proceedings of the chemical society. august 1963

PROCEEDINGS

OF THE

CHEMICAL SOCIETY

AUGUST 1963

PRESIDENTIAL ADDRESS* Contributions of X-Ray Analysis to Natural-product Chemistry

By J. MONTEATH ROBERTSON THE increasing power of the X-ray method as a means of solving molecular structures is now so well known that I feel it is hardly necessary for me to stress it. During the past four years my own group in Glasgow have determined the complete structure and stereochemistry of well over 20 natural products from various sources, whose chemical constitution was hitherto either unknown or only partially known. Many other laboratories throughout the world are now likewise engaged on work of this kind, and tremendous progress is being made. To mention only a few, A. McL. Mathieson in Australia, Maria Przybylska in Canada, R. Pepinsky and others in the United States, and the laboratories of Professor A. J. C. Wilson here in Cardiff have all made out- standing contributions to this field. I would also refer particularly to the beautiful work on the extremely complex molecules of the vitamin B,, group that has been carried out in the laboratories of Dorothy Hodgkin in Oxford and which has initiated a whole new field of chemical research. In the case of several even more complex biological molecules, in- cluding some of the globular proteins, exciting progress has recently been made and complete solutions for some of these are now in sight.

It would not be either possible or appropriate for me to attempt a survey of this whole field in the time that is available. Instead I propose to confine myself to a few results that have been obtained recently in my own laboratories in Glasgow. Because of the close and helpful collaboration of a number of

organic chemists it has been possible to choose structures that have been important in adding to our knowledge in the natural product field: &Ray crystal analysis is a powerful and often an essential tool in chemistry, but I am convinced that the fullest use of the potentialities that now exist can only be achieved by means of such collaboration between chemist and crystallographer.

For the crystallographer it is often unrewarding just to take a crystal from a bottle and try to determine its structure. This is all very well as an exercise in crystallography, but it may not add greatly to knowledge. For the chemist it may be equally un- profitable to spend years in trying to elucidate a structure when the solution may sometimes be obtained more quickly and directly by X-ray analysis.

Before I describe these structures I would like to outline very briefly the methods which have made work of this kind possible. It is well known that the central problem in X-ray analysis is that of determining the relative phases of the diffracted waves. Intensity measurements provide the amplitudes but not the phases. During the last 20 years many attempts have been made to solve structures from a knowledge of the amplitudes alone, and this mathematical approach has led to many valuable results. But when the number of atoms is considerable, the complexity of the problem is usually too great, even with the aid of fast electronic computers.

* Delivered at the Anniversary Meeting of the Chemical Society at Cardiff on March 28th, 1963. 229

Dow

nloa

ded

by U

nive

rsity

of G

uelp

h on

18

June

201

2Pu

blish

ed o

n 01

Janu

ary

1963

on

http

://pu

bs.rs

c.org

| doi:

10.103

9/PS9

630000

229View Online / Journal Homepage / Table of Contents for this issue

230 PROCEEDINGS

The methods that have proved effective in solving most of the very complex organic structures, in- cludmg some of the protein molecules, depend upon a chemical rather than a physical or mathematical approach, and this is why it is appropriate that I should mention them here They are the heavy-atom

f"

b,

FIG 1 Superposition of waves

and isomorphous-substitution methods, which were first developed and applied to organic structures during the 1930s, starting with Linstead's phthalocyanine compounds Once again this represented a fruitful and most important collaboration between

ference to give a resultant F of known amplitude but unknown phase If a centre of symmetry is present the phase may be represented by either a peak or a trough at a fixed reference point, as in the diagram We now perform a chemical experiment and add another atom at this point, preferably an atom containing a good many electrons We also assume that the new atom will not disturb the remainder of the structure to any large extent The contribution of this atom is represented by the dotted peak f,, and the resultant amplitude given by the heavy-atom derivative is FH We can only measure amplitudes, but by noting whether the resultant FH increases or decreases we can determine whether the original amplitude F was a peak or a trough We have thus trans- formed the unknown differences in phase, which cannot be measured, into amplitude differences which can be measured

This is the principle of the isomorphous-substitution method If the conditions are rigorously ful- filled and if accurate intensity measurements can be made, we can always determine the unknown phases The phthalocyanme structures: shown in Fig 2 as electron-density projections of the metal-free and nickel compounds, provided an almost perfect example, and these complex structures were deter-

._..- FIG 2 Electron-density projection of metal- free and nickel phthalocyanine

chemist and crystallographer (An earlier application of isomorphous substitution to inorganic alum structures by Cork in 1927 did not lead to very conclusive results )

The prmciple of this method is illustrated in Fig 1 The waves scattered by the atoms in the unknown structure combine by the principles of optical inter-

mined in this way without any chemical assumptions, and indeed without even assuming the existence of atoms in the molecules

However, very few structures, and especially very few natural-product structures, show the high degree of isomorphism and symmetry displayed by the phthalocyanines But the simple heavy-atom method,

Robertson, J , 1935, 615, 1936,1195, Robertson and Woodward, J , 1937,219

Dow

nloa

ded

by U

nive

rsity

of G

uelp

h on

18

June

201

2Pu

blish

ed o

n 01

Janu

ary

1963

on

http

://pu

bs.rs

c.org

| doi:

10.103

9/PS9

630000

229View Online

AUGUST 1963 23 1

developed at the same time, has a wider application although it is less exact. We now assume no symmetry and merely suppose that the heavy atom or some group of heavy atoms can be attached to the structure by any convenient chemical method. It will then usually be possible to find the position of this small group in space by a simple application of the Patterson method: and so calculate its contribution to the structure amplitude. As we are not assuming any symmetry it is now convenient to represent the vector equation by a diagram in the complex plane (Fig. 3a). Frepresents the contribution of the unknown part of the structure, and the end of this vector may lie anywhere on the circumference of the

white circle. f' is the known or calculable contribution of the heavy atom or atoms, and we assume this vector to be completely known. The addition offH to I; gives the structure factor of the heavy-atom derivative, FH, and the end of this vector is now constrained to lie somewhere on the circumference of the shaded circle. We see therefore that the phases of the structure factors of the heavy-atom derivative (FH) are restricted to a considerable extent by the known and calculable phases of the heavy atoms themselves CfH). By merely assigning these phases to the structure factors, F', we are therefore usually able to obtain a first very rough approximation to the structure of the heavy-atom derivative. This can later be refined as further recognisable atoms are included

ture factor IF[ from a group of about 34 randomly placed carbon atoms with a heavy-atom contribution I f ~ l from a single bromine atom. An iodine atom would be equally effective for a random group of about 78 carbon atoms, whereas a chlorine atom would only suffice for a group of about 8 carbon atoms on this basis. However, the heavy-atom contribution will in general be somewhat more effective than is indicated by this calculation, which does not take account of variation of scattering power with angle of incidence. Furthermore, the phases of the very large number of structure factors of less than average magnitude will in general be more effectively determined.

iff^ is larger (or F relatively smaller) the effect on our diagram is to move the shaded circle to the right (Fig. 3b) with the phases of F' now more closely restrained to those off'. On the other hand, iff^ is smaller (or Frelatively larger) the effect is to move the shaded circle to the left (Fig. 3c) and now the approximation is worse, with some of the phases of I;B very much in error. Such difficulties may be lessened by the use of a weighting function for the Fourier terms based on the probable magnitudes of the phase-angle errors, as discussed by sin^.^

If the heavy-atom derivative (Fir) is isomorphous with the parent compound, and if accurate measurements can be made, then the more powerful method of isomorphous substitution may be used equally well in the asymmetric case, although there are some complications. Here we may assume as before that f~ is completely known, representing perhaps the difference in scattering power of two successively substituted heavy atoms, or groups of such atoms. With f' completely known, but only the magnitudes 13'1 and IFa 1, there are two solutions for F in the vector equation 3"' = F + f ~ , as Bijvoet4 has shown. These are indicated in Fig. 4 at the points of intersection of the two circles at Uand V.

FIG. 4. lsomorphous substitution, showing two solutions for F in the asymmetric case.

in the phasing calculations. Fig. 3a assumes that If'[ is equal to IFI. This

would be approximately true for the average struc-

It is now possible to proceed by employing both values of F. In this case it can be shown that the result is equivalent to the heavy-atom method, but

Patterson, Phys. Rev., 1934,46, 372; 2. Krist., 1935,90, 517. Sim in "Computing Methods and the Phase Problem in X-ray Crystal Analysis," ed. Pepinsky, Robertson, and

Bokhoven, Schoone, and Bijvoet, Actu Cryst., 1951, 4, 275. Speakman, Pergamon Press, Oxford, 1961, p. 227.

Dow

nloa

ded

by U

nive

rsity

of G

uelp

h on

18

June

201

2Pu

blish

ed o

n 01

Janu

ary

1963

on

http

://pu

bs.rs

c.org

| doi:

10.103

9/PS9

630000

229View Online

232 PROCEEDINGS

with a very advantageous weighting factor attached to the structure amplitudes.6

However, a complete solution can be obtained if a third isomorphous derivative is available. This double isomorphous-replacement method has been outlined by Bokhoven, Schoone, and Bijvoet4 and very fully discussed by Harker.6 It is of great importance for by this means very direct solutions of some of the complex protein structures are now being achieved.

Returning to our diagram, let us now suppose that another isomorphous heavy-atom derivative is available, with structure factors F, and a known heavy- atom contributionf~. With centre at the end of the -fHp vector we now describe a third circle (Fig. 5 ) of radius IFH? 1. The point of intersection of the three circles ( U ) now defines the end of the F vector uniquely. If the isomorphism were perfect and all measurements could be made with sufficient accuracy, this method would yield solutions as complete and direct in the asymmetric case as those obtained for the phthalocyanines in the centrosym- metric case. In practice there are of course many difficulties, and only rarely will such a set of closely isomorphous derivatives be available. It is also difficult to carry out the many structure-factor measurements that are involved with a sufficient degree of accuracy. Instead of the clean intersection of the three circles shown in Fig. 5 we are therefore more

FIG. 5 . Double-isomorphous substitution. likely to obtain a scatter of points and the best value for the phase angle has to be assessed on a prob- ability basis.

