|
1. |
Proceedings of the Chemical Society. July 1963 |
|
Proceedings of the Chemical Society ,
Volume 1,
Issue July,
1963,
Page 189-228
Preview
|
PDF (5438KB)
|
|
摘要:
PROCEEDINGS OF THE CHEMICAL SOCIETY JULY 1963 TILDEN LECTURE* The Biosynthesis of Alkaloids By A. R. BATTERSBY (‘THE ROBERT LABORATORIES, ROBINSON UNIVERSITY OF LWERPOOL) EVERsince a sufficiently large number of alkaloidal structures had been worked out around the turn of the century there has been thought and speculation about the way in which plants construct these sub- stances and my purpose in this Lecture is to outline how the subject has developed. I shall deal with the speculative aspects and also more fully with recent experimental work on living plants. The aim of the experimental studies is to discover how the C,-unit carbon dioxide is converted say into morphine (1). A proper understanding of this pathway demands a knowledge of the substances which are involved as intermediates and also of the mechanisms by which the various transformations are carried out.I think it will be clear that this type of study is a major undertaking which must involve the extensive use of tracer methods. Indeed real progress in the study of alkaloid biosynthesis began when organic com-pounds labelled with carbon-14 and with other iso- topes became readily available. So we can under- stand why the fist publications covering tracer work on alkaloid biosynthesis began to appear only in the early nineteen-fifties.14 Since then however many groups of workers have tackled problems in this area and information is accumulating rapidly. But fist we should set this recent work into historical perspective.Nearly two thousand alkaloids have been isolated5 during the last 150 years or so and in most cases the molecular structures are known. This enormous effort has given a rich and varied collec- tion of structural types and it is this variety which makes the alkaloids so attractive and informative as subjects for research on biosynthesis. Many bio- synthetic processes of fundamental importance can be studied by tracer methods on this one group of compounds; examples of such processes are con- densation ring-closure methylation and demethyla- tion several different oxidation reactions dehydro- genation and so on. Yet organic chemists were not neglecting the field of biosynthesis long before tracers became available and their contribution then was to suggest possible biosynthetic routes to natural products.These ideas put forward over the last 50 years or so have been invaluable in guiding experi- mental work on living plants. In this Sir Robert * Delivered before the Chemical Society on February 25th 1963 at the Durham Colleges; on February 28th at Imperial College London S.W.7; on March 1st at the Royal College of Science and Technology Glasgow; and on March 14th at the University College Aberystwyth. 1 Kirkwood and Marion Canad. J. Chem. 1951,29,30. 2 Leete Kirkwood and Marion Canad. J. Chem. 1952,30 749. 3 Brown and Byerrum J. Amer. Chem. SOC.,1952,74 1523. 4 Dubeck and Kirkwood J. Bid. Chem. 1952,199,307. 6 Boit “Ergebnisse der Alkaloid-Chemie bis 1960,” Akademie-Verlag Berlin 1961.189 PROCEEDINGS Robinson6-s was a pioneer and he put forward many relation of its structure to that of tropinone (3) is important ideas on alkaloid biosynthesis. We shall clear. Dissection in each case as indicated exposes also see how fruitful have been the ideas of Barton units which Robinson suggestedQ might be derived and Cohen on phenol oxidation;1° and of the many from acetoacetic acid (5) and ornithine (6) or their other chemists who contributed ideas those of biological equivalents. With this as a guide experi- Schopf,ll Woodward,12J3 Wenkert,14 and Thomas15 mental study of hygrine biosynthesis could start by have been particularly stimulating to experimental feeding of labelled ornithine to a suitable plant. Such work on plants.Altogether there has been a con- a study starts appreciably along the biosynthetic siderable number of postulated biogenetic routes and pathway and similarly in the case of morphine which it is hardly surprising that some of them have been we considered at the outset much but not all of the shown to be incorrect; however it will become clear experimental work has involved labelled precursors in the sequel that the proposals have often been of Cg or greater size. Indeed for almost every remarkably close to the truth. alkaloid so far studied the possible precursors fed to Looking back at these speculations one can see the plants have ranged in size from C2 upwards. two main thought processes though they are often Among the substances tested have been acetic acid closely allied.One approach has been to seek and other simple fatty acids a-amino-acids carbo- common structural units within a group of alkaloids hydrates and shikimic acid (7) phenylpropanoid and to suggest the possible relationship of these units derivatives such as tyrosine phenylalanine and cin- to simpler natural products particularly to the namic acid and mevalonolactone16 (8). The bio- a-amino-acids and to acetic acid. The other has been synthetic pathways to these substances have been to correlate alkaloidal structures on the basis of a Me OH unifying reaction mechanism The fist ap roach can 00 (8) elucidated by the work of many investigators too numerous to list in a complete way. But the brilliant researches of Krebs,17Calvin,l* Davis and Sprinson,ls BloCh,2* LynenYa and Rudney= have been closely concerned with the building units we shall mainly be considering today.Experiments on alkaloid biosyn- thesis which use these substances as precursors are be illustrated by the cases of hygrine (2) and tro-thus based upon and take advantage of this vast pinone (3) the latter being derived from the tropane body of biosynthetic knowledge. alkaloids. If hygrine is re-written as (4) then the The hygrine-tropinone example illustrated the Robinson J. 1917 111,876. Robinson Internat. Congrew of Chemistry Madrid 1936 *‘The Molecular Architecture of Plant Products” ; J. Roy. SOC.Arts 1948 96,795; Internat. Congress Biochem. Abs. Communications to 1st Congress Cambridge 1949 p.32 (1950). * Robinson and Sugasawa J. 1931,3163; Robinson J. 1936,1079. @ Robinson “The Structural Relations of Natural Products,” Clarendon Press Oxford 1955. loBarton and Cohen “Fe(itschrift A. Stoll,” Birkhauser Basle 1957 p. 117. l1 Schopf Angew. Chem. 1937,50,787,797. l2 Woodward Nature 1948 162 155. l3 Woodward Angew. Chem. 1956,68,13. l4 Wenkert Experientia 1954,10,346; Wenkert and Bringi J. Amer. Chem. Soc. 1959 1474; Wenkert ibid. 1962, 84,98. l6 Thomas Tetrahedron Letters 1961 544. l6 Wolf Hoffman Aldrich Skeggs Wright and Folkers J. Amer. Chem. Soc. 1956,78,4499; 1957,79,1486; Wagner and Folkers Endeavour 1961,20 177. l7 Reviewed by Krebs and Lowenstein “Metabolic Pathways,” ed. Greenberg Academic Press 1960 Vol. I p. 129. 1* Calvin Angew.Chem. Internat. Edn. 1962,1,65. l@Davis and Sprinson “Symposium on Amino Acid Metabolism,” ed. McElroy and Glass Johns Hopkins Baltimore U.S.A. 1955; Davis Adv. Enzymol. 1955 16,247. fLoChaykin Law Phillips Tchen and Bloch Proc. Nut. Acad. Sci. U.S.A. 1958,44,998; Bloch Chaykin Phillips and de Waard J. Biol. Chem. 1959,234 2595; Yuan and Bloch J. Biol. Chem. 1959,234 2605 and earlier papers. 21 Lynen Eggerer Henning and Kessel Angew. Chem. 1958,70 738; Knappe Ringelmann and Lynen Biochem. Z.,1959,332 195 and many other papers. aa Rudney and Ferguson J. Biol. Chem. 1959,234,1076; Ferguson Durr and Rudney Proc. Nat. Acad. Sci. U.S.A., 1959,45,499. JULY 1963 first approach that of structural relations and we can now turn to the second approach the unifying reaction mechanism.One example is phenol oxida- tion10*23 and I shall deal with this in detail when we reach the appropriate alkaloids. The other example is Robinson’s that if the reaction illustrated below (A) could occur in plants then one could account for a wide variety of alkaloidal struc- tures. This type of reaction in which the carbonyl I1 I Ill (A) -C-+ C=O + NH +-C-C-N Ill I I\ I1 component is formaldehyde was later to become familiar to organic chemists as the Mannich reaction. The carbanion could be derived from an “activated” carbon atom or formally by the indicated shifts; the bottom line of formula (A) is simply to indicate that the carbonyl component and the secondary amine may be replaced by a (usually cyclic) Schiffs’ base.a ** (12) 00 An example of this sort of mechanism is the pro- po~ed~-~ route to lupinine (12). We need not concern ourselves with details but lysine (9) may be decar-boxylated and the resultant cadaverine (10) could be oxidised to give units which may yield the inter- mediate (11). The above mechanism could then operate to set up the bicyclic system of lupinine. If this scheme is correct in broad outline then a tracer experiment with [I ,5-1*C]cadaveMe should afford lupinine labelled at each of the asterisked positions. In fact when Schutte= fed the labelled cadaverine to Lupinus luteus plants he obtained radioactive lupinine; when he degraded this each of three of the asterisked carbon atoms carried one-quarter of the total activity-the carbon atom at position 6 has not so far been examined.This tracer experiment is the first of many I shall consider today selecting groups of alkaloids which have been most extensively studied by tracer methods. The experiments by this technique involve feeding plants with a labelled substance which is thought to be a precursor of the alkaloid under study; the alkaloid produced is then isolated and de- graded to determine the labelling pattern. Proper degradation is an essential part of the work and I cannot stress this too strongly. The step in which the labelled substance is introduced into the plant’s bio- synthetic system is often quite difficult but it is a technical problem which can be overcome and we shall consider it no further.When the feeding prob- lem is solved the organic chemistry begins. The theme which will runthrough the rest of the talk will be “Oxidation” though it will not be the only theme. The tropane alkaloid hyoscyamine (13) is an interesting case. When [2-lqC]ornithine (14) was fed’L5to Datura stramonium plants radioactive hyo- scyamine was isolated and a neat degradation2s showed it to be specifically labelled at only one of the two bridgehead carbon atoms (position 1 or 5). MeN -HOE Ph 2Y Vj:co tCH,9H Yph (13) c%?a 0s) t It is not yet known which one but this result shows clearly that ornithine can act as a specific precursor of the pyrrolidine part of the tropane system. Because of the specificity of the labelling no free symmetrical intermediate can lie on the pathway between ornithine and hyoscyamine.[1-l4-C]Acetic acid is also incorporated into the pyrrolidine portion of hyoscyamine and again only one of the two bridge- 23 Erdtman and Wachtmeister “Festschrift A. Stoll,” Birkhauser Basle 1957 p. 144. 24 Schutte Arch. Pharm. 1960,293 1006. 2s Lxete Marion and Spenser Cunud.J. Chem. 1954,32 11 16. 26 Leete J. Amr. Chem. Soc. 1962,84 55; cf. ref. 27. head positions was found to be labelled.27 This is as would be expected for a pathway leading from acetate by way of the tricarboxylic acid cycle to ornithine or its equivalent without the intervention of a symmetrical intermediate. The three carbon atoms at positions 2 3 and 4 of the tropane system arise from acetic acid almost certainly by way of acetoacetic acid.Thus [1J4C]- acetic acid was incorporated28 by Datura metel into hyoscyamine which carried most of its activity at position 3. In contrast [2-14C]acetic acid gave rise to a sample of the alkaloid which was heavily labelled at positions 2 and 4 whilst position 3 carried only a trace of activity. We must bear in mind of course that one of the two acetic acid units could well be incorporated by conversion first into malonyl co-enzymeA.29 These results are sufficient to indicate in broad outline the biosynthetic pathway to the tropane system and there is similar knowledge concerning the biosynthesis of the tropic acid (15) part of the mole- cule.Separate feeding experiments30 with [1-14C]- [2J4C]- and [3-14C]-phenylalanine (16) have shown that the carboxyl group of tropic acid can be derived from position 2 of phenylalanine and that the hydroxymethyl residue is provided by the carboxyl group of phenylalanine. There is evidence that this interesting rearrangement proceeds intramolecularly. Now let us turn to the alkaloids of the opium poppy Pupaver sornniferurn. This is a plant we have studied now for almost nine years and it has the attraction that it produces many different alkaloids. They are all related structurally to the 1-benzyliso- quinolines of which the so-called norlaudanosoline (17) is an example. The relation of papaverine (18) PROCEEDINGS to this system is obvious and for narcotine (19) a further one carbon unit is required to provide the carbonyl group of the lactone system.The benzyliso- quinoline system in cryptopine (20) is picked out with thickened bonds. However it is the hydro- phenanthrene alkaloids morphine (22; R = H), codeine (22; R = Me) and thebaine (23) which pose perhaps the most interesting biosynthetic problems. These too are structurally related to the benzyliso- quinolines. Indeed it was the possibility that such a relationship existed which led Gulland and Robin- son3l to propose their structure for morphine (22; R = H) which has been amply confirmed by later work; the structural relation can best be seen by R$$Me0 @Me R'\ HO"Zf=l (21) I (23) R'O+Me -re-writing the benzylisoquinoline system (cf.17) in the form (21). Robinson further post~lated~~~ that morphine and its relatives are formed in the plant by oxidative coupling of a suitable benzylisoquinoline precursor. A second important proposal came from Barton and CohenlO who postulated that the coupling step is the result of phenol oxidation; this is the second example of a unifying reaction mechanism *' Bothner-By Schutz Dawson and Solt J. Amer. Chem. Soc. 1962,84 52. 28 Kaczkowski Schutte and Mothes Biochim. Biophys. Acta 1961 46 588. 2B Bentley and Keil Proc. Chem. Soc. 1961 111; Bu'Lock and Smalley ibid. p. 209; Birch Cassera and Rickards Chem. and Ind. 1961,792. Leete J. Amer. Chem. Soc. 1960 82 612; Leete and Loudon Chem. and Ind.1961 1405; Loudon and Leete J. Amer. Clrem. SOC.,1962,84 1510. s1 Gulland and Robinson Mem. Proc. Manchester Lit. Phil. SOC.,1925 69 79. JULY 1963 which I mentioned earlier would be considered at the appropriate point. Their proposed mechanism can be illustrated with the diphenolic base (21) in which the groups R and R’ represent a residue possibly part of an enzyme surface which can be added to or taken away from a phenolic hydroxyl in order to provide protection adequate to ensure specific coupling of radicals. In its simplest form their pro- posal then involves oxidation of the base (21) by some one-electron transfer system to generate radicals which by coupling could yield the dienone (24). The fifth ring may then be formed as indicated and finally adjustment of the oxidation level is re-quired to afford the hydrophenanthrene alkaloids of opium.There is an attractive alternative sequence by which the biogenetic scheme may be continued from the dienone stage. It was pointed out by Bentley that the alkaloids which contained a hydrofuran ring are not oxygenated at position 7 whereas this position carries an oxygen atom in those alkaloids such as sinomenine (25) which lack the hydrofuran ring.32 This and several other facets of present knowledge of these alkaloids can be accommodated by a suit- able allylic elimination mechanism. The one shown below is based upon that of StorkB and involves NMe ;& RbI2 I NMe 0’ OMe (25) cox (26) (27) CH,-C r\_ ,CO,H p*H solution into the capsule.GO The research on morphine bio~ynthesis~*~~ was --c H02Cy-& guided by the relationship of this alkaloid to a H 1-benzylisoquinoline. Winterstein and Trier38 had OH earlier suggested that the latter system is built in (28) %I (29) Nature with units derivable from 3,4-dihydroxy- reduction of the dienone (24) formed by the phenol oxidation step to the dienol(26; X = H). This could then undergo elimination as indicated probably as the phosphate ester (26; X = phosphate) to give the complete pentacyclic system. An example of a similar process which is known in living systems is the formation of prephenic acid (29) from shikimic acid enof pyruvate (28).1993p We think it probable that this allylic elimination is an important step in the formation of a number of alkaloids and suitable examples will be considered later.However before leaving these biogenetic schemes I want to mention an interesting finding as yet unpublished and I am grateful to my friends at Imperial College% for per- mission to mention this result. This is that the dienone (24; R = R’ == Me) has been prepared and it is in the open form shown. The foregoing summary of biogenetic speculation about morphine codeine and thebaine covers only some of the schemes which have been published to the present day. I have selected those ideas which seem to be the important ones and have not obscured them with other proposals which are less attractive or which have been shown to be wrong by tracer experiments.These experimental studies have been yielding results over the period that some of the most recent schemes have been put forward. In summary the important proposals are that morphine codeine and thebaine are derived from a suitable 1-benzylisoquinoline; that the carbon-carbon bond joining positions 12 and 13 in these alkaloids is formed as a result of phenol oxidation; that the hydrofuran ring is closed by an allylic elimination mechanism. Let us now go back to 1954 to the start of experi-mental work on this problem. Such a study must cover the origin of the skeletal carbon atoms and the source of the methyl groups the nature of the various intermediate substances and the mechanisms of the conversion steps.Papaver somniferum var. Noordster was used for most of our own work and the radioactive precursors were administered to these plants by direct injection as an aqueous Phosphate phenylalanine. Since this amino-acid arises from 32 Bentley Experientia 1956 12 251. 33 Stork “The Alkaloids,’’ ed. Manske and Holmes Academic Press New York 1960 Vol. VI p. 233. 34 Davis Arch. Biochem. Biophys. 1958,78 794; Sprinson Adv. Carbohydrate Chem. 1960 15,235. 35 Barton Kirby and Steglich personal communication. 36 Battersby and Harper Chem. and Ind. 1958 364; Battersby Binks and Le Count Proc. Chem. SOC.,1960 287. 