''*'O' together with electrochemical measurement^.^^ This consider- able amount of activity arises from the possible applications of lithium nitrate batteries, where even small amounts of acetamide increase anodic current densities significantly and hence the utilization of metallic lithium.'02 The solubilities of five sodium and potassium halides in molten acetamide have been rep~rted."~-''~ Best values have been given.Io8 Phase diagram and re- fractive index measurements on the aqueous acetamide lithium chloride system have indicated two compounds (LiCI*2MeCONH2*H,0 m.p.48-50 "C and LiCI-MeCONH,*H,O 'melting over a broad range').''' Phase diagram "O-' l2 and crystallographic ''3*1l4 studies have also indicated the formation of other compounds (NaBr-2MeCONH2, NaI-2MeCONH2 and KI-6MeCONH2 melting points 144, 110, and 55 "C respectively) and their Raman absorptions have been reported.48 The ionic nature of acetamide solutions of alkali metal halides and pseudo-halides has been shown by their high conductivities '04*' ''9' ' and solu- bility decreases with addition of a common ion.lo4 Viscosities,' l6 densities,' '7,1 ' and enthalpies of solutions ' ' 9-1 " ha ve been measured over a range of concen- trations and of temperatures (85-1 15 "C) and indicate an increasing disruption of the spacial network of hydrogen bonds by the added ions as the temperature increased.Similar conclusions arose from paramagnetism measurements.'22 This ionic nature has also allowed successful quantitative polarographic determination of potassium in a number of complexes,95 and the successful ion exchange of all the 99 G. Berchiesi, G. Vitali. and A. Amico, J. Mol. Liq., 1986, 32, 99. loo D. A. Tkalenko, S. A. Kudrya, A. A. Rudnitskaya, and N. A. Chmilenko, Fiz Khim. Svoistua Raspkwl Trertl. Elektrolitor, ed. Y. K. Delimarskii, 1979, 66. S. A. Kudrya, D. A. Tkalenko, and L. P. Antropov, Vesrn. Kiec. Politrkh. Inst., Ser. Klzim. Mashinostr. Tekhnol., 1974, 11, 84. I01 D. A. Tkalenko, S.A. Kudrya, and A. A. Rudnitskaya, Elektrokhiniiya, 1978, 14, 140. Io3 A. G. Sarkisov, S. Z. Melamud, and N. F. Sakharova, Khim.Sh. Nauch Tr. Kuibyshet:. Politekh. Inst., 1969, 3. lo' A. G. Sarkisov and S. Z. Sakharova, Kliiniij*a, 1969, 3. R. A. Wallace, Inorg. Chem., 1972, 11, 414. 'Oh R. A. Wallace, Inorg. Nucl. Chenz. Lett., 1973. 9. 601. lo' R.A. Wallace, and P. F. Bruins, J. Electrochem. Soc., 1967, 114, 212. '08 B. Scrosati, Soluhilifj Data Ser., 1980. 11, 245. lo' X. Jiang, G. Wu, Y. Li, and Y. Chen, Kexue Tongboa (Foreign Language Edn.), 1985, 30,1041. 'I0 B. Menshutkin, J. Russ. P~J.Chmi. SO c.. 1909, 40, 1415. 'I' J. W. Walker and F. M. G. Johnson, J. Chern. Soc., 1905, 87, 1597. A. W. Titherley, J. Cheni. Soc., 1901, 79, 413.P. Piret, L. Rodrigue, Y. Gobillon, and M. van Meerssche, Actu Crjstallogr., 1966. 20, 482. 'I4 Y. Gobillon and P. Piret. Acra Cr)sfallogr.,1962, 15, 1186. ''5 L. Belladen, GKZ. Cliim. Itul., 1927, 57, 407. 'I6 S. Taniewska-Osinka and M. Woldan, Ac,ta Unic. Loci:., Ser. 2, 1978, 24, 53. 'I7 S. Taniewska-Osinka and M. Woldan, Acta Univ. Loci:., Ser. 2, 1976, 6, 47.'' M. Woldan, J. Cheni. Eng. Data. 1987, 32. 177. 'I9 S. Taniewska-Osinka and M. Woldan, Acta Univ. Loci:., Ser. 2, 1978, 24, 17. S. Taniewska-Osinka and M. Woldan. Acta Uniu. Lod:., Ser. 2, 1978, 24, 67. S. Taniewska-Osinka and M. Woldan, Zh. Fiz. Khim., 1974, 48, 2154. Y. V. Ergin. L. L. Kostrova. and D. Y. Samaiiov, Zh. Strukt. Khim., 1978, 535. The Chemistry of Molten Acetamide and Acetamide Complexes alkali metal cations (as chlorides) in molten acetamide on powdered zirconium phosphate.' 23 (ii) Alkaline Earth Metal Cations.Much information on the interactions of alkaline earth cations with acetamide has been obtained from aqueous solutions, mainly by phase diagram investigations '24-' 30 but also from density,' ' viscosity,' ' electrical conductivity,' 3' DTA,13, and TG '33*134 measurements (alkaline earth metal oxide being the final product). Naturally the compounds claimed have usually been acetamide hydrate complexes as well as anhydrous (the stoicheio- metries being listed in Table 2) with the proportion of acetamide in the complex increasing with the concentration of acetamide in the aqueous solution. Infrared spectroscopy indicated coordination of acetamide through oxygen.' 32*134 Inter-facial tension measurements between aqueous acetamide solutions and immiscible organic liquids (isoamyl acetate, n-hexyl alcohol) have also been claimed to show association,' 3s (the stoicheiometries being MCI,.2MeCONH2, MC1,-MeCONH,, 2MC12*MeCONH, in the case of SnCI, and BaCI,, but only the first two with MgCl2).Table 2 indicates that there is no overall trend in melting or decomposition temperatures with varying ratios of acetamide, or of acetamide/water, nor any apparently preferred coordination number though structure determination may reduce the ostensible spread of 3 to 7. Interestingly, problems attributed by the authors to hydrolysis were reported in establishing the composition of the solid phase formed with magnesium chromate and certainly magnesium hydroxide was precipitated.'22 However in view of the reactivity of chromium(v1) with acetamide melts (Section 3C, iv) it seems possible that the problems were due to some reduction to chromium(II1) with resulting increase in the basicity of the solution and consequent precipitation of magnesium hydroxide.Solutions of alkaline earth halides and of alkali metal halides have been the sub- ject of extensive physical measurements, including densities,' ' viscosities,' '' en-thalpies,' 19-12' Raman and infrared spectros~opy,~~ electrical conductivities,' 's and potentiometric titrations.' 36 123 G. Alberti, Ricercu Sci., 1960, 30. 2139.'24 S. Baichalova and B. Imanakunov, Ix. Aknd. Nauk. Kirg. SSR, 1968. 64. 12' B. Imanakunov, S. Baichalova, P. Yun, and A. Dzhunusov, Muter. Nuuclino-Tekli. Konf. Posu~slrch. lOO(Sto) Letiyu Period Zukonu D. J. Mendeleevu, 1969, 85. V. F. Tarakanov, Uch. Znp. Yarosl. Gos. Perlugog. Inst., 1976, 154, 37. 12' S. Baichelova, I. G. Druzhinin, and B. I. Imanakunov. Z/r. Neorg. Khim.. 1967. 12. 1381. 128 K. Abykeev, K. Sulaimankulov, and M. Ismailov, IT. Akarl. Nuuk. Kirg. SSR, 1974. 36. 129 N. N. Gustomesova and A. S. Karnaukhov. Zh. Neorg. Khim., 1974. 19. 525. I3O V. T. Orlova, V. I. Kosterina, and I. N. Lepeshkov. Zh. Neorg. Khim.. 1985, 30, 1877. 13' E. A. Gyunner, Zlr. Nrorg. Kliim., 1962, 7. 1431. 13' I. A. Borukhov. T. F. Kalinevich, and A.K. Umarov. Deposited Doc., 1976. VINITI, 2963. 133 M. N. Nabiev, 1. A. Borukhov. and 0.A. Momot. Zh. Neorg. Khim., 1976, 21, 1958. 134 V. T. Orlova, V. I. Kosterina, E. A. Konstantinova, and I. N. Lepeshkov, T/iermochim. Ac,rrr, 1985. 92, 709. 13' H. V. Barot, K. Hemlata, and C. M. Desai. Viclw J. Gujurcrt Unir.. 1958, 2. 73. 136 T. Chouleru, Ann. Fac. Sci. Mursrille, 1954, 23, 11. Kerridge Reactions have also been reported, calcium hydride solvolysing 91 and calcium metal reacting vigorously and forming hydrogen in molten acetamide. lo' ~Table 2 Complexes of ucetaniide ,t.ith alkaline earth metal salts st oicheiomet ries, dens it ies melting points, and decomposition temperatures) Melting Poin i Decomposition Stoiciieiometry Colour Density ("C) Temperature ("C) Ref: BeCl2.MeCONH,.3H,O 1.2712 311 ub BeCI,.3MeCONH2.2H2O 1.2224 380 a, h BeSO,.2MeCONH2.H2O 1.4460 136 231 a, c MgC1,.2MeCONH2 white 68 d MgC12.4MeCONH 153 e MgCI,.4MeCONH,.2H20 1.3790 15G-3 290 a, b.f MgBr2.6MeCONH, 169,170 e*gMgI2-6MeCONH, 177 g Mg(SCN),.4MeCONH2 h Mg(ClO,),.SMeCONH, i Mg(Cl0,),.7MeCONH2 i Mg(N03),-4MeCONH,.H,0 6G5 j, k Mg(NO,),-4MeCONH2*2H2O 45 390 n, o Mg(N03),.6MeCONH,.2H,0 110 410 n, o Mg(O,CMe),-2MeC0NH2 P MgS0,-no compound U M gC r0,-MeCONH 2.6H ,O 1 CaCl,.MeCONH, m CaC12.3MeCONH, 186 n1 CaC12.4MeC0NH, white 72, >240 >240 d, q, r CaC12-6MeCON H , 62.64 g, m CaCI2.2MeCONH,.2H,O b CaCl2.4MeCONH,.2H,O b Ca(O,CMe),-no compound S CaS0,-no compound U SrC1,-no compound a SrS0,--no compound a BaC12.6MeCON H , white 60 d BaC1,-no compound U BaS0,-no compound U 'S.Baichalova and B. Imanakunov. Izc. Akad. Nauk Kira. SSR. 1968. 64. 'B. Imanakdnov, S. Baichalova, P. Yun, and A. Dzhunusov, Mater. Nauchno-Tekh. Konc Posujshch. ZOO(Sto) Letiju Period Zcikorio D.J. Menrleleecn. 1969, 85. E. A. Gyunner, Zh. Neorg. Khim., 1962, 7, 1431. R. C. Paul and R. Dev. Itidioti J. Chem.. 1967, 5, 267. L. Kahovec and K. Knollmullor, Z. Physik. Chem., B, 1941, 51, 49. S. Baichelova, I. G. Druzhinin, and B. I. Imanakunov, Zh. Neorg. Khim., 1967, 12, 1381. B. Menshutkin, J. Russ. Phj. Chem. Soc., 1907. 39, 102. K. Abykeev, K. Sulaimankulov, and M. Ismailov, IT. Akriil. Nuuk Kirg. SSR, 1974, 36. ' V. F. Tarakanov, Uch.Zap. Yarosl. Gos. Pedagog. Inst., 1976, 154, 37. 'I. A. Borukhov, T. F. Kalinevich, and A. K. Umarov, Deposited Doc. 1976, VINITI, 2963. li M. N. Nabiev. I. A. Borukhov, and 0.A. Momot, Zh. Neorg. Khim., 1976,21, 1958. N. N. Gustomesova and A. S. Karnaukhov, Zh. .Veorg. Khim., 1974, 19, 525. B. Menshutkin, J. Russ. Plij. Cliem. Sac., 1909, 40. 1415. " V. T. Orlova, V. 1. Kosterina, and I. N. Lepeshkov, Zh. Neorg. Khim., 1985, 30, 1877. 'V. T. Orlova, V. I. Kosterina. E. A. Konstantinova, and I. N. Lepeshkov, Thermochim. Acta, 1985, 92, 709. B. Murzubraimov. G. I. Shtrempler, M. Ismailov, and D. J. Altybaeva, Zh. Neorg. Khim.,1985, 30, 1896. R. C. Paul, S. L. Chadha. and R. Dev, Indian J. Chem., 1965, 3, 364. 'P. I. Kuznetzov, J. Russ. PIijx C'hcwi.Soc.., 1909, 41. 379. 'B. Murubraimov. G. I. Shtrempler. and S. Madanov, Zh. Neorg. Khim., 1985, 30, 2438. The Chemistrj? of Molten Acetamide und Acetumide Complexes (iii) Group ZZIB Compounds. The boron trihalides react with acetamide to form I : I adducts, by mixing solutions of the reactants in inert solvents,'37 by conductio- metric titration in liquid hydrogen ~hloride,'~ or by absorbing boron trifluoride vapour on solid acetamide.' 38.139 The bromide and chloride adducts are solids (BBr,*MeCONH, m.p. 98.9 OC, BCI,*MeCONH, m.p. 75.5-76.5 oC)s2-137and the fluoride a colourless, non-fuming, liquid.' 38 Hydrolysis occurs readily, as does pyrolysis (eg. BCl,.MeCONH, -+ MeCONH-BCl, + HCI -+ MeCN + B(OCl),+ HCl) 137 though boric acid has been claimed as a product,'3* as well as tetrafluoroborate.' 39 Acetamide has also been substituted into a carboborane (7- MeCONH,-7-CBl,Hl 2).140 Aluminium trichloride has been shown to form a 1 : 1 complex by density and viscosity studies in nitrobenzene solution,'41 and a 1 :6 complex (white solid melting at 143 "C with decomposition) by refluxing in benzene or carbon tetra- ch10ride.l~~Evidence for both the 1 : 1 and 1 :6 complexes was obtained from measurements on the aqueous ternary solution,'43 as well as for mixed com- plexes [Al2Cl,*3MeCONH,, n-C,H,NO, when nitrobenzene was the third component; 144 AI(OH)(OOCMe),( MeCONH,),,, and AI(OH)(O,CMe),* (MeCONH,), from the quaternary Al(0,CMe)3-MeCONH,-H02CMe-O(OCMe), system.14' The electrical conductivities of aluminium trichloride solutions in molten acetamide have been measured.146 With aluminium tribromide 1 : 1 and 2: 3 complexes were reported.'47 In molten acetamide solvolysis has been found to occur (probably at 100°C)9'though solutions with much more AlCI, + 2MeCONH, -AICI(MeCONH), + 2HC1 (10) aluminium trichloride (only 10-300/;, acetamide) were found to be solvents for high molecular weight organics and polymers, and to act as solvent media for a number of reactions, alkylation, dehydrogenation, halogenation and insertion of carbonyl, sulphur, and amine groups.'48 N.m.r. (,'A1 and 13C) indicated only a 1 :1 complex of low stability between acetamide and aluminium nitrate nonahydrate in heavy water. '49 Indium trichloride solutions can be reduced electrochemically to indium 13' W.Gerrard, M. F. Lappert, and J. W. Wallis, J. Cliem. Soc.. 1960, 2141. H. Bowlus and J. A. Nieuwland, J. Am. Cliem. Soc., 1931. 53, 3835. 139 E. L. Muetterties and E. G. Rochow. J. Am. Cliem. Soc., 1953. 75. 490. 14" T. Jelinek, J. Plesek, S. Hermanek, and B. Stibr. Cokt. Czrcli. Clieni. Conimun., 1965, 50, 1376. 14' B. Y. Rabinovitch and A. G. Ponomarenko, Ix. Kier Politekli. Inst.. 1954, 14, 98. 142 R. C. Paul and R. Dev, Inrlicin J. Cliem., 1967, 5, 267. 143 B. Y. Rabinovitch. Zl7. Ohshcli. Kliim.. 1954. 24. 48. 144 B. Y. Rabinovitch and A. G. Ponomarenko, Sh. Stntri Oh.d7di. Kliini. Akud Nuuk SSSR, 1953,2, 11 18. 14' I. Zlatera and R. Ioncheva. Dokl. Bulg. Akarl.Nuuk. 1979. 32, 1040. 14'E. M. Golobchik, L. G. Koshechko, and K. A. Tikomitova, Izc. V~~.s.s/i.licliehn. Zacerl. Kliini. Khim. Trkhno!.. 1987, 30, 138. 14' B. Y. Rabinovitch and Y. S. Notkin, Ix. Kioc Politekli. Inst., 1954. 14, 88. 14'Badische Anilin u Soda Fabrik, Ger. Pat. 878647, 1953. 149 R. Caminiti, G. Crisponi. V. Nurchi, and A. Lai. Z. Naturforsch.. Teil A. 1984. 39, 1235. Kerridge metal 150and have been used to form indium alloy surface layers on magnesium and aluminium alloys.' 51 Thallium trichloride was found to form 1:1 and 1 :2 complexes. The latter, on the basis of infrared, molecular weight, and conductance measurements, was suggested to be dimeric with two chloride bridges between six coordinate thalliums, and the former to be tetrahedral ions (i.e.[TIC12*2MeCONH,] and [TlCI,] -).' 52+ Thallium(]) is soluble in molten acetamide at 87 "C and has been found to behave reversibly in d.c. polarography and cyclic voltammetry.' 53 (iv) Group IVB Compounds. As was mentioned in the Introduction (Section 1) molten acetamide is an excellent solvent for very many organic compounds and complexes can be formed e.g. 2: 1 with phenol (2PhOH*MeCONH2, m.p. 42.5 "C).' 54 Phase diagrams have also been given for the potassium formate and acetate systems together with viscosity and density measurements.' 55 Acetyl chloride was found to react with molten acetamide to give diacetimide [(MeCO),NH] and the 1 :2 hydrogen chloride ~omplex.~ Metal alkyl compounds solvolyse in the melt to the hydrocarbon3 e.g.LiBu molten MeCONH, + Li(NHC0Me) + C,H,, (1 1) molten MeCONH,Zn(Et), Zn(NHCOMe1, + 2C,H, (12) Quinones have been reduced polarographically with a two-electron step (reversi- bility increasing in the order benzoquinone < napthaquinone = phenanthra-quinone < anthraquinone) with 1 M sodium acetate as the supporting electro- lyte.' 5631 57 Many organic acids have been titrated in molten acetamide (98 "C) and dissociation constants have been reported.82 Silicon tetrafluoride was found to form an unstable 1 :2 complex when passed into a solution of acetamide in benzene. The white crystals, initially very soluble in benzene, reacted on drying becoming insoluble in organic solvents (C,H6, MeCN, CCI,, CHCI,). Reaction was also indicated by the wide spread in melting point reported (105-128 "C, decomposition 22k245 "C).The infrared spectrum did not show the band shifts expected for the more usual coordination through oxygen, and nitrogen coordination was claimed by the authors.' 58 A possibly important factor is that silica tetrafluoride-nitrogen bond energies are more than twice those to oxygen.' 59 By contrast, silicon tetrachloride was reported to form a 1 :4 and a lSi'I. G. Erusalirnchik and E. A. Efirnov, Elekrrokhim. Protessj Elektroosazhednii Anodnom Rastvorenii Mctal, 1969, 129. Is' E. M. Golubchik, Zashch. Met., 1984, 20, 286. K. C. Malhotra and Balkrishnan. J. Inorg. Nucl. Chem., 1977, 39, 387. Is' V. Bartocci, M. Gusteri, R.Marassi, F. Pucciarelli, and P. Cescon, J. Electroanal. Chem., 1978,94, 153. Is4 N. Z. Rudenko and D. E. Dioniseu, Zh. Ohshch. Khim., 1956, 26, 1866. lS5 F. Castellani, G. Berchiese, P. Pucciarelli, and V. Bartocci, J. Chem. Eng. Data, 1982, 27, 45. R. Narayan and K. L. N. Phani, Trans. SAEST, 1984, 19, 177. Is' K. L. N. Phani and R. Narayan, J. Electroanal. Chem., 1985, 187, 187. 15* A. A. Enan and L. A. Gavrilova, Zh. Neorg. Khim., 1977, 22, 124. ls9 J. P. Guertin and M. Onyszchuk, Can. J. Chem., 1968, 46, 987. The Chernistrji of Molten Acetaniide and Acetaniide Comp1e.ue.y 1:8 complex (cream and light yellow in colour) but the high melting points ( >240 "C) may indicate considerable p~lymerization.'~~ On the interaction of germanium and its compounds with acetamide, the only report is of the electrochemical behaviour of the element itself, specific con- ductances, streaming and zeta potentials being measured, which showed molten acetamide to be 'as electrically active as water' because of electron transfer between germanium and the melt.'60 Tin(r1) chloride was found to form a four-coordinate 1 :2 complex,'61 while a probably six-coordinate di-acetamide complex with tin(1v) chloride has been reported by several group^,^^-^ 1*142,162-'66 and described as white crystals (m.p.109 "C), as well as with tin(iv) bromide lh5.1h7 and diphenyl tin(1v) chloride and bromide,' 68 in each case the infrared spectrum indicated bonding through oxygen, though in more acidic solution an oxonium salt [(MeCONH:),SnCl,] was formed and also studied by infrared spectros~opy.~' Higher (1 :4)complexes (SnCl,.4MeCONH,, white crystals m.p. 61 OC, and SnBr4-4MeCONH,, a yellow viscous liquid) have also been claimed and their conductances re~0rted.l~~ The specific conductance of the tin tetrahalide solutions and their ability to be titrated with various bases has suggested ionization to a six-coordinate diacetamide complex [SnX,-(MeCONH),]2' may occur to some extent in molten a~etamide,~'.'~~ though cryoscopy indicated that the complexes were undissociated in nitro- methane lh3 (1 :1 and 1 :2 breaks were found in conductiometric titration curves of SnC1, and SnBr, against MeCONHNa corresponding it is said to the formation of acid and neutral salts respectively 165).Tin(I1) chloride solutions have been electrochemically reduced to produce alloy surface layers '51 as have lead(r1) chloride solutions.' 69 With lead(r1) nitrate a 1:3 complex has been claimed.' 70 Lead(1i) cations were soluble in molten acetamide at 87 "Cand the solutions have been studied by polarography and cyclic voltam- metry but irreversible behaviour was reported. Lead(I1) and (IV) compounds [PbCl,, (NH,),PbCl,] were found to react with basic acetamide solutions, con- taining sodium acetimide, to give precipitates; the amphoteric character of these and other cations varied in the order [Sn" < Pb" < Pb" < Cu", Ag'].3 (v) Group VB Compounds. Ammonium nitrate forms a low melting eutectic with acetamide (40:60 wt%, m.p.38 "C; 17' 39:61, 38,98 37:63, 37.5 '72), and density, I6O W. T. Eriksen and R. Caines. Ph~.r.Clii~ni.Solids. 1960. 14, 87. 161 K. L. Jaura. R. K. Chadha. and K. K. Sharma. Res. Bull. Pnnjtrh Unit>.Sc,i.. 1979, 27. 213. 16' D. S. Bystrov and V. N. Filimonov. Fi:. Prohl. Spektroskopii Akd Ntruk SSSR Muter. 13 90 Soveslicli. Lmingrud. 1960. 2, 49. 163 R. C. Aggarwal and P. P. Singh, Z. Atiorg. C'hetn., 1964. 332, 103. 164 R. C. Paul, B. R. Srennathan, and S. L. Chadha, J. lnorg. Nucl. Cheni.. 1966. 28. 1225. 165 R. C. Paul and R. Dev, Indian J. Clieni., 1969, 7, 392. 16' R. C. Paul, S. L. Chadha. and R. Dev, Indian J. Cliem., 1965, 3. 364. 16' R. C. Aggarwal and P. P. Singh, J. Innrg. Nucl. CIiem., 1966, 28, 1655. 168 T.N. Srivastava, S. K. Tandon, and B. Bajpai, Inorg. Chim. Acta, 1975. 13, 109. E. M. Golubchik, Izv. V~~ssh.C'cliehn. Zaved. Khim. Kkim. Teklinol.. 1984, 27. 59. P. T. Danilchenko and V. G. Ediger, Ann. Inst. Anaf. Pliys. Chim. USSR. 1935, 7, 255. L. S. Bleshchinskaya, K. Suliamankulov. and M. D. Davranov. Zh. Neorg. Kliim., 1983. 28, 1068. '12 0.K. Khaishbashev. Bull. Acd. Sci. C'RSS CI. Sci. Chim., 1945, 587. Kerridge viscosity, and conductance measurements gave no evidence of compound formation.' 73 Similar physical measurements have been made on acetamide solutions of alkali metal nitrates.174*'75 (See also Section 3Bi.) Magnesium nitrate complexes are listed in Table 2 and calcium nitrate is more than 30 mole% soluble.' 76 No reports of phosphorus compounds, other than orthophosphoric acid, have been made (cf:Section 3A).This acid in molten acetamide solution (at 98 "C)has been used in titrations, and its dissociation constant has been reported.' With arsenic, the trichloride reacted with acetamide in refluxing benzene or carbon tetrachloride solution to give a white solid 1 :3 complex (m.p. 108 "C) whose conductance in molten acetamide has been measured.'42 Antimony trifluoride solutions in molten acetamide have been electrochemically reduced to the element,' 50 and antimony(m) chloride solutions have been more extensively studied. solvated protons being thought to form as well as antimony complexes.82*'77 2SbC1, + 2MeCONH, eSbC1,SNHCOMe + SbCl, + MeCONH; (13) Antimony(m) bromide and iodide have only been reported to react in molten acetamide to give insoluble prod~cts.~ Antimony(v) chloride was more acidic in molten acetamide and was found to be dibasic suggesting a 1:2 complex (SbC1,- 2MeCONH2)'65 with unusual seven-coordinate antimony.This complex, together with 1 :1 and 1:5 complexes have all been claimed and stated to be white crystals m.p. 62 "C, a transparent viscous liquid 'b.p. 216212 "C', and a dark yellow oily liquid b.p. 198 "C respecti~ely.~~?~1,142,164.165 Coordination through oxygen was suggested by the infrared absorptions. 1423 64 However, reaction of antimony(v) chloride with acetamide in aqueous perchloric acid produced an oxonium salt, (MeCONH: )SbCl,, which crystallized. 