PROCEEDINGS OF THE CHEMICAL SOCIETY APRIL 1963 TILDEN LECTURE* Nuclear Magnetic Resonance By R. E. RICHARDS AN atomic nucleus of spin quantum number I possesses magnetic moment p which arises from the circulation of electric charge associated with the spin of the particle. If a substance containing these nuclei is placed in a strong uniform magnetic field Ho the energy of the magnetic moment depends on its orientation with respect to the lines of force of the applied field. The quantum theory requires that this energy can have only discrete values so that only certain orientations of the nuclear magnet in the applied field are allowed and a nucleus of spin quantum number l is permitted to occupy one of (21 + 1) orientations. The energy separation of ad-jacent levels is pHo/I,and if suitably polarised radia- tion is applied to the sample transitions of the nuclei among these energy levels can be induced in accordance with the equation where h is Planck’s constant vo is the frequency of the radiation and y is the magnetogyric ratio.Note that for a given nucleus there is a simple linear relation between frequency and applied field strength HO. The probability that the radiation will induce an upward or a downward transition of a nucleus be- tween the energy levels is the same. Therefore a net absorption of energy from the radiation occurs only if there is an excess population of nuclei in the lower energy level. When energy is transferred between the nuclei and the radiation a nuclear magnetic resonance spectrum may be observed.When H is 104 gauss the frequencies of radiation required to excite the nuclei from one energy level to another lie in the region of 10 Mc./sec. so that nuclear magnetic resonance is a branch of radio- frequency spectroscopy. When a similar experiment is performed with substances containing unpaired electrons an electron-spin resonance spectrum may be observed. Because the magnetic moment of the electron is of the order of 103 times those of nuclei the frequency of radiation required for electron-spin resonance (see equation 1) is of the order of 10,OOO Mc./sec. in the microwave region. The chemical interest in nuclear resonance lies in the complex ways in which the spectra are often affected by interactions between a nucleus and its surroundings.These interactions can be considered conveniently under three headings namely (a)inter-actions between the extranuclear electrons and the applied magnetic field I€ ; (b) interactions between neighbouring magnetic moments which can occur in a variety of ways; (c) interactions between nuclear electric quadrupole moments and electric field gradients at the nucleus. The usefulness of nuclear magnetic resonance spectra is limited in practice by the line width and by the signal-to-noise ratio. If the lines are broad in- formation depending on fine structure may be lost. The strength of a nuclear resonance at a given field No depends only on the number of nuclei causing it and their physical properties and is measured by the area under the resonance.The broader the resonance line the smaller will be its height for a given area * Delivered before the Chemical Society on October 30th 1962 at Queen’s College Dundee; on October 31st at the University Aberdeen; on November 15th at King’s College London W.C.2; on November 29th at the University Liverpool. 101 PROCEEDINGS and thus the lower the signal-to-noise ratio. Increas- induced moment often dominates the chemical shifts ing the excess population of one energy level over the of protons in the molecule. For example in copro- other (if there are only two energy levels as for a porphyrin-I methyl ether a strong current circulation proton) will give a stronger signal.is induced in the porphyrin ring. The diamagnetic (a) Chemical Sh$fs.--When a chemical substance moment opposes the applied field H,,in the centre is placed in a strong magnetic field ITo orbital cur- of the ring causing the resonances of the NH protons rents are induced in the electron clouds which set up to occur at low frequencies in a fixed field Ho (or at a weak diamagnetic moment proportional to the high fields at a fixed frequency); the lines of force of applied field. This diamagnetic moment opposes the the induced moment however assist No on the out- applied field so that the field actually experienced by side of the ring causing the proton resonances of the a nucleus is reduced by a factor a. We therefore He,( groups to occur at frequencies about 14 p.p.m.write He = Ho(l -a) and equation (1) becomes higher than the NH group (or at Iow fields at fixed frequency). hvo = pHO(1 -a)/l . . . (2) The effect of electric fields of bond type, and of where He is the field actually experienced by the unpaired electron spin density5 in certain paramag- nucleus. netic molecules can also be interpreted qualitatively and sometimes quantitatively and such studies have The value of a depends on the strength of the led to a great range of useful chemical applications. diamagnetic moments which H can induce in the ex- The empirical applications of chemical shifts are tranuclear electrons. This is affected by the density very extensive and of great value in chemistry. The and spatial distribution of the electrons and therefore use of the so-called high-resolution nuclear magnetic varies with the chemical environment of the nucleus.resonance spectra in organic chemistry has now The factor Q is known as the chemical shift and reached an advanced stage of sophistication and can varies over a range of as little as twenty parts per be applied to very complex molecules. million (p.p.m.) for proton resonances and as much The chemical shifts (a)of protons in a great range as 2% for cobalt resonances. of organic compounds have been correlated with the The theory of this diamagnetic shielding of the chemical environment and functional group contain- nucleus by the electrons has been given by Ramsey,l ing the protons. A measurement of the chemical but unfortunately at present it can be applied shifts of the protons in an unknown substance may rigorously only to the simplest molecules.Approxi- then permit many of its structural features to be mate approaches to the theory of diamagnetic identified and from the relative intensities of the moments of complex molecules have been attempted chemically shifted lines the relative number of pro- by valence-bond and molecular-orbital methods and tons in the different groups can be obtained. Recent there is every hope that useful progress will be made examples have been given by Shoolery.6 The applica- along these lines soon.2 At present however the uses tions to inorganic chemistry have not been so exten-of chemical shifts rest mainly on qualitative theories sive but even so many useful results have been and on empirical correlations.obtained and are reviewed by Muetterties and The electron density on the hydrogen atom is so Phillips.?For example the tin resonance of a mixture low that the induced diamagnetic currents are very of stannic bromide and stannic chloride has five small so that for protons a is very sensitive to the components due to SnBr, SnBr,CI SnBr2C12 influences of stronger magnetic moments induced in SnBrCl, and SnCl,. The relative intensities are con- other parts of the molecule. A particularly striking sistent with a simple statistical equilibrium involving illustration of this effect occurs in molecules with a substances with equal thermodynamic stabilities.8 conjugated ring system. In molecules such as benzene The 170 resonance of aqueous solutions of paramag- the applied magnetic field can produce a very strong netic ions has two components which are associated diamagnetic moment at right angles to the plane of respectively with the water malecules of the solvent the ring by inducing current circulation in the T-and with those of the hydrated ions.A limit can be electron system. The local field associated with this placed on the rate of exchange of water between Ramsev. Phvs. Rev.. 1950. 78. 699. McGafvey J. Chem Phys 1957,27,68; Stephen Proc. Roy. Soc. 1957 A 243,264; Pople Discuss.Faraday Soc., in the press; Kern and Lipscomb J. Chem. Phys. 1962,37,260 267 275 279. Frank and Gutowsky Arch. Sci. 1958 11 215; Marshall and Pople Mol. Phys. 1958 1 199. Bothner-By and Naar-Colin Ann.New York Acad. Sci. 1958 70 833. Benson Eaton Josey and Phillips J. Amer. Chem. SOC.,1961 83 3714. Shoolery Discuss. Faraday Soc. 1962 in the press. ’Muetterties and Phillips “Advances in Inorganic Chemistry and Radiochernistry,” Academic Press New York 1962 Vol. IV p. 231. Burke and Lauterbur J. Amer. Chem. Soc. 1961 83 326. APRIL1963 hydrated ion and solvent and from the relative in- tensities approximate hydration numbers can be cal-~ulated.~ The chemical shift of the 170 resonance in the hydrated rare-earth ions can be shown to arise from very weak covalent bonds involving the 2s (or 2s and 29) orbitals of the oxygen and the 6s orbitals of the rare-earth ion.9 (b)Interaction between magnetic dipoles.-let us now consider some of the ways in which interactions between the nuclear magnetic moments of a sample may affect the nuclear resonance spectrum.