37 Battersby Binks and Harper J. 1962 3534. Winterstein and Trier “Die Alkaloide,” Borntraeger Berlin 1910. I94 PROCEEDINGS tyr0sine,3~ the first experiments involved feeding ~~-[2-'~C]tyrosine (30; R = H) to the plants.These yielded radioactive morphine (31; R = H) together with active codeine (31; R = Me) thebaine (23) and papaverine (32). The unambiguous degradations of morphine have been published3' and it is sufficient to say that they showed the alkaloid to be specifically and almost equally labelled at positions 9 and 16. Leete also studied the incorporation of activity from DL-[2-14C]tyrosine into morphine and his results40 are in full agreement with ours. This specific labelling of the morphine molecule is interpreted as proof that the C1,-skeleton of mor-phine is built from two aromatic units having at least two carbon atoms attached to the ring (Ar-C-C) and that these units can be derived from tyrosine in the living system.It seems probable that the para-hydroxy-series of compounds (tyrosine and its relatives such as p-hydroxyphenylpyruvic acid) is converted first into the dihydroxy-series (3,4-di- hydroxyphenylpyruvic acid and its relatives) since ~~-3,4-dihydroxyphenyl [2-14C]alanine (30; R = OH) was also incorporated into morphine in P. somni-When 3,4-dihydroxy [l -14C]phenethylamine (dopamine) was fed to the plants radioactive mor- phine (22; R = H) and codeine (22; R = Me) were isolated which were shown42 to be labelled only in the ethanamine side-chain. Thus the two units in- volved in the construction of morphine are different they are probably at the 3,4-dihydroxy-state of oxidation and one of them is dopamine or some close relative.Subsequent experiments have given support to these results and their interpretation. The next problem is to determine whether or not a suitable 1-benzylisoquinoline can be converted into the morphine-codeine-thebaine group of alkaloids in the plant. Norlaudanosoline (33; R = H) was the first substance chosen for synthesis since it seemed probable that this is the initially produced system which can then be correctly N-and 0-methylated by the plant. There remained however the problem of administration of the labelled benzylisoquinoline for at this stage no precursors of this size had been introduced into the biosynthetic system of a plant producing alkaloids. In the event DL-[l-14C]nor-laudanosoline (33; R = H) was incorporated smoothly into all the hydrophenanthrene alkaloids and into papaverine with a higher incorporation than had been found with tyrosine.Degradati~n~~ of the morphine (3 1 ;R=H) proved it to be specifically labell- ed at position 9. ~~-[3-~~C]Norlaudanosoline (33;R= R'f? Me \ ""00 (334 H) was also synthesised and this was incorporated in a separate experiment into the same four alkaloids.43 Here the codeine (31 ; R = Me) and thebaine (23) were the former by elimination of the ethanamine side-chain to yield an inactive phen- anthrene derivative (as expected for codeine labelled specifically at position 16). So far I have spoken of degradation without giving an indication of the methods which can be used and so it is worth out- lining the economical sequence of reactions which was used to degrade the thebaine.This makes use of the proneness of thebaine to undergo rearrangement and the one which occurs44 by proton catalysis in / Rosenfeld Leeper and Udenfriend Arch. Biochem. Biophys. 1958 74 252. 40 Leete Chem. and Ind, 1958 977; J. Amer. Chem. Soc. 1959 81 3948. 41 Battersby Harper and Staunton unpublished work. 42 Battersby and Francis unpublished work. 43 Battersby and Binks unpublished work. 44 Knorr Ber. 1903 36 3074; Freund and Holtoff ibid. 1899,32 168; Gulland and Virden J. 1928,921. JULY1963 methanol yields methebenine (34). Full methylation and Hofmann degradation afforded the vinylphen- anthrene (35) which was oxidatively cleaved to the aldehyde (36) and formaldehyde.The former was inactive and the latter carried all the activity of the original thebaine which is thus shown to be specific- ally labelled at position 16. These experiments demonstrate that a l-benzyliso- quinoline system in this case norlaudanosoline can be converted by the poppy plant into morphine codeine and thebaine. But although it seems very probable that norlaudanosoline lies on the direct pathway to these alkaloids one can feel fairly sure that this 1 -benzylisoquinoline undergoes methyla- tion before the oxidative coupling step occurs. The reason for this is that it has been established by Rapoport and his colleagues45 and by ourselves46 that thebaine (23) is the first hydrophenanthrene alkaloid to be formed.Codeine (22; R = Me) is formed from it and a final demethylation step yields morphine (22; R = H). In these experiments on the order of formation of a related group of alkaloids studies of incorporation rate^^^^^^ and careful feeding experiments with radioactive alkaloids45 both played their part; these methods should certainly be informative in other cases. From the knowledge that thebaine is the alkaloid produced from the coupling reaction and subsequent steps (reduction and allylic elimination ?) one can infer that the benzylisoquinoline which is oxidatively coupled is suitably methylated to generate thebaine (23) directly. The base (33; R = Me) is the one required;this was synthesised4' by suitable modifica- tions of the published method48 to allow position 3 to be labelled with carbon-14.It was found4' that DL-protothebaine was incorporated still more efficiently than norlaudanosoline (33; R = H) into the hydrophenanthrene alkaloids. The corresponding secondary base (33; RO = MeO NR = NH) was incorporated less efficiently than the base (33 ;R= Me) but more efficiently than norlaudanosoline.49 How- ever when DL-tetrahydro [3-14C]papaverine (3 3A) was fed to the plants there was virtually no incor- poration into m~rphine?~ Thus the biosynthesis is blocked by methylation of both phenolic hydroxyl groups in the benzylisoquinoline (33; RO = MeO NR = NH). All the foregoing evidence interlocks and the case is a very strong one in favour of the biosynthesis of 195 morphine codeine and thebaine by coupling of two aromatic rings in a partially methylated l-benzyliso- quinoline.All present knowledge is in keeping with the scheme below. It seems to us37 that the tracer experiments on poppy plants in which radioactive carbon dioxide was used50 are not in conflict with the results or with the interpretation which I have out- lined. C02 -Shikimic acid -Prephenic acid 0 OH OH rc-OH OH Meoa H\ kpave rine (32) Morphine (22 ;F+H) t c Codeine (22;R=Me) -Thebaine (23) The N-methyl group of morphine and the N-and the O-methyl group of codeine and thebaine have been shown51 to be derived from the S-methyl group of methionine. Though much has been discovered about the biosynthesis of these alkaloids a great deal remains 45 Rapoport Stermitz and Baker J.Amer. Chem. Soc. 1960 82 2765; Stermitz and Rapoport Nature 1961 189 310; J. Amer. Chem. Suc. 1961,83,4045. 46 Battersby and Harper Tetrahedron Letters 1960 No. 27 21. 47 Battersby Binks and Ramuz unpublished work. Tomita and Kikkawa Pharm. Bull. Japan 1956,4 230; J. Pharm. SOC.Japan 1957 77 195; Jain J. 1962 2203. 48 Battersby Binks and McCaldin unpublished work. so Rapoport Levy and Stermitz J. Amer. Chem. Soc. 1961,83,4298. 5* Battersby and Harper Chem. and Ind. 1958 365. to be done. The direction these new researches should take is clear and they should lead to many interesting results. The 1-benzylisoquinolines have been considered in biogenetic theory to be important intermediates not only for morphine and its relatives but also for many other isoquinoline An experi- mental study of the biosynthesis of a benzyliso-quinoline alkaloid thus has in addition to its own interest a bearing upon future work in the iso- quinoline series.The radioactive papaverine isolated from the feeding experiments we have been dis- cussing above allowed a start to be made. It was found52 that the papaverine (32) isolated after DL-[2-14C]tyrosine had been administered was specific- ally and about equally labelled at positions 1 and 3 (cf. morphine above). Again this is interpreted as proof that the papaverine skeleton is built from two aromatic units each carrying a side chain of at least two carbon atoms (Ar-C-C) and that these units are derivable from tyrosine in the plant.Moreover DL-[1-14C]n~rlaudan~~~lineincorporated into was papaverine (32) which became specifically labelleds at position 1. Aromatisation of a tetrahydroiso-quinoline system is thus demonstrated. The alkaloids present in the Amaryllidaceae form another fascinating group of compounds and their biosynthesis has been intensively studied by several groups of workers. More than the whole of this Lecture would be required to cover the results fully and I shall therefore deal only with the main findings. Our own tracer work has been carried out in col- laboration with Professor W. C. Wildman (Ames Iowa) and Dr. H. M. Fales (Bethesda Maryland) and I want to record the pleasure we have had from this fruitful collaboration on the biosynthesis of lycorine (40) haemanthamine (41) haemanthidine tazettine crinamine and belladine.We shall discuss the first two which exemplify two of the three main skeletal types; an example of the third type is galanthamine (42). Barton and CohenlO made the highly satisfying proposal that all three skeletons arise by oxidation of the same type of diphenolic precursor based upon the so-called norbelladine (37; R = H). The norbelladine skeleton is written in three different ways (37) (38) and (39) so that one can see how aryl-aryl coupling and further plausible changes could lead to the three alkaloids. In the survey of experimental work which follows I shall not describe the degradation of the radio- active alkaloids.It is sufficient to say that unam- PROCEEDINGS biguous degradations were carried out in each case. QH -7 O -< &? 0 OH (39) Bearing in mind the work on the opium alkaloids one would expect that part of the lycorine molecule (40) drawn with thickened bonds to be derivable in the plant from tyrosine. In fact when DL-[~-~~C]- tyrosine (43; R = OH) was fed to double Narcissus plants radioactive lycorine (40) was isolated and this was shown54 to be specifically labelled at posi- tion 5. Similarly ~~-[3-~~C]tyrosine (43; R = OH) was used by Nerine bowdenii plants to afford ly~orine~~ labelled only at position 4. The above expectation is thus proved to be correct; further it is also established that the c6-Cl unit represented by ring A and carbon atom 7 of lycorine (40) cannot be derived from tyrosine and its close relatives.How- ever the radioactive lycorine which resulted when ~~-[3-~*C]phenylalanine (43; R = H) was the pre- cursor tested was found to be labelled56y57 only at position 7; position 4 was radioinactive. This means that phenylalanine and its close relatives can provide ring A and the carbon at position 7 whereas tyrosine and its equivalents provide the rest of the skeleton. The two sets of precursors are thus on quite separate pathways in these Amaryllidaceae plants.58 [3-14C]-Cinnamic acid55 (44)and [1-14C]tyramine55~cf*56 (45) were also used by the plants in separate experiments to form two samples of lycorine (40) labelled respectively at positions 7 and 5.52 Battersby and Harper Proc. Chem. SOC.,1959 152; J. 1962 3526. 53 Battersby and Parry unpublished work. 54 Battersby Binks and Wildman Proc. Chem. Soc. 1960 410. 65 Battersby Binks Breuer Fales and Wildman unpublished work. Suhadolnik Fischer and Zulalian J. Amer. Chem. SOC.,1962,84 4348. 57 Wildman Battersby and Breuer J. Amer. Chem. SOC.,1962 84 4599. 58 Cf. Neish Ann. Rev. Plant Physiol. 1960 11 55. JULY 1963 Similar experiments showed that those parts of haemanthamine591a0 (41) and galanthamheal (42) drawn with thickened bonds are also derivable from tyrosine. Also activity from the benzylamine (46) was incorporated without randomisation into galantha- mine (42) probably by way of the aldehyde (47) since the AT-methyl label was lost in the process.62 o^.c;"" R\ (43) fieOc=/^;.l H,N--" HOUw0 (45) (47) The intermediates must now occupy our attention.\ H~N--J / Amaryliidame alkaloids .. (48) Can a diphenolic base of the norbelladine type (cf. 37) give rise to these various alkaloidal structures? Several crucial experiments showed that this can indeed occur. Thus the Imperial College group proved62 that the base (39; R = R = Me) labelled as shown was incorporated without breakdown into galanthamine (42) and we founda3@ that the doubly labelled norbelladine (37; R = H) is incorporated intact into lycorine (40).Both groups of workers es ta bli~hed~*j~~ doubly labelled norbelladine that derivatives are similarly incorporated into haeman- thamine (41) and into crinamine which is stereoiso- meric with haemanthamine; there was no change in the ratio of activities of the two labelled positions in the course of the biosynthesis. The combined evidence is thus very strong that the coupling of two phenolic rings is a key stage in the biosynthesis of the Amaryllidaceae alkaloids. All the available information is collected in the scheme below and though the sequence of events shown is in keeping with the experimental results much more work will be required before all the details can be accepted as proved. Evidence has been obtained for the presence of phenolic norbelladine derivatives (cf.48) in Amaryl- lidaceae plants by radiochemical dilution analysis. Thus if these substances are present then labelled tyrosine fed to the plants should cause their pools to become radioactive. A particular substance can then 59 Battersby Fales and Wildman J. Amer. Chem. Soc. 1961 83 4098; Wildman Fales and Battersby ibid. 1962 84 681. 6o Jeffs Proc. Chem. Suc. 1962,80. Barton Kirby Taylor and Thomas forthcoming Communication. 62 Barton Kirby Taylor and Thomas Proc. Chem. Soc. 1961 254; 1962 179. 63 Battersby Binks Breuer Fales and Wildman Proc. Chem. Soc. 1961 243. '* Wildman Fales Highet Breuer and Battersby Pruc. Chem. Soc. 1962 180. 65 Barton Kirby and Taylor Proc. Chem. SOC.,1962,340. 66''AxelrodFales Mann and Mudd forthcoming Communication.and Tomchick J. Bid. Chem. 1958,233 702 and refs. therein. I \ be sought by using synthetic radioinactive material to act as a carrier for the biosynthetic product. In this way a weak but clearly positive result was obtained for the presence of the base61 (48; R = R' = Me) and the base55 (48; R = Me R' = H) in two of the plants used in the foregoing studies. The indication from these dilution experiments that such bases are present only in minute amounts is not at all surprising. All the experiments I have discussed so far have been carried out on living whole plants and it is of considerable interest that studies of isolated enzyme systems are now yielding the first results. Our col-leagues at Bethesda have isolated% a highly purified enzyme system from Nerine bowdenii which can be incubated with norbelladine (48; R = R = H) and (-)-5-adenosyl-~- [methyZ-14C]methionine to yield almost entirely the monomethyl ether (48; R = Me R' = H); this corresponds to para-methylation.There was also present about 4% of the isomer in which by contrast the rneta-hydroxyl had been methylated. The plant enzyme differs sharply from the mammalian enzyme catechol 0-methylpherase,G' which monomethylates 4-substituted catechols main- ly at the meta-hydroxyl group. The last group of alkaloids I shall discuss includes the protoberberine bases e.g. canadine (50) and the phthalide-isoquinolines such as narcotine (51 ;R = OMe). This group was considered by Robinson7p9 to be yet another which is derived in the plant from 1-benzylisoquinolines and he postulated the use of a one-carbon unit to provide the carbon at position 8 of canadine and berberine (52); this carbon atom has come to be known as the berberine bridge.He further proposed that hydrastine (51; R = H) and narcotine (51; R = OMe) may be formed biosyn- thetically by oxidation of the protoberberine skele- ton. An alternative biogenetic theory6* for the phthalide-isoquinolines involved the rearrangement of a hydrated form of prephenic acid to provide ring D and the lactone system (see 51). A simple benzylisoquinoline unit such as norlaudanosoline (17) is not on this theory a precursor. The results from tracer experiments are clearly in favour of the original proposal.It has been found that [2-14C]tyrosine (30) is used by Hydrastis canadensis to yields9 hydrastine (51; R = H) labelled specifically at positions 1 and 3. The same precursor was used by Papaver somniferurn to form narcotine70 (51; R = OMe) also labelled solely at positions 1 and 3. One can infer from these results that the building blocks for these alkaloids are two aromatic units carrying at least two-carbon side chains (Ax-C-C) which are derivable in the plant from tyrosine. Moreover the two units differ since 3,4-dihydroxy[ 1 -14C]phenethylamine (49) was used by H. canadensis to form hydrastine (51; R = H) labelleds9 only at position 3 This experiment and the similar one on morphine (p. 194) are thus in complete agreement.The berberine (52) isolated from the above experiment in which tyrosine was fed to H. canadensis was similarly shown71 to be labelled specifically at positicms 6 and 14 and the same interpretation can be given to this result. Strong support for the use of a l-benzyliso-quinoline as an intermediate on the way to narcotine comes from the incorporation of [ 1 -14C]norlaudano- PROCEEDINGS soline (17) into narcotine (51 ; R = OMe) which was shown70 to be specifically labelled at position 1. Moreover when sodium [14C]formate was fed to P. sumngerum plants the narcotine produced was radioactive and carried70 13 % of the total activity at the carbonyl group of the lactone residue.* This carbon atom which on the above hypothesis of a biosynthetic relation between the phthalide-iso-quinolines and the protoberberines corresponds to the berberine bridge is thus demonstrated to be derivable from the pool of C,-units in the plant.It does not follow that the C,-unit used is formic acid itself for when this substance in radioactive form is fed to a plant several related C,-units and more complex substances will all become labelled. What then is the precise nature of the extra carbon atom at the time it is built on to the 1-benzylisoquinoline system and how is it further modified? Let us speculate about this-indeed this Lecture can be rounded off with some ideas which will serve as pointers for future work. It seems probable that the berberine bridge is derived oxidatively from an N-methyl group and there are several mechanisms by which this could be accomplished.Rearrangement of an N-oxide (53) may occur as illustrated much as in the case of simple N-o~ides~~ to yield a carbinolamine. Ring- closure to the protoberberine system is then straight- forward. Or again direct oxidative attack at the N-methyl group of a 1 -benzyl-N-methylisoquinoline could generate the carbinolamine (54) or the related imine (55) without the intervention of an N-oxide. A further plausible route can be written which involves attack at the N-methyl group by a radical generated as the result of phenol oxidation. At this stage how- ever the details need not greatly concern us and the 1 t h-* Recent experiments indicate that this is probably a minimal figure; further refinements of technique are in progress.Wenkert Experientia 1959 15 165. 6s Gear and Spenser Nature 1961 191 1393; J. Ainer. Chem. Soc. 1962 84 1059. 70 Battersby and McCaldin Proc. Chem. SOC.,1962 365. 71 Spenser and Gear Proc. Chem. SOC.,1962 228. 