54 Bismuth(m) cations were found to be triacidic in molten acetamide at 98 "C by titration with acetimidate anions, the first two associations being strong and the third weak,'78 the product Bi(NHCOMe), being only slightly soluble (pK, = 7.9).'79 Bismuth trichloride was also reported to form a solid 1 :4 complex, white in colour with a melting point over 240 0C.142Bismuth triiodide has been reported to give an insoluble solvate with molten acetamide., (vi) Group VZBCompounds.Polarographic reduction of molecular oxygen at 85 "C in molten acetamide containing 1 molar sodium acetate showed two waves. The '73 M. A. Klochko and G. F. Gubskaya, 1:~'.Sekt. Fiz.-Khim. Anal. Inst. Ohshch. Neorg. Khim. Akad. Nauk SSSR, 1956, 27, 393. I74 D. A. Tkalenko, S.A. Kudrya, and A. A. Rudnitskaya, Fiz- Khim. Scoistea Rasplaci. Tcerd. Elektrolitov, 1979. 66."'M. Woldan, Acta D'nic. Lodz Folia Chim., 1985, 5, 105. 176 G. Berchiesi, G. G. Lobbia, V. Bartocci, and G. Vitoli, Thermochim. Acta, 1983, 70, 317. 17' V. P. Basov, Y. A. Karapetyan, and A. D. Krysenko, Zh. Fiz. Khim., 1978, 52, 1753. 17' M. Pournaghi, J. Devynck, and B. Trernillon, Anal. Chim. Acta, 1977, 89, 321. 179 M. Pournaghi, J. Devynck, and B. Trernillon, Anal. Chim. Acta, 1978, 97, 365. 199 The Chemistry of Molten Acetamide and Acetamide Comple.ues first reduction was to peroxide, but some hydrogen peroxide was thought to have thermally decomposed to water and oxygen (i.4. to have disproportionated). The second reduction wave was to water (or hydro~ide).'~~-'~~ The density and viscosity of aqueous solutions of acetamide have been measured,' 84 and density measurements on acetamide solutions of aqueous hydrogen peroxide at 7&75 "C have shown no evidence of compound formation.'85 Sulphur tetrafluoride reacts to form bis(acetimidato)sulphur [(MeCONH),S] but only in the presence of sodium fluoride as a hydrogen fluoride acceptor.'86 Sulphur monochloride forms the same product with acetamide.' 87 This product is soluble in water and has been characterized by mass and infrared spectroscopy.'86 Its melting point (221 "C) distinguishes it from its isomer [(NH,COCH,),S, m.p.162 0C].'87 Hydrogen sulphide with acetamide under 8 500 atmospheres pressure and 125-150 "C, only gave a low yield (10%) of diethyl disulphide together with elemental sulphur (1 90;,), the major part of product being unreacted acetamide (61%).18' Sulphur trioxide however forms 1 :1 and 1 :2 complexes exothermically (described as brown or red-brown, and lemon-yellow 'sticky masses' respective- ly)913'623164 and the infrared absorptions of the former [3 410 (vNH), 1 600 (ijNH), 1380 (ijcH), 1 105 (w), 1 102 (w)] were taken to indicate bonding via carbonyl 0~ygen.l~~However other reactions could have taken place and acetimidato groups may well be present.Certainly conductivity measurements in nitrobenzene solution were considered to arise by proton donation to solvent rnole~ules.~~ Some doubt also arises about the nature of both compounds because, although the analytical results for sulphur and nitrogen were close to theoretical, they were un- stable, could not be recrystallized, and decomposed 'at higher temperatures'.14',' 64 Aqueous solutions with persulphate reacted when irradiated (S,Oi --Sb'i MeCoNH2+ 'CH,CONH, + HSO,).' 89 No complex was found between potassium sulphate and acetamide in a phase diagram study.'" Lithium thiocyanate is soluble in molten acetamide to more than 26 mole% 176 and phase diagrams of acetamide with two other thiocyanates have been determined, that with the sodium salt showing an inflexion (32 molx NaSCN at 52°C) which may be due to a peritectic and would thus indicate compound formation, together with a eutectic (22.5 mol";;, at 15 0C).'91 Potassium thio- cyanate showed only a eutectic (quoted as 26 mol:/; at 25.5 "C 19' and 24.7 mol% ''O K.L. N. Phani and R. Narayan. J. Elwirormril. Clieni.. 1982, 134, 291. K. L. N. Phani and R. Narayan, Proc. In/. Symp. Molten Sdi Clieni. Techol. Is/. 1983. 157. K. L. N. Phani and R. Narayan, J. Eli~r~/rnanal.Cliem.. 1985, 193, 283. IH3K. L. N. Phani and R. Narayan, J. Electrounal. Cliem.. 1985, 189, 135. H. J. Christoffers and G. Kegeles, J. An!. Clwni. Soc,.. 1963. 85, 2562. '13' J. Barlot and S. Marsaule, C. R. Ac,ad. Sci. Puris. 1947, 225, 120. IH6R. D. Peacock and I. N. Rozhkov, J. Cliem. Soc. A. 1968, 107. "'P. Hope and L. A. Wiles, J. Chem. Soc., Suppl. No. 1, 1964, 5679. "'T. L. Cairns, A. W. Larcher, and B. C. McKusick, J.Org. Chem., 1953, 18, 748. M. J. Davies, B. C. Gilbert, C. R. Thomas, and J. Young, J. Clietn. Soc., Perkin Trans. 2. 1985, 1199. I9O A. V. Tolstousov and M. K. Balbaev, Deposited Doc. 1984, VINITI, 4441.''' F. Castellani, G. Berchicsi, F. Pucciarelli. and V. Bartocci, J. Chem. Eng. Dais, 1981, 26, 150. Kerridge at 28.5 "C).' 92 A ternary eutectic phase diagram study (acetamide-NaSCN-succin- imide) has also indicated the 1 :2 perite~tic.''~ Binary eutectics with both sodium and potassium thiocyanate have been the subjects of a number of investigations, longitudinal and shear impedance, density and viscosity measurements having been made with acetamide-sodium thiocyanate and interpreted as suggesting (un- specified) changes in liquid structure around 23 "C; these highly viscous solutions show a strong tendency to superco~l.''~-'~~ The anode deposits from acetamide- potasssium thiocyanate eutectic at high current densities were of insoluble orange- yellow parathiocyanogen [(SCN),] and have been shown to be photochemically active.However at low current densities, the oxidation products were soluble and suggested to be thiocyanogen and the trithiocyanogen anion [(SCN), and (SCN);] though positive identification was la~king.'~'~'~~ Tellurium tetrachloride forms a 1:2 adduct by direct reaction in boiling benzene, or carbon tetrachloride, which was described as a 'greenish-black semi-solid', 142 whose infrared spectrum suggests coordination by oxygen,' 1*200 and five co- ordinate tellurium (the far infrared 240 cm-' band is cited as diagnostic ,O0).Since its molar conductivity in nitromethane was nearly equal to that of a 1: 1 electrolyte an equilibrium was postulated. TeCl, + 2MeCONH2e[TeC13(NH2COMe)2]+ + C1-(14) (vii) Group VZIB Compounds. The halide anions are, as expected, stable in molten acetamide. Hydrogen fluoride solutions have been titrated and a dissociation constant reported', as has a 2: 1 compound.201 The complexes of hydrogen chloride were discussed earlier (Section 3A). Of the three elements studied, acetamide has been fluorinated during electrolysis of molten potassium hydrogen fluoride at 120 "C. A variety of gaseous products were formed (NF,, CF,, C,F,, COF,, N,O, CO,, N,, 0,).202,203The specific conductance of bromine in acetamide has been measured,204 and bromine is stated to interact with acetamide to form a 1:l complex (as indicated by density, 19' L.S. Bleshinskaya, K. S. Suliamankulov, and M. D. Davranov, Deposited Doc. 1983, VINITI, 121. lY3G. G. Lobbia and A. Amico, Thermochim. Acta, 1985, 87, 257. 194 G. Berchiesi, G. Vitali, P. Passamonti, and R. Plowiec, J. Chem. SOC.,Farada-v Trans. 2, 1983,79, 1257. Iy5 G. Berchiesi, M. A. Berchiesi, R. Plowiec, and F. Castellani, Calorim. Anal. Therm., 1983, 14, 174. Iq6 P. Passamonti, A. Amico, and G. Berchiesi, J. Chem. Soc., Faraday Trans. 2, 1985,81, 217. "' A. Amico, G. Berchiesi, C. Cametti, and A. Di Biosio, J. Chem. Soc., Faraday Trans. 2, 1987,83, 619. IqR F. Pucciarelli, V.Bartocci, F. Castellani, M. Gusteri, P. Cescon, and M. Bragadin, Ann. Chim. (Rome), 1983, 73. 697. 199 M. Bragadin, G. Scarponi, G. Capadoglio, F. Ossola, V. Bartocci, and F. Pucciarelli, Mol. Crj9st. Liq. Cryst., 1985, 121, 345."" K. C. Malhotra and K. K. Paul, Curr. Sci., 1969, 38, 266. '"I J. C. Balle, Fr. Pat. 1370827, 1964. 'O' A. Tasaka, H. Sakaguchi, R. Aki, H. Ihara, K. Saka, and T. Yamamoto, J. Fluorine Chem., 1985, 27, 23. '03 T. Sakaguchi and T. Yamamoto. Proc. In[. Sym. Molten Salt Chem. Technol. Is[, 1983, 305. 204 V. A. Plotnikov and S. I. Yakubson. J. Gen. Chem. USSR, 1935, 5. 1337. 201 The Chemistry qf Molten Acetamide and Acetamide Comple.xes viscosity, and conductivity measurements in nitromethane solution '05) which from galvanic cell studies was considered to ionize 206-'09 Br, + MeCONH, MeCONH,.Br, [MeCONH,Br]+ + Br-(15) +though other ions ([Br(MeCONH,),] and Br;) have also been postulated because the transference number of the cation is quite E.m.f.data indicate surprisingly that iodine in molten acetamide is less stable than bromine, the molar conductivity decreasing with dilution and increasing temperat~re,~'~and the infrared spectrum again suggesting coordination through oxygen.' lo Recent potentiometric and voltametric studies showed two waves in the oxidation of iodide (to 13 and 1') together with a maximum (ascribed to I+). In the absence of iodide, iodine in acetamide was again unstable and slowly reduced to iodide. However, in the presence of excess chloride, iodide was oxidized in one step to iodine dichloride (ICl;).2 l1 Iodine monochloride forms an adduct, suggested to be 1 :2 by viscosity measure- ments at 2545 "C, which has a very high conductance. The ionization ICl + 2MeCONH, ICl(MeCONH,), T=+ (MeCONH,I)' + C1-(16) is suggested, since iodine was found to be discharged at the cathode on electro- lysis.*l2 The compound possibly dissociates (to 1 : 1) in nitrobenzene ~olution,~' and is highly active in halogenation and oxidation reactions.' Hydrogen iodide is soluble in acetamide '' and potassium iodide almost completely dissociated.21 Of the oxyacids, perchloric acid has been studied (see Section 3A) and lithium perchlorate solutions, all apparently stable near 80 "C and hypobromous 959176 acid has been found to react with acetamide, initially to form a 1:1 adduct, which then dehydrated to acetimidatobromide (MeCONHBr)." Exchange of "0 with alkali metal bromates was almost completely absent in molten acetamide at 170 "C over 25 hours, but exchange of hydrogen between acetamide and water was comparatively rapid as the latter was added.Acetamide was found to have no effect on H,180 exchange when added to aqueous solutions of sodium and potassium bromates, as with these cations no ion pairs were considered to be formed. But it considerably reduced exchange with the lithium, magnesium, 'OS E. Y. Gorenbein and A. E. Gorenbein. Zli. Nrorg. Kliini.. 1968, 13. 161. '06 E. Y. Gorenbein and A. E. Gorenbein.USSR Pat. 182777, 1966. 207 A. E. Gorenbein and E. Y. Gorenbein. Eleclrokliiniiyo. 1967, 3, 628. '08 E. Y. Gorenbein and A. E. Gorenbein. Z17. Oh.did7. Khim., 1967. 37. 969. '09 E. Y. Gorenbein, A. E. Gorenbein. and A. A. Fominskaya. Zh. Ohslid7. Kliirn.. 1968, 38, 960. M. M. Gerbier and J. Gerbier. C. R. Am/. S(,i.Pnris. 1966, 263B. 1057. '11 V. Bartocci, F. Pucciarelli. and M. Gusteri, Ann. Chin?.(Rome), 1984. 74. 239. 2'2 Y. A. Fialkov and I. D. Mazyka, Z17. Ohslic,li.Kliiui.. 1950. 20. 385. Y. A. Fialkov and 1. D. Mazyka. Zli. 0hslrc.h. Kliim.. 1948. 18. 802. 2'4 Y. A. Fialkov, IT. Aknd. Nnuk SSSR Oldel Kliini. Nmk, 1954. 972. K. G. Khanapin, B. P. Beremzhanov. N. N. Narakhmetov. and R. S. Erkasov, Sh. Rah. Khini. KO:. Unit-..1973, 527.'16 G. Bruni and A. Manuelli, Z. E/ektroc~lirwi.,1904. 10, 601.'"E. Boismenu. Atin. Cliitti.. 1918. 9. 144. 202 Kerridge cadmium, zinc, copper, and nickel bromates, as with these salts the cations are thought to form ion pairs with bromate, and to have a common hydration sheath, part of which acetamide replaces.' l8-"' Astatine in oxidation states I and 111 (as AtX, where X = HSO,, NO,, ClO,, or MeCOO-, and as At3') does not seem to interact with acetamide (solid acetamide was contacted with n-heptane containing astatine, without significant absorption), although with other ligands (all insoluble in n-heptane) anionic or cationic com- plexes were formed (AtX, or AtL; with At').''' C. Transition Metals.-(i) Scandium Group. The interactions of compounds of these metals with acetamide has been little studied, as might be expected.However, scandium trichloride has been reported to form a 1:4 complex with acetamide, as pink crystals (m.p. 135 "C). This complex has an infrared absorption (quoted as 3 230, 3 090, and 1 660 cm-') which was 'considered to indicate co- ordination through nitrogen', but although the N-H stretching frequencies are slightly lowered, and therefore not in accord with oxygen coordination, that at 1 660 cm-' (and assigned to a C-C stretch) is of a similar energy to many other amide I band absorptions (and generally considered to be largely due to C=O stretching) for oxygen bonded complexes. Part of the explanation may indeed be due to instrumental effects, since the absorptions quoted for acetamide itself (3 290, 3 138, and 1 640 cm-')222 are in fact much lower than those usually found (<$ Table 1).Yttrium trichloride is also the subject of one report, a study of the ternary aqueous system (YC1,-MeCONH,-H,O) at 30 "C indicating one compound (YCl,~4MeCONH2~5H20)but no compounds were found in analogous studies with yttrium(rr1) ~ulphate.~~~ Reactions with lanthanum and lanthanide compounds are dealt with in a later sect ion. (ii) Titanium Group. The reaction of titanium tetrachloride was initially reported in 19343 and later found to form a 1: 1 complex (yellow, m.p. 133.4 "C) from methylene chloride solution, with an infrared spectrum indicating coordination through oxygen,224 but a 1 :1 complex obtained by direct reaction (10 hours at 60-70 "C) has been described as red with a melting point of 104 "C, though with similar infrared absorptions.22s A 1:2 complex has been prepared by reaction in benzene or carbon tetrachloride solution and was found to form conducting solutions in molten acetamide.It was considered to have a six coordinate octa- l'' T. S. Kuratova, M. D. Tereshkevich, 0.K. Skarre, and A. N. Baturin, Zh. Fiz. Khim., 1964, 38. 1535. lLYT. S. Kuratova. M. D. Tereshkevich, E. E. Golteuzen, E. Y. Pozhidaeva, and 0. K. Skarre. Zh. Fiz. Kliini., 1965. 39. 2365. 220 M. D. Tereshkevich and E. Y. Pozhidaeva, Zh. Fiz Kliim., 1966, 40, 27. 12' G. W. M. Visser and E. L. Diemer, Rndiochem. Actn, 1983, 33, 145. 12' N.L. Firsova, Y. V. Kolodyashni, and 0.A. Osipov. Zh. Ohshch. Klzim., 1969, 39, 2151. 223 G. A. Ashimkulova, K. Nogoev, and K. Suliamankulov, Zh. Neorg. Kliim., 1974, 19, 2588. 12' D. Schwartz and R. Hey, J. Itiorg. Nucl. Chem., 1967, 29, 1384. 225 N. Yoshino and T. Yoshino. Kogj.0 Kagaku Zosshi,1968, 71, 1025. The Chemistrj3 qf Molten Acrtumide umi Ac~tumide Comp1e.ue.y hedral structure with coordination of acetamide through oxygen. This compound on titration with basic solutions (sodium acetimidate, quinoline, or x-picoline in molten acetamide) gave 1 : 1 and 1 :2 endpoints illustrating the acidic nature of such complexes."*166 Solutions of titanium tetrachloride have been used for electro- plating,226.227 and solid titanium disulphide is reported to form intercalation compounds reversibly with acetamide.228 Zirconium tetrachloride was reported to form a 1 :4 complex as a 'gummy solid' with a melting point of 174.6 "C.No structural details were obtained but the analogous formamide complex was claimed to have six-coordinate zirconium with coordination through oxygen and two additional amide molecules being attached through hydrogen bonds.229 Zirconium oxydichloride however formed a 1 :2 complex on reaction in inert solvents, with the infrared spectrum again indicating coordination through oxygen. However the complex was also reported to have an extraordinary melting point of'over 800 0C',230which was elucidated in some later thermogravimetric results which stated weight loss began at 160 OC, reached a maximum rate of loss at 200 "C and dropped to zero from 41 5 "C 23 ' (i.e.constant weight) and thus suggesting complete loss of acetamide from this temperature. Isopropyl zirconium chlorides have been refluxed with acetamide in hexane, forming 1 :1 complexes with monoisopropyl zirconium trichloride and with di- isopropyl zirconium dichloride. Triisopropyl zirconium chloride also gave the first product [Me,CH)ZrCl,*MeCONH,] by disproportionation, but no adduct was formed with tetraisopropyl zirconium.232 However, after refluxing the latter reactant with acetamide in anhydrous benzene highly polymeric products with up to four acetamides per zirconium were reported.233 These polymers had con-siderable stability for they were found not to sublime under vacuum up to 250 OC, nor to decompose below 300 0C.234 No reactions with hafnium compounds are reported in the literature.(iii) Vunctdium Group. A vanadyl(1v) sulphate complex with two acetamides has been reported, the infrared spectrum indicating bonding of acetamide through oxygen and with bidentate ~ulphate.~~' Two vanadyl(v) compounds (VOP0,- 2H20 and VOAs0,-3H20) formed intercalation compounds, the acetamide forming hydrogen bonds through the amide group.236 22b N. K. Tumanova, N. M. Sarnavikii. M. U. Prikhodko, A. V. Chetverikov. L. V. Bogdanovitch. and I. M. Muk ha, Ohrazr.~j~Trrcrrrrij~jZnuki. I98 I, 40, 1 12. "' N. K. Tumanova. N. M. Sarnavikii, L. V. Rogdanovitch. V. N. Beldi, and G. N. Novitskaya, Ukr.Khini. Zh.. 1983. 49. 266. 22n A. Weiss and R. Ruthardt. Z. Na/ur/i)r.sc~li.,Toil B. 1969. 24, 355. 229 A. Clearfield and E. J. Malkiewich. J. Itzorg. Nid. Chmi., 1963. 25. 237. 230 R. C. Paul, S. L. Chadha. and S. K. Vasisht. J. Le.c.r-ConinionMe/., 1968. 16, 288. 23' R. C. Paul, A. K. Moudgil, S. L. Chadha, and S. K. Vasisht, Indion J. Cheni., 1970. 8, 1017. 232 Y. Nario and T. Yoshino, Kogjw KLI~U~UZasslii, 1969, 72. 2293. 2.13 K. R. Nadar. A. K. Solanki, and A. M. Bhandari, Indim J. Cheni., Src. A. 1980, 69. 234 K. R. Nadar, A. K. Solanki. and A. M. Bhandari, Z.Anorg. Clicwi., 1979. 449, 187.'-''Y. S. Usmankhodzhaeva. 0.F. Khodzhaev, N. A. Parpiev, K. Khodzhaeva. and Z. M. Musaev, Uzh. Khini. Zli.. 1983. 3. '3h M. M. Lora.L.. M. Real. A. J. Lopez. S. B. Gomei. and A. R. Garcia. Mrrtcv. Rex Bull., 1980. 21, 13. Kerridge Several niobium and tantalum pentahalide adducts have been claimed (MX,-n- MeCONH,, where n = 1,2, or 3 with NbCl,, n = 1 or 3 with NbBr, and TaBr,) with conductivity and dipole moment measurements indicating bonding through Niobium oxychloride formed a number of complexes (NbOCl,*n- MeCONH, where n = 24, 3, 4, 6, and 7) which were all 1: 1 electrolytes with oxygen bonded acetamide. In ethanol another complex was formed CNbOC1,- (OEt)(MeCONH,)4(EtOH)].238 The niobium and tantalum disulphides form many intercalation compounds (up to 2MS,-MeCONH,) when heated with acetamide in the absence of air.239 (iv) Chromium Group. Of chromium(v1) compounds, chromates, and chromium tri- oxide were early reported to be 'slowly reduced', though no further details were given.' Much later chromyl chloride was reacted with acetamide in carbon tetra- chloride solution to give a 'brown sticky mass' which nevertheless yielded sur- prisingly good analytical results corresponding to a 1:2 adduct, the infrared spectrum showing band shifts indicative of coordination by oxygen.240 Chromium(rI1) perchlorate was found to form a 1:6 complex, whose visible- ultraviolet spectrum in nitrobenzene solution [absorption bands at 16 450 cm-', molar extinction coefficient 53.0 1 mol-' cm-', assigned to 4A2, -+ 4T2gtransition; and 22 940 cm-' (41.3) to 4A2g -+ 4T,g(F)] suggested octahedral coordination by the acetamide oxygen^.^^' (These values of the spectroscopic parameters have been much quoted in data compilation^,^^^,^^^ and indicate that the ligand field of acetamide is slightly less than that of water but greater than those of urea or DMSO, the D,values for these-four ligands being 1645, 1740, 1 600, and 1 577 cm-' respectively). When an anion more coordinating than perchlorate was present less acetamide was found to be coordinated, for example with thiocyanate only two acetamides per chromium [KCr(SCN),(MeCONH,),, m.p.208 OC (d), was separated from a solution of KCr(S04),*12H,0 with excess KSCN and MeCONH,]. The molar conductivity of this compound (139 in acetone) indicates the presence of two ions, while infrared and electronic spectroscopy showed reversal (at least in part) of the thiocyanate coordination from nitrogen, which is usual with chromium(m), to sulphur bonding.The absorptions of coordinated acetamide quoted indicated bonding through oxygen.244 Acetamide forms an aqueous insoluble 'reineckate' in 0.5 M hydrochloric acid (of composition MeCONH: [Cr(NH,),(NCS),] -) which decomposes at 1368 "C. Most nitrogen bases form similar salts which are used as a basis for their forensic 237 J. R. Masagner Fernandez and M. R. Berrnejo, Ann. Quim., 1973, 69, 1099. 23R S. M. Sinitsyna and N. A. Razorenova, Koord. Khim., 1985, 11, 617. 239 R. Schoellhorn and A. Weiss, Z. Naturforsch., Teil B, 1973, 28, 172. 240 R. C. Paul, 0.Khosla, and R. Dev, Indian J. Chem., 1964, 7, 1254. 24' R.S. Drago, D. W. Meek, M. D. Joesten, and L. LaRoche, Inorg. Chem., 1963, 2, 124. 242 V. Gutrnann and G. Melcher, Monatsh. Chem., 1972, 103, 624. 243 M. Ban and J. Csaszar, Acta Chim. Acad. Sci. Hung., 1967,54, 133. 244 G. Contreras and R. Schmidt, J. Inorg. Nucl. Chem., 1970, 32, 127. 205 The Chemistry of Molten Acetamide and Acetamide Complexes identification and Acetamide has also been substituted for cyclo- pentadiene in chrom~cene.