For the smaller natural-product molecules with which we are concerned the attachment of a heavy atom usually causes a considerable change in the overall crystal structure. It is therefore difficult to find a suitable series of isomorphous derivatives with

crystal structures sufficiently similar. However, for some of the very complex protein molecules it is fortunately possible to prepare a considerable number of isomorphous heavy-atom derivatives. The attachment of one or two heavy-metal atoms to these giant molecules may not affect the overall crystal structure to any appreciable extent, but it is sufficient to cause measurable intensity changes in the diffraction spectra, as Perutz has shown in the case of hzmoglobin. For the structure of myoglobin to a resolution of 24 A, Kendrew and his co-workers8 employed a series of no less than five isomorphous heavy-atom derivatives, and a total of some 48,000 reflections were measured and analysed. For these molecules it is therefore possible to apply the isomorphous-substitution method at its full power and with the same directness as in the case of the original p h t halocyanine structures .

With the more usual natural-product molecules that we now describe, the heavy-atom method already outlined must generally be used, and this is usually successful if a single really suitable derivative can be found. Difficulties and ambiguities often occur depending on the particular symmetry that may be present, but instead of struggling with these it is usually better policy to search for another more suitable derivative. Owing to the partial and incom- plete nature of the phase determination the first electron-density distributions are often hard to interpret. But if a few recognisable atoms can be found and included correctly in a further set of phasing calculations, the picture generally clears and can then be refined and made more accurate by further application of the powerful Fourier technique.

I now wish to illustrate these methods by describ- ing some of our recent results in the natural-product field. For this purpose some of my favourite structures are alkaloids, where the ionic nature of the heavy atom halides often ensures a comparatively rigid and consequently well resolved molecule. How- ever, Dr. Sim has already described a number of these structures in a paper to one of the Symposia, so I propose to confine my illustrations to some of the terpenoids, bitter principles, and a few other molecules. I am indebted to a number of organic chemists for providing these interesting problems, but particularly to Professor Barton who has supplied most of them.

In the terpenoid field we have made quite a number of studies, beginning with /3-caryophyllene alcoholg (I) and one of its curious dehydration products, iso-

6 Sim, ref. 3, p. 315. * Harker, Acta Cryst., 1956, 9, 1. * Kendrew, Dickerson, Strandberg, Hart, Davies, Phillips, and Shore, Nature, 1960,185,422.

Green, Ingram, and Perutz, Proc. Roy. Soc., 1954, A , 225,287.

Robertson and Todd, J., 1955, 1254.

Dow

nloa

ded

by U

nive

rsity

of G

uelp

h on

18

June

201

2Pu

blish

ed o

n 01

Janu

ary

1963

on

http

://pu

bs.rs

c.org

| doi:

10.103

9/PS9

630000

229View Online

AUGUST 1963 233

cloven&O @I), whose structure was previously quite unknown. The conversion of the alcohol into iso- clovene presents an unusual problem in mechanism which is not yet fully understood and which calls for further chemical work. Although the X-ray method is extremely powerful in elucidating structures, the result very often, as in this case, does not end the problem but merely indicates the need for further chemical investigation.

of the 2-bromo-desmotropo-santonins (V), but this work is still in progress.

Earlier work by Barton and Levisallesf4 had established the constitution of the related sesquiterpenoid lactone geigerin (VI), and here the stereochemistry has been confirmed and made fully quantitative (VII) by the beautiful three-dimensional X-ray analysis of Hamilton, McPhail, and Sim,15 which is illustrated in Fig. 7.

Another sesquiterpenoid whose structure has to be revised as a result of our X-ray work is a-santonin (IV). The reversal of the accepted configuration at position 11 first became apparent in Asher and Sims studyll of the stereochemistry of isophotosantonic lactone, which was defined by the complete three- dimensional analysis of 2-bromodihydroisophoto- a- santonic lactone acetate (III). Superimposed sections of the electron-density distribution which define the stereochemistry are shown in Fig. 6. Bartons work12 showed that inversion of configuration at position 11 does not occur during the conversion of santonin

FIG. 6, Electron-density distribution in bromodi- hydroiso- a-photosantonic lactone acetate.

into this derivative, and this was confirmed by Asher and Sims later study13 of 2-bromo- a-santonin itself. To obtain still further evidence we are now engaged, at the suggestion of Professor Cocker, on an analysis

lo Clunie and Robertson. J.. 1961. 4382.

AcO Bt (VII)

In the diterpenoid series the stereochemistry of gibberellic acid (VIII) has now been established by the complete X-ray analysis of methyl bromogib- berelate ( IX ) by McCapra, Scott, Sim, and Young.16 The stereochemistry of cafestol (X) has also now been revised by their similar analysis1 of a bromo-

l1 Asher and Sim, Proc. &em. Soc., 1962, 11 1. l2 Barton, Miki, Pinhey, and Wells, Proc. Chem. SOC., 1962,151. l3 Asher and Sim, Proc. Chem. Soc., 1962, 335. l4 Barton and Levisalles, J., 1958, 4518. l6 Hamilton, McPhail, and Sim, J., 1962, 708. l6 McCapra, Scott, Sim, and Young, Proc. Chem. SOC., 1962, 185. l7 Scott, Sim, Ferguson, Young, and McCapra, J. Amer. Chem. SOC., 1962,84, 3197.

Dow

nloa

ded

by U

nive

rsity

of G

uelp

h on

18

June

201

2Pu

blish

ed o

n 01

Janu

ary

1963

on

http

://pu

bs.rs

c.org

| doi:

10.103

9/PS9

630000

229

View Online

234 PROCEEDINGS

Y .1

f I y

FIG. 7. Electron-density distribution in acetyibromogeigerin.

FIG. 8. Electron-density distribution in bromoepoxy- norcafestanone. (Reproduced, with permission, from Scott, Sim, Ferguson, Young, and McCapra, J. Amer. Chem. Soc., 1962, 34, 3197.)

Is Sim and Sutherland, unpublished results. l o Paul, Sim, Hamor, and Robertson, J. , 1962,4133.

derivative of epoxynorcafestanone (XI) which gives the well-defined electron density distribution shown in Fig. 8. Rosololactone has also been analysed as the dibromo-derivative (XII) by Sim and Suther- land.ls This work provides the stereochemistry shown in (XIII) and now establishes the position of the hydroxyl group. From these results and many related chemical studies a consistent picture of a tran~-anti-(9,IO)-backbone in the diterpenoids begins to emerge.

I would now like to mention briefly our work on a number of bitter principles, and the first two of these, clerodin and cascarillin, also belong to the diterpenoid series. Clerodin, the bitter principle from Ciei-odendrun in fortcmnatirm, was examined as the bromo-lactone (XIV), and with phasing based on the bromine atom the X-ray work established the constitution and stereochemistry of this derivative as shown in (XIV), and hence of clerodin as (XV). The X-ray workfs was carried out at a very early stage, even before the correct molecular formula for clerodin was established, but the number and kind of atoms in the molecule, as well as the entire geo-

Dow

nloa

ded

by U

nive

rsity

of G

uelp

h on

18

June

201

2Pu

blish

ed o

n 01

Janu

ary

1963

on

http

://pu

bs.rs

c.org

| doi:

10.103

9/PS9

630000

229

View Online

AUGUST 1963 235

metry, is shown clearly in the electron-density distribution (Fig. 9).

FIG. 9. Electron-density distribution in clerodin brorno-lac tone.

Our examination of the related bitter principle cascarillin, kindly supplied by Dr. Halsall, is not yet complete, but the structure and stereochemistry of the iodoacetate of the acetal, on which our work has been based, appears to be as shown in (XVI), which would indicate that the structure of cascarillin is (XVII). This is closely related to the structure suggested by Halsall apd his co-workers.20

O&H HOIH OHC

1cH;co.o

belonging to the triterpenoid family. The first of these to be determined was limonin, with an X-ray analysis based on the iodoacetate of epilirnonol. This was a major undertaking, because both the preparation of suitable crystals and the crystallography were difficult, with two complete molecules in the asymmetric unit and 228 positional parameters to determine. The electron-density distribution as finally analysed is shown in Fig. 10. It was a rewarding task, however, not only because of the interest of this structure, but because of the number of other structures that are now seen to be closely related. The structure and stereochemistry of limonin21 is shown in (XVIII), cedrelone22 in (XIX), and g e d ~ n i n ~ ~ in

ocpo & \ 0 : Q 0 Fi (xvlll) OH (XIX)

3 D

I c

(XVO (XVII)

We now come to a most interesting and important series of bitter principles and heartwood constituents

FIG. 10. Electron-density distribution in ep ihonol iodoacetate.

Birtwistle, Case, Dutta, Halsall, Mathews, Sabel, and Thaller, Proc. Chem. Suc., 1962, 329. Amott, Davie, Robertson, Sim, and Watson, J., 1961, 4183.

22 Grant, Hamilton, Hamor, Hodges, McGeachin, Raphael, Robertson, and Sim, Proc. Chem. Soc., 1961,444. 23 Sutherland, Sim, and Robertson, Proc. Chem. Suc., 1962, 222.

Dow

nloa

ded

by U

nive

rsity

of G

uelp

h on

18

June

201

2Pu

blish

ed o

n 01

Janu

ary

1963

on

http

://pu

bs.rs

c.org

| doi:

10.103

9/PS9

630000

229

View Online

236 PROCEEDINGS

ocx). They were all determined as iodoacetates, e.g., OMJ), with phasing based on the iodine atom. The results now show that they are all triterpenoids of the euphol type in different oxidation patterns, gedunin being intermediate between limonin and cedrelone.

The remaining problems whose solution I wish to describe all relate to important fungal metabolites with unusual and difficult structures. The constitutions of the first two of these, the antibiotics fumagillin and griseofulvin, were known, and our X-ray work was directed towards elucidating the stereochemistry quantitatively. Fumagillin was studied by McCorkindale and SimeM as the p-bromobenzene- sulphonate of the tetrahydroalcohol degradation product (XXII). The electron-density distribution (Fig. 11) shows the whole structure very clearly. Further refinement is proceeding, however, to measure certain interesting features more accurately, such as the internal hydrogen bonding between the tertiary hydroxyl group and the epoxide oxygen atom.

FIG. 11. Electron-density distribution in the furnagillin cierivative (XXII).

Griseofulvin, the antibiotic metabolite of Penicil- lium patulum, was analysed by Brown and Sim% as the 5-bromo-derivative, with phasing on the

24 McCorkindale and S h e , Proc. Chem. Soc., 1961, 331. 26 Brown and Sim, J., 1963, 1050. 26 Baldwin, Barton, Bloomer, Jackman, Rodriguez-Hahn,

bromine and the chlorine atom. The stereochemistry derived from this work is summarised in (XXIII).

