摘要:

Chemical Society Reviews Vol 6 No 4 1977 Page TILDEN LECTURE New Perspectives in Surface Chemistry and Catalysis By M. W. Roberts 373 Organoborates in Organic Synthesis: The Use of Alkenyl-, Alkynyl-, and Cyano-borates as Synthetic Intermediates By G. M. L. Cragg and K. R. Koch 393 CENTENARY LECTURE Systematic Development of Strategy in the Synthesis of Polycyclic Polysubstituted Natural Products :The Aconite Alkaloids By K. Wiesner 41 3 Properties and Syntheses of Sweetening Agents By B. Crammer and R. Ikan 43 1 MELDOLA MEDAL LECTURE N.M.R. Spectral Change as a Probe of Chlorophyll Chemistry By J. K. M. Sanders 467 Prostaglandins, Thromboxanes, PGX: Biosynthetic Products from Arachidonic Acid By K.H. Gibson 489 Corrigendum 511 1977 Indexes 513 The Chemical Society London Chemical Society Reviews Chemical Society Reviews appears quarterly and comprises approximately 25 articles (ca. 500 pp) per annum. It is intended that each review article shall be of interest to chemists in general, and not merely to those with a specialist interest in the subject under review. The articles range over the whole of chemistry and its interfaces with other disciplines. Although the majority of articles are intended to be specially commissioned, the Society is always prepared to consider offers of articles for publication. In such cases a short synopsis, rather than the completed article, should be sub- mitted to The Managing Editor, Books and Reviews Section, The Chemical Society, Burlington House, Piccadilly, London, W 1V OBN.Members of the Chemical Society may subscribe to Chernicaf Society Reviews at f5.00per annurn (beginning 1978, 26.00 per annum): they should place their orders on their Annual Subscription renewal forms in the usual way. Non-members may order Chemical Society Reviews for f 14.00 j$,O) [be-ginning 1978, E16.00($33)] per annum (remittance with order) from: The Publications Sales Officer, The Chemical Society, Blackhorse Road, Letchworth, Herts., SG6 lHN, England. 0Copyright reserved by The Chemical Society 1977 Published by The Chemical Society, Burlington House, London, W1V OBN Printed in England by Eyre & Spottiswoode Ltd, Thanet Press, Margate

ISSN:0306-0012    

DOI:10.1039/CS97706FP009

出版商:RSC    

年代:1977

数据来源: RSC

摘要:

TILDEN LECTURE* New Perspectives in Surface Chemistry and Catalysis By M. W. Roberts SCHOOL OF CHEMISTRY, UNIVERSITY OF BRADFORD, BRADFORD, YORKSHIRE BD7 1DP 1 Introduction Surface Chemistry is an area where information at the molecular level has been difficult to obtain and, in the main, we have had to resort to studying the gas phase and then extrapolating the information to develop models of the molecular events occurring at the solid surface. The problems facing the surface chemist have therefore largely revolved around the difficulty of finding experimental methods which can (i) monitor directly the outermost atoms at a solid surface as distinct from sub-surface atoms, (ii) then be able to distinguish one kind from another, and (iii) then monitor the different chemical environment of a given surface atom.These are three distinct milestones on the road to the complete understanding of molecular events at surfaces. Let us see how far along that road we can go. Fifteen years ago our knowledge of chemisorption on metals was confined largely to information relating to the patterns of reactivity, nitrogen exemplifying obviously rather stringent requirements for chemisorption, being only adsorbed by a few metals at room temperature, whereas oxygen is the most ubiquitous. Such an approach tells us nothing about the molecular nature of the surface species although other techniques such as flash-desorption developed in the late fifties and early sixties would suggest that nitrogen is dissociatively chemisorbed whereas carbon monoxide epitomizes molecular adsorption. We will return to this particular point later.What is missing is surface definition at the molecular level, both chemical and structural, and also information on the adsorbate; for example, what is the nature of interadatom bonding in a diatomic molecule or adatom-substrate bonding? 1 would like to consider how one particular experimental development, electron spectroscopy, has, over the past five years, enabled a more detailed picture of chemisorption to be developed. I shall confine myself in the main to just core-level and valence-level spectroscopies (Figure I ). The two relaxation processes, the Auger effect and X-ray fluorescence, which may accompany photoemission are not considered.Although Auger Electron Spectroscopy was developed in the early sixties as a surface sensitive technique its inherent com- plexity (it is a three-electron process) and the possible damaging influence of the *First delivered on 14 October 1976 at the University of Hull. New Perspectives in Surface Chemistry and Catalysis electron beam has limited its application compared with X-ray and U.V. photo-electron spectroscopy (XPS or ESCA and UPS). In photoelectron spectroscopy the kinetic energy EICof the electron emitted by a photon of energy hv is measured and using the relation EK = hv -(Einitiai -Efinai) = hv -EB the binding energy EBcan be calculated. If we invoke Koopmans’ Theorem then the eigenvalue of a given orbital is equal to its binding energy, i.e.it is the energy needed to extract the orbital electron to infinity as long as there is no readjustment of the other electrons in the atom or molecule (the frozen orbital approximation).The concept of single electron orbitals with specific binding energies is of course an ideality, but to consider a molecule as being built up of energy levels corres- ponding to the binding energy of its molecular orbitals is a highly productive one even though it contains some half-truths. Core levels UPS: VALENCE LEVEL SPECTROSCOPY XPS: CORE LEVEL SPECTROSCOPY J T X-RAY AUGER FLOURESCENC E PROCESS Figure 1 Photoelectron spectroscopy My interest in photoelectron spectroscopy stems from a study with C.M. Quinn of the energy distribution of photoelectrons from nickell using photons of energy 6.2 eV. This work illustrated that photoelectrons were emitted from a region very close to the surface (<20 A) and that, after oxygen interaction, their energies C. M.Quinn and M. W. Roberts, Trans. Faraday SOC.,1965, 61, 1775. Roberts were different from those from the clean metal. Clearly with such low-energy photons it was only possible to explore the band structure close to the Fernii level. But the enormous interest in the sixties in the X-ray and U.V. photoelectron spectroscopy of gases, largely through the elegant work of Siegbahn, Turner, and Price led us to consider the possibility of investigating the molecular nature of adsorbed species on well defined atomically clean surfaces.We were encouraged by two quite different types of experiment carried out by Siegbahn and his colleagues. First they had shown that, for example, the C(1s) binding energies for a series of halogenated hydrocarbons (e.g. CBr4, CHBr3, CHaBrz, CHBr3, or CH4) increased by roughly equal increments as each hydrogen was replaced by a bromine atom. In the case of fluorinated hydrocarbons the C(1s) shift can be as much as 8 eV. We have therefore a relationship between chemical shift and the electron density on the carbon atoms. Secondly, some experiments on iodostearic acid had shown2 clearly the possibility that X-ray photoelectrons could be sensitive at the monolayer level since a substantial I3dsp signal was observed for ostensibly a monolayer.It should, however, be recalled that iodostearic acid is a comparatively large molecule (ca. 50 A) whereas we would require sufficient sensitivity to study, at the sub-monolayer level, the adsorption of relatively small molecules such as nitrogen, carbon monoxide, and simple hydrocarbons. Two questions therefore needed to be resolved, first, whether the escape depth of photoelectrons was sufficiently small (a few Angstroms) to give a sensitivity at the sub-monolayer level. In this respect our work with nickell and the results of Bordass and Linnett3 who reported in 1969 a U.V. photoelectron spectrum for methanol adsorbed on tungsten were encouraging. Secondly there were the experimental difficulties associated with coupling the photoelectron technique with the stringent requirements for preparing well defined metal surfaces.A u.h.v.-compatible photoelectron spectrometer was constructed4 in con-junction with Vacuum Generators; the instrument was of a dual-chamber design with multiphoton (He and X-ray) sources. It allowed the in situ generation of atomically clean metal surfaces and the study of adsorption at these surfaces using both valence-level (He-I and He-I1 radiation) and core-level (AI-Ka radiation) spectroscopies, and was coupled to a Digico computer for storage and analysis of the experimental data. We therefore had facilities which enabled us, at least in principle, to investigate electron binding energies of both the valence and core levels, to distinguish between different chemical environments of a given atom by increasing chemical shifts, and lastly to have analytical information on the atomic composition of the surface.Whether each of these were to be realized in practice remained to be seen. 2 K. Siegbahn, C. Nordling, G. Johansson, J. Hedman, P. F. Heden, K. Hamrin, U.Gelius, T. Bergmark, L. 0. Werme, R. Manne, and Y. Baer, ‘ESCA Applied to Free Molecules’ North-Holland, Amsterdam, 1969. W. T. Bordass and J. W. Linnett, Nature, 1969, 222, 660. C. R. Brundle, M. W. Roberts, D. Latham, and K. Yates, J. Electron Spectroscopy,1974, 3,241. New Perspectives in Surface Chemistry and Catalysis By first studying the physical adsorption5 of such molecules as C02, HzO, and Hg on atomically clean gold surfaces the viability of the experimental approach was established.We then turned our attention to chemisorption. 2 Adsorption of Carbon Monoxide on Metals Carbon monoxide adsorption on molybdenum films was a particularly attractive system to start with because it illustrated some of the central features which needed to be resolved in order to establish the potential of photoelectron spectroscopy for investigating systems of interest to those of us concerned with chemisorption and catalysis. It was known that CO adsorption on such transition metals as molybdenum and tungsten6p7 at 295 K occurs with a high heat of adsorption (ca. 300 kJ mol-l); on cooling to 80 K further adsorption of carbon monoxide occurs6 with a comparatively low heat of adsorption (ca.60 kJ mol-l). The room-temperature adsorption is usually referred to as the ,&state and the adsorption at 80 K we designate as the y-state. Figure 2 shows the O(1s) spectrum of CO in the ,&state and that its binding energy is 530eV; the y-state has associ- He (I + 11) O(2p) 5.6 eV 7eV 11 eV 5a 4-IT 4a 1 1 530 535 eV Figure 2 Core-level O(1s) spectra together with data from valence-level studies using He radiation for CO adsorption on Mo at 290 K and 85 K ated with it an O(1s) spectrum with a range of binding energies up to a maximum of about 536 eV. When the valence levels are explored using He (I and 11) radiation the 13-state exhibits a single peak at about 5.6 eV below the Fermi level (EF)whereas the y-state is characterized by two peaks at about 7 eV and 11 eV below EF.The 5.6 eV peak can be assigned to the O(2p)level while the 7 eV and 11 eV peaks have counterparts in the spectrum of CO(g).If the adlayer (the Virgin state) is formed at low temperature (80 K) and then warmed to 395 K, the initial O(1s) binding energy which is at about 531 eV decreases to 530 eV at 295 K and the two peaks at 7 eV and 11 eV in the He spectrum disappear, being repIaced by a broad hump at about 5.5 eV below EF.We make the following observations : C. R. Brundle and M. W. Roberts, Proc. Roy. Soc., 1972, A331, 383. (a) R. R. Ford, Adv. Catalvsis, 1970, 21, 51 ; (b) D. 0. Hayward, in ‘Chemisorption and Reactions on Metallic Films’, ed.J. R. Anderson, Academic Press, London, New York, 1971. J. G. Little, C. M. Quinn, and M. W. Roberts, J. Catalysis, 1964, 3, 57. Roberts (i) CO adsorbed with different heats of adsorption is distinguished by different O(1s) binding energies. (ii) Strongly chemisorbed CO is characterized by substantially lower O(1s) binding energy than that which is weakly adsorbed. Furthermore the O(1s) value of 530 eV @-state) is the same as for chemisorbed oxygen on molybdenum. (iii) The p-state does not exhibit orbital structure in the He spectra which can be assigned easily to molecular carbon monoxide. The peak at 5.5 eV is attributed to the O(2p)level. (iv) The y-state exhibits peaks at 7 eV and 11 eV which can be assigned to the (50 and In)and 4aorbitals in CO(g).The assignment of these peaks gave rise to considerable debate and speculation8a~gJ0 but is now well established experimentally as a result of the synchroton radiation studies of Gustafsson et a1.8b The outer 50. orbital is primarily centred on the C atom and looks like a lone pair in that it extends out well beyond the C atoms. The IT level, which accommodates four electrons, is centred on the oxygen in that the charge distribution is about 3 :1 around the oxygen compared with the carbon. Lastly the 4a is like an oxygen lone pair and extends beyond the 0 atom. O(W 540 eV He PEAKS AT He PEAKS 7 eV & 10 eV ABSENT I DISSOCIAT ED535 eV weak inter- atomic bonding rn 530 eV n 50 100 300 kJ (AH) Figure 3 Correlation between O(1s) binding energy for adsorbed CO on diflerent metals, its heat ofadsorption (AH),and the presencelabsence ofpeaks in the He spectra characteristic of molecular CO There is therefore a prima facie case for considering carbon monoxide to be dissociated, or at least that the C-0 bond is very weak, in the /%state.The relationship between heat of adsorption and the O(1s) binding energy noted above led us to explore whether there was a more general correlation. (a)D. R. Lloyd, Furaduy Disc. Chem. SOC.,1974, No. 58, p. 136; (b) T. Gustafsson, E. W. Plummer, D. E. Eastman, and J. L. Freeouf, Solid State Comm., 1975, 17, 391. D E. Eastman and J. K. Cashion, Phys. Rev. Letters, 1971, 27, 1520.loT. A. Clarke, I. D. Gay, B. Law, and R. Mason, Chem. Phys. Letters, 1975, 31, 29. New Perspectives in Surface Chemistry and Catalysis Figure 3 shows O(1s) values as a function of the heat of adsorption of carbon monoxide on a number of metals.ll Furthermore we include iqformation from U.V. photoelectron spectroscopy using He (I and 11)radiation. The vertical line divides these situations where peaks in the He spectra at about 7 eV and 10 eV were observed from those where they were absent. It is evident that there is a correlation between the molecular state of CO (dissociatively or associatively adsorbed), the heat of adsorption, and the O(1s) binding energy. It should be noted that when CO is chemisorbed at 295 K with a heat of adsorption greater than about 260 kJ mol-l the O(1s) binding energy is invariant at 530 eV, a value typical of dissociatively chemisorbed oxygen on transition metals.What else can we deduce from this correlation? If we think of the decrease in the O(1s) value (Figure 3) as reflecting an increase in the electron density in the vicinity of the oxygen of the carbon monoxide molecule, then this increase can be associated with the accompanying increase in the heat of adsorption. Clearly it is attractive to think of the increased heat of adsorption as arising largely from back-bonding into the antibonding orbitals of CO. Support for this view is also inherent in the conclusions of Grimley’s12J3 theoretical work. As the extent of back-bonding increases, the heat of adsorption increases, the O(1s) value decreases, and the C-0 bond strength decreases until ultimately no orbital structure is observed in the He spectra.At this stage we consider the molecule to be ‘dissociated’ but whether the carbon and oxygen exist as separate adatoms is another matter. The threshold for dissociation of CO at 295 K is therefore reached when the extent of back-bonding results in the heat of chemisorption exceeding about 260 kJ mol-1. It should be noted that the threshold will vary with the substrate temperature. Now any correlation of this kind is only worthwhile if it is successful in predicting molecular processes and the system we chose to consider was the adsorption of CO on iron. The heat of adsorption of CO on ironl4 is close to the ‘threshold heat’; further- more isotope exchange data15 had suggested that the molecule was in an incipient state of dissociation at 190 K.In spite of this observation which was made some 20 years ago, and occasional reports of CO adsorption being infrared inactive,16 it should be emphasized that the consensus of opinion was wholly in favour of the adsorption of CO on metals being molecular at room temperature. Figure 4 shows the C(1s) spectrum for the CO adlayer formed on polycrystalline ironl7 at 85 K and during the warming of the adlayer to 295 K. A new C(1s) peak, reflecting ‘carbidic’ carbon emerges at 295 K; this is accompanied by a diminution in the intensity of the (50 and 1.n) and 40 orbitals. This diminution occurs slowly at l1 R.W. Joyner and M.W. Roberts, Chem. Phys. Letters, 1974,29,447. laT. B. Grimley, ‘Batelle Colloq. Mol. Processes on Solid Surfaces’, 1968. l3See also ref. 6u. l4 D. Brennan and F. H. Hayes, Phil. Trans. Roy. SOC.1965, A258, 347. l5 A. N. Webb and R. P. Eischens, J. Amer. Chem. SOC..1955,77,4710. l6 A. M. Bradshaw and J. Pritchard. Prac. Roy. SOC.,1970, A316, 169. K. Kishi and M. W. Roberts, J.C.S. Furuduy I, 1975, 71, 1715. 378 Roberts 295 K but is accelerated by warming to 350 K. On recooling to 295 K the ‘reacted’ surface, which can be considered to be essentially a ‘carbidic/oxide’ adlayer, is active in further adsorption of CO and the C(1s) spectra (Figure 4), the O(1s) spectra, and the He (I and 11) data indicate molecular adsorption.Menzel and his colleagues18 have reviewed the present position regarding CO adsorption on metals and also reported data for the adsorption of carbon monoxide on W(110). The evidence is unambiguous regarding the ready cleavage of the C-0 bond on metals and as a rule the ease of dissociation follows the correlation shown in Figure 3. -85 K 290 K I’ I I I 280 285 290 (ev) Figure 4 Thermally induced dissociation of carbon monoxide adsorbed on iron at 85 K. C(1s) data Carbon monoxide adsorption on copper19 has been studied extensively and, in general, there is good agreement between different studies. At low temperature there is unanimity of view that adsorption is relatively weak, with a heat of adsorption of about 60 kJ mol-l, the adlayer exhibits a (2/z x 42) R45” symmetry but at higher coverage the symmetry is hexagonal which is attributed to a ‘compression structure’.The adlayer formed at 85 K desorbs on warming to 295 K under a dynamic vacuum. At 295 K there has been some ambiguity as to whether or not copper adsorbs carbon monoxide, but by combining photoelectron spectroscopy with LEED we have recently reconsidered the question, the stimulus for this having come from observations we made on the (NO + Cu) system which we discuss later. Table 1 summarizes the data observed with Cu(100) at 85 K and 295 K. There is no doubt that if exposed to a relatively high pressure of CO (510-2 Torr), Cu(100) will adsorb CO in the molecular state20 at 295 K.The electron spectroscopic characteristics of the two states (85 K and 295 K) are, l8 E. Umbach, J. C.Fuggle, and D. Menzel, J. Electron. Spectroscopy, 1977, 10, 15. lS J. Pritchard, T. Gatterick, and R. K. Gupta, Surface Sci., 1975, 53, 1. 2o S. Isa, R. W. Joyner, and M. W. Roberts, J.C.S. Chem. Comm.,1977, 377. New Perspectives in Surface Chemistry and Catalysis Table 1 Characteristics of CO adsorption an Cu(100) at 80 K and 295 K (XPS, UPS, LEED) 80 K 295 K WS) 534 eV 532.8 eV He-II/eV 8.6, 11.8, 13.7 8.1, 11.1 S (sticking probability) -unity -10-6 LEED (symmetry of adlayer) (42 x 42)R45" (42 x 42)R45" Sharp diffraction Diffuse diffraction features 8 N 0.5 features 8 N 0.3 Desorption temp. (K) -180 -380 Bonding a-bonding d-rr*-bonding dominant: Cob+ dominant: Cob-however, very different, as also are their heats of adsorption and sticking probabilities.We designate the low-temperature state as CO", implying that o-bonding to the metal is dominant whereas we designate adsorption at 295 K as Cob-. Back-bonding into the antibonding CO molecular orbital is suggested to be most significant at room temperature; this is compatible with our correlation of O(1s) binding energy values versus heat of adsorption (Figure 3). The estimated coverage at 295 K is only about 0.3 and the symmetry of the adlayer (d2 x d2) R45";the LEED pattern is, however, rather diffuse implying appreciable disorder within the layer. The question that is immediately raised is why has irreversible adsorption of CO on copper at room temperature not been observed by infrared spectroscopy? One possibility is that COadss-is infrared inactive, either because of its particular molecular orientation with respect to the surface and the incident radiation (this is critical in the reflection-absorption technique) or because the C-0 bond has been appreciably weakened by back-bonding.* A pattern is therefore emerging for CO adsorption on metals where the following sequence of events has been recognized: co(g)3COads" COada8---+ Cads f Oads In the case of molybdenum we observe largely COaass- at 85 K but we go through to Cads + Oads at room temperature, whereas in the case of nickel we stop at coads8-at 295 K.With iron, dissociation occurs slowly at 295 K but can be completely arrestedL7 by preadsorbing submonolayer quantities of sulphur.With copper we see almost exclusively COads" at 85 K; this desorbs on warming to 295 K, but toads'-forms only slowly from CO(g) at room temperature, possibly because COads" is an essential precursor and its concentration at 295 K is, in view of its low heat of adsorption, very small. Alternatively (or in addition) it is electron redistribution leading to bonding that is rate-determining. In this discussion we have emphasized the relationship between charge, electron *Note added in proof: For CO adsorbed on Ni(100) recent calculations suggest that the molecule is tipped over at an angle of about 34"with respect to the surface normal (Dr. J.B. Pendry, personal communication). 380 Roberts spectroscopy, and the chemical reactivity of adsorbed carbon monoxide. The interdependence of charge and stretching frequency has of course been established in transition-metal chemistry where the stretching frequency of CO is not only related to the number of d-electrons but also to the presence of a total net charge. In the following complexes, for example, all have d6 electron configurations and the same co-ordination sphere but different formal charges: [Mn(CO)6]+ [cr(co)61 [v(co)~I-vco = 2095 cm-1 vco = 2000 cm-l vco = 1860 cm-1 The surface chemistry of carbon monoxide on atomically clean metal surfaces may therefore be considered in terms of the two-way electron transfer process, the molecule behaving as either an electron acceptor, COs-, or an electron donor, COs+.This is a point central to a number of infrared studies,2la to recent theoretical calculations of CO chemisorption and bonding in transition-metal complexes,21b and to certain homogeneous catalytic reactions. A particular example of the latter would be the carboxylation of alcohols where CO appears to behave as an electron acceptor. 3 Adsorption of Nitric Oxide on Metals Our interest in nitric oxide was stimulated by the conclusions we reached with carbon monoxide. The bond energy of NO (600 kJ) is appreciably smaller than that of CO (lo00 kJ) and if the model we used (Figure 3) to rationalize the molecular state of adsorbed CO is also valid for NO then we would anticipate analogous molecular events.There is an important difference between CO and NO and that is the inherent thermodynamic instability of NO, so that we will be exploring the conditions under which adsorbed molecular NO becomes kinetically unstable and dissociates. Although infrared studies22 have shown that at high pressures (several Torr) NO interacts with iron and other surfaces to give oxide, there is no information on the possible molecular subtleties of the interaction. Furthermore the data have been interpreted over the years in different ways so that it is difficult to obtain a satisfactory model for the interaction. Figure 5(a) shows that the molecule, in fact, dissociates at low coverage on a clean iron surface at 85 K but at higher coverage molecular adsorption predominate^.^^ If, however, the adlayer is thermally activated, by warming to 295 K, dissociation occurs to give a surface which is incipient oxide/nitride.Exposure to NO(g) at higher pressure ( Torr) gave evidence for further molecular adsorption. Although we only report here the N(1s) data, similar conclusions are reached from an examination of the O(1s) spectra. One further point to note is that when the N(ls) and O(1s) intensities are compared, making use of known ionization cross-section data, we find that at low coverage at 85 K some recombination of 21 (a) G. Blyholder, J. Phys. Chew., 1964, 68, 2773; (b) I. H. Hillier and V. R. Saunders, Mol. Phys,. 1971, 22, 1025. 22 (a) A. N. Terenin and L.M. Roev, Actes du Deuxieme Congres International de Catalyse,1961, 2, 2183; (6) A. Alekseev and A. N. Terenin, J. Caralysis, 1965, 4, 440. 23 K. Kishi and M. W. Roberts, Proc. Roy. SOC.,1976, A352,289. New Perspectives in Surjace Chemistry and Catalysis nitrogen adatoms leading to desorption of N2 occurs. By examining the surface species using He (I and 11) radiation [Figure 5(b)] the conclusions from core-level spectroscopy regarding the nature of the adsorption, dissociative or associative, He-I1 185 K 290 K I1 1 I I l l I I I I I I I I I I 0 5 10 15 (ev) (b) Figure 5 (a)Adsorption of NO on Fe at 85 K followed by warming the adlayer to 295 K,N(1s) spectra are recorded with increasing exposure (Curve 2:2 x Torr, 80 s; 4 (total):2 x Torr, 480 s) (b) He-I1 spectra for NO adsorbed on Fe at 85 K followed by warming the adlayer to 295 K [details of exposure as in (a)] Roberts are confirmed. Peaks in the U.V.spectrum at high coverage at 85 K can be assigned to the 277*, 177 and 50, and 40 orbitals. These are absent at 295 K and the density of states curve has similarities to ‘oxidized’ iron. Clearly it is necessary to explore in more detail the high binding N(1s) (Figure 5) and O(1s) regions, i.e. that part of the spectrum assigned to molecular NO adsorption. The comparatively large FWHM value (ca. 6 eV) suggests contri- butions from more than one type of NO bonding. With this end in mind we investigated the interaction of nitric oxide with copper and, by analogy with carbon monoxide, adsorption was anticipated to be weak and molecular in nature.Figure 6 shows the O(1s) spectrum for the adlayer formed on copper24 at 80 K; there are two distinct peaks. We assign the peak at 531.6 eV to NOa&-and that at 535 eV to NOa&&+. He (I and 11) spectra confirm the molecular nature of these species. Furthermore, intensity data suggest that the molecule is linearly bonded to a single copper surface atom. We believe that in the case of iron (Figure 5) these two species are not resolved and the broad envelope at high binding energy probably incorporates them. -1.6 eV .1l 290 K 220 K 183 K 143 K 110 K 80 K clean metal II111111 IIIIII1171 530 535 540 (eV) Figure 6 O(1s) spectra for NO adsorption on copper at 80 K followed by warming of the adlayer, in stages, to 290 K *IM.Matloob and M. W. Roberts, J.C.S. Faraday I, (in press); Physica Scipta, (to be published). New Perspectives in Surface Chemistry and Catalvsis This molecular adlayer on thermal activation transforms into a layer which has no surface nitrogen present (Scheme 1). The nitrosonium-like species desorb at about 150 K and we estimate their heat of adsorption to be about 60 kJ mol-l. The Noads'-or nitrosyl-like species dissociate above about 180 K, giving exclu- sively chemisorbed oxygen which is bridge-bonded to copper. The nitrogen adatoms recombine and desorb as molecular nitrogen, above about 180 K. NOb-NO8+ NOb-_ NOb--NO8-I1 -I I + NO(,, cu cu cu cu cu cu /"\ /"\ + N2m cu cu cu Scheme 1 On the other hand, when copper is exposed to NO at 295 K, dissociation occurs and both the nitrogen and oxygen are retained at the surface in what we believe is a bridge-bonded configuration.24 Clearly there is a small activation energy for nitrogen adatoms to be chemisorbed by copper and in the temperature range 150-200 K recombination and desorption is faster than chemisorption.By making use of recent ionization cross-section and electron escape depth data it has also been possible to calculate from the photoelectron intensity data "(Is), O(ls), and Cu(2p3p)I the surface concentrations of nitrogen and oxygen adatoms. It is on this basis that we conclude that bridge-bonding occurs at 295 K whereas at 85 K each surface copper atom is linearly bonded to an NO molecule.We are currently investigating the structural aspects of these processes using Cu single crystals and combining photoelectron spectroscopy with LEED. In the case of NO adsorption on dissociation to give nitrogen adatoms N8-occurs "(1s) = 397 eV] at 85 K followed at higher exposure by the emergence of two further N(1s) peaks at 403 eV and 406 eV. The former, as with copper, we designate NOaas'-and the latter Noads'+. Evidence for these assignments was obtained by exposing the adlayer at 85 K to water vapour, a strong electron donor, when the NO8+species was desorbed entirely and the N8-and the NO8-species remained substantially unaltered. We can, in a sense, therefore think of the nitric oxide molecule as probing the electronic nature of the adlayer formed on aluminium (and also copper and iron) when it is but a few Angstroms thick, NO8-providing information on electron- excess sites and NO8+being bonded at Lewis acid or electron-deficient sites such as Al3+.That Nosas-is the precursor to dissociation is in keeping with these A. Carley and M. W. Roberts, (to be published). Roberts assignments since the N-0 bond will have been considerably weakened. It also follows that we might anticipate (as we observe with copper) that NO*+, where the N-0 bond is effectively strengthened on adsorption, would desorb on thermal activation of the adlayer. The occurrence of nitrosonium-and nitrosyl-like species is entirely compatible with the chemistry of nitric oxide but what is perhaps significant from the present work is that it is possible to isolate their analogues at metal surfaces and that monitoring their behaviour during thermal activation provides a better understanding of surface reactivity.Stimulated by our results for nitric oxide and carbon monoxide interaction with, in particular, iron, we explored26 the dinitrogen + iron system by X-ray photoelectron spectroscopy. The data and conclusionsare summarized in Scheme2. Scheme 2 Species (1) and (2), observed only at 85 K, are in equilibrium with Nz(g) and are weakIy adsorbed [Figure 7(a)]. We suggest that these two species are precursors to the formation of nitrogen adatoms characterized by an N(1s) value of 397 eV [Figure 7(b)].The low concentration of the two molecular species that would obviously be present at 295 K would account for the observed low rate of for-mation of nitrogen adatoms; we estimate its overall sticking probability to be less than Studies of other nitrogen molecules27 (NZH4,NH3) and nitrogen complexes offer supporting evidence for the molecular processes suggested and assignments made for the N(1s) peaks observed. An interesting feature of this study26 and relevant to the catalytic activity of nitrogen was that XPS indicated that the electronic environment of a nitrogen adatom was perturbed by surface impurities such as chemisorbed oxygen. This work was carried out on poly- crystalline iron, but MasonZSa and his colleagues have recently reported similar results for dinitrogen chemisorption on Fe(ll1) at 295 K.Mossbauer spectra have recently been reported for S7Fe in a solid nitrogen matrix where at 4.2 K molecular nitrogen is considered to bond to an iron dimer. In fact Barrett and MontanoZ8* suggest that molecular nitrogen only bonds to iron when the latter is present as a dimer. If this is correct then our assignment of 26 K. Kishi and M. W. Roberts, Surface Sci., 1977, 62, 252. 27 M. Matloob and M. W. Roberts, see ref. 24. 28 (a) I. D. Gay, M. Textor, R. Mason, and Y.Iwasawa, (to be published); (b) P. H. Barrett and P. A. Montano, J.C.S. Faraday IZ, 1977, 73, 378. 385 New Perspectives in Surface Chemistry and Catalysis N 111 N 80 K (i-iii)I PN~ Torr (i)= px, = IO-’Torr (ii) f ‘ ~= ~ Torr (iii) Warm to e-- 295 K t t t 397 eV 400.2 eV 405.3 eV (4 N 295 K, P = 1 Torr for 1 min followed clean Fe surface 397.2 eV (b) Figure 7 N(1s) and assignments for the adsorption of nitrogen on iron at (a)80 K and (b)295 K the N(1s) peak at 405 eV will require further investigation. It is apposite to enquire as to why N2 dissociates less easily than CO on, for example, iron.The clue may well lie in the observation that a-donation and n-back-bonding is considerably less in metal-dinitrogen ligands than in metal-carbonyl ligands. This would certainly fit in with the general ideas proposed here to account for the reactivity of simple diatomic molecules on ‘clean’ metal surfaces.4 Metal Oxidation I have, as yet, not mentioned any example where information has been extracted from core-level spectra of the metal, and the explanation for this is that substrate core-level binding energies are generally rather insensitive to adsorption. However, in the interaction of oxygen29 with aluminium the Al(2p) spectra have provided useful information on the mechanism of oxidation. Figure 8 shows the emergence of a peak at about 2.4 eV higher binding energy than the ‘metal’ AI(2p) value. The variation of the FWHM of this second ‘oxide component’ peak with oxygen exposure and temperature, in the range 85-295 K, was complex and indicated quite clearly that it was composite.By deconvolution, removing both instrumental and X-ray broadening from the experimental data, two distinct as A. Carley and M. W. Roberts, (to be published). Roberts components a and p are solved. Furthermore, we suggest that the a-component is a precursor of the /%form. This is compatible with the observation that the ‘oxide component’ peak develops more rapidly at 85 K than at 295 K. If the ,&form, which we believe is analogous to A1203, can only develop via the a, and since the formation of the a is faster at the lower temperature, then the overall kinetic characteristics of the rvidation may be understood. It is important to note that the O(1s) data, althougn compatible with the behaviour of the Al(2p) peak during oxygen interaction, are not in themselves revealing. IItIIOIII I 1 I I I III1I 532 533 70 75 B.E.(eV) B.E. (eV) Figure 8 O(1s) and AI(2p) spectra during oxygen interaction with aluminium at 290 K (1 L zz Torr sec) One of the consequences of the emergence of such new and powerful techniques as electron spectroscopy is that one can reflect critically on the information obtained from more traditional techniques. The aluminium-oxygen system is one of them. A decade or so ago we investigated3O the interaction of oxygen with 30 M.W. Roberts and B. R. Wells, Surface Sci., 1969, 15, 325. 387 New Perspectives in Surface Chemistry and Catalysis aluminium by monitoring the change in work function. There is one aspect of that work which I would like to recall, namely the observation that oxygen incor- poration occurred with an activation energy of no more than a few kJ mol-1 and that a transitory low work function oxide was observed before the formation of the oxide of higher work function.These particular studies were made in the same temperature range (85-295 K) and at about the same oxygen pressure (ca. 10-6 Torr). It is, therefore, tempting to attribute the low work function ‘oxide’ to the a-component, and the higher work function ‘oxide’ to the /?-component which we believe is probably the ionic-spinel A1203. Furthermore, this would be compatible with the NO6+ and NO6-species observed when aluminium, oxidized to an estimated thickness of 9 A, is exposed to nitric oxide at 85 K.In contrast to the above, the electron spectroscopy of the interaction of oxygen with Pb(100) and Pb(ll0) single crystals was not as revealing31 as the information we obtained from LEED. The data were unusual in that from the lowest oxygen exposures used the extra diffraction spots which were observed remained un- altered during the continuous growth of the oxide (up to about 9 A, at 295 K). It is much more common during ‘oxidation’ at low temperature for the adlayer to appear either highly disordered or for facet structures to develop. We interpret the present data as reflecting the growth of four domains of orthorhomobic PbO with a = 89” and that growth is continuous from the lowest oxygen exposure used. A nucleation model is proposed for oxide growth for both Pb(100) and Pb(ll0).Epitaxial behaviour was, however, only observed above 120K indicating that although thermal activation is essential it need not be more than a few kilojoules per mole. There has been a tendency to study metal-oxygen systems in the hope that something new will emerge from electron spectroscopy, and generally speaking the outcome has been disappointing, previous ideas being merely confirmed. However, a few examples do exist where particular problems related to oxidation have been probed, and two of these, the oxidation of copper32 and cobalt,33 have relied on monitoring the shake-up satellites to decide on whether CuII or Cul (and CoII or CoIrl) are present. In the case of cobalt oxi- dation both oxygen pressure and temperature determine whether the surface, when no more than three or four oxide layers thick, is COO or Co30.1 and recourse is made to the detailed studies of McDowell et aZ.34C who had established the relationship between shake-up satellites and the valence state of cobalt with a range of different metal complexes.In the case of oxygen interaction with CU(~OO)~~~in addition to core-level spectroscopy and associated satellite peaks LEED has provided important complementary structural data giving convincing evidence on, for exampIe, the interplay between the two surface oxides involving the CuII and CuI valence states. 31 K. Kishi, R. W. Joyner, and M. W. Roberts. (to be published). 32 (a)S. Evans and J. M. Thomas, Fnraday Disc.Chenr. SOC.,1974, No.58, p. 97; (b)M.J. Braithwaite, R. W. Joyner and M. W. Roberts, Faraday Disc. Chem. SOC.,1975, NO.60, p. 89. 33 R. B. Moyes and M. W. Roberts, J. Catalysis, 1977, (in press). Roberts 5 Molecular Nature of Solids and the Role of the Reactants Finally, I want to discuss briefly some examples (Table 2) we have come across where new surface sites are created under conditions which illustrate the facile nature of many surface processes. Mechanistic aspects are not considered here. We see that the reactant molecules can control the molecular nature of the surface sites thus emphasizing the need to know more about the catalyst surface under reaction conditions when it may bear little relationship to the ‘unused’ catalyst.The temperature region in which the processes are observed (85-295 K) indicates that the activation energies of the molecular processes occurring can not be more than about 15 or 20 kJ mol-1. Perhaps the only example (Table 2) which needs any amplification is the iron-water interaction,37 the implications from the others being self-evident. After the formation of only a monolayer or so of what we suggest is the hydroxide FeO$OH, reaction becomes very slow. Furthermore the surface becomes passive to subsequent oxidation by molecular oxygen even though in the absence of the water vapour pre-treatment the oxide thickness would have been some four or five times greater at the oxygen pressure used. The role of water vapour in metal oxidation is far from understood and more work of this kind should be useful to extract some generalizations which may also be helpful in understanding some aspects of catalytic reactions. Table 2 Examples of facile surface processes and the creation of new ‘surfacesites’ Ref.100 K PbO(s) + H2S(g) +PbS(s) + H2O(g) lo-‘ Torr XPS/LEF.D 34 a,b 295 K PbO(s) + HCOOH(g) +Pb(s) + H2O(g) + C02(g) XPS/LEED Torr 35 295 K CUZO(S)+ NH3(g) +CUNH(S)+ H20(g) XPS 36 10-4 TO^ Fe(s) + H20(g) +FeO-OH 295 K 1O-‘ Tom XPS 37 One obvious point that emerges from the examples I have chosen is the dynamic nature of a catalyst surface. It seems, for example, rather futile to carry out a detailed experimental or theoretical study of an oxide catalyst with a view to understanding its catalytic behaviour if it is not recognized that under catalytic conditions its surface is not an oxide at all, but a chemisorbed sulphide layer on an underlying oxide, or an overlayer of imide-like species on an oxide (Table 2).In this general context it is apposite to mention that studies at high coverages are more likely to be relevant to heterogeneous catalysis while experimental studies (a) K. Kishi and M. W. Roberts, J.C.S. Faraday Z, 1975, 71, 1721; (b) R. W. Joyner,K. Kishi, and M. W. Roberts. (submitted for publication); (c) D. C. Frost, C. A. McDowell, and 1. S. Woolsey, Mol. Phys., 1974, 27, 1473. 35 S. A. Isa, R. W. Joyner, and M. W. Roberts, (to be published). 36 M. H. Matloob and M. W.Roberts, J.C.S. Faraday I, 1977, (in press). s7 M. W. Roberts and P. R. Wood, J. Electron Spectroscopy, 1977, (in press). New Perspectives in Surface Chemistry and Catalysis at low coverage, with possibly single crystals, are more relevant to theoretical calculations. These are not tight compartments, however, and the recent interest in cluster calculations may be one area where one can get close to the ‘real situation’. One particular cluster complex, Fes(CO)1 6C2-, iron carbonyl carbide, is the molecular analogue of carbon monoxide chemisorbed on iron carbide, the central carbon atom being co-ordinated octahedrally by iron atoms at almost the precise Fe-C distance characteristic of bulk iron carbide. Is this relevant to the fact that iron carbide is a good Fischer-Tropsch catalyst for the conversion of carbon monoxide into hydrocarbons? Are the carbon atoms formed during CO dissociation (Figure 3) relevant ? It is questions such as these that are providing much of the impetus to current research in surface chemistry.6 Conclusions What I have attempted to do in this lecture is to discuss some of our recent work, and in this sense it is a rather narrow viewpoint. But I hope it at least illustrates the very dramatic change that has taken place in the approach to tackling prob- lems in interfacial chemistry. The contribution made by electron spectroscopy, and to a lesser extent LEED, to a better understanding of the surface chemistry of some diatomic molecules (nitrogen, carbon monoxide, nitric oxide, and oxygen) has been considered, the interplay between associative and dissociative modes of adsorption emphasized, and the relative stabilities of molecular states on different metal surfaces discussed.A recurrent theme is the significance of the two-way electron transfer process (a-donation and n-back-bonding) and how this can account for the spectroscopic data as well as the overall surface chemistry. The role of the metal catalyst can be seen as a vehicle for pumping electrons into antibonding orbitals, thereby inducing potential reactivity in otherwise unreactive bonds. Although there has been a general tendency to overemphasize the ‘surface physics’ that has emerged from the more recently developed experimental methods we must not lose sight, however, of the relevance the new data have for those of us more interested in the chemical reactivity of solids, and in this context Mason and Textor38 showed how information from organometallic chemistry can facilitate a better understanding of the two-dimensional chemistry emerging from LEED and electron spectroscopy.The ‘new approach’ (see also, for example Spicer et a1.39 and Baron, Blakely, and Somorjai40) will provide insights hitherto inaccessible to the surface chemist, and although the development of a new catalyst will remain the prerogative of the catalytic chemist, its genesis may well be accelerated by an efficient flow of ideas across the ‘interface’. For example, the simple nature of the molecular events referred to in Table 2 is a timely reminder of the difficulties of defining the surface of a ‘working’ catalyst.It is already 38 R. Mason and M. Textor, in ‘Surface and Defect Properties of Solids’, (Specialist Periodical Reports), ed. M. W. Roberts and J. M. Thomas, The Chemical Society, London, 1976, vol. 5, p. 189. 39 W. E. Spicer, K. Y. Yu, J. Lindau, P. Pianetta, and D. M. Collins, in ‘Surface and Defect Properties of Solids’, (Specialist Periodical Reports), ed. M. W. Roberts and J. M. Thomas, The Chemical Society, London, 1976, vol. 5, p. 103. 4o K. Baron, D. W. Blakely, and G. A. Somorjai, Surface Sci., 1974, 41, 45. Roberts obvious that the ‘new’ data have provided a stimulus to the theoretician, and I am thinking here particularly of the cluster-type calculations of Johnson and Slater which are bringing the experimentalists and theoreticians closer together. Both groups, however, have a long way to go to satisfy the industrialist but I believe we are on the right road, with the three milestones I mentioned earlier in my talk well behind us. It is a pleasure to express my gratitude to those many collaborators who have played an essential part in our surface chemistry studies.