~~~,~~~ Chromium(rr1) chloride solutions in molten acetamide have been used for chromium plating, solutions of chromium(iI1) acetate having a lower con-du~tivity.,~~ Molybdenum compounds appear not to have been studied but for the report of a molybdenyl(v1) complex formed by ‘lengthy ageing’, presumably hydrolysis, of an acetonitrile solution [MoCl,(NO)(MeCN), in MeCN -, MoO,Cl,-(MeCONH,),].249 Tungsten hexachloride in ether solution is reported to form a 1:6 complex which is greenish-yellow in colour, is insoluble in common organic solvents, and decomposes in moist air.’ 50 Tungsten oxyfluoride in acetonitrile solution however gave a 1: 1 complex, considered from its infrared and Raman spectra to be trans octahedral and oxygen bonded.251 In contrast tungsten(v1) oxy- chloride in carbon disulphide, when mixed with acetamide dissolved in benzene was reported to give a 1:4 complex of ‘dirty grey’ colour and of ‘considerable stability’. Chlorides were considered to be displaced by oxygen-bonded acetamides ([WO(MeCONH2)4]C14).2 (v) Manganese Group.Manganese(1r) chloride was early reported to be highly dissociated in acetamide solution including the original water of hydratioq2 ’ and later a 1 :2 complex was observed to form on direct reaction in refluxing benzene or carbon tetrachloride, the light pink crystals (m.p. 188 “C) 142 having infrared and Raman spectra indicative of oxygen coordination., 53 The heat of formation has been calculated from solution data.254 Phase diagram studies on aqueous acetamide solutions indicated a number of complexes with manganese(1i) halides (MnC1,-2MeCONH2-H2O; MnC1,- 4MeCONH,*H,O; MnBr,.4MeCONH,-H2O; and MnI,-4MeCONH,),255 and with manganese(i1) acetate [Mn(O,CMe),~MeCONH,~H,O; Mn(OOCMe),* 2MeCONH,] where X-ray diffraction and infrared spectra showed the complexes to be octahedral with oxygen-bonded acetamide and bidentate acetate.256 However with manganese(1r) nitrate an anhydrous 1:6 complex [Mn(NO,),* 24s L.Kum-Tat, Anal. Chim. Acta, 1961, 24, 397. 246 L. Benes, J. Kalousova, and J. Votinsky, J. Organomet. Chem., 1985, 290, 147. 247 J. Kalousova, J. Votinsky, and L. Benes, Proc. 10th Conf: Coord. Chem., 1985, 189. 248 A. L. Hanson, D. Frokjer, and D. Mitchell, Metal Finishing, 1951, 48. 249 V. S. Sergienko, N. A. Ovchinnikova, M. A. Porai-Koshits, M. A. GIushkova, Koord. Khim., 1986, 12, 1650. ”O S. Prasad and K. S. R. Krishnaiah, J. Indian Chem. SOC.,1961, 38, 177. 251 Y. A. Buskev, A. Y. Tsivadze, Y. Y. Kharitonov, Y. V. Kokunov. and N. P. Gustyakova, Dokl. Akad. Nauk SSSR, 1977, 236, 1367.252 S. Prasad and K. S. Krishnaiah, J. Indian Chem. SOC.,1961, 38, 757. 2s3 A. Y. Tsivadze, Y. Y. Kharitonev, G. V. Tsintsadze, A. N. Smirnov, and M. N. Tevzadze, Zh. Neorg. Khim., 1974, 19, 3321. 2s4 M. S. Barvinok and L. V. Mashkov, Zh. Neorg. Khim., 1985, 30, 2972.”’B. Imanakunov, S. Baicholova, and K. Alymkulova, Muter, Nauchn. Konf. Posuyashch lOO(Sto)Letiyu Period. Zakona D. I. Mendeleeva, 1969, 143. 256 0.F. Khadzhaev, T. A. Azizov, and A. N. Parpiev, Koord. Khim., 1977. 3, 1495. Kerridge 6MeCONHJ forming triclinic crystals has been prepared,257 as have two hydrated complexes [Mn(N0,),-6MeCONH2-2H20 and Mn(NO,),-4MeCONH2-2H20] which were again oxygen bonded to manganese.258 No compounds of technetium or rhenium containing acetamide appear to have been prepared.(vi) Iron Group. An iron(][) chloride complex (FeC12*2MeCONH2) was made by the unusual route of reaction of an iron carbonyl, Fe,(CO),,, with acetamide in chloroform [possibly a Fe(CO),(MeCONH,), intermediate was first formed between the carbonyl and acetamide, which then reacted further with chloroform] which was initially considered to be monomeric and tetrahedral.259 Later intensive Mossbauer spectroscopy and magnetic measurements showed the structure to be octahedral and polymeric, perhaps with chloride bridges.260,261 There was agree- ment however that the acetamide was bonded through oxygen. With iron(m) chloride a 1:3 complex has been obtained by two groups. The compound, prepared in benzene or carbon tetrachloride, was yellow-brown (m.p.79 "C),' 42 infrared and Raman spectroscopy showing oxygen bonding and hence presumably monomeric octahedral molecules.2 -where N 0 = NHCOMe (1) The only report on ruthenium chemistry described the reaction of a dinuclear ruthenium acetate chloride, Ru2(00CMe),C1, with molten acetamide, to form a product also with the ruthenium atoms in formally different oxidation states [i.e. Ru,(NHCOMe),Cl] which the authors stated might in the solid consist of dimeric units (1) with chloride bridges forming zigzag chains. Electrochemistry in several non-aqueous solvents brought about oxidation of the initially (11, III) units to (111, 111) units [Ru2(NHCOMe)J2 or reduction to (II,II) units [Ru,(NHCOMe),] and+ possibly to (I, 11) units [Ru,(NHCOMe),] -.262 Unfortunately vibrational spectro- scopy was not undertaken and so no further information is available on this potentially most interesting family of complexes. Of the osmium compounds reported, one was a carbonyl hydride cluster with an acetimidate group [0s3(C0), ,H(NHCOMe)], made by the reaction of acetamide with the osmium carbonyl acetonitrile complex in refluxing cyclohexane and characterized by elemental analysis and mass spectroscopy.Infrared spectroscopy showed, besides nine bands assigned to carbonyl absorptions, one band at 1 576 cm-', in accord with nitrogen-bonded acetimidate. The suggestion was made that 257 M. Nordelli and L. Coghi, Ricerca Sci.,1959, 29, 134.258 V. T. Orlova, V. I. Kosterina, and I. N. Lepeshkov, Zh. Neorg. Kliim., 1986, 31, 1854. 259 P. P. Singh and R. Rivest, Can. J. Chem., 1968, 46, 1773.''(' T. Birchall, Can. J. Chem., 1969. 47, 1351."' T. Birchall and M. F. Morris, Can. J. Chem., 1972, 50, 201. 262 M. Y. Chavan. F. N. Feldman, Y. Q. Lin, J. L. Bear. and K. M. Kadish, Inorg. Cliem., 1984 23,2373. 207 The Chemistry of Molten Acetamide and Acetamide Comp1e.ue.Y CNS co Ncs' Me (2) the acetimidate was also coordinated through oxygen and thus bridging between two osmiums of the cluster. However, again proof that the acetimidate was bi- dentate and any further vibrational data are la~king.~~~,~~~ The others all con- tained an anionic carbonyl hydride cluster [(Os,H(CO), ,(p3-N(CO)CH3)- with Au(PPh,)+, Cu(PPh,)+, or N(PPh,),f] in which the crystal structure of the first showed a N(C0)Me group bridged to two edges of the Os, butterfly by nitrogen.265 (vii) Cobalt Group.An early report indicated that cobalt(I1) chloride dissolved in molten acetamide and was highly dissociated, but more recently a 1 :6 complex was reported (CoC1,-6MeCONH2), made by direct reaction in benzene or carbon tetrachloride, which was pink, and presumably octahedral, with a melting point of 58 OC.14, Other complexes have been suggested from interfacial tension measure- ments on aqueous acetamide solutions (CoC1,.2MeCONH2; and CoCI,* MeCONH,) 135 but obviously water could also be coordinated. The heat of formation has been calculated for the 1 :2 complex.254 Direct reaction of cobalt(I1) perchlorate in acetonelether solution also gave a 1 :6 complex which was studied by X-ray powder diffraction, infrared, and reflectance spectroscopy (u.v./visible bands at 19 000, 15 200, and 7 800 cm-').Octahedral coordination by acetamide oxygens was inferred and acetamide was placed in the spectrochemical and nephelauxetic series as follows (MeCONH, > MeC0,H > CO(NH,), and MeCONH, > CO(NH,), > MeC0,H respectively.266 The same 1:6 complex was deduced from phase diagram studies of the ternary system Co(ClO,),-MeCONH,-H,O and it was found to lose some acetamide above 143 OC, to melt at 198 OC, and to decompose at 278 0C.267 These studies also indicated another mixed complex Co(CIO,),~4MeCONH2~2H,O 267 which was presumably also octahedral with oxygen coordination of acetamide and water- although a 1 :2: 1 complex with cobalt(I1) chloride (m.p.62 "C) was claimed in 1886.268 An analogous mixed complex was found in the ternary system with 2h3 B. F. Johnson, J. Lewis, T. I. Odiaka, and P. R. Raithby, J. Organornet. Chem., 1981, 216, C56. 264 T. 1. Odiaka, J. Organomet. Chem.. 1985, 284. 95. 265 J. Puga, R. A. Sanchez-Delgado, J. Asconio, and D. Brag, J. Chem. Soc.. Chern. Commun.. 1986, 1631. 266 P. W. N. M. Van Leeuwen and W. L. Groenewald. Rev. Trm. Chim. Paj.s-Bas. 1968, 87, 86. 267 Y. A. Goryunov, Uch. Zap. Yarosl. Gos. Pedagog. Inst.. 1976, 154, 30. 268 G. Andre, Jahresher. Fortschr. Chem., 1886. 1303. 208 Kerridge cobalt(r1) nitrate at 25 "C [Co(N0,),~4MeCONH2~2H,0,density 1.61, m.p.89 "C, loss of water from 112 OC, explosive decomposition 226-253 "C to water-insoluble 'powder', presumably cobalt oxides], together with another complex [Co(NO,),- 6MeCONH2-2H2O, density 1.51, m.p. 96 OC, water loss from 129 OC, decomposition 229-252 "C] apparently having an anomalously high coordination, unless some acetamide was attached by hydrogen bonds.269-272 An anhydrous 1:6 complex [Co(NO3),-6MeCONH,] as triclinic crystals has also been reported., 57 With cobalt(1r) thiocyanate a different stoicheiometry was found [Co(NCS),.4MeCONH2], where again infrared spectroscopy data suggested bonding of acetamide through oxygen and of thiocyanate through nitrogen.253 Less coordination by acetamide was found with another cobalt(I1) thiocyanate complex (2) where tetrahedral geometry was postulated on the basis of infrared measurements which indicated coordination of acetamide through oxygen, as well as bridging and sulphur bonded thiocyanate.Thermogravimetric analysis showed a three-step decomposition (equation 17). 500 600°CCo(NCS),(NCSAg),(MeCONH,), Co(NCS),(NCSAg), -1 Ooo-1 150°CCo(CN), + Ag,S *Co + Ag (ref. 273) (17) Similar compounds with selenocyanate and thallium thiocyanate groups (formulated as M[Ag(SCN)(SeCN)],-2MeCONH2 and (MeCONH,),M(NCS),- (NCSTI), where M = Co, Ni, or Cu) have also been reported recently.2743275 Another tetrahedral complex ([Co(MeCONH,),Cl]Cl) has been reported recently which again has oxygen-bonded a~etamide.,~~ The aqueous ternary phase diagram with cobalt(l1) formate indicated a further different stoicheiometric ratio [~CO(O,CH),-M~CONH,~~H~O].~ Cobalt(n1) complexes are much rarer with acetamide, only three being reported, even though cobalt(II1) solutions [of Co(NH,),Cl, and K,Co(CN),] were early found to be deep blue.All reported complexes, however, are of considerable interest, the first (trans Na[CoSO,(DH),MeCONH, where DH, = dimethylgly-oxime) was unusual by having acetamide in an anionic complex and was prepared as orange rhombic crystals by treating the corresponding aquo complex with acetamide in aqueous alcohol. The coordination of the acetamide was weak how- ever, since hydrolysis occurred immediately in dilute aqueous s~lution.~ 78 Thermo-269 A.D. Dzhunusov, B. I. Imanakunov, M. K. Kydynov, and A. s. Karnaukhov. Uch. Zap. Yarosl. Gos. Pedagog. Insr., 1969, 66, 181.*'' A. D. Dzhunusov, B. I. Imanakunov, M. K. Kydynov, I. G. Druzhinin, and A. S. Karnaukhov, Zh. Neorg. Khim., 1970, 15, 532. 271 A. D. Dzhunusov, B. I. Imanakunov, M. K. Kydynov, and A. S. Karnaukhov, Khim. Kompleksn. Soedin. Redk. Sopur.stt.u.~nsiiL.liikh.Elem., 1970, 127. 272 A. D. Dzhunusov, B. I. Imanakunov, M. K. Kydynov,and P.T.Yun, I:u. Akad. Nauk Kirg. SSR, 1967,67. 273 S. B. Sharma, T. N. Ojha, S. A. Khan, and M. K. Singh, J. Indian Chem. Soc., 1984, 61, 476. 274 S. B. Sharma, M. K. Singh, and V. P. Singh, Indian J. Chem., 1986, 25A,335. 27s S.B. Sharma, V. P. Singh, and M. K. Singh, J. Indian Chem. Soc., 1985, 62, 721. 276 G. Narain and P. R. Shukla, J. Inst. Chem. (India), 1985, 57, 231. 277 G. K. Distanov and B. Dzhashakneva, Tr. Kirg. Unit.. Ser. Khim. Nauk, 1972, 21. 278 G. P. Syrtsova and N. N. Chaban, Zh. Neorg. Khim., 1971. 16. 2471. 209 The Chemistry of Molten Acetamide and Acetamide Complexes 5+ N C Me 4+ Scheme 1 gravimetry was reported later., 79 The second acetamide complex was unique in apparently having isomeric forms. It was formed from an acetimidate complex of cobalt(Ir1) ([NH,),CoNHCOMe](CIO,),) which was itself made by heating the aquo pentammino cobalt(II1) complex with acetamide in trimethyl phosphate solution, and was shown to be bonded to nitrogen by X-ray single crystal diffraction280 and by proton magnetic resonance.281 Treatment of the acetamidate complex with perchloric acid gave an acetamide complex ([(NH,),CoNH,-COMe](CIO,),) which as would be expected had very different ultraviolet absorptions [29 300 cm-’ (57.5 1 mol-’ cm-’) and 21 000 (61.5) as compared to 40 000 (2 400), 28 500 (83.5) and 20 600 (71.1) for the acetimidato complex] 281 and was highly acidic (pK, (25 “C) 2.16 compared to 14 at 20 “C for acetamide).282 Interestingly, this orange nitrogen-bonded complex underwent conversion in the solid into the pink oxygen-bonded complex.281.283 The third complex was of binuclear cobalt(rr1) with a bridging amide anion (NH,) which was also bridged by an acetimidate anion coordinated through both nitrogen and oxygen, which had been formed by hydrolysis of an acetonitrile complex.284 The reaction was thought to proceed as shown in Scheme 1.2’9 N. N. Chaban, G. P. Syrtsova, G. B. Seifer, and N. M. Thu, Koord. Khini., 1977. 3, 582. 280 M. L. Schneider. G. Ferguson, and R. J. Balahura, Cm. J. Chem., 1973, 51, 2180. 281 R. J. Balahura and L. Hutley, Can. J. Ckeni., 1973. 51, 3712. 2n2 R. J. Balahura, Can. J. Cheni., 1974, 52, 1762. 283 R. J. Balahura and R. B. Jordan, J. Am. Clieni. Soc., 1970, 92. 1533. 2*4 N. J. Curtis. K. S. Hagen. and A. M. Sargeson, J. Clirm. Soc.. Clieni. Coninirtn.. 1984, 23, 1571. 210 Kerridge Rhodium(II1) chloride hydrate did not react with acetamide at room temperature and on heating formed an ammine complex ([Rh(NH,),Cl]Cl,) 28s but binuclear rhodium(I1) compounds have been the subject of considerable study.Dirhodium tetraacetate, for example, undergoes a stepwise reaction in molten acetamide at 120 “C to form a series of acetimidate complexes [Rh2(0,CMe),(NHCOMe),-,, where n = &-4]which have been oxidized electrochemically to the 11,111 and to the 111,111 binuclear complexes. The electrochemical potentials and electronic absorption spectra change uniformly both with the number of acetamide ligands and with the non-aqueous solvent used.286-288 0ne such complex [Rh,(NH- COMe),2H20] has been the subject of X-ray diffraction with rhodiums 2.415 8, apart and bridged by four acetimidate ligands, each rhodium having a pair of nitrogen donor atoms in a cis arrangement.289 Such complexes form carbon monoxide ad duct^.,^' A crystal structure determination on a 11,111 complex [Rh2(NHCOMe),(H,0),C104] has given essentially the same geometry but with a significantly reduced distance (2.399 A) between equivalent rhodium~.”~ The formation of the tetraacetimidate at 160 “C and its electrooxidation to rhodium(II1) has been confirmed and further complexes formed of the type Rh,(NHCOMe),L, (where L = pyr, H20, DMSO, PPh,),292,293 and [Rh,- (NHCOMe),L,]NO,~H,O (where L = theophylline), the crystal structure of the latter indicating stabilization by hydrogen bonding between the acetamide and oxygen of the base.294 Rather similar complexes were formed on prolonged heating of dirhodium tetraformate with aqueous acetamide, but at room temperature a different product resulted, Rh,(0,CH),*2MeCONH2, which presumably now had acetamide as the axial ligands.Infrared spectroscopy again showed coordination ‘3,through oxygen., A carbonyl complex [(pCl),(Rh(CO)Cl-CH ,CONH,} ,C1] has also been reported.296 No compounds of iridium with acetamide seem to have been prepared. (viii) Nickel Group. Reactions involving acetamide and compounds of the elements of this group of the Periodic Table have been the most widely studied. Most of the attention has been directed towards platinum compounds, but a considerable amount has involved nickel. An early report stated that nickel(1r) chloride was 285 R. N. Shchelokov, A.G. Maiorova, G. N. Kuznetsova, I. R. Golovaneva. and 0.N. Evstafeva. Zh. Neorg. Khim., 1984, 29, 1335. 286 T. P. Zhu, M. Q. Ahsan, T. Malinski, K. M. Kadish, and J. L. Bear, Inorg. Ciiem., 1984, 23, 2. ”’ M. Y. Chavan. T. P. Zhu, X. Q. Lin. M. Q. Ahsan. J. L. Bear, and K. M. Kadish. Inorg. Chem.,1984,23, 4538. M. Q. Ahsan, Diss. .4bsrr. Int. B, 1984, 45, 1461. 289 M. Q. Ahsan. I. Bernal, and J. L. Bear, Inorg. Ciiem.. 1986, 25. 260. 290 M. Y. Chavan. M. Q. Ahsan, R. S. Lifsey. J. L. Bear, and K. M. Kadish, Inorg. Chem.. 1986, 25, 3218. 29’ I. B. Baranovskii. M. A. Golubnichaya, L. M. Dikareva. A. J. Rotov, R. N. Shchelokov, and M. A. Porai- Koshits, 2%. Neorg. Khini., 1986, 31, 2876. 292 M. Q. Ahsan. 1. Bernal, and J. L. Bear. Inorg. Chitn.Acro, 1986, 115, 135.293 I. B. Baranovskii and R. E. Sevastyanova. Zh. Neorg. Khim., 1984, 29, 1786. 294 K. Aoki. M. Hoshiro, T. Okada, H. Yamazaki, and H. Sekizawa, J. Ciiem. Soc., Ciiem. Commun., 1986. 314. 19’ V. N. Shafranskii and T. A. Malkova. Zh. Obshcii. Khim., 1975, 45, 1065. 296 Y. N. Kukushkin. V. K. Krylov. and M. Y. Romanov, Zh. Obshch. Kiiini.. 1983, 53. 867. 211 The Chemistry of Molten Acetumide and Acetamide Comple.ues soluble in molten acetamide and was highly dissociated, the water originally co- ordinated to nickel being displaced.2 l6 Much later the electronic spectroscopy of these pale green solutions showed octahedral coordination of nickel(i1) at 85 "C [the absorptions being 7 800(44), 12 800(3.9), 14 200(0.9), and 23 700(10) cm-'1 with an increasing shift to tetrahedral coordination (and a bluer solution) as the temperature was increased to 172 "C.The low-temperature spectrum was stated to be very like that of an acetamide aquo complex, [Ni(MeCONH2),(H,0),]C12, in a complex itself made from anhydrous nickel(r1) chloride and acetamide in absolute ethanol (the authors draw attention to the formation of the hydrate even though 'superficially dry conditions were maintained' 298). Infrared studies showed this complex to involve bonding through oxygen and an X-ray structural analysis indicated six oxygens coordinated to nickel at virtually equal distances.299 A green hexaacetamido complex has been made from nickel(r1) perchlorate in acetone/diethyl ether solution and its reflectance spectrum showed three main absorption bands [8 730, 13 800 (sh), 14 800, 21 800 (sh), and 25 100 cm-'1 with a pronounced shift to higher energies.266 The preparation of this complex had been attempted earlier but only an oil had been obtained,241 due it was later claimed to the presence of water, and overcome by the addition of a large concentration of ethyl orthoformate as a dehydrating agent.266 However, an acetone solution of the oil gave three absorption bands [8 240 cm-', assigned to 3A29-+ 3T2y;13 370 3A2g-+ 3T1&F);14 700 (sh); 24 510, 3A2,-+ 3T19(P)]which were again of higher energy than those of molten acetamide solution, but which might be attributed to a partially aquated nickel complex.A re-examination of the electronic spectra of nickel(I1) with acetamide ligands, together with scrupulous attention to purity and water content, is obviously very desirable.A solid hexaacetamide nickel(I1) chloride complex (m.p. 66 "C) has been reported to be made by refluxing the components in benzene or carbon tetra- chloride. Unfortunately no spectroscopy was carried out, but the solid complex was said to be blue in colo~r,'~~possibly indicating a packing effect. Nevertheless, a light green hexaacetamide with tetrafluoroborate anions has been reported with a magnetic moment of 3.49 BM, which is normal for octahedral coordination, and with infrared absorptions characteristic of oxygen-coordinated acetamide and ionic tetrafluorob~rate.'~~~~~~A triclinic 1 :6 complex with nickel(1i) nitrate has also been claimed2" and a 1 :4 complex with nickel(r1) thiocyanate whose vibrational spectra indicated bonding of acetamide through oxygen and of thiocyanate through nitrogen, suggesting an octahedral stereochemistry.' 53 The hexammine nickel(i1) bromide complex [Ni(NH,),Br,] dissolved in molten acetamide to give an intensely green solution, according to an early report.301 Spectroscopy would be interesting in order to decide if there had been ligand displacement by acetamide.Polarography and cyclic voltammetry have shown the "' M. E. Stone and K. E. Johnson. Ccin. J. Cheni.. 1971. 49. 3836. M. E. Stone and K. E. Johnson, Criti. J. Chem., 1973, 51, 1260. 299 M. E. Stone, B. E. Robertson, and E. Stanley, J.Clictn. Soc. A. 1971. 3632. 30n R. C. Paul. P. Kapila. A. S. Bedi. and S. K. Vasisht, J. Indicin Chcwz. Soc.. 1976, 53,768. L. F. Yntema and L. F. Audrieth, J. An?. Cheni. Soc.. 1930, 52, 2693. Kerridge reductions of nickel(r1) cations in molten acetamide to be irreversible.' 