(XXI I I)

Byssochlamic acid, the characteristic metabolite of BpsochZarnys fulva, and glaucanic and glauconic acids from Penicillium purpurogenum have been the subject of intensive chemical studies by Barton, Sutherland, and their Our X-ray work proceeded simultaneously and has resulted in a full determination of the constitutions and relative stereochemistries which are also in agreement with the chemical evidence, except that in glauconic acid ocxv) the oxygen substituent at position 4 is in the opposite configuration to that which was suggested.

A beautifully crystalline bis-p-bromophenyl- hydrazide of byssochlamic acid, which belongs to the

Y

FIG. 12. Electron-density distribution in byssochlamic acid bis-p-bromopknylhydrazide. (Repro- duced, with permission, from Hamor, Paul, Robert- son, and Sim, Experientia, 1962, 18, 352.)

and Sutherland, Experientia, 1962, 18, 345.

Dow

nloa

ded

by U

nive

rsity

of G

uelp

h on

18

June

201

2Pu

blish

ed o

n 01

Janu

ary

1963

on

http

://pu

bs.rs

c.org

| doi:

10.103

9/PS9

630000

229View Online

AUGUST 1963 237

tetragonal system, was the subject of our first X-ray determinati~n.~' The large molecule provides a well- defined electron-density distribution (Figs. 12 and 13) from which structure (XXIV) for byssochlamic acid can be derived.

P'

Y

'"40, W

FIG. 13. Atomic arrangement corresponding to Fig. 12. (Reproduced, with permission, from Hamor, Paul, Robertson, and Sim, Experientia, 1962, 18, 352.)

Glauconic acid was analysed% in the form of the rather simpler m-iodobenzoate derivative (XXV ; R = m-IC,H,.CO,). This yielded the full structure and relative stereochemistry as shown. The atomic arrangement in this derivative and the conformation of the nine-membered ring are illustrated in Fig. 14.

iium atrovenetum, was examined as the crystalline ferrichloride of atrovenetin orange trimethyl ether (XXVI). The crystal structure is complex and difficult to refine, but the results show that atrovenetin must now be represented by structure (XXVII) with the orientation of the ether ring reversed as com-

Finally, the structures of atrovenetin and the related compound, herqueinone, have now been revised by the X-ray work of Paul, Sim, and Morrison.29 Atrovenetin, the metabolite of Penicil-

27 Hamor, Paul, Robertson, and Sim, Experientia, 1962,18, 352. as Ferguson, Sim, and Robertson, Proc. Chem. Suc., 1962, 385.

Paul, Sim, and Morrison, Pruc. Chem. SOC., 1962, 352.

FIG. 14. Atomic arrangement in gfauconic acid m-iodobenzoate.

pared to the earlier formulation. It also follows that the structure of herqueinone should now be represented by (XXVIII), but the position of the methyl group may be at either R or R'.

0

(xxvr) (xxvr I) (XXVI 1 I)

This account is only a very condensed summary of our work in these fields during the past four or five years. There has not been time to discuss at all fully the many points of vital chemical interest that emerge, or to dwell upon the many difficulties and complexities of the crystal structure determinations

Dow

nloa

ded

by U

nive

rsity

of G

uelp

h on

18

June

201

2Pu

blish

ed o

n 01

Janu

ary

1963

on

http

://pu

bs.rs

c.org

| doi:

10.103

9/PS9

630000

229View Online

238 PROCEEDINGS

and the various ways in which they have been over- come. My object has been rather to present a brief picture of what has been accomplished to illustrate the potentialities of the method in the natural product field. When I add that during the same period we have also determined the structures and stereochemistries of nine complex alkaloids, the power of the method will be apparent.

In the terpenoid and fungal metabolite field my principal collaborator has been George Sim, who has been independently responsible for a great deal of the work that I have described. The names of our many:other collaborators who have played prominent

parts in this work are given in the list of references. In addition we have of course enjoyed helpful

co-operation and guidance from a very large number of organic chemists, not only from my own Depart- ment in Glasgow but from all over the country, who have suggested problems and prepared suitable derivatives.

Finally, I would like to mention that the very extensive numerical calculations involved were performed mainly on the Glasgow University DEUCE computer, and most of the programmes used were devised by Dr. J. S. Rollett and Dr. J. G. Sime.

COMMUNICATIONS

The Nucleophilic Reactivity of Akoxide and Mercaptide Ions towards Hydrogen By J. F. BUNNETT and ENRICO BACIOCCHI

(DEPARTMENT OF CHEMISTRY, BROWN UNIVERSITY, PROVIDENCE, R.I., U.S.A.)

SODIUM THIOETNOXDDE, C2H,SNa, was found by Bunnett, Davis, and Tanidal to effect elimination from aa-dimethylphenethyl chloride (I; X = Cl) about ten times as fast as does sodium methoxide. Similar observations have been made by other workers.2 These reactions were taken to be E2 PhCH,-CMe,.X + PhCH=CMe, + PhCH,*CMe = CH,

eliminations,3 and EtS- was judged to be a stronger nucleophile than MeO- towards hydrogen.

(1) (11) tlli)

MeO-

the transition state. When the C-H bond is largely broken, as in formation of menthone enolate ion, the alkoxide is the stronger reagent, whilst when the C-H bond is but slightly sundered, as in nearly El E2 eliminations, the mercaptide is the more reactive.

A test of this hypothesis was to determine the EtS-: MeO- rate ratio in elimination from structures (I) as a function of the leaving group, X. On the hypothesis formulated, this ratio should decrease as X is changed to a poorer leaving group. Accordingly,

EtS- k(EtS-)/k( MeO-) A

f . ,..-A -.--- T--L- 7 X Temp. k , (11) (%I (111) (%I k, (11) (%I (111) (%) (11) (111)

- - 6.5l 11*4l C1 75.8 SMe2+ 29.6 2 . 2 8 ~ 70.3 29.7 2 . 4 6 ~ 5.6t 2-27 0.86 0-79 SO,.Me 113-5 4.40 x 94.4 5.6 -2x lo-$ -100 -0 -0.05 V. small

limit the formation of (11).

__. - - -

* Extrapolated to zero ionic strength. 1 (I; X = SMe) is the main product. $ Slowness of reaction was shown to

However, Stauffer4 showed that sodium thioeth- oxide in ethanol is much less effective than ethoxide in effecting the isomerisation of menthone to iso- menthone, which presumably goes through an enolate-ion intermediate. The seeming contradiction between these observations finds interpretation in the hypothesis that the relative nucleophilicities of RS- and RO- reagents towards carbon-bound hydrogen depend on the degree of rupture of the C-H bond in

kinetics and products were determined for EtS- and MeO----induced eliminations from the tertiary sul- phonium salt (I; X = SMe,+) and the sulphone (I; X = S02-Me) in methanol. SMe,+ is a somewhat poorer leaving group than C1,3 and SO,.Me is much poorer. Rates were measured photometrically and products determined by gas-liquid chromatography. Results are given in the Table.

In fact, there was a great change in the direction

Bunnett, Davis, and Tanida, J. Amer. Chem. SOC., 1962, 84, 1606. de la Mare and Vernon, J., 1956,41; Eliel and Haber, J. Amer. Chem. SOC., 1959,81, 1249. Bunnett, Angew. Chem., Internat. Edn., 1962, 1, 225; Ingold, Proc. Chem. SOC., 1962,265. Stauffer, Sc.B. Thesis, Brown University, 1962.

Dow

nloa

ded

by U

nive

rsity

of G

uelp

h on

18

June

201

2Pu

blish

ed o

n 01

Janu

ary

1963

on

http

://pu

bs.rs

c.org

| doi:

10.103

9/PS9

630000

229View Online

AUGUST 1963 239

anticipated. The mercaptide reagent is slightly less effective than the alkoxide in reaction with the sul- phonium salt, and lags far behind in reaction with the sulphone. The extent of C-H breaking in the transition state, or of nucleophile to hydrogen bond formation, appears to be an important factor in- fluencing relative nucleophilicity towards hydrogen bound to carbon.

This research was supported in part by a grant from the Petroleum Research Fund administered by the American Chemical Society. Grateful acknow- ledgment is hereby made to the donors of this fund. One of us (E.B.) thanks the Italian National Research Council for a grant.

(Received, June 8th, 1963.)

The Total Synthesis of (-J-)-Clovene By P. DOYLE, I. R. MACLEAN, W. PARKER, and R. A. RAPHAEL

(CHEMISTRY DEPARTMENT, THE UNIVERSITY, GLASGOW, W.2) THE tricyclic structure (=I) assigned to clovene? one of the acid rearrangement products of caryophyllene, has now been confirmed by the following unam- biguous synthesis. The bicyclic acid2 (I) was trans- formed by the Amdt-Eistert method into the homo- logous ester (11) which was converted smoothly into a homogeneous lactone (111), m.p. 56-57", by treatment with selenium dioxide in acetic acid. Reduction of the lactone by lithium aluminium hydride gave the diol (IV) which was selectively oxidised to the hydroxy-ketone (V) by manganese dioxide. Oxida-

tion of this ketone(V) by chromium trioxide, followed by catalytic hydrogenation and esterification, led to the saturated keto-ester (VI) which was converted into the ketal-ester (VII) by treatment with ethylene glycol. frhe great susceptibility of the allylic lactone (111) to hydrogenolysis precluded its conversion into the ketal (VI) by seemingly more direct routes.]