ISSN:0306-0012    

DOI:10.1039/CS9770600373

出版商:RSC    

年代:1977

数据来源: RSC

摘要:

Organoborates in Organic Synthesis: The Use of Alkenyl-, Alkynyl-, and Cyano-borates as Synthetic Intermediates By G. M. L. Cragg and K. R. Koch DEPARTMENT OF ORGANIC CHEMISTRY, UNIVERSITY OF CAPE TOWN, RONDEBOSCH, SOUTH AFRICA 7700 1 Introduction It is now twenty-one years since H. C. Brown and Subba Rao demonstrated that alkenes readily react with diborane in the presence of ether solvents to give the corresponding organoboranes in high yield.1 The ready availability of organo- boranes has led to extensive studies of their chemistry and to the development of many reactions of great value in synthetic organic chemistry.2-4 Of prime importance in the development of many synthetically useful reac- tions has been the observation that organoborates, formed by the reaction of organoboranes with appropriately substituted nucleophilic species, tend to undergo spontaneous 1,Zmigrations of an organic group from boron to an acceptor atom as illustrated in reactions 1-4.&B + -02H -+ &B--O--OH + R2M-R + -OH (1) +M+ N-CHC02R’ -+ RS-CH(Na+)(C02R1) -+ R2B-CHRCOzR’ + N2 (3) The organoborate ions formed in the above reactions are generally formed as transient species which undergo rapid rearrangement. Jn the absence of a suit- able activating group, however, many nucleophiles and bases react with organo- boranes to give stable organoborates, which do not undergo further spontaneous reactions. Reaction of some such organoborates with suitable electrophiles can, however, result in synthetically useful transformations, and the application of these reactions has, in recent years, led to the development of many valuable l H.C. Brown and B. C. Subba Rao, J. Anrer. Chem. Soc., 1956, 78, 5694. H. C. Brown, ‘Boranes in Organic Chemistry’, Cornell University Press, Ithaca, New York, 1972. G. M. L. Cragg, ‘Organoboranes in Organic Synthesis’, Dekker, New York, 1973. * H. C. Brown, ‘Organic Syntheses via Boranes’, Wiley, New York, 1975. Organoborutes in Organic Synthesis new synthetic methods.5 The most useful organoborates studied to date have been the alkenyl-, alkynyl-, and cyano-borates, and it is the object of this review to illustrate the application of these compounds to the synthesis of a variety of organic compounds.2 Preparation of Alkenyl-, Alkynyl-, and Cyano-borates Lithium alkenyl-6 and alkynyl-trialkylborate~~may be readily prepared by the reaction of the corresponding alkenyl- or alkynyl-lithium reagents with trialkyl- boranes in an ethereal solvent (reactions 5 and 6). Et-0 0°C R3B + LiCH=CHz [R3B-CH=CHz] Li 1-(5) THF, 0°C -R13B + LiCECR2-[R13B-C~CR2] Li+ (6) Extension of the latter reaction to the preparation of ethynyltrialkylborates (R2= H) can be complicated by disproportionation of the product to ethyne and ethynylbis(trialky1borates)(reaction 7).8 -R3B + NaCECH Et 0 [R3B-CrCH] Na+ -+ [R~B-CGC-~R~]~N~++ HCECH (7)34°C Though lithium ethynyltrialkylborates have been prepared in good yield by reaction of ethynyl-lithium with trialkylborates at -78 "C (reaction S),9 the THF -7S'C -R3B + LiCECH A[R3B-CZCH] Li+ (8) danger of disproportionation can be obviated by use of ethynyl-lithium ethylenedi- amine complex in place of ethynyl-lithium (reaction 9).10 THF r.1 -R3B + LiC=CH,EDA -[RzB-CrCH] Li ' EDA (9) Alternative routes to lithiumll and sodium* alkynyltrialkylborates are shown in reactions 10 and 11; the latter route offers advantages in cases where the E.Negishi, J. Organometallic Chem., 1976, 108, 281. K. Utimoto, K. Uchida, and H. Nozaki, Tetrahedron Letters, 1973, 4527.'A. Pelter, T. W. Bentley, C. R. Harrison, C. Subrahmanyan, and R. J. Laub, J.C.S. Perkin I, 1976, 2419. P. Binger, G. Benedikt, G. W. Rotermund, and R. Koster, Annalen, 1968, 717, 21. H.C. Brown, A. B. Levy, and M. M. Midland, J. Amer. Chem. SOC.,1975, 97, 5017. lo M. M. Midland, J. A. Sinclair, and H. C. Brown, J. Org. Chem., 1974, 39, 731. l1 K. Utimoto, Y. Yabuki, K. Okada, and H. Nozaki, Tetrahedron Letters, 1976, 396. 394 Cragg and Koch -R'X[R~B-CECH] Li + &[R$--C=CLi] Li+ -* THF, -78°C THF-HMPA -78°C to r.t. Hexane, r.t. NaR13BH + HCrCR2-b [R136-C~CR2]Naf presence of base-labile groups precludes preparation of the alkynyl-lithium or sodium reagents. Disodium bis(trialky1borate) salts are prepared in high yield by reaction of ethyne with sodium trialkylborohydrides in benzene (reaction 12),* PhH, r.t.2NaR3BH + HCrCH +[R3&-C=C-BR3] 2Na+ while sodium trialkylcyanoborates are readily prepared by reaction of trialkyl-boranes with sodium cyanide (reaction 13).12 R3B + NaCN THF r.t.[R$-CN] Na+ 3 Synthesis of Alkynes In general, the synthetic approach to alkynes involves treatment of the alkynyl- borate with a suitable electrophilic reagent, thereby promoting migration of an organo-group from boron to carbon; the intermediate alkenylborane then undergoes either spontaneous or base-induced elimination of an organoborane species to yield the alkyne. The most commonly used electrophilic reagent is iodine, and the procedure is illustrated in Scheme 1. . I\,[R13B--CeC-R2] Lif ARl2B-CGC-R2 +R1,BC(R1)=C(R2)I 1 ii R'-C=C-R2 + R'?B-OH Reagents: i, I,; ii, NaOH Scheme 1 Monoynes.-The procedure outlined in Scheme 1 has been applied to the synthesis of internal alkynes (reaction 14).13 12 A.Pelter, K. Smith, M. G. Hutchings, and K. Rowe, J.C.S. Perkin I, 1975, 129. l3 A. Suzuki, N. Miyaura, S. Abiko, M. Itoh, H. C. Brown, J. A. Sinclair, and M. M. Midland, J. Amer. Chem. SOC.,1973, 95, 3080. Organoborates in Organic Synthesis Terminal alkynes (R2= H) are likewise prepared in 75-94% yields from the corresponding ethynyl borates,lO while 2-chloroethynylborates give symmetrical alkynes (reaction 15).14 I, THF -78°C R-CEC-R (48-78%)[R3&CrC--Cl] Li+ -L*(15) The use of methanesulphinyl chloride in place of iodine gives an intermediate alkenylborane which undergoes spontaneous elimination (reaction 16).'6 -MeSOCl[R13B-CrC-R2] Li+ ---+ R1zBC(R1)=C(R2)SOMe-+ R~-C~C-R~+ R'2BSOMe (55-82 %) (16) Attempts to achieve selective alkyl group migration by use of B-alkyl-9-borabi-cyclo[3,3,l]nonane or dialkylthexylborane (thexyl = 2,3-dimethyl-2-butyl) derivatives unfortunately gives mixtures of products resulting from random 1nigrati0n.l~ While migration of the B-cyclo-octyl moiety has been observed in organoborane reactions,lB competitive migration of the bulky thexyl group has not previously been reported.Dimes.-Reaction of dicyclohexyl-or disiamyl-chloroboranes (siamyl = 3-methyl-2-butyl) with two mole equivalents of an alkynyl-lithium furnishes the corresponding dialkyldialkynylborates which, on treatment with iodine, give symmetrical conjugated diynes (reaction 17).l7 TH F I?, -78'CRbBCI + 2LiCrC-R2 +[R1.i(CzsCR2)z] Li+ -R2-C~C-C=C-R2 (70-90 %) (17) (R1= c-CsH11 or MeKHCHMe) Attempted synthesis of unsymmetrical dialkynylborates by sequential reaction of dialkylchloroboranes with two different alkynyl-lithium reagents failed owing to preferential formation of the symmetrical dialkynylborate during the first addition.lS This difficulty was, however, overcome by replacement of the dialkylchloroborane by the dialkylmethylthioborane which gives the initial dialkylalkynylborane stabilized as the methylthio-complex; further reaction with the second alkynyl-lithium reagent then proceeds via dissociation of the complex to the active dialkylalkynylborane as shown in Scheme 2.18 l4 K.Yamada, N. Miyaura, M. Itoh, and A. Suzuki, Tetrahedron Letters, 1975, 1961.l6 M. Naruse, K. Utimoto, and N. Nozaki, Tetrahedron, 1974, 30, 2159. l6 Reference 3, p. 258. l7 A. Pelter, K. Smith, and M. Tabata, J.C.S. Chem. Comm., 1975, 857. la A. Pelter, R. Hughes,K. Smith,and M. Tabata, Tetrahedron Letters, 1976, 4385. Cragg and Koch -!+(c-C6H11)d3SMe [(c-CGHII)~B(~CR*)(SMe)] Li+ i=? LiSMe f An alternative route to unsymmetrical conjugated diynes involves formation of the methoxy-stabilized dialkylalkynylborane c~mplex.~g While the overall yields of diynes are slightly higher than those obtained using the methylthio- complex, this method suffers from the disadvantage that the reactive dialkyl- alkynylborane is only freed for further reaction on treatment with boron trifluoride (Scheme 3).I -iii Sia2BOMe-* [SiazB-CrCR1(OMe)] Li + 1:Sia2B-C=C-R' -+ -[Sia2B(CECR1)(CrCR2)]Lif R*--C=:C-C=C-R2 (Sia = MeKHCHMe) (61-73%; g.1.c. yields) Reagents: i, LiC=CR1, THF, -78°C; BF,, Et,O, THF, -78"G25"C; iii, LiC=CR2, -78°C; iv, I,, -78°C Scheme 3 Enynes-A similar approach to that used in the synthesis of conjugated diynes has been applied to the synthesis of conjugated trans-enynes as shown in Scheme 4.20 iii iv R'CHSH-CEC--K2 (60-747;; > 99XE) Reagents: i, THF, 0°C; ii, LiC=CRp, THF, -50°C; iii, 12, THF, -78OC-25"C; iv, NaOH Scheme 4 The method has been successfully applied to the synthesis of the pheremones, bombykol (1)20 and (7E, 9Z)-dodecadienyl-l-y1 acetate (2).21 l9 J. A. Sinclair and H. C. Brown, J. Org. Chem., 1976, 41, 1078.20 E. Negishi, G. Lew, and T. Yoshida, J.C.S. Chem. ComM., 1973, 874. E. Negishi and A. Abramovitch, Tetrahedron Letters, 1977, 41 1. Organoborates in Organic Synthesis The efficiency of the alkynylborate synthetic approach is highlighted by the fact that the overall yield of (2) obtained was 40 % ( > 98 % pure), compared with a yield of 10 % (-80 % pure) obtained in an earlier synthesis.22 Application of the above method to the synthesis of cis-enynes is not as efficient, only proceeding in yields of 30-50% because of competitive migration of the siamyl group23 4 Synthesis of Alkenes The conversion of alkynylborates into alkenes viaelectrophilicattackand protono-lysis of the intermediate alkenylborane has been thoroughly investigated by a number of groups.In general, mixtures of the E and 2 stereoisomers are ob- tained, but the stereoselectivity of the boron to carbon alkyl group migration has been found to be high in a number of cases. The transformation of alkynyl- borates to alkenes thus constitutes a valuable addition to the many methods available for the stereoselective synthesis of alkenes.24 Terminal A1kenes.-Protonation of the ethylenediamine-stabilized complexes of ethynylborates (reaction 9) with propanoic acid gives monosubstituted alkenes (reaction 18),25 while treatment of the intermediate alkenylborane with iodine and base gives the corresponding symmetrical disubstituted alkenes (reaction 19).9 EtCO H[R~~CFCH]Li+,EDAA R-CH-CH. (56-64%) (18) i HCI THF -78'C 11, NaOH, -78 C [R~B-CECH] Li+-*P RzB-C(R)=CHz NaOH'., R~C=CHB(76% R = C; CsHii) (19) R3B + Li-CMe=CH2 + [R&-CMe=CH?] Lif Unsymmetrical disubstituted alkenes may be prepared via alkylation with dihalo- genomethanes, followed by protonolysis (Scheme 5).26 22 J.N. Labovitz, C. A. Hendrick, and V. L. Corbin, Tetrahedron Letters, 1975, 4209. 23 E. Negishi, R. M. Williams, G. Lew, and T. Yoshida, J. Organometaflic Chem., 1975, 92, C4. 24 J. Reucroft and P. G. Sammes, Quart. Rev.,1971, 25, 135. 25 N. Miyaura, T. Yoshinari, M. Itoh, and A. Suzuki, Tetrahedron Letters, 1974, 2961. 26 A. Pelter and C. R. Harrison, J.C.S. Chem. Comm., 1974, 828. Cragg and Koch x(68-74 7:) Reagents: i, CH2X2,DG, 0°C; ii, PriCOzH Scheme 5 Iodination of trialkylisopropenylborates,followed by oxidation, provides a novel synthetic route to terpenoids (reaction 20).27 R3B + LiCMe=CH2 [R3E--CMe=CH2] Li+ i, 1:.-78°C to 0°C yp~ R-CMe=CH2 (73-100X) (20)11, H,O,, OH-The selective transfer of an alkyl group from the corresponding B-alkyl-9- borabicyclo [3,3,1 Inonane borate derivatives greatly enhances the synthetic utility of the method (Scheme 6).27 Reagents: i, Iz, -78 "C4"C;ii, HzOz,OH-Scheme 6 Disubstituted Internal A1kenes.-Protonation of alkynylborates generally gives mixtures of the E and 2 stereoisomers,* though the ratio of stereoisomers has been found to be dependent on the acid used as the proton source. Thus, with methanesulphonic acid the E-alkene predominates,28 whereas with propanoic acid the Z-alkene is preferentially formed.25 Another interesting feature of the reaction is the marked effect exerted by the presence of a phenyl group, either in the alkyne moiety25 or as the migrating group.28 In both these cases the Z-alkene is predominantly formed (reactions 21 and 22).27 N. Miyaura, H. Tagami, M. Itoh, and A. Suzuki, Chem. Letters, 1974, 1411. 28 A. Pelter, C. R.Harrison, C. Subrahmanyan, and D. Kirkpatrick, J.C.S. Perkin I, 1976, 2435. 399 Organoborates in Organic Synthesis EtCO,H, -7S'C [Bun&--CC-Ph] Li+-BunCH=CHPh (817;; 98 3; Z) (21) i, MeSO,H, DG, 0°C [Ph&-CeC-Bun] Li+ PhCH=CHBun (81 %, 98 :;Z) (22)11. HOAc Protonation of diaIkylalkynylthexylborates gives a predominance of the E-alkene irrespective of whether methanesulphonic or propanoic acid is used as proton source (reaction 23).2* [(Me2CHCMe2)ER1z(C=CR2)]Lii ,ZH+_R*CH=CHKZ (, > 70 "/, E) (23)ii, RC0,H For the phenylalkynylmoiety (R2 = Ph), however, the 2-alkene predominates to an extent exceeding 95 %.28 Stereospecific synthesis of Z-alkenes has been achieved using dialkylchloro- boranes29 or tributyltin chloride30 as the electrophilic reagents (reactions 24 and 25).R1CH=CHR2 (65-81 %; Z only) (24) R1CH=CHR2 (70-797,;; Zonly) (25) The stereoselective synthesis of E-alkenes by reaction of bis(trialky1) ethynyl- borates with one mole equivalent of cyanogen bromide proceeds with double migration of alkyl groups (reaction 26).31 BrCN Et 0 [R3&-C=C--BR3] 2Li+ RCH=CHR (46-88 %; E) (26)NsOMe.r.t. The application of alkenylborates to the synthesis of internal alkenes has been elegantly demonstrated in studies related to prostaglandin synthesis.32 The reac- tion sequence involves iodination of an intermediate boronic ester as shown in Scheme 7. '' P. Binger and R. Koster, Tetrahedron Letters, 1965, 1901. :lo J. Hooz and R. Mortimer, Tetrahedron Letters, 1976, 805. " N. Miyaura, S. Abiko, M. Itoh, and A. Suzuki, Synthesis, 1975, 669. 32 D. A. Evans, T. C. Crawford, R. C.Thomas, and J. A. Walker, J. Org. Chm., 1976, 41, 3947. Cragg and Koch (58%; '99 %El R = CH(OTHP)(n-C,HI1); THP = Tetrahydropyranyl Reagents: i, 2-LiCH=CHR, THF, -45OC; ii, Is, NaOMe, MeOH-THF Scheme 7 Use of the E-Iithioalkene gives the corresponding Z-cyclopentylalkene in 75 % ~ield.3~ Trisubstituted A1kenes.-The reaction of sodium trialkylalkynylborates with a variety of alkylating agents has been found to give, after hydrolysis, mixtures of E-and Z-alkenes (reaction 27)329 [R13B-CsC--K2 J Na+ i El+ R1CH=CR2RJ (E :Z -65 :35) (27)11, H,O A detailed investigation of the reaction of lithium and sodium trialkylalkynyl- borates with a variety of alkylating agents indicated that the E to 2 stereoiso-meric ratio (w 65 :35) is insensitive to the alkylating agent used, the nature of the solvent, and the counter cation present.' As observed in the case of disubstituted alkene synthesis (reactions 21 and 22), the presence of a phenyl group in the borate intermediate greatly enhances the stereoselectivity of the reaction, the E-stereoisomer being preferentially formed to an extent exceeding 82%.7 The reaction of dialkylalkynylthexylborates proceeds with greatly increased stereoselectivity, but, once again, the E :Z stereoisomeric ratio varies little with the nature of the alkylating agent or solvent (reaction 28).7 i.El+, DG, -78°C [(MezCHCMez)ERlz(CrCR2)] Li+-R1CH=CR2R3 11. PriCOIH (El+ = MezS04. EtsO+BF4-, R31, allyl or PhCHzBr; R3= alkyl, allyl, PhCH2) The major stereoisomer formed in each case is that in which the migrating group (R1)and thegroup introduced from the alkylating agent (R3)are cis to oneanother. In no instance was migration of the thexyl group observed, though it should Organoborates in Organic Synthesis be noted that the use of groups other than n-alkyl groups (R1)in the dialkyl- thexyl moiety was not reported.A valuable extension ofthe above reaction involves the use of a-bromo-ketones and -esters, iodoacetonitrile, and prop-2-ynyl bromide as alkylating agents (reaction 29).33 [R13fi-C=C-R2] Li i, DG, -78'C -+XCHzY+ to r.t. R1CH=C(R2)CH2Y (29) ii, HOAc (64-77 7;;R', CHzY cis) (X = Br or I; Y = COMe, COPh, CO.Et, CN, CrCH) The reaction proceeds in a stereospecific manner with all the reagents indicated (except a-bromo-esters for which 88-96 % stereoselectivity is observed) to give products in which the migrating group (R1)and the alkylating group (CH2Y) are cis to one another.The stereoselectivity of the reaction with a-bromo-esters is increased to > 98 % when dialkylalkynylthexylborates are used.33 Similar routes to allylic amine~~~ and methyl ethers35 have been reported. While equi- molar mixtures of E and 2 stereoisomers are obtained in both cases, the cor- responding Z-alkenylboranes are selectively hydrolysed as shown in Scheme 8. i, ii R1\ /CH,Y "qR'[R1,B-C=C-R2] Na+ --+ C=C, + R',B, H' R2 Y (X = Br; Y = NMe,; > 90%) (X = CI; Y = OMe; ,45-75%) Reagents: i, XCH,Y, Et,O; ii, H20, r.t. Scheme 8 The unchanged (E)-alkenylboranes may be converted into the corresponding allylic derivatives by treatment with triethylaluminium and acid in the case of the amines34 and acetic acid in the case of the methyl ethers.35 The protonation-iodination reaction sequence (reaction 19) has also been applied to the synthesis of trisubstituted alkenes (reaction 30).2*~36 HCl, THF I NaOMe [R%B-C=C-R2] Li+ ___+ R12BC(Rl)=CHR2 11,-70°C MeOH, -20°C R12C=CHR2 (64-79 %) (30) :13 A.Pelter, K. J. Gould, and C. R. Harrison, J.C.S. Perkin I, 1976, 2428. P. Binger and R. Koster, Chem. Ber., 1975, 108, 395. 35 P. Binger and R. Koster, Synthesis, 1974, 350. 36 G. Zweifel and R. P. Fisher, Synthesis, 1975, 376. Cragg and Koch A useful application of procedures discussed this far is illustrated by the synthesis of propylure, the sex attractant of the pink Bollworm moth (Scheme 9).3' [Prn3-&CzC-SiMe3] Li+ + TsO(CH2)2--CrC--(CH2)40THP -% 4 stepsPr%C=C(SiMes) (CH~)~-CGC--(CH~)~OTHP -+ Prn2C=CH(CH2)2CH &H(CH2)40Ac Propylure (30%overall yield) (THP = Tetrahydropyranyl) Reagents: i, THF, -78°C; ii A; iii, I,, NaOH, 0°C to r.t.Scheme 9 Epoxides also serve as electrophilic reagents in promoting the transformation of alkynylborates into alkenes; such reactions have been used in an elegant synthesis of hydroxy-alkenes (Scheme lO).38 Heating the reaction mixture results in exclusive formation of the cyclic borane (3); in dichloromethane or tetra- hydrofuran, however, mixtures of E and 2 isomers are formed.38a The reaction, Lit -R3 (3) R1CH=C(R2)CH,CH(OH)R3 -(70-98 "A; R', R2cis) Reagents: i, A 2 5 h; ii, HOAc, d Scheme 10 however, proceeds in low yield for more hindered epoxides, such as cyclohexene oxide.38 The reaction of acetyl chloride with alkynylborates proceeds via double migration of alkyl groups to give, after oxidation, afi-unsaturated ketones (reaction 31).39 O7 K.Utimoto, M. Kitai, M. Naruse, and H. Nozaki, Tetrahedron Letters, 1975, 4233. 38 M. Naruse, K. Utimoto, and H. Nozaki, Tetrahedron, 1974, 30, 3037. 38(3. K. Utimoto, T. Furubayashi, and H. Nozaki, Gem. Letters, 1975, 397. 39 M. Naruse, T. Tomita, K. Utimoti, and H. Nozaki, Tetrahedron, 1974, 30, 835. 2 403 Organoborates in Organic Synthesis -i, MeCOCI. TIIF, O'c[Ki.~B--CC-R2] K1K=C(R2)COMe (30-43n0) (31)Li ' -11, CrO,, Mr,CO Other electrophilic reagents investigated are the cationic metal complex (4)40 and N-acetylpyridinium chloride (5) (Scheme 11).41 The products are the OMe A(CO), I COMe I COMe (E + Z) (7) Scheme 11 alkenylborane (6) and 4-alkenyl-su bstituted- 1,4-dihydro-pyridines (7) respective-ly.Trisubstituted alkenes have also been prepared by the reaction of bis(trialky1)- ethynylborates with a 2: 3 molar ratio of cyanogen bromide (reaction 32).31 [K~~--C~C--BRR] i, 2BrCN, Et,O, NaOMc, 1.1.2Lif I~,rtcoltt K.C=CHK (SO 760/0) (33) The synthetic utility of this reaction appears to be limited by the formation of significant amounts of di- (reaction 26) and tetra-substituted alkenes as by-products. Highly stereoselective routes to trimethylsilylalkenylboranes are illustrated in reactions (33)42 and (34).43 The ready conversion of the trimethylsilyl group into other functional groups enhances the synthetic utility of these processes.*O A. Pelter, K. J. Gould, and L. A. P. Kane-Maguire, J.C.S. Chem. Comm., 1974, 1029. 41 A. Pelter and K. J. Gould, J.C.S. Chem. Comm., 1974, 347. O3 P. Binger and R. Koster, Synthesis, 1973, 309. 43 R. Koster and L. A. Hagelee, Synthesis, 1976, 118. 404 Cragg and Koch Tetrasubstituted A1kenes.-Reaction of bis(trialky1)ethynylborates with a three molar ratio of cyanogen bromide gives low yields of tetrasubstituted alkenes, together with large amounts of trisubstituted alkenes (reaction 32), the former only being formed with primary alkyl groups, whereas the cyclic intermediates (3; Scheme lo), formed in the reaction of alkynylborates with epoxides, react, with iodine to give alkenyliodides (reaction 33.58 i, d(-R’Li)(3; Rl = Bu”, R? = n-C5H11, R3 = Et) __j (see Scheme 10) ii, 12.NaOH R1C(I)=C(R~)CHKH(OH)R3(71 x;RI,R2cis) (35) Dienes.-2-Bromo-6-lithiopyridine readily reacts with trialkylboranes with stereospecific cleavage of the pyridine ring to give conjugated dienylnitriles as shown in Scheme 12.44 Treatment of the intermediate dienylborane with iodine I ..CN H N HH (60-939;: ; R = alkyl) Reagents: i, R,B,Et,O; ii, HOAc Scheme 12 and base gives the dienylnitrile, R&=CH-CH=CHCN, in 50-67 % yields.44 Mechanism and Stereochemistry of Alkene Formation.-The results discussed in the foregoing sections clearly indicate that the formation of alkenes is, in general, not stereospecific.This observation rules out the possibility of the trans- formations occurring by a concerted mechanism involving the migrating group and incoming electrophile being stereospecifically trans to one another. In fact, careful studies by Pelter et al. have shown that the stereochemistry of the major alkene formed in the alkylation of trialkylalkynylborates is that requiring the cis arrangement of the migrating group and the incoming alkyl group (reactions 28 and 29).7133 A mechanism involving the initial formation of a linear vinyl cation which is in equilibrium with two possible sp2-hybridized vinyl cations has been proposed by Suzuki et al. for the protonation of alkynylborates (Scheme 13).25 This general mechanism does not, however, account for the influence of a thexyl (reaction 23) or a phenyl group (reactions 21 and 22) on the stereo- chemistry of the reaction, nor for the effect of the nature of the acid used in the protonation step.25s28 In addition, conclusive evidence that the alkylation of 44 K.Utimoto, N. Sakai, M. Obayashi, and H. Nozaki, Tetrahedron, 1976, 32, 769. Organoborates in Organic Synthesis Scheme 13 alkynylborates proceeds via a largely kinetically controlled process has been presented (Scheme 14).' [(n-C,H,,),k2~C-C,H7-n] Li+ (68%; E:Z = 29:7)---k n-C,H,,CH=C(C,H ?)C,H,,-n 111, 1v(697;; E:Z = 71 :29) [( n-C ,H ,3)3 -C C -C H I3-n ] L i + Reagents: i, n-C6H131, THF; ii, H+; iii, n-C,H,I, THF; iv, H+ Scheme 14 Equilibration via a common vinyl cation species should lead to the same mixture of products from each reaction in Scheme 14.Thermodynamic control is thus clearly ruled out for the alkylation process; in addition, the stereochemistry is found to be insensitive to the nature of the solvent, the counterion, and the alkylating reagent used, but the rate of reaction is strongly influenced by the alkylating reagent. A two-stage mechanism involving rate-determining attack on the alkynyl moiety to give a bent vinyl cationic intermediate, which undergoes rapid exo- thermic rearrangement with preferred retention of stereochemistry at the migration terminus, has been proposed to account for the above observations (Scheme 15).7 Although the intermediate is formally represented as a zwitterion, MIND0/3 calculations indicate only small charges on the boron and carbon, thus support- ing the non-polar characteristics of the process.Calculations have also indicated Cragg and Koch 1 ;+ Rl,B l-' R',B\ 7' Rapid \ iR2 [R1,B-C=C-R2] Li+ (major)c=c -slow ;rate \determining + x-R3 R' R3 Reagent: i, R3X Scheme 15 an increase in stereoselectivity of reaction (with retention of configuration at the migration terminus being favoured) for rearrangements involving bulky substituents on boron, thus supporting the increase in stereoselectivity observed for thexyl derivatives (reaction 28).7 5 Synthesis of Ketones Organoborates have found extensive application in the synthesis of ketones and, of the stable organoborates, cyanoborates were the first to be extensively investigated.12 Of the electrophilic reagents studied, acylating agents, in par- ticular trifluoroacetic anhydride, proved to be the most effective in promoting migration of groups from boron to carbon (reaction 36).12 [R3&-CN] Na' I* (CF,CO)IO,DG.-78 C to O'C~RzCO(84--100%) (36) 11. H,O,, NaOH The relative migratory aptitudes of alkyl groups is in the order n > s > t, and migration of groups occurs with complete retention of configuration.45 Diglyme is found to be superior to tetrahydrofuran as solvent, since some triple migration occurs in the latter solvent to give trialkylmethanols (reaction 39), particularly when tri-n-alkylcyanoborates are used.Unsymmetrical ketones are readily synthesized by use of dialkylcyanothexylborates(reaction 37).12 The transformation occurs by the mechanism outlined in Scheme 16. 'j A. Pelter, M. G. Hutchings, K. Smith, and D. J. Williams, J.C.S. Perkin I, 1975, 145. Organoborates in Organic Synthesis [R,&C=N]Na'. (CF,CO),O + R,B-C=N-COCF, + e If R-BkRN H,O,, OH -R,CO Scheme 16 A most useful application is the ready synthesis of fused ring systems (Scheme 17).12 Reagents: i, ThBH,, THF, 0°C;ii, KCN, r.t.; iii, (CF,CO),O, -78°C to r.t.; iv, H,O,. NaOH Scheme 17 Use of acids as the electrophilic reagents generally gives inferior yields of ketones.46 A. Pelter, M.G.Hutchings, and K. Smith, J.C.S. Perkin I, 1975, 142. Cragg and Koch Oxidation of the intermediate alkenylboranes obtained upon treatment of alkynylborates with suitable electrophilic reagents (Scheme 15) gives ketones in high yields. Alkynylborates thus constitute extremely valuable intermediates in tbe synthesis of a wide variety of ketones (reaction 38). El + HZO2[R13S-CsC-R2] +R1zB-C(R1)=C(R2)EI +RlCOCH(R2)El (38)W+Z) NaOH Thus, many of the reactions discussed in Section 4 on the synthesis of alkenes have also been applied to the synthesis of ketones. Some results of these studies are given in the Table. TabIe Conversion of alkynylborates into ketones (reaction 38) Electrophilic reagent Ketonea Yield Ref. El + R1COCH(R2)EI % HCI Rf CO Me0 70-90 9 MeS03H R1COCH2R2 78-86 28 BdlaSnCl R1COCH2R2C 88 30 MeOTs R1COCH(Me)R2 88-98 7 Me2S04 R1COCH(Me)R2 81-85 7 EtsO +BF4- R1COCH(Et)R2 83-95 7 CH2=CHCH2Br R1COCH(R2)CH2CH=CHz 80-88 7 PhCHzBr RICOCH( R2)CHzPh 78-81 7 BrCHzCOzEt RICOCH( R2)CHzCOzEt 74-78 33 BrCHzCOMe R1COCH(R2)CH2COMe 70-75 33 BrCH2COPh R1COCH(R2)CH2COPhd 74 33 B~CH~CECH R1COCH(R2)CH2CrCH 75-80 33 ICHzCN R1COCH(R2)CH2CN 62-7 1 33 R1COCH(R2)CH2CH(OH)RS 59-95 38 0 The results listed in Table 1 serve to illustrate the tremendous synthetic utility of these reactions.Groups R1, R2, and El (reaction 38) all originate from simple and independent units which can be varied to give a wide range of products. A particularly valuable aspect of the reaction is the potential to achieve regio- specific a-substitution of a ketone, RlCOCHzRZ, by a group El whether it be alkyl, allyl, benzyl or any of the other groups listed in Table 1.6 Synthesis of Alcohols As in the synthesis of ketones (Section 5), cyanoborates were the first of the stable organoborates to be applied to the synthesis of alcohols.47 Treatment 47 A. Pelter, M. G. Hutchings, K. Rowe, and K. Smith, J.C.S. Perkin I, 1975, 138. Organoborates in Organic Synthesis with excess trifluoroacetic anhydride promotes migration of three groups from boron to carbon (Scheme 16; reaction 39). [R3B--CN] K i. Excess (CF,CO):O, DG, WO'C P RXOH (60-90%) (39)ii, H,O,, NaOH Ready transfer of n-alkyl group occurs, while migration of s-and t-alkyl groups is accelerated by the addition of 10% pyridine to the reaction mixture.47 The use of alkenyl- and alkynyl-borates is illustrated in reactions (40),9~48 (41),49 and (42).50 [R12kC=C-R2] llcl-l~>o Li+ +RIB(OH)--CR1~CH&Z H&* THF, A NaOH R'K(0H)CHzR~ (30-86%) (40) I ICI H10, +RlCH?C( BR3kle)R2RR:' +R1CH2C(OH)R2R3 (69-76 %) (41)Et,O NaOti 1,4-Diols may be synthesized by reaction of alkenylborates with epoxides as shown in reaction (43).g R1CH(OH)CH,CH,CH(OH)R2 (90-1007;) (431 7 Protonation, Protonolysis, or Dehydroboration ? Treatment of alkynylborates with organic acids (reactions 21 and 22) gives alkenylboranes which may be treated further as desired.Similar results are obtained on treatment with aqueous acid at -78 "C(reaction 19), but at higher temperatures further protonation and rearrangement of the alkenylborane occurs (reaction 40).On reaction with anhydrous acid (HCl), alkenylboranes undergo protonolysis to give alkene~;~g alkenylborates, however, undergo protonation and rearrangement to give trialkylboranes, which may react further (react ion 4 1). Treatment of alken yldialkylmethoxyborates with aqueous I* M. M. Midland and H. C. Brown, J. Org. Chem., 1975, 40,2845. G.Zweifel and R. P. Fisher, Synthesis, 1974, 339. 50 A. B. Levy and S. J. Schwartz, Tetrahedron Letters, 1976, 2201. 410 Cragg and Koch acid also results in protonation and rearrangement (reaction 42); in the case of a1 kenyl trial kyl borates, however, prot onat ion-rearrangement is followed by rapid dehydroboration to give alkenes (reaction 44).51*52 1 M-HCI[BunCH=C(Bun)~(c-C,H,,),Bun] Li+ -BunCH,C(Bun) In general, base treatment of alkenyltrialkylborates gives the corresponding alkenes, thus providing a useful alternative to the normal protonolysis procedure of treatment of alkenylboranes with carboxylic acids (reaction 45).52 BPLi NaOHR1:!B-C(R1)=CHR2 +[Bur1R1zE-C(R1)==CHR2] Li' --+ R1CH=CHR2 (86-95 ;<) (45) The above results clearly indicate that care must be exercised to controlling the conditions of reaction when using acids.8 WhyOrganoboron Derivatives? The synthetic versatility of organoboranes has been amply demonstrated;2-4 organoborates clearly possess equal promise.What properties give organoboron compounds such unique potential in organic synthesis ? Their relative unreactivity to many functional groups compared with other organometallic reagents, such as Grignard and alkyl-lithium compounds. The ready synthesis of organoboranes of differing structures.53 In this respect, the tolerance of many functional groups by the hydroboration reaction permits the synthesis of organoboranes and organoborates containing a wide variety of functionalities. The complete retention of configuration during the transfer of asymmetric groups in reactions proceeding by ionic co-ordination mechanisms. The broad scope of the reactions. Most of the reactions discussed in the fore- going article proceed in high yield with compounds containing alkyl, cycloalkyl, and aryl groups.Low yields resulting from the transfer of only one of the R groups from a R3B moiety may often be overcome by use of B-alkyl-9-borabi- cyclo[3,3,1 Inonane or dialkylthexylborane (thexyl = MezCHCMe2) derivatives in which selective transfer of the alkyl group is achieved.54 With the potential of organoboron compounds in organic synthesis firmly K.-W. Chiu, E. Negishi, and M. S. Plante, J. Organometallic Chem., 1976, 112, C3.'' E. Negishi and K.-W. Chiu, J. Org. Chem., 1976, 41, 3484. I3 K. Smith, Chem. SOC.Rev., 1974, 3, 443. 54 Reference 3, Section 8.2.2. 411 Organoborates in Organic Synthesis established it is hoped that these compounds will be fully exploited in devising new or improved synthetic routes to complex organic molecules. The authors are indebted to Professor Andrew Pelter for helpful comments on the manuscript.