(i) Static dipolar interaction. This occurs in crystals when the nuclei are rigidly fixed in space. The magnetic moment of each nucleus generates a weak magnetic field in its environment and at any point in the crystal all these “local” fields combine to pro- duce a net resultant HIocal directed along the lines of force of Ho. Hlmal can take on a range of values on either side of zero depending on the instantaneous orientations of the various nuclear moments causing it so that the nuclei at any site in the unit cell of the crystal will between them experience a range of fields of the order of Hlocalon either side of H, the applied magnetic field.The result is a broadening of the nuclear resonance by an amount of the order of H&al which is proportional to p/9,where r is the distance from the moment p. For hydrogen atoms which have large nuclear magnetic moments and small radii the broadening of the nuclear magnetic resonance lines in crystals is typically about 10gauss or IOOO p.p.m. in a field of lo4gauss; on the other hand for larger atoms such as thallium with small nuclear moments the broadening may be as small as 0.1 gauss or 10 p.p.m. The chemical shifts of proton resonances range over only 20 p.p.m. so it is clearly impossible to observe these effects in solids when the lines are far too broad. On the other hand for some inorganic substances the chemical shifts are greater and the broadening is smaller.Forexample the thal- lium resonance of crystalline Tl,Cl shows two re- solved components with intensities in the ratio of 3 1. These are chemically shifted resonances for thallous and thallichloride ions in accordance with the structure TIz mrWl,]. In simple crystals the broadening of the nuclear resonance which depends on CC(plpz/ri:) can ij sometimes be used to obtain internuclear distances,1° but this method has only limited applications. When molecular motion occurs in the crystal the local fields HI^^^^ may be more or less averaged out and a detailed study of this effect as a function of temperature can lead to information about the type of molecular motion occurring and the potential barriers which hinder it.lo In the liquid or gaseous state the molecular motion is so vigorous and random that no static component of the local field remains; the characteristic broaden- ing of the nuclear resonance lines of solids is com-pletely lost and line widths become dominated by the Heisenberg uncertainty principle and may be as small as 10“ gauss or 0.01 p.p.rn.at lo4 gauss. All that remains of the local magnetic fields are components fluctuating randomly in accordance with the thermal motion of the molecules in the liquid. The important effects of these fluctuations will be considered later. The very narrow lines which may occur in the nuclear magnetic resonance spectra of liquids reveal a further mode of interaction between nuclei which can pro- vide useful information for the chemist.(ii) Spin-spin coupling. The second type of inter- action between nuclear moments is transmitted only through chemical bonds. This scalar coupling arises from a weak spin polarisation of the electrons in a chemical bond by the magnetic moment of one nucleus which produces a small magnetic field at another nucleus joined to it through the chemical bond or bonds. The magnetic field produced at the nucleus depends on the orientation or orientations of the nuclei causing it so that the nucleus experiences a variety of situations which reflect some of the pro- perties of atoms to which it is joined. This scalar coupling is independent of molecular orientation in the magnetic field because it is transmitted through the chemical bonds and is therefore not affected by the molecular motion in the liquid.The result of the interaction is to produce the so-called “spin multi- plets” in nuclear magnetic resonance spectra. For example acetaldehyde shows two regions of proton resonances one at low fields due to the CHO proton and another at higher fields with three times the in- tensity due to the protons of the methyl group. A scalar interaction is transmitted between the nuclei through the chemical bonds and because carbon and oxygen have nuclei with zero magnetic moments only the protons are involved. These have a spin quantum number I = * so that the proton of the CHO group is allowed two orientations in H,.Each of these orientations transmits a slightly different effect through the chemical bonds to the protons of the methyl group the resonance of which therefore appears as a doublet; because the two orientations of the CHO proton are more or less equally probable the two components of the methyl doublet are of equal intensity. The separation of the two lines of the doublet is a measure of the strength of the scalar Swift and Connick J. Chem. Ph-vs. 1962 37,307. lo Richards “Determination of Organic Stntctures by Physical Methods,” ed. Phillips and Nachod Academic Press New York 1962 Vol. 11 p. 537. coupling and is equal to the spin-spin coupling con- stant J. The proton of the CHO group experiences a variety of magnetic fields determined by the various combinations of the two allowed orientations for each of the three CH protons.These combinations are 4$-ti+*; +4+*-+; +&-*-*;-4-4-2; +*-*+8 -*+*-+ -&+++&-*-i+8 The CHO resonance is therefore a quartet the two inner components being three times as strong as the outer ones because they correspond to arrangements of the CH protons which can be achieved in three times as many ways. In more complex molecules the spin-spin coupling leads to much more complicated fine structure. Some- times it is easy to interpret the spectrum in terms of the molecular structure but sometimes it is necessary to use rather complicated procedures1’ to evaluate the chemical shifts and coupling constants which characterise the spectrum.However in many cases the fine structure of high-resolution nuclear magnetic resonance spectra can be used to deduce information about the ways in which groups are joined together in a molecule; the nature of the functional groupings and the relative numbers of nuclei in them is ob- tained from the chemical shifts and relative intensities of the resonances. The values of the spin-spin coupling constants depend very sensitively on the stereochemical arrangement of the bonds joining the nuclei together. The theory of spin-spin coupling12 can be applied rigorously only in the simplest cases but valence- bond approximations have been used very success- fully by Karplus13 to interpret the variation with bond angle of spin-spin coupling between protons.A powerful method of studying spin-spin coupling depends on the simultaneous excitation of two or more nuclear resonances in the same sample. For example in a sample of 13CH,I the proton resonance is a d0ub1et.l~ The proton resonance is split by scalar coupling with the carbon-13 nucleus which has a spin 1 = Q; the iodine nucleus produces no effect because of very strong quadrupole relaxation. If whilst the proton resonance is constantly observed the 13C resonance is excited by simultaneous applica- tion of radiation of the appropriate frequency the 13Cnucleus will be flipped back and forth among its PROCEEDINGS two allowed orientations. If the irradiation is intense enough the 13Cnuclei are reoriented so rapidly that the protons are no longer able to distinguish their two different orientations and the proton resonance collapses to a singlet.As soon as the radiation ex- citing the 13C nuclei is removed the protons again experience the spin-spin coupling and the proton resonance regains the doublet structure. Double ir- radiation involving two groups of chemically shifted protons can be used as an aid to the analysis of com-plex high-resolution spectra and for the determina- tion of relative signs of spin-spin coupling con- stant~.~~ An elegant double-resonance method has been described by Andersonls which permits the determination of chemical shifts of 13C resonances in some compounds where direct observation is made difficult by very weak signals.The values of spin-spin coupling constants between protons in organic compounds are usually in the range 1-100 c./sec. i.e. of the order of 1 p.p.m. at 50 Mc./sec. Spin-spin interaction between protons can therefore be observed only in liquids where the resonances are narrow enough; the lines in solids are far too broad (-loo0 p.p.m.). Spin-spin coupling constants among atoms of high atomic number may however be very much greater than for protons. For example spin-spin coupling between nuclei of platinum and phosphorus in alkylphosphine com- plexes of platinum have been found1’ to be as large as 6 Kc./sec. In some solid compounds of thallium the spin-spin coupling between the two abundant isotopes and TlZo5 is so strong that it determines the line widths of the thallium resonances in these compounds.18 (iii) Relaxation times.We now turn to the third type of interaction between magnetic moments in the sample which may be called dynamic dipolar coup- ling. The local magnetic fields generated by the nuclear moments produce no “static” component in a liquid as explained above because of the random molecular motion. The fields still exist however as fluctuating components having a distribution of in- tensity F(v) with frequency corresponding to the molecular motion. The frequency distribution of the Brownian motion in a liquid has the form F(v) CC 2Tc/(1 + V2Tc2) . . . (3) where rc is the so-called correlation time for the motion and is a measure of the time taken for the molecule to change its orientation or position signi- l1 Pople Bernstein and Schneider “High Resolution Nuclea Magnetic Resonance,” McGraw Hill New York 1959.l2 Ramsey Phys. Rev. 1953 91 303. l3 Karplus J. Chem. Phys. 1959 30 11. l4 Royden Phys. Rev. 1954,96 543. l5 Maher and Evans Proc. Chem. SOC.,1961 208; Freeman A!o/. l6 Anderson J. Chem. Phys. 1962 37 1373. Pidcock and Richards Proc. Chem. SOC.,1962 184. l8 Freeman Gasser and Richards Mol. Phys. 1959 2,301. Phys. 1961.4 385. APRIL1963 ficantly. The form of the correlation function for various values of T~ is shown in Fig. 1. FIG.1 Intensity of field fluctuations as a function of frequency. The effect of these fluctuating magnetic fields is to modify the life-times of the nuclei in their various energy levels and hence by the uncertainty principle to change the nuclear resonance line-width.The Heisenberg uncertainty principle can be written approximately as AE-t x h/2v x h-Av-t . . . (4) or Au E 1/2nt where AE and du are the uncertainty or spread of the energy and line width and t is the lifetime of the particle in the energy level. The lifetime of a nucleus in an energy level is characterised in terms of two “relaxation times,” Tl and T2.TI is the spin-lattice relaxation time; it is a measure of the time taken for a nucleus to exchange energy with the thermal motion of the molecules around it; and T2is the spin-spin relaxation time and is a measure of the time taken for two nuclei to exchange a quantum of energy one falling from an upper to a lower energy level and the other using the quantum of energy to raise it from a lower to an upper energy level.Spin-lattice relaxa- tion is the process by which a nuclear-spin system heats up or cools down to the temperature of the “lattice,” which may be a gas liquid or solid. Spin-spin relaxation is an adiabatic process in which energy is passed through the spin system but not exchanged with other forms of energy. In liquids Tl and T2are usually approximately equal so only Tl will be considered in the discussion which follows. For nuclear spins the probability of a spontaneous transition from one energy level to another is nearly zero and for nuclei with I = * the only agency which can bring about a change from one energy level to another is a magnetic field fluctuating at the resonance frequency uo = pHo(l -o)/Ih.In a liquid such as water the only source of such a fluctuating magnetic field is the local fields of the proton magnetic moments fluctuating with the range of frequencies corresponding to the correlation function of the molecular motion. It is the component of these fluctuating fields at the frequency vo which provides a mechanism by which the nuclear spins can exchange energy with the thermal motion of the molecules and which permits spin-lattice relaxation. For most mobile liquids rc is of the order of 10-10-10-12 seconds. The value of F(u) (equation 3) therefore has a maximum at a frequency near lo1 c./sec.The significant frequency for relaxation is uo which is usually about lo7c./sec. for nuclei; at this frequency V070 < 1 so that F(u) oc rc,and it is therefore clear that F(uo) must be very weak. Spin- lattice relaxation is therefore very ineffective and the spin-lattice relaxation time T, is long (about 3 seconds in water). The line width of the nuclear resonance in the liquid is determined by the relaxa- tion time and is therefore only a fraction of a cycle and this is why high-resolution nuclear magnetic resonance is possible in diamagnetic liquids. In viscous liquids the molecules move more slowly T~ is larger F(uo) is greater relaxation is more effective Tlis shorter and the lines become broader. The intensity of the fluctuating component at vo depends not only on 7c,but also on the strength of magnetic moments in the liquid and on the mean approach distances to the nuclei concerned.These factors depend on molecular structure and on mole- cular interaction in the liquid. For example the relaxation time of the protons in the benzene ring of toluene in dilute solution in carbon disulphide is about 60 seconds whereas for the protons of the methyl group it is about 15 seconds. If the liquid contains a paramagnetic solute dramatic changes in the relaxation times may be brought about. The unpaired electron of the para- magnetic substance has a magnetic moment of the order of lo3 times greater than those of the nuclei so that although the correlation spectrum of field fluctuations may not be changed significantly the total intensity of the magnetic field fluctuations is much greater.Very small concentrations of para- magnetic solute can shorten nuclear relaxation times considerably and we can expect to observe differen- tial effects on the nuclei in different parts of a mole- cule which will reflect any preferred types of collision in the liquid. Double spin flips The fluctuating magnetic fields can also induce more complex transitions than the simple spin-lattice relaxation described above. One interesting effect is to induce two simultaneous transi- tions in which one magnetic moment falls from a higher to a lower energy level and another simul- taneously jumps from a lower to a higher energy level.If the first particle is an electron and the second a nucleus the process involves loss of energy by the spin system and can be regarded as a mode of relaxa- tion. If in such a solution the electron resonance is strongly excited so as to disturb the population of the electrons (S = 4) among their two energy levels a I 06 continuous flow of electrons from the upper to the lower energy level will be produced (Fig. 2). The I I jhY Jrrf I Nuclei I i i -1. Electrons FIG. 2. Double transitions causing upward jiimps of nuctei when electrons relax. relaxation process involves the relaxation of an elec- tron from the upper to the lower energy level with a simultaneous jump of a nucleus from a lower to an upper energy level; the excess energy is lost as thermal motion of the molecules.The result is to pump nuclei from the lower to the upper energy level and so to disturb the relative population of the levels; this disturbance is reflected directly in a change in the intensity of the nuclear resonance signal. As the rate of excitation of the electron resonance is slowly increased the nuclear resonance signal grows weaker and then vanishes when the populations of the two nuclear levels have become equalised. Further in- crease in the intensity of the irradiation of the elec- tron resonances causes the nuclear resonance to re- appear but now inverted as the population of the upper state becomes greater than that of the lower one.When the electron resonance is completely saturated the inverted nuclear resonance may be as much as ye/2y times stronger than normally where ye and yNare the magnetogyric ratios of the electron and the nucleus respectively; the ratio is about 330 for protons and greater for most other nuclei. This so-called dynamic nuclear polarisation has hardly been exploited at the present time. There are clearly great possibilities of its use for observing nuclear resonances which are otherwise too weak for example I3C resonances in natural abundance in organic compounds. Preliminary work shows19 that not all nuclear resonances in a given compound are equally affected during this type of experiment. This means that me cannot expect to retain the valuable "group-counting" uses of intensities in nuclear resonance spectra obtained by dynamic polarisation.On the other hand the variations may well lead to PROCEEDINGS valuable information about molecular interactions in solution.20 (c) Electric Quadrupole Interaction.-The final type of interaction to be considered here is of the greatest importance in determining line widths of nuclear resonances. Nuclei with spin quantum numbers greater than 8 usually have an aspherical distribution of electric charge which can be ex-pressed as an electric quadrupole moment. This moment can couple strongly with any electric field gradient set up at the nucleus by the electrons dis- tributed around it. If the molecule is tumbling over in a liquid this electric quadrupole coupling tends to reorient the nucleus into another energy level in the applied magnetic field.The effect is to shorten the lifetime of the nucleus in a given energy level and to reduce Tlaccording to an equation of the form where (eq Q)is the product of the electric quadrupole moment and electric field gradient 7 is the viscosity of the liquid and Tis the temperature. This quadru- pole coupling is often very strong so that the nuclear resonance may be characterised by very short spin-lattice relaxation times. The broadening of the nuclear resonance lines by this relaxation is a great limitation on the study of nuclei affected by it because the broadening is often so great as to make the signal undetectable. The effect can however often be turned to good account.For example when glycerol is added to dimethylformamide l/T for the nitrogen resonance is found to vary linearly with viscosity as expected from equation (4).21The slope of the straight line is a measure of the electric field gradient at the nitrogen nucleus. Protonation of the amide with a strong acid could occur at the nitrogen or at the oxygen of the amide; the former would make the electric field at the nitrogen more sym-metrical and reduce the quadrupole coupling and the latter would increase the n-bonding at the nitrogen and also the quadrupole coupling. In mixtures of di- methylformamide and trifluoroacetic acid the slope of the plot of 1/57' against viscosity is greater than in the absence of acid showing that protonation occurs predominantly at the oxygen atom of the amide.21 In this survey it has been possible only to discuss some of the properties of nuclear resonances which seem to me to be of interest to chemists.I hope nevertheless to have been able to show that although some aspects of nuclear magnetic resonance may have been widely exploited there remain many other effects which may reveal still further details about the microscopic and macroscopic properties of chemical substances. l9 Richards and White Proc. Chem. SOC.,1962 119. 2o Richards and White Discuss. Faraday SOC.,in press. 21 Herbison-Evans and Richards Trans. Faraday SOC.,1962 J8,845. APRIL1963 107 CHEMICAL SOCIETY MEETING The following papers were read and discussed at a Scientific Meeting which was held in the Society's rooms at Burling-ton House on Thursday March 14th 1963 at 6 p.m.Some o-phenanthroline Complexes of Yttrium and the the metal. Studies of the cations formed from Lanthanides. By F. A HART and F. P. LAMJNG. deuterium chloride show that carbon-3 is protonated THE complex chemistry of the lanthanides has so far and the following mechanism is suggested been limited with rather few exceptions to ligands containing anionic oxygen as a donor atom. There was an especial lack of information about complexes with ligands having nitrogen as the donor and because of this it has been often assumed that such compounds are relatively unstable. We have now prepared several types of complex c incorporating the uncharged nitrogen donor 0-phen- anthroline.The method of preparation is reaction in +. m Fe-$H anhydrous ethanol or ethyl methyl ketone between CH3 a metal salt and a suitable molecular proportion of the ligand when complexes having the following The reaction of a metal complex containing a stoicheiometries are obtained MC1,(H20)phen2 propargyl group M.CH,.CH=CH, with acids has MCl,phen, M(CNS),phen, M(CNS),phen, also been studied. Evidence is presented for the M(N03),phen2 and M(RCOO),phen. formation of an unstable cation containing what is In general they are crystalline and stable in dry formally an-allene group. The cyanoalkyl complexes air. They are decomposed by water but are stable C5H,Fe(C0),R where R =CH,.CN CH,*CH,.CN thermally some being unchanged at 300".Investiga-or CHMe.CN have been prepared and their re- tion of their structures is made difficult by their actions studied. The 1-cyanoalkyl complexes are relative insolubility in organic solvents and the de- quite soluble in water and readily protonate giving composition in water. However there is evidence cationic species. On the basis of analytical and particularly from infrared absorption studies that spectral evidence structures which contain a keten- they are mostly to be formulated as six-co-ordinate imine system bonded to the metal are proposed for complexes of an orthodox type. Thus the thio- the cations. H,COR, cyanates [Mphen,](SCN) show a CN stretching fre- The oxoal kyl complexes ,C5H ,Fe( CO)%-C quency at about 2050 cm.