The ketal-acid corresponding to (VII) was converted into the ketal-ketone (VIII) by treatment with ethyl-lithium, and the derived diketone cyclised under strongly basic conditions to the tricyclic cyclo- pentenone (IX). Birch reduction of the last compound gave the corresponding saturated alcohol which was then oxidised to the crystalline, sterically homogeneous cyclopentanone; this was converted into the gemdimethyl homologue (X) by the standard methylation procedure involving the methylanilino-

methylene blocking That this ketone possessed the stereochemistry as well as the carbon framework of clovene was shown by ozonolysis of its furfurylidene derivative whereby ( f)-clovenic anhydride (XI), m.p. 76-78", was obtained. The infrared spectrum of this material (in CClJ was identical with that of the anhydride (m.p. 50-51") obtained by oxidation of (-)-clovene.*

~~~~ Go (VIO (VI I I)

(XI) (X I0 Finally the ketone (X) was reduced with lithium

aluminium hydride, and pyrolysis of the carbonate of the resulting alcohol gave ( f)-clovene (XII). This material, when purified through the crystalline mixture of diastereoisomeric dibromides5 was indistin- guishable (Lr., n.m.r., mass spectra, and g.1.c.) from the naturally derived f -)-clovene.

We thank the Department of Scientific and Industrial Research (P.D. and 1.R.McL.) and the Salter's Institute (P.D.) for studentships.

(Received, June 13th, 1963.) Aebi, Barton, Burgstahler, and Lindsey, J., 1954,4660. Murray, Parker, Raphael, and Jhaveri, Tetrahedron, 1962, 18, 55. Birch and Robinson, J., 1944, 501. Gibson and Ruzicka, Hdv . Chim. A m , 1931, 570. Lutz and Reid, J., 1954, 2265.

Dow

nloa

ded

by U

nive

rsity

of G

uelp

h on

18

June

201

2Pu

blish

ed o

n 01

Janu

ary

1963

on

http

://pu

bs.rs

c.org

| doi:

10.103

9/PS9

630000

229View Online

240 PROCEEDINGS

Heat of Ionisation of Water By J. D. HALE, R. M. IZATT, and J. J. CHRISTENSEN

(DEPARTMENTS OF CHEMISTRY AND CHEMICAL ENGINEERING, BRIGHAM YOUNG UNIVERSITY, PROVO, UTAH, U.S.A.)

THE classical method of determining the standard heat of ionisation of water, AH", has been to measure calorimetrically at 25" the heat of reaction of a strong acid with a strong base at a particular ionic strength and to correct this measured heat of reaction to infinite dilution by use of appropriate heat-of-dilution data. This procedure, followed by several workers,l using A H data obtained in solutions of high ionic strength, gives AH" values ranging from 13.27 to 13.37 kcal./mole. The AH" value calculated from the temperature coefficient of the ionisation constant of water as determined from electrochemical-cell data1s2 is 13.52 kcal./mole. Since the experimental uncertainty of each of these independent methods is approximately f. 0.050 kcal./mole, the results are obviously not in good agreement .

Papee, Canady, and LaidleP were the first to measure the heat of ionisation of water, AH, in very low ionic-strength regions. Their AH" value, 13.50 -& 0-05 kcal./mole. obtained by extrapolation of d H values to infinite dilution, agrees well with values calculated from the temperature coefficient of the ionisation constant of water. The reactions studied by these workers were those of sodium hydroxide with hydrochloric and sulphuric acid. Vanderzee and Swanson4 report a AH" value calculated from A H values for the reaction of perchloric acid with sodium hydroxide in a low ionic-strength region to be 13.336 f 0.009 kcal./mole. This value, obtained by correction of A H values to infinite dilution with appropriate heat-of-dilution data, agrees well with the values obtained by the classical calorimetric procedure.

Vanderzee and Swanson4 carried out their calorimetric study in the same concentration range as that used by Papee et al.;3 however, the AH" values obtained in the two studies differ appreciably (200 cal.). It would not be expected that in this low ionic- strength region the difference between the heats of neutralisation of hydrochloric and perchloric acid with sodium hydroxide would be as great as this difference indicates.

An uncertainty in the value of AH" is particularly unfortunate since most reactions in aqueous solution involve pH changes. In addition, acid-base

reactions are frequently used to calibrate calorimetric equipment. An accurate knowledge of the standard heat of ionisation of water at infinite dilution and how this heat varies through low ionic- strength regions is of particular importance since it is in low ionic-strength regions that much calorimetric work is now being carried out.

We have just completed a study of the heat of neutralisation of both hydrochloric and perchloric acid with sodium hydroxide in a low ionic-strength region comparable with that used by Papee et aL3 and by Vanderzee and Swanson? The A H data were obtained by using a non-isothermal, constant- temperature-environment calorimeter. The results are given in the Table.

A H Values (kcaE./mole) for reaction of NaOH with HC1 and HClO,.

Values are averages of at least 10 determinations at each p value where initial and base concentrations

are equal to final salt concentrations.

HCl 0.00496 13.375 0.0171 13.405 0.0341 13-430

HCIO, 0.00496 13.370 0.0171 13.395 0-0341 13.410

Acid P AH

AH" values were obtained from the tabulated dlrr data both by extrapolation to p = 0 of the linear plot of d H against P**~, and by correction of the individual A H values to infinite dilution by using heat-of-dilution data. The AH" value obtained by either extrapolation method from determinations with both acids is 13.337 f 0.015 kcal./rnole, in excellent agreement with the AH" value reported by Vanderzee and Swanson.*

Acknowledgement is made to the United States Atomic Energy Commission and National Institutes of Health for financial support.


A review is found in C. T. Mortimer, "Reaction Heats and Bond Strengths," Pergamon Press, London, 1962, p. 166. Harned and Owen, "The Physical Chemistry of Electrolytic Solutions," 3rd ed., Reinhold Pub. Corp., New York,

1958, p. 754.

Vanderzee and Swanson, J. Phys. Chem., 1963, 67, 285. The experimental details have not yet been published; however, a full account is given by J. A. Swanson, Ph.D. Thesis, University of Nebraska, January 16th, 1962; cf. Dzss.

9 Papee, Canady, and Laidler, Canad. J . Chem., 1956, 34, 1677.

Abs., 1962, 62-2688.

Dow

nloa

ded

by U

nive

rsity

of G

uelp

h on

18

June

201

2Pu

blish

ed o

n 01

Janu

ary

1963

on

http

://pu

bs.rs

c.org

| doi:

10.103

9/PS9

630000

229View Online

AUGUST 1963 24 1

The Hexamethylbenzene Cation By ROGER HULME and M. C. R. SYMONS

(DEPARTMENT OF CHEMISTRY, THE UNIVERSITY, LEICESTER)

ALTHOUGH cations formed by loss of one n-electron from polynuclear aromatic hydrocarbons are well known, attempts to prepare cations from mono- nuclear hydrocarbons have been unsuccessful.lt2 In view of its relatively low ionisation potential (7-8 ev), hexamethylbenzene should form a stable cation. However, our attempts to prepare this ion by chemical oxidation have failed.

We now report the detection of a radical formed by photolysis of a solution of hexamethylbenzene in sulphuric acid by use of a high-pressure mercury lamp. The electron spin resonance spectrum of this radical is in good accord with expectation for the cation of hexamethylbenzene, provided that all the methyl groups are magnetically equivalent.

Only 13 of the expected 19 lines have been detected since we have not been able to accumulate sufficiently high concentrations of this radical. How- ever, relative intensities of these lines are very close to those expected for a radical with eighteen equivalent protons and definitely exclude radicals with twelve or fourteen equivalent protons.

Attempts to reduce the observed line width to below 250 milligauss failed, although the Varian spectrometer used has resolved lines with widths in the region of 30 milligauss. Further, as has been found for benzene anions: the spectrum is not saturated at high R.F. power levels. These results are characteristic of radicals with the unpaired electron in an orbitally degenerate level, as expected for this cation.

The isotropic hyperfine constant of 6-45 gauss is in fair agreement with expectation, corresponding to a value of Qs, in the relation aaH = &pC, of 38.7. Here, as in other instances of coupling to /%protons in radical hyperconjugation appears to be particularly important.

We thank the Department of Scientific and Industrial Research for financial assistance and Shell Research Ltd., for a grant to R.H.


Bolton and Carrington, Proc. Chem. SOC., 1961, 385. Brivati, Hulme, and Symons, Proc. Chem. Soc., 1961, 384. Tuttle and Weissman, J . Amer. Chem. SOC., 1958, 80, 5342. De Boer and Mackor, MoE. Phys., 1962,5493; Bolton, Carrington, and McLachlan, ibid., p. 31.

The Formation of Abnormal Valency States in the Radiolysis of Aqueous Metal-ion Solutions By G. E. ADAMS, J. H. BAXENDALE, and J. W. BOAG

(RESEARCH UNIT IN UDIOBIOLOGY, B.E.E.C., MOUNT VERNON HOSPITAL, NORTHWOOD, and CHEMISTRY DEPARTMENT, UNIVERSITY, MANCHESTER)

IT has been concluded from the effect of certain metal ions, such as Zn2+, Ni2+, Co2+, Cd2+, and Pb2+, on the yields of hydrogen from y-irradiated neutral aqueous 0.1M-methanol solutions, that the ions react rapidly with the hydrated electron, produced in these conditions, to give ions in abnormal valency states as transient intermediates1

We have now obtained (Fig. 1) the absorption spectra of some transient species using the technique of pulsed radiolysis.2 They were taken 2 psec. after solutions of Zn2+, Cd2+, Co2+, and Mn2+ in de- grated 0.1 M-aqueous methanol had been irradiated by a 2 psec. pulse of 2Mev electrons from a linear accelerator. No absorption is produced in the same conditions by 0.1M-methanol in the absence of the metal ions, and change from 0 . 1 ~ to 1 0 - 3 ~ methanol

has no effect on the spectrum given by Zn2+ solutions. Except for a decrease in density, the spectra from Zn2+ and Cd2+ are the same with methanol a b ~ e n t . ~ Hence they probably originate in the metal ions and are due to the lower valency states or some other form of association of M2+ with an electron.