ISSN:0306-0012    

DOI:10.1039/CS9770600393

出版商:RSC    

年代:1977

数据来源: RSC

摘要:

CENTENARY LECTU€W* Systematic Development of Strategy in the Synthesis of Polycyclic Polysubstituted Natural Products :The Aconite Alkaloids By Karel Wiesner NATURAL PRODUCTS RESEARCH CENTER, THE UNIVERSITY OF NEW BRUNSWICK, P.O. BOX 4400, FREDERICTON, NEW BRUNSWICK, E3B 5A3, CANADA 1 Introduction During approximately the last 12 years my students and I at the University of New Brunswick have been involved in systematic studies which have had as the final goal the development of relativeIy simple and highly efficient methods for the synthesis of delphinine-type alkaloids. We wished to develop a synthetic strategy in which all the many functional groups of the delphinine (1) system would materialize in the correct positions and configurations simultaneously with the construction of the hexacyclic polybridged skeleton. It was my belief that this exercise, i.e.a systematic search for the simplest possible method to construct a complicated compound, would significantly contribute to the art of synthesis and advance the day when compounds of the complexity of delphinine (if sufficiently important and expensive) might be produced on the industrial scale. A development of this type was seen after the second world war in the total synthesis of the far simpler steroids. This development motivated by the medicinal importance of steroid hormones and initiated by the classical syntheses of Robinson and Woodward finally lead to the design of the extremely simple and highly practical synthetic strategies of Torgov and J0hnson.l While many laboratories participated in the progressive simplification of steroid synthesis over many years, we came to regard this process as the very objective of our work, similar to a chess problem in which the task is to defeat the opponent in the smallest number of moves.* Based on the Centenary Lecture of the Chemical Society given in Lancaster in March 1977. For comprehensive discussions of steroid total synthesis see: (u) A. A. Akhrem and Yu. A. Titov, ‘Total Steroid Synthesis’, Plenum Press, New York, 1970; (6) R. T. Blickenstaff, A. C. Ghosh, and G. C. Wolf, ‘Total Synthesis of Steroids’, Academic Press, New York, 1974; (c) R. Pappo in ‘The Chemistry and Biochemistry of Steroids’, ed. N. Kharasch, IntraScience Research Foundation, Santa Monica, Calif., 1969, vol.3, No. 1, pp. 123-140; (d)G. Saucy and N. Cohen, MTP Znternat. Rev. Sci., Ser. One, 1973, 8, 1-26. Synthesis of Polycyclic Polysubstituted Natural Products The technique which we used in the search for simplicity and efficiency is a common one in engineering. The key reactions of a first generation synthetic design were tested and modified on model compounds and then the synthesis proper was carried out. On the basis of the experience gained and accidental discoveries made in the first generation synthesis a second generation synthetic design was worked out, tested on models and carried out and the process then repeated. Thus, the third generation synthesis is now finished and the fourth and final generation is under way.It will be seen that this technique attains to a high degree the objectives and standards that we have initially set for ourselves. The choice of the aconite alkaloids as our target naturally gave us a consider- able advantage. Our degradation studies which resulted in the first correct structure proposals for this class of natural products2 gave us a good under- standing of the chemical properties of these systems. However, the advantage was hardly unfair. A similar understanding may also be clearly gained by a thorough study of the literature and thus our choice was motivated mainly by the pleasure a chemist can derive from the synthesis of ‘his own’ compounds which had seemed so formidable when their structures were first ~larified.~ OMe Q (3) (4) Scheme 1 cf.for example K. Wiesner, Pure and Applied Chemistry, 1975, 41, 93. R, B. Woodward, ‘XIVth International Congress of Pure and Applied Chemistry’, Main Congress Lectures, Birkhauser Basel and Stuttgart, 1955. Wiesner 2 The Aromatization Product of Delphinine Our first approach to the delphinine system is illustrated in Scheme 1. It was still lengthy, classical, and was undertaken before I had clearly formulated my present ideas. The aromatization product (2) obtainable in high yield by degradation of delphinine (1) was intended to serve as a relay compound.* The starting material was the substituted tetralone (3) which was oxidized to the aldehyde (4).Base catalysed aldolization of this last product gave the synthon (5) in a high yield. The disposition of functional groups in (5) is very convenient and it permitted a relatively easy elaboration of this intermediate to the desired compound (2). Finally, a differential reaction of the racemate (2) with 1-camphorsulphonyl chloride resulted in a very efficient resolution and completed the total synthesis of the optically active relay (2).5 The routes from (2) to (1) which we studied on model compounds,6 while entirely feasible, were unattractive, pedestrian, and if undertaken, would have absorbed the major part of our energy. Thus, this approach was abandoned in favour of more sophisticated methods to be described later. 3 The Synthesis of Talatisamine Many years ago Professor Cookson and ourselves proposed independently that delphinine-type alkaloids might originate from the atisine system by a loss of a carbon atom, followed by bridging and then a rearrangement.7 The order of these steps in the biosynthesis was, of course, unknown and the rearrangement could occur before or after the bridging of ring B.The first laboratory implementation of the postulated rearrangement was accomplished by Overtons who reported the formation of the rearranged ketone (7) on high temperature pyrolysis of the atisine derivative (6). However, no way was found to introduce the ring B bridge (i.e. a bond between the asterisked carbons) in compound (7). We studied the rearrangement of the readily obtainable model compound (8) and found that this material gives a high yield of the products (9) and (10) in equal amountsg when heated to 180"C in dimethylsulphoxide with tetramethyl guanidine.Since in degradation products of delphinine-type alkaloids which are analogous to compound (9) the ring B bridge can be easily closed,lo we decided to base our second generation synthesis on this model. K. Wiesner, M. Gotz, D. L. Simmons, and L. R. Fowler, Coll. Czech. Chem. Comm., 1963, 28,2462. K. Wiesner, E. W. K. Jay, T. Y. R. Tsai, C. Demerson, Lizzie Jay, T. Kanno, J. Kiepinskf,A. Vilim, and C. S. Wu, Canad. J. Chem., 1972, 50, 1925. K. Huber and J. Poslusny, unpublished data. cf. K. Wiesner and Z. Valenta in 'Progress in the Chemistry of Organic Natural Products', ed.L. Zechmeister, Springer, Vienna, 1958, vol. XVI, p. 26. 'J. P. Johnston and K. H. Overton, J.C.S. Perkin I, 1972, 1490. H. J. Wu, Ph.D. Thesis, University of New Brunswick, 1977. K. Wiesner, M. Gotz, D. L. Simmons, L. R. Fowler, F. W. Bachelor, R. F. C. Brown, and G. Buchi, Tetrahedron Letters, 1959, 15; 0.E. Edwards, L. Fonses, and Leo Marion, Canad. J. Chem., 1966, 44, 583; 0. E. Edwards, Chem. Comm., 1965, 318. 415 Synthesis of Polycyclic Pofysubstituted Natural Products I I TsO Ic-j+) Me0 Me0 f As the target we chose talatisaminell (1 l), a somewhat simplified version of delphinine with two substituents missing. The tetracyclic ‘aromatic intermediate’ (13) was elaborated starting with the addition of the dienophile (12) and trans,trans-l,4-diacetoxy-l,3-butadiene.The conversion of (13) to the tetrasubstituted ‘noratisine’ intermediate (1 8) was performed along the lines of the synthesis of atisine which we published some time ago.l2 Photoaddition of allene to compound (14) [obtained by Birch- reduction, reacetylation, and acid treatment of (13)] gave stereospecifically the adduct (15) predictable by our addition rule.13 Compound (1 5) was transformed to the acetal (16) and this material gave the product (17) by ozonolysis and borohydride reduction of the ozonide.Finally, treatment of (17) with aqueous acid resulted in the unmasking of the keto group, retroaldol reaction and an immediate stereospecific aldol condensation to (1 8).This last compound was now elaborated by standard methods to the highly crystalline key intermediate (19) (Scheme 2) the structure of which was corroborated by an X-ray crystallo- graphic determination. The rearrangement of (19) proceeded as in our model study and gave a 45 % yieid of the intermediate (21) besides an equal amount of the useless by-product (20). l1 M. A. Khaimova, M. D. Palamareva, N. M. Mollov, and V. P. Krestev, Tetrahedron, 197 1 27, 819. la R. W. Guthrie, Z. Valenta, and K. Wiesner, Tetrahedron Letters, 1966, 4645. l3 K. Wiesner, Tetrahedron, 1975, 31, 1655. 416 Wiesner HO HV (15) R’= 0; R2 = =CH2 (16) R1=OCH2CHPO ; R2 = =CH2 (17) R1= OCH2CH20; R2 = H and OH Deacetalization of compound (21) followed by lithium aluminium hydride reduction and mercuric acetate oxidation of the product yielded finally compound (22) which cyclized spontaneously to talatisamine (1 l).2J4 4 The Synthesis of Chasmanhe and Napelline Although talatisamine was the first delphinine-type alkaloid to be synthesized, it was quite clear that its construction did not meet our requirements of sim-plicity and high efficiency.The main shortcomings of the synthesis were as follows: (i) The formation of the useless by-product (20) in equal amounts together with the synthetic intermediate (21) on rearrangement of the ‘noratisine’ derivative (19). (ii) The low yield (40%) of the mercuric acetate introduction of the ring B bridge, this low yield is clearly due to the non-regiospecscity of the oxidation.(iii) The rather lengthy process which we had to use to introduce the,!3-tosyloxy group in compound (19). l4 K. Wiesner, T. Y. R. Tsai, K. Huber, S. E. Bolton, and R. Vlahov, J. Amer. Chern. SOC., 1974,96,4990. 417 Synthesis of Polycyclic Polysubstituted Natural Products Me0 I HO 4 (iv) The typical ring B substituent of delphinine-type alkaloids is missing in talatisamine; this fact facilitated considerably the planning of the synthesis. All these serious flaws might be removed (and, in fact, were removed) by introducing the ring B bridge at the beginning rather than at the end of the synthesis. Thus, we decided to carry out the total synthesis of chasmaninel5 (28), an alkaloid in which only one of the seven delphinine substituents is missing, via the main stages (23), (25), and (27).Later on as a result of model work and practical synthetic experience we modified the plan to the sequence (24), (26), (27), and (28). The first problem to be solved was the synthesis of the ring B bridged ‘aromatic intermediate’ (24) with all its substituents appearing readily in the correct positions. In order to expIain how we arrived at a synthetic strategy, which satisfied aZmost all of our original principles, it is necessary to digress and to mention l5 S. W. Pelletier, Z. Djarmati, and S. LajgiC, J. Amer. Chem. SOC.,1974, 96, 7817. 418 Wiesner Me0 H Me0 I \OMe (23) R = COMe (25) R = OTs (26) R = Rr (24) R = C//O n briefly our synthesis of napelline (32).16 The key intermediate in the construction of (32) was the aromatic compound (31) obtained by elaboration of the tricyclic sulphonamide (30).This last product was prepared by a stereo- and regio-specific acetolysis of the aziridine (29). The mechanism of this process is portrayed by the arrows in formula (29). The regiospecificity of the rearrangement (29) -+ (30) is caused by the aromatic methoxy group which enhances the migratory aptitude of the substituent in the para position to such an extent that the alternative rearrangement initiated by the opening of the other aziridine C-N bond cannot successfully compete. Me0 Ho---H 0' IH Meo 'OMe SO, (28) I Ph (29) MeG l6 K.Wiesner and A. Philipp, Tetrahedron Letters, 1966, 1467; I(. Wiesner, P. T. Ho. R. C, Jain, S. F. Lee, S. Oida, and A. Philipp, Canad. J. Chem., 1973, 51, 1448; K. Wiesner. P. T. Ho, D. Chang, Y. K. Lam, C. S. J. Tsai, and W. Y. Ren, Canad. J. Chem., 1973, 51, 3978; K. Wiesner, P. T. Ho, C. S. J. Tsai, Canad. J. Chem., 1974, 52, 2353; K. Wiesner, P. T. Ho, C. S. J. Tsai, and Y. K. Lam, Canad. J. Chem., 1974, 52,2355. Synthesis of Polycyclic Polysubstituted Natural Products If we compare the structures of the chasmanine intermediate (24) and the napelline intermediate (31), we see at once that the technique used in the con- struction of (31) is capable of providing the entire array of aliphatic substituents in (24), since the primary methoxy group of (24) can be introduced into the starting material.Thus, the only difficulty to be overcome was the location of the aromatic methoxy group in (24)and its anticipated unfavourable influence on the aziridine rearrangement in the step corresponding to (29) (30).Although it includes novel features I shall not describe the conversion of (31) into napelline (32), as it is not relevant to the theme which I wish to develop in the present article and is also somewhat lengthy. I shall mention a far better synthesis of the same alkaloid now in progress in a brief discussion of the fourth generation approaches. The starting material for the aromatic intermediate (24)17 was the methoxy- indanone (33). The enol ether (34) prepared from it may be readily carboxylated in the presence of n-butyl-lithium and thus the dimethoxyindene ester (35) was obtained.Compound (35) is in a thermal equilibrium with the quinonoid tauto- mer (36) and as a consequence yields practically quantitatively a maleic anhydride adduct. The adduct was decarboxylated by the method of Trostls and the tricyclic intermediate (37) was obtained in a yield of over 80%. The ester (37) was then modified to the corresponding aldehyde (38) and the stage was set for the introduction of the nitrogen and the annelation of ring A. The following three steps converted the aldehyde (38) into the unstable aziridine (39) which re- arranged immediately to the diketone (40). (i) Treatment with the Grignard reagent prepared from 1-bromo-3-benzyloxy-4-methoxybutane, (ii) oxidation Me0 SO2 -Ph ?vleO O-CH,-Ph (37) R = COOMe OMe (38) R = -C /p ‘H l7 S.F. Lee, G. M. Sathe, W. W. Sy, P. T. Ho, and K. Wiesner, Canad. J. Chem., 1976, 54, 1039. la B. M. Trost and F. Chen, Tetrahedron Letters, 1971, 2603. Wiesner with chromic acid, and (iii) treatment with a large excess of benzenesulphonyl- azide in acetic acid. The function of the methoxy group attached to one bridgehead of the system is to accelerate the desired rearrangement [as shown by the arrows in formula (39)] while the keto group at the other bridgehead slows down the competing reaclion. This competing rearrangement, initiated by the opening of the other aziridine carbon nitrogen bond, requires the development of a partial positive charge on the carbon directly bonded to the keto group and consequently the energy of its transition state is increased.The combined action of the two bridgehead groups succeeded to overcome the unfavourable accelerating influence of the aromatic methoxy group and to steer the rearrangement with moderate regioselectivity (60:40) in favour of the diketone (40).The absence of regiospecificity in this step is the only serious flaw of the synthesis and it can be corrected by a temporary replacement of the aromatic methoxy group by the electron withdrawing mesyloxy group (cf.ref. 16). The several steps required for this operation however make it profitless from the point of view of yield.The final generation synthesis does not suffer from this defect since it requires a substitution of the aromatic ring favourable to the desired rearrangement (vide infra). The diketone (40)was now modified in a few simple steps to the @-unsaturated ketone (41). Photochemical addition of vinylacetate to this compound yielded, with complete regio- and stereo-specificity, the adduct (42), the stereochemistry of which was again predictable by our addition rule.13 Saponification of the acetoxy group in (42) resulted in a retroaldol reaction and gave the homoaldehyde (43) which was degraded in very high yield to the ester (45) via the enolether (44). All these derivatives including the ester (45) turned out to prefer the A/B cis configuration in which the ester group is incapable of forming a lactame.However, reflux with absolute niethanolic alkali caused, predictably, a gradual epimerization of the ring junction with the simultaneous formation of the lactame and loss of both acetyl groups in a high yield. Finally, chromic acid oxidation of the product completed the synthesis of the diketolactame (46). Compound (46) was now stereospecifically reduced with tri-butoxyaluminium hydride and the resulting diol (47) was methylated to the dimethoxy-N-methyl lactame (48). Reduction of (48) with lithium aluminium hydride and oxidation of the product (49) under carefully controlled conditions with permanganate gave a high yield of the aromatic intermediate (24). Before proceeding with the conversion of the ‘aromatic intermediate’ (24) into chasmanine (28) we performed several studies,lS which involved the synthesis of the tetracyclic compound (51) from the model starting material (50).I shall not discuss these studies in the present article, but I wish to emphasize the important role which they played in our ultimate success. While not every method worked out on the model system can be applied without modification in the synthesis proper, it is clear that the second part of the chasmanine synthesis would have lDK.Wiesner, P. T. Ho, W. C. Liu, and M. N. Shanbhag, Canad.J. Chem., 1975,53,2140; K. Wiesner, I. H. Sanchez, K. S. Atwal, and S. F. Lee, Canad.J. Chem., 1977, 55, 1091. Synthesis of Polycyclic Polysubstituted Natural Products OMe 5 AcO 0 (44 II (43) R1= H; R' = CH,-C-H (44) R' = Ac; R' = CH=CHOMe (45) R1= Ac; R2 = C0,Me OMe (46) (47) R = H (49)(48) R = Me been almost impossible to accomplish without the preliminary experience on models.An additional advantage, which we derived from our preliminary studies, was the n.m.r. spectra of our model intermediates. These served us as a reliable and precise simulation of the corresponding n.m.r. patterns in the synthesis proper and helped us to recognize the various synthetic intermediates without a shadow of doubt. The conversion of the 'aromatic intermediate' to chasmanine was accomplished as follows. Reduction of compound (24) with lithium in liquid ammonia followed by acetylation and acid treatment gave the @-unsaturated ketone (52).In this product the new chiral centre was created as expectedlg by exoprotonation with respect to the bicycloheptane system and is consequently epimeric to the corres- ponding chiral centre in chasmanine. Compound (52) is somewhat more stable than the desired epimer and equilibration requires very drastic conditions which cause extensive destruction of the material. Consequently, the third variantlg of our model synthesis (50) + (51) was successfully utilized. Photochemical addition of allene to (52) gave a high yield of the single product (53), the configuration of which was predictable by our addition rulel3 and in agreement with the corresponding model compound whose structure had been corroborated by X-ray crystallographic studies.19 The ethylene glycol acetal(54) prepared from the photo adduct was now converted into the acetoxy ketone (55) in an overall yield of 72% by the following three steps.(i) Ozonolysis followed by borohydride reduction of the ozonide. (ii) Acetylation. (iii) Removal of the acetal group by mild acid treatment. Wiesner Me0 (53) R = =O (54) R = -O1-0 Bromination of (55) gave the monobromide (56) and dehydrobromination of this compound yielded 87% of the unsaturated ketone (57). The way was now prepared for the skeletal transposition to the nordenudatine system with the simultaneous disappearance of the ‘offending’ enantiomeric chiral centre. Treatment of compound (57) with mild alkali caused saponification of the acetoxy group followed by immediate reverse aldol reaction and conjugate aldol con- densation and gave a 90% yield of the epimeric aldols (58).Acetylation yielded the acetates (59) and this material was transformed to the single crystalline keto- acetal (60) in high yield by a stereospecific a-hydrogenation with rhodium on alumina followed by a simple modification of the functional group system. Compound (60) was now simply and stereospecifically converted into our long OR AcOoho Me0 ; H H ‘OMe (55) R = H (57) (58) R = H (56) R = Br (59) R = AC RO H ‘OMe (61) R = H (62) R = Me (63) R = H (64)R = Br 423 Synthesis of Polycyclic Polysubstituted Natural Products envisaged intermediate (26).Reduction with borohydride yielded the alcohol (61) which was methylated to (62). Deacetalization of this material gave the ketone (63) and this compound was brominated to the monobromide (64). Finally, the bromoketone (64) was converted into the intermediate (26) using the acetalization method of Professor Barton20 with diethylene orthocarbonate. I might mention that no other acetalization method seemed to work in this case and if Barton's paper had not appeared in time, the synthesis might well be still unfinished. As mentioned above, we were quite certain about the structures of all these intermediates since the corresponding model derivatives obtained in the same way gave n.m.r. spectra which were qualitatively and quantitatively identical in the relevant parts.As in the model series, however, the configuration of the bromine in (64) followed from the success of the subsequent rearrangement which requires antiplanarity rather than from an interpretation of the n.m.r. spectra. The rearrangement of the bromoacetal (26) to the oxopyrochasmanine deri- vative (27) was performed in the presence of a strong base in a mixture of xylene and DMSO. It proceeded, as expected, very cleanly in a yield of 85% and the racemate (27) was identical in its spectral and chromatographic properties with the corresponding optically active compound prepared from chasmanine. The synthesis of racemic chasmanine was completed with the synthetic racemate (27) without the use of a relay, but each remaining intermediate was again compared and found identical with the corresponding naturally derived chasmanine deri- vative. The remaining steps were simple: oxymercuration of the double bond, deacetalization, and a stereospecific reduction of the liberated keto group by lithium aluminium hydride, completed a stereospecific total synthesis of chasmanine (28).21 5 The Fourth Generation: Synthesis of the Denudatine System by Diene Addition The synthesis of chasmanine has clearly demonstrated the superiority of the nordenudatine route (26) -+ (27) + (28) over the noratisine route, which we used three years ago for the synthesis of talatisamine.While very little can be improved in this sequence, fundamentally better methods could be developed for the synthesis of the nordenudatine intermediate (26). I started thinking about such possibilities several years ago in connection with the synthesis of the relatively simple alkaloid denudatine (67), the structure of which we had clarified at the Ayerst Laboratories.22 A simple and obvious way to construct the system of (67) [or (26)] would be an addition between a diene of the type (65) and a suitable dienophile.However, one would first have to know whether such an addition would be stereospecific and if so, whether the dienophile would add to the p-or a-face of the diene (65). The well-known preferential exoreactivity of the bicyclo [2,2,l]heptene system in ionic reactions makes one favour the first ao D. H. R. Barton, C. C. Dawes, and P. D. Magnus, J.C.S.Chem. Comm., 1975,432. 'IT. Y. R. Tsai, C. S. J. Tsai, W. W. Sy, M. N. Shanbhag, W. C. Liu, S. F. Lee, and K. Wiesner, Heterocycles, (Woodward Issue), submitted for publication. aa M. Gotz and K. Wiesner, Tetrahedron Letters, 1969, 4369. Wiesner possibility, but it is necessary to find out with certainty before starting a full scale total synthesis. Thus, we first synthesized compound (65) using our aziridine rearrangement method and then studied its addition reaction with maleic anh~dride.~3The result was the stereospecific formation of the p-adduct (66) in high yield. 0 H H This type of adduct can clearly never serve as an intermediate in the con- struction of (67) since it lacks substitution in the a-branch of the bicyclo [2,2,2]-octane system and there is no way in which substituents can be introduced into the required positions.Consequently, in order to obtain adducts capable of further development, it is necessary to introduce substituents into the six- membered ring of (65) which contains the diene system. The compound which immediately comes to mind and which fills all the requirements is an o-quinone with the two carbonyl groups located at the asterisked carbons of (65). Because of the notoriously unpleasant properties of o-quinones, I decided to try a masked o-quinone system which would be more stable and at the same time allow a regiospecific addition of unsymmetrical dienophiles. As the masking functionality I selected the spirolactone group which was previously used success- fully by Deslongchamps24 in his approach to ryanodine.Before proceeding with actual total syntheses and investing much labour, time, and money, we decided to do one more preliminary study and to synthesize rapidly and efficiently compound (74), the c, D ring system model for denudatine (67). The starting material (68)25was oxidized with N-bromosuccinimide and the resulting spirolactone (69) was immediately allowed to react with ethylvinyl sulphide. The two epimeric adducts (70) were obtained in a yield of 85% based on compound (68). Mild hydrolysis of (70) with methanolic potassium carbonate gave the single highly crystalline diketone (71) in a 92% yield. Compound (71) reacts selectively with Grignard reagents, thus, treatment of this material with a large excess of trimethylsilylmethyl magnesium chloride gave exclusively the product (72) in high yield. Desulphurization of (72) with 2J K.Wiesner, P. T. Ho, and S. Oida, Canad. J. Chem., 1974, 52, 1042. D. Berney and P. Deslongchamps, Canad. J. Chem., 1969, 47, 515. 25 K. P. Nambiar, Ph.D. Thesis, University of New Brunswick, 1977. Synthesis of Polycyclic Polysubstituted Natural Products 0+-jjM.0Ho%Me0 I CHZ 0 COOH (72) (73) (74) Raney nickel followed by an acid catalysed elimination of the trimethylsilyl group yielded 90% of the dienone (73). Finally, protection of the ap-unsaturated ketone system, a stereospecific hydroboration, and deprotection gave hydroxyketone (74), the structure of which was corroborated by X-ray crystallographic studies of the p-bromobenzoyl derivative.26 If we now apply the above sequence to the aromatic intermediate (73, there is every reason to believe that a simple synthesis of denudatine (67) will result.Because of our limited resources, this problem is temporarily at a standstill. I shall now turn to the description of the work already accomplished in the direction of a new, more efficient, shorter, and completely regio- and stereo- specific synthesis of chasmanine. In view of what I said in the introduction about the main purpose of our work, this is closer to my heart than denudatine and for this reason denudatine has to wait. The aromatic intermediate (76) was oxidized to a spirolactone as in the model system (68) + (69) and addition of benzylvinyl ether gave the adduct (83) in high yield.The analogy of compound (83) and of the ‘old’ chasmanine intermediate (59) is striking, as is the difference in the number of steps which the two similar products required to be synthesized. I shall first comment briefly on the synthesis of (76) and then outline the further development of the adduct (83) to chasmanine. The compound (76) which was transformed to the nordenudatine derivative (83) in virtually one step was obtained in high yield (80-85%) from the standard ‘aromatic intermediate’ (77) as follows. The aromatic methoxy group was z8 K. Wiesner, T. Y. R. Tsai, G. I. Dmitrienko. and K. P. Nambiar, Canad. J. Chem., 1976, 54, 3307.Wiesner (77) R = Me (75) R = H (78) R = H (76) R = OMe (79) R = CH,COOMe COOMeCOOMeMe0 ' Me0 P b --H N Acij Me0 ; H I AcO H 'OMe selectively cleaved by reflux with sodium thioethoxide in DMF to the phenol (78). Alkylation of (78) with methyl bromoacetate yielded (79). Under carefully controlled conditions the methyl group in (79) was oxidized with chromic acid to an aromatic aldehyde in a 90% yield. Finally, perbenzoic acid oxidation of the aromatic aldehyde followed by basic hydrolysis gave 85% of the phenolic acid (76).The preparation of compound (77)27 is much more efficient than the synthesis of the 'old' chasmanine intermediate (24). The reason for this is the positions of the aromatic substituents which are favourable to a completely regiospecific aziridine rearrangement.Furthermore, several improvements in individual steps have contributed to a large increase in the overall yield and a significant decrease in the time and labour involved. The most significant of the innovations2* was the use of trimethylsilylazide29 for the direct transformation of the tricyclic ester (80) to the acetyl aziridine (81). This product rearranged stereo- and regio-specifically on acetolysis to the synthon (82). The rearrangement portrayed by the arrows in formula (81) is accelerated by the methoxy group situated orrho to the migrating benzene bond, while the potentially competing rearrangement is slowed down by the ester carbonyl. Thus, complete regiospecificity results. Compound (82) was converted *' Unpublished work by T.Y. R. Tsai, A. Feicht, R. Marini-Bettolo, and D. Krikoryan. 28 Unpublished work by T. Y.R. Tsai. *9 L. Birkofer, A. Ritter, and P.Richter, Chem Ber., 1963, 96, 2750. Synthesis of’Polycyclic Polysubstituted Natural Products into the ‘aromatic intermediate’ (77) by our standard methods. The development of the ‘nordenudatine intermediate’ (83) to chasmanine is foreshadowed only by model work.30 The model-analogue of (83), compound (84), was converted into the tosylhydrazone (85). Borohydride reduction of (85) followed by acidic hydrolysis yielded the epimeric ketols (86) which were con- verted into the corresponding mesylates (87). Reduction of (87) with calcium in ammonia followed by acetylation yielded the intermediate (88) which is the same as in the photochemical approach.It was transformed to the bromoketal (89) and hence to the chasmanine model (51) as in the photochemical synthesis of chasmanine. 0‘ (84) (85) H (86) R = H (87) R = MS The flaw that a substituent present in (84) has been removed and sub- sequently reintroduced was made necessary by our inability to control stereospecifically the configuration of the ketol (86). Moreover, in the nordenu- datine system bromine is required as a leaving group and we were unable to introduce the p-bromo-substituent by displacement. However, the flaw is only aesthetic, the yields of all steps were almost quantitative and introduction of the bromine by displacement would decrease the number of operations by only one.Besides the increased efficiency in the synthesis of the aromatic intermediate (77) the construction of the C,D ring system in the potential fourth generation chasmanine synthesis has been shortened by seven steps and I believe that the entire process is now close to our initial objective. 30 Unpublished work by K. S. Atwal and 1. Sanchez. Wiesner There is another advantage inherent in our fourth generation methods as compared with the already completed photochemical synthesis of chasmanine. It seems that it would be possible to place a substituent on the aromatic ring of the ‘aromatic intermediate’ which would end up on the bridgehead of the C,D ring system and be ultimately converted to a hydroxy group.Thus, delphinine (1) itself might be reached with about the same effort as chasmanine. To illustrate the versatility of our fourth generation methods I wish to mention in conclusion a second very efficient synthesis of napelline (32) which is now in progress and is presently at the stage of compound (90).31 The construction of the C,D ring system as foreshadowed by a model study32 proceeded as follows: the spirolactone (84) was converted by unexceptional methods to the derivative (91) with a tosyloxy group suitably disposed for rearrangement, acetolysis of (91) gave a mixture of the epimeric acetates (92) which by saponification and oxi- dation yielded the diketone (93), which was identical to the corresponding intermediate in our ‘old’ napelline model work33 and is an anaIogue of the intermediate (94) which we converted some time ago into napelline.16 0 0 Me Q Me% NH 0H!Me In conclusion, I wish to thank my younger colleagues who struggled with devotion, courage, and ingenuity to make this synthetic development possible.Their names are recorded in the references and it is perhaps interesting to point 31 Unpublished work by W. W. Sy and C S. J. Tsai. 32 Unpublished work by R. Marini-Bettolo. 33 Unpublished work by C. S. J. Tsai. 429 Synthesis of Polycyclic Polysubstituted Natural Products out how few there were and how individual names are repeated in several of the references quoted. This testifies to the hard work these young men and women were doing, but perhaps also to the efficiency of our methods.I also wish to thank my secretary, Miss Judy Briggs, for the typing of the manuscripts and general help in my English composition not only in this article but in all the references quoted. FinaIly, it is a pleasure to thank the National Research Council, Ottawa, and the Hoffman-La Roche Company, Nutley, and Vaudreuil for supporting our studies in synthetic strategy over many years, and the Merck, Sharp and Dohme Company, Montreal, for a grant in the current year.