53 The heat of formation of a 1:2 complex has been calculated.254 Aqueous acetamide complexes have also been prepared from the ternary systems. A 1 :2: 1 complex with nickel(I1) chloride was claimed at a very early date (1886),268 while with nickel(l1) nitrate two complexes Ni(N03),*4MeCONH2* 2H20 and Ni(NO3),4MeCONH2*2H20) were found by phase diagram studies on which DTA and TGA were carried out (melting being at 89-98 and 80-90 "C respectively, loss of water at 112 "C and 111 OC, and decomposition to water- insoluble products at 260-294 and 220-298 0C).263,269.270-272An aquo complex of the first of these stoicheiometries was also made by partial dehydration of the hydrated nickel(i1) salts in methanol or butanol with a four times excess of acetamide, as well as with nickel(l1) chloride, bromide, and iodide.These com- pounds had a visible spectrum suggesting distorted octahedral coordination, and significantly the octahedral splitting (10 D4)varied with the anion, so some inter- action between anion and nickel was occurring. Electrical conductivities suggested the stability constants were in the order C1- > Br-> I-> NO;.302 Again the infrared absorptions indicated bonding of acetamide by oxygen.270,302 Spectral and conductiometric measurements on the chloride have also been made indepen- dently.297.298A more unusual octahedral complex was claimed with a nickel(i1) thiocyanate, which infrared measurements showed had bridging and sulphur- bonded terminal thiocyanate with oxygen-bonded acetamide (3).This structure was thus plainly polymeric though claimed as 'monomeric'. (However, a molecular weight was not determined because of 'insolubility in common organic solvents' and 'low solubility in DMSO'. Molar conductance values were quoted for the latter solution which indicated only a small concentration of ions to be present.273) Only three palladium-acetamide complexes have so far been reported. Palladium(I1) (as K2PdC1,) reacts directly with acetamide to give an acetamidate product, Pd,(MeCONH),(OH),, which is polynuclear and appears to have partially oxidized metal centres according to infrared, n.m.r., and e.s.r.spectro- scopic measurements.303 A dipalladium complex has been claimed as the final product of a facile hydration of acetonitrile in the presence of a skilfully designed ligand. The ease of hydration under very mild conditions was considered to arise '02 M. A. A. Beg and M. R. Farooqui, Pak. J. Sci. Ind. Res., 1971, 14, 336. 303 S. Durand, G. Jugie, and J. P. Laurent. 7runsirion Met. Chem.. 1982. 7.310. 213 The Chemistrjt of’ Molten Acetaniidr and Acetanzide Complexes MeI Me Me MeI I I Scheme 2 because of activation of acetonitrile to nucleophilic attack when coordinated to one palladium combined with the reactant nucleophilic hydroxide, being held in close proximity and bound to the second palladium.304 The reaction was considered to proceed as shown in Scheme 2. A 1 :4 complex, [Pd(MeCONH2)4](BF4)2, has been prepared and the solid has been shown to be easily reduced to black palladium by carbon monoxide, thus having potentiality as a visual detector.30s Such complexes can also act as sulphur dioxide and ozone detectors.306 The small amount of known palladium chemistry contrasts markedly with the very large number of papers on platinum acetamide complexes.The massive interest in this area arises because ‘platinum blues’ are important anti-cancer drugs which act by interfering with cell division.Most now contain acetamide derivatives, though the earlier and extremely financially successful drug ‘cisplatin’ [cis-Pt(NH,),Cl,] did not. The numerous papers on platinum-acetamide complexes are unfortunately frequently mutually contradictory and, since general agreement has not been reached on many points, their findings will now be considered on a broadly chronological basis. One of the first reports, in 1908, was of a preparation when a platinum(r1) nitrile complex was hydrolysed in the presence of silver nitrate (or sulphate) to a deep blue acetimidato complex [written as Pt(NHCOMe), and called ‘platinblau’] considered to contain platinum(i1) since, besides the stoicheio- 304 M. Louey. C. J. McKenzie, and R.Robson. lnorg. Chim. Acra. 1986. 111, 107. 305 J. L. Lambert, Y. L. Liaw, J. V. Paukstelis, and Y. C. Chiang. Environ. Sci. Technol., 1987, 21. 500. ’06 Y. L. Liaw. Din. Ahsrr. lnr. B, 1986, 47, 167. 2 14 Kerridge metry, the acetamide could be replaced on treatment with concentrated hydro- chloric acid and potassium chloride to give potassium tetrachloroplatinate(~~).~~’ The same acetimidato stoicheiometry was also claimed when the latter complex was treated with aqueous a~etamide.’~ However a hydrate (4)with platinum four- (4) coordinated with both nitrogen and oxygen from bidentate acetimidate had earlier been claimed from this same reaction [and also from that of cis-Pt(NH,),Cl, with acetamide] 308 and an unusual three-coordinate platinum(I1) hydrogen-bonded complex (5) (where X = C1-, I-, NO,, NO,, or SCN-).,09 Ammine complexes X 0 ‘Pt i/ ‘C-MeI0f ,NH I Me-C -N ---H / H (5) ([Pt(MeCONH,)(NHCOMe)NH3]X, where X = C1-, Br-, I-, or NO,, and also the corresponding compounds with en replacing NH,} which were potentially four-coordinated, were also claimed.All these latter compounds were highly resistant to oxidizing agents, though less so than the non-electrolytes, but on treatment with chlorine, platinum(rv) complexes were formed (e.g. brick red H[Pt(MeCONH,)(NHCOMe)enC14] from [Pt(MeCONH,)(NHCOMe)en]X) whose electrical conductivity indicated four ions 31 and thus possibly an interesting six-coordinate structure, but certainly deserving of further study.Alter- natively, treatment with bromine gave a red compound NH,[Pt(MeCONH,)Br,], but prolonged boiling with aqua regia apparently only produced the platinum(I1) complex [PtMeCONH,(NHCOMe)C1].3 lo The acetimidato hydrate Pt(NHCOMe),H,O was later claimed to be polymeric by its chemistry and spectroscopy (a distinct red-blue dichroism was found, similar to other complexes with Pt-Pt interactions in chains) with the water incorporated into the crystal by hydrogen bonding.311 An extensive investigation then 307 K. A. Hoffmann and G. Bugge. Ber. Bunsenges. Plijs. Ckem., 1980, 41, 312. 308 I. I. Chernyaev and L. A. Nazarova, 1~.Sekt. Plariny Drugikh Blagorodti. Met. Inst. Obshch. Neorg. Khini. Akad. Nnuk SSSR, 1951, 26, 101. 309 I. I.Chernyaev and L. A. Nazarova, Izc. Sekr. Platin), Drugikh Blagorodn. Me!. Inst. Ohshch. Neorg. Kliim. Akad. Nauk SSSR, 1952, 27, 175. 310 I. I. Chernyaev and L. A. Nazarova, Ix. Sekt. Platin?, Drugikh Blagorodn. Met. Ins!. OhshcA. Neorg. Kllini. Akad. Nauk SSSR, 1955, 30, 2 1. .3 1 I R. D. Giliard and G. Wilkinson, J. Clietn. Soc., 1964, 2835. 215 The Chemistry of’ Molten Acetamide and Acetamide Complexes showed 312,31 that the original ‘platinblau’ contained much silver sulphate as well as chloride, and that it could be purified by chromatography. A related compound, Pt(NHCOMe),Cl,, also deep blue in colour was prepared by reacting the nitrile complex Pt(MeCN),Cl, with molten acetamide in contact with air. Unfortunately these blue solids were amorphous, as were most of the analogous compounds pre- pared from substituted acetamides.However one compound, Pt(Me,CCONH,),- CI,, was crystalline and a single crystal suitable for X-ray diffraction analysis was obtained. A partial refinement showed a non-linear chain of three platinum atoms each attached to two chlorides but did not give the positions of the light atoms. Then, however, chromatography was employed which showed the presence of three distinct compounds. On the basis of n.m.r. and infrared measurements these were considered to have the structures (6)-(8). CI OH\ N-Pt-N IIyellow, Pt Me,C-C // I ‘H \ CI OH (6) OH CI \,CMe, HI yellow, pt19 M e3C- C/N-rN\\o CI ( 7) L‘ O=C-CMe,I/ /0-Pt-Ni deep blue, PtIP Me,C-C<N/ I H CI (8) The triplet chain of three platinums in the original blue-green needles is made up of either compounds (7) or (8) randomly at the centre, with compound (6) on either side, and is apparently not constrained together by more than packing consider- ations in the mixed crystals.The blue compound (8) was considered to be the model for platinblau, which would thus also be a platinum(1v) complex, since analysis of the chloride-free material agreed ‘almost perfectly’ [with Pt(MeCONH,),(OH),] and moreover mass spectroscopy showed the molecular weight of the parent ion to be 345 as calculated, rather than the 329 required for the hydrate originally postulated. However, no evidence was advanced that these molecules might not be 3’2 D.B. Brown, M. B. Robin, and R. D. Burbank, J. Am. Ciiet?i.Sor,., 1968.90. 5621. Kerridge stacked in the crystal.,l 2,31 Shortly afterwards repetition of the chromatographic separation was reported to yield two blue phases 314 and not two yellow and one blue as originally reported. These findings do not appear to have been further elaborated. Polymeric structures were however soon supported, together with the new suggestion that platinum had fractional oxidation states, the latter on the basis of titration with vanadium(I1). Platinum-platinum interaction was considered to be diminished due to the greater metal-metal distances with nitrogen-substituted acetamides, where only yellow complexes had been obtained {cis-[Pt(NH,),-(H,0)2]S04 reacted with a three-times excess of acetamide to give an orange-red compound, 20 800 cm-'}.,' The blue compound precipitated from aqueous acetamide by other workers [K,PtCl, --+ Pt(NHCOMe),Cl] was also considered to be polymeric with platinum wholly or partially in oxidation state III; infrared analysis indicated coordination by nitrogen as well as by oxygen.3033316 Many similar compounds with other anions (Br-, I-, SCN-, NO,, NO,) replacing chloride were also prepared; dimeric structures were suggested with the two metal atoms joined by four bridging acetimidato groups.31 In another investigation the same reactants were reported to give a further deep blue compound in which planar cis-Pt(MeCONH,),(OH)Cl units, with three Pt-0 bonds, were stacked to form polymers.318 The analysis figures reported do not support this stoicheiometry, but suggest instead Pt2C6H1 ,N,O,Cl-or close to that expected for the acrylamide complex also claimed in this paper.The quoted analysis figures for the latter are much nearer, but not close to, those calculated for the acetamide-platinum complex. Thus the stoicheiometry of the acetamide- platinum complex may well not be that claimed. Dimers rather than indefinite polymers were thought more probable by other authors, the diamagnetic measure- ments suggesting variable oxidation states (i.e. Pt"-PtIv rather than Pt"').319 At about this time titrations with cerium(1v) indicated an average oxidation state of 2.25,320 and provided support for the chain of four platinums postulated by analogy with that found in a single crystal X-ray diffraction study of cis-diammine platinum x-pyridone blue [Pt,(NH,)4(C5H40N),],(N03)5,which is another rare example of a crystallizable compound from this family of platinum blue com- pounds.Extensive X-ray photoelectron spectroscopy (XPS) measurements have however been made on many of these compounds, and the acetamide complex was found to have similar binding energies to the x-pyridone complex and suggested platinum in oxidation state 11, though some samples of platinblau had small .'I3 D. B. Brown, R. D. Burbank, and M. B. Robin, J. Am. C'hem. Soc.. 1969, 91. 2895. 'I4 D. F. Cahen, Dhs. Ahstr., 1973 ..4,34B. 4266. 'Is C. M.Flynn. T. S. Viswanathan, and R. B. Martin, J. Inorg. Nucl. Chem.,1977. 39, 437. .'I6 S. Durand. G. Jugie, and J. P. Laurent, C. R. Acad. Sci., Ser. C, 1980, 290, 145. -'I7 R. N. Shchelokov, A. Y. Tsivadze. A. G. Maiorova. and G.N. Kuznetsova, Zii. Neorg. Khim.,1978.23, 1036; 1979. 24. 1279. 31x G. Schmuckler and B. Limoni, J. Inorg. Nucl. Chem.. 1977, 39, 137. 'I9 V. 1. Nefedov, Y. V. Salyn, and I. B. Baranovskii, Zh. Neorg. Kliirn., 1980. 25. 216. "" J. K. Barton, C. Caravana. and S. J. Lippard. J. Am. Chem. Soc., 1979, 7269. The Chemistry qf Molten Acetamidc and Acctamidr Complexes amounts of higher oxidation state con tarn in ant^.^^' Other XPS (ESCA) measure- ments have also indicated equivalent platin~ms,~~~,~~~ though the assertion that these were in oxidation state III,,~ is ill-founded, since the overlap of Pt 4J,,, binding energies is such that oxidation states II and IV are not excluded.In contrast, magnetic susceptibility studies further supported the suggestion of one unpaired electron per tetramer chain (i.e. 0s 2.25) and single-crystal e.s.r. studies indicated this resided in a molecular orbital derived from atomic d,z orbitals, directed along the platinum chain axis. The visible blue colour results from a broad absorption band centred at 14 500 cm-' with an extinction coefficient around 1300 MP;l cm-', and attributed either to an amide, to metal charge- transfer,, ',or alternatively to platinum-platinum interactions. To sum up, it is clear that platinum blues are by no means completely characterized and that further investigation is to be welcomed.However, despite this, it seems likely that there are platinum-platinum interactions and that the oxidation state is greater than two. Further, it may be that polymeric and dimeric structures are interconvertible, though experimental conditions for this have not been defined. Some organometallic complexes, possibly somewhat similar to platinum blues but with platinum also bonded to carbon, have been made by hydrolysis of organometallic nitrile complexes (equation 19).324p326 [PtR(NCMe)L,]BF, Pt(NHCOMe)RL, (19) where R = Me, Ph and L = PEt,, PMe,Ph, PMePh,, PPh, More generally, platinum(I1) hydride complexes [trans-PtHC1( PR,), where R = MR or Et] have been shown greatly to increase the rate of hydration of acetonitrile in basic aqueous solution (as opposed to the conventional sulphuric acid-catalysed hydration).though with a different mechanism from that of the dipalladium complex. Three intermediates were identified spectroscopically [PtH(H,O)(PEt,):, PtH(NCMe)(PEt,);, and PtH(NHCOMe)(PEt,),] and proton transfer from solvated water to the coordinated N-carboxamide ligand was found to be rate limiting.327 An undoubted compound with platinum(rv) has been reported [(MeCONH,),- PtCl,, prepared from H,PtC16 and aqueous acetamide in sulphuric acid solution],54 and the probably analogous hydrate (H,PtCI,~2MeCONH2-2H20) reported much earlier from a similar preparation was said to form fine, long golden- brown, prisms with a melting point of 83.4 0C.328 32' J.K. Barton, S. J. Lippard. and R. A. Walton, J. Am. Cliem. Sor... 1978, 3785. 322 J. Salins, V. L. Nefedov. A. G. Maiorova, and G. N. Kuznetsova. Zli. Neorg. Kliim.. 1978. 23. 829. 323 V. L. Nefedov and J. Salins. 1nor.g. Cliini. Acttr. 1978. 28. L135. 324 M. A. Bennett and T. Yoshida, J. Ani. Cliem. Soc.. 1978, 100, 1750. 325 D. P. Arnold and M. A. Bennett, J. Orpt?ome[.Cliem., 1980, 199. 119. 326 D. P. Arnold and M. A. Bennett, J. Orgcimniet. Clien?..1980, 202. 107. 32i C. M. Jensen and W. L. Trogler, J. Ani. Cliem. Soc., 1986. 108. 723."'R. Fricke and F. Ruschhaupt. Z. Anorg. Clieni.. 1925. 146, 141. Kerridge (ix) Copper Group.Copper(r1) chloride is soluble in molten acetamide and was early reported to be highly dissociated, and in contrast to other transition metal dichlorides the hydrated water was partially retained (i.e.forming CUCI,=~H,O).~ l6 Interfacial tension measurements gave breaks indicating two complexes (CuCl,* xMeCONH,, x = 1 or 2).'35 Potentiometric titration of copper(I1) chloride, or nitrate, in molten acetamide at 98 OC, showed the green solutions to be weakly diacidic towards acetimidate anions, forming Cu(NHCOMe),.Stability constants were calculated.' 78 Earlier phase-diagram studies of the ternary system CuC1,-MeCONH,-H,O suggested the formation of an anhydrous complex CU(M~CONH,),C~,.~~~~~~~ The same stoicheiometry which had been reported in 1886 268 was also found by a preparation involving refluxing in benzene or carbon tetrachloride with a reported melting point of 138 "C,14, and was also prepared by precipitation from alcoholic solutions (when the product was stated to be light green, decomposing at 110 "C, magnetic moment 2.07 BM,253.332 and 4 charge-transfer and 20 infrared bands indicated coordination through oxygen). The heat of formation has been cal- ~ulated.,~~Copper(1r) bromide under similar conditions gave a different stoicheio- metry (CUB~,~~M~CONH,~~H,O).~~~With copper(I1) sulphate the ternary phase diagram (CuS0,-- MeCONH,-H,O) indicated no ' but density and viscosity measurements suggested two complexes, 1:2 and 1 :4.333 A complex of 1:2 stoicheiometry has been isolated, infrared and X-ray diffraction showing coordination through oxygen and bidentate ~ulphate.~~, A later investigation however proposed a 1 : 1.5 complex from refluxing methanol solutions, and this light blue compound decomposed on heating at 21S240 "C to a 1 :0.5complex which lost all acetamide at 245-260 0C.335 An anhydrous copper(I1) nitrate 1 :6 complex (triclinic) has been reported 257 and a hydrated 1 :6 complex Cu(N03),-6MeCONH,-2H,0 from phase diagram st~dies.~~'.~~~Co pper(I1) acetate gave a 1: 1 anhydrous which may of course be dimeric as is the case with the much better known monohydrate; coordination was certainly through oxygen.Aqueous solutions gave a hydrated complex CU(OOCM~),~M~CONH,~H~O.~~~A 1: 1 ratio was also found with butcopper(r1) monochl~roacetate,~~~ not with tetrafluoroborate [light blue Cu(BF4),*4MeCONH,, magnetic moment 2.02 BM, i.r.indicating bonding through oxygen, distorted octahedral, with weakly bonding fluoroborate in axial positions].300 An acetimidate of copper(i1) has been made, Cu(NHCOMe),, and could be titrated with excess sodium acetimidate to give a deep yellow-green anion (equation 20).4 319 N. V. Saleeva, M. K. Kydynov, and B. I. Irnanakunov. Zh. Prikl. Kliini. (Leningrad). 1969. 42. 544. 330 N. V. Saleeva. M. K. Kydynov, and I. G. Druzhinin, Deposited Doc. 1973, VINITI, 6271. 331 N. V. Saleeva, M. K. Kydynov, and B. I. Imanakunov. IY. Akad. Nauk Kirg. SSR, 1968, 51. 332 M. A. A. Beg and M. A. Hashrni, Pak. J. Sci.Itid Res.. 1971, 14, 458. 333 S. S. Ahrned.S. A. Khan. and A. R. Khan, Pak. J. Sci. Ind. Rex, 1970, 13, 45. 334 M. S. Barvinok and L. V. Mashkov, Zli. Neorg. Khirn.. 1979. 24, 2833. 33s J. E. House and P. D. Dunlop, Tlzerniochirn. Acra, 1981, 47, 113. 336 T. A. Azizov, 0.F. Khodzhaev, and N. A. Parpiev, U:h. Khini. Zh., 1976, 6. 33' R. C. Paul, P. Singh. H. S. Makhni. and S. L. Chandha. J. Inorg. Nucl. Climi., 1970. 32, 3694. 219 The Chemistry qf' Molten Acetamide and Acrtamide Comp1e.w~ Table 3 Solubi1itie.y of silver(r) sults in molten ucetamide Solute Solubility constunt (K,) at 98 "C" at 87 OCb AgCl 1.95 k 0.05 x lo-' 1.9 i0.5 x AgBr 4.1 & 0.1 x lo-'' 8.5 i0.1 x lo-" AgI 1.1 * 0.1 x lo-', 2.2 i0.2 x 10 l2 AgSCN 1.3 _+ 0.3 x 'S. Guist, Anti. Chin?.(Priris), 1969 4.235. M. Gusteri. V. Bartocci, and F. Castellani, J. E/t,c~frocinuI. Ciiem.. 1979. 102, 199. Cu(NHCOMe), + 2NaNHCOMe -Na,[Cu(NHCOMe),] (20) The only copper(1) compound so far reported, Cu(NHCOMe)PPh,, contains an acetamidate group and the infrared absorptions indicate coordination through nitrogen.338 Silver([) nitrate was initially found to form a 1 : 1 complex with acetamide 17' which was confirmed by a binary phase diagram study, though no breaks were present in density, viscosity, and conductivity isotherms.339 However, some decom- position was suspected. Reaction was later definitely found with molten acetamide at 98 "C;the brown colour initially formed underwent progressive bla~kening.,~' This reaction was due to the high oxidizing power of silver(1) and nitrate together, since silver(r) cations without nitrate are stable and have been titrated potentio- metrically against sodium acetimidate. Breaks w-ere found at 1 : I and 1 :2 ratios, the latter stoicheiometry being claimed as brown in c~lour,~ and values for the stepwise stability constants and for the acid dissociation constant of acetamide were reported.340 In fact silver([) cations are more solvated in acetamide than in water at 98 OC, though the solubilities determined (Table 3) were somewhat smaller than in water at 20 "C.Further coordination occurred in molten acetamide, with halide and pseudo-halide anions, and formation constants have been An early report stated that gold(1Ir) formed a complex (HAuC14*2MeCONH,) as yellow needles which decomposed before melting,328 and solutions have been found to be electroreduced in a two-electron process.' " (x) Zinc Group.Conductivity measurements have suggested that zinc(r1) halides in solution in molten acetamide at 94 "C are incompletely dis~ociated.~ A number of complexes have been reported. By refluxing in inert solvent a 1 :2 complex with zinc(I1) chloride (ZnC12-2MeCONH2) was obtained 142 (described as a glassy solid of m.p. 173 "C), and the same stoicheiometry was obtained by the solubility isotherm technique at 25 "C from aqueous acetamide solutions,343 though this was 3'R T. Yamamoto. Y. Ehara. and M. Kubota, Bull. Clicwi. Soc. Jpn.. 1980. 53,1299. 339 M. A. Klochko and G. F. Gubskaya. Zli.Ntwrg. Kiiini.. 1960, 5,2491. ''O S. Guiot and B. Tremillon, J. Elt~c~trotrritrl.C/ll'I77.. 1969. 22. 147. 341 S. Guiot. Ann. Chitn. (Pmiy), 1969. 4. 235. s3'' M. Gusteri. V. Bartocci, and F. Castellani, J. El~~c~tro~itzr~l.Chetn., 1979. 102. 199. 343 I. G. Druzhinin, B. K. Dzhashakueva. N. V. Makhonina. and T. S. Kozhanova. Zii. Ntwrg. Khini.. 1982, 27, 23 1. 220 Kerridge described as triclinic crystals, density 2.10, with a melting point of 68 "C. The latter technique also suggested two orthorhombic zinc(I1) bromide complexes (1 :2, d. 3.16, m.p. 59 "C; and 1 :3, d. 2.86, m.p. 46 "C) both with bonding through oxygen according to infrared spectra.343 The heat of solution of the 1 :2 chloride complex has been measured and the heat of formation calculated.254 Zinc(I1) nitrate forms two hydrated complexes from such solutions [Zn(NO,),-4MeCONH2*2H,O, d.1.51, m.p. 7&80 "C; and Zn(N03),*6MeCONH,*2H20, d. 1.62, m.p. 78-85 0C].27 1*272*344 The zinc(I1) acetate complex Zn(02CMe),-2MeCONH2 was shown by X-ray diffraction and infrared spectroscopy to have bidentate acetate groups with monodentate acetamide coordinated through oxygen, and to be octa- hedra1.256 Some evidence for the formation of chlorozincate anions was found from specific conductance measurements on zinc(I1) chloride dissolved in molten acetamide saturated with potassium chloride. 