72 Craig Dwyer Glazer and Horning J. Amer. Chem. SOC.,1961 83 1871. JULY 1963 aim must be to test the main proposal;73 this is now in progress. An example of the generation of a methylenedi-oxy-group by oxidative attack at an O-methyl group comes from the demonstration65 that O-methylnor- belladine (38; R = Me) labelled at the @methyl group and in the skeleton is incorporated into the two labels. Our without change in the supports haemanthamine (41) own work55 strongly ratio of this pathway to the methylenedioxy-group in that [methyZ-14C]methionine fed to double Narcissus plants gave rise to lycorine (40) labelled solely at the methylene carbon of this group.Norpluviine (58; R = H) was also isolated and this carried all its activity in the 0-methyl group. Furthermore the plants con- verted tritiated norpluviine (58; R = H) in 10% yield into tritiated ly~orine~~ (40). Not only does this result support the foregoing work but it also gives the first indication that the second hydroxyl group in ring c of lycorine (40) arises by allylic oxidation. In considering the relation of the protoberberines to the phthalide-isoquinolines it is interesting to see the same series turning up in the Amaryllidaceae alkaloids.For example ophiocarpine (56) corres-ponds to pluviine (58; R = Me) rhoeadine (57) is the analogue of lycorenine (59) whilst narcotine (51 ; R = OMe) and hydrastine (51; R = H) are at the Me& R\ Me (58) (60) oxidation state Of a carboxylic acid and can be corn-pared with homobrine (60). It seems highly prob- able that we are looking here at two trios which are related biosynthetically in the same way. During the discussion of thebaine codeine and morphine I mentioned that reactions involving allylic elimination probably occur quite widely in the alkaloid field. Isolated examples have been con- sidered previously notably by Boit in his valuable book but this step can reasonably amount for Suite a number of different alkaloids.For example the Mo H~ -c ~~ Meo/ HO (61) 0 (62) c %-X alkaloids isothebaine (64)and stephanine (65) have unusual oxygenation patterns but they can be con-sidered to arise from the normally oxygenated I-benzylisoquinoline (61) by phenol oxidation to form the dienone (62). If this undergoes reduction to the dienol (63; X = H) then allylic elimination can occur probably as the phosphate ester (63; X = phosphate) with the illustrated rearrangement to generate isothebaine (64).Were bond (b) to be the one undergoing migration rather than the illustrated case involving (a) the product would be close to stephanine. Only the methylenedioxy-system would remain to be generated.Clearly these ideas can be tested experimentally and this is in hand. So we end by looking to the results from future experiments and they will add still more to our knowledge of the way plants build complex mole- cules and of the chemistry which plants can carry out. Moreover biogenetic thinking and experimental work on the living material have already led to simple and often mild methods for the synthesis of molecules otherwise difficult to construct Future wozk with plants is likely to assist this in vitro approach. 73 Professor D. H. R. Barton and Dr. G. W. Kirby independently reached this same conclusion and kindly informed me of it at the Summer School on “Biogenesis,” Milan September 1962. Much remains to be done in the study of alkaloid biosynthesis in vivo and the deepest understanding will come from the combined efforts of chemists biochemists and enzymologists.Yet my aim in this Lecture was to show that many of the problems in- volved in this field Cali for the particular skills of the organic chemist and i am sure that his contribution to such a team effort will continue to be a very important one. PROCEEDINGS My own team of organic chemists has been and is a most enthusiastic one and I am very pleased to be able to acknowledge their skilful work especially that of my colleague Dr. R. Binks who has made important contributions to almost all of it. Finally I record my sincere thanks to Professor Wilson Baker F.R.S. for his help and encouragement at the University of Bristol.It was there that most of our own researches were carried out. COMMUNICATIONS Paramagnetic Shifts in Fluoro-aromatic Compowids BYA. J. R. BOURN and E. W. RANDALL D. G. GILLIES (DEPARTMENT QUEEN LONDON,E.1.) OF CHEMISTRY MARYCOLLEGE IN view of interest in pentafluorophenyl derivatives of metals,l we measured the 19Fchemical shifts of a number of pentafluorophenyl compounds with results given in the Table. C6F5X * (Ortho) 19FC s(Para) hemical hifts** (meta) 104-3 157.5 161.2 115.6 158.7 163.7 117.1 152-6 159.6 121-8 150.8 159.8 119.1 152.4 159.6 133.3 155.4 161.5 141-0 156-6 161.5 163.7 163.7 163.7 128.7 146.0 161.1 * Measured in p.p.m. (L-0-2p.p.m.) relative to CClsF as internal standard.The 19F chemical shifts of the +fluorine atoms in C6F5X compounds depend markedly on the nature of X. Moreover 6 (ortho) shifts to lower field when in a given group of the Periodic Table the atomic number of X increases or when the pentafluoro- phenyl group is bonded to a transition metal rather than to a main-group metal. Similar chemical-shift relations have been reported for perfluoroalkyl derivatives C,F2,+,X [X = halogen Mn(CO)5 Sat3 et~.)~,~ and have been attributed2 to the mesence of unoccupied low-lying orbitals in the gr&p X. The magnetic field causes mixing of low-lying excited states of appropriate symmetry with the ground state of the molecule. This mixing produces paramagnetic contributions to the chemical shifts which are greatest at the fluorine atoms closest to X.For main-group atoms of high atomic number the excitation energies are smaller than for lighter atoms and larger paramagnetic shifts are expected in the former cases. Even smaller energies and consequent- ly larger paramagnetic shifts should occur in transition-metal compounds. From the results given in the Table we suggest 6 (ortho) for pentafluoro- phenyl derivatives is governed in part by these paramagnetic effects. Similar downfield shifts should occur in substi- tuted monofluorobenzenes. A previous- thorough analysis of 19Fshifts in these compounds showed that even after allowance for mesomeric and induc- tive effects a large additional shift occurs in ortho-substituted monofluorobenzenes.* This additional shift has SO far not been satisfactorily explained.We suggest that it arises through the Presence of low- lying unoccupied orbitals in the €TOUP which is ortho to the flUorine atom. We are indebted to the U.S. Department of the Army through its European Research Office for support of this research and to the Department of Scientific and industrial Research for Research Studentships to A.J.R.B. and D.G.G. (Received May 23rd 1963.) Holmes Peacock and Tatlow Proc. Chem. SOC.,1963 108; Treichel Chaudhari and Stone J. Orgonometallic Chem. in the press; Massey Park and Stone Proc. Chem. SOC.,1963 212. Pitcher Buckiagham and Stone J. Chem. Phys. 1962,36 124. Tiers J. Amer. Chem. SOC.,1956,78,2914. Gutowsky McCall McGarvey and Meyer J.Amer. Chem. SOC.,1952,74,4809. JULY 1963 201 High-resolution Electron-Nuclear Double-resonance Spectraof Solutions of a Free Radical By R. E. RICHARDS and J. W. WHITE (PHYSICALCHEMISTRY OXFORD) LABORATORY WHENthe electron resonance is saturated in a solu-tion of a paramagnetic substance changes may occur in the nuclear res0nances.l Resonances of nuclei which are scalar-coupled to the unpaired electrons may be enhanced in intensity and may also be shifted; resonances of nuclei which are dipolar coupled to the electrons may be reduced in intensity or even re~ersed.~~~ These effects are illustrated in the experiments described below in which both types of behaviour are found in the same sample.RG.1 1 Figure la shows the proton resonance at 53.15 Mc/sec. of a solution of 2,4,6-tri-t-butylphenoxy-radical in benzene. The strong resonance at low fields arises from the protons in the solvent and the resonance at high fields comes from the protons of the t-butyl groups. The resonances of the aromatic protons of the radical are obscured beneath the benzene peak. The weak resonances between the two main peaks arise from minor con- stituents formed during the preparation of the radicals and there is a side-band resonance of the benzene peak which occurs 400 c/sec. to higher fields for calibration purposes. When the electron resonance of the radical is excited by 8-mm. micro-wave radiation of the appropriate frequency changes occur in the spectrum which are shown in Figures 1 b lc and Id representing the effects of a progressive increase in microwave power applied.The solvent resonance (as well as the weak peaks of the minor constituents) is first decreased in intensity as the microwave power is increased and then inverted until it reappears as an even stronger resonance than in the unperturbed spectrum. On the other hand the resonance of the t-butyl groups in the radical is increased in intensity. The reversal of the solvent resonance is consistent with the effect of dipoIar coupling between the nuclei and electrons and the enhancement of the nuclear resonance of the radical is as expected if the nuclei are scalar-coupled to the unpaired electrons. Under higher resolution the resonances of the t-butyl protons are split into a number of compo- nents.These are chemically shifted as the splittings are field dependent. When microwave power is applied some components are more strongly en- hanced than others. The detailed dependence of the nuclear resonance intensities on applied microwave power depends on (a) the scalar loss factor (leakage of spin polarisation of scalar-coupled nuclei),2 (6) the free electron spin densities at various sites in the free radical (c) the modes of molecular collision between radical and solvent molecules and (d) three-spin processes slr FIG.2 1 Overhauser Phys. Rev.,1953,92,411. a Abragam “Principles of Nuclear Magnetism,” Oxford University Press 1961. a Richards and White Pruc.Roy. Suc. 1962 A 269,287 301. PROCEEDINGS between different nuclei and the electrons in the sample.* These power dependences have been studied in detail and are discussed theoretically in a forthcoming p~blication.~ The different behaviour of nuclei coupled to an unpaired electron by scalar and by dipolar processes has also been observed at low fields and even under conditions of very poor resolving power. Figure 2a shows the proton resonance of a solution of the same phenoxy-radical in p-fluoronitrobenzene. This spec- trum was obtained at 13.4 Mclsec.; the spectrum is the first derivative of the resonance as the magnetic field was modulated with a frequency and amplitude smaller than the line-width. When the electron resonance is excited by microwave radiation of about 3 cm.wavelength the resonance changes its shape (Figures 2b-2i) and two components can now be distinguished. The low field component (from the protons of the solvent) is strongly inverted but a high field component remains with positive sign which causes the asymmetry of the derivative curve. The proton resonance of the t-butyl groups of the radical is enhanced by the application of the micro- wave power because these protons are scalar-coupled to the electrons and the resonances due to the solvent are inverted because the protons are dipolar-coupled to the electrons. It seems probable that experiments of this kind will lead to valuable information about weak mole- cular interactions in solution and about the details of electron-nuclear spin coupling in radicals.The experiments at 3 cm. and 13 Mclsec. were performed on an apparatus described previ~usly.~ The high resolution spectra at 8 mm. and 53*15 Mc/sec. were carried out with a new apparat~s.~ We express our gratitude to the Paul Fund of The Royal Society for a grant towards the construction of the high-resolution double-resonance apparatus. (Received April 29th 1963.) Richards and White Discuss. Faraday SOC.,1962 1962 34 96. Richards and White to be published. Photochemistry of Diphenylamine Solutions By E. J. BOWENand J. H. D. ELAND (PHYSICAL LABORATORY, CHEMISTRY OXFORD) IT was noted by Parker and Barnesx that carbazole is a product of the ultraviolet irradiation of solutions of diphenylamine.The reaction has been examined in more detail. The loss of diphenylamine and production of carbazole are readily followed by measurements of fluorescence and absorption spectra or by separation of reactant and product by extraction with 50% v/v sulphuric acid solution. Solutions in deoxygenated methanol propan-2-01 or hexane at ordinary temperature illuminated by mercury radiation at about 3000 A show formation of carbazole with a quantum yield of 0-1 independ- ent of concentration from 5 x to 7 x M. In the early stages of the reaction the loss of diphenyl- amine is within about 10% of the formation of carbazole but later some polymer is formed. Dissolved oxygen has an inhibiting effect on the reaction and also quenches the fluorescence of di- phenylamine the yield of which is about 0.05.The constants of these effects are of similar magnitude and if diffusion is the controlling factor they indicate an excited state of true radiational mean life of at least lO-’sec. The integrated area of the absorption band corresponds to an excited state of one-thirtieth of this life. Chemical reaction therefore in-volves a different state from the original excited level but not the triplet because there is no inhibition by butadiene. In viscous paraffin solution the quantum yield of reaction is larger and in light petroleum it passes through a maximum of about 0.2 as the temperature is lowered becoming zero at -190” where mole- cular movements are restricted.Molecular hydrogen in amount comparable to the carbazole formed has been detected mass-spectrometrically which dis-tinguishes this reaction from the apparently similar one of the photo-formation of phenanthrene from cis-stilbene where oxygen is required to remove the hydrogen as water.2 Photochemical cleavage of the N-H bond occurs in hydrocarbon solvents only with a very low efficiency.3 (Received May 13th 1963.) Parker and Barnes Analvst 1957. 82. 606. Mallory Gordon and Wood J. Amer. Chem. SOC.,1963 85 828; Moore Morgan and Sternitz ibid. p. 829. Lewis and Lipkin J. Amer. Chem. Soc. 1942 64 2801 ;Lewis and Bigeleisen ibid. p. 208; Tarutina “Primenie Metodov Spectroscopi i Prom. Prodovol’sten,” Leningrad 1955. JULY1963 203 1-Benzylisoquinolinesas Precursors of the Opium Alkaloids Tracer and Stereochemical Studies By A.R. BATTERSBY R. J. FRANCIS, R. BINKS,D. M. FOULKES D. J. MCCALDIN and H. RAMUZ (THEROBERT LABORATORIES OF LIVERPOOL ROBINSON UNIVERSITY and THE CHEMISTRY UNIVERSITY DEPARTMENT OF BRISTOL) IT has been firmly established1s2 that [l-14C]nor-R = H) and (I; R = R’ = Me) were synthesised’ Iaudanosoline (I; R = R = H) is incorporated with I4C-labelling at position 3; these were fed specifically into morphine (11; R = H). [3-14C]Nor-separately to poppies. The incorporations into mor- laudanosoline has now been synthesisedl and fed phine (11; R = H) were 3-2 % and 7.3 % respectively to Papaver somniferum plants to yield radioactive (cf.incorporation of norlaudanosoline 2.2 % in a alkaloids. Degradation of thebaine (111) by way of parallel experiment). (&)-[3-14C]Tetrahydropapa-methebenine3 proved specific labelling at position 16. verine [(&) -IV; R = Me R = HI was not how- The codeine (11; R = Me) lost all its radioactivity ever incorporated significantly into morphine; the when the ethanamine chain was eliminated4 to yield biosynthesis is thus blocked by methylation of the 0-acetylmethylmorphol. two phenolic groups required for phenol ~oupling.~ Resolution of the (&)-base [(k)-IV; R = RO / CH,Ph R = Me] gave the (+)-and the (-)-form [alD44”(c 1.0in CHCl,). Acid-catalysed debenzyl- ation yielded the hydrochlorides of the resolved ’a) bases (IV; R = H R’ = Me and enantiomer).RO \ Ho%Rf Methylation of the (+)-form [a], + 73” (hydro- OH (1) HO chloride; c 1.0 in water) with diazomethane gave (+)-laudanosine [aID+ 54” (c 1.0 in CHCl,) of M/ known absolute configuration.8 It follows that the absolute configuration of the (-)-base (IV; R = H 0 R’ = Me) is as shown; this then is the enantiomer Meo\ which corresponds to thebaine (111) codeine (11; R M‘SR’ Me0\ = Me) and morphine (11; R = H). Feeding experi- OR (IV) ments are in hand with the optically active bases (nn) (IV; R = H R = Me and enantiomer) which carry For reasons summarised el~ewhere,~~~~~ the multiple 14C and tritium labels. methylated (f)-1 -benzylisoquinolines (I; R = Me (Received May 14th 1963.) Battersby and Binks Proc. Chem. SOC.,1960 360. a Battersby and Binks Quart.Rev. 1961 15 277. Knorr Ber. 1903 36,3074. Pschorr and Dickhauser Ber. 1911,44 2633. Barton and Cohen “Festschrift A. Stoll,” Birkhauser Basle 1957 p. 117. Battersby The Donegani Lectures on Biosynthesis Milan September 1962 in press; Battersby Tilden Lecture Proc. Chem. Soc. 1963,189 and refs. therein; Barton Hugo Muller Lecture Proc. Chem SOC., in press and refs. therein. We are grateful to Professor D. H. R. Barton for kindly sending LZS a copy of this lecture before publication. ’Tomita and Kikkawa Pkarm. Bull. Japan 1956 4 230; Gopinath Govindachari and Viswanathan Chem. Ber. 1959,92 1657; Jain J. 1962,2203 and refs. therein. Corrodi and Hardegger Helv. Chim. Acta 1956,39 889. The Biosynthesis and Synthesis of Morphine Alkaloids G.W. W. STEGLICH,and G. M. THOMAS By D. H. R. BARTON (IMPERIAL LONDON, COLLEGE S.W.7) IT was suggested1 that morphine alkaloids are furnish thebaine (V). Work so far reported2f3 biosynthesised by oxidation of (I; R = R’= Me) to supports the correctness of this scheme. give the dienone (11; R = H),which after isomerisa- The benzylisoquinoline (I; R = R’ = Me) was tien to (HI) reduction and dehydration would synthesised4 labelled with 14C at the N-methyl and Barton and Cohen “Festschrift A. Stoll,” Birkhauser Basle 1957 p. 117. aBattersby Tilden Lecture Proc. Chem. Soc. 1963 189 and refs. therein. We thank Professor A. R. Battersby for kindly providing us with a copy of his lecture prior to publication. Barton Hugo Miiller Lecture Proc. Chem.Soc. in press and refs. therein. Tomita and Kikkawa Pharm. Bull. Japan 1956,4,230;J. Pharm. SOC.Japan 1957,77,195; Gopinath Govindachari and Viswanathan Chem. Ber. 1959,92 1657; J. Kunitomo J. Pharm. SOC.Japan 1961,81 1253; Jain J. 1962,2203. PROCEEDINGS with tritium as indicated. It was fed to Papaver reduction and conversion into radioactive thebaine. SomnifeYurn plants and the radioactive morphine As judged by radioactivity a third,1° and long (VI;R = H) isolated (incorporation 0.14%; cf. sought,ll synthesis of the morphine alkaloids has tyrosine incorporation at the same time 0.08%) thus been completed. with the same 14Cto tritium ratio as in the precursor (I; R = R’ = Me) and with all the 14Cactivity in the N-methyl The important intermediate dienone (11; R = H) has been synthesised in the following way.Oxidation of “phenolic dihydrothebaine”6 acetate with selenium Me0 dioxide followed by treatment with manganese di- OH (I’ ,//’ / oxide gave the acetate (11; R = Ac) m.p. 171” / I’ [a] + 120” (c 1-26) which on mild alkaline hydrolysis furnished the dienone (11; R = H) m.p. 197” [a] + 111’ (c 1.69). Unlike its analogues,’ (11; R = H) showed no tendency to isomerise to (111). It has been sugge~ted~~*~~ that reduction of the dienone (11; R = H) would afford via the allylic alcohol (IV) an alternative route to thebaine (V) and thence to codeine (VI; R = Me) and morphine (VI; R = H). In fact we find that two allylic 0 alcohols are formed on reduction of the dienone (11; R%:e R = H) with sodium borohydride and that both give thebaine (V) at pH4 (and lower) at room temperature.HOV (VI) The chemical results therefore strongly support the Satisfactory spectral analytical and nuclear mag- modified route (11; R = H) -(N) 3(V). netic resonance data have been obtained for all new Manganese dioxide oxidation of the benzyliso- compounds. [a],are in ethanol. quinoline (I; R = R’ = Me) labelled with tritium gave after dilution of the product with excess of (11; We thank the Deutsche Forschungsgemeinschaft R = H) radioactivity corresponding to a yield of (W.S.) and the Salters’ Institute (G.M.T.) for 0.012% (equivalent to 0.024% of racemate) of Fellowship support. dienone (11; R = H). This yield was confirmed by (Received May 14th 1963.) This result was first reported at the Summer School in Biogenesis Milan Italy Sept.1962 (Barton and Kirby in press). The incorporation of singly labelled precursor (I; R = R = Me) has also been observed by Battersby and his collaborators. “ntley and Robinson Experientia 1950 6 353; Bentley Robinson and Wain J. 1952 958; Stork J. Amer. Chem. SOC.,1951,73 504; 1952,74,768. Pummerer’s ketone Barton Deflorin and Edwards J. 1956 530. Narwedine Barton and Kirby Proc. Chem. SOC.,1960 392; J. 1962 806. Battersby Summer School in Biogenesis Milan Italy Sept. 1962 (Battersby in press). Ginsburg “The Opium Alkaloids,” Interscience Publishers New York 1962 p. 91. loFor earlier syntheses see Gates and Tschudi J. Amer. Chem. SOC.,1952 74 1109; 1956 78 1380; Elad and Ginsburg ibid.1954,76 312; J. 1954 3052. l1 Sir Robert Robinson “The Structural Relations of Natural Products,” Oxford University Press London 1955. The Structureof Potassium Tetranitritomercurate(I1)Nitrate By D. HALLand R. V. HOLLAND DEPARTMENT OF AUCKLAND, (CHEMISTRY UNIVERSITY NEW ZEALAND) YELLOW crystals of potassium mercurinitrite were but details of the analytical procedures were not first prepared by Langl in 1862 by evaporation of a given. Later analyses4 for mercury potassium and solution of mercuric nitrate with an excess of potas- nitrite led to the formula K3Hg(N02),,H20 but sium nitrite. Lang and other nineteenth-century chem- again the method for nitrite analysis was not speci- i~ts~,~ claimed that the formula was K,Hg(NO,), fied nor was any direct evidence (such as weight-loss Lang J.prakt. Chem. 1862 86 295. Rammelsburg Pogg. Annalen 1863 118 249. a Fock 2.Krist. 