-l corresponding with ionic where R = H or Me have also been prepared and SCN groups.An additional peak at about 2080 cm.-l their reaction with acids has been studied. Again appears in the complexes [M(SCN),phen,]SCN as cations are formed and the evidence indicates that expected if there are both ionic and co-ordinated they contain a vinyl alcohol system bonded to iron. SCN groups present. The lightest lanthanides show exceptional behaviour in that if conditions of preparation are not strictly anhydrous the hydrated 0-and n-Banding from Cyclic Ligands to Transition- complex chlorides MCl,(H,O)phen are obtained. metal Ions in Organometallic Compounds. By M. R. M. GERLOCH, The water molecule is believed because of the CHURCHILL and R.MASON. thermal stability of the hydrate to be co-ordinated THE molecular structure of n-cyclopentadienyl-to the metal. tetrakis(trifluoromethy1)cyclopentadienonecobalt (I) and n-cyclopentadienyl-1-phenylcyclopentadiene-Stabilisation of Some Organic Systems by Complexing cobalt (11) have been determined by three-with Transition Metals. By J. K. P. ARIYARATNE,dimensional X-ray analyses. (T-and n-bonding to the M. L. H. GREEN, and P. L. I. NAGY. cobalt ion is observed for both the cyclopentadienone ORGANIC systems not isolable in the free state may and cyclopentadiene ligands as demonstrated by sometimes be stabilised by bonding them to transi- their conformation and the metal-carbon and tion metals. We have shown that complexes con- carbon-carbon bond lengths which have been deter- taining a metal-ally1 group MCH,CH= CH, mined with estimated standard deviations of 0.01 A react readily with hydrogen chloride or mineral acids and 0.025 A respectively; the stereochemistry of the and that the ally1 group is protonated.The cations cobalt ion is essentially octahedral. A comparison formed with various metals may be isolated as with wcyclopen tadien yl te t rame t hylcyclopen tadien- crystalline salts and chemical and spectral evidence onecobalt illustrates the dependence of the extent of shows them to contain a propene group bonded to 0-and n-electron donation from the ligand on the nature of the substituent while the bonding of the cyclopentadiene is entirely similar to that reported @ @ PROCEEDINGS for 2,4,6-triphenyltroponeiron tricarbonyl.A rationale for these structures is suggested in terms of the relative delocalisation energies of the ligands and stability of the octahedral cobalt(@ ion vis h vis its five-co-ordinate cobalt(1) analogue. The nuclear magnetic resonance spectra of substi- tuted cyclopentadiene complexes and related mole- cules suggest that a pseudo Diels-Alder description of these molecules may be more valid than has been previously assumed. COMMUNICATIONS Some Fluoroaromatic Derivatives of Tin By J. M. HOLMES and J. C. TATLOW R. D. PEACOCK (DEPARTMENT THE UNIVERSITY OF CHEMISTRY BIRMINGHAM) ALTHOUGH alkyl and aryl derivatives of tin are well known of the corresponding fluorocarbon series only the fluoroalkyls have been iso1ated.l We now report the preparation and characterisa- tion (Table) of a number of fluoroaromatic deriva- tives of tin including the fully substituted tetrakis- pentafluorophenyltin(rv).In addition the trichloro- compound has been prepared but has not been fully ammonium chloride-with the latter one of the products is the hexachlorostannate [Me,N],SnCI,. These compounds are more easily prepared than the fluoroalkyl derivatives mainly because the Crignard reagent2 is more accessible. However unless reaction conditions and proportions of start- ing materials are carefully controlled the chief product is the tetrakis-compound. Preliminary measurements of the recoilless emis- characterised. Compound M.p.Sn(C6F5)4 SnMe,(C6F& SnPh(C6F5)3 Sn(C6F5)&1 Sn(C6F5)&12 221 O liq. 100-102" 106" liq. TABLE Method Notes 1 Stable to water 2 3 Stable to water; b.p. 94-96"/1.7 Stable to water mm. 1 Stable to water 1 4 Reacts with water; b.p. 130"/2mm. Methods 1 Grignard (C,F,MgBr) 5 SnC14.2 Grignard + Me,SnCl,. 3 Grignard + PhSnC1,. 4 SnMe,(C,F,) + SnCl (Carius). Tetrakispentafluorophenyltin(1v) is stable to its melting point. It is unreactive probably because of steric effects and crystallises in needles with a pseudo-cubic unit cell as determined from a Guinier X-ray powder photograph of 17.45 A linear dimen- sion. The chloro-compounds are more reactive; they are hydrolysed even by weak bases and undergo a number of cleavages with the loss of the pentafluoro- phenyl group e.g.with oxine and with tetramethyl- sion and absorption of gamma rays (Mossbauer effect) in tetrakispentafluorophenyltin suggest that the electronegativity of the pentafluorophenyl group is approximately the same as bromine? we are grateful to the Ministry of Aviation for their support of this work. (Received January 3 lst 1963.) Kaesz Phillips and Stone Chem. andlizd. 1959 1409; Stone and Treichel ibid. 1960 837; Clark and Willis J. Amer. Chem. Suc. 1960 82 1888; Suddens M.Sc. Thesis Birmingham 1960; Kaesz Stafford and Stone J. Amer. Chem. SOC.,1960,82 6228. Nield Stephens and Tatlow. J. 1959 166. M. Cordey-Hayes private communication. APRIL1963 109 A Molecular Prw~ in the Radiation-induced Oxidation of Hydrocarbons By G.DOBSONand G. HUGHES (THEUNIVERSITY, LXWRPOOL) ALCOHOLS and carbonyl compounds are known to be amongst the products of the radiation induced oxida- tion of The following sequence of reactions has been suggested to account for their formation in oxygenated cyclohexane CyClO-C6H12 CyClO-C6H11*f H. . . (1) CyClO-c6H11* f 02 -f CyClO-C6H1102* . . (2) 2 CyCIO-C6H11Og*-+ CYClO-C6H11*OH CyClO-C6H1,0 f 02 . (3) We have examined the effect of added scavenger on the y-radiolysis of oxygenated cyclohexane. The concentration of oxygen was 10-2~.In the presence of a suitable scavenger it might be expected that cyclo-C6Hl102* radicals would be scavenged CyClO-C6H1102. SH -t CyClO-C6H11O2H S*. .(4) Addition of iodine would probably lead to scaveng- ing of cyclo-C6Hllo rather than cyclo-C6Hl1020 radicals3 Either process should lead to a reduction in the yields of ketone and alcohol. However even at the highest concentrations of scavenger we find that the yield of ketone is never decreased to zero but reaches a limiting value of G(cyc1ohexanone) = 0.46 independent of the scavenger used as shown in the Figure. Yields are in all cases independent of dose. A W 6 2.0 I I I 1.0 2.0 Concn. of scavenger Efect of scavengers on the yield of cyclohexanone in the y-radiolysis of oxygenated cyclohexane. Scavenger concentrations (M) are multiplied by a nor- malisation factor to bring the curves into coincidence x pure- cyclohexane .Chromatographic experiments confirm that the ketone product is solely cyclohexanone and that no ketone is formed from the scavenger. An additional process involving non-radical inter- mediates would then seem to be responsible for this limiting ketone yield. It is possible that an excited- molecule reaction might be occurring though it is not possible to distinguish between this reaction and the formation of an intermediate excited cyc~o-C6Hll-00H which then decomposes to ketone. Reaction (5) could occur in some photo-oxidation processes though it is likely to require considerable excitation energy. The limiting yield of cyclohexane is decreased to 0.38 in =rated solution. Thus there may be competi-tion between reaction (5) and some other process in which cyc1o-C6Hl2* is destroyed.Our results indicate that there is no apparent effect of iodine concentra- tion on reaction (5). However in dezrated cyclo- hexane solution iodine quenches both scavengeable and non-scavengeable processes with comparable efficiency.* Quenching of cyclo-C6H12* by iodine must be relatively much less efficient than reaction with oxygen. Indirect confirmation of reaction (5) is obtained from measurements of the alcohol yields. Dispropor- tionation of cyclo-C6Hl102* radicals would lead to production of equal amounts of cyclohexanone and cyclohexanol. Our results indicate that the yield of cyclohexanol is always less than that of cyclo-hexanone by an amount equal to the limiting value of G(cyc1ohexanone).These results are contrary to the findings of others2 though their results were obtained at much greater total doses. We have found evidence for the existence of a similar molecular oxidation process in n-hexane and experiments are in progress to investigate the possibility of such processes in other organic systems. One of us (G.D.) thanks the Department of Scientific and Industrial Research for a maintenance allowance. (Received January 31st 1963.) Bach and Popov Symposium on Radiation Chemistry Acad. Sci. U.S.S.R. 1955 156. McCarthy and MacLachlan Trans.Furaduy Soc. 1961,57,1107; Dewhurst J. Phys. Chem. 1959,63,813. a Fessenden and Schuier J. Amer. Chem. SOC.,1957,79 273. Dyne and Jenkinson Canud.J. Chem. 1961,39 2163. I10 PROCEEDINGS Precise Measurements with the GlassEleclrode the Time Variation of E.M.F.By W. H. BECK J. CAUDLE and W. F. K. WYNNE-JONES A. K. COVINGTON DEPARTMENT COLLEGE OF DURHAM, (CHEMISTRY KING’S UNIVERSITY NEWCASTLE-UPON-TYNE, 1) IT is now no longer disputed that the potential of a glass electrode newly placed in a solution may vary with time.l There is some confusion whether this is due to a slow response to proton activity when a steady value should be taken as the correct e.m.f. or whether it is due to other reactions at the electrode surface when an initial value should be taken. Beck and Wynne-Jones2 concluded that the glass electrode responds instantaneously to proton activity and ascribed any time variation to asymmetry potential changes due to other potential-influencing reactions occurring at the glass-solution interface.Using this concept Covington and be3 extrapolated the linear e.m.f.