ISSN:0306-0012    

DOI:10.1039/CS9770600413

出版商:RSC    

年代:1977

数据来源: RSC

摘要:

Properties and Syntheses of Sweetening Agents By B. Crammer and R. Ikan DEPARTMENT OF ORGANIC CHEMISTRY, NATURAL PRODUCTS LABORATORY, HEBREW UNIVERSITY, JERUSALEM, ISRAEL 1 Introduction Research for inexpensive non-toxic non-nutritive sweetening agents has been intensified since the withdrawal of cyclamate from the consumer’s market in 1970. There is doubt whether the other commercial sweetener, saccharin, will continue to be marketed because of toxicological problems that have been encountered. Further, the increase in cost of sugar has also prompted a search for novel cheap sweeteners. The number of new sweeteners that have emerged for public consumption over the past decade has been severely restricted by the stringent requirements of the American Food and Drugs Administration (F.D.A.) and the British Food Standard Committee. A major difficulty in the search for the ideal sweetener is that no satisfactory theory or mechanism has been elucidated to explain the relation of molecular structure to sweetness.It is known, however, that sensory responses to sweet substances are related to chemical specificity. A very slight modification in the molecule’s structure can result in a tasteless or even bitter product. For example saccharin (1) is very sweet but the N-methyl derivative (2) is tasteless.’ Similarly, sodium cyclamate (3) is sweet, but the N-methyl derivative2 (4) is tasteless. Dulcin,p-ethoxyphenylurea(3, is 250 times sweeter than sucrose, the ortho derivative (6) is tasteless, whereas the thiourea analogue (7) is bitter.A. F. Holleman, Rec. Trav. chim., 1923, 42, 839. a R.J. Wicker, Chem. and Ind., 1966, 1708. 431 Properties and Syntheses of Sweetening Agents OC,H, (Sweet) (Tasteless) (Bitter) (5) (6) (7) Stereoisomers and anomers may have different tastes. For example L-glucose has a slightly salty taste whereas D-glucose is sweet. Similarly, the two anomers of D-mannose have different tastes, a-D-mannose is sweet whereas /3-D-mannose is bitter (some sugars, for example cellobiose, are tasteless). Many of the D-isomers of amino acids are sweet, while the L-isomers generally are tasteless (L-isoleucine is bitter and D-isoleucine is sweet). These differences in taste are due to a change in the conformation of the molecule and indicate that discrimination between tastes is in part a recognition of the spatial structure of molecules.So far none of the more promising sweeteners which have been tested in recent years have satisfied all the requirements for an ideal sweetener, hence the search for such a sweetener continues. This review intends to survey the most recent trends in the search for sweetening agents. There have been other review^^-^ on sweeteners but they are not concerned with complete examination of all the classes of sweeteners now known. 2 Gustation Taste or gustation is comprised of four basic taste sensations; sourness, saltiness, sweetness, and bitterness. Sourness is related to the concentration of hydrogen ions in the media.Saltiness is caused by the presence of Group I cations Lif, Na+, and K+ (and only in a minor way to halogen anions such as the chlorine anion). Sweet and bitter sensations are more complex and their mechanism is not under- stood. There is no satisfactory explanation why, for example, in the non-nutritive sweeteners cyclamate (3) and saccharin (1) there should be a marked decrease in their sweetening potency as their concentration in aqueous solutions increases. Surprisingly, at high concentrations these sweeteners become bitter.6 Magidson and Gorbatschow7 showed, over 50 years ago, the comparative decrease in sweet- ness of saccharin with respect to sugar (Table 1). Later Sjostrom and Cairncross8 observed that the addition of 0.5% saline solution to a 5% sucrose solution increased the sweetening potency.Recently it has been reported by Seyboldg that even salt can taste sweet at a sufficiently low concentration. Bitterness is often found in such natural products as alkaloids, bile salts, and glycosides such W. Herzog, Fortschr. Chem. Ztg., 1929, 53, 99..I E. Newbrun, Iniernat. Dental J., 1973, 23, 328. G. A. Crosby, CRC Rev. Food Sci. Nurririon, 1976, 7, 297. M. Brook, Royal SOC.Healrh J., 1969, 89, 140. 0. J. Magidson and S. W. Gorbatschow, Ber., 1923, 56B,1810. * L. B. Sjostrom and S. E. Cairncross, Adv. Chem. Ser., 1955, 12, 108. P. G. Seybold, Chemistry, 1975, 48, 6. 432 Crammer and Ikan Table 1 Concentration of Sucrose Sweetness % Concentration of sucrose solutionlconcentration of saccharin solution as sodium salt 15 190 10.2 261 6.2 317 3.9 399 2.7 553 1.65 675 as the saponins.It will be seen in this review that only minor differences distinguish bitterness from sweetness and that in many cases a reversible relationship exists between them. 3Natural Sweeteners The earliest natural sweetener known to man was honey. Evidence from rock paintings indicates that it was used in the Neolithic period. The manna mentioned in the Bible was possibly the secretion of the insects Trabutina mann@ara and Naiacoccus serpentinus which feed on desert Tamarisk trees. Analysis shows that this secretion contains sucrose and invert sugar as the principal ingredients.lo Table 2 shows the sweetness of a variety of sugars based on sucrose as the standard.*J1-13 The discrepancies in the sweetness factor are due to differences in: (a) concentration of the sweetener, (b) temperature, (c) pH of the sugar solution, (d) nature of the medium, (e) sensitivity of the human taster.The world sugar production has increased from 8 million metric tons in 1900 to 73 million metric tons in 1970. Sugar is by far the cheapest natural sweetener currently available and it will not be easy to replace it as an important food ingredient. It should be appreciated that sugar is not the panacea to man’s sweet taste as it can have a detrimental effect on health (especially during pregnancy) and can cause (a) obesity, (b) dental diseases, particularly dental carries in man and (c) coronary diseases.14J5 It is because of these serious side effects of sugar that the search for non-caloric sweetening agents has been dramatically intensified during the past decade.4 Glycyrrhizin Glycyrrhizin, the potassium or calcium salt of glycyrrhizic acid (8) (3p,2Op)-20-carboxy-1 1 -0xo-30-noro~ean-12-en-3-yl 2-O-p-~-glucopyranosyl a-D-giuco- l°K. S. Brown, Chem. SOC.Revs., 1975, 4, 277. l1 H. M rk, J. Mcketta, and D. Othmer Encyclopedia of Chem. Technol., 1969, 19, 594, John Wiley. 12 L. W. Aurand and A. E. Woods Food Chemistry, Carbohydrates, 1973, 101. 13 A. Biester, M. W. Wood, and C. S. Wahlin, Amer. J. Physiol., 1925, 73, 387. l4 0. Paul, A. Macmillan, and H. Park. Lancet, 1968, 2, 1049.W. E. Waters, S. Moore, and P. Sweltnam, Lancet, 1970, 1014. 433 Properties and Syntheses of Sweetening Agents Table 2 Sugar Sweetness based on sucrose 100 Raffinose 2313 a-Lact ose P-Lact ose 3912 4011 a-Mannose a-Galactose Maltose Rhamnose D-Xylose 7011 Xy1i to1 Mannitol 7011 Sorbitol (D-glucitol) 544 Dulcitol a-Glucose 744J3 /?-Glucose Sucrose Invert sugar (Glucose + Fructose) 70-9011 1304913 D-Fructose 11412 P-Fruct ose 1804 HO,C Glycyrrhizic acid (8) pyranosidouronic acid has been known for at least 4000 years. It is found in certain rhizomes and licorice roots such as Glycyrrhiza glabra L., G. hersuta and G. Uralensis. Glycyrrhizin is about 50 times sweeter than sucrose.The rhizomes contain from 3 to 23 % of glycyrrhizic acid salts.16 The ammonium salt of gly-cyrrhizic acid is also 50 times sweeter than sucrose and when it is mixed with lG A. Muravev, Farm. Polska, 1967 798. Crammer and Ikan sugar a potentiating effect occurs which doubles its sweetening potency. Ammon-ium glycyrrhizate is not suitable as a marketable sweetener because of its lingering licorice flavour, however, this lingering after taste is repressed by the presence of specific nucleotides such as 5’ inosinic acid or 5’ guanylic acid without any loss of sweetening power of the ammonium glycyrrhizate.17 5 Stevioside Stevioside, [1~-(2-~-~-~-g~ucop~anosy~-~-~-~ucop~anosy~)oxy]ka~-~6-en-18-oic p-D-glucopyranosyl ester (9) was first isolated by two French chemists HO H OH H OH Stevioside (9) Bride1 and Laviellel8 in 1931 from the leaves of the shrub Steviu rebaudiana Bertooi (also known as Yerba duke) which grows wild in Paraguay. Stevioside is 230to 300times sweeter than sucrose.4~5 MacAndrews and Forbes Co., Israel P.42,86211975.M.Bridal and R.Lavielles, J. Pharm. Chim., 1931, 14, 99. 3 435 Properties and Syntheses of Sweetening Agents 6 Sweet Proteins A. Monel1in.-Monellin is an intensely sweet protein found in the red berries (known as ‘serendipity berries’ or ‘guinea potatoes’l9) of the tropical West African plant, Dioscoreophyllum comminsii.19 It is a single polypeptide which, on a weight basis, is approximately 3000 times sweeter than sucrose.It is reputed to be the sweetest natural product known and is also the first protein known to exhibit the remarkable property of potent sweetness. It has an unusual sweetness which develops in a few seconds and lingers for a few minutes. Unfortunately its sweetening potency disappears within 24 hours at room temperature and it is therefore not suitable for use in foods and beverages. The complete structure of monellin has recently been elucidated.20-22 Subunit A Arg-Glu-Ile-Lys-Gly -Tyr -Glu-Tyr-Gln- Leu-Tyr-Val-Tyr-Ala-Ser-Asp-Lys-Leu-Phe-Arg -Ala -Asn -1le -Ser -Gln -Asn -Tyr-Lys -Thr -Arg -Gly -Arg -Lys -Leu -Leu -Arg -Phe-Asx-Gly-Pro-Val-Pro-Pro-Pro. Subunit B Gly-Glu-Trp-Glu-Ile-Ile-Asp-Ile-GIy-Pro-Phe-Thr-Gln-Asn-Leu-Gly-Lys-Phe-Ala-Val-Asp-Glu-Glu-Asn-Lys-Ile-Gly-Gln-Tyr-Gly-Arg-Leu-T~-Phe-Asn-Lys-Val-IleArg-Pro-Cys-Met-Lys-Lys-Thr-Ile-T~-Glu-Glu-Asn. Complete structure of monellin B.Thaumath-Thaumatin, another sweet-tasting protein has been isolated by van der WeP from the fruit of Thaumato coccus Danielli Benth, a plant found in West Africa. Thaumatin is a basic protein with a molecular weight of about 14,000 and an iso-electric point of 12.0.It is 750 times as sweet as sucrose with a slight licorice after taste. It is also a single polypeptide chain, but about twice as long as the monellin molecule. Heat denaturation and the splitting of the disulphide bridges of thaumatin both result in the complete disappearance of sweetness implying the importance of the tertiary structure of the protein for its taste.24 C.Miraculh-A native shrub of Tropical West Africa yields a small red berry which, when chewed, causes sour substances to taste sweet.19 This berry first came to European attention in 1852when an English surgeon, Daniell, noted that the commander of a British fort in Dahomey enjoyed ‘constant opportunities of testing the wonderful effects of this fruit’.The so-called ‘miracle berry’ was 1) G. E. Inglett and J. F. May, Econ. Bot., 1968, 22, 326. lo G. Frank and H. Zuber, Z. Physiol. Chem., 1976,357,585. l1Z. Bohak and Shoei-Lung Li, Biochim. Biophys. Acta, 1976,427, 153. la G. Hudson and K. Biemann, Biochem. Biophys. Res. Comm., 1976,71,212. H. van der Wel, FEBS Letters, 1972, 21, 88.l4 H. van der We1 and K. Loeve, European J. Biochem., 1972, 31,221. Crammer and Ikan rediscovered in the 1920’s by a U.S. Department of Agriculture collecting expedition. In 1968 Beidler and K~rihara~~ and independently van der We1 and coworkers26 showed that the active constituent of miracle fruit, Synsepalum dulcifcum, is a basic glycoprotein with an approximate molecular weight of 44,OOO. The modifying effect usually lasts from 1 to 2 hours. The sweetness of miraculin is gradually lost on being heated, which again implies a relationship between tertiary structure and taste. The approximate amino acid composition of miraculin is known but not its amino acid sequence. 7 Chlorogenic Acid and Cynarin The taste-modifying property of artichoke (Cynarascolymus)has been known for some Tests have shown that solutions of different taste qualities (sucrose, citric acid, quinine hydrochloride and sodium chloride) are all sweetened to some degree by aprior mouth rinse with artichoke extract.The artichoke induced sweet- ness, however, lasts only for about 4 to 5 minutes. Chlorogenic acid (3-caffeoyl quinic acid) (10) and cynarin (1,5-dicaffeoyIquinic acid) (1 1) have been OH / H OOCCH=CH Ho2@HHO H OH Chlorogenic acid (10) OHI IH O H b-CH=CHCOO shown to be the two principal active compounds responsible for the taste- modifying property of artichoke. Cynarin also has a sweet taste. Artichokes L. Beidler and K. Kurihara, Science, 1968, 161, 1241. H. van der Wel, J.N. Brouwer, A. Francke, and G. J. Henning, Nature, 1968, 220, 373. 97 L, M. Bzirtwibuk, C. H. Lge, and R, Scarpellino, Science, 1972, 178, 988. Properties and Syntheses of Sweetening Agents are on the G.R.A.S. (Generally Regarded As Safe) list of the F.D.A. and hence these two compounds may show promise as sugar substitutes. 8 Perillarthe l-Perillaldehyde (12) the essential oil of the plant Perilla nankinensis is 12 times sweeter than sucrose. On the other hand the syn-oxime of 1-perillaldehyde known as perillartine (13) is 2000 times sweeter than sucrose.lL The isomer, originally believed to be the &znti-oxime (14) was in fact the tert-chloride formed by the OH I OH I CHO H \/N H Lc&N Markovnikov addition of HCI to the isopropenyl moiety of (12). Numerous analogues of perillartine have been synthesized by Acton and coworkers28-30 and were found to be sweet.The most promising was 4-(methoxymethy1)-cyclohexa-1,4-diene-l-carboxaldehyde,syn-oxime (15) known as SRI oxime V. It is 450 times sweeter than sucrose and has no undesirable after taste of saccharin.31 If kept above pH 3 this sweetener is stable in most foods and sweet concentrates. SRI oxime-V was ~ynthesized3~ according to Scheme 1. 9 Osladin Osladin (16), a steroidal saponin, has recently been isolated from the rhizomes of Polypodium vulgare, the first sweet substance isolated from the steroid It is approximately 3000 times sweeter than sucrose. Osladin has recently been partially synthesized from solasodine to osladin agIy~one.~* 10 Phyllodulcin Phyllodulcin (17) is a sweetener which was first isolated, in 1916, from the tea 28 E.M. Acton, K. Yamamoto, and H. Stone, U.S.P. 3,833,65411974. 29 E. M. Acton and H. Stone, Science, 1976, 193, 584. 30 E. M. Acton, M. W. Lerom, and H. Stone, U.S.P. 3,952,144/1976. 31 Chem. and Eng. News, 1975,53, No. 34,27. saE.M. Acton, M. W. Lerom, and H. Stone, U.S.P. 3,919,318/1975. 33 J. Jizba, L. Dolejs, V. Herout, and F. Sorm, Tetrahedron Letters, 1971, 1329. 34 M. Have1 and V. Cerny, Coll. Czech. Chem, Comm., 1975, 40,1579. 438 Crammer and Ikan NC OH 4 CH20CH3 CH,0CH3 CH20CH3 CH20CH3 1iv /OHCH=N (vi0 fj Jv (j CH20CH3 CH,OCH, CHBOCH, Reagents: i, Li-EtNH2; ii, oxidation-HNO, aq.; iii, KCN-H+; iv, -HaO; v, Bui,AcH; vi, NHaOH.Scheme1 0HO HO I HOty Osladin (16) leaves of Hydraugea macrophylla Seringe.35 It is 200 to 300 times sweeter than sucrose.36 Phyllodulcin may be synthesized by condensing 3-hydroxyphthalic 3b Y.Asabina and S. Ueno, J. Pharm. SOC.Japan, 1916, 146. 36 Y. Ariyoshi, Kagaku To Seibutsu, 1974, 12, 274. Properties and Syntheses of Sweetening Agents anhydride with homoisophthalic acid to 7,3’-dihydroxy-4’-methoxybenzal-phthalide which undergoes simultaneous reduction and lactone ring expansion in the presence of ethanolic sodium borohydride and aqueous alkalis7 Various Japanese workers have successfully synthesized phyllodu1~in~~-~~ (Scheme 2). OH OH 0 OH 0 (1 7) Reagents: i, KAc; ii, NaBH,-EtOH; iii, OH-Scheme 2 11 Sweetener from the Chineese Fruit ‘Lo Han Kuo’ Recently a natural sweetener was isolated from the dried fruit (Lo Han Kuo) of Momordica grosvenori Swingle which grows in Southern China.This sweetener of unknown composition is about 150 times sweeter than sucrose with a licorice after taste similar to the after taste of stevioside and the dihydrochalcones. Preliminary studies indicate the sweetener to be a glycoside of a triterpene.5 12 Chlorinated Sucrose A very sweet sucrose analogue was accidentally discovered by Hough and his coworkers of Queen Elizabeth College, London.41 The analogue, 1,4,6,6’- tetrachloro-1’,4,6,6’-tetradeoxygalactosucrose(1 8) was obtained by chlorination of sucrose (Scheme 3).It was 500-600 times sweeter than sucrose without the licorice after taste associated with stevioside and aspartame or the bitter after taste of saccharin. In the reaction the glucose ring of sucrose inverts to a galactose 37 Ternyo Tsuji, Sap. P. (Kokai) 74/110,667. say. Naoi, S. Higuchi, H. Ito, T. Nakano, K. Sakai, T. Matsui, S. Wagatsuma, A. Nishi, and S. Sano. Org. Repn. Proced. Int., 1975, 7, 129. 30T.Nakano H. Ito, Y.Naoi, S. Higuchi, Y.Takahashi, K. Sakai, T. Matsui, A. Nishi, and S. Sano, Jap. P. (Kokai) 75/35,167. 40 T. Nakano, M. Murase, K. Ochi, and S, Tabinaga, Chem. Comma,1976, 20, 820, dl Chern. Eng, ?vewso 1976, 30, Crammer and Ikan CH,OH HO + HO HO Scheme 3 ring. It is known that galactosucrose is tasteless and the reason is probably because of intramolecular hydrogen bonding on the 4-carbon hydroxy group.This bonding is not present in the conformational structure of sucrose or its chlorinated product. 13 The Dihydrochalcones It is known that citrus fruits contain bitter ingredients known as flavone glyco- sides. Naringin (19) is the bitter ingredient of grapefruit, neohesperidin (20) is the bitter ingredient of unripe Seville oranges and Prunin (21) is the bitter ingredient of Prunus wood. In 1963 Horowitz and Gentili accidentally found in the course of study on the structure-taste relations of citrus flavanones that catalytic reduction of naringin (19) and neohesperidin (20) gave dihydrochalcones (DHC) which were found to be surprisingly sweet42 (Scheme 4).This discovery led to numerous analogues of dihydrochalcones being synthesized and one of them displayed particular promise as ahighly potent non-nutritive sweetener, /3-neohesperidin dihydrochalcone (23) which was 1500 times sweeter than sucrose. The synthesis of dihydrochalcones from naringin has been reported43~44 (Scheme 4). R. M.Horowitz and B. Gentili, J. Agr. Food Chem., 1969,17,696. 43 L. Krbechek, G. Inglett, M. Holik, B. Dowling, R. Wagner, and R. Riter, J. Agric. Food Chem., 1968, 16, 108. 4g L. Givandan and Co., B.P. 1,443,310/1976. Properties and Syntheses of Sweetening Agents (19) Naringin R = P-neohesperidosyl* X = H Y = OH (20) Neohesperidin R = /3-neohesperidosyl X = OH Y = OCH, (21) Prunin R = P-D-glUCOSYl X = H Y = OH */?-neohesperidosyl = 2-O-cu-~-rhamnopyranosy~-~-g~ucopyranosy~ (19) J Naringin dihydrochalcone OH 0 (22) Reagent: i, H,(Pd-C) Scheme 4 0Jy OH ,CH2G0CH3 CHsOCO (23) Although the intense sweetness of neohesperidin dihydrochalcone (22a) permits the use of extremely low amounts to be used as a non-caloric additive, the lingering menthol after taste presents some problems.It has been found that a mixture of (22a) with saccharin gives a more acceptable sweet taste. The relative sweetness of some dihydrochalcone derivatives compared to sucrose are shown in Table 3.*3 Dihydrochalcones are stable in acids at normal temperatures and can be used in fruit and carbonated beverage products.45 46 G.E. Inglett, L. Krbechek, B. Dowling, and R. Wagner, J. Food Sci., 1969, 34,101. Crammer and Ikan Table 3 R = p-Neohesperidosyl group Compound Benzaldehyde substitute X Y Z Relative sweetness sucrose = 1 2,4,6,3-tetrahydroxy-dihydrochalcone 4’-/3-neohesperidoside H OH H 100 Naringin dihydrochalcone (22) H H OH 100 Neohesperidin dihydrochalcone (22a) H OH OMe 1500 Homoneohesperidin dihydrochalcone H OH OEt loo0 2,4,6,3-Tetrahydroxy-4-n-propoxy- dihydrochalcone 4’-/3-neohespidoside H OH OPrn 2000 Horowitz and Gentili42 replaced the neohesperidosyl group of the dihydro- chalcones of naringin and neohesperidin with other sugar moieties such as glucose, galactose, and xylose.46 These dihydrochalcone glycosides were found to be sweet having 5 to 10% the sweetening power of (22a) and were found to be less soluble.Esaki and coworkers47 further modified the sugar moieties of DHC and found that certain DHC glycosides such as p-sophorosyl DHC and /3-D-glucosyl- (1 -2)/3-~-galactosyl DHC were devoid of any sweetness. This contradicts the hypothesis of Horowitz and Gentili42 that in the DHC‘s of naringin and neo- hesperidin neither the C-2 or C-6 hydroxyl of the glucose molecule is essential for sweetness since the sweetness does not disappear when these groups are blocked with a 2-0-a-~-rhamnosyl group or the 6-0-methyl substituents. However, the C-3 and C-4 hydroxyl groups of glucose are necessary which suggests that the rhamnose moiety is essential for sweetness.The problem of lingering menthol after taste was overcome by a Hungarian company48 who synthesized DHC analogue (23) which substituted a sugar moiety by an acetyl moiety (Table 4). 14 Synthetic Non-Nutritive Sweeteners A. Sulphamic Acid Derivatives Sodium cyclohexyl subhamate (sodium cyclamate).-The sweetness of sodium cyclamate (3) was accidentally discovered by Michael Sveda of the University of Tllinois in 1937 while he was investigating the antipyretic properties of sulphamic 46 R. M. Horowitz and B. GentiIi, U.S.P. 3,890,298/1975. 47 S. Esaki, S. Kamya, and F. Konishi, J. Agric. Biol. Chem., 1975, 39, 1385. L. Farkas, M. Nogradi, A. Gottsegen, and S. Antus, G. P. (Offen.) 2,258,304/1973.Table 4 Relative sweetness of some dihydrochalcone analogues P 5 %$ 2 tl6’ 5’ 2 “LCompound-2 4 6 2’ 3’ 4’ Relative References 2 common name sweetness G sucrose = 1 c % Naringin DHC OH NeoH* OH OH 100 42, 43, 49, 50, 51 $Neohesperidin DHC OH NeoH OH OH OMe lo00 42, 43, 49, 50, 51 Neoeriocitrin DHC OH NeoH OH OH OH Slightly? sweet 42, 43, 49, 50, 51 3 $0OH NeoH OH OH 100 43 OH NeoH OH OH OEt lo00 43 b00OH NeoH OH OH OPP 2000 43 9 OH NeoH OH OH OPri lo00 42 2 OH (a) OH OH OMe Sweet? 54 OH NeoH OH OH OMe Sweet? 50 OH NeoH OH OMe Sweet? 55 OMe NeoH OH OH OMe Sweet? 51 OH (b) OH OH OMe 80 42, 50, 51 OH (c) OH OH OMe 160 51, 52 OH (4 OH OH OMe 140 300-50053 51, 53 OH 6-0-Me OH OH OMe lo00 51 NeoH OH (d OH OH OMe 1 R.M. Horowitz Personal communication 1973 OH OCHzCOzH OH OMe 180 48 OMe NeoH OH OH Sweet? 51 OH (f) OH OH loo0 51, 56 Hesperedin DHC OH OH OH OH OMe -i-R. M. Horowitz Personal communication 1973 OH OMe 180 48 OH OMe p 57 * NeoH = ,!3-neohesperidosyl (a) 4-0-a-glucosyloxy t Not available (b) ,!3-D-glUCOSJ’lOXy (C) @-D-XYlOSylOXJ’ (d) /I-D-galactosyloxy (e) 6-O-a-~-rhamnosyl-NeoH (fl 2-O-a-~-rhamnosyl-,!3-~-ga~actosyloxy 4g R. M. Horowitz and B. Gentili, U.S.P. 3,087,821/1963. 6o R. Horowitz and B. Gentili, in ‘Sweetness and Sweeteners,’ ed. G. G. Birch, L. F. Green, and C. B. Coulson, Applied Science Publishers, London, 1971. 61 R. Horowitz and B. Gentili, in ‘Sweeteners’ Symposium, ed.G. E. Inglett, Avi Publishing, Westport, Conn., U.S.A., 1974, chapter 16. 6p R. M. Horowitz and B. Gentili, U.S.P. 3,826,856/1974. 6s R. M. Horowitz and B. Gentili, U.S.P. 3,890,296/1975. 64 S. Okada, Kagaku To Seibutsu, 1974, 11, 712. s5 L. Farkas and M.Nogradi, Hung.P. 4026/1972. 66 D. M. Van Niekerk and B. H. Koespen, Experientia, 1972, 28, 123. 67 L. Farkas, M.Nogradi, T. Pfliegel, S. Antus, and A. Gottsegen. G.P. (Offen.) 2,506,356/1975. L. F. Audrieth and M. Sveda, J. Org. Chem., 1944,9, 89. 6g L. F. Audrieth and M. Sveda (E. I. Du Pont de Nemours & Co., Inc.), U.S.P. 2,275,125/1942. Properties and Syntheses of Sweetening Agents acid derivatives when he found that his cyclamate contaminated cigarette tasted sweet.The sodium salt of cyclohexyl sulphamate was found to be 30 times sweeter than sucrose. Aqueous solutions of sodium cyclamate are neutral whereas cyclamic acid, the free base of sodium cyclamate, has a sweet-sour taste and the pH of a 10% aqueous solution is between 0.8 and 1.6. The cyclamates are stable under most conditions encountered in food proces- sing. The sweetness of cyclamate salts, usually in the sodium or calcium form, relative to sucrose may vary from 25 to 140, depending on the type of food that is used. The potassium and magnesium salts were also found to be sweet but are not commercially used. The syntheses usually involve sulfonation of cyclohexylamine in the presence of a base. Sulfonation is often carried out in excess cyclohexylamine in order to isolate the cyclohexylammonium cyclamate. This double salt is readily converted to the desired sodium or calcium salt by treating it with either NaOH or Ca(OH)2, respectively.Audrieth and Sveda's original method used chlorosulfonic acid. This is a simple and nonexpensive method of synthesizing cyclamate^^^^^^ (Scheme 5). 0-[NH, + CISOBH -&-c).HS03H3N<>] (3) Reagents: i, CC14@ 5 "C;ii, NaOH Scheme 5 Catalytic reduction66 of the aromatic sulphamate was carried out by the Abbott Laboratories (Scheme 6). Various other processes for synthesizing (3) have been patented but none of them have any commercial prospects60-67 (Scheme 7). Ru or RhONHSQN~+ [HI -(3) Scheme 6 6o D. J. Lodger (E. I. Du Pont de Nemours & Co., Inc.), U.S.P.2,804,472/1957.H. S. McQuaid (E. I. Du Pont de Nemours & Co., Inc.), U.S.P. 2,804,477/1957.Abbott Laboratories (USA); Israel P. 18,631/1965. 63 P. Mueller and R. Trefzer (Ciba Corp.), U.S.P. 3,060,231/1962. 