1889 17,.188. Rosenheim and Oppenheim 2.anorg. Chem. 1901,28 173. JULY 1963 on heating) obtained for the water molecule. Other investigators5 agreed with this formula which appears still to be acceptedV6 We have determined the crystal structure of the compound and shown that it should be formulated as K,[Hg(NO,),]NO,. The crystals are ortho-rhombic with a = 12.12 6 = 10.58 c = 9.28 A there being 4 molecules per unit cell. The axial ratio agrees with previous determination^.^^^ The space group is either Pnma or Pn2,a. The heavy-atom loca- tions were consistent with the centrosymmetric space group and the structure analysis proceeded satis- factorily on that assumption.The mercury atom was located from Patterson projections and the potassium atoms from subse- quent Fourier projections. A three-dimensional difference Fourier synthesis then revealed the light atoms. Four nitrite ions are grouped about the mercury atom in a roughly tetrahedral manner. The near neighbours to the mercury atom are the eight oxygens at 2.4 A but their distribution about the mercury is not at all regular. The remaining four light atoms occur about 5 A from the mercury in an approximately planar trigonal grouping with bond lengths 1.2 A. This can only be interpreted as a nitrate ion. Structure factors calculated on this model at the present stage of refinement (isotropic tempera- ture factors only) show a reliability factor of 17%.Analysis of the crystals by Dr. A. D. Campbell confirms the above composition [Found K 21.4; total N 12.4; NO2- (addition of excess of per- manganate followed by potassium iodide and titra- tion with thiosulphate) 32.6. K3Hg(N02),N03 requires K 20.75 total N 12.0; NO2- 32-6%]. (Received May 30th 1963.) Kohlschutter Bet-. 1902 35 489; Ray J. 1907 91 2031. * E.g. Durrant and Durrant "Introduction to Advanced Inorganic Chemistry," Longman Green and Co. London, 1961 p. 483. Chromium Pentafluoride and Chromium Oxide Tetrafluoride By A. J. EDWARDS (DEPARTMENT THEUNIVERSITY, OF CHEMZSTRY BIRMINGHAM) THEREare three brief reports of chromium penta- flu~ride,'-~~~ in which it is recorded only that the compound is red and volatile at 100".The most recent of these reports3 describes the preparation of the pentafluoride and of chromium hexafluoride by high-pressure techniques. An investigation of some properties of the pentafluoride is now being made and the direct fluorination of metallic chromium has been studied as a convenient preparative method. The red volatile product of the fluorination was separated by vacuum-sublimation into chromium pentafluoride (Found Cr 35.5; F 64.5. Calc. for CrF5 Cr 35.4; F 64*6%),and the previously un- reported chromium oxide tetrafluoride (Found Cr 36.0; F 53.8. CrFO requires Cr 36.1; F 52.8%). Complete separation of the two compounds has been achieved only with difficulty (cf.the pentafluorides and oxide tetrafluorides of molybdenum4 and ruthenium5). Chromium pentafluoride is a crimson solid which melts at 30" to a dark red viscous liquid. The vapour is crimson. The oxide tetrafluoride is very dark red in both the solid and the liquid phase (m.p. 55") and Wartenbern Z. anow. Chem. 1941 247 135. forms a red vapour. Both compounds are immedi- ately hydrolysed by water and dissolve the penta- fluoride forming a yellow-green solution and the oxide tetrafluoride a yellow solution. The two compounds can be distinguished by their X-ray powder and single-crystal photographs. Chromium oxide tetrafluoride has a monoclinic unit cell a = 12.3,b = 5-4,c = 7.3 A fi = 104",whereas the unit cell of the pentafluoride is orthorhombic a = 5.5 b = 7.4 c = 16.3 A.The detailed crystal structures of the two compounds are under investiga- tion. The unit cell of the pentafluoride is very similar to that of technetium pentafluoride6 (a = 5.8 b = 7.6 c = 16.7 A). I am indebted to Dr. R. D. Peacock for his advice and encouragement to the Royal Society for provi- sion of X-ray equipment to Imperial Chemical In- dustries Limited for the loan of a fluorine cell and to the Department of Scientific and Industrial Research for a Research Fellowship. (Received,June 5th 1963.) Huss and Klemm ZI anorg. Chem.; 1950 262 25. Glemser Roesky and Hellberg Angew. Chem. 1963 75 346. Edwards Peacock and Small J. 1962,4486. Holloway and Peacock J.1963 527. Edwards Hugill and Peacock unpublished work. PROCEEDINGS The Structures of Rotundifohe and Mitragynol By G. M. BADGER (DEPARTMENT CHEMISTRY OF ADELAIDE, OF ORGANIC THE UNIVERSITY AUSTRALIA), L. M. JACKMAN (DEPARTMENT UNIVERSXTY AUSTRALIA), OF CHEMISTRY OF MELBOURNE R. SKLARand ERNEST WENKERT OF CHEMISTRY UNIVERSITY, (DEPARTMENT INDIANA U.S.A.) INa recent interpretation of the recorded chemistry folane (111) m.p. and mixed m.p. 231-232" [alas of the Mitragyna alkaloids,l Hendrickson2 suggested 113" (c 0.85). Its p.m.r. spectrum indicated the loss the mitraphylloid structures (I; Y =OH and OMe) of both methoxy-groups and the olefinic 17-H of the for mitragynol and rotundifoline respectively. alkaloids. Finally- its optical rotation p& 5-40,* and However the following observations now prove these thin-layer chromatographic behaviour revealed it as bases to be rhyncophylloid stereoisomers (11).a stereochemical relative of rotundifoline. -Y & \)OH Hde 0 VEt Et (m> Et While no evidence is available for rigorous I assignment of the stereochemistry of the alkaloids at (1) (a) o%d this time some incidental observations reflect on The proton magnetic resonance spectra of the two their steric environments. Thin-layer chromato-alkaloids reveal diffuse multiplets at 0.7-0.9 p.p.m. graphy and pKa results* (mitragynol 7.30 ca. 11 -1 ; (in deuteriochloroform solutions with tetramethyl- rotundifoline 5.25 ; rhyncophylline 7.25; isorhynco-silane as internal standard) characteristic of the phylline 6.05) revealed a close similarity between the methyl-hydrogen atoms of a secondary ethyl group alkaloid pairs mitragynol-rotundifoline and rhynco- as found in dihydrocorynantheine corynantheidine phylline-isorhyncophylline.However thermal equi- and rhyncophylline but in contrast to the distinct libration (without a solvent) of the two sets of bases, methyl doublet exhibited by the proton magnetic involving 3,7-bond rupt~re,~~~ led to rotundifoline resonance (p.m.r.) spectra of ajmalicinoid3 or mitra-exclusively in the case of the first set but a mixture phylloid4(1) alkaloids. Comparison of the aromatic of the two bases (isorhyncophylline predominating) hydrogen signals (6-3-7-2 p.p.m.) in the p.m.r. in the second set.Further after short heating of spectra of the Mitragyna bases with those of the four mitragynol just above its m.p. thin-layer chromato- ar-methoxyoxindoles showed the alkaloids to be graphy indicated starting material and production of 9-oxygenated. The 0-methyl singlets appeared at rotundifoline as well as of a new product of inter- 3.74 and 3.82 p.p.m. in the spectrum of mitragynol mediate mobility. This mixture was transformed and at 3.64 and 3.74 p.p.m. in that of rotundifoline fully into rotundifoline on exposure to 220-250" for while the 17-H singlets showed at 7.40 and 7-39 10 min. Thus while rhyncophylline and isorhynco- p.p.m. respectively.* phylline are merely 7-epimers,"y7 mitragynol and O, Heating mitragynol m.p. 199-201 [a]? -5.6" rotundifoline may be isomeric at both positions (c 1.0 in chloroform) without a solvent at 220-250" 3 and 7.for 10 min. transformed it into rotundifoline m.p. m.m.p. 240-242" [a]? 129"(c 1-0). Acid hydrolysis R.S. and E.W. are grateful to the Ciba Pharma- and decarboxylation of both alkaloids and Wolff- ceutical Company (Summit New Jersey U.S.A.) for Kishner reduction of the resultant aldehydes5 led to financial support of this work. the identical crystalline degradation product rotundi- (Received April 25th 1963.) * The pKa measurements were made on 33% aqueous dimethylformamide solutions. The authors are grateful to Mr. L. Spangle (Eli Lilly and Co.) for these determinations. Barger Dyer and Sargent J. Org. Chem. 1939,4,418; Badger Cook and Ongley J. 1950 867.Hendrickson Chem. and Ind. 1961 713. Wenkert Wickberg and Lejcht J. Amer. Chem. Soc. 1961 83 5037; Shamma and Moss ibid. 1962 84 1739. Wenkert Wickberg and Leicht Tetrahedron Letters 1961 822. Cf. Seaton and Marion Canad. J. Chem. 1957 35 1102. Wenkert Udelhofen and Bhattacharyya J. Amer. Chern. Soc. 1959,81,3763; Seaton Nair Edwards and Marion Canad. J. Chem. 1960,38 1035. Finch and Taylor J. Amer. Chem. SOC.,1962,84 3871. JULY 1963 207 The Green Form of B~~i~phenylpho~~e)~bromo~ckei(~) :an Interailogon Compound By B. T. KILBOURN and J. A. C. DARBYSHIRE H. M. POWELL (CHEMICAL CRYSTALLOGRAPHY LABORATORY, OXFORD UNZVERSITY) A NEW complication in stereochemistry has been found. Examples are already known of nickel@) compounds in which four bonds are directed from the metal atom either towards the corners of a square or towards the vertices of a tetrahedron.Some series of these compounds include both stereochem- ical f~rms.l-~ The occurrence of one form or the other can be related to the nature of the ligands. Further it has been recognised that in a particular compound the difference of stability of the two forms may be so small that each can exist separately in the crystalline For example the compounds bis(benzyldiphenylphosphine)dihalogenonickel(II) where halogen = C1 Br or I were each obtained in two crystalline varieties which differ in colour and magnetic properties one set being diamagnetic and the other paramagnetic. A reasonable assumption might be that the mole- cules in the diamagnetic crystals have square planar nickel bonds and those in the paramagnetic crystals have tetrahedral bonds.The compound bis(benzyldipheny1phosphine)di-bromonickel(n) crystallises in a red diamagnetic form and in a green paramagnetic form of moment 2.7 B.M. A three-dimensional X-ray crystal-structure determination with 3500FRRt values has been made og the green form. The triclinic unit cell space group P1 contains three single molecules. The possibility of a trimer was eliminated. One molecule has its nickel atom in a centre of symmetry and has the trans square planar configuration. The other two molecules are related to each other by a centre of symmetry; their nickel bonds are arranged tetra- hedrally.The whole is a molecular compound Ni(PBzPh,) ,Br2[square],2Ni(PBzPh,) ,Br,[tetra-hedral] where Bz = benzyl Ph = phenyl. Allowance for one third of the molecules being diamagnetic gives a recalculated magnetic moment for the tetra- hedral form of approximately 3-3 B.M. This agrees with the moment1 of the corresponding chloro- compound which in its paramagnetic crystalline form presumably has the tetrahedral arrangement only. Published magnetic moments of a number of related nickel@) compounds are not all easily inter- preted; there has been confusion between tetrahedral and octahedral structures. In view of the results given above magnetic moments must be regarded with further caution if there is a possibility of the sub- stance being a crystalline molecular compound of different magnetic forms.Unambiguous space-group determination may be sufficient to show the equi- valence of all the molecules without a full structure analysis. It was desired to establish firmly that in this com- pound there is a form of isomerism which depends on the existence of two different stereochemical arrangements of a fixed number of bonds to a central atom. Alternative suggestions might be that the two sets of nickel atoms had acquired different ligands or had different co-ordination numbers. Although refinement is not yet complete the R value is already reduced to 13 %; all atoms excluding hydrogen have been located and are at satisfactory separations; no atom in any molecule is within 4 A of a nickel atom in any other molecule and apart from the bonded phosphorus and bromine atoms the closest approaches (Q=3.4 A) to a nickel atom are those of some CH groups of its own ligands; a difference map computed at the R value 13 % shows that at no point in the unit cell is there any electron density that requires further atoms other than hydrogens.The conclusion is therefore that both nickel atoms are 4-co-ordinatedY the ligands in the two forms are identical apart from incidental conformational differences and the isomerism is proved. In this molecular compound there is a fixed ratio between the magnetically different molecules. Certain ratios not necessarily simple could lead in similar cases to false though plausible conclusions con- cerning the number of unpaired electron spins.The compound bears some analogy to other crystalline substances containing fixed ratios of isomers e.g. racemates which do not reveal the true optical activity of their component molecules. Isomers which depend on the existence of two different stereochemical arrangements of the same number of bonds from the same central atom may be called diogons (Gr. alios other different and gonia angle). The new stereochemical complication-the occurrence of two such isomers together in the same crystal-may then be concisely described as in the title. It is possible to foresee a great variety of related molecular compounds in which the two stereochem- ically distinct components differ also in the ligands.Formation of these from suitable mixtures may be more probable than that of the interallogon com- pound itself since special geometrical packing con- Browning Mellor Morgan Pratt Sutton and Venanzi J. 1962 693. Hayter and Humiec J. Amer. Chem. SOC.,1962,84,2004. Goodgame and Goodgame J. 1963 207. ditions may be more readily satisfied if the restriction to identical ligands is removed. Perhaps some normally unobtainable stereochemical forms might be stabilised in crystals of this kind. PROCEEDINGS We thank Miss M. C. Browning and Dr. L. M. Venanzi for the material used and the Oxford University Computing Laboratory for facilities. (Received Maj?17th 1963.) CIeavage Reactions of Pentaffuorophenyl Derivatives of Tin and Boron By R.D.CHAMBERS and T. CHIVERs (DEPARTMENT UNIVERSITY OF CHEMISTRY OF DURHAM,SOUTHROAD,DURHAM) THE preparation of some pentafluorophenyl deriva- tives of tin has been reported recent1y.l Trifluoro- rnethyb2 and to a lesser extent trifluorovinyl- derivatives3 of three-covalent boron are unstable decomposing to boron trifluoride and corresponding polyfluoroaromatic derivatives were hitherto un-known. We now report the preparation of pentafluoro- phenyl derivatives of tin from the appropriate metal halide and the Grignard reagent an unusual halide- ion catalysed hydrolysis of these compounds and their reactions with boron halides to give penta- fluorophenylboron derivatives. Three series of tin compounds have been prepared Me,,Sn(C,F,), Ph4-,Sn(C6F,), where x = 1-4 and (p-tolyl),,Sn(C,F,), where x = 2 and 3.The most remarkable property of these compounds is that when pure they are stable in aqueous ethanol but that they are rapidly hydrolysed on addition of catalytic quantities of halide or cyanide ion giving pentafluorobenzene. The reaction most probably involves initial co-ordination of halide ion to tin followed by hydrolysis. e.g. Me3Sn-C2 + X-+ [Me,(C,F,)SnX]-(X = F Cl CN) H2O 4 Me,Sn.OH +C,F,H +X-+ [Me,C,F,Sn(OH,)X]-The overall relative order of cleavage of groups from tin by the electrophiles hydrogen chloride and boron halides is p-tolyl > Ph > C,F5 > Me. Cleavage of p-tolyl.Sn(C,F,) by hydrogen chloride gives (C,FS),SnC1 exclusively which can be hydrolysed to tris(pentafluoropheny1)tin oxide by ammonium hydroxide without significant cleavage of pentafluorophenyl.Cleavage of the methyl compounds with boron trichloride and trifluoride yields pentafluorophenyl- boron dihalides. No Me$n.C,F + 2BCI3 -+ Me,SnCl + C,Fs.BCI solvent + MeBCl CCl* Me,Sn*C,F + 2BF3-Me3Sn*BF + C,F5*BF2 The latter reaction differs from the reaction of Me3SnCF3 with boron trifluoride which gives [Me3Sn][CF3,BF,3,5 and this indicates that in contrast to trifluoromethylboron difluoride? pentafluoro- phenylboron difluoride is a weaker Lewis acid than boron trifluoride. Pentafluorophenylboron difluoride stored as liquid for one month underwent 40% con-version into boron trifluoride but the dichloride could be distilled (b.p.123-124"/760 mm.) in nitrogen with slight decomposition. Pentafluorophenylboron dichloride forms a stable 1:l complex with pyridine and can be hydrolysed with an exactly equivalent quantity of water in acetone at -80" to pentafluorophenylboronic acid (m.p. 290"). However aqueous ethanol quickly cleaved pentafluorophenyl from the boronic acid and such ready hydrolytic cleavage is quite unknown for other arylboronic acids ;,preliminary results indicate that the hydrolysis is inhibited by acid. These observations indicate that the anion derived from pentafluorophenylboronic acid is unstable readily losing pentafluorophenyl as a carbanion and this would account for the fact that it was not possible to isolate a diethanolamine ester which is usually the most stable derivative for the characterisation of arylboronic acids.' Unlike phenylboronic acid pentafluorophenylboronic acid is not readily de- hydrated to a boroxine even at 140"/0.01mm.We thank Professors C. E. Coates and W. K. R. Musgrave for their interest. (Received May 7th 1963.) 1 Holmes Peacock and Tatlow Proc. Chem. Soc. 1963 108. 