-time graphs obtained when low-resistance glass electrodes were transferred rapidly with washing between dilute hydrochloric acid solu-tions to the time of transfer the contribution of the asymmetry potential to each cell then being the same and on subtraction cancelling. Clever and Reeves4 recently used high-resistance electrodes and this technique but changed the concentration of hydrochloric acid within one cell instead of using two solutions and separate reference electrodes. The e.m.f. wasfollowed with time until the change became linear and this linear portion was extrapolated to the time the acid concentration was changed.Caution must be exercised in applying extrapola- tion techniques and a short account of our investiga- tion of the problem of e.m.f. variation is presented for the guidance of users of glass electrodes. We have tested various commercial electrodes conditioned in distilled water by transferring them between acid solutions of various concentrations each containing hydrogen reference electrodes. If a glass electrode behaves as a hydrogen electrode the e.m.f. of all cells should be identical. If a time varia- tion occurs this can be described as a time-variable error being the difference between the e.m.f. at a certain time in the second solution and the final e.m.f. in the first solution.Careful attention to shield- ing and to the prevention of electrical leakage is necessary to avoid spurious effects. Recorder traces of the output of a vibrating condenser electrometer were obtained on which the e.m.f. was measurable to 0.1 mv. The e.m.f.-time variation may contain the follow- ing features singly or combined (A) a rapid change over the first few minutes which is largely independ- ent of solution composition and concentration and is probably electrical rather than electrochemical in origin; (B) a zero or slight and approximately linear change; and (C) a rapid change the rate of which slowly decreases but a steady value is rarely being reached even after many hours and occasionally turning values are observed.There is evidence that other ions and probably water participate in reac- tions at the glass surface during (C). Broadly the characteristics of the electrodes tested fall into two classes depending upon which of the above are observed Class I. Electrodes recommended for use in the pH range 1-9 of resistance < 200 and believed to be made from MacInnes-Dole (soda) glass (Corning015). Feature (A) is absent. At low and intermediate acid concentrations (B) is observed and extra-polation to the time of transfer shows no error (Ifl 0.1 mv). Above a molality dependent upon the acid anion [HCI(lm) HBr(Sm) sulphuric(7m) phos- phoric acetic and perchloric acid (lOm)] charac- teristic behaviour (C) is found. The error becomes more negative with time and increasing acid con- centration.Reasonable extrapolation of the non- linear trace leads to smaller but apparently non-zero errors. Class ZI. Electrodes usually described as “general purpose” (1-13 pH) believed to be made from lithia glass and of resistance > 200 MG. The characteristic e.m.f.-time curves consist of both features (A) and (B). Only at very high concen- trations are (A) and (C) observed. The linear portion (B) when extrapolated may give an error of f0-3 mv but the error at any time in (23) also does not exceed & 0-3 mv. The electrodes used by Clever and Reeves4 apparently showed (A-B) behaviour unless the initial change had its origin in their titration tech- nique. In either case they were correct in ignoring the initial change.Pretreatment5 of the glass electrode in a medium similar to that it is to be used in reduces time varia- tions i.e. asymmetry potential changes and may alter the characteristic pattern of the time variation from (C) to (B) making precise measurements I Mattock “pH Measurement and Titraticn,” Heywood London 1961 pp. 104-106. a Beck and Wynne-Jones J. Chiin.phys. 1952 49 C97. Covington and Prue J. 1955 3696; Covington J. 1960 4441. Clever and Reeves J. Phys. Chem. 1962 66,2268; also Baes and Meyer Inorg. Chem. 1962 1 780. Bacareila Grunwald Marshall and Purlee J. Phys. Chem. 1958 62 856. APRIL 1963 possible. The success of work3 with hydrochloric acid depended upon pretreatment of the electrodes in 0.lm-hydrochloric acid.If low-resistance elec- trodes are pretreated in distilled water asymmetry- potential changes are rapid and extrapolation is difficult. Accurate measurements are most likely to be possible if adequately tested glass electrodes on transfer between solutions are subjected to only Small changes in pH ionic environment and water activity. (Received. February Jth 1963.) Reaction of Olefin-Palladium(II) Chloride Complexes with Nucleophiles Mechanistic Considerations By E. W. STERN (THEM. W. KELLOGG CITY,N.J. U.S.A.) Co. JERSEY OLEFIN-PALLADIUM(II) chloride complexes react with a variety of nucleophiles in non-aqueous media to yield vinyl compounds.l The following is a discussion of recent observations which bear on the mechanism of this reaction.The reactions of propene and [2-2H]propene with acetic acid in “iso-octane” at initial pressures of 2 atmospheres and room temperature in the presence of palladium(r1) chloride and disodium hydrogen phosphate gave results summarised in the Table. The apparent reaction velocity of propene is greater than that of [2-2H]propene by a factor of 2.8 indi- cating rupture of the C-H bond at the central carbon atom of propene in the rate-determining step. The free from deuterium indicating that vinyl alcohol was not a reaction intermediate.3 The lack of ex- change cited in this case is consistent with a direct 1,2 shift of h~dride.~ However transfer involving an intermediate hydridopalladium complex cannot be rigorously excluded at this time.The absence of diacetates from the products of reactions of olefins with an excess of acetic acid at room temperature1 suggests strongly that formation of free carbonium-ion species after nucleophilic attack3 is unlikely. Formation of 1,l -diacetates when reactions were carried out at higher temperatures or products of room temperature reactions were sub- jected to heat as in distillation can be explained by Propene [2-2H]propene Total acetate yield (mole % calc. on PdC12) 15 10 Acetate (%) as isopropenyl acetate 64.0 63.6 Acetate (%) as propenyl acetate 36.0 36.4 Apparent reaction velocity (acetate yieldihr.) 0.31 0.11 Acetate (%) containing deuterium observed retention of deuterium by propenyl acetates formed from [2-2H]propene rulesout proton loss from the carbon undergoing nucleophilic attack2 as a mechanistic possibility.The extent of retention supports a mechanism involving a 1,2 shift of hydride from the attacked carbon followed by pro- ton loss from an adjacent carbon atom. In the absence of an appreciable isotope effect in the latter step a minimum deuterium retention of 71 % would be expected. Under similar conditions cis-but-2-ene reacted with acetic acid to yield on hydrogenation s-butyl acetate as the sole product. It is concluded from this that methyl migration does not occur and cannot be used to rationalise the formation of propenyl acetate from propene. Evidence for hydride migration has also been pre- sented in the reaction of the complex (C,H,PdCI,) with deuterium oxide which yielded acetaldehyde -75 addition of acetic acid to the vinyl acetate formed initially.The observation that the inclusion of water in reaction media leads to the production of carbonyl compounds indicates a greater intrinsic rate for the water reaction. In view of the relatively low nucleo- philicity of water toward either carbon or pal-ladium(i1) complexes in replacement reactions this difference in rates may reflect the relative ease of solvation of the olefin-palladium(l1) chloride com- plex,6 the ease of proton loss from an 0-H relative to a C-H bond in the final step a steric phenomenon or an effect of the nucleophilic ligand on electron distribution in the olefin complex.The author is grateful to Mr. A. S. Logiudice for assistance with experimental work and to Dr. H. P. Leftin for discussions. (Received Janrruty 19th I963.) Stem and Spector. Proc. Chent. Sor.. 1961. 370. Moiseev Vargaftik and Syrkin Doklady Akad. Nauk S.S.S.R. 1960. 133 377 ’ Hafner Jira Sedlmeier and Smidt Chem. Ber. 1962 95 1575. Cannel and Taft J. Amer. Chent. Sue. 1956 78. 5812. ’ Edwards and Pearson J. Aimr. Chem. Sue. 1962 84 16. Harris Livingstone and Reece J. 1959 1505. PROCEEDINGS The Stereochemistry and Conformationof v-Cyclopentadienyi-1-phenylcyclopentadienecobalt By M. R.CHURCHILL and R.MASON CRYSTALLOGRAPHY IMPERIAL OF SCIENCE (INORGANIC LABORATORY COLLEGE AND TECHNOLOGY LONDON S.W.7) INFRARED and nuclear magnetic resonance studies of a number of transition-metal-cyclopentadienecom-pounds have indicated the non-equivalence of the methylene protons? In the case of mcyclopentadi- enylcyclopentadienecobalt an intense band at 2781 cm.-l in the infrared spectrum is assigned to the 1 -Ha stretch Ha being the methylene-hydrogen atom Ccl)-H in the endo-position and assumed to lie in close proximity to and interact with the cobalt atom; the high-resolution nuclear magnetic resonance spec- trum of this compound shows the methylene-protons as two doublets the higher of which (T 8.1) is assigned to the H resonance.The absence of a carbon-hydrogen stretching vibration in the 2780 cm.-l region2 and of the Ha resonance3 in T-cyclopenta- dienyl- 1 -phenylcyclopentadienecobalthas led to the suggestion2 that the phenyl group occupies the endo-position the cyclopentadiene ligand being assumed to be planar and bonded to the cobalt ion by two v-bonds.We have carried out a complete single-crystal X-ray analysis of this compound. It has space group Pbca (I);:; no. 61) with eight molecules in a unit cell of dimensions a = 29.64 b = 7.70 c = 10.68A. The crystal structure has been determined by Patterson Fourier and least-squares analyses of 750 three- dimensional reflexions the refinement analysis con- verging to the present discrepancy index ii = 0.106 after five cycles of refinement. The Figure shows the essential geometry of the molecule. The phenyl group occupies the exo-position and the “cyclopentadiene” ring is non-planar with C-1 bent away from the cobalt atom (by 36”) resulting in a cobalt-carbon distance of 2-55 A compared with an average cobalt-carbon distance of 2.01 A (estimatd standard deviation 0.03 .