The absorption produced with Mn2+ seems sur- prising since it had no effect on the yield of hydr0gen.l The higher concentration used here is probably responsible since absorption is not obtained in similar conditions3 with 2 x lo-*~-Mn~+.

The absorptions decay at about the same rate and we have examined the Zn2+ solution in most detail (Fig. 2). The absorption maximum, at 3000 A decreases according to second-order kinetics but after it has disappeared (100 psec.) there is still consider-

Baxendale and Dixon, Proc. Chem. Soc., 1963, 148. Hart and Boag, J. Amer. Chem. Sac., 1962, 84,4090. Baxendale, Fielden, and Keene, following communication.

Dow

nloa

ded

by U

nive

rsity

of G

uelp

h on

18

June

201

2Pu

blish

ed o

n 01

Janu

ary

1963

on

http

://pu

bs.rs

c.org

| doi:

10.103

9/PS9

630000

229View Online

242 PROCEEDINGS

> .c

0 8

0.6 x 4 u)

W 0

.-

0.4 .- 4

0-2

28 32 36 40 Wave length (IOQA ).

RG. 1. Transient absorption of lO-%-solutions in 0-1M-detzrated aqueous methanol ( IO"M for Zn2+) afer irradiation by an electron pulse.

able absorption below 2700 8,. The presence of 1 O-2N-sulphuric acid decreases considerably the absorption at 3000 8, (Fig. 2), as might be expected since H+ and Zn2+ will compete for the electron. Acid also removes the steeply rising absorption below 2700 8,.

These observations may be understood in terms of the further reduction in neutral solution, of Zn+ to the atom (Zn metal has been observed as a product1) either by dismutation or by reaction with the radical

I I I I r'\c I I I 1 22 24 26 28 30 32 34 36 38

Wavelength (lo"&) FIG. 2. Transient absoration of 10-2M-Zn2+ in

1O3wdetzrated aqueous methanol "at various times after the electron pulse.

CH,.OH which will be present in almost equal concentration. Thus at short delay times the absorption spectrum. of the solution is a composite of those of Zn+ and ZnO, the latter being responsible for the rising absorption at the lower wavelengths. In lo-%- acid this component is absent since Zn+ is oxidised back to Zn2f by H+, and Zno is not formed.

(Received, May 7th, 1963.)

Absolute Rate Constants for the Reactions of Some Metal Ions with the Hydrated Electron By J. H. BAXENDALE, E. M. FIELDEN, and J. P. KEENE

(CHEMISTRY DEPARTMENT, UNIVERSITY, MANCHESTER 1 3, and PATTERSON LABORATORIES, CHRISTIE HOSPITAL, MANCHESTER)

RECENT work1 has shown that the hydrated electron We have now followed the decay of this absorp- produced by y-radiation in aqueous solution will tion, produced by a 2 psec. pulse of 4 MeV electrons reduce Zn2+, Cd2+, Co2+, Ni2+, and C0(NH3);f ions. in the presence of the above metal ions, using a sensi- Other work2y3 has established that the hydrated tive photomultiplier/oscilloscope technique: and electron in water has a broad absorption in the find that the decay rate is considerably increased. We visible spectrum with a maximum at 7000 A. have been able to obtain conditions where the decay

Baxendale and Dixon, Proc. Chem. SOC., 1963, 148. Boag and Hart, J. Aier. Chem. Soc., 1962,84,4090.

Keene, to be published. a Keene, Nature, 1963, 197, 47.

Dow

nloa

ded

by U

nive

rsity

of G

uelp

h on

18

June

201

2Pu

blish

ed o

n 01

Janu

ary

1963

on

http

://pu

bs.rs

c.org

| doi:

10.103

9/PS9

630000

229

View Online

AUGUST 1963 243

of the hydrated electron in pure water is negligible compared with that in the presence of 50-500 ,UM- metal ions which is a first-order process. From these and similar data we have obtained the bimolecular rate constants given in the Table, referring to the hydrated electron-ion reaction.

Metal ion .. .. Znw COS 10-lo k(mo1e-l 1. sec.-l) . . 0.17 1.35

- ' * ~ n

2 Y \\. 400 5 0 0

Wavelength (my)

Transient ultraviolet spectrum produced in 5.3 x 10-5~-NiS04 solution, taken 2 psec. after the ead of the electron pulse.

The rate for Cu2+ compares well with the previously determined value of 3.3 f 0.3 x 1 0 ' O mole-l l.sec.-l.6 The results are fairly consistent with the semi- quantitative data: and we also confirm that the Mn2+ reaction is at least an order of magnitude slower than the slowest measured.

Ni2+ c u 2 + Cd2+ Co(NH&+ 2.3 3.0 5.8 > 9

Transient absorption has been observed at - 3000 A in these solutions, and ascribed6 to the unstable valency states produced in these reactions, e.g., Zn+, We have confinned this using the photomultiplier technique and have also observed the ultraviolet transient absorption spectrum for Ni+ shown in the Figure.

In support of this interpretation we find that all these absorptions build up in the time-period observed for the enhanced electron decay. Further, the effect of the ions on the build up of the hydrated- electron absorption during the pulse confirms the order of reactivity of the ions given in the Table. It seems probable that the transient ultraviolet absorptions are the charge-transfer spectra of the reduced ions, the molar extinction coefficient of the Zn+ species being at least 1-3 x l@.

(Received, May 7th, 1963.) Gordon, Hart, Matheson, Rabani, and Thomas, in the press. Adams, Baxendale, and Boag, Proc. Chem. Soc., 1963, preceding communication.

A Direct Oxidation of cis-4Hydroxycinnamic Acid to Umbelliferone By M. B. MEYERS

(DEPARTMENT OF CHEMISTRY, THE UNIVERSITY, GLASGOW) THE conversion of 4-hydroxycinnamic acid into glucoside, which after isomerisation to the cis-acid is umbelliferone (I) and its derivatives in plant systems1 dehydrated to the lactone.6 involves an apparent oxidation meta to the estab- It is now shown that treatment of aqueous solu- lished phenolic group. Of the several theories proposed, to rationalise this mechanistically unfavour- able process, two involve oxidative cyclisation

(III), where subsequent migration of the inserted oxygen would lead -to umbelliferone. Alternatively a direct cyclisation by carboxyl radical or cation species has been envisaged.* In the biosynthesis of coumarin it is known that ortho-hydroxylation prob-

m* 0-p p 2 H 0~

- HO\ through a quinol intermediate2 (11) or a spirolactone3 (1) (n)

Meo' OGlU - 0 ably takes place in trans-cinnamic acid to give a (m) (IV)

Brown, Science, 1962, 137, 977. Waworth, J., 1942,448. Grisebach and Ollis, Experientia, 1961, 17, 4; Scott, Proc. Chem. Soc., 1962, 207. Chambers, Kenner, Temple-Robinson, and Webster, Proc. Chem. SOC., 1960,291. Brown, Canad. J. Biochem. Physiol,, 1962, 40, 607; Stoker and Bells, J. Biol. Chem., 1962, 237, 2303; Kahnt,

Naturwiss., 1962, 49, 207; Brown, Towers, and Wright. Canad. .J. Biochem. Physiol., 1960, 38, 143.

Dow

nloa

ded

by U

nive

rsity

of G

uelp

h on

18

June

201

2Pu

blish

ed o

n 01

Janu

ary

1963

on

http

://pu

bs.rs

c.org

| doi:

10.103

9/PS9

630000

229View Online

244 PROCEEDINGS

tions of cis-4-hydroxycinnamic acid in the presence of ferrous-ethylenediaminetetra-acetic acid complex and ascorbic acid (phosphate buffer pH 4-6) with air or oxygen at room temperature (19) for 12-18 hours produces umbelliferone (0.1 % yield) and 6,7-dihydroxycoumarin, the latter in similar but variable yields. Practically no umbelliferone is detected in the absence of ascorbic acid and, as the metal ion-ascorbic acid system is a hydroxylating reagent,s participation of the carboxyl group may be excluded. Under these mild conditions the plausible intermediate (III)* does not undergo rearrangement and known compounds structurally similar to (11) are likewise insensitive to acids. Moreover, the forced acidic rearrangement of compound (111) leads almost entirely to 6-hydroxycoumarin through carbon migration, and an acid (IT) would be expected to behave in the same way: no 6-hydroxycoumarin was found among the products. Haworth2 has suggested that umbelliferone could be formed from an acid (11) by lactonisation on to the b-position of the a/hnsaturated ketone, followed by dehydration. This is unlikely as treating the spirolactone (111) with one equivalent of sodium hydroxide gives no trace of umbelliferone but a mixture of 6-hydroxy-

coumarin and the hydroxy-acid corresponding to 01).

It is not very probable that the second oxidation product, 6,7-dihydroxycoumarin, is formed by further oxidation of umbelliferone as the latter is not appreciably oxidised under these conditions. Its likely mode of formation involves hydroxylation ortho to the phenol group of cis-4-hydroxycinnamic acid to give cis-caffeic acid which is then further hydroxylated to 6,7-dihydroxycoumarin. The latter conversion has been observed in these circum- stances.8

The above results suggest that the biosynthesis of 7-oxygenated coumarins is related to that of coumarin, without participation of the phenol or carboxyl group in the first instance, and that it is unnecessary to use quinol intermediates. This view is supported by Browns isolation1 of the 2-glucoside of 2- hydroxy-4-methoxy-cicinnamic acid (IV).

The author thanks Mr. D. J. Austin for assistance with the experimental work. This investigation was carried out during the tenure of an I.C.I. Fellowship.

(Received, June 12th, 1963 .)

* Preparation of this compound by electrolysis of cis4hydroxycinnamic acid will be described later.9

* Butler and Siegelman, Nature, 1959,183, 181 3 ; Van Sumere, Parmentier, and Van Poucke, Naturwiss. , 1959,46,668. Norman and Radda, Proc. Chem. Soc., 1962, 138; Green, Ralph, and Scofield, Nature, 1963, 198, 754. Bamberger, Ber., 1900,33, 3652.

Scott, Dodson, McCapra, and Meyers, J. Amer. Chem. SOC, in the press.

The Effect of Cup& and Thallous Ions on the Radiolysis of Aqueous Solutions of Ethylenediamine, Propane-i,Z-diamine, and Glycine

By M. ANBAR (ISOTOPE DEPARTMENT, THE WEIZMANN INSTITUTE OF SCIENCE, REHOVOTH, ISRAEL)

and P. RONA (ATOMIC ENERGY COMMISSION, SOREQ RESEARCH ESTABLISHMENT, REHOVOTH, ISRAEL)