2 Parsons Baker Burg and Juvinall J. Amer. Chem. SOC.,1961 83 250. Stafford and Stone J. Amer. Chem. SOC.,1960,82 6238. Nield Stephens and Tatlow J. 1959 166; Brooke Chambers Heyes and Musgrave Proc. Chem. Soc. 1963 94. 6 Chambers Clark and Willis J. Amer. Chem. SOC.,1960 82 5298. Gerrard "The Organic Chemistry of Boron," Academic Press London 1961 and refs.therein. Musgrave and Park Chem. and Ind. 1955,1552; Letsinger and Skoog J. Amer. Chem. SOC.,1955,77,2491. JULY 1963 209 The Structure of Ergoflavin By J. W. APSIMON N. G. CREASEY, J. A. CORRAN K. Y.SIM,and W. B. WHALLEY (THESCHOOL LONDON) OF PHARMACY,THEUNIVERSITY WEhave previously1 defined the general properties and functional groups of the pigment ergoflavin (I). Additional work in conjunction with the following Communication,2 establishes the complete structure as (I; R = OH R' = H). Oxidation with potassium permanganate gives (-)-methylsuccinic acid and alkali degradation of tetra-0-methylergoflavinl (11) [a] -28.6" (in CHCl,) furnishes a biphenyl derivative1 which has now been defined as (111). The structure of this pro- duct combined with the general characteristics and spectral properties including the nuclear magnetic resonance (all in CDCl,) of compounds (I) and of (11) which includes a doublet at T 8.84 (J 6 c.sec.) (:CHMe) a pair of doublets at T 2.48 and 3.02 (J 8 c. sec.) (two pairs of ortho-aromatic protons) and two extensively coupled methylene residues clearly show that ergoflavin is a symmetrical mole- cule derived from two identical Cl5Hl30, units (cf. ref. 1). The production of the biphenyl (III) together with the shift to higher frequencies of the o-hydroxy- carbonyl absorption band in the infrared spectrum of e.g. ergoflavin vmax. 1645 cm.-l and di-0- methylergoflavin,' vmax 1620 cm.-l to the 1680 cm.-l region in the ether GI) and its analogues (cf.the similar behaviour of o-hydroxyacetophenone~~ and of 5-hydroxyfla~anones~) clearly indicates the Presence of two ~h~~~anone (rather than chromone) rings in ergoflavin. [~~ [*9 '..Me~ ..& 3\ 7 40 \ c4-I (a t;i (1) 2 HH Me-co O W\O CO-Me ~~. OMe OMe OV) (a * Oxidation of the ether (11) with Jones's reagent5 gives a tetraketone (IV) m.p. 290" (decomp.) [a] + 190" (in CHCl,) vmax 1733 1800 cm.-l. Its degradalton by alkali yields the biphenyl (111) (60%) and oxalic acid hence it must contain the residue (V) (cf. rotenonones) in which the carbonyl group marked * is derived from the XH-OH residue of the ether (11). The downfield shift of the CMe doublet in the ketone (IV) T 8.74 (J 6 c.sec.) com- pared with the ether (11) shows that this residue is adjacent to the 5-carbonyl group. Since the methylene residue must be adjacent to the :CHMe grouping [to produce methylsuccinic acid] and the terminus of the y-lactone must be located on the remaining carbon atom it follows that ergoflavin may be repre- sented as (I; R = OH R' == H) or (I; R = H R' = OH). An unequivocal decision between these alter- natives could not be made. This structure is com- patible with the high y-lactone frequency in these compounds (vmax. ca. 1800 cm.-l) with vmax. 1733 cm.-l of the 5-carbonyl group in the ketone (IV) and with the considerable shift in [aIDwhich accompanies the transformation of the ether (11) into the ketone (IV). The isolation of (-)-methyl- succinic acid defines the absolute stereochemistry at position 6 since the 5-hydroxyl group in the ether (11) is sterically hinderedl it must be axial.Definition of the structure of ergoflavin as (I; R = OH R' = H) and of the absolute stereochem- istry have been provided by a complete three- dimensional X-ray analysis2 of the di-(p-iodo-] benzoate) of the for their co-operation. The unusual whom we thank ether (11) by the Glasgow group acidity of the alcoholic 9-hydroxyl group is attributed to the diaxial interaction with the 5-hydroxyl group (cf. the similar behaviour of methyl picrotoxate'). Ergoflavin is thus the first representative of a new class of bixanthonyl derivatives several complex .~2 pigments associated with ergoflavin appear to repre- sent variants of the ergoflavin molecule the defini- tion of their structures will be reported shortly.We believe that the co-occurrence* of ergoflavin and the anthraquinone endocrocin is of biogenetic significance since ergoflavin could be derived from endocrocin by an unexceptional and acceptable bio- genetic sequence (cf. ref. 9). The occurrence of fungal Eglinton King Lloyd Loder,.Marshall Robertson? and Whalley J. 1958 1833. Asher McPhail Robertson Silverton and G. A. Sim following Communication. Cf. Baker Finch Ollis and Robinson J. 1963 1480. Unpublished information from this laboratory. Bowers Halsall Jones and Lemin J. 1953 2548. @ La Forge J. Amer. Chem. SOC.,1932 54 3377. Burkhill Holker Robertson and Taylor J.1957 4945. Franck and Reschke Chem. Ber. 1960,93,347. Gatenbeck Svensk kem. Tidskr. 1960,72 188. PROCEEDINGS bianthraquinonylslO is relevant in this context. Smith Ltd. Edinburgh for certain extraction We thank Messrs. Burroughs Wellcome & Co. for facilities and the Wellcome Trust for financial the provision of crude pigment Messrs. T. & H. support. (Received May 31st 1963.) loCf. Miller “The Pfizer Handbook of Microbial Metabolites,’’ McGraw-Hill Book Co. Inc. New York 1962 pp. 267-271. T&eStructure of Ergoflavin By J. D. M. ASHER,A. T. MCPHAIL ROBERTSON and G. A. SIM J. MONTEATH J. V. SILVERTON DEPARTMENT GLASGOW, (CHEMISTRY THEUNIVERSITY W.2) A NUMBER of crystalline pigments have been isolatedl from the colouring matter of ergot the sclerotia produced by the fungus Claviceps purpurea when grown on rye.One of the pigments ergoflavin C30H26014, first isolated in 1912 by Freeborn,2 has since been investigated intensively by Professor W. B. Whalley and his c~llaborators.~~~ In order to estab- lish unambiguously the constitution and stereochem- istry of ergoflavin we undertook at the suggestion of Professor Whalley an X-ray crystal-structure analysis of tetra-0-methylergoflavin di-p-iodo-benzoate. Our results define the structure of the iodobenzoate as (I). The absolute configuration I-Y i I X The second three-dimensional electron-density dis- tribution over one molecule of tetra-o-methylergo-flavin di-p-iodobenzoate shown by mems of super-imposed contour sections drawn parallel to (0015.shown was determined by Bijvoet’s anomalous- dispersion method4 and by oxidation of ergoflavin to (-)-methylsuccinic acid.3 The di-p-iodobenzoate crystallises in the ortho- rhombic system space group P2,2,2 (I);),with four molecules of C48H,120, in a unit cell of dimensions a = 13.23 b = 38.70 c = 9.37 A. From equi- inclination Weissenberg photographs about 2900 independent structure amplitudes were evaluated. The co-ordinates of the iodine atoms were obtained initially from Patterson syntheses and the 64 carbon and oxygen atoms in the molecule were then located in three-dimensional electron-density distributions. Superimposed contour sections taken from the second electron-density distribution and covering the region of one molecule are shown in the Figure.The value of R is now 21 % and refinement of the atomic co-ordinates is continuing. The symmetrical nature of ergoflavin makes it likely that oxidative phenolic coupling5 is involved in the biogenesis of ergoflavin (cf. ref. 3); the phenolic precursor appears to arise from a ,k.?-poly- keto-acid chain6 folded as in (II). a) For the calculations on the Glasgow University DEUCE computer programmes7 devised by Dr. J. S. Rollett and Dr. J. G. Sime were employed. We are grateful to Dr. G. Eglinton for helpful discussions. (Received May 3 1st 1963 .) For review of earlier work see Eglinton King Lloyd Loder Marshall Robertson and Whalley J. 1958 1833. Freeborn Pharm.J. 1912,88 568. ApSimon Corran Creasey K. Y. Sim and Whalley preceding Communication. * Bijvoet Peerdeman and van Bommel Nature 1951 169 271. ti Barton and Cohen “Festschrift A. Stoll,” Birkhauser Basle 1957 p. 117. Birch Proc. Chem. Soc. 1962 3. ’ “Computing Methods and the Phase Problem in X-Ray Crystal Analysis,” ed. Pepinsky Robertson and Speakman Pergamon Press Oxford 1961; Rollett p. 87; Sime p. 301. JULY 1963 21 1 The Absolute Configuration of Metal Complexes from the Optical Rotatory Power of the Ligand Transitions By A. J. MCCAFFERY and S. F. MASON (CHEMISTRY THE UNIVERSITY DEPARTMENT OF EXETER) THE optical methods proposed hitherto for the assignment of absolute configuration to dihedral transition-metal complexes have been based upon the sign of the optical rotatory power of the metal- ion d-+dabsorption bands in the visible-wavelength regi0n.l Such methods are empirical and they relate the absolute configuration of a given metal complex to that of the (+)-Co ion determined by X-ray diffraction.2 In a number of metal complexes con- taining conjugated ligands such as 1 ,lo-phen- anthroline (I) the metal-ion d-td absorption bands are not observed as they are overlaid by the more intense charge-transfer bands (see Figure).However the absorption bands of the free ligand are preserved largely unchanged in such complexes and the helical disposition of the ligands in the complex confers an optical rotatory power upon the ligand transitions.A non-empirical method for determining the chirality of metal complexes containing conjugated ligands is now proposed based upon the sign of the rotatory power of the ligand transitions in the complex with- out reference to a standard molecule having a known absolute configuration. The electronic transitions responsible for the moderate and the high-intensity absorption bands of aromatic molecules in the near-ultraviolet region are polarised in the molecular plane? In 1,lO-phen-anthroline (I) such transitions are either y-polarised along the two-fold symmetry axis or x-polarised along the long axis of the molecule. The y-polarised transitions in a tris-1,lo-phenanthroline-metal com-plex (11) have no zero-order rotatory power as they have only an electric moment although they acquire a small first-order rotational strength by mixing with magnetic-dipole d4d transitions of the metal ion whereas the x-polarised transitions have both an electric and a magnetic moment and thus a large zero-order rotational ~trength.~ The circular dichro- ism spectrum (Figure) of the complex [II; M = Ru(II)] indicates therefore that the a-and the @-band of 1,lO-phenanthroline at 3000 and 2300 A respectively are y-polarised whilst the p-band at 2700 A is x-polarised in conformity with molecular- orbital theory.The p-band excitations of the 1,lO-phenanthroline ligands couple in the complex (II) to give two electronic transitions one with A2 symmetry in D3 directed along the three-fold axis of GI) and the other with E symmetry directed in the plane per-pendicular to the C3 axis of 01).The particular geometry of the octahedron and coupled oscillator theory indicate that the dipole strength D and the rotational strength R,of the A and the E transition derived from the p-band excitations of the ligands in the complex (11) are related by Wavelength (x) 5000 4000 3000 2500 I 4 -20 0 3* CI 2 i -7 20,000 v (cm:') 40,000 The circular dichroism (--) and the electronic absorption spectrum (----) of (-)-tris-(1,lO-phenanthroline)ruthenium(n) perchlorate in aqueous solution. Ballard McCaffery and Mason Proc. Chem. Soc. 1962 331 and refs. therein. Nakatsu Shiro Saito and Kuroya Bull. Chem. Soc. Japan 1957 30 158.Mason Quart. Rev. 1961 15 287. Mason Quart. Rev. 1963 17 20. D(A2) = 2D(E) = 2D(l,lO-phenanthroline). .(1) and R(A2) = -R(E) * * (2) where D(l ,lO-phenanthroline) is the dipole strength of thep-band given by the free ligand. The large negative circular-dichroism band at 37,500 cm.-l in the spectrum of the complex [II; M = Ru(II)] lies at the same frequency as the extinction maximum of the p-band of the complex whilst the large positive circular dichroism band at 38,700 cm.-l equal in area to the adjacent negative band lies at a frequency where the extinction coefficient of the Orgel J. 1961 3683. PROCEEDINGS p-band in the complex has one-half of its maximum value indicating from equations (1) and (2) that the large negative and positive circular dichroism bands are due to the A and the E transition respectively.These assignments are supported by Orgel’s coupling coefficients for n-bonding in metal chelates contain- ing conjugated ligand~.~ Since the A2 transition has a negative rotational strength the ligands in (-)-Ru( 1 ,lO-phenanthroline),2+ are disposed in the form of a left-handed helix DI; M = Ru(II)] the complex having the same chirality2 as (+)-Co en3%. (Received May 29th 1963.) T~s~n~fluoropheny1)~ron By A. C. MASSEY, A. J. PARK,and F. G. A. STONE OF CHEMISTRY QUEEN LONDON, (DEPARTMENT MARYCOLLEGE E.1) KNOWLEDGE compounds having fluorocarbon of groups bonded to boron1 is limited and the few known substances of this class all release boron tri- fluoride relatively easily accounting for difficulties encountered in preparation as well as unsuccessful attempts to isolate other fluorocarbon-boron com-pounds.We now report tris(pentafluorophenyl)boron which although air-sensitive is thermally stable at ambient temperatures and from which numerous other pentafluorophenylboron compounds can be obtained. Initially we attempted to obtain (C6F5)3B by treating pentafluorophenyl-lithium prepared in ether,2 with boron trichloride but difficulties were encountered in attempting to remove the solvent. This problem was overcome by using pentafluoro- phenyl-lithium as a fine suspension in pentane. In- deed formation of the lithium reagent in this medium rather than in ether is advantageous in other syntheses.Under nitrogen n-butyl-lithium in pentane was slowly added to a vigorously stirred pentane solution of pentafluorophenyl bromide. Boron trichloride (BC13:C6F5Li = 1:3.5) was added to the stirred suspension at -78” and after 5 min. the mixture allowed to warm slowly to room temperature. Separation of the lithium chloride afforded an air-sensitive solution of tris(pentafluoropheny1)-boron. Addition of an equimolar ether solution of pentafluorophenyl-lithium to this pentane solu- tion immediately precipitated air-stable water-soluble lithium tetrakis(pentafluorophenyl)boronate Li( C,F,),B. Tris(pentafluoropheny1)boroncan be obtained as a white solid by evaporation of its pentane solution. It will sublime at 150” in vacuo but simultaneous decomposition is observed.Tris(pentafluoropheny1)-boron shows marked acceptor properties a,s indicated in the chart. / in pentane\ (C6F5)3B vii C6F5H M%m(C6F5)3 LE 4N][(C6F5)4B] i NH,. ii HCI (gas) at 120”. iii Me,N. iv Ph,P. v C,F5Li in Et,O. vi aqueous Et,NCI. vii aqueous KCI. We are indebted to the U.S. Department of the Army for support of this research through its European Research Office. (Received May 9th 1963.) Treichel and Stone “Advances in Organometallic Chemistry,” Academic Press N.Y. Vol. I in press; Goubeau and Rohwedder Annalen 1957 604 168; Parsons Baker Burg and. Juvinall J. Amer. Chem. Soc. 1961 83 250; Stafford and Stone ibid. 1960 82 6238; Chambers Clark and Wlllis ibid. 1960 82 5298; and Chambers Clark Reeves and Willis Canad.J. Chem. 1961 39 258. Coe Stephens and Tatlow J. 1962 3227. JULY 1963 213 The Preparation of t-Butyl Hypochlorite By C. P. C. BRADSHAW and A. NECHVATAL (QUEENS’ UNIVERSITY DUNDEE) COLLEGE OF ST.ANDREWS THE preparation of t-butyl hypochlorite by the inter- action of chlorine and t-butyl alcohol has been described and reference made to the unstable nature1* of the product. We now record that after two uneventful prepara- tions following the procedure described,l a third re- action on a 1-2-mole scale exploded towards the end of the reaction; the vessel was shattered fume cup- board glass broken and one of us (C.P.C.B.) cut about the hands by fragments of flying glass. Urg.Synth. 1952 32 20. Chem. Eng. News 3962 October 22nd p. 62. As a result of our experience we suggest that careful control of the temperature of the reaction mixture is necessary. The reaction vessel should be fitted with a thermometer that dips into the reaction mixture. The rate of flow of chlorine must be regulated so that the temperature of the reaction mixture never exceeds 20° a point which is not emphasised in the original description. With these precautions subsequent preparations on a 1 e4-mole scale have proved uneventful. (Received May 27th 1963.) Orientation Reactions of Chloropentafluorobenzeneand Related Compounds R. D. CHAMBERS, By G. M. BROOKE J. HEYES,and W. K. R. MUSGRAVE (THEDURHAM IN THEUNIVERSITY COLLEGES OF DURHAM) THEREare many examples1 of further nucleophilic substitution of polyfluoro-aromatic compounds C6F5X (X = H Me NH, NHMe NO, and OMe) but in only one of these the action of ammonia on C6F,*N0,,2has any significant amount of the ortho- isomer been formed.We now report the reactions of chloropentaflu~robenzene~ with four nucleophilic reagents (LiAlH, NH, NaOMe and NH,.NH,). Fluorine replacement occurred in each case giving in high overall yield mixtures of all three isomers the approximate ratio in each case being o :m:p = 25 :5 :70. Since the chlorine atom is still available for Grignard formation3b this now makes possible the preparation of ortho-difunctional tetrafluorobenz- enes. The isomer proportions were determined by vapour-phase chromatography ortho- and para-isomers being isolated on a preparative scale; structures were then determined unambiguously by nuclear magnetic resonance spectroscopy.The pro- duct obtained with hydrazine was reduced by hydriodic acid4 to a mixture of chlorotetrafluoro- anilines identical with that obtained from chloro- pentafluorobenzene and ammonia. Only in the reaction between chloropentafluoro- benzene and lithium aluminium hydride was replace- ment of chlorine observed; in addition to the isomeric chlorotetrafluorobenzenes pentafluoro-benzene and 1,2,4,5-tetrafluorobenzenewere ob-tained. The proportion of the last two compounds depended on the reaction conditions. 1,2,4,5-Tetra- fluorobenzene can arise from further reaction between pentafluorobenzene and the nucleophilel or from the replacement of chlorine in the initially formed 1-chloro-2,3,5,6-tetrafluorobenzene.The lat- ter undergoes chlorine replacement predominantly when treated with lithium aluminium hydride. l-Chlor0-2,3,5,6-tetrafluorobenzeneand 4-chloro- 2,3,5,6-tetrafluoroanisoleform Grignard reagents3b which with acid give the known 1,2,4,54etrafluoro- benzene1 and 2,3,5,6-tetrafluoroanisole,re~pectively.~ 4-Chloro-2,3,5,6-tetrafluoroanisolereacts with more sodium methoxide to give 1 -chloro-2,3,5-trifluoro-4,6-dimethoxybenzene thus illustrating again al- though it may not be the only factor involved the artho-directing power of the chlorine atom since further reaction of pentafluoroanisole with sodium methoxide gives predominantly tetrafluoro-l,4-dimethoxybenzene.6 We thank Dr.J. W. Emsley for the nuclear magnetic resonance measurements and the Imperial Smelting Corporation for a maintenance grant (to J.H.). (Received May loth 1963.) Brooke Burdon and Tatlow J. 1962 3253 and earlier papers cited. Brooke Burdon and Tatlow J. 1961 802. (a)Mobbs and Musgrave Chem. and Inn. 1961 1268; (b)Brooke Chambers Heyes and Musgrave Proc. Chem. SOC.,1963,94. * Birchall Haszeldine and Parkinson J. 1962,4966. Stephens and Tatlow Chem.and Ind. 1957,821;Brooke Forbes Richardson and Tatlow forthcoming publication. Godsell Stacey and Tatlow Nature 1956 178 199; Nield and Tatlow Tetrahedron 1960 8 38. PROCEEDINGS Copaene By G. B CHI and S.H. FEAIRHELLER DEPARTMENT INSTITUTE CAMBRIDGE, (CHEMISTRY MASSACHUSETTS OF TECHNOLOGY MASS.,U.S.A.) and P. DE MAYOand R. E. WILLIAMS DEPARTMENT OF WESTERN ONTARIO, (CHEMISTRY UNIVERSITY ONTARIO,LONDON CANADA) COPAENE is a tricyclic unsaturated hydrocarbon of periodate followed by potassium permanganate in the cadalene series the structure of which has been acetone (MgSO,) gave the keto-acid the methyl partially defined by amongst other transformations ester (CH,N2) [a] + 27" of which was further conversion into (-)-cadinene dihydroch1oride.l A oxidised with peroxytrifluoroacetic acid to the di- number of structural suggestions have been made,2,3 ester [a] + 31 ".Alkaline hydrolysis and oxidation as for instance (I) which clearly were to be rejected with potassium permanganate then gave (VII) since they necessitated non-Markownikoff opening (vmax.1780 1720 and 1420 cm.