64 0. G. Birsten and J. Rosin (Baldwin-Montrose), U.S.P. 3,366,670/1968. 66 A. Calder and J. Whetstone (I.C.I. Ltd.), B.P. 1,395,497/1975.M. Freifelder and B. Meltsner (Abbott Laboratories), U.S.P. 3,194,833/1965. 67 W. W. Thompson (E. I. Du Pont de Nemours & Co. Inc.), U.S.P. 2,800,501/1957. Crammer and Ikan NHS0,Na0-NH2-Reagents: i,60 H,NSO,Na; ii,61 H2NS0,NH4; iii,62 S03-N,; iv,03 H2NS0,H; v,04 ClSO,H, EtsN; vi,64*66 M%NS03 Scheme 7 In 1970 cyclamate was thecheapest non-nutritive sweetener on the market.On August 14, 1970 the sale of cyclamate was forbidden on the American market. The decision came about because certain experiments on the bladder of rats showed that it was possibly carcinogenic. Other countries including England and Canada followed the American decision to ban cyclamate. Various groups retested the initial findings and were convinced that cyclamate was not car- cinogenic.68~69 In March 1976 the National Cancer Institute Committee presented their findings to the F.D.A. and concluded that 'the evidence to date does not establish the carcinogenicity of cyclamates'. Chemical structure and sweetness in certain sulphamates.-There are no specific rules or even explanations relating structure of certain sulphamate molecules to sweetness.Audrieth and Sveda5* investigated cyclamate and its derivatives in order to determine which part of the molecule causes sweetness. They first determined whether the cyclohexane ring was essential for sweetness by synthe- sizing the normal straight chained hexyl analogue, n-hexyl sulphamate, which they found to be tasteless. Later the Nitto Labs70 found a branched chain n-isoamyl sulphamate Me-CH-CH,CH,NHSO,NaI Me to be 10 times sweeter than cane sugar. Recently, Nofre71 discovered that the 68 F. Coulston, E. W. McChesney, and L. Goldberg, Food Cosmet. Toxicol., 1975, 13(2), 297. OY B. L. Oser, S. Carson, G. E. Cox, E. E. Vogin, and S. S. Sternberg, Toxicology, 1975,4(3), 315. 'O H. Yamaguchi, Nitto Laboratories, Jap.P.(Kokai) 881 5/1960. "C. Nofre and F. Pautet, Bull. SOC.chim.France,1975, 3-4, 686. Properties and Syntheses of Sweetening Agents n-butyl derivative is 50 times sweeter than saccharose. Sveda et al. further found that the sweetness disappeared when the hydrogen atom on the sulphamyl moiety (NHS03Na) was replaced by a methyl, ethyl or a cyclohexyl group (Table 5). After preparing the benzene analogue and finding that it was tasteless, Table 5 Sodiurn sic/p/iarnatc aiinlog~ienoii-swrrteners RI-NSOJNa I R2 RI R2 Reference Phenyl Cyclohexyl Me(CH2)5--CH, y2 = 172J5 , 272, 1072 H 72,75 Cyclohexylmethyl H 72 Methyl Bui 72 Cycloheptyl Me 72 2-Methylcyclopent-1-yl H 72 2-Me thylcyclohexen- 1-y 1 H 72 Adam ant y 1 H 72 Piperidin-4-yl €3 72 Tetrahydropyran-3-yl H 72 Me(CH*)XH-(n = 4-16) €1 72 Isopropyl H 72 they decided that retention of the sulphamate moiety and the presence of a saturated cycloaliphatic ring was essential for sweetness in sulphamate molecules.S~illane7~confirmed that various saturated cycloaliphatic rings did not destroy the sweetening power of these sulphamates. He prepared a number of sulpha- mates of the formula GI421 >(CH,),NHSO,Na CH, and also showed that if these rings were substituted by methyl or dimethyl groups, the sweetness was retained (Table 6). 72 G. A. Benson and W. J. Spillane, J. Medicin, Chern., 1976, 19, 869. 76 B. Unterhalt and L. Boschemeyer, Z. Lebensm.-Untersuch.-Forsch., 1971, 145, 93.Crammer and Ikan Table 6 Sodium alkyl and cycloalkyl sulphamote sweeteners R-NHS03Na R Relative sweetness Reference sucrose = 1 n=2 10* 72 n=3 41*, 30 72 n=4 34 * 72 n=5 28* 72 30 73CH,r( Mc(CHZ)~CHZ-r1 = 1 0.5” 74 I1 = 2 3.5* 74 n=l 2.9” 70 n=2 10* 2 * The sulphamates were compared lo a 3 % w/v sucrose solution A recent study of Benson and Spillane72 concluded that in the structure- activity relationships of acyclic sulphamates sweetness appears to be present in the system C -N -SO,-II HR 73 K. M. Beck and A. W. Weston, LJ.S.P. 2,785,195/1957.‘’ B. Unterhalt and L. Biischemeyer, Z. LcDenun.-Untersuclz.-Fol.sch.,1972, 149, 227 Properties and Syntheses of Sweetening Agents where there can be one or two a-hydrogens.They argued that the presence of this system is a necessary but not a sufficient condition for sweetness. N-n- Butylsulphamate was three and a half times sweeter than sucrose but N-n- propylsulphamate was only one fifth as sweet and other straight chain sulphamates were not sweet. It appears that from the various examples in Table 5 that there is no relation between the hydrocarbyl moiety and sweetening power. Possibly the proton of the sulphamyl moiety -NHSO? is related to sweetening power in certain sulphamate molecules. 15 Cyclic Sulphamates Saccharin.-The earliest known commercial synthetic sweetening agent was saccharin or 3-oxo-2,3-dihydro-l,2-benzisothiazol-l,l-dioxide(1). Saccharin was discovered accidentally by Remsin and Fahlberg76 at the Johns Hopkins University in 1879during an academic study of the oxidation of o-toluene sulphonamides.Fahlberg noticed that during his evening meal his bread tasted sweet and he traced this the following day in his laboratory to a certain benz- isothiazole derivative which was later known as saccharin. It has been employed as a non-nutritive sweetener for more than 80 years. The original synthesis used by Remsin and Fahlberg nearly 100 years ago is still the principal industrial process of today. The synthesis starts from toluene and it is shown in Scheme 8. An alternative method is used by Maumee Company in the U.S.A.77 starting from anthranilic acid. SQ,Cl Scheme 8 Saccharin is approximately 300 to 550 times as sweet as cane sugar and has a bitter metallic after taste.The sweetness of saccharin and cyclamate relative to sugar is not constant. For example, as the concentration of saccharin in aqueous solution increases, the relative sweetness compared with sugar appears to decrease. '~3 I. Rernsen and C. Fahlberg, Arner. Chem. J., 1879, 1,426. 77 Maumee Co., Chem. Eng., 1954, 61(7), 128. 450 Crammer and Ikan Also the calcium salt is less sweet than the sodium salt. Bitterness becomes apparent with a saccharin solution as the concentration is increased. Furthermore, relative sweetness is affected by acidity, temperature, and the type of food in which it is being measured. So far 17 different reports have been published dealing with the mutagenicity of saccharin.78 Mainly tested as its sodium salt, saccharin has shown mutagenic effects in Salmonella, Drosophila, and in mice.Saccharin was removed from the list of GRAS food additives after it was dis- covered that it can also cause bladder cancer in rats.79 As with cyclamate, there have been detailed studies in order to find the effect of changes in the structure of saccharin on its sweetness. Saccharin forms salts, many of which are sweet, and it is reasonable to assume that the saccharin anion is an essential part of the structure required for the property of sweetness. This is supported by the fact that N-alkyl derivatives of saccharin are tasteless. Holleman and coworkers1 carried out an intensive investigation into the effect of altering the saccharin molecule and discovered some interesting results.If the heterocyclic ring is opened to give the corresponding o-carboxy-benzenesulphonamidethen the sweetness disappears. Further, if the sulphonyl group is replaced by a carbonyl group the product is the tasteless phthalimide analogue, On the other hand if the carbonyl group in saccharin is changed from a sulphonyl group to benzodithioimide then the sweetness is retained, but with a bitter after taste. Duplication of the heterocyclic ring on the other side of the benzene ring also causes loss of sweetness. Substitution in the benzene ring of saccharin with various substituents modifies the taste, In the case of electron donating halogen group, the change from sweetness to bitterness progresses as the change from F to C1 to I.Substitution with the NO2 group and other electron withdrawing groups resulted in the disappearance of sweetness with the introduction of bitterness. The p-ethoxy group, essential for sweetness in the dulcin molecule, is not effective in the saccharin molecule. Since perillartine, which is the oxime of perillaldehyde, is sweeter than perillaldehyde, and the oximes are known to be sweet, it might be expected that the introduction of the oxime group on the carbonyl would modify the sweetness, but this is not so in the case of saccharin when sweetness disappears on oximation. Table 7 shows the relation of saccharin analogues and sweetening potency. In the pursuit of other types of saccharin analogues Clauss and JensenSo accidentally discovered a different class of non-toxic sweetening agents, 1,2,3- oxathiazin-4(3H)-one 2,2-dioxides (Table 7).Their alkaline salts were very soluble in water, and like the cyclamates, were stable to hydrolysis. They were prepared according to Schemes 9 and 10. Very recently BASF87 AG have discovered that the replacement of the benzene ring of saccharin with thiophene does not enhance its sweetening power. Further the 3 analogues (24,25,26) that were prepared were found to have no unpleasant taste and were non-toxic (Scheme 11). 78 P. G. N. Kramers, Mutation Research, 1975, 32, 81. 7s G. T. Bryan and E. Erturk, Science, 1970, 167, 996. s°K. Clauss and H. Jensen, Angew. Chem. Internat. Edn., 1973, 12, 869.BI0. Hromatka and D. Binder (BASF AG.), G.P. (Offen). 2,534,689/1976. sz ch Table 7 Saccharin and heterouzine analogues and their sweetening potency V R2".Iy2,Y' R1 R2 X Y Yl Y2 Relative sweetness Reference sucrose = 1 Na co so2 0 50 5 Bitter after taste H co so2 NH Tasteless 80 H so2 co NH Sweet 80 H so2 co 0 Tasteless 80 H co so2 0 100 81 H co so2 Bond 500 76 H co co Bond Tasteless 2 H so2 so2 Bond Sweet 1 Bitter after taste CH=CH-CH=CH H HC=N-OH so2 Bond Insipid taste 108 CH=CH-CH=CH Substituted by F, C1, H co SO2 Bond Sweet 1 Me, NH2 H H Na CH2 so2 0 10 80 H Me Na CH2 so2 0 130 80 Me H Na CH2 so2 0 20 80 Me Me Na CH2 so2 0 130 80 H Et Na CH2 so2 0 150 80 Et H Na CH2 so2 0 20 80 Me Et Na CH2 so2 130 80 Et Me Na CH2 so2 250 80 H Bun Na CH2 so2 30 80 Prn Me Na CH2 so2 30 80 Et Pm Na CH2 so2 70 80 Me Me Double CHCOMe SO2 Tasteless 80 Bond with Y H Me Me co so2 0 Tasteless 80 H Me H co so2 NH Tasteless 80 H Me H co so2 NMe Tasteless 80 H H H co so2 0 10 81 Me H H co so2 0 Sweet 82 Me H H 0 so2 co Sweet 83 81 K.Clauss, H. Jensen, and H. Luck (Farb. Hoechst AG.), G.P. (Offen.) 2,228,423/1973. 82 K. Clauss, H. Jensen, and H. Schnabel (Farb. Hoechst AG.), G.P. (Offen.) 2,237,804/1973. 83 K. Clauss and H. Jensen (Farb. Hoechst AG.), G.P. (Offen.) 2,264,235/1974. lo*A. Mannessier-Mameli,Gazzetta., 1932, 62, 1062. Properties and Syntheses of Sweetening Agents 0 0 0 0 0 Reagents: i,O0 MeCCMe; ii, H,O; iii,80 MeCH,COMe; iv,84 MeCOMe; v,~~MeCOCH,-COCOCMe,; v,"" MeCOCH,COsH Scheme 9 I0NHSOaR CHs I CH3CO2C=CHCONHS0,R __f HsC Reagents :i,80 EtCOMe; ii,06 MeCO,C(Me)CH, Scheme 10 84 Farb.Hoechst AG., B.P. 1,340,911/1973.e5K.Clauss, H. Jensen, E. Schmidt, and H. Pietsch (Farb. Hoechst AG.), G.P. (Offen.)2,434,54711976. K. Clauss, E. Schmidt, H. Jensen, and H. Pietsch (Farb. Hoechst AG.), G.P. (Offen.) 2,434,54911 976. Crammer and Ikan (24) 2,3-dihydro-3-oxothieno[2,3-d]isothiazole-1,l-dioxide 02 (25) 2,3-dihydro-3-oxothieno 1,l-dioxide[3,2-d]isothiazole-0 (26) 2,3-dihydro-3-oxothieno[ 1,l-dioxide3,4-d]isothiazole-S0,CI CO,CH, Reagents: i, NaHSO,; ii, Hf; iii, MeOH; iv, NH,; v, NaOCH, Scheme 11 16 Dipeptide Esters Aspartame.-Aspartame or ~-aspartyl L-phenylalanine methyl ester (27) was also accidentally discovered by James Schlatters8 of G.D. Searle in the early sixties, and in 1966 by Davey and coworkers of I.C.I.89while they were involved in the synthesis of gastrin and its tetrapeptide analogues, but the sweetness was un- noticed. The surprisingly potent sweet taste of this dipeptide is completely unexpected and could not have been predicted from its chemical structure. The 88 R. H. Mazur, J. M. Schlatter and A. H. Goldkamp, J. Amer. Chem. SOC.,1969, 91, 2684. J. M. Davey, A. H. Laird, and J. S. Morley, J. Chem. SOC.(C), 1966, 555. Properties and Syntheses of Sweeteniiig Agents sweetness may be related to the polarity of the molecule since the free carboxylic acid (28) is tasteless.H,NCH -CONH -CH -C02CH3 H2NCH-CONH -CH-CO,HI I CH2C02H CH,CO,H (27) (28) The sweetening property of (27) is dependent on the stereochemistry of the individual amino acids, i.e. the aspartyl and phenylalaninyl moieties from which it is derived. Each of the two amino acids can exist in two optically isomeric forms. Aspartame, the L,L optical isomer, is sweet whereas the corresponding D,D; D,L and L,D isomers are tasteless. The combination of isomers (racemates) which contain the L,L form i.e. DL-aspartyl-L-phenylalanine, L-aspartyl-DL-phenylalanine and DL-aspartyl-DL-phenylalanineare also sweet. Aspartame is conveniently prepared by suitable processes involving the coup- ling of amino acids (Schemes 12 and 13) and a method for the large-scale manu- facture of aspartame has been reported by Ariyoshi and cow~rkers.~O 0CH,OCONH -CH -C02 0CH2-TH -C0,CH3I N H,HC1 CH2C02CH2 1 CH2Ph I Ii Aspartame CONH-CH -C0,CH3 -1(27) CH -NHC02CH2 -Ph I Reagents: i, 65 "C, 24 hr; ii, Pd-ACOH, H2 Scheme 12 NH2 Reagents: i, THF, N, ( <0 OC); ii, PhCH,CH(NH3C0,CH3, Et,N-THF (<0 "C) Scheme 13 @O Y Ariyoshi (Ajinomoto Co.Inc.), Kaguku to Sebutsu, 1974, 12(3), 189. Crammer and Ikan Aspartame has a sweetening potency of about 180 times that of sucrose. This is surprising since neither L-aspartic acid nor L-phenylalanine is sweet. In fact, while a number of small peptides are known to be bitter, aspartame is the only known peptide to elicit a sweet taste.The potency of aspartame relative to sucrose decreases with increasing concentrations of sucrose. It was found to be about 180 times sweeter than 2% sucrose solutions, but only 40 times sweeter than 30% sucrose. There is evidence of a synergistic increase in sweetening potency when it is combined with other ingredients in foods. The unpleasant after taste (e.g. bitter, metallic, saline) characteristic of saccharin and cyclamate does not occur with aspartame. It would be expected that as aspartame is derived from naturally occurring amino acids it would be devoid of any toxicity and extensive studies have shown this to be except for persons affected by phenylketonurea.92 Like many sweeteners aspartame is more acceptable and relatively sweeter in low concentrations than in high concentrations.It is unstable in acidic solutions and therefore not suitable for use in carbonated beverages and fruit products: also, it decreases in sweetness during storage and during hot pro~essing.~ Schlatter and coworkers investigated the aspartame molecule in order to specify the sweetening site of the molecule. They found that the C-terminal amino acid could be changed to a considerable degree even to the extent of replacing the methoxycarbonyl moiety by a methyl group without any significant loss in sweetness potency (Table 8). It appeared, however, that the presence of the free unsubstituted amino group and the carboxy group of the aspartamine moiety as well as the minimum distance between them, and the absolute configuration of the asymmetric carbon atom of the aspartyl moiety were necessary for sweetness.The L configuration of the phenylalaninyl moiety of aspartame was also an essential factor. It was shown (Table 8) that the sweetening potency decreased markedly with increasing size of the ester group. The ethyl ester of (27) had 25 % sweetening potency whereas the n-propyl ester had only 1 % sweetening potency. The sweetening potency was increased if the benzene ring was hydrogenated implying a conformational structural relationship between the non-planar cyclohexane ring and the planar benzene ring (Table 8). The problem of explain-ing the diversity of structure giving a sweet taste has not yet been satisfactorily solved.Various theories based on the assumption of a single complex receptor site have been made by Shallenbergerg3 and more recently by Kier.94 No prediction could have been made that the ester group could be replaced by alkyl groups without major change in biological activity. The essential requirement for sweetness appears to be an a-amide of L-aspartic acid in which the nitrogen atom of the amide group is attached to an asymmetric carbon bearing two unidentical groups. When the groups are identical no sweetness results. When the groups are S. L. Halpern, in ‘Scientific Review of a New Sweetener,’ American CoIlege of Nutrition, First Annual Interim meeting, 11, November 16-18, 1974.(12 ‘Aspartame,’ Technical Bulletin No. 600 (060473), Searle Biochemics, G. D. Searle, Arlington Heights, Ill., U.S.A. ”? R. S. Shallenberger and T. E. Acree, NatzirP, 1967, 216, 480. ”’ L. B. Kier, J. Pharm. Sci., 1972, 61, 1394. Properties and Syntheses of Sweetening Agents Table 8 Aspartame analogues H2N-CH-CONH4H-RI I I CH,C02H R2 Ri R2 Relative sweetness sucrose = 1 1 CO2Me 100-200 2 COsEt 25-50 3 COzMe 225CHz 0 4 Me (Laevo) 50 5 Me (Dextro) 0 6H 0 7H Bitter 8 Me 10 9 Me 20 10 Me 10 11 Me 30 12 Me 20 13 Me 50 14 Me (Laevo) 100 15 Me 10 16 COzMe (Laevo) Me 0 458 Reference 95,96 96 97 97 97 97 91 97 97 97 97 97 97 97 97 97 Crammer and Ikan Table 8 (continued) Ri R2 Relative Reference sweetness sucrose = 1 17 CO2Me (Laevo) Me(CH& 70 97 18 C02Pri Me 125 97 19 C02Pr' 0 97 20 COrMe :z2+7J (Fenchyl) 22,200-33,200 98, 99 21 COzMe c02 300-600 98 22 COzMe (tram) 5450-7300 98 CH 3 23 COzMe 556-880 98 sufficiently different sterically then there is a definite possibility that one of the diastereoisomers will be sweet.The requirement of a particular absolute con- figuration follows from the fact that the receptor site is asymmetric. Compounds 16 and 17 (Table 8) show that amino acid methyl esters with alkyl side chains do not yield sweet derivatives until the receptor site contains an alkyl group of at least 4 carbon atoms (when R2 = butyl, R1 = COzMe the relative sweetness is 40).On the other hand isopropyl esters (R1 = CO2Pri) require C1-4 alkyl groups to ensure sweetness. The compound R2 = n-pentyl, R1 = C02Pri is tasteless because the bulky alkyl group of R2is too near to the ester moiety CO2Rl for sweetening potency. Appreciable potency is detected when the alkyl group for R2has no more than 3 carbon atoms. The most intense sweeteners yet known were discovered by M. Fujino's Japanese gr0up~89~~ working for Takeda Chemical Industries. The compounds were L-aspartyl amino malonic diesters with ester groups selected from alcohols such as fenchyl alcohol and cycloalkanols. The most potent sweetener, methyl R. H. Mazur, A. H. Goldkamp, P. A. James, and J. M. SchIatter, J.Medicin. Chem., 1970, 13, 1217, 5. M. Schlatter, G. D. Searle & Co., Israel P. 30,266/1972. y7 R. H. Mazur, J. A. Reuter, K. A. Swiatek, and J. M. Schlatter, J. Medicin. Chem., 1973,16, 1284. O8 M. Fujino, M. Wakimasu, K. Tanaka, H. Aoki, and N. Nakajima, Naturwiss ,1973, 60, 351. QQ M. Fujino, M. Wakimasu, N. Nobuo, and H. Aoki, Takeda Chemical Industries, Japan B.P 1,434,043/1976. Properties and Syntheses of Sweetening Agents fenchyl L-aspartylaminomalonate (29),was synthesized in the same manner as aspartame (Scheme 14). CH20CONHCHC02H +I CH,CO,CH, H,NCHCONHCHCO2CH, yONHCHCO,CH, I I ii If--CHNHCO,CH,I CH,C02CH2 Reagents: i, DCCD; ii, Pd-AcOH Scheme 14 Again by accident, Milton Lapidus of Wyeth tasted a peptide ~-3-(2,2,2-trifluoroacetamido)-succinanilicacid (30) and found it to be 12 times sweeter than sugar.It was synthesized according to Scheme 15. /OCHCI >O(CF3C0)20 4-NHzCHCOzH CF,CONHCHCI CH,CO,H CF,CONHCHCONHI CH2COZH (30) Reagent: i, PhNHl Scheme 15 looM.Lapidus and M. Sweeney, J. Medicin. Chem., 1973, 16, 136. Crammer and Ikan The L-form of (30) was tasteless. Furthermore the acetylation of aspartame by the trifluoroacetyl group yielded the trifluoroacetyl aspartame which was found to be 120 times sweeter than sugar. This derivative is slightly less sweet than aspartame but its potency invalidates Mazur's assumption that a free amino group is a necessary requirement for sweetness in the aspartyl dipeptide ester analogues. 17 The Tryptophanes In 1968 Kornfeld and coworkers,lol while working on pyrrolnitrin derivatives, accidentally discovered that dl-6-trifluoromethyltryptophanewas intensely sweet.Like Horowitz and his coworkers who had prepared dihydrochalcone analogues, Kornfeld and his coworkers synthesized various tryptophane analogues in order to determine the sweetening site of the molecule. They found that substituting different groups102 in position 6 of the dl-tryptophane molecule gave particularly good results (31). These tryptophane sweeteners were synthesized by well known methods, for example via the Mannich reaction (Scheme 16) or via the R, = H and Rz = CF,, CI, Br. F, CH,O, or (CH,),CH R, = R, = CH, or CI (31) + HCHO + (C,H,),NH -&- RQJ- CH,N~C,H,~, R, H R* H c ii (31) &-RyJ-CH2iK02C2Hd* RZ NHCHOH Reagents: i, AcOH ; ii, (C,H,C02)2CHNHCHO; iii, AcOH-HCI Scheme 16 Reimer-Tiemann route (Scheme 17).Kornfeld and coworkers resolved these dl-tryptophanes into their d and I forms. It was found that D-6-chlorotryptophane was lo00 times sweeter than sugar. The L form, as expected, was tasteless but was lbl E. C. Kornfeld, J. M. Sheneman, and T. Suarez, Eli Lilly and Co., G.P. (Offen.) 1,917,844/ 1969. 1Cs E. C. Kornfeld, J. M.Shcneman, and T. Suarez, Eli Lilly & Co., B.P. 1,269,851/1972. Properties and Syntheses of Sweetening Agents "DT+ CHCI, R2 H H Reagents: i, Hydantoin-Ac20; ii, Na-Hg, [HI; iii, H+ Scheme 17 found to have antidepressant activity instead.Chibata and coworkers success- fully separated the D form of 6-chlorotryptophane from the racemic mixture by addition of an enzyme extract of wheat bran and CoC12, 6H20103 (Scheme 18). Various other heterocyclic compounds have been (accidentally) found to be sweet. CH2CH -C02H c1 H I COCH, dl (II CH,CH-CO,H iii dform CI AH2 H d Reagents: i, AcOH; ii, enzyme extract of wheat bran and CoCl,, 6H,O; iii, OH- Scheme 18 I. Chibata, T. Tosa, T. Mori, and Y. Iwasawa, Jap.P. (Kokai) 75/58060. 462 4- Crammer and Ikan 18 Thiazolo[3,2-b]-a-Triazolesl0* The 3-methyl derivative was prepared according to Scheme 19. -N-NH N-NH+ CH,COCH,Rr (NA Scheme 19 19 Tetrazole Derivatives (a) 4-Amino tetrazoles105 (32) were synthesized according to Scheme 20.(b) 5-(m-Hydroxyphenoxy)-tetrazole~O~(c) 5-Amino tetrazoles (33). H I i RNHCNH, + CHJ -RN =C-NH,HI RN=C-NH,HI II I I S SCH, NHNH, ii, iii, ivJ. Reagents: i, N,H4, H20; ii, HNO,; iii, AgNO,; iv, NaNO,; v, Reflux/xylene Scheme 20 20 3,4-Dehydro-3-Hydroxy-2-(l~-indol-3-ylme~yl)-l-me~yIpiperidin~4-carboxylic acid (34) Hauth and H0fmann1~~ of Sandoz have found that (34) is 500 to loo0 times sweeter than cane sugar. It was synthesized according to Scheme 21. lo4S. Kano, 0.Nomura, and T. Taniguchi, Jap.P. (Kokai) 73/68589. lo6 R. M. Herbst, Eli Lilly & Co., B.P. 1,170,590/1969. loE W. L. Garbrecht, Eli Lilly & Co., B.P. 1,221,115/1971.lo7 H. Hauth and A. Hofmann (Sandoz Ltd.), Swiss P. 574,439/1976. Properties and Syntheses of Sweetening Agents -i CH,NHCH2C02H + Br(CH2),CO2C,HS HOICCH2N(CH2),CO2C&1 CH3 ""6 I CH3 CHI CH, (34) *eReagents: i, Et,N-C6H6; ii, NaOEt; iii, ~~~-~~~~ iv,iH+;v, Ac20-; ~ vi, hydrolysis H Scheme 21 21 Conclusions The search for the ideal sweetener to replace sugar is primarily to satisfy the diabetic who desires a non-nutritive sweetener having no undesirable side effects such as obesity. The ideal sweetener should satisfy the following requirements : 1 It must be reasonably sweet, at least as sweet as sugar, with no lingering after taste. 2 It must be economical to produce and be cheaper than sugar for the consumer market.3 It should be preferably non-caloric and have no nutritive value. 4 It must be non-toxic with no dangerous side effects such as carcinogenicity or teratogenicity. It should not have any synergistic detrimental effects with drugs or in the presence of foods or beverages. 5 The metabolite must also be non-toxic with none sfthe side effests mentioned above Crammer and Ikan 6 It must be thermostable and not decompose during cooking or in the presence of sunlight. 7 It must be soluble in water. So far no sweeteners satisfy all the above requirements. One of the main reasons for being unable to find the ideal sweetener is the lack of understanding of the mechanism responsible for causing the sensation of sweetness. Various theories and hypotheses have been proposed but the most satisfactory to date is Shallenberger’s hydrogen bonding hypothesis.This hypothesis is based on the fact that sweetness depends on a lock-key fit of hydrogen bonds between the sweet molecule and a receptor site with a separation between them of 2.5 to 4A. For successful hydrogen bonding to take place a specific conformative and spatial arrangement of atoms must be present in the molecule. Such an arrangement is present in many sweeteners, natural and artificial, but cannot explain why such an arrangement is also found in non-sweetening molecules. The diversity of struc- tures of sweet compounds suggests that the taste bud protein responsible for initiating a sweet sensation has more than one active receptor site.