$) for the remain- ing four carbon atoms of this system.The plane of these four atoms and that of the 7r-cyclopentadienyl ring are essentially parallel (f4”). The average cobalt-carbon distance in the rr-cyclopentadienyl ring is 2-06 Ifi 0-03A indicating that the bonding of the cyclopentadiene to the metal is slightly stronger than that for the n-cyclopentadienyl a feature indi- Green Pratt and Wilkinson J. 1959,3753. Fischer and Herberich Chem. Ber. 1961 94 1517. Jones and Wilkinson personal communication. 4 Dahl and Smith J.Arner. Chem. Soc. 1961,83,752. Gerloch and Mason,in preparation. * Boston Sham and Wilkinson J. 1962 3488. cated also by studies on n-cyclopentadienyl(tetra-methylcyclopentadienone)cobalt4 and its tetrakis- trifluoromethyl analogue.s ;> I The short bond length of 1-38 A for the 3,4-bond in the cyclopentadiene ring and the deviation of C-1 from the plane of the ring indicate that the cyclo- pentadiene is bonded to the cobalt by one v-and two o-bonds an unusual mode of bonding already postu- lated by Wilkinson et aZ.1*6for tricarbonyl(tetrakis- trifluoromethylcyc1opentadienone)iron(the structure of which will be published elsewhere’). In simple valence-bond terms three bonds to points in the n-cyclopentadienyl ring complete the expected octa- hedral co-ordination of the d6cobalt ion which has a formal oxidation state of +3 as has its precursor di-rr-cyclopentadienylcobaltiodide.The cobalt-methylene-hydrogen distance (on an assumed C(,)-H bond length of 1.1 A) is ca. 3.0 A so that although the hydrogen is in a formal endo- conformation it is not close enough to the cobalt for interaction to occur (this would result in the charac- teristic C(,]-H stretch at 2780 cm.-l). It is possible that in those cyclopentadiene derivatives for which this stretch is observed C-1 will be displaced in the opposite direction thereby allowing Co-H inter-7 Bailey and Mason in preparation. Fischer and Pfab 2.Naturforsch. 1952,7b 752; Fischer and Fritz “Advances in Inorganic Chemistry and Radio- chemistry,” Academic Press London and New York 1959 Vol.I p. 94. APRIL1963 action. Such a scheme has already been proposed for a number of n-cyclohexadienyl complexes of transition-metal ions.s We are grateful to Professor G. Wilkinson for @ Jones Pratt and Wilkinson J. 1962,4458. discussions Drs. 0. S. Mills J. S. Rollett and R. Sparks for computer programmes and the Depart- ment of Scientific and Industrial Research for a Research Studentship (to M.R.C.). (Received January 29th 1963.) Covalent Bonding in Fluoro-salts of the Trimethyltin Group By H. C. CLARKand R. J. O’BRIEN (DEPARTMENT UNIVERSITY VANCOUVER, OF CHEMISTRY OF BRITISH COLUMBIA CANADA) EVXDENCE has accumulated rapidly that trimethyltin derivatives Me,SnX have structures in which the Me3Sn group interacts strongly with X conferring five co-ordination on the tin atom.This has been observed in Me,SnF,l Me3SnC104,2 and M~,SIBF,.~ Moreover Me,SnClO and Me,SnN03 react with ammonia to form the five co-ordinate cation Me3Sn(NH3),+.4 In extending our own studies of these compounds we have now examined two compounds in which the Me,Sn group may interact with an octahedral ion. These are trimethyltin hexafluoro-arsenate and -antimonate prepared as white crystalline solids from trimethyltin bromide and the corresponding silver salt in liquid sulphur dioxide. The weights of silver bromide recovered and the analyses were consistent with the formula=. If these compounds are polymers similar to the perchlorate then planar Me3Sn groups are interacting with the MF6 octahedra through either cis- or trans-fluorine atoms.In the former case the symmetry of MF becomes CZuand inlatter D4&. The infrared spectrum of Me,SnAsF shows the following main features. (a) the v3 vibration ob- served5 for the free A@,- ion at 700 cm.-l is split into two bonds at 710 and 675 cm.-l the former being about twice as intense as the latter; this is con- sistent with DPA symmetry. (b) Bands associated with the Me,Sn group are located at 2900-3000 (C-H stretch) and 795 cm.-l (CH rocking) while there are no bands in the 500-590 region in which Sn-C vibrations are usually found. Instead a band of moderate intensity is found at 603 cm.-l and is pre- sumably a Sn-C vibration; this shift indicates con- siderable interaction with the ASF group.(c) No bands of any intensity were present in the 400-590 cm.-l region. We therefore believe that the AsF group interacts through trans-fluorine atoms with Me,Sn to form a chain structure in which the tin atom is five co-ordinate. For Me,SnSbF, the infrared spectrum shows the following features (a) the v vibration observed5 for the free SbF6- ion at 660 cm.-l is resolved into three components at 675,656 and 640 cm.-l. This number of components is consistent with Czzlsymmetry. (b) A new band is present at 875 cm.-l and a doublet at 472 454 cm.-l appears. The appearance of such bands particularly those at low frequency is also consistent with lower symmetry for the SbF group.(c) Theband observed at 603cm.-l in the AsF com-pound is found at 606 cm,-l in the hexafluoro- antimonate and is associated with the Me,Sn group. Again there must be strong interaction between Me,Sn and SbF6. The spectra provide strong evidence that as a result of this interaction the hexa- fluoroantimonate group has a cis-configuration while the hexafluoroarsenate group has a trans-configura- tion. This is consistent with the structure of SbF in which cis fluorine atoms act as bridges between SbF octahedra. We are now attempting to confirm these configurations by nuclear magnetic resonance methods. It has therefore not yet been possible to obtain an Me3SnX compound in which there existed a free Me,Sn* ion; in all cases strong interaction occurs and the tin atom is essentially five co-ordinate.We are grateful for the assistance of Dr. C. Reid and for the financial support of the National Research Council and the award of a Canadian Industries Ltd. Fellowship to R.J.O’B. (Received February 25th 1963.) Clark O’Brien and Trotter Proc. Chem. Soc. 1963 85. Okawara Hathaway and Webster Proc. Chem. Sac. 1963 13. Hathaway and Webster Proc. Chem. SOC.,1963 14. * Clark and O’Brien Jnorg. Chem. 1963 in press. Peacock and Sharp J. 1959,2762. Hoffman Holder and Jolly J. Phys. Cheni. 1958,62 364. PROCEEDINGS On the Photochemical Dearbonylation of a Homoallylic Conjugated Aldehyde1 By J. IRIARTE,J. HILL,I(.SCHAFFNER, and 0. JEGER LABORATORIUM TECHNISCHEN (ORGAMSCH-CHEMISCHES DER EIDGEN~SSISCHEN HOCHSCHULE ZURICH) WEreport a light-induced unimolecular decarbonyla- tion of a steroidal &unsaturated aldehyde which shows properties of homoallylic conjugation.3,17- Bisethylenedioxyandrost-5-en-19-al(I1; R = H) m.p. 169-171" (decomp.) [a] -249" (C A,,,. 226 and 310 mp (E 1255 and 113 in EtOH)? was prepared in excellent yield by oxidation of the alcohol (I; R == H)* with chromium trioxide in pyridine. Irradiation in ethanol at room temperature using a high-pressure mercury lamp with Pyrex filter caused evolution of carbon monoxide and conver- sion of the aldehyde (11; R H) into compound -1 (111; R = H) m.p. 131-132" [a] -13" (c 1.41) in more than 90% yield.The reaction was not sensi- tive to the presence of oxygen but when a quartz vessel and a low-pressure mercury lamp (ca. 90% emission at 254 mp) was used the rate of decarbonyl- ation was significantly slower. The structure of the photoproduct (III; R = H) was established as follows :Acid-catalysed hydrolysis of (III; R = H) quantitatively gave oestr-4-ene-3,17- dione (IV).&The lop-configuration of the photo- product was confirmed by hydrogenation of (111; R = H) with palladium on charcoal in ethanol to a mixture of saturated 3,17-bisethyIenedioxyoestranes and acid hydrolysis of the major component to 5/h~tran-3,17-dione (V).s A blank experiment in the absence of hydrogen demonstrated that (HI; R I= H) remained unchanged under the conditions of hydrogenat ion.In order to trace the origin of the hydrogen atom at position 10 of compound (111) the deuterated aldehyde (11; R = D) m.p. 180-182" [a] -241" (c 0.53) Amax. 226 and 310 mp (E 1380 and 113 in EtOH)3 was irradiated. This compound was syn- thesised from methyl 3,t 7-dioxoandrost-4-en- 19- oate8 by ketalisation to methyl 3,17-bisethylenedi- oxyandrost-5-en-19-oate m,p. 185-186" [a] -113" (c 0-39) reduction with lithium aluminium deuteride to the alcohol (I; R == D) m.p. 196" [r] -57" (c =0-38) and oxidation with chromium trioxide in pyridine. On irradiation of the aldehyde (11; R = D) under conditions identical to those described in the first experiment 3,l 7-bisethylene- dioxy[lO~-D]<lestrJ-ene (111; R D) m.p.133". =_5 [a],-11" (c 0-51) was formed in equally excellent yield. Mass-spectrometric comparison of the starting aldehyde (11; R = D) and the photoproduct (111 R = D) revealed no loss of deuterium during the photochemical step (deuterium content 98-99 -1 1 % in both compounds). Preliminary experiments with the saturated aldehyde (VI) m.p. 97-103" which was obtained by oxidation with chromium trioxide in pyridine of the amorphous 3,17-bisethylenedioxy-5~-androstan-19-01,~[a] + 5" (c 049) indicate that comparable irradiation conditions lead to a complex mixture of products. Decarbonylation appears to take place only to a small extent. Thus the important role of the double bond in the conversion of the aldehydes (11) into the product (111) is demonstrated.* Photochemical Reactions Part 21 ;Part 20,Helv. Chini. Acta 1963 46 67s. M.p. are uncorrected a11 [a]~ refer to CHCI solutions. An analogous chromophore is e.g.. present in dianhydrodihydrostrophanthidin;for a discussion of its ultraviolet absorption cf. Cookson and Wariyar J. 1956 2302. * Amorosa Caglioti Cainelli lmmer Keller Wehrli MihailoviC Schaffner Arigoni. and Jeger Heh. Chiru. Acfn, 1962,45,2674. Wilds and Nelson J. Anzer. Chenz. SOC.,1953 75 5366. Rapala and Farkas J. Amer. Chenz. SOC.,1958 80 1008. 