-l) which was of the cyclopropane ring in the formation of cadinene esterified (CH2N,) to the corresponding ester dihydrochloride. The recent suggestion3 that copaene (Vmax. 1775 and 1740 cm.-l) [a],+ 41 ". contained a cyclopropane methylene group would indicate structure (11) or (111) rather than (I) but the present work has rendered these observations so far as copaene is concerned irrelevant; evidence is here presented that copaene has the structure and absolute stereochemistry shown in (IV). The copaene used [a] -6.3" was isolated (P. de M. and R.E.W.) from cedrelu toonu Roxb.* and (G.B.and S.H.F.)from a chloranthus oil (chloranthus p spicutus ?) of Chinese origin.The hydrocarbons from the two sources were identical and gave the same (viI I) crystalline diol m.p. 73-73.5" [a] + 0.3" (V; R = H OH) with osmium tetroxide. The hydro- carbon gave no spectral evidence of cyclopropane / ('") Ill hydrogen but was optically transparent in the ultra- violet region confirming the presence of the remain- ing double-bond equivalent as an alicyclic ring. Oxidation (Cr0,-pyridine) gave the corresponding ketone (V; R = 0),m.p. 71-72" [a] + 28" whose absorption (vmax 1718 and 1408 cm.-l) in- dicated a cyclohexanone with a flanking methylene group. \ Treatment of the alcohol (V; R = H OH) with refluxing formic acid gave an optically active phenol (Amax 280 mp E 2300) characterised as the 33-dinitrobenzoate.All physical data were indicative of the structure (VI) and in particular the ester n.m.r. spectrum showed the presence of two singlets (J small for paru-coupling) at T 2.82 and 3-05.This suggested mechanistically that one terminus of the remaining These facts together with the known absolute ring was a to the tertiary hydroxyl in (V; R = H stereochemistry of (-)-cadinene dihydr~halides~ OH). The other terminus was indicated by the (derived from copaene by a process not involving presence of a singlet methyl peak in the n.m.r. of C8) lead to (IV) as the representation of copaene. (V; R = H OH) for the CI methyl. This substance is possibly formed biogenetically by Oxidation of (V; R = H OW) with sodium meta- cyclisation in the ion (VIII); this same ion is a sug- For a summary of early work see Simonsen and Barton "The Terpenes," Vol.111 Cambridge Univ. Press 1952 p. 88. Briggs and Taylor J. 1947 1338. Vonasek Herout and Sorm Coll. Czech. Chem. Comm. 1960 25 919. * Copaene has been previously isolated from this source and characterised as copaene keto-carboxylic acid (and methyl ester) semicarbazone; cf. Pillai and Rao J. SOC. Chem. Ind. 1931,220~. Sykora Herout and Sorm Coll. Czech. Chem. Comm. 1958,23 2181. JULY 1963 215 gated intermediate in the genesis of helmintho-and another (R.E.W.) to the National Research sporal.6 Council of Canada for a Bursary. G.B. and S.H.F. thank Firmenich and Cie Geneva for hancial One of us (S.H.F.) is indebted to the National support and Dr.K. Weinberg for preliminary Science Foundation for a Predoctoral Fellowship experiments. (Received May 24th 1963.) 'Mayo Robinson Spencer and White Experientia 1962 18 359. A Large Primary Hydrogen Isotope Effect in the Mercuration of Benzene By A. J. KRESGE and J. F. BRENNAN (DEPARTMENT CHICAGO 16 ILLINOIS) OF CHEMISTRY ILLINOIS INSTM"J5 OF TECHNOLOGY A NUMBER of primary hydrogen isotope effects have effects are also observed in the reaction of 1,3,5-tri-but not with recently been found in electrophilic substitution of t-butylbenzenesb and 4-bromodureneGC benzene derivatives. We now report the first instance benzene of a primary isotope effect for the reaction of The present work on mercuration shows that benzene itse1f.l The mercuration of benzene can be another factor is the strength of the new carbon- carried out heterolytically under conditions where electrophile bond.Thecarbon-mercury bond is known the substituting agent is positively charged mercury? to be unusually weak estimates of the bond-dis- In this situation hexadeuterobenzene is mercurated sociation energy lie in the range 13-52 kcal./mole.8 considerably more slowly than ordinary benzene This is much below the strength of a carbon-k,/k = 6-0& 0.1 at 25O.4 This indicates that proton hydrogen bond and it is reasonable that in mercura- removal occurs in the rate-determining step. tion v-l be larger than v2 even in absence of a steric It is generally accepted that electrophilic substitu- effect. Some additional evidence to support this tion in aromatic substrates occurs by a two-step influence of bond strength on the relative values of mechanism with a phenonium ion intermediate.5 v- and v2 can be adduced from isotope effects on other substitution reactions.Sulphonation and *I iodination are the only two common electrophilic ArH + X+ -~1 v2 ArHX+ -j. AxX + H+ aromatic substitution reactions which are reversible V-1 and in which therefore the two bonds carbon- On this mechanism the primary isotope effect is a hydrogen and carbon-electrophile are of similar relative measure of the rates v,~and v,; it indicates strengths. It is significant that these two reactions which of the two bonds carbon-hydrogen or account for more than half the reported isotope carbon-electrophile in the intermediate phenon- effects in electrophilic aromatic substitution.Similar- ium ion is the more easily broken. It seems worth- ly the bromination of 3-bromodurene shows an while to inquire what the factors are which govern isotope effect whereas chlorination of the same mole- the relative ease of these two processes. cule does and carbon-chlorine bonds are The suggestion has already been made that steric stronger than carbon-bromine bonds. hindrance is one of these factors.6 In diazo-coupling an isotope effect is observed only when the reaction This investigation was supported by a Public site is crowded and there is obstruction to formation Health Service Research Grant. of a nitrogen-aromatic carbon bond.6u Isotope (Received May 2nd 1963.) l It has been reported (ref.2) that [3H]benzene reacts with fuming sulphuric acid appreciably more slowly than ordinary benzene but the possible existence of a primary isotope effect in this reaction has never been confirmed. Melander Arkiv Kemi 1950 2 21 1. (a) Schramm Klapproth and Westheimer J. Phys. Colloid Chem. 1951,55,843 ;Brown and McGary jun. J. Arner. Chern. Suc. 1955,77,2306; (6) Kresge and Brown to be published. * We had first indication of this phenomenon when we studied aromatic mercuration in acetic acid solution (ref. 36). Since then Westheimer and Perrin have reported a similar isotope effect on mercuration in a different solvent (Abs. 144th National Meeting her.Chem. SOC. Los Angeles Calif. April 1963 49~). Kresge and Chiang J.Arner. Chem. Soc. 1961,83,2877; however for an apparent exception see Olah Kuhn and Flood ibid. p. 4571 4581. (E) Zollinger Helv. Chim. Acta 1958 41 2274; (6) Myhre Acta Chern. Scand. 1960 41 219; (c) Baciocchi, Illuminati and Sleiter Tetrahedron Letters 1960 No. 23 30. de la Mare Dunn and Harvey J. 1957 923. * Sidgewick "The Chemical Elements and their Compounds," Oxford. 1950,298 ;Gowenlock Polanyi and Warhurst Proc. Roy. SOC.,1953 A 218 269. 2 16 PROCEEDINGS A Short Hydrogen Bond in a Basic Salt By H. H. MILLS and J. C.SPEAKMAN (CHEMISTRY DEPARTMENT W.2) THE UNIVERSITY GLASGOW COO^ has recorded the infrared spectrum of the structure analysis3 based on two projections and pyridine 1-oxide salt (C5H5NO),HAsF, and de- giving 0-0= 2.5 A.One of us (H.H.M. at the duced from it that this material contains a very short Roswell Park Memorial Institute Buffalo) has now hydrogen bond which may be symmetrical. HadZi2 collected 1,374 three-dimensional data by counter had previously reported a similar spectrum with 2- methods and a refinement is in progress. These data picoline 1-oxide hemihydrobromide (C6H7NO) ,HBr already gave an R-value of 21.8 %when tested against and made a similar prediction. At his suggestion the structural parameters derived from the two-dimen- we have been studying the latter salt by X-ray sional work. We conclude that Hadii’s prediction is diffraction. The unit cell belonging to the space correct and that a short and probably symmetrical group C2/c contains four molecules.This evidence hydrogen bond is present here as in certain acid that the crystal includes the symmetrical unit salts.4 [C,H,NO* .H- -oNC6’H7]+ was supported by a (Received May 25th 1963.) Cook Chem. and Ind. 1963 607. Hadii J. 1962 5128. Mills Thesis Glasgow 1962. Speakman and Mills J. 1961 1164 and earlier papers. Biosynthesis of the Amaryllidaceae Alkaloids. Part IV.l The Incorporation of Cinnamic p-Coumaric and Caffeic Acids into Haemanthamine and Lycorine By R. J. SUHA~~LNIK and J. ZULALIAN LABORATORIES OF BIOCHEMISTRY EINSTEIN CENTER, (RESEARCH DEPARTMENT THE ALBERT MEDICAL PHILADELPHIA, PENNSYLVANIA) and protocatechuic aldehyde are Compound trans- [3J4C] p-[3-14C] ~HENYLALANINE incorporated into the Amaryllidaceae alkaloids as Cinnamic acid Coumaric acid c,-c1 unit~.l-~ The incorporation of trans-Activity Activity [3-14C]cinnamic acid p-[3-14C]coumaric acid and (mpc/mmole) (mpc/mmole) tritium-labelled caffeic acid into haemanthamine and 10-6 2.82 lycorine by whole plants and floral primordia have Haemanthamine We now report that cinnamic acid Oxohaemanthamine been rep~rted.~ 10.6 2.8 1 and p-coumaric acid are directly incorporated into Oxohaemanthamine methiodide 10.6 -haemanthamine and serve as C,--Cl units.The administration of the radioactive compounds and N-Methyl-N-( 6-phenyl- the isolation and degradation of haemanthamine piperony1)glycine 11.4 3.43 have been described.lP2 The results of the degradation sodium salt of haemanthamine show that 100% and 86 % of the 2-Me t h yl-4,5-me thylene- 10-7 2-42 radioactivity from the [3-14C]cinnamic acid and dioxybiphenyl acid, p-[3-14C]~~~mari~ respectively reside in 2-methyl-4,5-methylenedioxybiphenyl(see Table).coumarin. The results obtained from our experiments Neish5 points out that the phenylalanine cinnamic provide the first direct evidence that cinnamic acid acid and p-coumaric acid are important precursors and p-coumaric acid also play an important role in in the biosynthesis of lignin flavanoids coumarins the biosynthesis of alkaloids. That caffeic acid and phenolic glucosides. It has been demonstratedG8 appears to be incorporated into lycorine lends addi- that phenylalanine cinnamic acid and hydroxylated tional support to the importance of the hydroxylated cinnamic acids serve as precursors of lignin and cinnamic acid pathway for the biosynthesis of ring A Part 111 Suhadolnik Fischer and Zulalian Proc.Chem. Soc. 1963 132. Suhadolnik Fischer and Zulalian J. Amer. Chem. SOC.,1962 84 4348. Wildman Battersby and Breuer J. Amer. Chem. SOC.,1962 84 4599. Suhadolnik Fischer and Zulalian Biochem. Biophys. Res. Comm. 1963 11,208. Neish Ann. Rev. Plant Physiol. 1960 11 55. McCalla and Neish Canad.3. Biochem. Physiol. 1959,37 537. Kosuge and Conn J. Biol. Chem. 1961,236 1617. Brown Science 1962 137 977. JULY 1963 and the benzylic carbon atom of the Amaryllidaceae alkaloids. Enzymes involved in the deamination of phenyl- alanine and tyrosine to cinnamic acid and p-coumaric acid respectively have been rep~rted.~tl~ We have succeeded in isolating from the meristematic tissue of N.pseudonarcissus the enzyme that deaminates phenylalanine to cinnamic acid.* In view of the many intermediates that have been reported in the metabolism of aromatic compounds Koukol and Conn J. Biol. Chem. 1961,236,2692. lo Neish Phytuchemistry 1961 1 1. in higher plants additional enzyme studies are needed to provide information about the exact nature of the compounds involved in the bio-synthesis of these alkaloids. One of us (R.J.S. Career Development Awardee) thanks the National Science Foundation and the United States Public Health Service for support of this work. (Received May 13th 1963.) The Dehydration of Dimethylphenanthrene-9,lO-dioI:A New Route to Quinodimethanes By IANT.MILLAR and K. V. WIL,SON (DEPARTMENT THEUNIVERSITY, OF CHEMISTRY KEELE,STAFFS.) QUINODIMETHANES have been postulated as inter- mediates in several reacti0ns.l in particular Cava et aL2have shown that the 1,2- and 2,3-naphthoquino- dimethanes are intermediates in the formation of cyclobut [a]- and cyclobut [b]-naphthalene by pyro- lysis of the corresponding sulphones. We have dehydrated 9,lO-dimethylphenanthrene-9,lO-diol (I)over alumina in boiling diethyl phthalate and have isolated the adduct (11) of the resulting 9,10-phenanthroquinodimethane and maleic an-hydride. The structure of the adduct (Il) m.p. 301-302" has been confirmed by aromatisation and decarboxy- lation to give triphenylene.The sulphone (IlI) m.p. 246-249" (decomp.) which shows strong infrared absorption at 1300 (S= 0symmetric stretching)? 1 120 (S= 0 asymmetric stretching) and 752 and 719 cm.-l (4 adjacent aro- matic C-H) has been synthesised by the route shown. In boiling diethyl phthalate containing maleic an- hydride it gave the adduct (ll) identical with that obtained from the diol (I). We have also examined the pyrolysis of 9,lO-di- methylphenanthrene the reaction of 9,lO-bischloro-methylphenanthrene with magnesium in tetrahydro- furan and the thermal decomposition of trimethyl- 10 -methyl -9 -phenanthrylmethyl ammonium hydroxide in various conditions all these reactions appear to yield the p henan t hr oquin odime t hane since its spirodime? and higher polymers of it were isolated.The general scope of the dehydration of a vicinal tertiary diol to give a quinodimethane is now being investigated especially in the naphthalene and anthracene series. We are indebted to Unilever Ltd. for a Research Studentship to K.V.W. (Received May 31st 1963.) Mann and Stewart J. 1954,2826; Errede J. Amer. Chem. Soc. 1957,79,4952; Tedder Ann. Reports 1961,58,225. Cava and Shirley J. Amer. Chem. SOC.,1960 82 654; Cava Shirley and Erickson J. Org. Chem. 1962 27 755. Gardner and Hossein Sarrafjzadeh J. Arner. Cham. SOC.,1960 82,4287. PROCEEDINGS Crystal and Molecular Structure of Cyclohexane-1,edione By P. GROTH and 0. HASSEL (CHEMISTRY THE UNIVERSITY NORWAY) DEPARTMENT BLINDERN-OSLO CONSIDERABLE interest attaches to the fact that cyclohexane- 1,4-dione exhibits an electric dipole moment in solution and that the molecules cannot therefore all be present in a centrosymmetrical con- f0rmation.l A two-dimensional crystal structure analysis (which will be followed later by a three- dimensional analysis) has already revealed that in the crystal a twisted "boat" form of the molecule is present with an angle between the two C-0 bonds of about 152".The crystals are monoclinic the space group is P2, and the unit cell containing two molecules has the parameters a = 6.96 b = 6-34 c = 6.73 A p = 99". The atomic co-ordinates deduced from Fourier maps with projections along [loo] [OlO] and [Ool] lead to a mean value of the C-C bond length of 1.54 A and of the C-0 bond length of 1.23 A.The mean value of the two CC(O)-C angles is 118" that Hassel and Nieshagen Tidrskr. K'emi 1930 10 81. Le Fhvre and Le Fhvre J. 1935 1696. of the four C(0)C@3,)C(H2) angles 110". The angle between the two C-0 directions (152") is in good agreement with that to be expected from the dipole moment found2 in solution (1-3 D) if all molecules have the same configuration. The R factors for the three zones mentioned above are 0.09,0.09,and 0.1 1 respectively. It is easily seen that distances between neighbour- ing hydrogen atoms will be more acceptable in the twisted than in the more symmetrical classical boat configuration but the problem still remains as to why the twisted boat form is more stable than a centrosymmetrical chair configuration.There are no indications of unusually strong atomic vibrations in the crystal and the agreement between the measured dipole moment in solution and that calculated from the molecular structure reported above may indicate that the structure is fairly rigid. (Received May 3rd 1963.) Addition of Organomanganese Pentacarbonyls to Fluoro-olefins By J. B. WILFORD and F. G. A. STONE P.M. TREICHEL OF CHEMISTRY QUEEN MARY COLLEGE UNIVERSITY €3.1) (DEPARTMENT OF LONDON LONDON SOME transition metal carbonyl hydrides add at room temperatures to certain fluoro-olefins thereby affording fluoroalkyl-transition-metalcar-bonyls for example (CHFClCF2)Mn(CO),1 and (n-C,H,)W(CO),(CF2.CHF&2 We have now ex-tended these reactions to certain alkyl- and aryl- transition-metal carbonyls and have observed significant differences in behaviour.Unlike hydridopentacarbonylmanganese penta-carbonylmethylmanganese does not add to tetra- fluoroethylene below 80" even in the presence of solvent but above 90°/- 1-2 atm. the pale yellow compound (CH,CF2.CF2)Mn(CO), m.p. 41-42' is formed in high yield. Pentacarbonylphenyl-manganese and tetrafluoroethylene at -100" afford (Ph-CF,CF,)Mn(CO), but the yield is only 5-1 0%. Also pentacarbonylmethylmanganese does not add to chlorotrifluoroethylene below -90° at which temperature the desired product is unstable. These observations prompted a study of the effect of ultraviolet light on reactions between tetrafluoro- ethylene or chlorotrifluoroethylene and pentacar- Treichel Pitcher and Stone Inorg.Chem. 1962 1 511. Treichel Morris and Stone J. 1963 720. bonylmethylmanganese in pentane at ambient temperatures and at 1-2 atm. Tetrafluoroethylene afforded the complex (CH3CF2CF2)Mn(CO) quan- titatively and chlorotrifluoroethylene gave a com- pound (C,H,F,Cl)Mn(CO), m.p. 53-54" in 30% yield (after chromatography). Examination of the chemical shifts and fine structure in the proton and in the fluorine nuclear magnetic resonance spectra of the latter complex showed that it was produced predominantly as the isomer (CH,-CF,CFCI)Mn(CO),. We thus reinvestigated the reaction between hydridopentacarbonylman-ganese and chlorotrifluoroethylene previously reported1 to give (CHFCI-CF,)Mn(CO) (I) when carried out at 5 atm.in a steel bomb at 25". Under the ultraviolet radiation conditions which gave (CH,-CF2-CFCI)Mn(CO), we obtained a mix- ture of the isomers (I) and (CHF2-CFC1)Mn(CO) (11). Nuclear magnetic resonance studies showed that the latter predominated. Repetition of the earlier work1 showed that the direction of hydride addition is pressure-dependent. In a steel bomb formation of JULY 1963 isomer (11) can be inhibited to a point where its detection by nuclear magnetic resonance spectro- scopy is difficult or even impossible. It appears that two mechanisms may operate in the reactions of these carbonyls with