ISSN:0306-0012    

DOI:10.1039/CS9770600431

出版商:RSC    

年代:1977

数据来源: RSC

摘要:

MELDOLA MEDAL LECTURE” N.M.R. Spectral Change as a Probe of Chlorophyll Chemistry By J. K. M. Sanders UNIVERSITY CHEMICAL LABORATORY, LENSFIELD ROAD, CAMBRIDGE. CB2 1EW 1 Introduction The shift reagent experiments which were so new and exciting in 1970 can be seen in retrospect to be part of a continuing trend in n.m.r. spectroscopy; it is often possible to derive more detailed information about molecular structure or reaction mechanism from induced changes in a spectrum than from the original spectrum itself. This article is concerned with various techniques of n.m.r. spectral change which have been used to study the chemistry of chlorophylls, but it also includes a section on shift reagents. The emphasis throughout is on the application of these techniques in Organic Chemistry as a whole.In the early days, n.m.r. spectral parameters were regarded almost as funda- mental and immutable. Chemical shifts and coupling constants were tabulated,l and organic chemists began to feel that n.m.r. spectroscopy had reached a useful but boring plateau as was apparently the case for U.V. and i.r.2 Aromatic solvent- induced shifts were a nuisance until it was realized that they could not only improve spectral resolution but also give qualitative geometrical information.3 However, these shifts were small and suffered a lack of selectivity. Paramagnetic effects had long been known, and had found some application4 but were not generally useful. The breakthrough came in 1969. 2 Lanthanide-induced Changes A.Shift Reagents.-In 1969, Hinckley discovered5 a lanthanide shift reagent, the pyridine adduct of Eu(d.p.m.)s.t This /3-diketone complex was a Lewis acid *Delivered 1st April, 1977 at the Annual Chemical Congress, University College, London. td.p.m. is dipivaloylmethane : ButCOCH,COBut L. M. Jackman and S. Sternhell, ‘Applications of NMR Spectroscopy in Organic Chemistry,’ Pergamon, Oxford 1969. a Recently 13Cn.m.r. has added valuable detail, but has contributed little conceptually except in the area of biosynthesis: M. Tanabe in ‘Biosynthesis,’ ed. J. D. Bu’lock (Specialist Periodical Reports), The Chemical Society, London, 1976, Vol. 4, p. 204. 5. Ronayne and D. H. Williams, Annual Reviews of n.m.r. Spectroscopy, 1969, 2, 83. W.D. Phillips, C. E. Looney, and C. K. Ikeda, J. Chem. Phys., 1957, 27, 1435; W. D. Horrocks, R. C. Taylor and G. N. LaMar, J. Amer. Chem. Soc., 1964, 86, 3031. C. C. Hinckley, J. Amer. Chem. SOC.,1969, 91, 5160. 4 467 N.M.R. Spectral Change as a Probe of ChIorophy!l Chemistry which formed adducts with basic lone pairs in organic molecules and which caused selective shifts with little broadening. For the first time it became possible to characterize a nucleus not only in terms of its immediate chemical environ- ment but also by its geometrical relationship with one or more distant functional groups. We soon found that pyridine-free Eu(d.p.m.)s was an even better shift reagent and obtained a first order spectrum of n-hexanol (Figure 1),6 equivalent to thousands of MHz in resolution (but not sensitivity!) for a trivial cost... 1.... I.. .I I...:,.... f , .. 1 ..:,..._I,_.., ;.:.;...::I: :,..:.;.:.I ,....f 10 8 6 4 2 Figure 1 100MHz lH n.m.r. spectrum of n-hexanol in CC14 after the addition of 0.29 equivalents of Eu(d.p.m.), (Reproduced from Chem. Comm., 1970, 422) This beautiful result is seen because the shifts are primarily caused by the pseudo-contact7~*mechanism. In effect the lanthanide ion, which is bound to the hydroxy group, exerts a magnetic field which decreases with distance. More precisely, the shift (AH) is given by equation 1, where R and +are defined in Figure 2. C is a constant for any given adduct at a fixed temperature, but AH = C.R-3(3 COS~~1) (1)-both its sign and magnitude depend on the magnetic anisotropy of the metal ion; for Eu(d.p.m.)s shifts are normally downfield, but for Pr(d.p.m.)s they are ~pfield.~~~ Note, however, that (3 cos2$ -1) changes sign when r$ = 54.7" J.K. M. Sanders and D. H. Williams, Chem. Comm., 1970,422.'E. de Boer and H. Van Willigen, Prog. N.M.R. Spectroscopy, 1967, 2, 111. J. Reuben, Prog. N.M.R. Spectroscopy, 1973, 9, 1. D. R. Crump, J. K. M. Sanders, and D. H. Williams, Tetrahedron Letters, 1970, 4419; J. Briggs, G. H. Frost, F. A. Hart, G. P. Moss, and M. L. Staniforth, Chem. Cornmi., 1970, 749. Sanders Symmetry A xis M3’ i”..\\\ R \\ \ \ \ \\\-\\ -2H Figure 2 Schematic view of a shift reagent-substrate adduct so that reversals in shift direction can be seen.8J0 With shift reagents available, the n.m.r.spectra of many compounds could be modified to yield a wealth of new information, and applications abounded.8 The interpretation of results was relatively simple for monofunctional compounds, but we felt that an approach to polyfunctional cases was necessary. It was apparent, both qualitatively11 and quantitatively,lO that a group’s affinity for loJ. K. M. Sanders, S. W. Hanson, and D. H. Williams, J. Anier. Chem. SOC.,1972,945325. l1 J. K. M. Sanders and D. H. Williams, J. Anter. Chem. SOC.,1971, 93, 641. N.M.R.Spectral Change as a Probe of Chlorophyll Chemistry a shift reagent was in part a measure of the basicity of its lone pair. Therefore, a functional group should be ‘turned off’ by suitable derivatization; e.g.hydroxy (-OH) can usually be converted to trifluoroacetate (-OCOCF3) or tosylate.l* Also, since lanthanides are ‘hard’ metals with little affinity for second row elements, ketones can be protected as ethylene thioketals.12 However, derivatization is not always possible and we must be able to study molecules with multiple binding sites of varying affinity. At low shift reagent concentration, the highest affinity site is partly complexed and its associated protons shift. As the shift reagent concentration increases, the high affinity site becomes saturated, the low affinity site becomes more complexed and the shifts reflect these changes. Figure 3 illustrates this process for the bicyclic 32 / H-1 28 /I /I /-24 20 r-‘ E 2 16 Y-€ dd 12 m 8 4 0 0 0.4 0.8 1.2 1-6 2.0 moles of Eu(dpm), per mole of substrate Figure 3 The effect of added Eu(d.p.m.), on the lH (n.m.r.)spectrum of (1) (Reproduced by permission from TetrahedronLetters.1971, 3733) la D. R.Crump, J. K. M. Sanders, and D. H. Williams, Tetrahedron Letters, 1970,4949. Sanders compound (l).13The same competition effect can be used to study subtle isotope effects on Lewis basicity.14 OH B. Aquo1ons.-SimpleLanthanide ions can be used in polar solvents as shift + reagents for carbo~ylates~~J~ or R3X-0-(where X is N, P, or As).1OJ6 The resulting shifts are normally upfield for Eu3+, presumably due to a change in magnetic anisotropy.Figure 4 shows the effect of europium nitrate on a solution Figure 4 100 MHz lH n.m.r. spectrum of (2) in [2H4]methanol aftev the addition of 0.4 equivalents of Eu(NO,),, 6*H20 of triphenylphosphine oxide (2) in [2H6]acetone. The normally uninformative spectrum is rendered amenable to first order analysis: J(P-Ho) = 11.4 Hz, J(P-HnL) = 3.0 Hz.10 ('1 l3 I. Fleming, S. W. Hanson, and J. K. M. Sanders, Tetrahedron Letters, 1971, 3733. l4 J. K. M. Sanders and D. H. Williams, Chem. Comm., 1972, 436. l5 F. A. Hart, G. P. Moss, and M. L. Staniforth, Tetrahedron Letters, 1971. 3389. l6 J. K. M. Sanders and D. H. Williams, Tetrahedron Letters, 1971, 2813. 471 N.M.R. Spectral Change as a Probe of ChlorophyIl Chemistry Lanthanide ion-induced shifts provide a powerful method of mapping metal binding sites in proteins.17 Complexes such as Ln(edta)- may also be useful.18 C.Relaxation Effects.-Lanthanides can also induce changes in relaxation times which can be used to probe molecular shape.lg The principles of relaxation times are discussed in Section 4B. 3 Radical-anion-induced Changes An amine-catalysed dimerization of 2-methyl-l,4-naphthoquinone(3), to give the pentacene derivative, (4),was discovered in this laboratory, but attempts to extend the reaction to quinones such as (5) failed.20 The only dimeric product isolated was (6), the stereochemistry being determined using a shift reagent.21 0 0 (3) R = H (5) R = Me or PI1 In an attempt to study the mechanisms of these dimerizations we ran n.m.r.spectra of the reaction mixtures. Addition of small amounts of t-butylamine to a solution of (3) in [2H4]methanol caused selective broadening in the n.m.r. spectrum (Figure 5) without affecting solvent or reference signals. Various experi- ments22 convinced us that the broadening was due to a small amount of semi- l7 I. D. Campbell, C. M. Dobson, R. J. P. Williams, and A. V. Xavier, Annals New York, Acad. Sci., 1973, 222, 163. G. A. Elgavish and J. Reuben, J. Amer. Chem. SOC,1977, 99, 1762. l8 J. W. Faller, M. A. Adams, and G. N. LaMar, Tetrahedron Letters, 1974, 699. 2o 1. Baxter, D. W. Cameron, and R. B. Titman, J. Chem. SOC.(0,1971, 1253. p1 I. Baxter, D. W. Cameron, J. K. M. Sanders, and R.B. Titman, J.C.S. Perkin Z, 1972, 2046; I. Baxter, J. K. M. Sanders, and G. E. Evans, J.C.S. Perkin I, 1974, 2574. sa J. K. M. Sanders and 1. Baxter, J.C.S. Chem. Contm., 1974, 255. Sanders 0 0 Figure 5 100 MHz ‘H n.m.r. spectrum of (3) in [2H4]methanon (upper trace) and in the presence of a trace of (3.) (lower trace) quinone, (3r)7 undergoing fast electron exchange with (3). Similar effects occur with radical cations. The spectrum exhibits characteristics of both species, and the protons closest to unpaired spin density broaden most. Thus we had a beautifully visual way of mapping unpaired spin density and molecular orbitals in an organic molecule.23 This technique had been known for many yea~s,2~ and the theory was well established.25 Its use in assigning e.s.r.spectra had been advocated26 but mostly ignored.27 We used it successfully on a variety of small molecules2* and then decided to look at a large molecule where e.s.r. fails, the chlorophyll radical cation. The remainder of the article is concerned with chlorophyll but I should 23 The relationships between spin density, molecular orbital coefficients, and chemical reactivity are described in I. Fleming, ‘Frontier Orbitals and Organic Chemical Reactions,’ Wiley, London, New York, Sydney, 1976. 24 C. R. Bruce, R. E. Norberg, and S. I. Weissman, J. Chem. Phys., 1956,24,473. a5 E. de Boer and C. MacLean, J. Chem. Phys., 1966, 44, 1334. *6 G. T. Jones and J. N. Murrell, Chem. Comm., 1965, 28. 27 See however, C. S.Johnson and R. Chang, J. Chem. Phys., 1965,43,3183; H. Angad-Gaur,C. H. Wijtzes, and J. Smidt, Mol. Phys., 1969, 17, 179; J. Chaudhuri, S. Kume, J. Jagur-Grodzunski, and M. Swarc, J.Amer. Chem. SOC.,1968,90,6421. Quinone and nitro-aromatic radical anions and aminium radical cations: 1. Baxter and J. K. M. Sanders, unpublished work. 473 N.M.R.Spectral Change as a Probe of Chlorophyll Chemistry point out here that we are still no wiser about the mechanisms of the quinone dimerizations ! 4 Chlorophyll Radical Cations A Introduction.-The fundamentally important step in photosynthesis is light- induced generation of chemical oxidizing and reducing power. This is achieved by a chlorophyll-mediated electron transfer in the following manner hv chlorophyll +chlorophyll* (2) Acceptor + chlorophyll* -+chlorophyll: + acceptor-(3) Donor + chlorophyll: -+ chlorophyll + donor+ (4) Summation of equations (2)-(4) gives (5) hvDonor + Acceptor -+ Donor+ + Acceptor-(5) Clearly donor+ is an oxidizing agent and acceptor- is a reducing agent but their exact chemical nature remains unclear.29 These reactions take place in a photo- synthetic reaction centre and involve only a small proportion of the chlorophyll n, (7).The bulk of chlorophyll a, and all the chlorophyll by(8)y is simply used as an antenna, harvesting incident sunlight and transferring it to the reaction centre. Many photosynthetic bacteria use the closely related bacteriochlorophyll (9) in photosynthesis. OPhytyl I C0,Me 0 Phytyl (7) R = Me (S) R = CHO (9) A.J. Bearden and R. Malkin, Quart, Rev. Biophysics, 1975, 7, 131. Sanders Equation (3) above, shows that the radical cation (7:) is a crucial intermediate, but until very recently little was known of its properties; the e.s.r. spectra of (7:) and (9;) are single signals containing over lo9unresolved lines! ENDOR, a double resonance technique, has very recently30J1 yielded detailed spin distribu- tions but when we began our work ENDOR was rather limited.32 ENDOR and e.s.r. suffer the fundamental disadvantage that peak assignments must be made chemically whereas n.m.r. spectra can be assigned spectroscopically (Sections 4B and 4C).33 We set out, therefore, to study (71) via its effect on the n.m.r. spectrum of (7) which is rich in information (Figure 6).The very richness of the spectrum poses the first problem-assignment of resonances. The spectrum had been assigned previously using various analogues and chemical shift arguments, and aggregation-induced shifts had been most informative.34 We decided to develop more general spectroscopic methods for assignment which would be useful for a wide range of porphyrins and chloro- phylls. Note that these methods are based on spectral changes which yield structural information. B. Spin-lattice Relaxation Times (TI).-In the n.m.r. experiment, we observe transition from a lower to an upper nuclear energy level, which requires electro- magnetic radiation at the correct frequency.The radiationless return to the lower level is spin-lattice relaxation and requires fluctuating magnetic fields of the same frequency. These fields can arise from many SOU~C~S,~~ but for organic compounds, in practice they are due to the tumbling of other protons in the same molecule; this is the dipole-dipole intera~tion.~~ The intensity of fields of the correct frequency at the relaxing proton depends both on its distances from other protons and on their tumbling times. This is illustrated in Figure 7 and expressed mathematically in equation (6) TI-1 cc r cr-s (6) TI-1 is the relaxation rate and is the reciprocal of the relaxation time (TI), r is an effective correlation time,37 and r is the distance from another proton. Clearly, if two protons are close in space they will relax each other efficiently and have short TI’S.A proton attached to a rigid framework (porphyrin, steroid, ?O H.Scheer, J. J. Katz, and J. R. Norris, J. Amer. Cheni. Sac., 1977, 99, 1372. 31 D. C. Borg, A. Forman, and J. Fajer, J. Amer. Chem. SOC.,1976, 98, 6889. 3z J. R. Norris, H. Scheer, M. E. Druyan, and J. J. Katz, Proc. Nat. Acad. Sci. U.S.A., 1974, 71, 4897. 33 The n.m.r. experiments yield only relative spin densities and ENDOR is necessary to provide absolute values. 34 H. Scheer and J. J. Katz in ‘Porphyrins and Metalloporphyrins,’ ed. K. M. Smith, Elsevier, Amsterdam 1975, p. 399. 35 T. C. Ferrar and E. D. Becker, ‘Pulse and Fourier Transform NMR,’ Academic Press, London, New York, 1971, Chapter 4.36 For details of when this is true, and for a good review of proton relaxation in general see L. D. Hall, Chem. SOC. Rev., 1975, 401. 37 7 Contains contributions both from the overall rate of molecular reorientation and from internal rotation: H. B. Coates, K. A. McLauchlan, I. D. Campbell, and C.E. McColl, Biorhem. Biophys. Acta, 1973, 310, 1. N.M.R. Spectral Change as a Probe of Chlorophyll Chemistry -----ij 61U t Lo co a3 0 c Sanders H=*.H H........H short long H H short long Figure 7 The dependence of proton TI on environment or alkaloid) will tumble slowly and have a short TI, whereas if it is part of a small molecule or it has motional freedom within a large one then T will be short and TI long.Steric hindrance of C-Me bond rotation will also lead to short TI'S. It is not always easy to separate the-r and r contributions. However, within methyl groups each proton relaxation is dominated by interaction with the other protons on the carbon to which they are attached. The Cr-6 term is therefore constant and TI is simply a measure of T. Thus, we can predict that as we move along a flexible alkyl chain away from a rigid framework, TI should increase smoothly. Our results (Table 1)show that this prediction is correctF8 and that TImeasure Table 1 Typical TI values for metaalloporphyrins in co-ordinating solvents at 310 K TI (sec) Substituent CH2 CH3 -CH3 - 0.5 -CHzCH3 0.3 0.7 -CHzCH2C02Me4 a 0.3 /3 0.5 1.1 ments could provide a method of sorting protons into groups according to their distance from a macrocycle.Absolute TIvalues are heavily dependent on solvent, temperature, and aggregation state but relative values within a molecule have been reliable (so far!). Relaxation times are rnea~ured3~ by applying a '180" pulse' to the sample, s8 I. S. Dennis, J. K. M. Sanders, and J. C. Waterton, J.C.S. Chem. Cumm., 1976, 1049; 1977, 192. 477 N.M.R. Spectral Change as a Probe of Chlorophyll Chemistry which inverts all the spins, and monitoring their return to equilibrium. A normal spectrum taken with a 90" pulse immediately after the 180" pulse shows all peaks to be negative. As the spins relax towards equilibrium the negative-going peaks decrease in intensity, pass through a null point (at roughly 0.69 7'1) and grow in a positive sense.It is therefore possible to measure TI by observing 'partially-relaxed' spectra at various times after the 180"pulse. Figure 8 illustrates A A I P P CO, Me 5-0 sec )), ~ I 0.55 n d 1 II I 3'85 4 Figure 8 Partially-relaxed 100 MHz IH n.m.r. spectra of uroporphyrin-ZI-octanzethyl ester, 6 mM in CDCI,. Each spectrum is displayed with a different vertical scale. A = -CCH,CO,Me, P = -CH2CH,CO2Me results for a typical porphyrin. 0.15 Seconds after the pulse only the protons directly attached to the macrocycle are up. After 0.55 seconds, only the pro- Sanders pionate methyl signal is still inverted as it has the longest TI.Therefore, solely on this distance basis we can assign not only the a and /3 methylenes of the pro- pionate side chain but also the acetate and propionate methyl groups.This is a useful technique for unravelling new natural product structures, partly through the sorting of groups by distance, and also because partially relaxed spectra with some nulled peaks are simpler to interpret. TI measurements on chlorophylls give the expected resulW but do not allow distinction between methyl groups in similar positions, and these assignments require Nuclear Overhauser Effects (see next section). Note, however, that all the ~PSOprotons in (9) have the same r but H, has few proton neighbours (due to the acetyl group). It therefore has a much longer TIthan HFor H, and can be confidently assigned.38 Relaxation times also tell us something of the conforma- tional mobility of the phytyl group.39 As deuterium has a smaller magnetogyric ratio than protium, it is much less effective in causing proton or 13C relaxation and leads to lengthening of TI’S in its vicinity.This can be a useful technique for locating the position40 and stereo- chemi~try3~of deuterium in a molecule. On the other hand, paramagnetic ions generate intense magnetic fields, causing TIto shorten by an amount proportional to r-6. This is the basis of the mapping method mentioned in Section 2C. C. Nuclear Overhauser Effect (NOE).-If proton A causes proton B to relax, then saturation of the A resonance will cause the B resonance intensity to increase by up to 50%;41 this is the NOE.It arises through the dipole-dipole interaction, and if several protons are close to that being observed then their relative NOE contributions are proportional to each r-6. The sum of their NOE’s cannot exceed 50%. The rigidity of the porphyrin macrocycle makes it ideal for these experi- ments, and the positive observation of an NOE is powerful evidence that two nuclei are close to each other. Table 2 confirms the value of the technique for Table 2 Nuclear Overhauser Eflects for (lO)aJ Signal irradiated % increase Ha HI H, H,vinyl-CH 15 15 -CH2 --30 --CH3 20 20 -30 a J. K. M. Sanders, unpublished results; In COC13/[2H5]pyridine the model compound (10).Extension to chlorophylls is in principle simple; the unambiguous signals of (7) are the 8-methyl doublet at 1.75 6 and the 2a-vinyl double doublet at 8 6 (see Figure 6). Using these starting points it is possible to ‘walk’ round most of the macrocycle, obtaining connectivities from successive 39 J. C. Waterton and J. K. M. Sanders, unpublished results. *O L. M. Jackman and J. C. Trewella, J. Amer. Chem. SOC.,1976, 98, 5712. 41 J. H. Noggle and R. E. Schirmer, ‘The Nuclear Overhauser Effect,’ Academic Press, London, New York, 1971. N.M.R, SpectralChange as a Probe of Chlorophyll Chemistry R Me C0,Me C0,Me NOE experiments.39 In addition, these experiments, when carried out with utmost care, reveal in (9) small long-range couplings which were previously unsuspected and which are invaluable in making assignments.39 The experiments are, however, quite difficult to do, and in the long run it may be that NQE methods will be replaced by Hall's selective TI method, which is based on the same relaxation phenomenon but is numerically more reliable.42 D.Distribution of Unpaired Spin Density.-With assigned spectra available, we could interpret the line broadening effects of (7f), generated by oxidation of (7). This can be done electrochemically,31 or with iodine30 but for our work the most convenient method is with the crystalline radical cation (lQ43yM Addition of (12) to a chlorophyll solution sets up an equilibrium (equation 7), the spectro- scopic result of which is shown in Figure 9.R3N' + chlorophyll + R3N + chlorophyllf (7) It is immediateIy obvious that "K. Bock, R. Burton, and L. D. Hall, Canad. J. Chem., 1977, 55, 1045; L. D. Hall and H. D. W. Hill, J. Amer. Chem. SOC.,1976, 98, 1269. *I F. A. Bell, A. Ledwith, and D. C. Sherrington, J. Chem. SOC.(C), 1969, 2719. We have also used (13) and the perchlorate salt of (tetraphenylporphyrin) t;81,43these give essentially the same results as (12). Sanders N i 4 =.’7 i i I‘\--------i 481 N.M.R.Spectral Change asaProbe of Chlorophyll Chemistry (i) the amine signals (-7 6) are sharp, confirming that the equilibrium lies to the right. (ii) some chlorophyll signals have disappeared, some are broadened, and others unaffected.(iii) the very broad Hp resonance has moved upfield.45 The broadened spectrum of Figure 9 arises from Iess than 1 % oxidation of (7). Solutions containing even less radical cation show highly selective effects. The various proton resonances broaden in the following order, covering a range of ca. 340 fold between 7 and 4b at ambient temperature46 7 > 8 > 5a > la,3a > 8 > a > 4a > 10 > 2b > 2a > fl > 4b The broadening arises through the electron transfer reaction in equation 8. In the radical cation, protons are coupled to the unpaired electron with k (7) + (7:) +(7.) + (7) (8)k hyperfine coupling constants, aH, of up to 110 MHz. At ambient temperatures the exchange rate k is fast compared with UH and each line is broadened by an amount proportional to the square of aH.25 From relative broadenings we can therefore get relative values for all significant hyperfine couplings in (7:), (st), (9:) and their derivatives.39~~6 The hypefine coupling can be related to spin density, p, by the McConnell relation (equation 9), where Q is a ‘constant’ of notorious variability having a value of roughly 70 MHz* for aH= Q.p (9) a proton attached to a r-system (14) and 75 MHz* for a methyl group attached to a r-system (15).*’ Due to some confusion in porphyrins4* we synthesized (1 1) in order to actually measure Q values.49 If the alkyl substituents have a negligible effect on the spin distribution, then p should be the same for each position and relative CLH’S would reflect relative Q values.Thus from observed relative broadenings in (1 1:) we find that QH :QM~:QCH, = 1.O :1.28 :0.54 The very low figure for QCH~indicates that the propionate side chains are out of *2.8 MHz = 1 G 45 This shift is unobservable at 100 MHz. Access to the 270 MHz instruments at Oxford and PortsmoEth is gratefully acknowledged. 46 J. K. M. Sanders and J. C. Waterton, J.C.S. Chem. Comm., 1976, 247. J. R. Bolton in ‘Radical Ions,’ ed. E. T. Kaiser and L. Kevan,Wiley, London, New York, Sydney, 1968, p. 1. 48 K. Wuthrich, B. J. Wyluda, and W. S. Caughey, Proc. Nat. Acad. Sci. U.S.A., 1969, 62, 636; R. G. Shulman, S. H. Glarum, and M. Karplus, J. Mol. Biol., 1971, 57, 93. 48 C. G. Newton and J. K.M. Sanders, unpublished work. Sanders plane leading to minimal overlap of the C-H a-bonds with the wsystem (16). Armed with QH and Q we can compare our observed spin densities with MO calculations. The agreement is qualitatively reasonable but calculations entirely fail to predict the highly asymmetric distribution that we46 and Katz30 find. Spin distributions for (8:) and (9:) and various derivatives are also poorly predi~ted.3~ Solutions containing radical cations are reasonably stable, although in some cases specific decomposition products can be i~olated.~O The great amplifica- tion of radical effects given by the electron transfer process (equation 8) means that we can detect the paramagnetic species at concentrations as low as mol 1-l.E. Electron Transfer Rates.-When chlorophyll solutions containing radical cation are cooled, the hyperfine broadening increases. On further lowering of the temperature, we observe sharpening,39 which is caused by the electron transfer rate, k, becoming slow on the NMR time scale so that we are no longer seeing an averaged spectrum. Using standard equations25 we can calculate the expected dependence of broadening on k (Figure 10) and compare the results with experi- mental observations. Particularly diagnostic are the coalescence temperatures (point of maximum broadening) which are dependent on the hyperfine coupling. From these results we can derive electron transfer rates and activation energies and investigate how they, and therefore the mechanism, depend on structure, solvent, and other factors.This work is at an early stage, but looks very promising. F. Radical-induced Pseudocontact Shifts.-The upfield shift of Hg in (7 +)(Figure 9) is too large to be entirely due to the contact mechanism,25 and we decided to test whether a shift-reagent type pseudocontact effect could operate. We chose the capped porphyrin (17) as the protons of the alkyl ‘cap’ are shifted upfield by the ring current.51 6o R. G. Brereton and J. K. M. Sanders, unpublished work. 61 M. Turnbull and A. R. Battersby generously provided the magnesium-free porphyrin: A. R. Battersby, D. G. Buckley, S. G. Hartley, and M. D. Turnbull, J.C.S. Chern. Conzm., 1976, 879. 5 483 N.M.R. Spectral Change as a Probe of Chlorophyll Chemistry FAST EXCHANGE > log (exchange rate) Figure 10 The dependence of broadening on electron transfer rate for meso protons of chlorophyll a Slight oxidation of (17) using (12) (41 %) gave exchange broadening of the hypefine-coupled peripheral protons in the usual way. Further oxidation causes the alkyl cap protons to shift strongly downfield;52 the highest field resonance at -0.6 8 is paramagnetically shifted downfield by an amount corresponding to at least 21 p.p.m.in pure (17:). Other cap resonances are similarly shifted I* J. K. M. Sanders and J. C. Waterton, submitted for publication. Sanders downfield by an amount proportional to their original upfield ring current shift. Clearly these protons experience a magnetic field which has the same geometri- cal properties as the aromatic ring current but which is ten times larger and has the opposite sense.This we believe to be the first observation of a large radical- induced pseudocontact shift. The reasons why such shifts have not previously been seen are instructive and are discussed elsewhere.52 In principle these results may allow the chlorophyll radical cation to be used as a natural spin label which, through its broadening and shift powers, could be used to map photo- synthetic reaction centres but in practice the difficulties will be formidable. Exhaustive experiment^^^ on the concentration dependence of the Hsshift in (7:) unfortunately are not consistent with contact, pseudocontact or a combina- tion of these mechanisms, and its causes remain a mystery.G.Modified Chlorophy1ls.-Spectroscopic evidence suggests that in the photo- synthetic reaction centre a pair of chlorophyll molecules, held together by water as a non-covalently-bound dimer, is the true catalyst of equations (2)-(4).31953 We therefore synthesized a dimer (18) in which two chlorophylls were joined 0 (18) [substitution pattern is same as in (7)] via the carbonyls normally carrying the phytyl group.= Chlorophyll is sensitive to acid, base, heat, light, and air so the synthesis (Scheme 1) avoided all these reagents. The corresponding dimer using an ethylene glycol bridge can only be synthesized in the magnesium-free series followed by metal replacement.53-55 Our shorter sequence gives a much better overall yield of dimer.The dimers fold up in the presence of water to give a reaction centre model. The folding is accompanied53-55 by dramatic changes in n.m.r., e.s.r. and elec- tronic spectral properties which should yield detailed information on molecular conformation. Electron transfer studies on (18:) should also prove illuminating. We have also made (Scheme 2) the tailed chlorophyll (21) in which the imida- 53 M. R. Wasielewski, M. H. Studier, and J. J. Katz, Proc. Nat. Acad. Sci., U.S.A., 1976, 73,4282. 54 I. S. Denniss and J. K. M. Sanders, unpublished results. I.5 S. G. Boxer and G. L. Closs, J. Amer. Chem. SOC.,1976, 98, 5406. N.M.R. Spectral Change as a Probe of Chlorophyll Chemistry CO,H Reagents: i, s inach beet chlorophyllase; ii, dicyclohexylcarbodi-imide(DCC)/HS SH ;iii, bcc/l, Scheme 1 Reagents: ClC0,Et; ii, Scheme 2 zole can bind intramolecularly to the magnesium.54 This binding is accompanied by large upfield shifts of the 'tail protons which sit over the macrocycle.An added ligand such as pyridine or N-methyl imidazole can either displace the tail or bind to the metal ion from the other side of the ring (Figure 11). Simultaneous n.m.r. monitoring of the ligand and 'tail' resonances enables the relative import- ance of these processes to be determined.54 In principle, the temperature depend- ence of the equilibria allows separation of the enthalpy and entropy contributions to the inter- and intra-molecular reactions but the detailed molecular interpreta- tion of these results is difficult.5 Conclusions N.m.r. spectral change induced by complexation, double irradiation or pulse methods can give much detail about the structure, stereochemistry and mobility Sanders L Figure 11 The binding of ligands to (21) of molecules. More exciting in the long run is actually watching chemistry in the n.m.r. tube indirectly via electron transfer broadening, CIDNP,S6 and related effects,57 or directly via flow n.m.r.58 I am grateful to the Society of Maccabaeans and the Royal Institute of Chemistry for the award of the Meldola Medal, and to the Science Research Council for generous support. I am deeply indebted to Iain Denniss and John Waterton for their dedication, ideas and enthusiasm in the chlorophyll work and to Dudley Williams who was my supervisor and is now my most effective critic.56 ‘Chemically Induced Magnetic Polarisation,’ ed. A. R. Lepley and G. L. Closs, Wiley,London, New York, Sydney, 1976. 67 S. G. Boxer and G. L. Closs, J. Amer. Chem. SOC.,1975, 97, 3268. 68 C. A. Fyfe, A. Koll, S. W. H. Danuji, C. D. Malkiewich, and P. A. Forte, J.C.S. Chem. Comm., 1977, 335.