'The structure proof for the photoproduct (111; R = H) provides indirect evidence in favour of the Ion-configuration of the (different) 3,17-bisethylenedioxy~str-5-ene. m.p. 135-1 37" which had been obtained earlier via another routc and to which formula (111; R = H) had been provisionally assigned (cf.ref. 4 p. 2683 compound no. 52). Hagiwara Noguchi and Nishikawa Ghenr. Pharin. Bull. (Jizpon) 1960 8,84. The preparation ofthis compound will be reported elsewhere. APRIL 1963 This photochemical decarbonylation of a &un-saturated aldehyde represents a novel and efficient method for stereospecific labelling of an allylic posi- tion. Its mechanism and scope are currently under investigation. From all new compounds satisfactory analyses and spectroscopic data (infrared and nuclear mag- netic resonance spectra) have been obtained. We are indebted to Dr. J. Seibl for providing the mass- spectrometric data. Generous financial support from CIBA Aktien- gesellschaft Basel is gratefully acknowledged.Two of US thank SYntex S.A. Mexico (J.1.1 and DsIR/ NATO (J.H-1 for fellowships* (Received February 25th 1963.) Generationof Tetraphenylcyclobutadiene and its Adducts from its Palladium Chloride Complex By R. C. COOKSON and D. W. JONES (THEUNIVERSITY, SOUTHAMPTON) WE have used Mala tes ta's te trap hen ylc yclo bu t adi- ene-palladium chloride complex conveniently made from diphenylacetylene,l as a source of tetraphenyl- cyclobutadiene. Because of the publication of pre- liminary work on the same complex2 we report some of our own results now. Like Maitlis and Stone2 we had found that treat- ment of the complex with triphenylphosphine in benzene gives in good yield bistriphenylphosphine- palladium dichloride m.p.297-298" (decornp.) and the tetraphenylcyclobutadienedime$-5 of m.p. 430". When the reaction was carried out in the presence of methyl phenylacetylenecarboxylate the inter-mediate cyclobutadiene was trapped6 as methyl pentaphenylbenzoate m.p. 355-360" identical with a sample prepared from methyl phenylacetylene- carboxylate and tetraphenylcy~lopentadienone.~ When the reaction was conducted in cyclopenta- diene an adduct was formed m.p. 184-1 87" (Found C 93.6; H 6.2%; M 400). The proton magnetic resonance spectrum measured at 80 Mcfsec. in deuterochloroform showed clearly that a bicyclo- [2,2,1 Jheptene unit was absent; while the spectrum was consistent with the presence of a cyclopentene unit having a broad doublet centred on 7-15 T from the methylene group (C-3) a multiplet centred on 6.5 T from the 4-proton a broad doublet at 5.76 7 J = 10.4 cfsec.from the 5-proton and two multi- plets centred on 4.4 and 4.09 7 from the olefinic protons. The gross structure was confirmed by syn-thesis of the same hydrocarbon by loss of carbon monoxide at 140" from the adduct of tetraphenyl- cyclopentadienone and cyclopentadiene. Neither the magnetic resonance nor the ultraviolet spectrum" * The spectrum of 1,2,3,4-tetraphenylcyclobutene,A,,,. sufficiently different to favour structure (11). * Malatesta Santarella Vallarino and Zingales Angew. Chem. 1960,72 34. Blornquist and Maitlis J. Amer. Chem. SOC.,1962,84,2329;Maitlis and Stone Froc. Ckm. Soc. 1962,330.Braye Hiibel and Capiier J. Amer. Chem. SOC.,1961,83 4406. Freedman J. Amer. Chem. SOC.,1961,83 2195. Freedman and Petersen J. Amer. Chem. SOC.,1962,84,2837. Cf.Berkoff Cookson Hudec and Williams Froc. Chem. SOC.,1961 312 and ref. 4. ' Dilthey Thewalt and Trosken Ber. 1934,67B 1959. Freedman and Frantz J. Amer. Chem. SOC.,1962,84,4165. (Amax. in ethanol 312 mp E 9450) clearly dis- tinguishes between the valency tautomers (I) and (11) although earlier workers have always assumed with- out rigorous proof that analogous products from tetraphenylcyclopentadienone had structures cor-responding to (1). The very ready aromatisation of the dihydro-derivative to the indane (111) with N-bromosuccinimide at room temperature does favour the cyclohexadiene structure (I).The problem is of some interest since it might give evidence on whet her tetraphenylcyclobu tadiene reacts in the closed (IV) or open form (V). So far we have not succeeded in isolating an adduct in which the four- membered ring is retained. Ph . ,- .. Ph Ph (IV) (VO On the basis of the absence of C=C absorption from its Raman spectrum and the Zsymmetry of its unit cell Freedman and Petersen5 have assigned the structure "octaphenylcubane" to the dimer of tetra- 303 mp E 19,500 which has just been recorded,s is phenylcyclobutadiene. The ultraviolet spectrum which we have confirmed (Am&x. 267 mp E 44,600 in tetrahydrofuran) would have to be explained by conjugation involving the bent bonds. Another possibility which would be consistent with the alter- nating tetrad axis with its thermal stability and with its ultraviolet spectrum would be the tautomeric to give the diphosphine-palladium chloride complex Chatt and Hart J.1960 1378. PROCEEDINGS and tetraphenylcyclobutadieneadducts or dimer. In benzene or chloroform at 20” a coloured solution is immediately formed (green or red respectively) that gives a strong and well-resolved electron spin resonance spectrum which is now being analysed and must be attributed to at least two radicals. (Received February 14th 1963.) The Configuration of the Anion in CsReCb J. E. FERGUSSON, By WARDT. ROBINSON and BRUCE R. PENFOLD (DEPARTMENT OF CHEMISTRY UNIVERSITY OF CANTERBURY CHRISTCHURCH NEWZEALAND) LIGAND field theory has been used to predict that the anion [ReCI,]- is regular tetrahedral1 with a d$ electron configuration for Re(@.The magnetic moment2 of RbReC1 does not contradict this pre- diction. With a view to confirming the stereochem- istry of rhenium in this anion we have carried out an X-ray analysis of the compound of empirical composition CsReCI . The crystals are orthorhombic with the non-centrosymmetric space group Ama2. There are 12 CsReC1 units in a unit cell of dimensions a = 10.66 b = 14-08 and c = 14.02 A. All atoms have been located with the aid of three-dimensional electron-density maps and for the incompletely refined structure the R factor for all observed (hkl) reflections is 0.21. The [ReCI,]- anion is not tetrahedral but is tri- meric with a triangle of bonded rhenium atoms.Each of these is bonded to two bridging chlorine atoms and one terminal chlorine atom in the plane of the triangle and also to two chlorine atoms on opposite sides of this plane (see Figure). Rhenium atoms are therefore seven-co-ordinate and each is at the centre of a distorted pentagonal bipyramid which has as its apices two chlorine atoms. If there are to be sufficient bonding orbitals it is clear that a seven-co-ordinate rhenium(m) atom requires a spin-paired d4electron ~onfiguration.~ The space group requires of the ion only a plane of symmetry. This passes through one rhenium and its three attached terminal chlorines and also the bridging chlorine to which it is not bonded.However the ion possesses symmetry 32 (= D3)within experi- mental error the three-fold axis being normal to the triangle of rhenium atoms. Mean values for bond lengths are Re-Re = 2.50 A Re-Cl (bridging) = 2.43 d; Re-Cl (terminal in plane of rhenium atoms) = 2-60 A and Re-Cl (terminal off plane of rhenium atoms) = 2.35 A. A detailed account of the crystal structure analysis will be presented later together with a full discussion of this most interesting and previously unreported stereochemical arrangement. All calculations were carried out on the University of Canterbury IBM 1620 computer using pro- grammes written by us and by Dr. D. van der Helm of the Institute for Cancer Research Philadelphia 11 Pa. U.S.A.The work was supported by the New Zealand Universities Research Committee by grants for equipment and a Research Fellowship (to W.T.R). (Received February 8th 1963.) 1 Orgel “Quelques Prob. de Chemie Minerale,” 10th Solvay Conf. 1956 289. Klemm and Frischmuth,2.anorg. Chem. 1943 13 253. Nyholm Proc. Chem. Soc. 1961 273. APRIL1963 117 ElementaI Organic Compounds. Part VII.' A Bridged Bisbenzenechromium n-Complex* By M. TSUTSUI and M.N. LEVY (NEW YORK UNIVERSITY DIVISION OF CHEMICAL RESEARCH DEPARTMENTS ENGINEERING AND CHEMISTRY NEW YORK 53 U.S.A.) MANYbridged metallocenes have been reported.2 We aluminium hydride gave 1,2,3,4- tetraphenylbu tane now report preparation of the first bridged bisarene proving the presence of this hydrocarbon as ligand.n-complex (I) in a one-step synthesis; it results from No trace of this was detected in the original reaction the reduction of a transition-metal halide by triethyl eliminating the possibility that it was a reaction aluminum in an arene environment. coupling product. The magnetic susceptibility of the The iodide or tetraphenylborate of the complex (I) complex kindly determined by the Gouy method by was obtained on reaction of chromic chloride Dr. W. Cave Monsanto Chemical Company St. (1 mol.) trans-stilbene (1 mol.) and triethyl-Louis Missouri was 1-96 B.M. The assignment of aluminum (3 mol.) in boiling n-heptane. The iodide complex-formation between the 1- and the 4-phenyl was precipitated and it crystallised from chloroform group rather than between those in other positions and n-pentane as golden-orange crystals m.p.is based on conformational feasibility from mole- 145-146" )Imax. 275-285 mp (E 2.5 x lo4) cular models. In addition our proposed mechanism [Found C 60-75; H 4-7; Cr. 9.7%; M 550 557 of cyclic reductive dimerisation of trans-stilbene with (thermoelectric measurement in dibromomethane). simultaneous n-donation suggests the 1-and the C2,H2,CrI requires C 61.0; H 4.7; Cr 9.4%; M 4-phenyl group as those most favourably disposed. 5411. PyroIysis of the iodide or its reduction3 by lithium Acknowledgment is made to the donors of The Petroleum Research Fund administered by the American Chemical Society and the Arakawa Forest Chemical Company Osaka Japan for support of this research.(Received,Januury 2 1st I 963.) * Presented in part at the 142nd National Meeting of the American Chemical Society September 1962. Paper VI Tsutsui and Chang Cunad.J. Chem.,in the press. Rinehart Curby Gustafson Harrison Bozak and Budlitz. J. Amer. Chem. SOC.,1962 84 3263; Rosenblum Banerjee Danieli and Herrich Tetrahedron Letters 1962 423 ;Schlogl and Reterlik ibid. p. 573 ;Gustafson Diss. Abs. 1962 23 71; Schlogl and Seiler Monatsh. 