345 Zinc acetimidate Zn(NHCOMe), is amphoteric, dissolving in basic molten acetamide containing excess sodium a~etimidate.~ Cadmium(I1) cations in molten acetamide at 87 "C (1 M in K0,CMe) have been studied by d.c.polarography and cyclic voltammetry, but metal deposition was not reversible.'53 The conductivity of solutions of the halides at 96 "C suggests they are incompletely dis~ociated,~ but their solubilities in molten acetamide follow the same sequence as in aqueous solution (CdBr, > CdI, > CdC12).346 Several authors from as long ago as 1886,268 agree that an anhydrous 1:2 complex with cadmium(I1) chloride is formed, though this is variously described as white crystals (m.p. >200 "C) 142 and as 'colourless, long monoclinic prisms'.347 X-Ray diffraction investigations showed the cadmium to be four-coordinate square planar, with two acetamides and two chlorides which were nearly equidistant to neighbouring metals, thus giving an effectively octahedral geometry.348 A complex of the same stoicheiometry has also been found through phase diagram studies.'25 The heat of solution in water of this complex, and of the corresponding bromide and iodide complexes, has been measured and their thermal decomposition has been found to occur in three stages (equation 21).349 CdX2*2MeCONH2---CdX,*MeCONH, -CdX,.xMeCONH, -CdX, (21) where .Y = 0.54.7 A 1 :1 cadmium(I1) bromide complex has also been rep~rted.~".~~' Infrared spectroscopy has shown that the 1 :2 complexes of the three cadmium(I1) halides 344 A.D.Dzhunusov, B. I. Imanakunov, M. K. Kydynov, and A. S. Karnaukov, Izc. Vpsh. Uchebn.ZaGed. Kliini.Kliini. Tekhnol., 1969. 123, 13. 34s R.A. Wallace, J. Inorg. NucI. Chem.. 1973. 35, 3641. 346 L. Belladen, Gnx. Chini. Ital., 1927. 57. 412. 347 M. Nardelli. L.Cavelca, and L. Coghi. Ricercri Sci., 1957, 27, 2144. 34R L. Cavalca. M. Nardelli, and L. Coghi, NUOL~OCimento, 1957, 6, 278. 349 T. A. Azizov. 0.F. Khodzhaev, and N. A. Parpiev, Uzb. Kliim. Zh., 1977, 27. 350 D. Usubaliev. B. Irnanakunov, and P. T. Yun. Moter. Nauchn. Konf: Posvyoshch IOO(Sto)Leriyu Period. Znkona D. 1. Medeleecn 1969, 31, (1970). "I D. Usubaliev, M. Batkibekova. V. D. Yusupov. and P. T. Yun, Tr. Frunz. Polirekh. Insr., 1974,79, 105. 22 1 The Chemistry of Molten Acetamide and Acetctmide Comple.xes and of the thiocyanate to have oxygen-bonded acetamide and halide, or pseudo- halide, bridges.352 From aqueous acetamide solutions 1: 1 : 1 complexes were formed with cadmium(1r) chloride, though not apparently with cadmium(r1) bromide, but the colourless crystals were not found suitable for X-raydiffraction.125 347.3 SO Cadmium(I1) nitrate has also been much studied.An anhydrous 1 :6 complex forming triclinic crystals has been reported which decomposed thermally in two reactions (equation 22),257*349 as well as a 1 :6 hydrate (Cd(N03),-6MeCONH2* 2H,O, d. 1.62, m.p. 82-91 "C) from phase diagram studies of the ternary system.27 1,272,344 By contrast, cadmium(i1) acetate formed a 1 :2 complex, characterized by X-ray diffraction, infrared, electrical conductivity, and thermal studies, where the acetamide was bonded through oxygen and the acetate was bi- dentate, giving an octahedral ge~metry.,~~,~~~ Cd(N03),.6MeCONH, -Cd(N03),*2MeCONH,-Cd(N03), (22) Mercury compounds have received mention in many reports and mercury(l1) acetimidate has been known since 1909 as a yellow compound (m.p.195 "C) prepared by reacting red mercury(i1) oxide with molten acetamide at 18CL 200 0C,48 also prepared from reaction with aqueous a~etamide.~ The stoicheio- metry Hg(NHCOMe), was confirmed by potentiometric titration of mercury(I1) cations with acetimidate anions in molten acetamide at 98 "C; the first association was strong and the second was weak.'78*179 Bonding of mercury to the imide nitrogen is clearly indicated by n.m.r. studies,35 and single crystal X-ray diffraction analysis gave a structure showing discrete planar centrosymmetric molecules with mercury forming two trans bonds to nitrogen with two longer bonds to the oxygens of neighbouring acetimidate anions (thus chelation does not occur) with further linking by hydrogen bonds along the c Voltammetry in molten acetamide solutions at 87 "C containing 1M potassium acetate produced a mercury acetate complex on the surface of the mercury electrode at the anodic limit.153 Mercury(i1) was shown to be the only stable oxidation state from the voltammetric and potentiometric studies, mercury(r) being found to dissociate to grey metallic mercury and mercury(Ii).' 79 Mercury(i1) acetimidate was also reported to dis- proportionate in the solid or in methanol/acetone solution, but in this case the anion reacted to form free acetamide and a polymer, which was claimed to be Hg(NCOMe),.Mercury(1i) halides have been found to dissociate only incom- pletely in molten acetamide, possibly also because of disproportionation (actually claimed for HgCl,'), but have also been reported to form complexes (HgCl,. 6MeCONH,, m.p. 68 OC, and HgBr,*4MeCONH2, m.p. 110"C; both as white crystalline solids) by reaction in benzene or carbon tetrachloride solutions. 142 Earlier, two other stoicheiometries were claimed-HgCl,*MeCONH,, m.p. 125 "C 352 A. Y. Tsivadze, Y. Y. Kharitonov, G. V. Tsintsadze, A. N.Smirnov, and M. N.Tevzadze, Zh. Neorg. Khim., 1974, 19, 2621. 353 D. B. Brown and M. B. Robin, Inorg.Chim. Actn. 1969, 3. 644. 354 B. Kamenar and D. Grdenic, Inorg. Chim. Acrn, 1969, 3. 25. Kerridge and HgCl,-xMeCONH, m.p. 1 18 "C 268-and the 1 : 1 stoicheiometry (HgC1,- MeCONH,) prepared from aqueous solution has also been claimed more re~ent1y.l~~A mercury(I1) cadmium(I1) iodide complex 355 (given as HgI,-2Cd12* lOMeCONH,, m.p. 85 "C) prepared from molten acetamide could well be mixed crystals {e.g. Hg12*6MeCONH,, 2[Cd12*2MeCONH,]) but no structural details have been reported. Studies have also been made of the interaction of mercury(1r) chloride with acetamide in neutral and basic aqueous solution (the formation constants varied HgCl-NHCOMe > Hg(NHCOMe), > Hg(OH)NHCOMe).3 Mercury(I1) nitrate forms a 1 :2 acetamide complex which has been characterized by X-ray diffraction and infrared techniques, though the melting point was quoted as the wide range 54-70°C.271,272 The same stoicheiometry was found with mercury(1r) acetate, where these techniques showed bidentate acetate and bonding through the oxygen of acetamide, resulting in octahedral c~ordination.~~~,~~~ D.Lanthanide Group.-Acetimidates of four lanthanide(r1r) cations, M(NHCOMe), where M = La, Pr, Nd, or Sm, were prepared by refluxing the lanthanide(II1) isopropoxide with acetamide in benzene. The complexes had the 'usual' colours (white, light green, pink, and yellow respectively) and electronic (.f-f') absorption bands were reported for praseodymium and neodymium com- pounds only. Infrared absorption indicated bonding nitrogen as well as some coordination through oxygen, which may support the suggested polymerization deduced from the insolubility of these complexes in common organic solvents and melting points 'above 300 "C'.Notwithstanding the latter point, thermogravi- metric analysis showed considerable weight losses above 100 "C, ammonia and 'nitrile' being evolved with the eventual formation of lanthanide oxides and oxy~arbonates.~~~Several solubility isotherm (phase diagram) measurements on ternary aqueous acetamide solutions have been made. In the case of lanthanum 358trichloride two stoicheiometries (LaC13~5MeCONH2~5H20 and LaCl,. 4MeCONH,-SH,O 359) have been reported, the former also being found with cerium trichloride (CeCl,~5MeCONH2~5H,O 360) but not with the tribromide (CeBr,-5MeCONH,*3H,0).361Erbium tribromide, however, gave an anhydrous complex (ErBr3.4MeCONH, 362).Cerium(1v) sulphate gave another stoicheiometry [Ce(S04),-4MeCONH,* 4H20, d. 2.151 which lost water at 120 "C and also showed three other (un- identified) thermal effects between 180 and 350 0C.363 Another study showed acetamide was oxidized by cerium(Iv), the measured rates indicating that this was uia inner-sphere cerium(1v) acetamide complexes.364 35s S. Prasad and P. I). Sharrna. J. Indian Chem. Soc., 1958, 35, 565. 3s6 R. 0.Gould and H. M. Sutton, J. Chem. Soc. (A), 1970, 1184. 357 A. M. Bhandari and A. K. Solenki, Synth. React. Inorg. Met.-Org. Chem., 1981, 11, 267. 35R Z. Zholalieva, K.Sulaimankulov, and K. Nogaev. Zh. Neorg. Khim., 1976. 21, 2290. 3s9 Z. Tang, Y. Wen, T. Li, and Y. Chen. Gaodeng Xuexiau, 1983, 4, 426. Z. Zholalieva and K. Sulaimankulov, Zh. Neorg. Khim., 1978, 23, 1206. Z. Zholalieva, K. Sulairnankulov, and M. Isrnailov, Zh. Neorg. Khim., 1976, 21, 2583. 362 K. Aitirnbetov, K. Sulairnankulov. K. Nogaev. and L. Kovalenko, Zh. Neorg. Khim., 1977, 22, 11 16. 363 V. A. Golovnya and L. A. Pospolova, Zh. Neorg. Khim., 1961, 6, 636. 364 S. Sondu, B. Sethuram, and T. N. Rao, Osid Commun.. 1984, 7, 223. 223 The Chemistrjt qf' Molten A cetamide and Acetamide Comp1e.ur.s A number of complexes have recently been reported containing an acetamido- borane ligand {LnL,(H,O), where Ln = La, Pr, Nd, Sm, Eu, Gd, Dy, Ho, Er, Tm, Yb, Lu and [Ln(pyrO),]L, where Ln = Nd, Sm, Dy, Eu, Yb, Lu and HL = H(B, ,H, ,NH,COMe)) which was coordinated to the lanthanide via oxide oxygen in the solid state.These complexes dissociated extensively in water or E. Actinide Group.-Actinium complexes have not been reported so far and thorium(1v) has been the subject of only one investigation, when thorium(1v) sulphite tetrahydrate was reacted with molten acetamide, two complexes being found after extraction with acetone which were characterized by elemental analysis [white Th,(S03),(OH),~MeCONH2~5H20and pale yellow Th,(SO,),(OH),- 2MeCONH2*4H,O]. The infrared absorptions indicated bonding through oxygen. Thermal analysis showed water to be lost at 70-175 "C and 65-115 "C respectively, and acetamide at 240-330 "C and 205-250 "C with formation of thorium(rv) oxide and sulphate (probably air was the oxidant) with loss of sulphur trioxide above 470 0C.366 At first sight it is surprising that uranium-acetamide complexes with a wide variety of ligands have been the subject of so many investigations.However this interest arose largely from an early claim of the nuclear energy industry that uranium(v1) could be effectively extracted from organic solutions by aqueous acetamide, from which it could be recovered by heating.367 Uranium(v1) oxide dihydrate reacts with acetamide in ethanol to give a 1:l:l complex UO,. MeCONH,-H,O, whose infrared spectrum suggested coordination through oxygen,368 which decomposed thermally in two stages (at 210°C to UO,(OH)- (0,CMe) i.e.oxidation of acetamide, and then at 280-300°C to uranium oxide(s). The final stoicheiometry depended on the oxygen partial pressure. Uranyl fluoride hydrate also complexed with acetamide to give a product, (UO,F,MeCONH,),, which was thought to be polymeric. Infrared and Raman spectra, laser luminescence, and thermal stability have been reported for this comp~und.~~~-~'Uranyl chloride in methanol solutions of acetamide formed green crystals of U0,CI,~2MeCONH,~H,0.372 Uranyl nitrate is soluble in molten acetamide and the ultraviolet absorption spectrum at 85 "C had 18 sharp bands.373A 1:2 complex, UO,(NO3),-2MeCONH2, was isolated from ethanol solution which decomposed thermally above its melting point (195 "C), losing acetamide, rapidly from 220 "C and exploding at 370 "C, forming uranium(vr) 365 G.Zhang, F. Jiang, and L. Zhang. New Fronr Rare Enrlli Sci. Appl. Proc. Int. Con/. Rare Etrrrli Der. Appl., 1985, 1, 187 (ed. G. Xu and J. Xiao, Sci. Press, Beijing). '" V. A. Golovnya. A. K. Molodkin, and V. N. Tverdokhlebov, Zh. Neorg. Kliim., 1967. 12. 2729. 367 Aktiebalaget Atomenergi, Br. Pat. 316628, 1959. 36n V. Z. Kolesnik, A. L. Zhirov. and K. M. Dunaeva, Zh. Nmrg. Khini., 1981, 26, 1849. 369 R. N. Shchelokov, I. M. Orlova, and G.V. Podnebesnova. Koord. Kliim.. 1975, 1, 119. 370 R. N. Shchelokov and A. Y. Tsivadze, Koord. Khim., 1978.4, 313. 371 R. N. Shchelokov, A. Y. Tsivadze, I. M. Orlova. and G. V. Podnebesnova, Inorg. Nud. Chem.Lett., 1977, 13, 367. 372 V. P. Markov and I. V. Tsapkina, Zli. Neorg. Kliim., 1962, 7, 2045. 373 T. Nakai, Bunko Kenkw, 1954, 3. 225. Kerridge oxide, carbon dioxide, and nitrogen. It was noted that oxygen ligands gave less stable complexes than with nitrogen ligand~.~~~ But in tributyl phosphate solution acetamide is bonded to uranyl nitrate, because the changes in its absorption spectrum indicate displacement of nitrate or tributyl phosphate.375 Uranyl sulphate formed three complexes, two in aqueous acetamide solutions of varying concentration (U0,S04~3MeCONH,-H,0 and green U0,S04-MeCONH,-H,O) and a further green complex (U0,S0,-2MeCONH2) was formed from ethanol The uranyl cations in these complexes were shown to have temperature-independent Van Vleck ~aramagnetism.~~~ Uranyl sulphite gave two different stoicheiometries (U0,S03~1.5MeCONH,-H,0 and U0,S03~MeCONH,~1.5H,0),377the latter also being claimed with smaller, to zero, proportions of water.378 Infrared spectroscopy indicated bonding through oxygen and probably bidentate or bridging sulphite.No detailed structures were proposed, but thermal decompositions were studied. The perhaps unexpected uranyl phosphite complexes have been made by reaction of uranyl nitrate in aqueous acetamide solutions containing phosphorous acid [greenish-yellow crystals of U0,(HP03)~MeCONH,~H,0 and U02(HP0,)- 2MeCONHJ. Acetamide was lost at >200 "C and ultimately uranium(rv) and (VI) phosphates were formed.379 The structure of the uranyl chromate acetamide complex UO2(CrO,)-2MeCONH, has been established by X-ray diffraction.The uranium is surrounded by a pentagonal bipyramid of oxygens, three equatorial oxygens coming from chromate and two from the acetamide molecules, which lie at -55" to the equatorial plane, the uranyl group being linear and at right angles to this pentagon. The (UO,)O, groups are joined into chains by chromate bridges.380 Thermal decomposition started at 19&220 "C with loss of acetamide; the uranyl chromate which formed decomposed at >600 0C.381A uranyl isothiocyanate complex has also been prepared from aqueous solution [UO,(NCS),-MeCONH,. H,O, m.p. 105"Cl which was also concluded to have a penta-cordinated uranium(v1) cation. Though acetamide was fairly readily replaced by 1 :10phenan-throline, it was more stable than the coordinated water, since dehydration occurred at 15&160 "C and decomposition at 240-420 0C.3823383 Uranyl oxalate, however, gave an anhydrous complex (U0,C20,-MeCONH,) from aqueous solutions.372 Several uranium(1v) complexes have also been made.The chloride gave a 1 :6 374 G. Siracusa, L. Abate, and R. Maggiore, Thermochim. Acta, 1982, 56, 333. 37s S. Minc and L. Werblan, Roczniki Chem., 1958, 32. 1419. 37h V. I. Beltva, Y. K. Syrkin, V. P. Markov, and I. V. Tsapkina, Zh. Neorg. Khim., 1961, 6, 495. 377 A. Y. Tsivadze, A. N. Smirnov, G. T. Bolotova, N. A. Golubkova, and R. N. Shchelokov, Zh. Neorg. Khim.. 1979. 24. 1635. 378 R. N. Shchelokov, N.A. Golubkova, and G. T. Bolotova, Koord. Kliim., 1975, 1, 113. 37q K. A. Avdnevskaya, N. B. Ragulina, I. A. Rozanov, Y. N. Mikhailov. A. S. Kanishcheva, and T. G. Grevtseva, ZIT.Neorg. Khim., 1981, 26, 1011. 380 Y. N. Mikhailov, I. M. Orlova, G. V. Podnebesnova, V. G. Kuznetsov, and R. N. Shchelokov, Koord. Kliim., 1976, 2, 1681. 3n1 R. N. Shchelokov. I. M. Orlova, and G. V. Podnebesnova. Zh. Neorg. Klzim., 1974, 19, 1581. 3R2 R. N. Shchelokov, I. M. Shulgina. and I. I. Chernyaev, Dokl. Akad. Nauk SSSR, 1966, 168, 1338. Jx3 R. N. Shchelokov. I. M. Shulgina. and I. I. Chernyaev, Zh. Neorg. Khim., 1967, 12. 1246. The Chemistrq’ oj Molten Acetamide and Acetamide Complexes stoicheiometry (Cs,UCl, + hot acetone solution of acetamide + UCl,.6MeCONH,) which the infrared spectrum showed had oxygen-coordinated acet- amide. A more stable 1:2.5 complex was also made which was presumed to be dimeri~.~~~The sulphate, however, in molten acetamide, gave a bright green 1:4 complex [U(S04),-4MeCONH,, m.p. 180 “C] which decomposed at 240-260 0C.385 Two other complexes with ammonium cations [(NH,),U(SO,),-2MeCONH, and (NH4),U(S0,)3- 2MeCONH2-4H,0] have also been prepared which were considered to contain octa-coordinated uranium, all the ligands bonding through oxygen.386 Uranium(1v) bromide solutions in molten acetamide have been found to electrodeposit uranium, but at a very low current efficiency, probably because of the reactivity of uranium(~~i).~~’ Little work has been done with transuranic elements, but neptunium(1v) and plutonium(1v) complexes (MC14-6MeCONH, and MCl,.2.5MeCONH2 have been reported and were analogous to the corresponding uranium(rv) com-pound~.~’~ 4 Applications As mentioned earlier, platinum-acetamide compounds are already used as anti- cancer drugs, and there is a possible new large-scale use of acetamide eutectic mixtures in thermal energy ~t~rage.~~,~~~*~~~ Besides these, two other areas of application are being actively considered. The first, electrodeposition, arises because molten acetamide solutions are convenient in providing high metal-cation concentrations, with an absence of hydrolytic reactions. Frequently acetamide-urea eutectics have been used to achieve even lower melting points [e.g.with 39 mole% CO(NH,),, m.p.56 0C].390-392 An early (1930) study showed that seven metals could be electroreduced (Zn, Cd, Pb, Sn, Co, Ni, Te) and that good quality deposits could be obtained even in the presence of small quantities of water.301 Later, reduction of titanium(1v) chloride solutions was reported 2249227 and also reduction of solutions of chromium(v1) and (111) compounds [K2CrzO7, Cr(O,CMe),]. Copper, iron, and nickel have been electrodeposited from ammonium ni tra te-acetamide-urea me1 ts. 93-3 With the less electrochemically active metals, deposition (‘cementation’) is possible without an externally applied current (e.g. InCl, and SnCl, deposit alloy layers on the surface of Mg and A1 alloy substrates).’s1 Surface oxide layers have also been 384 K.W. Bagnall. A. M. Deane, T. L. Markin, P. S. Robinson. and M. A. A. Stewart, J. Chrm. Soc.. 1961, 161I. 385 V. A. Golovnya and G. T. Bolotova, Zli. Neorg. Kliint., 1961, 6, 566. 386 G. T. Bolotova and V. A. Golovnya. Konipleksn. Soerlin. Uranu Akacl. Nnuk. SSSR Inst. Ohslicii. Neorg. Kliim., 1964, 393. 387 E. D. Eastman and B. J. Fontana, USAEC TID-5290, Book 1, 1958. p. 206. 388 K. K. K. Eslen and S. Kiya. Jpn. Kokui Tokhj~iKojo JP 6006781, 1984. 389 J. G. Dunn, H. C. Smith, and R. L. Willis. Tlirrrnocliirn. Actu. 1984. 80. 343. 390 G. E. McManis, A. N. Fletcher, D. E. Bliss, and M. H. Miles. J. Appl. Elec~frochc~ni.,1986, 16, 101. 3y1 G. E. McManis, A. N. Fletcher, and D. E. Bliss, J. Elec~~mimml.Chern., 1985. 190. 171. 392 A. N. Fletcher, G. E.McManis. and D. E. Bliss. US Pat. 4555455. 1986. 393 G. E. McManis, A. N. Fletcher, D. E. Bliss, and M. H. Miles. J. Appl. Elec~~ocliern..1986. 16, 229. 3y4 G. E. McManis. A. N. Fletcher, and D. E. Bliss, US Pat. 4624753. 1986. 395 G. E. McManis, A. N. Fletcher, D. E. Bliss, and M. H. Miles, J. Appl. Elr~~trochem..1986, 16. 920. Ke rridge produced on metals, for example steels can be anodized in alkali metal nitrate- acetamide solutions.396 The other potential application of acetamide solutions under investigation is in thermal batteries. In one example zinc and silver(1) chloride were selected as the reactants with a 0.1 M zinc(1r) chloride in acetamide solution as the electro- lyte.' 073397 Alkali metal nitrate-acetamide solutions have also been considered for battery applications with iron and cobalt electrode^.^^^'^^*^^^ The solutions have been shown to cause very little corrosion, even to inexpensive steels.398 Such batteries with lithium anodes give high discharge rates when combined with many silver salts (13 Ag' compounds were tested) or with cerium(1v) as cathode^.^^^.^'^ Moderate additions of ammonium nitrate (acidic) to the acetamide-urea melts were helpful in depassivating the anode but highly acidic melts themselves reacted rapidly with the lithi~rn.~" AcknoM.ledgement.~-Grateful thanks are expressed to the University of Delhi and to the University Grants Commission, New Delhi, for appointment as Visiting Scientist (January-April 1985) at the Centre of Advanced Studies Chemistry Department, Dehli University; to my host Professor H.C. Gaur; and to the British Council for a travel grant. During this sabbatical term most of the initial literature survey for this review was made. 396 L. L. Antropov, D. A. Tkalenko, S. A. Kudrya, E. L. Kozlov, A. A. Rudnitskaya, N. M. Voropai, and N. A. Chmilenko, USSR Pat. 800249. 1981. 39i R. A. Wallace and P. F. Bruins, J. Electrochent. Soc., 1967. 114, 209. 39H T. M. Panasenko, S. A. Kudrya, and L. V. Yatsenko, Geliotekhnika, 1983, 43. 39y G. E. McManis, A. N. Fletcher, and D. E. Bliss, US Pat. 4624754, 1986. 4"o G. E. McManis. A. N. Fletcher, and D. E. Bliss, Elerrrochim. Ac/a, 1986, 31. 1271.