chlorotrifluoro- ethylene-a free-radical and a four-centre mechan- ism.The former is favoured by ultraviolet irradiation and low pressures. Acknowledgement is made to the donors of the Petroleum Research Fund administered by the American Chemical Society for support of this research. We thank Dr. Randall and his co-workers for recording nuclear magnetic resonance spectra the University of London for a Final Studentship (to J.B.W.) and the U.S. National Science Foundation for a Postdoctoral Fellowship (to P.M.T.). (Received June 5th 1963.) A Convenient New Method for the Reduction of Organic Halides By D. BRYCE-SMITH and (in part) E. T. BLUES B. J. WAKEFIELD (THE UNIVERSITY READING) MAGNESIUM reacts readily with primary alcohols such as ethanol under the influence of a catalytic amount of iodine to give the dialkoxide.If an organic halide is present little reduction of it occurs. We have noted that magnesium does not react with secondary or tertiary alcohols unless an equivalent proportion of a halogen or halogen compound is present. The alkoxymagnesium halidel and the reduced product are formed thus RHal + Mg + ROH -RH + R'O-MgHal. This reaction has provided the basis of extent. As may be seen from the Table halogen can be reduced in the presence of certain other functional groups; but nitro-compounds seem to inhibit the reaction. Propan-2-01 either alone or in decahydro- naphthalene has normally served as the alcoholic component but preliminary results suggest that the readily prepared 1 -methoxypropan-2-01 may be even more reactive.The reactions can be initiated by a trace of iodine and evolve much heat. The overall process could be considered to involve Reduction of halides with magnesium andpropan-Zol. Halide Temp. Reduction product (%) Other products (%) Carbon tetrachloride 80* Methane (47) Traces of chloroform dichlorornethane chloro- methane 1-Chlorobutane l5Ot Butane (95) - 2-Chloro-2-methylbutane Chlor oc yclohexane 4 150 150 2-Methylbutane (32) Cyclohexane (83) 2-Methylbut-1 -ene (1 3) 2-methylbut-2-ene (39) Cyclohexene (1 0) Fluorocyclohexane Isopropyl chloroacetate Iodobenzene 150 150 150 Cyclohexane (33) Benzene (95-5) Isopropyl acetate (63) Cyclohexene (58) Biphenyl (trace) - Bromo benzene 150 Benzene (89) Biphenyl (trace) Chlorobenzene 1 -Bromonaphthalene p-Bromost yrene p-Bromoaniline p-Bromophenol 150 80 80 80 80 Benzene (89) Naphthalene (90) Styrene (72) Aniline (61) Phenol (66) Biphenyl (trace) -- * Reactions at 80" were conducted in an excess of refluxing propan-2-01.t In the reactions at 150" the halide and propan-2-01 were added to magnesium powder in decalin. $ An equimolar amount of 1-bromonaphthalene was used as entraining agent. a convenient and apparently widely applicable pro- cess for reduction of organic halides. The reduction of fluorides occurs less readily than that of the other halides but is accomplished in the presence of an easily reduced halide such as 1-bromonaphthalene. Even polytetrafluoroethylene is attacked to a limited Meenvein and Schmidt AnnaZen 1925,444 236.* Blues and Bryce-Smith Chem. and hd. 1960 1533. alkoxide-promoted formation of an organornag-nesium compound,2 followed by alcoholysis but this may well be an over-simplification. D.S.I.R. is thanked for financial support. (Received June 13th 1963.) PROCEEDINGS New Potent AnaIgesics in the Morphine Series By K. W. BENTLEY and D. G. HARDY (THE RESEARCH J. F. MACFARLAN LABORATORY AND Co. LTD. WHEATFIELD 11) ROAD EDINBURGH MANY attempts have been made to produce an analgesic as good as morphine but one not producing its unwanted side effects by the synthesis of com-pounds having simpler and more flexible structures. We believed however that this would be more easily achieved by the preparation of more rigid and complex structures which might be expected to be more selectively adsorbed than morphine at the receptor surfaces presumably closely similar associated with the various effects.Ready access to a series of such compounds is provided by the Diels- Alder addition of dienophils to thebaine; indeed a derivative (I) of the benzoquinone adduct has been shown to be about as active an analgesic as pethidine.l We have prepared by addition of certain ap-un- saturated ketones to thebaine and in poor yield by the interaction of arylmagnesium halides with the thebaine-acrylonitrile adduct2 or of alkylmagnesium halides with the adduct of thebaine and ethyl acryl- ate a number of ketones of the tetrahydro-6,14- ethenothebaine series of general structure (11).Further we have prepared by the action of Grignard reagents on the adducts of thebaine with acraldehyde (11; R = H),3 methyl vinyl ketone (11; R = Me):,* phenyl vinyl ketone (11; R = Ph),2 and ethyl acrylate (11; R = OEt) a series of secondary and tertiary alcohols (111; R = Me) and most of these by de- methylation under alkaline conditions have been converted into the corresponding phenols (111; R = H). In this way we have produced analgesics of unprecedented activity. The effect of 0-acetylation of the phenols on activity is marginal. The alcohols (111) are readily rearranged under acid conditions to the related 14-alkenyl-codeinones and -rnorphinones of general structure (IV) and these in turn are reducible to the dihydro- (V) and tetrahydro-compounds(VI) by chemical and catalytic means respectively; and the ketones of all three series(IV-VI) may be reduced at the carbonyl group by sodium borohydride to give the corresponding derivatives of morphine and codeine.The series of alcohols (111; R = Me) prepared had a wide variety of groups R’ and R”. When R’ and R” were not identical two diastereoisomers were obtained as a result of the Grignard reaction in roughly equal amounts and these afforded the same Bentley and Thomas J. 1956 1863. 14alkenylcodeinone on acid-catalysed rearrange- ment. Most of the above alcohols were converted into the corresponding phenols and the bases of both series were transformed into bases of series (IV-VI) . cm The bases obtained have analgesic activities rang- ing from the barely detectable to the unprecedented level of almost 10,OOO times that of morphine when determined by the subcutaneous route in rats (tail- pressure method); high activities have been con-firmed in other animals.The therapeutic index of the most active compound is 25,000 and of the next most active 65,000 (LD5@ determined by the intra- venous route in mice). (Received June 7th 1963.) Bentley and Ball J. Org. Chem. 1958,23 1720. Kanewskaya and Mitryagina J. Gen. Chem. (U.S.S.R.) 1947,17 1203. * B.P. 902,659. JULY 1963 221 An Anomalous Photoisomerisationin the Cycloheptatriene Serial By 0.L. #IAPMAN and G. W.BORDEN (DEPARTMENT IOWA OF SCIENCE OF CHEMISTRY STATEUNIVERSITY AND TECHNOLOGY AMES,IOWA, U.S.A.) IRRADIATION tropolone ethers cyclohepta- 1,3- of dienes cycloheptatriene and methyl 5,5,-dimethyl- cyclohepta-l,3,6-trienecarboxylatein solution gives in each case a bicyclic valence tautomer without skeletal rearrangement.2 Vapour-phase irradiation of cycloheptatriene in contrast to irradiation in solution gives primarily tol~ene.~ Wereport an efficient photo- isomerisation of 7-alkoxycycloheptatrieneswhich is anomalous both in solution and in the vapour phase.Irradiation of 7-methoxycyclohepta-l,3,5-triene (Ia; 8.0 g.) in ether (4 1.) for 6 hr. with a mercury-arc lamp in a quartz immersion well gives after removal of the ether and distillation the bicyclic photoisomer (IIa; 7.4 g. 92%). Similar irradiation of (Ib) gives (IIb; 91%).The photoisomers show no high-intensity ultraviolet absorption above 220 mp and the expected nuclear magnetic resonance spectra. Reduction of the photoisomers (IIa) and (IIb) with two equivalents of hydrogen over platinum gives (IIIa) and (IIIb) which do not show any proton resonance characteristic of the >CH.OR grouping. Pyrolysis of (IIa) and (IIb) gives the ethers Ova) and (IVb). The pyrolysis products were identified by direct comparison with authentic sample^.^ The initial 7-alkoxycyclohepta-l,3,5-trienes(Ia and Ib) also give the ethers ova) and (IVb) on pyr~lysis.~ Irradiation of (IVa) and (IVb) in ether gives the photoisomers (Ha) and (IIb) in normal fashion. Structure (IIa) is established by conversion into cycloheptane- l,.l-dione (VII).Reduction of (IIa) with one equivalent of hydrogen selectively saturates the cyclobutene double bond giving (V). Hydroboration of (V) followed without isolation by oxidation with aqueous chromic acid gives the methyl ether 071) which is converted very rapidly in acid into cyclo- heptane-l,4-dione (VII) (identity proved by com-parison of bisdinitrophenylhydrazones). The speci- ficity of the hydroboration suggests that reaction proceeds via a borane-ether complex. Irradiation of (Ia) in the vapour phase gives starting material (10 %) 1-methoxycyclohepta-1,3,5-triene (IVa; 40 %) and 1-methoxybicyclo [3,2,0]- hepta-3,6-diene (IIa ;50 %) ;vapour-phase chromato- graphic analysis failed to detect any toluene derivative.These photochemical rearrangements pose several interesting problems. In view of the results of the vapour-phase irradiation of (Ia) it is reasonable to assume that the rearranged bicyclic photoisomers @I) are formed from the 1-alkoxycyclohepta-1,3,5-trienes (IV). Collisional loss of excess vibrational HC J/ a R= CH b R= $Hs A energy should stabilise the bicyclic product more efficiently in solution than in the vapour. The re- arrangement of (Ia) to (IVa) photochemically is at least formally analogous to the thermal isomerisation of 7-deuterocyclohepta-1 ,3,5-triene6 possibly a result of two successive transannular hydrogen transfer processes. Perhaps both photochemical and thermaI isomerisations of cycloheptatrienes proceed via a vibrationally excited ground state.3 An alternate explanation can be based on analogy with the photo- isomerisation of 1,2,3,4,5-~yclohexa-1,3-diene.~ A similar light-induced hydrogen shift in (I) would lead to the norcaradiene tautomer of (IV).The mechanism of the photochemical transformations of the 7-alkoxycyclohepta- 1,3,5-trienes is under investiga- tion. Satisfactory analyses have been obtained for all new compounds except (V). Structure (1Ib) corrects our preliminary statement (footnote 3 J. Org. Chem. 1962,27,2291). Chapman “Advances in Photochemistry,” Ed. Noyes Hammond and Pitts Interscience Pub]. jn the press. Srinivasan,J. Amer. Chem. SOC.,1962 84 3432. Parham Soeder and Dodson J. Amer. Gem. SOC., 1962,84 1755.ti We are indebted to Professor Dauben for informing us of his findings on the thermal isomerisation of 7-alkoxy- cyclohepta- 1,3,5-trienes. ter Borg Klossterziel and Van Meurs Proc. Chem. SOC.,1962 359. Evanega Bergman and English J. Org. Chem. 1962 27 13. PROCEEDINGS This investigation was supported by grants from G.W.B. thanks the Department of Health Education the National Science Foundation and the Petroleum and Welfare for a pre-doctoral fellowship. Research Fund of the American Chemical Society. (Received May lst 1963.) Evidence for Paiiwise Trapping of Photolytic Free Radicals By P. W. ATKINS,M. C. R. SYMONS and P. A. TREVALION (DEPARTMENT THEUNIVERSITY, OF CHEMISTRY LEICESTER) IRRADIAITON of single crystals of potassium per- sulphate at room temperature with light of wave- length 3650 A results in the formation and trapping of a radical which has been identified as SO4- slightly distorted from a tetrahedral structure? The same radical is one of the products of y-irradiation but the spectra differ in that the single line expected for SO4- and found in y-irradiated crystals is sym- metrically flanked by two sets of doublets in the spectra of crystals exposed to ultraviolet light (see Figure).The separation (2AH)of the components of each doublet accurately follows the equation AH = K(3 cos2 8 -l) where 8 is the orientation of the crystal to the external field direction. electron spin spectrum ofa crystal of potassium persubhate after exposure to ultraviolet radiation.This behaviour is characteristic of a dipolar mag- netic coupling then = g/3s/r3,r being the separa- tion of the dipoles and can be understood in terms of isolated centres consisting of pairs of radicals with doublet spin states separated by about 13 A and 19 A the axis of separation having a common direction for all sets of radicals. This novel situation would arise if after photolytic cleavage of the oxygen-oxygen bond one or both of the SO4-radicals were to form a new peroxide linkage with their nearest neighbouring persulphate ions thereby forming new persulphate ions and leaving isolated SO4-radicals. This can be approxi- mately depicted as Sop so SO4-\ \ \ SO4-SO4-SO4-I -+ SO4-SO4-SO4-/ SO4-SO4-/ SO4-This conclusion is completely compatible with the crystal structure,2 from which we estimate that the sulphur to sulphur distances for the two sets of pairs would be 11.5 A and 18.5 A if there were no relaxa- tion of fragments after reaction.Similar interactions have been studied for rela- tively concentrated “solutions” of paramagnetic ions in diamagnetic host ~rystals,~ and the possibility that painvise trapping of photolytic radicals can grossly modify their spin resonance spectra has been envisaged previously for cases where such interaction be random.4 We acknowledge a grant from the Department of %ierrtific and Industrial Reear& to P.w.A. and a bursary from the U.K.A.E.A to P.A,T. (Received May 6th 1963.) Atkins Brivati Horsfield Symons and Trevalion Sixth International Symposium on Free Radicals Cambridge, 1963.Keen 2.Krist. 1935 91 129. Owen J. Appl. Phys. 1962 Suppl. Vol. 33 No. 1 355; Owen Brown CoIes and Stevenson J. Phys. Soc. Japan, 1961 17 Suppl. B-1 428. Symons and Townsend J. 1959 263 ;Smith and Wyard Nature 1960,186,226. JULY 1963 223 The Cyclic Tetramerisatian of o-Aminobenzaldehydein the Presence of Metal Ians By GORDON and DARYLE A. MELSON H. BUSCH AND MCPHERSON LABORATORIES UNIVERSITY, (THE EVANS CHEMICAL THEOHIOSTATE COLUMSUS 10 OHIO U.S.A.) PREVIOUS investigations of the reactions of metal ions with o-aminobenzaldehyde have usually included in the system molecules capable of Schiff-base forma- tion with the aldehyde gr0up.l Eichhorn and Latif? however investigated the reactions of o-amino-benzaldehyde alone with metal ions.They reported their products to be complexes of a trimeric con- densate trianhydro-o-aminobenzaldehyde(I). In conjunction with a general programme con- cerned with the influence of metal ions on the course of the reactions of materials that may function as ligands we have repeated and extended the observa- tions of Eichhorn and Latif.z From analytical and infrared data we conclude that the organic portion of the product is a closed macrocyclic tetramer (cf. 11), C28H20N4, functioning as a tetradentate ligand tetrabenzo [b,f,j,n] [195,9,1Sltetra -azacyclohexadec-ine. This has been established for three metal ions copper(II) nickel@) and cobalt@).HCQ b'" (a) M = Cu(ii),Ni(rt),or CO(II) J1 These compounds are remarkably stable toward mineral acids suffering no decomposition in boiling concentrated nitric acid and merely undergoing metathesis to the corresponding perchlorate salts when heated with 60% perchloric acid. The magnetic moment of the metal atom in the copper derivative is 1.84 B. M. indicating the presence of copper(r1) rather than of copper(1) as previously reported. The magnetic moment of the nickel atom in the tetra- azacyclohexadecineis small establishing the expected planar arrangement of the donor atoms. Reflectance and solution absorption- spectra in the visible and the near-infrared region confirm the planar structure of the nickel complex.Two factors have probably contributed to the erroneous conclusions previously reported for these systems. The great stability of the complexes caused some analytical difficulties particularly the deter- mination of nitrogen. Also a second product is formed by the self-condensation of o-aminobenz- aldehyde at least in the presence of nickel salts. The magnetic moment and spectral properties of the second product indicate octahedral co-ordination about nickel(n). The observed tetramerisation of o-aminobenz-aldehyde in the presence of metal ions bears a close resemblance to the cyclic tetramerisation of phthalo- nitrile in the presence of metals or metal ions to form phthal~cyanines.~ This suggests further that the reduction products of the new macrocyclic complexes should be of much interest.Indeed reduction with for example sodium borohydride produces addi- tional new substances (these are under investigation). In the absence of metal ions o-aminobenzaldehyde undergoes a slow self-condensation to form the trimer trianhydro-o-amin~benzaldehyde.~ SeideP also obtained this condensate in the absence of metal ions but in the presence of zinc chloride under anhydrous conditions he obtained a compound reported as C,H,N,&ZnCl,. This should probably be formulated as [ZnL][ZnCl,] where L is the tetra- meric macrocycle C28H20N4, that we have obtained. The reactions reported here provide an additional example of the influence of a metal ion in producing an organic molecule not obtained in its absence.