ISSN:0306-0012    

DOI:10.1039/CS9770600467

出版商:RSC    

年代:1977

数据来源: RSC

摘要:

Prostaglandins, Thromboxanes, PGX :Biosynthetic Products from Arachidonic Acid By K. H. Gibson ICI PHARMACEUTICALS DIVISION, MERESIDE, ALDERLEY PARK, MACCLESFIELD, CHESHIRE 1 Introduction During the two years since the last review on prostaglandins (PG’s) appeared in this series1 significant advances have been made in this field. The identification of thromboxane A2 (TXA2; 5) and prostacyclin (PGX or PGI2; 6) and the unprecedented potencies of these compounds in their effects on blood platelets and blood vessels has given additional impetus to the rapidly expanding area of prostaglandin research and, in addition, revealed new areas in which to search for therapeutically useful drugs. The intention of this review is to consider the biosynthesis of prostaglandins, thromboxane A2, and prostacyclin, and, in addition, it is hoped to convey some impressionof, (i) the wide range of biological processes in which rhese compounds have important physiological roles, (ii) the possible ways in which control is exerted over the levels of prostanoid material involved in these processes, (iii) the pathological involvement of prostaglandins and related compounds in various disease states, and (iv) the potential uses which can be envisaged, regarding the manipulation of prostaglandin levels or the use of synthetic analogues or antagonists.The major biosynthetic transformations which will be considered in detail are shown in Figure 1. The natural prostaglandins are a family of ‘hormone-like’ substances nominally described as oxidized derivatives of prostanoic acid (7) and characterized by a five-membered ring with two side chains containing, in total, twenty carbon atoms.The nomenclature is derived from the oxidation pattern and is depicted in Figure 2. It appears that prostaglandins are not stored in mammalian tissue but are elaborated in response to various stimuli and virtually every mammalian tissue examined has been shown to have some capacity to synthesize prosta- glandin-related material. Their ubiquitous occurrence, wide range of biological effects, and high potency, have stimulated much research into the natural materials and major pharmaceutical effort has been devoted to the search for analogues which are more selective in their actions and which may have longer- lasting biological effects than the rapidly-metabolized natural prostaglandins.2 History and Development The biological actions of prostaglandins were first reported in the early 1930’s E. W. Horton, Chem. SOC.Rev., 1975,4,589. Prostaglandins etc. :Biosynthetic Products jiom Arachidonic Acid Membrane phospholipid 5. (I) Arachidonic acid (2) PGG,J I' (6) PGX ?H @ -CO,HaoZHIdH IOH OH (5) TXA, Figure 1 Major pathways of arachidonic acid metabolism Gibson (7) Prostanoic acid Figure 2 Prostaglandin nomenclature when it was shown that a constituent of seminal plasma had the properties of lowering blood pressure and stimulating smooth muscle. Because it was thought to derive from the prostate gland the active principle was named prostaglandin by von Euler.2 In fact, most seminal prostaglandin probably comes from the seminal vesicles and although virtually all tissues contain prostaglandin syn- thesizing enzymes, the misnomer ‘prostaglandins’ has now been universally accepted for this class of compound.However, it was not until 1962 that the first isolation, purification, and structural identification of a natural prostaglandin were reported by the Swedish biochemist Professor Sune Bergstrom and his colleague^.^ Subsequently, he and fellow workers at the Karolinska Institute in Stockholm have played a leading role in the rapid increase in the volume of knowledge relating to the prostaglandins, their biosynthesis and metabolism, and their biological properties.The structures of the prostaglandins soon prompted the speculation that they a U. S. von Euler, Klin. Wochensch., 1935, 14, 1182. a S.Bergstram, R. Ryhage, B. Samuelsson, and J. Sjovall, Actu Chem. Scand., 1962, 16, 501. 6 491 Prostaglandins etc.: Biosynthetic Products from Arachidonic Acid were derived biosynthetically from the naturally-occurring, polyunsaturated, 20-carbon atom fatty acids (e.g. arachidonic acid). This was indeed shown to be correct and during the early years after the fist structural identifications, research quantities of prostaglandins were provided by using enzyme systems to convert polyunsaturated fatty acids into the corresponding prostaglandin products.4 However, the discovery of another remarkable natural source of prostaglandin material has provided valuable quantities for research purposes.A species of Caribbean coral, Plexauru homomalla, was found to contain relatively vast amounts of prostaglandins, the major component being either the 15-acetate methyl ester of PGAz (8) or its 15-epimer (9) depending, apparently, on the exact location of the coral c~lonies.~ The prostaglandin content can be as much as 3 % of the dry weight of the coral and commercial exploitation of this source is being investigated. The reason why the coral should produce such vast amounts of prostaglandin is still unknown but the enzyme system responsible will only function in solution of marine-like ionic strength.-co 2 M e I bCOMe OCOMe In 1964 the first papers on prostaglandin biosynthesis appeared6 and sub- sequently Hamberg and Samuelsson7 proposed a radical mechanism leading to the endoperoxide intermediate PGG1. Knowledge of PG involvement in disease processes can be considered to date from 1971 when Vane8 and others showed that aspirin-like drugs inhibit PG synthesis and Vane suggested that this was the biochemical mechanism of their anti-inflammatory activity. As well as PG involvement as mediators in a wide range of very diverse disease processes there may also be disorders which are very specifically attributable to defects in PG metabolism. The Karolinska workers9 have identified a blood-clotting disorder which is caused by a deficiency of blood platelet cyclo-oxygenase, one of the PG biosynthesizing enzymes.A D. I. Weisblat, Drug Comet. Ind., 1972, 111(1), 34. W. P. Schneider, R. D. Hamilton, and L. E. Rhuland, J. Amer. Chem. SOC., 1972,94,2122. a D. A. van Dorp, R. K. Beerthius, D. H. Nugteren, and H. Vonkeman, Biochim. Biophys. Acta, 1964, 90,204; S. Bergstrom, H. Danielsson, and B. Samuelsson, Biochim. Biophys. Am, 1964, 90, 207. M. Hamberg and B. Samuelsson, J. Bid Chem., 1967, 242, 5336. a J. R Vane, Nature New Biol., 1971,231,232; J. B. Smith and A. L. Willis, Nature New Biol., * C. 1971,231, 235. Malmsten, M.Hamberg, J. Svensson, and B. Samuelsson, Proc. Nat. Acad. Sci., U.S.A., 1975,72, 1446. Gibson disorder known as Bartter’s syndrome characterized by hypokalaemia, hyper- reninaemia, hyperaldosteronism, juxtaglomerular hyperplasia, normotension with resistance to the pressor effects of angiotensin I1 seem to be caused by a primary lesion which is the overproduction of prostaglandin by the kidney.10 Evidence is now emerging of PG involvement in many normal physiological processes.The most widely known example is the natural luteolytic effect of PGF2, in certain species11 but effects on the circulation, on respiration, neuro- transmission, immune response, temperature control, pain receptors, etc., are now being reported and evaluated. 1here is also a certain amount of speculation12 as to how they function and what role they take on, e.g. are they local hormones or bioregulators ? Regardless of where PG’s occur, the biosynthetic route appears to be basically the same in all mammalian tissues.Thus, for example, PGEz (3) is derived from the endoperoxide intermediate PGGz (2) which in turn is derived from arachi- donic acid (1). Although it is possible that the enzymes of different tissues may be different, or present in different concentrations, and their co-factor requirements may be different, thereby giving rise to the differing proportions of the various prostanoid products which are characteristic of any one particular tissue, the biosynthetic chain of events leading to prostaglandins provides us with a common theme. Furthermore, although we are here concerned principally with metabolites of arachidonic acid, the biosyntheses of prostaglandins of the PG1 and PG3 series have a number of close analogies with the corresponding conversions in the PG2 series.The former are derived from dihomoy-linolenic acid (10) and the latter from 5,8,11,14,17-eicosapentaenoicacid (11). (10) Dihomo-y-linolenic acid (1 1) 5,8, I I, 14,17-Eicosapentaenoic acid 3 Availability of Arachidonic Acid The major source of arachidonic acid is membrane phospholipid, principally lecithin (Figure 3). It is claimed, for example, that of the arachidonic acid released from platelets stimulated by thrombin, 70 % derivatives from phos- phatidyl choline (lecithin), 25 % from phosphatidyl inositol, and 5 % from phosphatidyl serine.l3 There does seem to be some possibility that in special cases some arachidonic acid may derive from triacyl glycerols (adipose tissue) or cholesterol esters (adrenal and ovarian tissue).lo L. Norby, R.Leutz, W. Flamenbaum, and P. Ramwell, Lancet, 1976,11,604. l1 E. W. Horton and N. L. Poyser, Physiol. Rev.,1976, 56, 595. l2 E. W. Horton, J. Pharm. Pharrnacol., 1976, 28, 389. l3 J. B. Smith and M. J. Silver, Biochim. Biophys. Acra, 1976,424, 303. Prostaglandins etc. :Biosynthetic Products from Arachidonic Acid 0-+ NMe, -Phospholipase A, (I) Arachidonic Acid + R Lysolecithin R = saturated fatty acid side chain Figure 3 Phospholipase Aa action The enzyme responsible for cleaving the arachidonic acid from the 2-position of lecithin is phospholipase A2. Many different phospholipase A2 enzymes have been obtained from various sources and quite a lot is known about them.14 Roughly they fall into two types.Type I is membrane bound, is stimulated by Caz+, has optimum pH 7-8, and its activity is very closely related to membrane stability. Type I1 phospholipase A2 enzymes are soluble enzymes inhibited by Ca2+ and with pH optimum in the range 4-6. These enzymes are found in lysosomes in adrenals, spleen, macrophages and similar enzymes are also found in the venoms of bees, wasps, snakes, scorpions, and gila monsters. The active form of soluble (Type 11) phospholipase A2 is formed from an inactive pre- phospholipase A2 and the porcine pancreatic prephospholipase Az has been obtained in a crystalline form and its structure determined by X-ray crystallo- graphic ana1y~is.l~ It has 132 amino-acids and can be activated by trypsin which removes the first seven amino-acids to give the active enzyme. A possible mechan- ism of action is proposed in which the enzyme binds the lecithin by the phosphate to a Caz+ atom and Arg-108 and the fatty acid side-chains sit on a lipophilic area of the enzyme.The reacting parts of the enzyme are probably the carboxylate of Asp-56 which functions as the nucleophile, His-55 which stabilizes the tetrahedral intermediate, and Tyr-35 which provides a proton for the lysolecithin. 14 W. E. M. Lands and L. H. Rome, in ‘Advances in Prostaglandin Research. Prostaglandins: Chemical and Biochemical Aspects’, ed. S. M. M. Karim, M.T.P. Press, Lancaster, 1975, p.87. Is J. Drenth, C. M. Enzing, K. H. Kalk, and J. C. A. Vessies, Nature, 1976, 264, 373. Gibson It is still necessary, however, to explain why mammalian phospholipase A2 is unable to attack lecithin within the tightly packed structure of the native mem- brane, although it can attack lecithin in the form of monomeric units or in loosely packed micelles. In this respect it is interesting to note that phospholipase A2 from snake venom can attack lecithin in native membrane to release fatty acid and lysolecithin (Figure 3) which is a very potent haemolytic agent and will destroy both red and white blood cells.16 Some of the toxic effects of venom may be the result of haemolysis caused by the lysolecithin released by phospholipase A2 activity.However, the arachidonic acid released also warrants consideration as some of the toxic effects of venom17 are not unlike effects which are attribut- able to arachidonic acid metabolites. (An intravenous injection of only 1.4mg/kg body wt of arachidonic acid is lethal in the rabbit,l8 the cause of death being intravascular platelet aggregation leading to blockage of pulmonary arteries). The mammalian phospholipase A2 may need, however, some assistance before it can release arachidonic acid from lecithin within the membrane and this may be where other activating factors are implicated. Vane19 has reported on an RCS-releasing factor (Rabbit Aorta Contracting Substance [RCS] activity is mainly due to thromboxane A2-vide infra) and RCS-RF is probably a peptide of less than 10 amino-acid units.It is released from sensitized guinea pig lung upon immunological challenge and causes the release of arachidonic acid from lung tissue. Jt is apparently not stored in the lung tissue but is elaborated in response to the immunological stimulus. Somehow it is involved with the stimulation of phospholipase A2 activity but whether this is stimulation of active phospholipase A2 from prephospholipase A2, or action as a cofactor, or is in some way related to making phospholipid in the membrane more available for phospholipase A2 action remains to be determined. Interest- ingly, the action of RCS-RF is inhibited by the anti-inflammatory steroids with relative potency very similar to their relative anti-inflammatory action but only in intact cells and the effect of steroids is not demonstrable on cell fragments.The naturally occurring vasodilator peptide, bradykinin, will also release ara- chidonic acid from lung tissue but this action is not sensitive to inhibition by the anti-inflammatory steroids. Slow reacting substance of anaphylaxis (SRS-A), another compound released from lung tissue during immunological challenge, is apparently similar to, but not identical with, RCS-RF. SRS-A also has the ability to release prostanoid material from lung tissue and furthermore the release of SRS-A is apparently subject to negative feedback control by product prostaglandin because blockage of the formation of prostaglandin causes potentiation of the release of SRS-A.*20 * Note added in proof: evidence has been recently presented (M.K. Bach, J. R. Brashler, and R. R. Gorman, Prostaglandin,, 1977, 14, 21) that SRS-A is itself a biosynthetic product from arachidonic acid. l6 A. T. Tu, R. B. Passey, P. M. Toom, Arch. Biochem. Biophys., 1970, 140,96. P. R. Larson and J. Wolff, Biochem. Pharmacol., 1968, 17, 508. M. J. Silver, W. Hoch, J. J. KOCS~S,C. M. Ingerman, and J. B. Smith, Science, 1974, 183, 1085. l9 F. P. Nijkamp, R. J. Flower, S. Moncada, and J. R. Vane, Nature, 1976, 263, 479. ao D. M. Engineer, P. J. Piper, and P. Sirois, Brit. J. Pharmacol., 1976, 57,460P. Prostaglandins etc.: Biosynthetic Products from Arachidonic Acid Clearly the availability of the precursor arachidonic acid is subject to com- plicated biological control mechanisms.2l It may also be the locus from which hypersensitivity conditions (e.g.certain types of asthma) arise, a thesis supported by the apparent relationship between the activity of anti-inflammatory steroids and their effects on arachidonic acid release.4 Prostaglandin Synthetase A. Cyclo-oxygenase.-The conversion of arachidonic acid into prostaglandins is referred to loosely as a process brought about by PG synthetase (PGS) although it is clearly a complicated process involving a number of enzymes. In 1967 Hamberg and Samuelsson7 proposed a radical mechanism (Figure 4) leading to an endoperoxide intermediate and the appropriate labelling studies have shown that it is specifically the pro4 proton at C-13 (labelled as 3H in Figure 4) which is removed and studies22 with isotopically-labelled oxygen have shown that one molecule of oxygen gives rise to both oxygen functions at C-9 I 0,H PGH, PGG, Figure 4 Cyclo-oxygenaseactivity and C-11.Subsequently, both endoperoxide intermediates PGGz and PGH2 have been isolated.23 Although the PG synthetase activity is membrane bound, it can be solubilized and resolved into two fractions.24 Fraction I, referred to as cyclo- oxygenase (it also contains peroxidase activity) because it leads to the formation of the cyclic endoperoxides PGG2 and PGH2, has the following properties: it uses haeme as a cofactor; the peroxidase activity requires tryptophan (or sero-R.J. Flower and G. J. Blackwell, Biochem. Pharmacol., 1976, 25, 285. 22 B. Samuelsson, J. Amer. Chem. SOC.,1965, 87, 301 1. 23 D. H. Nugteren and E. Hazelhof, Biochim. Biophys. Acta, 1973, 326, 448; M. Hamberg, and B. Samuelsson, Proc. Nut. Acad. Sci., U.S.A., 1973,70,899 ; M. Hamberg, J. Svensson, T. Wakabayashi, and B. Samuelsson, Proc. Nat. Acad. Sci., U.S.A., 1974, 71, 345. 84 T. Miyamoto, N. Ogino, S. Yamamoto, and 0.Hayaishi, J. Biof. Chem., 1976, 251, 2629; ibid., 1977, 252, 890. Gibson tonin, or epinephrine, i.e. a hydrogen donor); it is irreversibly inhibited by peroxides; it is reversibly inhibited by glutathione peroxidase; the action of the enzyme appears to catalyse its own destruction and in this way it is similar to fatty acid dioxygenase, e.g.from soybeans; there is some need for an ‘essential activator’ which is produced during the oxygenation, i.e. some form of positive feedback;25 it is inhibited by non-steroidal anti-inflammatory agents and also by poly-ynoic acids such as eicosatetraynoic acid. Fraction 11, the isomerase part of the PG synthetase complex, causes the isomerization of the endoperoxide moiety of PGGz or PGH2 to the p-hydroxyketone moiety present in PGEz (vide infva). This fraction is fairly unstable (t+ N 30 min at ambient temperature) although it can be stabilized by thiol groups and it actually uses glutathione as a cofactor.24 The cyclo-oxygenase part of the biosynthetic pathway has been investigated in some detail but no complete definitive picture has yet emerged.Lands and Rome14 have proposed detailed kinetics to fit the observed data and O’BrierP has proposed that peroxidase activity in the synthase gives rise to hydroperoxides and then eventually the peroxy radical intermediate of the Hamberg and Samuelsson mechanism. O’Brien2? has also proposed that singlet oxygen (as a metal complex) is involved as the initiator of the synthase. Singlet oxygen may also be responsible for the PGS inactivation because bilirubin, a singlet oxygen scavenger, inhibits the synthase reaction and also protects the synthase both from inactivation by peroxides and from self-catalysed destruction. These ideas are combined schematically in Figure 5. 0 RH F i d T ROHo, PGS -0,’-RodH donor RH R--RO,.-cyclization Bilirubin-RO,H R HX PGS deactivation R = fatty acid substrate Figure 5 Possible involvement of singlet oxygen and peroxidase mechanism for prostaglandin biosynthesis Peroxidation at C-11 is not the only pathway by which arachidonic acid is metabolized. For example, in blood platelets a large proportion or arachidonic acid is metabolized by an alternative pathway28 which gives rise to 12-hydro-peroxyeicosatetraenoic acid (HPETE) and 12-hydroxyeicosatetraenoic acid zs W. L. Smith and W. E. M. Lands, Biochemistry, 1972, 11, 3276. aa P. J. O’Brien and A. Rahimtula, Biochem. Biophys. Res. Comm., 1976,70, 832. 27 A. Rahimtula and P. J. O’Brien, Biochem. Biophys. Res. Comm., 1976, 70, 893. M. Hamberg and B.Samuelsson, Proc. Nat. Acad. Sci., U.S.A., 1974, 71, 3400. Prostaglandins etc. :Biosynthetic Products from Arachidonic Acid (HETE) (Figure 6). The conversion into 15-hydroperoxyarachidonic acid (15-HPAA)which is a potent inhibitor of the formation of prostacyclin may be of great importance to the vasculature (vide infra). H ,O HPETE CO,Hcd", HO-O2H HETE 15-HPAA Figure 6 Alternative metabolic oxidations of arachidonic acid B. Cyclo-oxygenase Inhibitors.-As mentioned earlier, in 1971 Vane, and Smith and Willis,8 showed that aspirin-like drugs inhibit PG synthesis and Vane suggested that this is probably the biochemical mechanism by which they exert their anti-inflammatory activity. In the PG synthase pathway it is the cyclo- oxygenase step which is most sensitive to inhibition by this type of anti- inflammatory agent and a brief consideration of cyclo-oxygenase inhibitors may be worthwhile (Figure 7).14 In the case of aspirin itself there is evidence to show that some of the inhibition is caused by irreversible acetylation of the cyclo-oxygenase enzyme; radio- labelling studies support this.With indomethacin, although an analogous p-chlorobenzoylation is mechanistically feasible, the available evidence29 does not support this and indomethacin must be included within a large number of acidic anti-inflammatory compounds of salicylic acid types (e.g. diflufenisal), the fenamates (e.g.flufenamic acid), and the acetic acid derivatives (e.g.sulindac), where one can merely indicate a vague similarity of size and shape between the inhibitor and arachidonic acid or the hydroperoxy radical intermediate (Figure 4) and thereby suggest some interaction which can lead to inhibition of the cyclo- oxygenase.There may be some ability within the non-steroidal anti-inflammatory agents to react with radicals and a mechanism has been proposed along these 29 N. Stanford, G. J. Roth, T. Y. Shen, and P. W. Majerus, Prostaglandins, 1977, 13, 669. Gibson C0,H -Aspirin: + Enz.NH, Enz-NH-C0l4CH, (+SA) \ FWC0,H F OH Diflufenisal Indoinethacin CO,H CF,H Flufenamic acid Me -1 0 Sulindac HO OH 2,7-Dihydroxynaphthalene 0:J ETYA Reaction with enzyme Glutathione peroxidase _I) Removes radicals? Figure 7 Cyclo-oxygenaseinhibitors 499 Prostaglandins etc.:Biosynthetic Products from Arachidonic Acid lines.30 2,7-Dihydroxynaphthaleneis an extremely potent inhibitor of cyclo- oxygenase.31 Acetylenic acids, e.g. eicosatetraynoic acid (ETYA), will inhibit the cyclo- oxygenase a~tivity.3~ These acetylenic acid inhibitors may be acting as kcat inhibitors33 whereby the enzyme accepts them as a substrate and converts them into a highly reactive allene (Figure 7). This allene is formed at the active site of the enzyme and rapidly reacts with some nearby function and thereby renders the enzyme inactive. This is a mechanism well characterized for a number of other enzyme inhibitors. The enzyme is also inhibited by glutathione peroxidase, possibly by lowering the concentration of activator (radical ?) and recent evidence indicates that this may be a natural inhibitor of PG synthesis.34 C.Reactions of PGGz.-From PGGz there are two steps required in order to obtain PGE2 or PGFza, i.e. a peroxidase step to convert the 15-hydroperoxy into a 15-hydroxy function and an isomerase step to isomerize the endoperoxide moiety to a ,&hydroxyketone, or alternatively a reductase step to directly cleave the endoperoxide to a 1,3-diol (Figure 8). In the 8000 x g supernatant from sheep seminal vesicles and in the absence of added glutathione the major products are 15-hydroperoxy-Ez and 15-keto-EZ and this is part of the evidence which leads Samuelsson35 to suggest that the isomerase followed by the peroxidase is prob- ably the preferred pathway in sheep seminal vesicles; both pathways may be operative.24 With the 8000 x g sheep vesicular gland supernatant system and added glutathione PGFza is the major product.Mechanistically, the isomerization of PGHz to PGEz requires only the loss of the proton from C-9 and the opening of the endoperoxide bridge as shown by the arrows. The analogous isomerization involving loss of the C-11 proton and the opening of the endoperoxide bridge in the alternative manner leads to PGD2 (Figure 8) which is a very potent inhibitor of aggregation of human platelets.36 An enzyme which converts PGG2 into PGDz is present in human serum. Not only can PGFza be formed by direct reduction of PGGz but it can also be formed by the reduction of PGEz by the action of an enzyme 9-keto-reductase and the widespread importance of this enzyme is now being rec~gnized.~~-~~ In certain situations PGEz and PGFza have antagonistic effects, e.g.PGEz is a vasodilator whereas PGFza is a vasoconstrictor. 9-Keto-reductase enzymes can Bo D. W. Cushman and H. S. Cheung, Biochim. Biophys. Acta., 1976, 424,449. 31 C. Takeguchi and C. 5. Sih, Prostaglandins, 1972, 2, 169. 32 D. T. Downing, D. G. Ahern, and M. Bachta, Biochem. Biophys. Res. Comm., 1970, 40, 21 8. 33 R. R. Rando, Science, 1974, 185, 320. 34 H. W. Cook and W. E. Lands, Nature, 1976, 260, 630. M. Hamberg, J. Svensson, and B. Samuelsson, in ‘Advances in Prostaglandin and Throm- boxane Research’, eds.B. Samuelsson and R. Paoletti, Raven Press, New York, 1976, vol. 1, p. 19; B. Samuelsson, ibid., p. 1. 30 J. B. Smith, M. J. Silver, C. M. Ingerman, and J. C. Kocsis, Thromb. Res., 1974, 5, 291. 37 P. Y-K. Wong, D. A. Terragno, N. A. Terragno, and J. C. McGiff, Prostaglandins, 1977,13, 1113. 38 K. J. Stone and M. Hart, Prostaglandins, 1976, 12, 197. 39 Y. Friedman, S. Levasseur, and G. Burke, Biochim. Biophys. Acta, 1976, 431, 615. Gibson PG D, cndoperoxide reductase M DA HHT 9-keto-reductaseg.. J QH CO,H2dH AH PGF,, Figure 8 Reactions of PGGz Prostaglandins etc.:Biosynthetic Products from Arachidonic Acid be obtained from arterial and veinous tissues and both enzymes are activated by cyclic-GMP.However, the enzyme from veinous tissue is much more sensitive to cyclic-GMP stimulation than is the enzyme from arterial tissue. Kinins also stimulate the 9-keto-reductase and as they also stimulate the availability of arachidonic acid so they have tremendous potential to affect the amount and the ratio of the E and F prostaglandin^.^^ The 9-keto-reductase from kidney is reported to be inhibited by non-thiazide diuretics (furosemide, chlorthalidone) ; other inhibitors of this enzyme are hydralazine (vasodilator), phentolamine (a-blocker), indomethacin (anti-inflammatory) and ethacrynic acid (di~retic).~~ It seems Iikely that the control of the PGE :PGF ratio has physiological import- ance and it is possible that some pharmacological effects are manifestations of the ability of certain drugs to cause variations in this ratio.PGGz can also decompose to malondialdehyde (MDA) and hydroxyhepta- decatrienoic acid (HHT, Figure 8). The formation of MDA has been widely used as a quantitative measure of the formation of PGG2 but the reaction may simply be a thermal decomposition and biologically it is of dubious significance. 5 Thromboxanes The biosynthetic products formed from endogenous arachidonic acid, upon treatment of washed human platelets with thombin are shown in Figure 9.3a Washed thrombin HETE (1.2--2.2 1J.g ml-I)human -. -platelets HHT (1.1 -2.4 p.6 111Ik') ?" (02H HO I bH TXB, (I -2.3 pg nil 1) Figure 9 Arachidonic acid metabolites from human platelets Only very small amounts of PGE2 and PGFza are formed in this reaction.Some of the arachidonic acid is converted viu hydroperoxyeicosatetraenoic acid (HPETE) into hydroxyeicosatetraenoic acid (HETE ; Figure 6). Hydroxyhepta-decatrienoic acid (HHT) is formed by retro-Diels-Alder reaction which splits off malondialdehyde (MDA) from the endoperoxide (Figure S), and a similar amount of material is found as the novel compound, originally referred to as PHD, now known as thromboxane B2 (TXBs), in which the cyclopentane ring has been converted into a tetrahydropyranyl ring. It was reasoned that TXBz was derived by rearrangement of PGG2 (or PGHr) with subsequent addition of a molecule of water and in accord with this hypothesis it was found that the inter- Gibson mediate could be trapped by added nucleophiles such as MeOH, EtOH, or N3-(Figure 10).These products were identified and this led to the proposal that the intermediate was an oxetane and it was named thromboxane A2 (TXA2) because of its extreme potency in inducing platelet aggregation. PGG, OH WCO?H OH X = OMe, OEt, or N, AHkjb--j -1 -? L J Figure 10 Formation of thromboxane Aa Thromboxane A2 can be obtained by very brief incubations of arachidonic acid with human platelets or by incubation of preformed PGGz with horse platelet microsomes. It is extremely labile and hydrolysis to form TXBz proceeds with a half-life of about 36 s in water at 37 "C.TXA2 is a highly potent stimulant of the contraction of certain smooth muscle preparations, e.g.it is many times more potent than PGFza in the stimulation of guinea-pig trachea and in its effect on guinea pig lung insufflation pressure. The effect of TXA2 on certain vascular tissue was known long before its structural identification by Samuelsson. As 503 Prostaglandins etc. :Biosynthetic Products from Arachidonic Acid early as 1969 Piper and Vane40 had described the release of RCS from guinea pig lung during anaphylaxis. Vane subsequently showed that aspirin-like drugs inhibit PG synthesis and also inhibit the release of RCS and it was suggested that RCS might be an endoperoxide intermediate.41 However, when the pure endo- peroxides became available, Samuelsson showed that although they were active on rabbit aorta they were too long lived and of insufficient potency to be identical with RCS.It now transpires that RCS is in fact a mixture of TXA2 together with some endoperoxides, but the majority of the biological activity is due to the TXAP Thromboxanes have now been found in platlets, lung, leucocytes, umbilical artery, spleen, brain, kidney, and seminal vesicles; indeed, in some of these tissues this is the major route of arachidonic acid metabolism.42 Their widespread occurrence is the subject of much speculation regarding their physio- logical importance, particularly that of TXA2 in cardiovascular incidents, because of its ability to aggregate platelets and to cause vasoconstriction. TXAz pro- duction by anaphylactically shocked lung tissue is highly suggestive of a possible involvement in lung dysfunction.The search for selective inhibitors of thom- boxane synthase (the enzyme responsible for the production of TXAz from endoperoxide) is underway and the effects of inhibitors of this enzyme are being evaluated.43 It has been reported44 that benzydamine (12) is a more selective inhibitor of thomboxane synthase (inhibitory at 100 pg ml-l) than of cyclo-oxygenase (inhibitory at 250 pg ml-1). By comparison indomethacin inhibits cyclo-oxygenase at 5 pg ml-l and thomboxane synthase at 100 pg ml-l. (12) Benzydamine TXBZis reported to be chemotactic for leucocytes and this may be significant to inflammatory processes.45 Mechanistically, the conversion of the endoperoxide to TXA2 can be considered as a polarization of the peroxide bridge,46 as shown in brackets in Figure 10, leading to a carbon to oxygen migration with the formation of a secondary 40 P.J. Piper and J. R. Vane, Nature, 1969, 223,29. 41 R. Gryglewski and J. R. Vane, Brit. J. Pharmacol., 1971, 43,420. 4z C. R. Pace-Asciak, Prostaglandins, 1977, 13, 811. 43 Scrip, 1976, 237, 21. 44 S. Moncada, P. Needleman, S. Bunting, and J. R. Vane, Prostaglandins, 1976, 12, 323. 46 J. R.Boot, W. Dawson, and E. A. Kitchen, J. Physiol. (London), 1976, 257, 47P. 46 Discussion by G. Fried and G. Just, in Prostaglandins, 1976, 11, 431. Gibson carbonium ion (stabilized as the oxonium ion canonical form) which then collapses by formation of the oxetane ring to give TXA2.Due to the relatively weak oxygen-oxygen single bond, the endoperoxide to oxetane transformation is energetically quite favourable because a simple summation of the bond energies gives an enthalpy change of about -38 kcals mole-2. It is interesting to note that in the 1-series, from dihomoy-linolenic acid, the corresponding conversion of PGH1 into TXAl does not occur in platelet microsomes.47 6 Prostacyclin (PGX or PGI2) In 1971 Pace-Asciak48 reported that incubation of arachidonic acid with rat stomach homogenate gave the d7-6(9)-oxy-PGFla (13) shown in Figure 11. When the incubation was performed with 5,6,8,9,11,12,14,15-octadeutero-arachidonic acid only six of the dueterium atoms were retained in thed7-product which could infer that the d5-compound (14) is an intermediate to the d7-compound.During 1975 and 1976 6-keto-PGFla, which can also exist in the lactol form (Figure 1 l), was reported from several sources: Pace-A~ciak~~ showed it to be present in the products from incubation of arachidonic acid with homogenate of rat stomach fundus; Poyser50 claimed that it was the major prostaglandin present in the uterus of the five day pseudopregnant rat and it was also claimed to be present in the sheep uterus; Dawson51 showed it to be released from guinea pig lung during anaphylaxis; and it was also shown to be produced by bovine seminal ve~icles5~ and microsomes of human and rabbit kidney c0rtex.5~ In late 1976 Vane and co-workers54 described an unidentified unstable substance, PGX, formed by incubation of PGG2 or PGH2 with microsomes from rabbit or pig aorta.PGX, unlike TXA2, does not contract rabbit aorta; also unlike TXA2, it relaxes rabbit mesenteric and coeliac artery. PGX will contract rat stomach strip, chick rectum, guinea pig ileum and trachea but with very much reduced potency relative to PGG2 or PGH2. But the most remarkable property of PGX is its potent ability to inhibit platelet aggregation55 and even to bring about the reversal of platelet aggregation.56 Collaboration between Vane’s and Upjohn’s workers57 resulted in the determination of the structure of PGX (14). This is, of course, the putative intermediate to Pace-Asciak’s d7-6(9)-oxy-PGF~a(13). It is P. Needleman, M.Minkes, and A. Raz, Science, 1976, 193, 163. 48 C. Pace-Asciak and L. S. Wolfe, Biochemistry, 1971, 10, 3657. 48 C. Pace-Asciak, Biochim. Biophys. Acta, 1976, 424, 323. 50 R. L. Jones, N. L. Poyser, and N. H. Wilson, Brit. J. Pharmacol., 1977, 59,436P. b1 W. Dawson, J. R. Boot, A. F. Cockerill, D. N. B. Mallen, and D. J. Osborne, Nature, 1976, 262,699. 52 W. C. Chang and S. Murota, Biochim. Biophys. Acta, 1977, 486, 136. 53 A. R.Whorton, M. Smigel, J. A. Oates, and J. C. Frolich, Prostaglandins, 1977, 13, 1021. 54 S. Moncada, R. J. Gryglewski, S. Bunting, and J. R. Vane, Nature, 1976, 263, 663. 55 R. J. Gryglewski, S. Bunting, S. Moncada, R. J. Flower, and J. R. Vane, Prostaglandins,1976, 12, 685. 56 S. Moncada, R. J. Gryglewski, S. Bunting, and 5.R. Vane, Prostaglandins, 1976, 12, 715. 57 R. A. Johnson, D. R. Morton, J. H. Kinner, R. R. Gorman, J. R. McGuire, F. F. Sun, N. Whittaker, S. Bunting, J. A. Salmon, S. Moncada, and J. R. Vane, Prostaglandins,1976, 12, 915. 505 Prostaglandins etc. :Biosynthetic Products from Arachidonic Acid (14) PGI, f \ '\\ OH I , ;OH OH 6H I OH 6H Figure 11 Formation and hydrolysis of PG?, Gibson also an enol ether obviously readily hydrolysed to 6-keto-PGFia (Figure 11) and quite probably in the situations in which 6-keto-Fla has already been found, the true active principle may be PGX. With the structure established, PGX was renamed prostacyclin and then given the notation PGI2 in the alphabetically-based semi-systematic nomenclature.The formation of prostacyclin is the major metabolic pathway in vascular tissue in man and other species58 and is no less a subject for widespread specu- lation59 than is the formation of TXA2. In the vasculature we have the remarkable possibility that the common intermediate endoperoxides, produced by the plate- lets, are converted by the platelets into the potent vasoconstricting and platelet aggregating TXA2, but that the vessel walls convert the endoperoxides into the potent vasodilating and platelet aggregation inhibiting prostacyclin. Thus the balance of the antagonistic properties of TXA2 and PGI2 may maintain blood vessel tone, platelet functionality and thereby be critical for thombi formation. Prostacyclin is the major prostaglandin produced by the isolated perfused heart of rabbit or rat, apparently without the necessity for preformed endoperoxide from platelets, and it may function to protect the coronary circulation against the formation of blood platelet aggregates which could occlude the vulnerable coronary microvasculature.60 Similarly, the generation of the PGI2 by human placenta may be important in the maintenance of placental circulation;F1 pulmonary circulation could be influenced by lung prostacyclin production; implantation of the fertilized ovum into the uterus may require protection from platelet aggregation ; possible involvement in bone resorption has been suggested,G2 and so has the protection of the stomach against ulcer f0rmati0n.j~ The conversion of PGG2 to PGI2 is brought about by the enzyme prostacyclin- synthetase and this enzyme is inhibited by 15-hydroperoxyarachidonicacid (1 5-HPAA; Figure 6) at a very low concentration.55 The possible implication of this to diseases of the cardiovascular system has been speculated upon by Vane. Regarding the mechanism of the formation of PGI2, polarization of the endo- peroxide in the opposite sense to that required for TXA2 formation could lead to participation of the 5,6-double bond thus leading to the secondary carbonium ion (15) which by loss of the proton from position-6 gives PGl2.46 The isomeriz- ation of PGI2 to d7-6(9)-oxy-PGFla is somewhat baffling at present, both from the mechanistic point of view and also from the point of view of the biological reason for such an isomerization.7 Prostaglandin Metabolism Transformations of PGE2 into PGC2 and PGB2 and the 19-hydroxy compounds63 68 G. J. Dusting, S. Moncada, and J. R. Vane, Prostaglandins, 1977, 13, 3. 59 J. L. Mam, Science, 1977, 196, 1072. 0o E. A. M. de Deckere, D. H. Nugteren, and F. Ten Hoor, Nature, 1977, 269, 160. 61 L. Myatt and M. G. Elder, Nature, 1977, 268, 159. 82 L. G. Raisz, J. W. Dietrich, H. A. Simmons, H. W. Seyberth, W. Hubbard, and J. A. Oates, Nature, 1977, 267, 532. G3 P. L. Taylor and R. W. Kelly, Nature, 1974,250,665; R. W. Kelly, P. L. Taylor, J. P. Hearn, R. V. Short, D. E. Martin, and J. M.Marston, Nafurc, 1976, 260, 544. Prostaglandins etc. :Biosynthetic Products from Arachidonic Acid PGEz PGA2 I nu 6H I9-HO-PGEz PGC2 1 CO,HIOH PG B, Figure 12 Other PGE, transformations are conceptually simple reactions (Figure 12). However, in vivo these transfor- mations are probably enzymically controlled and therefore of importance either as giving products with a different range of activities or as deactivation mechan- isms.As mentioned in the previous review, prostaglandins are very rapidly metabolized by degradative enzymes. The primary deactivation step is the oxidation of the 15-hydroxy function by the enzyme prostaglandin-1 5-dehydro- genase which is particularly active in the lungs: greater than 95% of PGFsar present in the blood stream is deactivated by this enzyme on one passage through the lungs.64 Subsequent stages of metabolism involve reduction of the 13,14-double bond, two sequences of p-oxidation of the acid (top) side chain, and w-oxidationof the alkyl side chain leading eventually to the major human urinary metabolite (1 6) which thus represents the end product of arachidonic acid 61 B.Samuelsson, Federation Proc., 1972, 31, 1442. Gibson metabolism via this pathway.65 PGE2 undergoes similar metabolism but with a keto function at C-9instead of the hydroxy function of PGFw In monkeys the major urinary metabolite of TXB2 is reported to be the product (17) from one sequence of P-oxidation.G6 ?H ?H -COZH G C 0 2 H I OH 0 bH 8 Concluding Remarks The wide variety of physiological activities associated with the compounds within this field offers potential therapeutic use in the many aleas which have been mentioned.The very short biological half-lives of natural PG’s are a severe restriction on their mode of administration and modified prostaglandins have been produced which are much less susceptible to the degradation processes. An obvious modification such as a 15-methyl group, as in 15-methyl-PGFza methyl ester (18) renders the molecule no longer susceptible to the prostaglandin-15- dehydrogenase enzyme, and this compound (18) has been used extensively for the induction of abortion in women. Synthetic analogues can also offer greater selectivity in their actions, for example cloprostenol (19) is a highly selective luteolytic agent on sale for the induction of luteolysis in .I OH 6H dH OH (18) (19) Cloprostenol The recent identification of TXA2 and PGI2 means that biological processes which have previously been considered only in terms involving the classical B. Samuelsson, E. Granstrom, K. Grkn, M. Hamberg, and S. Mammarstrom, Ann. Rev. Biochem., 1975,44,669. e6 H. Kindahl, Prostaglandins, 1977, 13, 619; L. J. Roberts, B. J. Sweetman, J. L. Morgan,N. A. Payne, and J. A. Oates, Prostaglandins, 1977, 13, 631. 13’ M. J. Cooper and A. L. Walpole, in ‘Advances in Prostaglandin Research. Prostaglandins and Reproduction’, ed. S. M. M. Karim, M.T.P. Press, Lancaster, 1975, p. 309. Prostaglandins etc. :Biosynthetic Products from Arachidonic Acid prostaglandin types must now be re-evaluated in the light of the possible involvement of these potent but labile molecules.The unstable nature of TXA2 and PGI2 is a strong incentive for the production of stable analogues as potential therapeutic agents. Alternatively, selective inhibitors of thomboxane synthase or prostacyclin synthase can also be envisaged as potentially useful drugs. The involvement of TXA2 and PGIa in the functioning of the vasculature and their effects on blood platelets opens up a whole new area of great interest in the field of cardiovascular disease. However, their probable involvement in pulmonary and reproductive tissues and their possible involvement in many other tissues will require a high order of selectivity to be achieved for analogues to become clinically successful.It is a fact that the impact of the prostaglandins in human clinical practice is, as yet, meagre considering the vast amount of effort which has been expended in the area. This failure is largely due to the lack of sufficient selectivity in those analogues which have undergone clinical trials.