THE rate and mode of single-electron reactions of organic compounds are modified on complex-formation with metal ions.lY2 The extent of radiolytic de- carboxylation of salicylic acid, in highly concentrated solutions, increases in the presence of ferric and cupric ions.3 However, since the effect of appropriate scavengers was not investigated, it is hard to interpret these results.

In the present study, thallous or cupric sulphate in dilute aqueous solution is shown to induce radiolysis of certain organic compounds in the presence of scavengers, which otherwise co.rlpletely inhibit their decomposition.

Oxygen-saturated solutions of ethylenediamine.

propane-l,2-diamine, and glycine in triply distilled water, were irradiated with sCo-gamma rays, at a dose rate of 12,400 rad./min., with total doses of 100,000---150,OOO rads. The radiolysis of ethylenediamine in dilute aqueous solution is characterised by the production of ammonia, and was shown, by mercuric oxide pre~ipitation,~ to be the sole volatile base produced. The Conway microdiffusion method5 was used for the standard determination of the yields of ammonia. Another major product is glycinealdehyde, which was identified and determined after oxidation to glyoxal. Glyoxal was shown not to be present, as such, in the irradiated solution. The radiolysis yields of glycinealdehyde and hydrogen

Gritter and Patmore, Proc. Chem. SOC., 1962, 328. Anbar, Guttman, and Friedman, Proc. Chem. SOC., 1963, 10. Sugimori and Tsuchihashi, Bull. Chem. SOC. Japan, 1960,33, 713. Weber and Wilson, J. Bid. Chem., 1918, 35, 385. Conway, Microdiffusion Analysis and Volumetric Error. Lockwood, London, 1947.

Dow

nloa

ded

by U

nive

rsity

of G

uelp

h on

18

June

201

2Pu

blish

ed o

n 01

Janu

ary

1963

on

http

://pu

bs.rs

c.org

| doi:

10.103

9/PS9

630000

229View Online

AUGUST 1963 245

peroxide were determined spectrophotometrically:6 G(NH,CH,CHO) = 1.05 and G(H,O,) = 2.35 were found in 0.3~~ethylenediamine solutions. No organic hydroperoxide was detected in the irradiated solutions as the peroxide yields determined by the titanium sulphate and the iodide methods were identical. The G-value of ammonia increased with the ethylenediamine concentration up to 0 .03~; at

the compounds studied. However, substantial nonscavengable G(NHJ values were found in the presence of low concentrations of either thallous sulphate or cupric sulphate.

As shown in Table 1, propan-2-01 and formic acid inhibit the formation of ammonia completely, and benzoic acid to a great extent. When thallous or cupric sulphate was added, the G(NH,) values were

TABLE 1. Yield of ammonia, in oxygen-saturated amine solutions at pH = 5, irradiated with 120,750 rads of 6oCo-gumma rays.

(en = ethylenediamine; pn = propane-1,Zdiamine; gly = glycine)

Scavenger Metal ion

None None Tl,S04 (0406~)

&so4 (0012M) None

Propan-2-01 (0.03~) Tl,SO, (0.006~) CUSO4 (O'O12M) None

Formic acid (0.03~) Tl,S04 (0.006~) CUSO4 (0012M) None

Benzoic acid (0.015~) T12S04 (0 .006~) cUso4 (0.012M)

en (0.03~)

2.75 2.75 3.6 0.0 2.0 3-6 0.0 2.0 2.7 0.0 0.9 3-4

G(NHd* Pn

(0.03~) 2.75 2-75 2.75 00 1.8 2.3 0.0 1 *6 1.1 033 2.7 4.7

g l Y (O'WM)

1.33 1.33 2.45 0.0 0.44 0.66 0.0 0.0 0 7 0 4 1.55 2.3

* Each G-value quoted is the result of at least four analyses.

TABLE 2. The effect of EDTA andof acidity on the action of metal ions on ethylenediarnine solutions.

(Experimental conditions are as in Table 1) Scavenger Metal ion G(NHJ

EDTA (0.005~) None 0-0 EDTA (0.005~) n2s04 (0.006M) 0.0 H2S04 ( 0 . 8 ~ ) None 1.6 H,S04 ( 0 . 8 ~ ) T12S04 (0 .06~) 0.0 H2S04 ( 0 . 8 ~ ) &SO4 (0.03~) 1.6

H2S04 ( 0 . 8 ~ ) + Propan-2-01 (0.03~) &SO4 (0.03~) 0-4 H2SOp ( 0 . 8 ~ ) + Propan-2-01 (0.03~) None 0.0

this concentration, a plateau value of G(NH$ = 2.75 f 0.30 was obtained. This yield was independent of acidity between pH 3.0-5.0. In radiolysis of solutions saturated with nitrous oxide G(NH$ increased to 6.3, so that the reaction seems to be initiated by OH radicals,7 in agreement with previous results on amines.8

The addition of various scavengers, at equimolar concentrations, completely inhibited the radiolysis of

close to those obtained in the absence of scavengers. T l s could not be detected in the irradiated solutions by spectrophotometric analysk9 EDTA inhibited the action of thallous sulphate. In @8~-sulphuric acid cupric sulphate affected the scavenging action of propan-2-01 to a smaller extent than in neutral solution, whereas thallous sulphate was a scavenger even in the absence of propan-2-01 (Table 2).

These metal ions produced similar effects when Neuberg and Straws, Arch. Biochem., 1945, 7 , 211; Satterfield and Bonnell, Analyt. Chem., 1955, 27, 1174;

Hocwadel, J. Phys. Chem., 1952,56,587.

Jayson, Scholes, and Weiss, J., 1955,2594; Jayko and Garrison, J. Chem. Phys., 1956,25,1084. Bode, 2. Analyt. Chem., 1955,144, 165.

'I Dmton, Peterson, and Sills, Discuss. Faradby SOC., 1960,29, 257.

Dow

nloa

ded

by U

nive

rsity

of G

uelp

h on

18

June

201

2Pu

blish

ed o

n 01

Janu

ary

1963

on

http

://pu

bs.rs

c.org

| doi:

10.103

9/PS9

630000

229

View Online

246 PROCEEDINGS

added to ethylenediamine solutions which were irradiated in the absence of oxygen (by saturation with argon). The presence of nonscavengable G(NH3) could not be demonstrated in ethylenediamine solutions containing nickel or caesium sulphate or lead perchlorate. The organic scavengers (Table 1) inhibited radiolysis completely in propane-l,3-diamine and in butane-lP-diamine solutions also in the presence of thallous sulphate. Cupric sulphate had no effect on the radiolysis of ethylenediamine solutions saturated with nitrous oxide, in the presence or absence of scavengers. Thallous ions, on the other hand, inhibited the scavenging action of propan-2-01 in the presence of nitrous oxide.

lo Sworski, Radiation Res., 1956, 4, 483.

These results indicate that metal ions act by complex formation with the substrate undergoing rddio- lysis. The absence of any effect of nickel ions implies that this is not due merely to complex-formation. It seems likely that thallous and cupric ions become involved in electron-transfer processes. It is suggested that cupric ions act by reduction to Cu+ since they have no effect under nitrous oxide. Thallous ions, on the other hand, may act via the T1+2 state, formed by their reaction with OH radicals.1

The authors are indebted to Mr. R. A. Munciz for

(Received, March 28th, 1963.) help in performing part of the analyses.

Structure of Amorphigenin, the Aglycone of the First Natural Rotenoid Glycoside By L. CROMBIE and R. PEACE

(DEPARTMENT OF CHEMISTRY, UNIVERSITY OF LONDON KINGS COLLEGE, STRAND, LONDON, W.C.2)

AMORPHIN, a glycoside which occurs in the insecti- cidal seeds of Amorpha fruticosa,l gives glucose, arabinose, and amorphigenin on acid hydrolysis. The genin is also formed by hydrolysis with p-glucosidase and gives a positive Durham test, like rotenone. Russian authors2 recently have presented further evidence of its rotenoid character and suggest that it has the molecular formula C22H2oO7 (Acree and co- workers1 proposed C22H22O7) and possesses a D/E system as in (I) with the hydroxyl group unassigned and the rest of the molecule as in rotenone. Our results lead to structure (11) for amorphigenin, C23H2207 : amorphin, with the sugar attachment at 8, is thus the first rotenoid glycoside.

Amorphigenin forms tenaceously solvated crystals (nuclear magnetic resonance evidence) from methanol, benzene, and aqueous acetone, and the molecular formula was established from analysis of derivatives, mass-spectral molecular weight (410), and nuclear magnetic resonance proton counts on amorphigenin and its derivatives. Infrared and ultraviolet spectra simulate closely those of rotenone except for hydroxyl absorption (vmax, 3498 cm.-l in CCl,, c < 0 . 0 0 5 ~ ; intramolecular bonding to 1-0), and the nuclear magnetic resonance spectrum shows two methoxyl groups with an aromatic proton pattern as for rotenone? The 1- and 4-hydrogen atoms give peaks at r 3-19 and 3.52, respectively, and the 10,l l-hydrogen atoms give rise to the characteristic quartet, r 3.47 and 2.12 (J 9 c./sec.), leaving no

doubt that the E ring is angularly fused as shown.

Chemical description of the B/C rings is given by 6a, 12a-dehydrogenation which occurs with characteristic shift of the carbonyl frequency (1672 cm.-l for amorphigenin, 1634 cm.-l for dehydroamorphi- genin) and r value of the l-hydrogen atom (1.90). Treating the dehydro-compound with nitrous acid affords the keto-lactone (III), verifying the presence of the 5,6,6a,12a,12-system as in (11). The 6a,12a- dehydro-compound is hydrolysed to the derrisic acid analogue (IV), which is recyclised by hot acetic anhydride to the 8-acetate of the 6a, 12a-dehydro- compound.

Acree, Jacobson, and Haller, J. Org. Chem., 1943, 8, 572.

Crombie and Lown, J., 1962, 775.

a Kondratenko and Abubakirov, Doklady Akad. Nauk S.S.R., 1962,146, 1340; Uzbek. Khim. Z h r . , 1962,6,60, 7 3 ; 1961,5, 66; Doklady Akad. Nauk Uzbek. S.S.R., 1960, 35.

Dow

nloa

ded

by U

nive

rsity

of G

uelp

h on

18

June

201

2Pu

blish

ed o

n 01

Janu

ary

1963

on

http

://pu

bs.rs

c.org

| doi:

10.103

9/PS9

630000

229View Online

AUGUST 1963 247

The primary nature of the hydroxyl group is shown by two pieces of information. First, the 8-methylene band at r 5.73 in amorphigenin (and its derivatives) moves to 5.33 on acetylation (the unsplit band also demands no proton on the carbon bearing the CH,-OH group). Secondly, on careful oxidation with manganese dioxide amorphigenin is oxidised to a 6-formyl-l2-ketone, with an aldehydic hydrogen peak at r 0.46. The corresponding hydroxyroten- one (cf. 111) gives an analogous aldehyde when oxidised. The oxidation conditions, and increased absorption at 225 mp, support an allylic primary alcohol system, and amorphigenin absorbs one mol. of hydrogen over a palladium catalyst. In addition, amorphigenin has a band at T 4.73 assignable to the two vinyl protons (cf. rotenone). With this in mind, the two carbon atoms unassigned must form a di- hydrofuranoid ring E. The whole 5-carbon addendum to ring D is isoprenoid, since attachment of the acyclic residue is at 5 and not 4. This is so because the three hydrogen atoms at 5 and 4 form multiplets below T 5.1 (S, one proton) and near 6.8 (4, two protons) with a splitting pattern similar to the corresponding situation in rotenone.