$ These investigations are being continued with a view to complete elucidation of the complexes formed by Sacconi Gazzeta 1953 83 884; J.1954 1326; J. Amer. Chem. Soc. 1954 76 3400; Pfeiffer Hesse Pfitzner Scholl and Thielert J. prakt. Chem. 1937 149 217. a Eichhorn and Latif J. Amer. Chem. Soc. 1954,76 5180. Dent and Linstead J. 1934 1027. Barnburger Ber. 1927 60 314; Seidel and Dick ibid. p. 201 8. Seidel Ber. 1926 59 1894. Thompson and Busch J. Amer. Chem. Soc. 1962,84 1762. PROCEEDINGS metal ions with the self-condensation products of The financial support of the National Institutes of o-aminobenzaldehyde and of the reactions of the Health United States Public Health Service is ligands so formed and the metal ions to which they gratefully acknowledged.are co-ordinated. (Received May 27th 1963.) Ring Confraction of N-HydroxyIactams to Heterocyclic Bases By G. DI MAIO and P.A. TARDELLA (ISTITUTO DI CHIMICA ORGAMCA DELL’UNIVERSITA ROME ITALY) WEhave recently reported a new ring-enlargement reaction wherein cyclic carbonyl compounds are con- verted into cyclic hydroxamic acids on treatment with Piloti’s acid (Ph-SO,.NH*OH) in alkaline media.l12 We have now found that on treatment with acidic reagents these N-hydroxy-lactams may under- go ring-contraction to yield non-oxygenated hetero- cyclic bases. Depending on the compound two distinct rearrangements may occur (a and b) with elimination of either carbon monoxide or carbon dioxide.1-Hydroxy-2-piperidone (I) is readily prepared from cyc1opentanone.l When heated with polyphos- phoric acid at 175-195” it yielded carbon monoxide (up to 45 %) but no detectable amount of carbon di- oxide. After dilution of the reaction mixture with water neutralisation with potassium hydroxide and distilfation under reduced pressure 1-pyrroline was / (n) i solated by conversion into the picrate (11) m.p. and mixed m.p. 166-167” (correct infrared spectrum). In confirmation reduction of a portion of the above distillate with tin and hydrogen chloride yielded pyrrolidine. On the other hand when trans-octahydro-2-hydroxyisocarbostyri12(111) was treated with poly- phosphoric acid under similar conditions it gave only carbon dioxide (50 %) and no detectable amount of carbon monoxide.After neutralisation of the re- action mixture and extraction of the basic fraction trans-octahydroindole (IV) was isolated in 45 % yield (picrate m.p. and mixed m.p. 150-152” correct infrared spectrum). &-m-(m (IV) &(-0 In our opinion the above results suggest that the two steps in sequence (a) are involved in a single synchronous process for which the coplanarity of bonds Q! and ,8 in (I) is essential. This being im- possible in compound (III) the reaction follows sequence (b). We thank Dr. F. Booth for a sample of trans-octahydroindole picrate and Consiglio Nazionale delle Ricerche for support. (Received June 1 lth 1963.) Panizzi Di Maio Tardella and d’Abbiero Ricerca Sci.Rend. Part 2A 1961 31 312. Di Maio and Tardella Gazzetta 1961,91 1124. JULY 1963 225 ~-The Extent of Hydrogen-isotope Exchange in the Elimination Reactions of Trimethyl-&-methyl-benzylammonium and -phenethylammoniumIons By D. V. BANTHORPE and J. H. RXDD (UNIVERSMY Gowm STREET LONDON,W.C.1) COLLEGE THE occurrence of hydrogen-isotope exchange in an olefin-forming elimination has rarely been observed;l the main evidence for it involves elimination of hydrogen fluoride from a dichlorotrifluoroethane2to form complex products presumably derived from an olefin. We find that elimination from trimethyl-a- methylbenzylammonium and -phenethylammonium ions is accompanied by significant hydrogen-isotope exchange between the unchanged substrate and the medium.When these reactions are studied in deuterated solvents and stopped after 25-30% of reaction the extent of deuterium uptake is as shown in the Table where it is expressed as a percentage of that for complete exchange at one C-H position. Thus for the reaction with Oe137M-sodium meth- oxide the water obtained from combustion of the phenethyl isomer (as the tetraphenylboron salt) con- tained 0.207 & 0.005 % D,O in excess of the natural abundance. The corresponding value for complete exchange at one position would be 2.63%. MeCHPh-NMe$ - group more readily than proton abstraction from the medium. The amount of exchange should then de- pend on the activity of methanol and this activity will be reduced as the concentration of sodium methoxide is increased.In contrast exchange and elimination in the a-methylbenzyl isomer should be independent reactions ; it is therefore reasonable that the percentage exchange should not decrease with increasing concentration of sodium methoxide. A difficulty with this interpretation is that Ayrey and Bourns* have reported a considerable nitrogen isotope effect (30%of the theoretical maximum) for elimination from the phenethyl isomer in ethanolic solution ([EtO-] = 0.1~);this result is apparently inconsistent with a rate-determining C-H heterolysis. Also the rate of proton loss from this isomer relative to the a-methylbenzyl compound appears unex-pectedly fast even after allowance for steric inhibi- tion of resonance in the conjugate base of the latter.We suggest therefore that under our conditions,* Ph.CH,CH,-NMe$ /---.-, r-A -A--7 Solvent (ROH) MeOD EtOD MeOD EtOD /-~---[OR-I (MI 0.83 2.98 0-16 0.137 0.98 3-83 0.16 Temp. 100" loo" loo" 60" 45" 45" 45" Exchange (%) 7.6 15.4 1-6 7.9 2.3 0.8 12.0 The exchange in the a-methylbenzyl isomer (A) the transition state for proton loss from the phen- presumably involves the benzylic a-hydrogen atom ethyl isomer derives some stabilisation from partial for this hydrogen should be the most acidic. How- breaking of the C-N bond and partial double bond ever exchange at the a-position or in the methyl character in the Ca-CB bond. Other authors have groups3 would be too slow to explain the deuterium emphasised that there is no sharp distinction between uptake observed with the phenethyl isomer (B) for the E2 and the ElcB mechanisms of elimination our elimination from this (and hence the concurrent ex- results suggest that some of the characteristics of the change) is much faster than from its isomer; at 60" E2 mechanism can persist when the conjugate base the ratio k2*/k2*is -100.To account for the of the substrate is formed as a transitory inter- exchange in the phenethyl isomer and its marked mediate. decrease with increasing concentration of sodium methoxide we suggest that both exchange and The authors thank Professor Sir Christopher elimination involve the common intermediate (I) Ingold Professor A. N. Bourns Professor J. F. Bunnett and Dr.C.A. Bunton for helpful discus- PhCH,CH,.NMef +-CHPhCH,.NMef 3 PhCH=CH sions and Professor s. Nperger for the com-(9 munication of unpublished results. and that this zwitterion undergoes loss of the NMe (Received,June 1 lth 1963.) * Mperger and his co-workers5 have recently shown that the nitrogen isotope effect for elimination from the phen- ethyl isomer in aqueous solution ([OH-] = 0.1~)is almost unity; this implies that the zwitterion (I) is then formed without significant stretching of the C-N bond. Bunnett Angew. Chem. (Znternat. Edn.) 1962 1 225. Hine Wiesbieck and Ramsay J. Amer. Chem. Soc. 1961 83,1222. Von Doering and Hoffman J. Amer. Chem. SOC.,1955,77 521. 4 Ayrey and Bourns quoted by Buncel and Bourns Canad. J. Chem. 1960,38,2457. Mperger Klasinc and PapiE personal communication.PROCEEDINGS NEWS AND ANNOUNCEMENTS Birthday Honours List.-The names of the following Fellows were included in the Birthday Honours List. Knights Bachelor Dr. G. E. Archey Director of the Auckland Institute and Museum; Dr. R. HoZroyd Imperial Chemical Industries Limited; Professor E. R. H. Jones University of Oxford; Dr. P. F. R. Venables Birmingham College of Advanced Technology; Dr. N. C. Wright Ministry of Agriculture Fisheries and Food. C.B.E. Dr. A. W. Chapman University of Sheffield; Dr. J. C. Kendrew Peterhouse Cambridge; Dr. M. F. Perutz University of Cambridge; Dr. F.Sanger University of Cambridge; Dr. B. C. Saunders Magdalene College Cambridge. 0.B.E. Lt.-Col.F. J. Grifin The Society of Chemical Industry; Dr. R. W.Pittman Birkbeck College London. Dexter Award in the History of Chemistry.- Professor Douglas McKie Head of the Department of the History of Science University of London has been chosen as the recipient of the Dexter Award in the History of Chemistry for 1963. The Award will be presented at a meeting of the Chemical Society in the Autumn. Local Representatives.-Bristol The Council has appointed Dr. L. Rough as Local Representative for Bristol in succession to Dr. W.D. Ollis who has been appointed to the Chair of Organic Chemistry of the University of Sheffield. hst Anglia The Council has decided to appoint a Local Representative for East Anglia and Dr. R. A. Y,Jones (shortly of the University of East Anglia) is to be the first Local Representative.Liaison Officer.-Dr. W.J. Geavy has been appointed Liaison Officer at the Lanchester College of Technology Coventry in succession to Dr. W.R. McGregor who has been appointed Senior Lecturer at Brighton College of Technology. Deatbs.-We regret to announce the deaths of the following Dr. H. Baines (7.6.63) Scientific Liaison Officer for Kodak Ltd.; Mr. P. Blackman (4.4.63), Chingford a Fellow since 1919; Mr. J. E. Brimley (13.3.63) formerly Chief Chemist Sanitas Co. Ltd. ; Professor E. D. Hughes (30.6.63) Professor of Chemistry at University College London W.C. 1; Professor W. H. Lewis (25.5.63) Emeritus Professor of Chemistry University of Exeter; Dr. W. B. Pollard (27.5.63) formerly Research Chemist Government Laboratory Cairo.Election of New FeUows.-2€0 Candidates were elected to the Fellowship in June 1963. International Symposia etc.-An International Symposium on Physical Chemistry of Biogenic Macromolecules will be held in Jena on September 18th-21sty 1963. Further enquiries should be addressed to Institut fi Mikrobiologie und Experi- mentelle Therapie Deutsche Akademie der Wissen- schaften Beuthenbergstr. 1 1 Jena Germany. A Symposium on Modern Instrumentation in Science and Industry arranged jointly by the London Sections of the Institute of Biology the Institute of Physics and Physical Society and the Royal Institute of Chemistry is to be held in London on January 17th 1964. Further details will be announced later.A Symposium on Neutron Dosimetry for Radio- logical Purposes arranged on behalf of the Joint Health Physics Committee is to be held in London on January 24th 1964. Further enquiries should be addressed to the Administration Assistant The Institute of Physics and The Physical Society 47 Belgrave Square London S.W.l. The Fifteenth Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy will be held in Pittsburgh on March 2nd-6th 1964. Further en- quiries should be addressed to Dr. William A. Straub Programme Chairman c/o Applied Research Laboratory United States Steel Corporation, Monroeville Pennsylvania. The 14th Chemical Engineering Exhibition and ACHlEMA Congress will be held in Frankfurt-am- Main on June 19-27thY 1964.Further enquiries should be addressed to DECHEMA Frankfurt-am- Main 7 Germany. The Sixth International Congress of Biochemistry sponsored by the International Union of Biochem- istry will be held in New York on July 26th-31stY 1964. Further enquiries should be addressed to the Secretariat c/o American Society of Biological Chemists 9650 Wisconsin Avenue Washington 14 D.C. The Fifth International Symposium on the Re- activity of Solids sponsored by the I.U.P.A.C. and the I.U.P.A.P. will be held in Munich on August &loth 1964. Further enquiries should be addressed to Dr. Rudoph Morf Secretary General I.U.P.A.C. c/o F. Hoffmann-La Roche et Cie Basel 2 Switzer- land. The Tenth International Symposium on Com-bustion sponsored by the Combustion Institute JULY 1963 Pittsburgh Pennsylvania will be held in Cambridge on August 17th-21st 1964.Further enquiries should be addressed to the Secretary Combustion Sym-posium Local Committee Department of Physical Chemistry Lensfield Cambridge. Persond.-Professor C. E. H. Bawn has been appointed acting Vice-Chancellor of the University of Liverpool. Mr. E. Bishop University of Exeter has been appointed Senior Lecturer from October 1st. Dr. A. Carrington Downing College Cambridge has been appointed an Assistant Director of Research in the Department of Organic and Inorganic Chem- istry for five years from October 1963. Professor N. B. Chapman of the University of Hull who is at present in the United States has accepted an invitation to lecture in the University of British Columbia Vancouver during July and August 1963.Dr. E. W. Crunden Crawley College of Further Education has been given two years leave of absence to act as Adviser on the teaching of Chemistry at the University of the Philippines under the Technical Co-operation Department of the Colombo Plan. Professor Joseph de Heer now Professor of Chem- istry at the University of Colorado is to spend the academic year 1963-64 on leave on a Fulbright Grant at the Department of Physical Chemistry University of Copenhagen Denmark. Mr. G. H. Dudd has been elected a Scholar of the House in Natural Sciences at Trinity College Dublin. Dr. J. P. Elder formerly at the Royal Institute of Technology Stockholm has taken up a research post in the Chemical Engineering Division of the Argonne National Laboratory Illinois.Professor H. J. Emelius has been elected a Fellow of the Imperial College of Science and Technology London. Dr. J. W.Emsley has been appointed Lecturer in Chemistry at the Durham Colleges. Lord Fleck has been elected President of the Royal Institution. Dr. F. Franks Bradford Institute of Technology is to take up a National Aeronautics and Space Administration Research Associateship for one year at the University of Pittsburgh. Dr. M. L.H. Green has been elected to an Official FeIlowship in Inorganic Chemistry at Balliol College Oxford from September 1963. Dr. A. K. Holliuky has been promoted to Reader in Inorganic Physical and Industrial Chemistry at the University of Liverpool as from October lst 1963.Dr. L. Hough Lecturer in Organic Chemistry Dr. R. Parsons Lecturer in Inorganic and Physical Chemistry and Dr. F. S. Stone Lecturer in Chem- istry have been appointed to Readerships at the University of Bristol from August lst 1963. Mr. C. W.Maplethurpe has been elected President of the Pharmaceutical Society of Great Britain. Dr. R. L. Mitchell has been awarded the Royal Agricultural Society’s Research Medal for 1963 in recognition of his work in the field of trace-element studies and spectrochemical analysis of soils plants and related material. Dr. R. H. Prince has been appointed University Lecturer in Organic and Inorganic Chemistry at St.John’s College Cambridge. Sir Eric Rideal has been elected a Fellow of King’s College London. Dr. €? G. A. Stone Queen Mary College has been appointed Professor of Inorganic Chemistry at the University of Bristol as of August 1st. Dr. I. 0.Sutherland has been appointed Lecturer in Chemistry at the University of Sheffield. Lord Todd has accepted an invitation to serve as a Trustee of the Ciba Foundation. The University of Padua Italy has conferred an Honorary Laureate in Engineering Science on Professor A. R. Ubbelohde. Dr Alan Wisemanis to join the M.R.C. Toxicology Unit at Carshalton Surrey in September next and later the M.R.C. Unit in the Department of Bio-chemistry Imperial College of Science and Tech- nology London S.W.7.ADDITIONS TO THE LIBRARY Directory of British scientists 1963. Pp. 1289. Benn. London. 1963. Imperial College Inaugural Lectures. 1960-61 and 1961-62. Pp. 157. Imperial College. London. 1963. (Presented by the Rector.) Infrared absorption spec troscopy-pract ical. K. Nakanishi. Pp. 233. Holden-Day. San Francisco. 1962. The shift and shape of spectral lines. R. G. Breene Jr. Pp. 323. Pergamon Press. Oxford. 1961. Schmelzpunkttabellen organischer Verbindungen (Melting point tables of organic compounds). W. Utermark and W. Schicke. 2nd edn. Pp. 715. Friedr. Vieweg und Sohn. Braunschweig. 1963. Comprehensive treatise on inorganic and theoretical chemistry. J. W. Mellor. Vol. 2 Supplement 3.Pp. 1459-2599. Longmans. London. 1963. Ion flotation. F. Sebba. (Elsevier Monographs.) Pp. 154. Elsevier. Amsterdam. 1962. Thermodynamics of gasification and gas-synthesis reactions. N. V. Lavrov V. V. Korobov and V. I. Filippova. Translated by G. H. Kinner. Pp. 116. Per- gamon Press. Oxford. 1963. Liquid extraction. R. E. Treybal. 2nd edn. Pp. 621. McGraw-Hill. New York. 1963. Surface chemistry theory and industrial applications. L. I. Osipow. (American Chemical Society Monograph Series no. 153.) Pp. 473. Reinhold. New York. 1962. Solvents guide. Edited by C. Marsden and S. Mann. 2nd edn. Pp. 633. Cleaver-Hume. London. 1963. Electrochemistry theoretical principles and practical applications. G. Milazzo. Translated by P. J. Mill.Pp. 708. Elsevier. Amsterdam. 1963. Metall-m-Komplexe mit di-und oligo-olefinischen Liganden. E. 0. Fischer and H. Werner. (Monographien zu “Angewandte Chemie” and “Chemie-Ingenieur-Technik” Nummer 80.) Pp. 142. Verlag Chemie. Wein- heim. 1963. Cahiers de synthhe organique mhthodes et tableaux #application. J. Mathieu and A. Allais; edited by L. Velluz. Vol. 10. Pp. 560. Masson et Cie. Paris. 1963. (Presented by the publisher.) Mises au point de chimie analytique organique pharmaceutique et bromatologique. Edited by J.-A. Gautier and P. Malangesu. 1 lth series. Pp. 252. Masson et Cie. Paris. 1963. Methods in carbohydrate chemistry. Edited by R. L. Whistler and M. L. Wolfrom. Vol. 3. Pp. 407. Academic Press. New York. 1963. Textbook of polymer science.F. W. Billmeyer Jr. Pp. 601. Interscience Publ. Inc. New York. 1962. Five- and six-membered compounds with nitrogen and oxygen (excluding oxazoles). Edited by R. H. Wiley. (Chemistry of Heterocyclic Compounds. Edited by A. Weissberger.) Pp. 493. Interscience Publ. Inc. New York. 1962. The chemistry of nucleosides and nucleotides. A. M. Michehon. Pp. 622. Academic Press. London. 1963. Textbook of physiology and biochemistry. G. H. Bell, J. N. Davidson and H. Scarborough. 5th edn. Pp. 3 11 7. E. and S. Livingstone. Edinburgh. 1961. Industrial applications of the organometallic com- pounds a literature survey. J. H. Harwood. Pp. 451. Chapman and Hall. London. 1963. (Presented by the publisher.) The thermochemical properties of uranium compounds.M. H. Rand and 0. Kubaschewski. Pp. 96. Oliver and Boyd. Edinburgh. 1963. Quantitative chemical micro-analysis. C. J. van Nieuwenburg and J. W. L. van Ligten. Translated by C. G. Verver. (Elsevier Monograph.) Pp. 181. Elsevier. Amsterdam. 1963. Residue reviews residues of pesticides and other foreign chemicals in foods and feeds. Vol. 2. Pp. 156. Springer Verlag. Berlin. 1963. New chemical engineering separation techniques. Edited by H. M. Schoen. Pp. 439. Interscience Publ. Inc. New York. 1962. Re-use of water in industry a contribution to the solu- tion of effluent problems. Pp. 247. Butterworths. London. 1963. Advances in mass spectrometry. Vol. 2 :Proceedingsof a conference organised by the American Society for Testing Materials Committee E-14 and the Hydrocarbon Research Group of the Institute of Petroleum and held in Oxford 1961.Edited by R. M. Elliott. Pp. 628. Pergamon Press. Oxford. 1963. The physics and chemistry of high pressures papers read at the High Pressure Symposium held at the 3rd International Congress of the European Federation of Chemical Engineering London 1962 with the discussions that followed; sponsored by the Society of Chemical Industry The Institution of Chemical Engineers and the Institute of Physics and the Physical Society. Pp.247. Society of Chemical Industry. London. 1963. Proceedings of the symposium on the handling of solids held at the 3rd Congress of the European Federa- tion of Chemical Engineering London 1962 under the auspices of the Institution of Chemical Engineers.Edited by P. A. Rottenburg. Pp. 100. Institution of Chemical Engineers. London. 1962. (Presented by the Institution.) Fuel cells. Vol. 2 a symposium held by the Divisions of Fuel Chemistry and Petroleum Chemistry of the American Chemical Society at the 140th National Meeting in Chicago 1961. Edited by G. J. Young. Pp. 225. Weinhold. New York. 1963. Saline water conversion-I1 based on symposia sponsored by the Division of Water and Waste Chem- istry at the 139th and 141st Meetings of the American Chemical Society 1961 and 1962. (Advances in Chem- istry Series no. 38.) Pp. 199. American Chemical Society. Washington. 1963. (Presented by the publisher.) Some general problems of paper chromatography relations between paper chromatographic behaviour and chemical structure-attempts at systematic analysis ; a symposium organised by the Chromatography Group of the Czechoslovak Chemical Society at Liblice 1961.Edited by I. M. Hais and K. Macek. Pp. 220. Publishing House of the Czechoslovak Academy of Sciences. Prague. 1962. Lectures on gas chromatography 1962 based on lectures presented at the advanced sessions of the Fourth Annual Gas Chromatography Institute held at Canisius College Buffalo 1962. Edited by H. A. Szymanski. Pp. 282. Plenum Press. New York. 1963. Organic semiconductors proceedings of an inter-industry conference sponsored by the Armour Research Foundation of Illinois Institute of Technology and Electronics.Edited by J. J. Brophy and J. W. Buttrey. Pp. 243. Macmillan. New York. 1962. Collagen. Proceedings of a symposium sponsored by the Central Leather Research Institute Council of Scientific and Industrial Research Madras India 1960. Edited by N. Ramanathan. Pp. 579. Interscience Publ. Inc. New York. 1962.
ISSN:0369-8718
DOI:10.1039/PS9630000189
出版商:RSC
年代:1963
数据来源: RSC
|
|