ISSN:0306-0012    

DOI:10.1039/CS9881700181

出版商:RSC    

年代:1988

数据来源: RSC

摘要:

Chern. SOC.Rev., 1988, 17, 229-267 Vinylcyclopropane Rearrangements By Z. Goldschmidt and B. Crammer DEPARTMENT OF CHEMISTRY, BAR-ILAN UNIVERSITY, RAMAT CAN 52100, ISRAEL 1 Introduction Vinylcyclopropanes (vcp) can undergo three fundamental types of bond-reorgan- ization processes (Scheme 1): (1) cis-trans isomerization; (2) ring opening to pentadienes; (3) ring enlargement to (a) cyclopentenes and (b) methylcyclobutenes. These rearrangement processes can occur by thermal, photochemical, and catalytic routes. RYR Ry Scheme 1 The first reported vcp rearrangement was the pyrolysis of 2,2-dichlorovinyl- cyclopropane (1) to a mixture of 4,4-dichlorocyclopentene (2), monochloro- cyclopentadiene (3), and unidentified chlorinated hydrocarbons, carried out by Vinylcyclopropane Rearrangements %'+ Y." + Scheme 2 Neureiter in 1959 (equation 1).' However, unobserved vcp rearrangements probably occurred as early as 1922 during the preparation of vcp by drastic methods2 On repeating Harper and Reed's acid-catalysed esterification of trans-chrysanthemic acid an additional product, the labile methyl ester of lavandulylic acid, was identified (Scheme 2).3 This review reflects the recent developments in vcp rearrangements under thermal, acid-catalysed, photochemical, and metal-promoted conditions.2 Thermal Rearrangements of Vinylcyclopropanes During the past two decades several reviews 4-1 relating to thermal rearrange- ments of vcp have appeared. The most recent and the most comprehensive relates only to thermal rearrangements of vcps to cyclopentenes.' ' (a) N.P. Neureiter, J. Org. Chem., 1959,24, 2044; (b)U.S. Patent No. 2,951,878 (1960); (c) US.Patent No. 2,981,756 (1961). 'N. J. Demjanow and M. Dojarenko, Chem. Ber., 1922,SSB, 2718. S. H. Harper and H. W. B. Reed, J. Sci. Food Agric., 1951, 414. [The mixture of esters obtained was analysed by n.m.r. and it was found that the methyl rrans-5-methyl-2-propen-2-yl-hex-3-enoatewas formed in 10% yield-B. Crammer (19SS)l. H. M. Frey, 'The gas phase pyrolyses of some small ring compounds), in Aduances in Physical Organic Chemistry, 1966,Vol. 4,Ed, V. Gold, Academic Press. S. Sarel, J. Jovell, and M. Sarel-Imber, Angew. Chem., Int. Ed. Engl., 1968,7, 5. Goldschmidt and Crammer Vcps can undergo two kinds of thermal reactions: cleavage of the cyclopropane ring o-bond followed by H-migration, and a formal 1,3-sigmatropic shift to cyclopentenes (Scheme 1, processes 2 and 3a).The ring opening process to the dienes (Scheme 1, process 2) occurs with activation energies E, 52-66 kcal mol-', which is characteristic of the cyclo- propane ring to the alkene process. Cyclopentene formation (Scheme 1, process 3) occurs with E, 31-63 kcal mol-'. Thermolysis of the parent vcp (4) under adiabatic conditions affords principally cyclopentene (5).129'3 Minor products that have also been isolated are cis-1,3-pentadiene (6), trans-1,3-pentadiene (7), isoprene (8), 1,4-pentadiene (9) and cyclopentadiene (10) (equation 2).Table 1 Cyclopentene formation from (1 1) Substituent * E, kcal mol-' ReJ lla R3 = Me 50.9 13 llb R' = Et 50.0 14 llc R' =Re = Me 54.6 15 lld R4 = Me 49.4 16 lle R3 = Pr 51.1 17 llf R3 = R4 = Me 50.5 15 Ilg R5 = R6= F 40.3 9,18 llh R5 = Ph 41.0 19 lli R4 = OMe 44.7 19 llj R5 = OMe 38.7 19 Ilk R' = CHSH, 44.5 20 111 R3 = OSiMe, 63.1 21 llm R5 = NMe, 31.1 22 lln R' = Me,R3 = Ph 55.0 23 * Remainder R = H (a) M. R. Willcott, 111, R. L. Cargill, and A. B. Sears, Prog. Phys. Org. Chem., 1972, 9, 25; (b) M. R. Willcott, 111and V. H. Cargle, J. Am. Chem. SOC.,1967, 89, 723.'R. Breslow, Molecular Rearrangements, 1963, 4, 233. H. M. Frey and R. Walsh, Chem. Rev., 1969,69, 103. W. R. Dolbier, Jr., Acc. Chem. Res., 1981, 14, 195.lo J. J. Gajewski, 'Hydrocarbon Thermal Isomerizations', in Organic Chemistry, A Series of Monographs, Ed. H. A. Wasserman, Academic Press, 1981, p. 5. T. Hudlicky, T. M. Kutchan, and S. M. Naqvi, Org. React. (N.Y.), 1985, 33, 247. R. J. Ellis and H. M. Frey, J. Chem. SOC.,1964, 3547. l3 H. M. Frey and D. C. Marshall, J. Chem. Sac., 1962, 3981. l4 R. J. Ellis and H. M. Frey, J. Chem. SOC.,1964,4188. C. S. Elliot and H. M. Frey, J. Chem. SOC.,1965, 345. I6 R. J. Ellis and H. M. Frey, J. Chem. SOC.,1964, 959. G. R. Branton and H. M. Frey, J. Chem. Soc., (A), 1966, 1342. W. R. Dolbier, Jr. and S. F. Sellers, J. Am. Chem. SOC.,1982, 104, 2494. l9 J. M. Simpson and H. G. Richey, Jr., Tetrahedron Lett., 1973, 27, 2545. 2o H. M. Frey and J.A. Krantz, J. Chem. SOC.,(A), 1969, 1159. B. M. Trost and P. H. Scudder, J. Org. Chem., 1981, 46, 506. 22 D. W. Shull, Diss. Abstr. B., 1976, 37, 775. 23 C. S. Elliot and H. M. Frey, J. Chem. Soc., 1964, 900. 231 Vinylcyclopropane Rearrangements Mono- and di-substituted vcps (1 1) undergo thermal rearrangement to substi- tuted cyclopentenes (12), Table 1. Vcps substituted by methoxy and amino groups (1 li,j,m) undergo thermal rearrangement to the corresponding cyclopentenes more readily than the alkyl substituted derivatives. l9 Alkenyl- and phenyl-substituted vcps (1 1h,k) likewise rearrange to cyclopentenes with lower activation energies. R3-substituted vcps (11a,e,l, and n) apparently hinder cyclopentene formation and require higher activation energies.(11) (12) Vinylcyclopropanol (13) on heating did not rearrange to the expected cyclopentanone but, instead, produced the 2-methylcyclobutanone (14) as the sole product (equation 3).24 cis-2-Methylvinylcyclopropane (15) rearranged on heating via a homo-f,5-hydrogen shift (retro-ene reaction) to give, in quantitative yield, cis-hexa- 1,4-diene (16).25 At higher temperatures the trans isomer (17) gave (16) as the main product [probably uiu initial &/trans isomerization to (15)] together with 8% of 4-methylcyclopentene (18).26 It was also found that the rearrangement of the difluorocyclopropane (1 9) to cis-3,3-difluoro- 1,4-hexadiene (20) occurred at even lower activation energies (equation 4).27 24 J.Ollivier and J. Salaiin, Tetrahedron Lelt., 1984, 25, 1269. 25 R. J. Ellis and H. M. Frey, Proc. Chem. Soc.. 1964, 221. 26 (a)R. J. Ellis and H. M. Frey, J. Chem. SOC.,1964,5578;(b)Personal communication from H. M. Frey: the 3-methylcyclopentene was incorrectly reported in reference 26a and the correct product should be 4- methylcyclopentene. S. W. Benson and H. E. O’Neal, ‘Kinetic Data on Gas Phase Unimolecular Reactions’, NSRDS-NBS 21, 1970, 239, U.S. Dept. of Commerce, Washington, D.C. ’’W. R. Dolbier, Jr. and S. F. Sellers, J. Org. Chem., 1982, 47, I. 232 Goldschmidt and Crammer X = H (15) X = H (17) X = F (19) (4) X = H (16) X = H (1 8) 8'10 X = F (20) X F (21) 66'10 When the 2-cis substituent is meth~xy,'~ f~rmyl,~'dimethylamino,28 a~etyl,~~ or ~henyl,~' the corresponding 3-substituted cyclopentenes are mainly obtained.It was observed that the retro-ene pathway generally competes with other rearrangements when the 2-cis substituted groups contain at least one hydrogen atom at the or-position, as in compounds (19, (17), and (19). Similarly the cis-(2- vinylcyclopropy1)carbinol (22) rearranges to cis-4-hexanal (24), presumably via a retro-ene pathway to (23) (equation 5).32,33 More drastic conditions are required when chrysanthemol (25) undergoes rearrangement to lavandulol (26) by a 1,Shydrogen shift from the cis-3-methyl (25) with no cyclopentene formation (equation 6).34 A. Mechanism of Thermal Rearrangements of vcp.-Although the mechanism of the vinylcyclopropane-cyclopentene thermal rearrangement has been studied in detail for more than twenty years, there is still no definite agreement on the question of concerted versus biradical mechanism of the reaction. The difficulty in distin- guishing between the two mechanistic pathways by kinetic studies lies in the fact ''H.G. Richey, Jr. and D. W. Shull, Tetrahedron Lett., 1976, 575. 29 P. Y. Bahural, A. Menet, F. Pautet, A. Poncet, and G. Descotes, Bull. SOC.Chim. Fr., 1971, 6, 2215. 30 E. Vogel, Angew. Chem., Int. Ed. Engl., 1963, 2, 1. 31 E. N. Marvel1 and C. Lin, J. Am. Chem. Soc., 1978, 100, 877. 32 A. F. Thomas and B. Willhalm, Tetrahedron Lett., 1964, 49, 3775. 33 L. H. Zalkow, D. R.Brannon, and J. W. Uecke, J. Org. Chem., 1964, 29, 2786.34 G. Ohloff, Tetrahedron Leti., 1965, 42, 3795. 233 Vinylcyclopropane Rearrangements that the activation energies for the rearrangements are not low enough to exclude a nonconcerted biradical mechanism. Thus the E, of the cyclopropane ring cleavage is about 65 kcal mol-', decreasing to ca. 50 kcal mol-' for vinylcyclopropane. This is close enough to support the involvement of the biradical intermediate (27), since the resonance stabilization energy of the ally1 radical is 13 kcal m~l-'.~' Further support for a biradical mechanism comes from a study of the stereomutation of labelled vinylcyclopropane. Thermolysis of anti-cis-[2,3,5-2H3]vinylcyclopropane (28) at 325 "C in the gas phase gave an equilibrium mixture of stereoisomers in a ratio independent of the extent of the reaction.This is consistent with the intervention of the intermediate biradical (29) (equation 7).35 It should be noted, however, that although cleavage of the weak allylic o-bond is the most plausible mechanism, the same kinetic behaviour would result if the reaction involved cleavage of the C(2)-C(3) bond, provided that the internal rotations in the biradical were fast relative to cy~lization.~~ Similarly, formation of a mixture of deuterated cyclopentenes (33) and (34) from ~inyl-[2-~H,]-cyclopropane (32) may result by regiochemical homolytic cleavage of the allylic bonds (equation Q6 The same biradical mechanism may explain the formation of 1,4-dienes from vinylcyclopropane 24,37 (equation 9). Since, however, the activation energies of these reactions are in the range of 3&35 kcal mol-' only, an alternative concerted mechanism may be invoked which involves the thermally allowed suprafacial 1,5-H migration (retro-ene reaction).Whereas kinetic studies may favour a biradical pathway for the vinylcyclo- propane-cyclopentene rearrangement, stereochemical analysis of this reaction suggests the occurrence of several concerted processes. Thus, pyrolysis of (+)-35 K. W. Egger, D. M. Golden, and S. W. Benson, J. Am. Chem. SO~.,1964, 86, 5420. 36 M. R. Willcott, 111 and V. H. Cargle, J. Am. Chem. SOC.,1969, 91, 4310. 37 (a) W. von E. Doering and K. Sachdev, J. Am. Chem. SOC.,1974,%, 1168; (b)ibid.,J. Am. Chem. SOC., 1975, 97, 5512. Goldschmidt and Crammer /= (8)4 "?(" (34) (1 S,2S)-trans,trans-2-methyl-1-propylcyclopropane (35) gave an optically active mixture of four 3,4-dimethylcyclopentenes (36)-(39) (equation The high degree of optical purity in the products eliminates the possibility of freely rotating 38 G.D. Andrews and J. E. Baldwin, J. Am. Chem. SOC.,1976, 98, 6705. Viny Icyclopropane Rearrangements biradical intermediates in the reaction. Consequently, it has been suggested that these 1,3-sigmatropic shifts occur by way of four competing concerted proces~es.~~,~~The processes allowed by the conservation of orbital symmetry are inversion-suprafacial (is) (36) and retention-antarafacial (ra) (38), whereas those allowed under control of subjacent orbitals are retention-suprafacial (rs) (37) and inversion-antarafacial (ia) (39). Similar results have been reported with the optically active isomers of 2-cyano-isopropenylcyclopropane.376 The mechanism of the degenerate vinylcyclopropane rearrangement in bicyclic compounds was elucidated using the optically active, labelled (-)-[2,2,3-'H3]-A3-thujene (40)."' Thermolysis of (40) resulted in both enantiomerization and rearrangements to the racemic mixture of [5,6,6-2H3]-A3-thujenes (equation 11).Determination of partitioning among the three products reveals that the major reaction pathway is forbidden by orbital symmetry. A high degree of stereo- chemistry is nevertheless observed in this reaction, suggesting the presence of a small (1.2 kcal mol-') barrier to conformational eq~ilibrium.~' The parent bicyclo[3.1.O]hexene (41) undergoes a thermal ring-expansion to a mixture of cyclohexadienes (42) and (43) uia the usual vinylcyclopropane-pentadiene pathway (equation 12).41 In larger bicyclic systems the rearrangement may take another course.For example bicycloC6.1 .O)non-2-ene (44) undergoes a reversible homo-[ 1,5]-sigma- tropic hydrogen shift to cis,cis-1,4-cyclononadiene (45) (equation 13).42*43The reversed process is the intramolecular, possibly concerted, ene reaction.44 The thermolysis of l-vinyl-spiro[2,4]hepta-4,6-diene (46) at 150 "C gave two products (47) and (48), resulting from an initial 1,3-sigmatropic rearrangement (equation 14)."' B. Thermal vcp Rearrangements under Basic Conditions.-If an auxiliary group which has an acidic hydrogen at the =-position is attached to the cyclopropane ring, then treatment with base will produce an anion which is capable of accele- rating the thermal vinylcyclopropanesyclopentene rearrangement.Thus, the lithium salts of 2-vinylcyclopropanols (49) undergo a facile rearrangement at 25 "C to the corresponding 3-cyclopentenols (50) (equation 1 9."' 39 W. R. Dolbier, Jr., 0.T. Garza, and B. H. Al-Sader, J. Am. Chem. SOC.,1975. 97, 5038. 40 W. von E. Doering and E. K. G. Schmidt, Tetrahedron, 1971, 27. 2005. 41 R. J. Ellis and H. M. Frey, J. Chem. SOC.,(A), 1966, 553. 42 D. S. Glass, R. S. Boikess, and S. Weinstein, Tetrahedron Lett., 1966, 999. 43 R. B. Woodward and R.Hoffmann, 'The Conservation of Orbital Symmetry', 1970, Verlag Chemie, Weinheim. 44 H. M. R. Hoffmann, Angew. Chem., In[. Ed. Engl., 1969, 8, 556. 4s R. L. Danheiser, C. Martinez-Davila, and J. M. Morin, J. Org. Chem.. 1980, 45, 1340. Goldschmidt and Crammer A A (40) (-1 -I 2,2,3 -Hgl (-1-5,6,6 -* H, (+I-t 5,6,6 -2 H, I (+I -[2,2,3-*H31 (-)-A-Thujene R = Me,Bu', Ph (49) (50) The starting cyclopropanols are usually obtained without isolation by treating the corresponding 2-chloroethyl ethers with Bu"Li in THF-hexane-HMPT. The rearrangement generally proceeds with remarkably high ~tereoselectivity.~~ For 46 R. L. Danheiser, C. Martinez-Davila, R. J. Auchus, and J. T. Kadonaga,J. Am. Chem. SOC.,1981,103,2443. 237 Vinylcyclopropane Rearrangements example, addition of 2-(ch1oroethoxy)carbene to cycloheptadiene gave a mixture of the exo and endo bicyclic isomers (51), which rearrange to the endo-bicyclo[3.2.1] octan-8-01 (52) as the only product (equation 16).45 This is in agreement with a mechanism of either a concerted 1,3-sigmatropic shift, or a stepwise pathway in which the intermediate 45 cyclizes faster than conformational interconversion.a-Sulphonyl carbanions also accelerate vinylcyclopropane rearrangements stereo~electively.~~Thus, exposure of the 1-[(phenylsulphonyl)methyl]-2-vinylcyclopropane (53), obtained from trans,trans-2,4-hexadiene,to 1.2 equivalents of Bu"Li in 5:1 THF-HMPT at -78 "Cprovided the corresponding lithium salt, which rearranged smoothly upon warming to -30 "C to afford the transpans- dimethylsulphone (54)as the only product.This sulphone was trapped by 4-methyliodopent-3-ene, affording, after desulphonylation, the trisubstituted cyclopentene (55) (equation 17). A related vinylcyclopropane-cyclopentene ring enlargement is observed when 2,2-dibromovinylcyclopropane (56) is treated with MeLi at -78 OC, affording cyclopentadiene (57) and the allene 1,2,4-pentatriene (58) (equation 18).48,49 The reaction is believed to proceed by initial formation of a vinylcyclopropylidene intermediate which rearranges either to the allene (58) or to the intermediate 0-(55) 47 R. L. Danheiser, J. J. Bronson, and K. Okano, J. Am. Chem. SOC.,1985, 107,4579. '* L.Skattebd, Tetrahedron, 1967, 23, 1107. 49 R. Brun, D. S. B. Grace, K. H. Holm, and L. Skatterbd, Actu Chem. Scand., Ser. B, 1986, 40, 21 M. S. Baird and I. Jefferies, Tetrahedron Lett., 1986, 22, 2493. 238 Goldschmidt and Crammer Br Br (56) (59) +T +H+ ___) (-ti+ C02Me (60) (20) C02Me C02Me (61) (62) C0,Me C02Me (63) (62a) Vinylcyclopropane Rearrangements cyclopentenylidene. The latter subsequently produces cyclopentadiene (57) by a 1,2-hydrogen shift. This mechanism has been confirmed recently by labelling studies.50 3 Acid-catalysed Rearrangements of vcps Protonation of vinylcyclopropane leads to the formation of the cyclopropane- stabilized carbonium ion (59).5 This cyclopropylcarbinyl cation may undergo one of the following rearrangements, before a deprotonation or reaction with nucleophiles (usually the solvent) takes place (equation 19): (a) cyclopropylcarbinyl-homoallyl rearrangement, (b) cyclopropylcarbinyl-cyclobutylrearrangement, and (c) cyclopropylcarbinyl~yclopropylcarbinylrearrangement. The chemistry of cyclopropanes, including vinylcyclopropanes, has been reviewed *re~ently.~ We shall mainly focus on acid-catalysed rearrangements of vinylcyclopropanes under non-solvolytic conditions. Chrysanthemyl derivatives are perhaps the best known vinylcyclopropane substrates used for studying the cyclopropylcarbinyl-homoallyl rearrangement. The special interest in these compounds stems from their possible biogenetic relationship with irregular acyclic terpenes, squalene, and higher terpenoid~.~~ Methyl trans-chrysanthemate (60) rearranges at room temperature, in 50% H,SO,-heptane, to a mixture of lavandulyl esters (61)-(63).54 The kinetic and deuterium exchange studies reveal that (61) and (62) are the primary products of the reaction, obtained from cyclopropylcarbinyl cation (60a) by deprotonation and water addition, respectively.Lavandulyl alcohol (61) and diene (62) equilibrate via the tertiary homoallylic carbonium ion (62a), which ultimately deprotonates to the isomeric diene ester (63) (equation 20). Further confirmation for this mechanism comes from the Lewis acid catalysed rearrangement of cis-ethyl chrysanthemate Lewis acid7 I MeS03H I C0,Et C02Et (64) (65)+ (21) CO,Et C0,Et (67) (66) 51 (a) H. G.Richey, in ‘Carbonium Ions’, Vol. 111, Ed. G. A. Olah and P. von R. Schleyer, Wiley- Interscience, New York, 1972, 1201. 52 For a recent review see, J. Salaun, ‘Rearrangements involving the Cyclopropyl Group’ in ‘The Chemistry of the Cyclopropyl Group, Part I’, ed., Z. Rappoport, Wiley, Chichester, 1987, Ch. 13. 53 R. M. Coates, Progress in the Chemistry of’Organic Natural Products, 1976, 33, 73.’’Z. Goldschmidt, B. Crammer, and R. Ikan, J. Chem. Soc., Perkin Trans. I, 1984, 2697. Goldschmidt and Crammer (64),in aprotic solvents, to the trans-isomer (65) and the two lavandulyl esters (66) and (67) (equation 2l).’’ It was also established that the chiral carbon of (64) retains its configuration during the rearrangement of optically active ethyl chrysanthemate with methanesulphonic acid as catalyst.’ ’ Treatment of trans-chrysanthemic acid (68) with pyridine hydrochloride at 210 “C gave a complex mixture of lavandulyl, artemisyl, and santolinyl derivatives, which result from cleavage of either one of the three cyclopropane bonds (equation 22).56 CO,HPy-HCIJ210 “C (G8) Y + This loss of specificity, achieved under drastic conditions, may result from competitive reaction pathways following protonation at either the double bond or the carboxylic group. Thus protonation at the double bond will lead to the lavandulyl derivative (69) by C( 1)-C(3) cleavage, whereas protonation at the carboxylic group will result in cleavage of either the C(l)-C(2) bond to the artemisyl derivative (70), or the C(2)-C(3) bond to form santolinyl derivatives (71) and (72). Specific protonation of the carbonyl group has been reported in the chrysanthemic aldehyde (73a) and ketone (73b) to give the corresponding artemisia dienal (74a) and dienone (74b) (equation 23),” and the thiochrysanthemate (75) rearranged under acid catalysis to the thiosantolina diene ester (76) (equation 24).’* Other mechanistic pathways, such as the direct protonation of the cyclo- propane ring and thermal 1,5-H migration from the gem-dimethyl groups have also been ~uggested.’~ Chrysanthemol (25), like the corresponding carbonyl compounds, rearranges following a preferential protonation at the carbinol functional group, to give (after ’’G.Suzukamo, M. Fukao, and M. Tamura, Tetrahedron Lett., 1984,25, 1595. ’‘ D. A. Otieno, G. Pattenden, and C. R. Popplestone, J. Chem. SOC.,Perkin Trans. 1, 1977, 196. 24 1 Vinylcyclopropane Rearrangements dehydration) the analogous artemisia triene (77)57,59(equation 25). When, however, the carbinol is allylic (78), protonation occurs at the allylic position, producing the lavandula triene (79) (equation 26).57 More recently, the solvolytic rearrangements of chrysanthemol derivatives (80) have been studied as a model of the biosynthesis of non-head-to-tail (irregular) 7-[ “OH 1-m’(23)COR (73a) R = H (74a) (73b) R = Me (74b)y+y]ir^e-+COSPr PrS ‘OH COSPr (241 (75) (76)TH%[T]+?(25) (25) (77) OHYH&[9q-yOH (26) (78) (79) terpenes.60961A plethora of products was observed depending on the leaving group and the solvolytic conditions (equation 27).For example, hydrolysis of N-methyl-4-chrysanthemyloxypyridiniumiodide (80, X = OPyI) gave, in addi- tion to chrysanthemol (25) and the expected artemisyl derivatives (8 1)-(83), santolinyl triene (84)and dienol (85). The interesting head-to-head dienol (86) was also detected in trace amounts and is believed to be obtained via a cyclopropylcarbinyl-cyclopropylcarbinyl rearrangement.60 57 L. Crombie, P. A. Firth, R. P. Houghton, D. A. Whiting, and D. K. Woods, J. Chem. Soc., Perkin Trans. i, 1972, 642. 58 N. F. Elmore, J. E. Roberts, and G.A. Whitham, J. Chem. Res. (S), 1985, 98. 59 C. D. Poulter, J. Am. Chem. Soc., 1972, 94, 5515. 6o C. D. Poulter, L. L. Marsh, J. M. Hughes, J. C. Argyle, D. M. Satterwhite, R. J. Goodfellow, and S. G. Moesinger, J. Am. Chem. Soc., 1977, 99, 3816. Goldschmidt and Crammer X OPyI (80) (811 + (84)+ The presence of a carbonyl group at position 2 of the vinylcyclopropane, which can react with the acid catalyst, opens up a new rearrangement route-the incipient homoallylic cation can reclose to give a 3-cyclopentenylcarbinyl intermediate. For example, arylvinylcyclopropane-2-carbonylchloride (87) rearranges in the presence of a Lewis acid such as BBr, or AlCl, to the corresponding 2-arylcyclopent-3-enecarbonyl chloride (88) (equation 28).62 A similar pathway was utilized by Corey and Myers63 in the synthesis of the plant hormone Antheridiogen-An (equation 29).Here a lactone vinylcyclopropane (89) underwent this rearrangement pattern to the crucial lactone cyclopentene (90).63 Yet another 61 C. D. Poulter and J. M. Hughes, J. Am. Chem. Soc., 1977, 99, 3830. 62 Y. Sakito and G. Suzukamo, Chem. Lett., 1986, 621. 63 E. J. Corey and A. G. Myers, J. Am. Chem. SOC.,1985, 107, 5574. Vinykyclopropane Rearrangements interesting variation of this rearrangement is exemplified in vinylcyclopropane dicarboxylate (91), which gave vinyl butyrolactone (92) when treated with the strong Lewis acid bis(trimethylsily1) sulphate as the catalyst (equation 30).64 The initially formed homoallylic carbonium recloses on the oxygen of the other carboxylate group, to form the lactone (92), instead of the a-carbon which would lead to cyclopentene (equation 30).Cyclopropylcarbinyl~yclobutylring expansions in acid-catalysed vinylcyclo- propane rearrangements are facilitated by the presence of oxygen-bonded groups at the allylic position 1. Thus the parent system 1-vinylcyclopropanol (13) undergoes a pinacol-pinacolone type of rearrangement to cyclobutanone (14) in the presence of p-TsOH or HBr (equation 31).66 Similarly, Ollivier and Salaiin showed that the diol (93) rearranges in the presence of a Lewis acid catalyst, with dehydration, to 2-vinylcyclobutanone (94) (equation 32).24 Analogous reactions 64 Y. Morizawa, T. Hiyama, and H.Nozaki, Tetrahedron Lett., 1981, 22, 2297.‘’ K. B. Wiberg, B. A. Hess, and A. J. Ashe, ‘Carbonium Ions’, Vol. 111, Ed. G. A. Olah and P. von R. Schleyer, Wiley-Interscience, New York, 1972, p. 1295. 66 H. H. Wasserman, R. E. Cochoy, and M. S. Baird, J. Am. Chem. Soc., 1969, 91, 2375. Goldschmidt and Crammer %OH BF3’Et20) < (32) WOOHI(33)VO (acac l23(95) R (96) (34)eer X = Halogen (351I SOH = Solvent (36) CHMe *pi-dHMeOAC-(99) Me I cis and trans cis and trans (101) (100) 245 Vinylcyclopropane Rearrangements with THP 67 and thioaryl 68 ethers have also been reported. A modification of these reactions for the synthesis of spiro-cyclobutanones was reported by Trost and M~o,~' using electrophiles other than proton or Lewis acids.Treatment of vinylcyclopropanol(95) with Bu'OOH in the presence of VO(acac), afforded the hydroxy-spiroketone (96) (equation 33).69 Analogously, reaction of vinylcyclopropane silyl ether (97) with bromine gave 5-bromo-spiro[3,4]octan-1-one (98) (equation 34).69 An exceptional case of cyclopropylcarbinykyclobutyl rearrangement is observed when the 4-halogenovinylcyclopropanes(equation 35, X = halogen, SOH = solvent) is solvolysed in the presence of Ag+. The solvolysis of either cis or trans iodovinylcyclopropane (99) with AcOH-AgOAc is non-specific, giving, in addition to both the cis and trans isomers of (99), a mixture containing acetate derivatives of vinylcyclopropane (loo), 2-methylene-cyclobutanols (lOl), cyclo- butene carbinol(102), and the open chain allenol acetate (103) (equation 36).70 The migratory aptitude of the o-bond as a function of the substituents was revealed from studies of the hydrolysis of 1,2-dimethyl-4-chlorovinylcyclopropane(104) (equation 37)." Of the complex mixture of cyclobutanols (106)-(109) obtained in the reaction, more than 84% were derived from the 2,4-dimethyl-methylenecyclo-butyl carbonium intermediate (105a), indicating a regioselective migration of the more substituted cyclopropane C(2).6' J. P. Barnier and A. J. Salaun, Tetrahedron Lett., 1984, 25, 1273. B. M. Trost and L. N. Jungheim, J. Am. Chem. Soc., 1980, 102, 7910. 69 B. M. Trost and M. K. Mao, J. Am. Chem. Soc., 1983, 105, 6753. 'O D.R. Kelsey and R. G. Bergman, J. Am. Chem. SOC.,1971,93, 1941. Goldschmidt and Crammer 4 Vinylcyclopropane Photorearrangements The cyclopropane ring behaves as a moderate auxochrome, causing a 8-15 nm bathochromic shift of the olefinic absorption band. The parent vinylcyclopropane (vcp) (4) and its alkylated derivatives all have absorption maxima for the n -n* olefin band in the 189-201 nm region.72 It is therefore not surprising that the majority of vinylcyclopropanes whose photochemistry has been studied, bear additional chromophoric functional groups which extend the electronic absorption to the experimentally more accessible region of >254 nm.73 A limited number of photochemical studies of the parent vcp in the gas phase have nevertheless been rep~rted,~~,~~and the photorearrangements of simple alkyl derivatives, in solution, have also been de~cribed.~~.~~ The gas phase vacuum-u.v.photolysis of vcp, using xenon (147.0 nm), krypton (123.6 nm), and argon (106.7-104.8 nm) resonance excitation, gave only fragmentation In contrast, the infrared multiphoton irradiation of gaseous vcp, with a CO, TEA laser, led to a mixture containing mainly two isomers, 1,4-pentadiene (9) and cyclopentene (5).75 A substantial amount of cyclopentadiene (10) was also formed, presumably uia dehydrogenation of ‘hot’ cyclopentene. This fragmentation process competes with the collisional de- activation of excited (5) to the ground state (equation 38). Isopropenylcyclopropane (1 la) is the simplest vcp derivative which has been photolysed in solution.76 Upon direct irradiation in hexane, (1la) rearranges to methylcyclopentene (12a) in 55% yield, with a zero-order rate of 1.1 x lop3Mh-’ (equation 39).The photosensitized reactions of (1 la) with either benzene or acetone gave only traces of (12a) and none was observed with naphthalene as sensitizer. Thus (12a) is derived from the singlet excited state of (lla). Direct far-u.v. irradiation (185-228 nm) of 2-norcarene in pentane solution gave a mixture of isomeric At 214 nm seven isomers were identified, (111)-(117), which constituted 91% of the mixture. The quantum yield for total product formation (at 185 nm) was CD = 0.26 f0.02. A similar mixture of the 71 M. Santelli and M.Bertrand, Tetrahedron, 1974, 30, 243. 72 C.H.Heathcock and S. R. Poulter, J. Am. Chem. SOC., 1968,90, 3766. 73 S. S. Hixon, Org. Photochem., 1979,4,218;T. Hudlicky, T. H. Kutchan, and S. M. Naqvi, Org. React., 1985, 33, 247. 74 E. Lopez and R. D. Doepker, J. Phys. Chem., 1979, 83, 573. 75 W.E.Farneth, M. W. Thomsen, N. L. Schultz, and M. A. Davies,J. Am. Chem. SOC., 1981,103,4001;W. E. Farneth, M. W. Thomsen, and M. A. Berg, J. Am. Chem. SOC., 1979, 101, 6468. 76 R.S. Cooke, J. Chem. SOC., Chem. Comrnun., 1970, 454. l7 W.J. Leigh and R. Srinivasan, J. Am. Chem. SOC., 1983,105,514. Vinylcyclopropane Rearrangements corresponding dideuterated isomers was obtained when [7,7-'H2]-2-norcarene was irradiated under the same conditions.The deuterium position in the products is shown in equation 40. Photolysis of (1 10) in the presence of naphthalene as sensitizer did not change the product mixture, indicating that all the products were derived from the singlet excited state. However, isomers (1 1 1) and (1 13) are also products of the triplet state, arising from toluene-sensitized photolysis of (110) in deoxygenated pentane solution (equation 41). Prolonged irradiation resulted, as expected, in cis-trans isomerization of (1 1 1). Mechanistically, it is obvious that except for (114) all the observed re-arrangements proceed by cleavage of either the internal C( 1)-C(6) cyclopropane bond of (1 lo), or the allylic C(l)-C(7) external bond. However, in the absence of stereochemical labels in (1 10) it is impossible to distinguish whether the reactions are concerted or stepwise.In a stepwise mechanism, the initial intermediates formed are the diradicals (1 18) and (1 19). Radical (1 18) may disproportionate by either a sigma-bond cleavage to (1 11) or allylic coupling to (1 13). Alternatively, 1,2-H migrations may lead to either (1 16) or (1 17). In a concerted pathway, (1 1 1) results from a o2 + o2 + 7c2 cycloreversion, and (113) by an alkyl 1,3-sigmatropic shift.78 Similarly, (1 12) is obtained either from diradical (1 19) by recombination, or concertedly via 1,3-alkyl migration, whereas (1 15) may be derived by a 1,2-H shift. There remains the exceptional rearrangement of (110) to allene (114). Since no Goldschmidt and Crammer intermediates could be trapped, it was concluded77 that (114) is formed concertedly via a formal 1,2-migration of the vinyl hydrogen.Alternatively, a short- lived carbene intermediate (1 20) may first develop, which subsequently rearranges to (114) by a 1,2-H shift. 7 Except for a few cases, diradicals, analogous to (118)-(120) represent the three major mechanistic pathways which account for most photochemical vcp rearrangements. Clearly, however, the ratio between the products is expected to depend heavily on thenature ofthesubstituents,and usually also on themultiplicity of the reactive excited states (vide infra). In addition, fragmentation products are often observed. These result from cleavage of the cyclopropane ring to vinyl carbene and olefin.There are two additional competing processes for the excited state, which cannot be observed in the absence of appropriate substitution. These are the cis-trans isomerizations about an acyclic double and about the three- membered ring.8 1-89 Vcp esters undergo geometrical isomerization at least 10 78 R. B. Woodward and R. Hoffmann, ‘The Conservation of Orbital Symmetry’, Verlag Chemie, Weinheim, 1970. 79 M. J. Jorgenson, J. Am. Chem. Soc., 1969,91, 6432. P. H. Mazzocchi and R. C. Ladenson, J. Chem. Soc., Chem. Comm., 1970, 469. 81 H. Prinzbach, H. Hagemann, J. H. Hartenstein, and R. Kitzing, Chem. Ber., 1965, 98, 2201. W. H. Pirkle and G. F. Koser, Tetrahedron Lett., 1968,129;W. H. Pirkle, S. G. Smith, and G.F. Koser, J. Am. Chem. SOC.,1969, 91, 1580. 83 H. E. Zimmerman and G. E. Samuelson, J. Am. Chem. Soc., 1969,91, 5307. 84 D. L. Garin and K. 0.Henderson, Tetrahedron Lett., 1970, 2009. 85 H. E. Zimmerman and G. A. Epling, J. Am. Chem. Soc., 1972, 94, 8749. 86 T. Sasaki, S. Eguchi, and M. Ohno, J. Org. Chern., 1968, 33, 676; J. Org. Chem., 1970, 35, 790. 87 K. Ueda and M. Matsui, Tetrahedron, 1971, 27, 2771. 88 M. J. Bullivant and G. Pattenden, Pyrethrum Post, 1971, 11, 72. L. 0.Ruzo and J. E. Casida, J. Chem. Soc., Perkin Trans. 1, 1980, 728. 249 Vinylcyclopropane Rearrangements times faster than irreversible product formation (equations 42 and 43).” Similarly, styryl cyclopropanes undergo cis-trans isomerization faster than rearrangement to cyclopentenes (equation 44).*’Interestingly, however, the esters undergo both isomerization and rearrangement to cyclopentenes upon triplet +LC0,Et (122) + (42) c02et (125) (124) (123) d (126) CO,Et + CO,E t ?!I+ 4A (129) (128) tx+Tco2EtC0,Et Me (130) (131) + (133) (132) + ,COOEt Goldschmidt and Crammer sensitization, whereas the styryl cyclopropanes only isomerize under these conditions.Upon direct irradiation, esters afford a variety of rearrangement products, including ring enlargement to cyclopentenes (121) and (127), double bond migration to (122) and (128), bicyclo[2.1.O]pentane formation (129) 90 and ring cleavage to diene-esters (125), (131), (132), and (133). In addition, fragmentation of the cyclopropane ring to ethylene and carbene gave cyclopropenes (123) and (130) 2-ethoxyfuran (124), and allene ester (134).The occurrence of cis-trans isomerization during ring cleavage has been reported in a variety of chrysanthemyl derivatives.86*88 Photochemical isomerization of the natural (+)-trans-chrysanthemic acid (135) gave a racemic mixture of all four possible diastereomers, presumably via the triplet excited state.89 This involves a homolytic cleavage of the 1-3 sigma bond, followed by rotation and recombination (equation 44). Ring closure may also occur on oxygen, to give lactone (136). Alternatively, a hydrogen shift takes place in chrysanthemyl alcohol (137) to give the lavandulol (1 38) (equation 45).86 Similarly, irradiation of spirocyclo- hexadienone (1 39) at 254 nm resulted initially in cis-trans isomerization, followed R' @+ Ph R2 \ R2 I I < CO, H CO2H C02H (135) (45) (137) (138) 90 H.Kristinsson and G. S. Hammond, J. Am. Chem. SOC.,1968, 89, 5970. 25 1 Vinylcyclopropane Rearrangements by photochemical 1,2-migration ofeither hydrogen (140) or methyl (141) (equation 46).82 When (139) is irradiated at lower energies (350 nm) or sensitized by acetophenone, only cis-trans isomerization is observed. Extensive studies have been carried out on the multiplicity dependent photo- chemistry of condensed cyclopropane systems having either an exocyclic83 (equation 47) or endocyclic 85 (equation 48) double bond. The ratio between cis-trans isomerization (142) #(143), hydrogen migration (144) and phenyl migra- tion (145) in 2-methylenebicyclo[3.l.0]hexanes is strongly dependent on the multiplicity of the excited state.Similarly, a multiplicity dependence was observed in the cis-trans isomerization (146) -(147) us. the 1,3-sigmatropic migration (147) e(148) and (149) --+ (146) of bicyclo[3.1.0]hex-2-enes (equation 48). 0 0 (1 39) 0 + b \I] (1411 (140) Ph bPh Ph (144) (142) @PhPh I Ph 252 (145) (143) Goldschmidt and Crammer In general, cis-trans isomerizations are favoured from the triplet excited state, 1,3-sigmatropic rearrangements occur from both the singlet and triplet excited states, and hydrogen migrations take place mainly from the singlet state.The selectivity of the 1,3-sigmatropic shift was recently utilized as the key step in the efficient synthesis of grandisol (152), the boll weevil sex pheromone, from (+)-carene (150) (equation 49).’l The photoproduct (151) proved to be the racemic mixture, indicating the involvement of a triplet diradical intermediate. By extending the vcp system with an additional chromophore, such as an olefinic or a carbonyl group, a substantial change in the reaction course may occur, especially when the chromophore absorbs in the excitation region. Also, the nature of the excited state (e.g. n7c* or m*)may often affect the chemoselectivity. One striking example is revealed when the photochemistry of bicycloC3.1 .O]hexen-2- ones is compared with that of the corresponding methylene analogues (homo- fulvenes).The ketone (1 53) undergoes rearrangement to cyclohexadienone (1 55) and phenol (156) (equation 50) presumably via the zwitterionic intermediate (154).92p95In contrast, homofulvene (157) rearranges mainly to a mixture of 91 H. R. Sonawane, B. S. Hanjundiah, and M. U. Kumar, Tetrahedron Lett., 1984,25,2245; 1985,26,1097. 92 P. J. Kropp, J. Am. Chem. Soc., 1964, 86, 4053; Org. Photochem., 1967, 1, 1. 93 H. Durr and P. Heitkamper, Liebigs Ann. Chem., 1968, 716, 212. 94 H. Hart and D. C. Lankin, J. Org. Chem., 1968, 33, 4398. 