The stereochemistry of amorphigenin is assigned as follows. The r value of the 1-hydrogen atom in

amorphigenin indicates a CiS-B/C fusion: and the positive Cotton effect (first extremum 357 mp) sup- ports this and indicates that the absolute conflgura- tion at positions 6a,12a is as in rotenone? As both derrisic acid and the corresponding compound from amorphigenin (cf. IV) [both of which possess asym- metry only at 51, have similar negative plain curves, the rotenone configuration is assigned at this centre.5

This information defines completely the structure and stereochemistry of amorphigenin. The conclu- sions have been confinned by hydrogenolysing the 8-hydroxyl group in amorphigenin and hydro- genating the 6,7-olefinic link. The product is identical with (-)-6,7-dihydror~tenone~ from natural rotenone. (-)-12-Deoxy-6,7-dihydroro- tenone is also formed in the hydrogenolysis by elision of the keto-group.

We are indebted to Dr. J. W. Lown (University of Alberta) for the optical rotatory dispersion curves and for some preliminary nuclear magnetic resonance spectra. Dr. R. I. Reed (University of Glasgow) kindly provided the mass spectral molecular weight. One of us thanks the D.S.I.R. for a postgraduate award and we are grateful to Shell Research Ltd. for financial support. (Received, May 23rd, 1963.)

Djerassi, Ollis, and Russell, J., 1961, ld8. Cf. Crombie and Peace, J., 1961, 5445. LaForge and Keenan, J. Amer. Chem. SOC., 1931,53,4450.

An 8-Co-ordinate Compound of Zinc(@ By D. P. GRADDON and D. G. WEEDEN

(DEPARTMENT OF INORGANIC CHEMISTRY, UNIVERSITY OF NEW S o m ~ WALES, SYDNEY,

SEVERAL first-row transition metals have recently been shown by structural studies to form compounds in which the co-ordination number exceeds six; thus, manganese and iron have been shown to form 7-co-ordinate complexes with ethylenediamine-tetra- acetic acid: and titanium tetrachloride to form an 8-co-ordinate adduct with o-phenylenebisdimethyl- arsine.,

By crystallisation of bis(dibenzoyhethanato)- zinc(@ from 4-methylpyridine we have obtained a tetra-(4methylpyridine) adduct as yellow needles, m.p. 95 (decomp.) (Found: C, 73.6; H, 5.9; N, 6.3; Zn, 7.3. C,,H,,N,O,Zn requires C, 73.3; H, 5.7; N, 6.3; Zn, 7.4%).

This apparently 8-co-ordinate compound could achieve a co-ordination number of six if two of the molecules of 4-methylpyridine filled spaces in the

AUSTRALIA) crystal lattice or if the chelate rings were opened. However, determination of the molecular weight in benzene gave a mean value of 789 (theor. for mono- mer 884), showing that only slight dissociation occurs, and the infrared spectrum in the 6 p region showed two very strong bands at 1520 and 1600 cm.-l, characteristic of chelated p-diketones; open- ing of the chelate rings would be expected to result in the disappearance of these bands and the appear- ance of a new band near 1700 cm.-l, characteristic of the conjugated carbonyl group then present, as found recently for some /?-diketone complexes of mercury(@, for which the open enolate structure has been ~uggested.~

We have also obtained the corresponding nickel@) compound.

(Received, June 7th, 1963.) Hoard, Pedersen, Richards, and Silverton, J. Amer. Chem. SOC., 1961, 83, 3533; Hoard, Lind, and Silverton, ibid.,

Clark, Lewis, NyhoIm, Pauling, and Robinson, Nature, 1961, 192, 222. Nonhebel, J. , 1963, 738.

p. 2770.

Dow

nloa

ded

by U

nive

rsity

of G

uelp

h on

18

June

201

2Pu

blish

ed o

n 01

Janu

ary

1963

on

http

://pu

bs.rs

c.org

| doi:

10.103

9/PS9

630000

229View Online

248 PROCEEDINGS

The Synthesis of a New Fragmentation Product of a Cephalosporanic Acid Derivative By S. H. EGGERS, T. R. EMERSON, V. V. Urn, and G. Lowe

(THE DYSON PERRINS LABORATORY, OXFORD UNIVERSITY)

THE elucidation of the structure of cephalosporin C (I; R = D--O,C.CH(NH~+).[CH,]~CO)~ an antibiotic from a species of Cephalosporium, revealed a close similarity to the penicillins (11; R = acyl) and in particular to penicillin N* (11; R = D--O,CCH(NH,+)+ [CH,],-CO ).2 Unlike the penicillins it was not inactivated by penicillinase,3 but had much weaker antibacterial activity? It has been shown, however, that replacement of the a-amino- adipyl side chain by other acyl residues can greatly increase this ac t i~ i ty .~ .~ The elegant method now available for the preparation of 7-aminocephalo- sporanic acid (I; R = H) has made a wide range of acylated derivatives easily available.6 We report the transformation of one of these, namely, 7- phenylacetamidocephalosporanic acid6 (I ; R = PhCH,CO), into the 6H-1,3-thiazine (IV) and describe its synthesis.

( I ) C02H

I I- -

+

The inactivation of the penicillins by alcohols is known to be due to the cleavage of the p-lactam ring to give the corresponding penicilloic ester (III);8 it is

claimed that metal ions are required for this reaction. Similar treatment of the acid (I; R = Ph*CH,CO) in the presence or absence of metal salts failed to cleave the p-lactam ring.

When the acid (I; R = Ph-CH,CO) was treated with two equivalents of sodium benzyl oxide in benzyl alcohol, however, it was smoothly trans- formed in 1 hr. at 15 into the acid (IV; R = PhCH,-CO, R = H) which was isolated in 60% yield. The acid (IV; R = PhCH,CO, R = H) and its benzyl ester (IV; R = Ph.CH,*CO, R = PhCHa were identical with synthetic materials in ultraviolet, infrared, and nuclear magnetic resonance spectra, melting point and mixed melting point, and elemental analysis. Their nuclear magnetic resonance spectra indicated that they exist largely (> 80%) in the enolic form, as shown.

This transformation of the cephalosporanic acid nucleus can be regarded as an extended fragmentation reaction (cf. ref. 9), the product of which under- goes a prototropic rearrangement to give the 6H-lY3-thiazine system. The facility of the reaction and the high yield of the fragmentation product is unusual for a molecule in which the leaving group is at a primary centre. It seems likely therefore that the fragmentation is a concerted reaction, the driving force arising largely from the cleavage of the p-Iactam ring.

The synthesis of this new fragmentation product was achieved by the annexed route.

(9 (ii)

(iii)

CH,(CN)*COg*CH,Ph -+ H0.N = C(CN)*CO,.CH,Ph +

Ph*CH2*CO*NH*CH(CN).CO2CHaPh -+

Ph.CHg.CO*NH.CH(CS.NHJ*COa*CHgPh \ (V) p+ (tV)

(iv) / CH2= CMeCH(OH)*C02R -+ CH,=CMeCOCO,R

(W (I) (i) NaN0,-AcOH. ( i i ) At-Hg, Ph-CH,-COCt. (iii) H2S. (iv) MnO,. (v) HCI.

* Also known as cephalosporin N and synnematin B, see ref. 2. Abraham and Newton, Biochem. J., 1961, 79, 377; Hodgkin and Maslen, ibid., P. 393. Abraham, Newton, and Hale, Biochem. J., 1954, 58, 94; Newton and Abraham, ibid., p. 103; Abraham, Phmm.

Abraham and Newton, Biochem. J., 1956,63, 628. Newton and Abraham, Biochem. J., 1956,62, 651. Loder, Newton, and Abraham, Biochem. J., 1961, 79,408.

Rev., 1962, 14, 473.

Chauvette, Flynn, Jackson, Lavagnino, Morin, Mueller, Pioch, Roeske, Ryan, Spencer, and Van Heyningen, Morin, Jackson, Flynn, and Roeske, J. Amer. Chem. Soc., 1962,84, 3400. * The Chemistry of Penicillin, ed. C!arke,,, Johnson, and Robinson, Oxford Univ. Press, 1949.

J. Amer. Chem. Soc., 1962, 84, 3401.

Grob in Theoretical Organic Chemistry, KekulC Symposium, Buttenvorths, London, 1958, p. 114.

Dow

nloa

ded

by U

nive

rsity

of G

uelp

h on

18

June

201

2Pu

blish

ed o

n 01

Janu

ary

1963

on

http

://pu

bs.rs

c.org

| doi:

10.103

9/PS9

630000

229View Online

AUGUST 1963 249

The hydroxy-ester1 (VI; R = Et) on transesteri- fication with benzyl alcohol in the presence of toluene-p-sulphonic acid gave the hydroxy-ester (VI ; R = PhCH,). Both hydroxy-esters were oxidised to the corresponding keto-esters (VII; R = PhCH, and Et) with manganese dioxide. The keto-ester (VII

Selective hydrolysis of both the thiazine-esters (IV; R = PhCH,CO, R' = Et and Ph-CH,) was effected with sodium carbonate solution to give the thiazine- carboxylic acid (IV; R = Ph.CH,*CO, R' = H), m.p. 21 1-21 3 ".

R =PhCH,) and the thioamide (V) were dissolved in dioxan saturated with hydrogen chloride at 15". The thiazine (IV; R = PhCH,CO, R = PhCH,) isolated after 2 days, crystallised in two forms, m.p. 155-156" (needles) and 173-175" (cubes). The keto-ester (VII; R = Et), when treated with the thioamide (V) under similar conditions, gave the thiazine (IV; R = Ph-CH,-CO, R' = Et), m.p. 138".

lo Vogel and Schinz, Helv. Chim. Acta, 1950, 33, 116.

We thank Glaxo Research Ltd. for a supply of 7-phenylacetamidocephalosporanic acid, for Fellow- ships (to T.R.E. and V.V.K.), and for many discus- sions. We also express our thanks for a Common- wealth Scholarship (to S.H.E.) and to Professor Sir Ewart R. H. Jones, F.R.S., for his interest and advice.

(Received, May 27th, 1963 .)

The Structure of the Tetraphenylcyclobutadiene Dimer By H. H. FREEDMAN and R. S. GOHLKE

(THE Dow CHEMICAL COMPANY, EASTERN RESEARCH LABORATORY, FRAMINGHAM, MASS., U.S.A.)

UNTIL X-ray results are complete, structure (I) (octaphenylcubane) for the dimer, m.p. 430", of tetraphenylbutadienel cannot be considered as established. However, we now report evidence that excludes the alternative (11) proposed by Cookson and Jones.,

which is considered to arise by loss of a phenyl group from the molecular ion of m/e 712, strongly favours structure (I) since production of this ion from a compound (11) requires cleavage of two bonds with concomitant transfer of a proton to the leaving

mle 43 75 76 77 89

166 178 190 233 245 257 269 28 1 293

Mass spectrtrm of dimer Rel. int. mie

100 356 - 368 - 380 54 392 36 404 32 457 22 469 15 48 1 31 546 15 558 42 623 27 635 15 712 7 713

Rel. int. 46 20 14 10 1 1 9

15 11 10 9 5 9

94 50

group. The loss of phenyl from structure (11) is considered to be no more likely than that from a compound such as 9-methylfluorene (IIl), since both (11) and (111) have potentially migratory a-protons, and it has been found that phenyl loss from 9-methyl-