ISSN:0306-0012    

DOI:10.1039/CS9770600489

出版商:RSC    

年代:1977

数据来源: RSC

摘要:

Corrigendum Vol 6 No 3 1977 “Chemicals which Control Plant Growth” by R. L. Wain. Plate 7 (in section between pp. 274 and 275): in the caption (2nd and 3rd lines) for “right” read “left”, and for “left” read “right”.

ISSN:0306-0012    

DOI:10.1039/CS9770600511

出版商:RSC    

年代:1977

数据来源: RSC

摘要:

INDEXES Volume 6, 1977 The indexes in this issue cover Volumes 1-6 (Figures in bold type refer to the volume number) Index INDEX OF AUTHORS Aarons, L. J., 5, 359 Ahluwalia, J. C., 2, 203 Allen, N. S., 4, 533 Baker, A. D., 1, 355 Bartlett, P. D., 5, 149 Beattie, I. R., 4, 107 Bell, R. P., 3, 513 Bentley, P. H., 2, 29 Berkoff, C. E., 3, 273 Bird, C. W., 3, 309 Blandamer, M. J., 4, 55 Blundell, T. L., 6, 139 Bradshaw, T. K., 6,43Braterman, P. S., 2, 271 Breslow, R., 1, 553 Brown, K. S., jun., 4, 263 Brundle, C. R., 1, 355 Buchanan, G. L., 3,41 Burdett, J. K., 3, 293 Burgess, J., 4, 55 Burrows, H. D., 3, 139 Cadogan, J. I. C., 3, 87 Carabine, M. D., 1, 411 Cardin, D. J., 2, 99 Carless, H.A. J., 1, 465 Cetinkaya, B., 2, 99 Chamberlain, J., 4, 569 Chatt, J., 1, 121 Chivers, T., 2, 233 Clark, G. M., 5, 269 Collins, C. J., 4, 251 Connor, J. N. L., 5, 125 Corfield, G. C., 1, 523 Cornforth, J. W., 2, 1 Cotton, F. A., 4, 27 Coulson, E. H., 1, 495 Coyle, J. D., 1, 465; 3, 329; 4, 523 Cragg, G. M. L., 6, 393 Cramer, R. D., 3, 273 Crammer, B., 6, 431 Cross, R. J., 2, 271 Dack, M. R. J., 4, 211 Dainton, F. S., 4, 323 Dobson, J. C., 5, 79 Doyle, M. J., 2, 99 Drummond, I., 2, 233 Eschenmoser, A., 5, 377 Evans, D. A., 2, 75 Fenby, D.,V., 3, 193 Fenton, D. E., 6, 325 Ferguson, L. N., 4, 289 Fisher, L. R., 6, 25 Flygare, W. H., 6, 109 Fry, A., 1, 163 Georghiou, P. E., 6, 83 Gibson, K.H., 6, 489 Goodings, E. P., 5, 95 Gray, B. F., 5, 359 Green, C. L., 2, 75 Greenhill, J. V., 6, 277 Greenwood, N. N., 3,23 1 Griffiths, J., 1, 481 Grossert, J. S., 1, 1 Groves, J. K., 1, 73 Guilford, H., 2, 249 Gutteridge, N. J. A., 1, 381 Haines, R. J., 4, 155 Hall, G. G., 2, 21 Hall, L. D., 4, 401 Hall, T. W., 5, 431 Halliwell, H. F., 3, 373 Harmony, M. D., 1, 211 Harris, K. R., 5, 215 Harris, R. K., 5, 1 Hartley, F. R., 2, 163 Hartshorn, S. R., 3, 167 Henderson, J. W., 2,397 Hepler, L. G., 3, 193 Hinchliffe, A., 5, 79 Horton, E. W., 4, 589 Hudson, M. F., 4, 363 Huntress, W. T., Jr., 6, 295 Hutchinson, D. W., 6, 43 Ikan, R., 6, 431 Isaacs, N.S., 5, 181 Isbell, H. S., 3, 1 Jaffk, H. H., 5, 165 Jamieson, A. M., 2, 325 Jenkins, J. A., 6, 139 Johnson, A. W., 4, 1 Johnson, S. P., 5, 441 Jotham, R. W., 2,457Kemp, T. J., 3,139 Kennedy, J. F., 2, 355 Kennewell, P. D., 4, 1S9 Kenny, A. W., 4, 90 Kirby, G. W., 6, 1 Koch, K. R., 6, 393 Kresge, A. J., 2,475 Lappert, M. F., 2, 99 Lee-Ruff, E., 6, 195 Leigh, G. J., 1, 121; 4, 155 Leznoff, C. C., 3, 65 Lindoy, L. F., 4, 421 Linford, R. G., 1, 445 Lipscomb, W. N., 1, 319 Lynch, J. M., 3, 309 McKellar, J. F., 4, 533 McKervey, M. A., 3, 479 Mackie, R. K., 3, 87 Maitland, G. C., 2, 181 Manning, P. G., 5, 233 Maret, A. R., 2, 325 Mason, R., 1, 431 Mayo, B. C., 2, 49 Meadowcroft, A.E., 4. 99 Menger, H. W., 2, 415 Midgley, D., 4, 549 Millen, D. J., 5, 253 Mills, R., 5, 215 Moore, H. W., 2, 415 Morley, R., 5, 269 Morris, J. H., 6, 173 Mulheirn, L. F., 1,259 Munn, A., 4, 87 Newman, J. F., 4, 77 North, A. M., 1,49 Oakenfull, D. G., 6, 25 Page, M. I., 2, 295 Perkins, P. G., 6, 173 Pletcher, D., 4, 471 Poliakoff, M., 3, 293 Pratt, A. C., 6, 63 Ramm, P. J., 1, 259 Rao, C. N. R., 5, 297 Rattee, I. D., 1, 145 Redl, G., 3,273 Richards, D. H., 6, 235 Ritch, J. B., jun., 5, 452 Roberts, M. W., 6, 373 Robinson, F. A., 5, 317 Roche, M., 5, 165 Rose, A. E. A., 6, 173 Rouvray, D. H., 3, 355 Sanders, J. K. M., 6,467 Index Sarma, T. S., 2, 203 Satchell, D.P. N., 4, 231 ;6, 345 Satchell, R. S., 4, 231 Senthilnathan, V. P., 5,297Simpson, T. J., 4, 497 Singh, S., 5, 297 Smith, E. B., 2, 181 Smith, K., 3, 443 Smith, K. M., 4, 363 Smith, W. E., 6, 173 Stacey, M., 2, 145 Suckling, C. J., 3, 387 Suckling, K. E., 3, 387 Sutherland, R. G., 1, 241 Sutton, D., 4, 443 Symons, M. C. R., 5,337Taylor, J. B., 4, 189 Theobald, D. W., 5, 203 Thomas, T. W., 1,99 Thompson, M., 1, 355 Tincknell, R. C., 5, 463 Toennies, J. P., 3, 407 Tolman, C. A., 1, 337 Truax, D. R., 5, 411 Twitchett, H. J., 3, 209 Underhill, A. E., 1,99 Wain, R. L., 6, 261 Walker, E. R. H., 5, 23 Walker, I. C., 3, 467 Waltz, W. L., 1, 241 Ward, I. M., 3, 231 White, A.J., 3, 17 Whitfield, R. C., 1, 27 Wieser, H., 5, 411 Wiesner, K., 6, 413 Yoffe, A. D., 5,51 Index INDEX OF TITLES Absorption bands in the spectra of stars, a crystal field approach,5,233 Across the living barrier, 6, 325 Acylation by ketens and isocyanates, a mechanistic comparison, 4, 231 --Friedel-Crafts, of alkenes 1, 73 Adamantane rearrangements, 3, 379 Affinity chromatography, chemical aspects of 3, 249 Alcohols and amines, conformational analysis of, 5,411 Alkali-metal complexes in aqueoussolution 4, 549 Alkaloids, aconite, synthesis of, 6,413 Alkenes, the Friedel-Crafts acylation of, 1,73 Aluminium phosphates, the chem-istry and binding properties of,6, 173 Amines and alcohols, conformational analysis of, 5,411 Aphids and scale insects, their chern- istry, 4,263 Application of electrochemical tech- niques to the study of homogeneous chemical reactions, 4,471 Aqueous mixtures, kinetics of re-actions in, 4, 55 -solution, micelles in, 6, 25 Aryldiazonium cations, co-ordination chemistry of, 4, 443 Atmosphere, interactions in, of drop- lets and gases, 1,411 Azidoquinones and related com-pounds, chemistry of, 2,415 Azobenzene and its derivatives, photo- chemistry of, 1,481 Bile pigments, 4, 363 Binding of heavy metals to proteins,6, 139 Binding properties and chemistry of aluminium phosphates, 6,173 Biomimetic chemistry, I, 553 Biosynthesis of sterols, 1,259 Biosynthetic products from arachi-donic acid, 6, 489 -studies, carbon-13 nuclear mag- netic resonance in, 4,497 Bredt’s rule, 3, 41 Brernsted relation-recent develop-ments, 2,475Butadiene, polymerization and co-polymerization of, 6, 235 Calciferols, hormonal : chemistry of ‘vitamin’ D 6, 83 Calorimetric investigations of hydro- gen bond and charge transfer complexes, 3, 193 Cancer and chemicals, 4, 289 Carbohydrate-protein complexes,glycoproteins, and proteoglycans,of human tissues, chemical aspects of 2,355 Carbon-1 3 nuclear magnetic resonance in biosynthetic studies 4,497Carbonium ions, carbanions, and radicals, chirality in, 2, 397 Carbonyl compounds, photochemistry of, 1,465Catalysis and surface chemistry, new perspectives 6, 373 -homogeneous, and organo-metallic chemistry, the 16 and 18 electron rule in, 1,337 -of the olefin metathesis reaction, 4, 155 CENTENARY LECTURE.Biomimetic chemistry, 1,553CENTENARY Light scattering LECTURE. in pure liquids and solutions, 6, 109 CENTENARY LECTURE. Quadruple bonds and other multiple metal to metal bonds. 4, 27 CENTENARYLECTURE. Rotationally and vibrationally inelastic scattering of molecules, 3, 407 CENTENARY LECTURE. Systematic development of strategy in the synthesis of polycyclic polysub-stituted natural products: the aco- nite alkaloids, 6, 413 CENTENARY LECTURE. Three-dimen-sional structures and chemical mech- anisms of enzymes, 1, 319 Charge transfer and hydrogen bond complexes, calorimetric investiga-tions of, 3, 193 Chemical applications of advances in Fourier transform spectroscopy,4, 569 -aspects of affinity chromatogra- PhY9 2,249 --of glycoproteins, proteo- glycans, and carbohydrate-protein complexes of human tissues, 2, 355 -interpretations of molecular wave functions 5, 79 Chemicals in rodent control, 1, 381 -which control plant growth,6, 261 Chemistry and binding properties of aluminium phosphates, 6, 173 -and the new industrial revolution, 5, 317 -, a topological subject, 2, 457 -of aphids and scale insects,4,263 of azidoquinones and re-lated compounds, 2,415 -of dyeing, 1, 145 -of homonuclear sulphur species, 2,233 --of the production of organicisocyanates, 3,209 -of transition-metal carbene com- plexes and their role as reaction intermediates, 2, 99 --of ‘vitamin’ D: the hormonal calciferols, 6, 83 -some considerations on the phhosophy of, 5,203 Chirality in carbonium ions, car-banions, and radicals, 2,397Chlorophyll chemistry, n.m.r.spectral change as a probe, 6,467 Chromatography, affinity, chemical aspects of, 2,249 Cis-and trans-effects of Iigands,2, 163 Complexes, alkali-metal, in aqueous solution 4,549 Complex hydride reducing agents,the functional group selectivity of, 5, 23 Conductivity and superconductivityin polymers 5, 95 Conformational analysis of some alcohols and amines: a comparison of molecular orbital theory, rota- tional and vibrational spectroscopy, 5,411 Conformation of rings and neigh-bouring group effects, development of Haworth’s concepts of, 3, 1 Index Conformational studies on small mol-ecules, l, 293 Contribution of ion-pairing to memory ‘effects’, 4,251 Co-ordination chemistry of aryldia-zonium cations :aryldiazenato (ary- lazo) complexes of transition metals, and the aryldiazenato-nitrosyl an- alogy, 4,443 Corrin synthesis, post-Biz problems in, 5,377 Crystal field approach to absorption bands in the spectra of stars, 5, 233 Cyclopolymerization, 1, 523 Dielectric relaxation in polymer solu- tions, 1, 49 Diffusion in liquids, the effect of isotopic substitution on, 5, 215 Difluoroamino-radical, gas-phase kin- etics of, 3, 17 Droplets and gases, interactions in the atmosphere of, 1,411 Drug design, quantitative, 3, 273 Dyeing, chemistry of, 1, 145 Echinoderms, 1, 1 Education, chemical, a reassessment of research in, 1, 27 Effect of isotopic substitution on diffusion in liquids 5, 215 Electrochemical techniques, applica- tion of to study of homogeneouschemical reactions, 4,471 Electron as a chemical entity, 4, 323 -scattering spectroscopy, thres- hold, 3,467 -spectroscopy 1,355Electronic properties of some chain and layer compounds, 5, 51 -transitions, vibrational inten-sities in, 5, 165 Electrons, solvated, in solutions of of metals, 5,337 Electrophilic aromatic substitutions, non-conventional, and related re-actions, 3, 167 -C-nitroso-compounds, 6, 1 Elimination reactions, isotope effect studies of, 1, 163 Enaminones, 6,277 Energetics of neighbouring groupparticipation, 2, 295 Enumeration methods for isomers,3,355 Index Environmental protection in the dis- tribution of hazardous chemicals, 4, 99 regulation: an international view, 5,431Enzymes, immobilized, 6, 215 in organic synthesis, 3, 387 ,the logic of working with, 2, 1 -, three-dimensional structures and chemical mechanisms of, 1, 319 Experimental studies on the structure of aqueous solutions of hydrophobic solutes, 2, 203 FARADAY The electron as a LECTURE. chemical entity, 4, 323 5-Substituted pyrimidine nucleosides and nucleotides, 6, 43 Fixation of nitrogen, 1,121Forces between simple molecules, 2, 181 Formation of hydrocarbons by micro- organisms, 3, 309 Fourier transform spectroscopy, chem- ical applications of advances in, 4, 569 Four-membered rings and reaction mechanisms 5, 149 Friedel-Crafts acylation of alkenes, 1, 73 Functional group selectivity of com- plex hydride reducing agents 5, 23 Gas-phase kinetics of the difluoro-amino-radical, 3, 17 Gases, and droplets, interactions in the atmosphere of, 1,411Glycoproteins, proteoglycans, and carbohydrate-protein complexes of human tissues, chemical aspectsof, 2,355Growth of computational quantumchemistry from 1950 to 1971, 2, 21 Handling toxic chemical s-envi ron-mental considerations, 4, 77 HAWORTH MEMORIAL TheLECTURE.consequences of some projects in- itiated by Sir Norman Haworth, 2, 145 HAWORTHMEMORIALLECTURE.The Haworth-Hudson controversy and the development of Haworth’s con- cepts of ring conformation and of neighbouring group effects, 3, 1 Health hazards to workers from industrial chemicals, 4, 82 Homogeneous catalysis, and organo- metallic chemistry, the 16 and 18 electron rule in, 1,337 --chemical reactions, application of electrochemical techniques to the study of, 4,471Hydrocarbon formation by micro-organisms, 3, 309 Hydrogen bond and charge transfer complexes, calorimetric investiga-tions of, 3, 193 -isotope effects, kinetic, recent advances in the study of, 3, 513 Hydrophobic solutes, experimentalstudies on the structure of aqueous solutions of, 2, 203 Imines, photochemistry of, 6, 63 Immobilized enzymes, 6, 215 Importance of solvent internal pres- sure and cohesion to solution phenomena 4,211Infrared and Raman vibrational spec- troscopy in inorganic chemistry,4, 107 INGOLD LECTURE.Four-membered rings and reaction mechanisms 5, 149 Inorganic pyro-compounds Ma[(X~07)bl, 5, 269 Insect attractants of natural origin,2, 75 Interactions in the atmosphere of droplets and gases, 1,411 -, metal-metal, in transition-metal complexes containing infinite chains of metal atoms, 1, 99 Introducing a new agricultural chem- ical 4, 77 Ion-molecule reactions in the evo-lution of simple organic molecules in interstellar clouds and planetary atmospheres, 6, 295 Ion-pairing, contribution to ‘memory effects’, 4, 251 Isocyanates and ketens, a mechanistic comparison of acylation by, 4, 231 -, organic, chemistry of the pro- duction of, 3, 209 Isomer enumeration methods, 3, 355 Isotope effect studies of elimination reactions, 1, 163 Isotopic substitution effects on dif-fusion in liquids, 5, 215 JOHN JEYES LECTURE.Chemicals which control plant growth, 6, 261 KELVIN LECTURE.Across the living barrier, 6, 325 Ketens and isocyanates, a mechanistic comparison of acylation by, 4, 231 Kinetics, gas-phase, of the difluoro- amino-radical, 3, 17 Kinetics of reactions in aqueousmixtures, 4, 55 P-Lactams, synthetic routes to, 5, 181 Lanthanide shift reagents in nuclear magnetic resonance spectroscopy,2, 49 Laser light scattering, quasielastic,2.325 Lasers, tunable, 3; 293 Ligands, cis-and trans-effects of, 2, 163 LIVERSIDGE Recent advances LECTURE. in the study of kinetic hydrogen isotope effects, 3, 513 Macrocyclic ligands, synthetic, trans- ition-metal complexes of, 4, 421 Mechanisms, chemical, and three-dimensional structures of enzymes, 1, 319 MELDOLA ChemicalMEDAL LECTURE. aspects of glycoproteins, proteo- glycans, and carbohydrate-protein complexes of human tissues, 2, 355 MEDALLECTURE.MELDOLA Molecular collisions and the semiclassical ap- proximation, 5, 125 MELDOLA MEDAL LECTURE. N.m.r. spectra1 change as a probe of chlorophyll chemistry 6,467 Metal-ion-promoted reactions of organo-sulphur compounds 6, 345 Metalloboranes and metal-metal bon- ding, 3, 231 Metal-metal bonding and metallo-boranes, 3,231 -bonds, multiple (especially quad- ruple), 4, 27 -interactions in transition-metal complexes containing infinite chains of metal atoms, 1, 99 Metals, binding to proteins, 6, 139 Micelles in aqueous solution, 6, 25 Index Molecular collisions and the semi-classical approximation 5, 125 orbital theory, comparison with rotational and vibrational spec-troscopy in conformational analysis of alcohols and amines, 5, 411 wavefunctions, chemical inter- pretations of, 5, 79 Multistability in open chemical reac- tion systems, 5,359 Natural products from echinoderms, 1, 1Natural products, polycyclic poly-substituted, systematic development of strategy in, 6, 413 Neighbouring-group effects and ring conformation, development of Haworth’s concepts of, 3, 1 -participation, energetics of, 2, 295 New perspectives in surface chemistry and catalysis 6, 373 Nitrogen fixation, 1,121C-Nitroso-compounds, electrophilic, 6, 1Non-conventional electrophilic aro-matic substitutions and related reactions, 3, 167 Nuclear magnetic resonance and the periodic table 59.1 ---,carbon-13, in biosynthetic studies, 4,497 ---spectral change as a probe of chlorophyll chemistry 6,467 ---spectroscopy, lanthanide shift reagents in, 2, 49 ----:spin-lattice relaxation, 4, 401 Nucleosides and nucleotides, pyrim- idine, 5-substituted, 6, 43 NYHOLM MEMORIAL LECTURE.For-ward from Nyholm’s Marchon Lecture, 3, 373 NYHOLM LECTURE.MEMORIAL Growth, change, challenge, 5, 253 Olefin metathesis and its catalysis,4, 155 Olefinic compounds, photochemistry of, 3, 329 Organic chemistry of superoxide,6, 195 Organoboranes as reagents for organic synthesis, preparation of, 3, 443 519 Index Organoborates in organic synthesis : the use of alkenyl-, alkynyl-, and cyano-borates as synthetic inter- mediates, 6, 393 Organometallic chemistry and homo- geneous catalysis, the 16 and 18 electron rule in, 1,337Organo-sulphur compounds, metal-ion-promoted reactions of, 6, 345 Organo-transition-metal complexes:stability, reactivity, and orbital correlations, 2, 271 PEDLER LECTURE.Porphyrins and related ring systems, 4, 1 Phase boundaries, reactivity of organicmolecules at, 1,229 Philosophy of chemistry, some con- siderations, 5, 203 Phosphates, aluminium, the chemistry and binding properties of, 6, 173 Phosphorus compounds, tervalent, in organic synthesis, 3, 87 Photochemistry of azobenzene and its derivatives, 1,481 of carbonyl compounds, 1,465 -of imines, 6, 63 -of olefinic compounds, 3, 329 -of organic sulphur compounds, 4, 523 -of the uranyl ion, 3, 139 of transition-metal co-ordination compounds-a survey, 1, 241 Photodegradation and stabilization of commercial polyolefins, 4, 533 Plant growth, control by chemicals 6, 261 Polymer solutions, dielectric relaxation in, 1, 49 -supports, insoluble, use in organic chemical synthesis, 3, 65 Polymerization and copolymerization of butadiene, 6,235Polymers, conductivity and super-conductivity in, 5, 95 Polyolefins, commercial, photodegra- dation and stabilization of, 4, 533 Porphyrins and related ring systems, 4,. 1Post-Bl2 problems in corrin synthesis, 5,377Preparation of organoboranes: re-agents for organic synthesis, 3,443PRESIDENTIALADDRESS1976.Chem-istry and the new industrial revo- lution, 5, 317 Properties and syntheses of sweetening agents, 6,431Prostaglandins, tomorrow’s drugs 4, 589 -, thromboxanes, PGX: biosyn-thetic products from arachidonic acid, 6,489Prostanoids, total syntheses of, 2, 29 Proteins, binding of heavy metals to, 6, 139 Pyrimidine nucleosides and nucleo-tides, 5-substituted, 6, 43 Pyro-compounds, inorganic, Mar(X207)bl, 5, 269 Quadruple bonds and other multiple metal to metal bonds, 4, 27 Quantitative drug design, 3, 273 Quantum chemistry, computational, growth of from 1950 to 1971, 2,21 -mechanical tunnelling in chem- iSt?, 1,211Quasielastic laser light scattering,2, 325 Radioactive and toxic wastes: a comparison of their control and disposal, 4, 90 Raman and infrared vibrational spec- troscopy in inorganic chemistry, 4, 107 Reaction mechanisms, four-membered rings and, 5,149Reactivity of organic molecules at phase boundaries, 1, 229 Recent advances in the study of kinetic hydrogen isotope effects, 3, 513 Research in chemical education: a reassessment, 1, 27 ROBERT LECTURE.ROBINSON Post-Blz problems in corrin synthesis, 5, 377 ROBERT LECTURE.ROBINSON The logic of working with enzymes, 2, 1 Rodent control, chemicals in, 1,381 Rotationally and vibrationallyinelastic scattering of molecules, 3, 407 Scale insects and aphids, chemistry of, 4, 263 16 and 18 Electron rule in organo- metallic chemistry and homogeneouscatalysis, 1, 337 Small molecules, conformation studies on, 1,293 Solids, surface energy of, 1,445 Solute-solvent interactions, spectro- scopic studies of, 5,297 Solution phenomena, the importance of solvent internal pressure and cohesion, 4,211 Solutions of metals: solvated electrons, 5,337Solvent internal pressure and cohesion, importance to solution phenomena, 4,211Some considerations on the philosophy of chemistry 5,203Some recent developments in chemis- try teaching in schools, 1,495 Spectra of stars, absorption bands in, a crystal field approach, 5, 233 Spectroscopic studies of solute-solvent interactions, 5,297 Spectroscopy, electron, 1,355 -, Fourier transform, chemical applicationsof advances in, 4, 569 -, rotational and vibrational, comparison with molecular orbital theory in confirmational analysis of alcohols and amines, 5,411 -, threshold electron scattering,3,467Spin-lattice relaxation: a fourth di- mension for proton n.m.r.spec-troscopy, 4,401Stability, reactivity, and orbital cor- relations of organo-transition-metal complexes, 2,271 Sterols, biosynthesis of, 1,259Structure of aqueous solutions of hydrophobic solutes, experimental studies on, 2, 203 Sulphoximides 4, 189 Sulphur compounds, organic, photo- chemistry of, 4, 523 Sulphur, organic compounds of, metal- ion-promoted reactions of, 6, 345 -species, homonuclear, chemistry of, 2,233Superconductivity and conductivity in polymers 5, 95 Superoxide, organic chemistry of, 6, 195 Surface chemistry and catalysis, new perspectives, 6,373Surface energy of solids, 1, 445 Sweetening agents, properties and Index syntheses of, 6,431Syntheses and properties of sweetening agents, 6, 431 -, total, of prostanoids, 2, 29 Synthesis, of corrins, post-Blz probl- lems in, 5,377 -of polycyclic polysu bstitu ted natural products, systematic de-velopment of strategy in, 6, 413 -, organic, enzymes in, 3, 387 9 , preparation of organo-boranes as reagents for, 3,443 9 , tervalent phosphoruscompounds in, 3, 87 Y , use of inorganic poly- mer supports in, 3, 65 -7 9 , the use of organo-borates as synthetic intermediates, 6, 393 Synthetic routes to P-lactams, 5, 181 Systematic development of strategy in the synthesis of polycyclic poly- substituted natural products: the aconite alkaloids, 6,413 TATE AND LYLE LECTURE.Spin- lattice relaxation: a fourth di-mension for proton n.m.r. spec-troscopy, 4,401Teaching of chemistry in schools, some recent developments in, 1,495 Tervalent phosphorus compounds in organic synthesis, 3, 87 Three-dimensional structures and chemical mechanisms of enzymes, 1,319Threshold electron scattering spec- troscopy, 3,467Thromboxanes, prostaglandins, PGX : biosynthetic products of arachidonic acid, 6,489TILDEN LECTURE.Electrophilic C-nitroso-compounds, 6,.1TILDENLECTURE.New perspectives in surface chemistry and catalysis6, 373 TILDEN Valence in transition- LECTURE. metal complexes, 1, 431 Topological subject-chemistry, 2, 457 Transition-metal carbene complexes, chemistry and role as reaction intermediates, 2, 99 -complexes, containing infinite chains of metal atoms, metal-metal interactions in, 1, 99 Index Transition-metal complexes of syn-thetic macrocyclic ligands, 4, 41 -complexes, valence in, 1, 431 -co-ordination compounds, photo- chemistry of, 1,241Tunable lasers, 3,293 Uranyl ion, photochemistry of, 3, 139 Use of insoluble polymer supports in organic chemical synthesis, 3, 65 Valence in transition-metal complexes, 1,431Vibrational infrared and Raman spec- troscopy in inorganic chemistry, 4, 107 Vibrational intensities in electronic transitions, 5, 165 Vibrationally and rotationally in-elastic scattering of molecules, 3,407 ‘Vitamin’ D, chemistry of: the hormonal calciferols, 6, 83

ISSN:0306-0012    

DOI:10.1039/CS9770600513

出版商:RSC    

年代:1977

数据来源: RSC