95 H. Perst, Tetrahedron Lett., 1970, 3601. 253 Vinylcyclopropane Rearrangements isomeric spiro cyclopentadienes (158) and (159), uia a walk process, termed the ‘bicycle rearrangement’ 96-g8 (equation 5 1).Analogous walking processes were also reported in the norcaradiene 99 (equation 52) and the bicycloC4.1.1 .]octa-2,4-diene systems (equation 53). O0 On the other hand, cleavage of vinylcyclopropanes to aromatic systems was reported in the photochemical interconversion of the strained benzvalenes to benzenes (equation 54).’ O 1,1O2 Finally, an interesting case is reported where a change in the relative position of the additional chromophore about the vcp group led to the same formal reaction, but from a different excited-state manifold. Thus, irradiation of bicyclo[5.1.0]oct-5- ene-2-one or its 2-methylene analogue resulted in vcp-cyclopentane rearrangement from the corresponding singlet n,x* and x,x* excited states (equation 55).However, the isomeric bicyclo[5.l.0]oct-2-ene-4-oneand its 4-methylene analogue undergo (153) (154) J0 (156) (155) Ph Ph Ph Ph (157) (1 5 8)-anti (159)-syn 96 H. E. Zimmerman, Chimia, 1982, 36, 423. 97 H. E. Zirnrnerrnan, D. F. Juers, J. H. McCall, and B. Schroder, J. Am. Chem. Soc., 1971,93, 3662. 9B T. Tabata and H. Hart, Tetrahedron Lett., 1969, 4429. 99 M. Regitz, Angew. Chem., Inl. Ed. Engl., 1975, 14, 222. loo W. T. Borden, J. G. Lee, and S. D. Young, J. Am. Chem. SOC.,1980,102,4843. Io1 M. G. Barlow, R. H. Hazeldine, and R. Hubbard, J. Chem. Soc., C, 1970, 1232. lo’ K. E. Wilzbach and L. Kaplan, J. Am. Chem. SOC.,1965, 87, 4004. Goldschmidt and Crammer the same formal 1,3-sigmatropic shift from the triplet n,n* and n,n* states, respectively '03 (equation 56). cF3 hV (54)Q R6 R6 R = CF3 5 Metal-promoted Rearrangements of Vinylcyclopropanes Metal-promoted vinylcyclopropane rearrangements can be divided into three categories: (a) metal-cataiysed rearrangements, in which the metal is bonded to the organic molecule only during intermediate stages of the reaction; (b) stoicheio- metric rearrangements of metal-complexed vcp yielding coordinated isomeric products; and (c) metal-induced rearrangements of non-coordinated vinylcyclo- propanes giving metal complexed isomers.Reactions of the type (b) and (c) are frequently useful for mechanistic studies of metal-catalysed rearrangements of vcp.They may also be utilized for synthetic purposes, if the metal fragment is removed subsequent to rearrangement."~'04~'05 The major pathways of metal-catalysed rearrangements of vcp involve ring cleavage followed by hydrogen migration, presumably via a metal hydride inter- mediate. These reactions closely resemble the metal-catalysed isomerizations observed in saturated small-ring compounds but differ from the vcpcyclo- pentene rearrangements that are usually encountered in the absence of a catalyst. A wide variety of metals has been utilized to induce vcp rearrangements. These usually belong to late transition metal groups. Simple vinylcyclopropanes have been shown to isomerize readily in the presence of catalysts such as lo3 L. A. Paquette, G.V. Meehan, R. P. Henzel, and R. F. Eizember, J. Org. Chem., 1973, 38, 3250. Io4 S. Sarel, Acc. Chem. Res., 1978, 11, 204. Io5 H. Alper, Isr. J. Chem., 1981, 21, 203. lo' K. C. Bishop, 111, Chem. Rev., 1976, 76, 461. Vinylcyciopropane Rearrangements X X=O,CH, XBX = O,CH, Rhlv-+ (58) + (Bu",P),NiC1,/Bui,AIC1 and [(o-tolyl),P],Ni(C,H,),/HCl (equation 57) lo' and [Rh(CO),Cl], (equation 58).'08 With the Ni catalysts, the same mixture of conjugated dienes is obtained from both the cis and the trans vcp isomers. The Rh' catalyst gave a mixture of conjugated and non-conjugated dienes, all of which were primary products of the reaction. However, substitution in the vcp ring with a phenyl or carbethoxy group led only to the corresponding conjugated dienes lo' P.A. Pinke, R.D. Stauffer, and R.G. Miller, J. Am. Chem. SOC.,1974, 96, 422. Goldschmidt and Crammer (equation 59).'08 The latter reaction could also be induced by Pt" and Rh,(OAc), catalysts.'0g R R = Ph,C02Et In both simple and condensed vcp systems cis-trans isomerization has been shown to be faster than rearrangement.'O* The exo isomer (equation 60) is more reactive than the endo. The reaction is therefore not concerted, and both isomers give essentially the same mixture of cyclohexenyl propenes. However, the re- arrangement is highly selective, providing only the three non-conjugated dienes shown in equation 60, which neither interconvert nor rearrange to other isomers under the reaction conditions.Rh'-catalysed rearrangements also occur when the vcp double bond is endo- cyclic, providing non-conjugated cyclic dienes in which the bridging cyclopropane sigma bond is preferentially cleaved (equation 6 1). It appears that the typical metal-catalysed rearrangements of vcp proceed in a stepwise mechanism which involves initial double bond coordination to the metal, R. G. Salornon, H. F. Salornon, and J. L. C. Kachinski, J. Am. Chem. Soc., 1977, 99, 1043. M. R. Doyle and D. van Leusen, J. Org. Chem., 1982, 47, 5326. L. A. Paquette and M. R. Detty, Tefrahedron Lett., 1978, 713. Vinylcyclopropane Rearrangements followed by cyclopropane ring cleavage to a a,n-allylic intermediate. Subsequent intramolecular hydride shift, presumably uia a metal hydride complex, affords the products (Scheme 3).Scheme 3 Metal-catalysed valence isomerizations often compete with rearrangements via hydride migration when an additional double bond is introduced into the vcp molecule. In these cases, the familiar 1,3-sigmatropic ring enlargement to cyclopentene, observed in thermal and photochemical rearrangements of free vcp, predominates. A variety of catalysts has been used to induce this reaction, including Nio (equation 62),' ' Ni" (equation 63),' ' Rh' (equations 64-65),' '3*1l4 and Pdo (equation 66).' ' Coordinated vinylcyclopropanes, like their free counterparts, undergo molecular rearrangements with or without hydrogen migration. As in catalysed reactions, hydride shifts are likely to occur when no additional unsaturation is present, whereas valence isomerizations predominate in highly unsaturated vinylcyclo- propanes. However, since the majority of stable organometallic complexes have more than one unsaturated functional group coordinated to the metal, the number of cases where hydride migrations occur is small.A unique example of a 1,3-H migration, which leads to the deconjugation of a vcp function, has been reported for (q2-9,9-dichlorobicyclo[5.1.O]oct-2-ene)Fe(CO), (equation 67).'16 When an additional double bond is present in the q4-bicyclic ''I M. Murukami and S. Nishida, Chem. Lett., 1979, 927. W. von E. Doering and W. R. Roth, Tetrahedron, 1963, 19, 715. R. Grigg, R. Hayes, and A. Sweeney, J. Chem.SOC.,Chem. Commun., 1971, 1248. '14 T. Hudlicky, T. M. Kutchan, F. J. Koszyk, and J. P. Sheth, J. Org. Chem., 1980, 45, 5020. 'I5 Y. Morizawa, K. Oshima, and H. Nozaki, Isr. J. Chem., 1984, 24, 149. J. C. Barborak, L. W. Dasher, A. T. McPhail, J. B. Nichols,and K. D. Onan, Inorg. Chem., 1978,17,2936. Goldschmidt and Crammer R R = atkyt , cyclopropyl , Ph SIlica + (-JJ I ;1 0 R = CO,Et Fe(CO), complex, a circumambulant degenerate rearrangement takes place at 75 "C involving hydrogen, carbon, and metal migrations (equation 68)' ' In each reaction step, an exchange between a sigma-bonded carbon or hydrogen, and a metal fragment occurs over a chain of five conjugated carbon atoms. These reactions may thus be termed sigmahaptotropic rearrangements of the order [5,5].* At higher temperatures, an additional rearrangement involving hydrogen migration takes place, giving q4-(bicyclo[4.2.0]octa-2,4-diene)Fe(CO)3. The isomeric o,n-allylic complex (equation 69) similarly undergoes a degenerate rearrangement at low temperature (40 "C), and a skeletal rearrangement at higher temperatures (6@-70 OC)."' Formal [5,5]-sigmahaptotropic rearrangements of (q6-bicyclo[6.1.0]nona-triene)M(CO), (M = Cr, Mo, W) have been reported (equation 70).' 199120Recent work 12' on the Moo complex has shown that the starting material has the cyclopropane ring in a syn rather than the anti geometry proposed earlier."' R. Aumann, Chem. Ber., 1976, 109, 168. 'I8 Z. Goldschmidt, H. E. Gottlieb, and D.Cohen, J. Organornet. Chem., 1985, 294, 219. '19 W. Grimme, Chem. Ber., 1967, 100, 113. A. Saber, J. Organornet. Chem., 1976, 117, 245. F. J. Liotta, Jr. and B. K. Carpenter, J. Am. Chcm Soc.. 1985, 107, 6426. Vinylcyclopropane Rearrangements Hence the proposed pathways for the reaction all involve an intramolecular rearrangement to the unstable endo-(q2,q4-bi~y~lo[4.2.1]nonatriene)Mo(CO),, which attains the final em stereochemistry by a subsequent facile intermolecular exchange process. Deuterium labelling studies showed that a competitive degenerate 1,7-carbon shift also occurs, which does not involve the metal. p:: Fe(C 0 6c-Fe (CO), Fe(C 01, Fe(CO1, 40 "C 60 "C -c--+ ae(c0)3FetCO), (70) -.M(CO), \/M(C013 M = Cr, Mo,W The q ',q2-Rh(acac) complex of rndo-6-vinylbicyc10[3.1 .O]hex-2-ene undergoes a thermal skeletal rearrangement above 80°C to a mixture of 85% of bicyclo[3.3.0]octa-2,6-dienecomplex and 150/, of the bicyclo[3.2.l]octa-2,6-diene Goldschmidr and Crammer complex (equation 71) 12* in a formal vcpcyclopentene (1,3-shift) and Cope (3,3- shift) rearrangement, respectively.The related Rh(hexafluor0-acac) complex rearranges more readily, providing only the first product. The Ir(acac) complex does not rearrange under similar conditions. It is often observed that metal complexation reduces the energy barrier to rearrangements. An interesting case of the opposite effect has been reported for the electrocyclic ring-opening of both the e.xo and endo isomers of (q4-norcaradiene)Fe(CO), complex to (q4-cycloheptatriene)Fe(CO), (equation 72).12, Here it was found that the energy differences between the cycloheptatriene complex and the isomeric norcaradienes amounts to 21 kcal mol-', while the difference between the free-valence isomers is estimated '24 to be only ca.4 kcal molt'. Finally, we mention two unusual examples of intramolecular oxidative additions in which a cyclopropyl ring carbon migrates to a metal carbonyl ligand. Thus, (q2-benzvalene)Fe(CO), undergoes a facile rearrangement above 10 "C to the M M M M = Fe(C0I3 isomeric q ',q3-Fe(CO), o,~-allylic complex (equation 73).' 25 Similarly, the q4-coordinated (spiro[4.2]heptadiene)Fe(CO), compounds were transformed into the corresponding sigma bonded q ',q 5-(cyclopentadienyl)Fe(CO)2complexes by a cyclopropane carbon shift to the ligand carbonyl (equation 74).26 Although coordination between vcp and Pd" complexes was detected as early as \ (73)P-Qy Fe(C014 Fe(CO l3 V. Ark, J. M. Brown, J. A. Conneely. B. T. Golding, and D. H. Williamson. J. Chem. Soc., Perkiti Trans. 2, 1975, 4. W. Grimme and H. G. Koser, J. Am. Chem. Soc., 1981, 103. 5919. lZ4 P. M. Warner and S.-L. Lu, J. Am. Chrm. Snc., 1980, 102, 331. R. Aumann and W. Worman. Ciiem. Bu.,1979, 112, 1233. P. Eilbracht and LJ. Mayser, Chem. Ber., 1980, 113, 2211. 261 Vinylcyclopropane Rearrangements &-Fe(CO15 Fe(C0I3 R = Aryl ,cyclopropyl M = Fe(C0) n = 0,1,2 (77) - Fe (CO) M Fe(C0I30I + &I M M + + 1 Fe,(CO Goldschmidt and Crammer 1967127-129 elucidation of the product structure has remained obscure.On the other hand, a series of (diene)Fe(CO), complexes, formed by treatment of acyclic vinylcyclopropanes with iron pentacarbonyl, has been characterized (equations 75-76).' 31 The reaction mechanism was elucidated by low-temperature irradiation of vcp and Fe(CO),, giving two unstable Fe(CO),-coordinated isomers, the q2-vcp and q1-acyl-q3-allylic complexes shown in equation 77.132 At temperatures above 10 OC, these complexes readily lose a CO molecule to give the ring-opened o,n-allylic-Fe(CO), complex. The latter subsequently rearranges by hydrogen migration to the final diene complexes.The reaction of bicyclo[6.1.0]nonatriene with Fe,(CO), also gave a mixture of mono and bicyclic coordinated dienes (equation 78)' 33*134 Formally, the formation of these complexes requires neither a o,n-allylic intermediate nor a hydrogen migration. However, evidence from analogous rearrangements of Co (equation 79),',' Mo (equation 70),12' and Rh (equation 71) '22 complexes suggests that o,n-allylic intermediates are indeed involved in this reaction. Polycyclic olefin systems containing a 1,2-divinylcyclopropane moiety react with Fe,(CO), to give stable mono- and di-nuclear o,n-allylic complexes, some of which show interesting dynamic properties. Thus, ring-cleavage reactions of this type occur in semibullvalene (equation 80),136*137barbaralone (equation 81),' ,**' 39 and bullvalene (equation 82).140 Bis(ethylene)rhodium(I) complexes likewise react with exo-6-vinylbicyclo-C3.1 .O)hex-2-ene to give o,n-allylic complexes (equation 83) which, depending on the auxiliary ligand, form stable monomers, dimers, or tetramers.We conclude this section with the reactions of iron and nickel carbonyls with spiroC4.2lheptadienes and its derivatives. The parent compound reacts readily with Fe,(CO), at 25 OC, affording an acyl-bridged cyclopentadienyl complex as the major product, together with an alkyl-bridged dinuclear complex (equation 84). '42 The same reaction occurs in the more complicated spiro-norcaradiene derivative shown in equation 85.14, In the reaction of 1-vinylspiro[4.2]heptadiene with Ni(CO),, both alkyl-and acyl-bridged cyclopentadienyl complexes were observed ''' A.D. Keteley and J. A. Braaz, J. Organomer. Chem., 1967, 9, P5. T. Shono, T. Yoshimura, Y. Matsumura, and R. Oda, J. Org. Chem., 1968, 33, 1876. E. Vedejs, J. Am. Chem. SOC., 1968,90, 4751. 130 S. Sarel, R. Ben-Shoshan, and B. Kirson, J. Am. Chem. Soc., 1965, 87, 2417; R. Ben-Shoshan and S. Sarel, J. Chem. SOC.,Chem. Cornmun., 1969, 883. S. Sarel and M. Langbeheim, J. Chem. SOC., Chem. Commun., 1979, 13. 13' R. Aumann, J. Am. Chem. SOC., 1974, %, 2631. 133 E. J. Reardon, Jr. and M. Brookhart, J. Am. Chem. SOC.,1973,95,4311. 134 G. Deganello, H. Maltz, and J. Kozarich, J. Organomet. Chem., 1973, 60, 323. 13' H. R. Beer, P.Bigler, W. von Philipsborn, and A. Salzer, fnorg. Chim. Acta, 1981, 53, L49. R. M. Moriarty, C.-L. Yeh, and K. C. Ramey, J. Am. Chem. SOC.,1971,93, 6709. 13' D. Entholt, A. Rosan, and M. Rosenblum, J. Organomet. Chem., 1973,56, 315. 13' A. Eisenstadt, Tetrahedron Lett., 1972, 2005. 139 A. H. Wang, I. C. Paul, and R. Aumann, J. Organomet. Chem., 1974, 69, 301. I4O R. Aumann, Chem. Ber., 1975. 108, 1974. 14' N. W. Alcock, J. M. Brown. J. A. Conneely, and D. H. Williamson, J. Chem. SOC.,Perkin Trans. 2, 1979, 962. 14* P. E. Eilbracht and P. Dahler, Chem. Ber., 1980, 113, 542. 143 R. M. Moriarty, K. N. Chen, M. R. Churchill, and S. W. Y. Chang, J. Am. Chem. SOC.,1974,96, 3661. Vinylcyclopropane Rearrangements a9-0I/ cpco cpco FP (CO) -..M = Fe(CO), + + X = cyclopentadienyl (monomer) X = acac (dimer) X = hexafluoro-acac (tetramer) 264 Goldschmidt and Crammer (equation 86).'44 In these reactions, as in those of the non-spiro counterparts, the initial step consists of coordination of the organic 7c-system with the metal. This is followed by cyclopropane ring-cleavage giving either alkyl or acyl o,z-allylic complexes. Here, the allylic group and the additional double bond join to form the cyclopentadienyl ligand. m--.Fe (CO) (85) Ni(C014 (86) CO co 6 Conclusions Recently, rearrangements of the vcp radical-cations have become a subject of interest. Collisional ionization studies of vcp (4) in the gas phase indicate cleavage of the cyclopropane ring uia pathway (a) to the 1,3-pentadiene radical-cation (7) and not by pathway (b) to the isoprene radical-cation (8) (Scheme 4).14' 144 P.E. Eilbracht, Chem. Ber., 1976, 109, 3136; P. E. Eilbracht, U. Mayser, and G. Tiedtke, Chem. Ber., 1980, 113, 1420. 14' C. Dass, D. A. Peake, and M. L. Gross, Org. Mass Spectrom., 1986, 21, 741. Vinylcyclopropane Rearrangements Vinylcyclopropyl radical-cations were proposed as intermediates in the cis-trans isomerization of cis-1-p-anisyl-2-vinylcyclopropane(160) to the trans isomer (16 1) by one-electron chemical oxidants such as, for example, p-BrPh,N' +SbF6 -at temperatures as low as -90 "C (equation 87). 146 The same one-electron oxidant caused the rearrangement of either the cis-vcp (162) or the trans-vcp (1 63) to the 1,2-dimethyl-4-p-anisyl cyclopentene (164) at 59°C in 86% yield (equation 88)14' It should be noted that the thermal rearrangement of either (162) or (163) in the absence of such oxidants proceeded only above 200 OC, to yield (164).'"' p-Br P h3N*+ S bF6--78 "C (87) Me0 Me0 (1601 (161) 84'1.59 "C p-BrPh3N*+SbF;1 (164) 86'10 It can be seen that vcp rearrangements have been extensively studied for more than two decades, principally from a mechanistic viewpoint. Recently, however, the thermal pathways involving vcp-cyclopentene rearrangements have been utilized in the total syntheses of antheridi~gen-An,~~ aphidi~olin,'~' ~izaene,'~~ and 1 1-deoxyprostaglandin E.' 50 Acid-catalysed rearrangement of chrysanthemic acid esters are important intermediates to lavandulols which are used as fragrants.51 A new route utilizing trimethyl iodide (TMSI) which induces the vinyl-cyclopropane-cyclopentene rearrangement has been developed. For example the vcp (165) under carefully controlled reaction conditions either rearranged 146 J. P. Dinnocenzo and M. Schmittel, J. Am. Chem. Soc., 1987, 109, 1561. 14' J. P. Dinnocenzo and D. A. Conlon, J. Am. Chem. Soc., 1988, 110, 2324. 14* B. M. Trost, Y. Nishirnura, K. Yarnarnoto, and S. S. McElvain, J. Am. Chem. Soc., 1979, 101, 1328. 149 E. Piers and J. Bauville, J. Chem. SOC.,Chem. Commun., 1979, 1138. 150 J. Salaun and J. Ollivier, Nouu. J. Chim., 1981, 5, 587. lS1 G. Suzukamo and M. Tamura, Surnitomo Chemical Co., U.S.Patent No. 4,547,586 (1985). Goldschmidt and Crammer to the annulated cyclopentene (166) or to the bicyclo[3.2.l]octene (equation 89).lS2 C0,Et (166) (165) + (i)TMS1,HMDS (11) H$IJ OnCO, Et This review has attempted to highlight the broad versatility and utility of vcp rearrangements. It is hoped that such rearrangements will encourage further developments in the use of vcps as precursors in organic and natural product syntheses. 15’ A. Fleming, G. Sinai-Zingde, M. Natchus, and T. Hudlicky, Tetrahedron Lett., 1987, 28, 167.

ISSN:0306-0012    

DOI:10.1039/CS9881700229

出版商:RSC    

年代:1988

数据来源: RSC