(n) The prominent ion fragments occurring in the

mass spectrum of the dimer, as obtained at 180" by a technique described elsewhere: are tabulated below. The observation of an ion fragment of m/e 635,

a Cookson and Jones, Proc. Chem. SOC., 1963, 115. Freedman and Petersen, J. Amer. Chem. Soc., 1962, 84,2837.

Gohlke, Chem. and Ind., in the press.

Dow

nloa

ded

by U

nive

rsity

of G

uelp

h on

18

June

201

2Pu

blish

ed o

n 01

Janu

ary

1963

on

http

://pu

bs.rs

c.org

| doi:

10.103

9/PS9

630000

229View Online

250 PROCEEDINGS

fluorene (III) does not occur to a significant extent (< 1 % rel. int.). Similarly, the presence of a high intensity (54%) phenyl-ion peak at mfe 77 is consistent with structure (I) but not with (11).

conclusion. Oxidative degradation of the dimer should produce benzoic acid if formula (I) is correct,

but a compound (11) should yield diphenic and, possibly, phthalic acid. Prolonged exposure of the dimer to chromic anhydride in acetic acid causes its dissolution and yields benzoic acid (30 %) as the only

A classical chemical technique leads to the Same base-soluble constituent. The residual oxidative fragments appear to be mainly ketonic.

(Received, June 14th, 1963)

PUBLICATIONS OF THE SOCIETY WHEN presenting the Accounts of the Society at the Annual General Meeting in Cardiff on March 28th last, the Honorary Treasurer referred (Proceedings, 1963, 157) to the continuous upward trend of the cost of producing the Journal and Proceedings, and the increasing number of pages to be printed at steadily increasing cost per page. Since 1959, the increase in size of the Journal and Proceedings has resulted in a 50% increase in cost of production although, during that period, the selling price remained constant. The present situation demands that the Society should take steps to protect its income and, whilst every effort is made to contain the cost of production, the Council has decided most regretfully that the selling prices must be increased.

For similar reasons the selling price of Current Chemical Papers must be raised but the prices of the Annual Reports on the Progress of Chemistry and Quarterly Reviews are to remain unchanged.

Fellows will still receive Proceedings free of charge and an increase in the Annual Subscription is not contemplated.

As from the issues for 1964, the prices of the publications to Fellows, payable with the Annual Subscription, and sold under declaration regarding use will be as follows (the price to non-Fellows is indicated in parenthesis) : Proceedings Free (E6 . 0.0) Journal &12. 0.Op.a. (&30. 0 . 0

sold only with Pro- ceedings)

Current Chemical Papers Ordinary Edition &3 . 0 . 0 p.a. (E7 . 10. 0) Edition printed on

one side of the Paper &4. 1 0 . 0 p.a. (&lo. 0.0)

Air Mail Edition E7 . 0 . 0 p.a. (El2 . 0.0) Progress of

Annual Reports on the

Chemistry 15s. p.a. (52. 0.0) Quarterly Reviews 15s. p.a. (&2. 0.0)

The prices of certain back-number issues have also been increased.

NEWS AND ANNOUNCEMENTS Election of New Fellows.49 Candidates were

elected to the Fellowship in July, 1963. Deaths.-We regret to announce the deaths of the

following: Mr. J. H. Young (March, 1963), Bourne- mouth, a Fellow since 1898; Dr. B. Lambert (1.7.63), Emeritus Fellow of Merton College, Oxford.

International Symposia.-The First Canadian Wood Chemistry Symposium sponsored jointly by the Chemical Institute of Canada and the Technical Section, Canadian Pulp & Paper Association, will be held in Toronto, Ontario on September 4--6th, 1963. Further enquiries should be addressed to the Chemical Institute of Canada, 48 Rideau Street, Ottawa 2, Ontario, Canada.

An International Symposium on Microbiology of Crude Oil, will be held in Greifswald, on October

1 st-6th, 1963. Further enquiries should be addressed to Professor Dr. W. Schwartz, Institut fur Mikro- biologie, Ludwig-Jahn-Str. 15, Greifswald, Germany.

The Eighteenth Plastics-Paper Conference, ar- ranged by the Technical Association of the Pulp and Paper Industry, will be held in Cleveland on October 14-16th, 1963. Further enquiries should be addressed to the Executive Secretary, 360 Lexington Avenue, New York 17, N.Y.

An International Conference on the Impact of Modern Physics on Materials, will be held in Philadelphia on February 3rd, 1964. Further enquiries should be addressed to the American Society for Testing and Materials, 1916 Race Street, Philadelphia, Pennsylvania.

A Symposium on the Application of Advanced

Dow

nloa

ded

by U

nive

rsity

of G

uelp

h on

18

June

201

2Pu

blish

ed o

n 01

Janu

ary

1963

on

http

://pu

bs.rs

c.org

| doi:

10.103

9/PS9

630000

229View Online

AUGUST 1963 25 1

and Nuclear Physics in the Testing of Materials, sponsored by the International Union of Testing and Research Laboratories for Materials and Structures, will be held in Philadelphia on February 3rd-6th, 1964. Further enquiries should be addressed to T. A. Marshall, Jr., Executive Secretary, American Society for Testing Materials, 191 6 Race Street, Philadelphia 3, Pennsylvania.

An International Congress on Documentation and Scientific Technical Information, will be held in Rome on February 2nd-llth, 1964. Further enquiries should be addressed to the Executive Secretary, Viale Regina Margherita, 83D, Rome, Italy.

A Conference on The Measurement of High Temperatures will be held in London on May ll-l3th, 1964. Further enquiries should be addressed to the Administration Assistant, The Institute of Physics and The Physical Society, 47 Belgrave Square, London, S. W. 1.

Personal.-Dr. C. A. Barson and Dr. M. H. B. Hayes have been appointed Senior Research Fellows at the University of Birmingham.

Dr. D. Betteridge has been appointed Lecturer in Chemistry at the University College of Swansea.

Mr. H. C. Butcher has taken a post as Technical Officer with the Federation of British Industries.

Dr. J. W. Clark-Lewis, Reader in Organic Chem- istry at Adelaide University, has been appointed to the Chair of Chemistry at the Bedford Park Division of that University.

Mr. R. B. Collins and Dr. G. F. DuBn have been appointed to the Board of Minnesota 3M Research Limited.

Dr. R. Colton has resigned his Research Fellow- ship at A.E.R.E., Harwell to accept a post as Lecturer in Inorganic Chemistry at the University of Melbourne.

Dr. L. Crombie has been appointed to the Chair of Organic Chemistry at the University College of South Wales and Monmouthshire.

Mr. J. Davis has resigned his post as Senior Research Chemist, Ever Ready Co. (G.B.) Ltd., to take up the position of Technical Director, Bright Star Industries of New Jersey, U S A.

Dr. J . E. Drake has been appointed Lecturer in Chemistry at the University of Southampton from October 1st.

The University of Oxford has awarded the Turner and Newall Official Tutorial Fellowship in Chemistry to Dr. B. E. F. Fender as from October 1st.

Dr. G. W. A. Fowles and Dr. A . C. Riddiford have been promoted to Readerships in the Department of Chemistry at the University of Southampton, from October 1st.

Dr. J. Fried has been appointed Professor in the

Ben May Laboratory for Cancer Research at the University of Chicago.

Dr. E. S. Hedges, Director of the International Tin Research Council, has been elected a Membre Correspondant de 1AcadCmie des Sciences dOutre- Mer. Membership in the Classe des Sciences Tech- niques is limited to ten persons outside the former Belgian territories.

Dr. H. G. Heller has been appointed Lecturer in the Department of Chemistry at the University Col- lege of Aberystwyth.

Dr. J. F. Herringshaw has resigned from his post as Lecturer in Inorganic Chemistry at the Imperial College of Science and Technology and has joined D. W. Kent-Jones and A. J. Amos, Analytical and Consulting Chemists.

Dr. H. S. Hirst, a Director of Imperial Chemical Industries Limited, Billingham Division, and General Manager of the Severnside Works, is to retire in September.

The University of Wales have granted the grade and title of Reader to Dr. J. W. Keyser, Senior Lecturer in Chemical Pathology, Welsh National School of Medicine; Dr. D. A . Long, Senior Lecturer in Chemistry, University College, Swansea; Dr. C. B. Monk, Senior Lecturer in Chemistry, University College, Aberystwyth.

Professor Sir Hans Krebs has been awarded an Honorary Degree of D.Sc. at the University of Leices ter .

Dr. R. Long has been appointed Senior Lecturer in Organic Chemistry at the Borough Polytechnic, London.

Dr. C. W. F. McCZare has been appointed to Lecturer in Biophysics at Kings College, London.

Dr. J. W. Mitchell, Professor of Physics in the University of Virginia, U.S.A., is returning to the United Kingdom on being appointed Director of D.S.I.R.s National Chemical Laboratory. He succeeds Dr. J. S. Anderson who is taking up the Chair of Inorganic Chemistry at Oxford University. His appointment takes effect from October 1 st.

Dr. W. D. Ollis, Reader in Organic Chemistry at the University of Bristol, has been appointed to succeed Professor R. D. Haworth to the Chair of Organic Chemistry at the University of Sheffield from October 1st.

Professor F. T. G, Prunty has been appointed to the Chair of Chemical Pathology at St. Thomass Hospital Medical School, London, from October 1st.

Dr. J. B. Raynor has been appointed Lecturer in Chemistry at the University of Leicester.

Dr. T. S. Stevens, Reader in Organic Chemistry, has been appointed to a personal Chair of Chemistry at the University of Sheffield.

Mr. W. I.. Thomas has retired as Chief Chemist

Dow

nloa

ded

by U

nive

rsity

of G

uelp

h on

18

June

201

2Pu

blish

ed o

n 01

Janu

ary

1963

on

http

://pu

bs.rs

c.org

| doi:

10.103

9/PS9

630000

229View Online

252

and Technical Director at Woolcombers Limited, Dr. D. A . Walker has been appointed Reader in Bradford. Botany in respect of his post at Queen Mary College,

Dr. J. C. J. Tynne and Mr. C. W.

proceedings of the chemical society. august 1963

Documents