1960 91 79; Tetrahedron Letters 1960 No. 7 4. * Zeiss and Tsutsui J. Amer. Chem. SOC., 1957,79 3062. NEWS AND ANNOUNCEMENTS Chemical Society Awards.-The Council has 26-29th 1963 the Council of the Society awarded awarded the following Medals for 1963 grants from the Ethel Behrens Fund to the following Longstaff Medal to Lord Todd.Fellows R. G. S. Banks (Balliol College Oxford) Flintoff Medal to Professor H. Raistrick. R. A. Dawe (St. Catherine's College Oxford) J. V. Chemical Society Lectureships.-The Council has Ramsbottom (Balliol College Oxford) and R. G. made the following appointments for 1963-64 Thorp (Balliol College Oxford). Liversidge Lectureship .. Professor J. S. Anderson Election of New Fellows.-1 52 Candidates were Simonsen Lectureship .. Professor G. Ourisson elected to the Fellowship in March 1963. (Strasbourg) Deaths.-We regret to announce the deaths of Tilden Lectureships .. Dr. V. M. Clark Mr. A. 7'. Dann (15.12.62) of the Annual Health ProfessorA. F.Trotman-Research Laboratory Parkville Victoria ; Dr. P. Dickenson Hudson (9.2.63) of the University of Sydney Centenary Lectureship ,.Professor C. Djerassi N.S.W. ;and Dr. A. D. Mitchell (25.3.63) llford (Stanford California) Assistant Editor 1926-1962. Robert Robinson Lecture- Professor R. B. Wood-Library.-The Library will close for the Whitsun ship .. .. .. ward (Harvard) Holiday from 6 p.m. Friday May 31st until Ethel Behrens Awards.-In connection with the 9.30 a.m. Wednesday June 5th 1963. Anniversary Meetings held at Cardiff on March Royal Society.-The following were included I18 amongst those elected to the Fellowship of the Royal Society on March 21st Dr. Thomas Stevens Stevens Reader in Organic Chemistry in the University of Sheffield. Distin- guished for his original researches in organic chem- istry particularly for his work on the mechanism of molecular rearrangements.Dr. Theodore Morris Sugden Reader in Physical Chemistry in the University of Cambridge. Distin- guished for his work on the detection and identifica- tion of transient species radicals ions and electrons in flames and on the reaction mechanisms involved. Professor Arthur James Cochrane Wilson Professor of Physics in the University of Wales University College of South Wales and Monmouthshire Cardiff. Distinguished for his contributions to X-ray analysis in the elucidation of the atomic structure of crystals. National Chemical Laboratory-Standard Samples of Organic Compounds.-Additions have recently been made to the fist of pure compounds which are available from the National Chemical Laboratory.The samples are suitable for calibration of spectro- meters and for precise physicochemical measure- ments. Most of the 78 compounds now listed have a certified purity of greater than 99.9 per cent. Further information on the supply of these Standard Samples can be obtained from The Director National Chemical Laboratory Teddington Middlesex. The American Crystallographic Association is to publish as Monograph No. 5 the 2nd Edition of “Crystal Data (Determinative Tables),” covering the literature from 1912-1960. It is obtainable from the Polycrystal Book Service G.P.O. Box 620 Brooklyn 1 N.Y. Price $20.00. Personal.-Dr. M. P. Barnett Associate Professor of Physics and Director of the Co-operative Com- puting Laboratory at the Massachusetts Institute of Technology has been appointed to the Readership in Information Processing tenable at the Computer Unit.Professor N. B. Chapman has been appointed Foundation Governor of Batley Grammar School until March 24th 1964 in place of Professor D. C. Johnson who has resigned. Dr. G. C. Culling formerly of the University of Chicago is now a Research Chemist at the Jackson Laboratory of E. I. Du Pont de Nemours and Company Wilmington Delaware U.S.A. Dr. J. F. J. Dbpy has been appointed Vice- Principal of Chelsea College of Science and Tech- nology in addition to his present appointment as Head of the Department of Chemistry. Dr. H. E. Haallam has been given a year’s leave of absence from the University College of Swansea to become Adviser in Physical Chemistry to the new University of Nigeria at Nsukka.PROCEEDINGS Dr. A. B. Hart is now Head of the High-tempera- ture Chemistry Section Chemistry and Biology Division Central Electricity Generating Board. The Title of Reader in Cytochemistry in the University of London has been conferred on Dr. J. S. Hoft in respect of his post at Middlesex Hospital Medical School. Dr. P. Johnson formerly University Research Fellow at Birmingham Medical School has been appointed Visiting Lecturer in the Biochemistry Division of the Chemistry Department University of Illinois Urbana U.S.A. Mr. K. A. R. Julian is now Technical Sales Manager €or the North R. and J. Dempster Ltd. Mr. B. H. Kingston formerly of A.Boake Roberts and Company Limited has joined the staff of Proprietary Perfumes Limited Dr. R. H. Ottewill has been promoted to Lecturer in Physical Chemistry within the Department of Physical and Inorganic Chemistry at the University of Bris tol. Dr. E. C. Potter is now adviser to the Chemistry and Biology Division of the Central Electricity Generating Board on scientific matters with par- ticular reference to electrochemistry and corrosion. The title of Reader in the History and Philosophy of Science has been conferred upon Dr. W. A. Smeaton in respect of his post at University College London. Dr. E. S.Stern is to succeed Dr. J. Chatt as Head of the Imperial Chemical Industries Ltd. organo- metallic research team at the Petrochemical and Polymer Laboratory at Runcorn Heath formerly located at the Frythe.Mr. J. R. WilinJield C.B.E. inventor of “Tery- lene” and a Director of Imperial Chemical lndustries Limited (Fibres Division) has retired. Symposia,etc.-A Conference on Prediction and Assessment of Paint Performance will be held at Scarborough Yorks. on June 25-29th. 1963. Further enquiries should be addressed to the Secretary Oil and Colour Chemists Association Wax Chandlers’ Hall Gresham Street London E.C.2. The Sixth International Conference of Ionisation Phenomena in Gases sponsored by the International Union of Pure and Applied Physics will be held at Orsay France on July 8-13th 1963. Further enquiries should be addressed to P. Hubert CENFAR Boite Postale 6 Fontenay-aux-Roses (Seine) France.The Third International Conference on the Physics of Electronic and Atomic Collisions will be held in London on July 22nd-26th 1963. Further enquiries should be addressed to the Conference Secretaries Physics Department University College Gower Street W.C. 1. APRIL 1963 The Sixth Congress of the European Molecular Spectroscopy Group sponsored by I.U.P.A.C. will be held in Budapest on July 22nd-27th 1963. Further enquiries should be addressed to the Hungarian Academy of Sciences Budapest Hungary. The Sixth International Congress on Nutrition will be held in Edinburgh on August 9-15th 1963. Further enquiries should be addressed to Dr. A. B. Meiklejohn Department of Clinical Chemistry Royal Iniirmary University of Edinburgh Edin- burgh.The Eleventh International Congress of Refrigera- tion will be held in Munich on August 27th- September 4th 1963. Further enquiries should be addressed to International Institute of Refrigeration 177 Boulevard Malesherbes Paris 17e; or RE-FRICONGRESS Pfeufferstr. 2 Munich 25. A Symposium on Oxygen in the Animal Organism sponsored by the International Union of Biochem- istry and International Union of Biological Sciences will be held in London on September lst-5th 1963. Further enquiries should be addressed to Professor F. Dickens Courtauld Institute of Biochemistry Middlesex Hospital Medical School London W.l. An International Symposium on Nitro-compounds arranged by the Polish Academy of Sciences under the auspices of the I.U.P.A.C.will be held in Warsaw on September 18-20th 1963. Further enquiries should be addressed to Sekretariat Miedzynarodowego Sympozjonu o Nitrozwiazkach Palac Staszica Warsawa Nowy Swiat 72 Poland. The Eleventh International Conference on Spectro- scopy sponsored by the Society for Applied Spectro- scopy will be held in Belgrade from September 30th-October 4th 1963. Further enquiries should be addressed to Dr. V. VukanoviC General Secre- tary Prirodno-matematicki C Fakultet FiziCko-hemijski zavod Studentski trg. 16 Blok C Belgrade Yugoslavia. An International Conference on Beryllium Oxide sponsored by the Australian Atomic Energy Com- mission will be held in Lucas Heights (near Sydney) Australia on October 21st-25th7 1963.Further enquiries should be addressed to the Conference Secretary,A .A. E.C. Research Establishment ,Private Mail Bag Sutherland N.S.W. Australia. The Thirty-seventh Congress of the Australian and New Zealand Association for the Advancement of Science will be held in Canberra Australia from January 2&24th 1964. Further information may be obtained from the Joint Secretaries Section B A.N.Z.A.A.S. Department of Chemistry School of General Studies Australian National University Box 197 P.O. Canberra City A.C.T. Australia. An International Symposium on the Chemistry of Natural Products arranged by the Science Council of Japan under the auspices of the I.U.P.A.C. will be held on April 12-18th7 1964.Further enquiries should be addressed to the General Secretary of the Organising Committee Professor Kyosuke Tsuda University of Tokyo Japan. The Third International Wool Textile Research Conference organised by the Textile Institute of France under the patronage of the International Wool Secretariat and the International Wool Textile Organisation with the support of the Central Wool Committee of France is to be held in Paris on June 28th-July 9th 1965. Further enquiries should be addressed to Institut Textile de France 59 Rue de la Faisanderie Paris (16e). FORTHCOMING SCIENTIFIC MEETJNGS London Thursday May 9th at 6 p.m. Hugo Miiller Lecture “The Biogenesis of Phenolic Alkaloids,” by Professor D. H. R. Barton D.Sc.F.R.S. to be given in the Lecture Theatre The Royal Institution Albemarle Street W. 1. Thursday June 6th at 6 p.m. Meeting for the Reading of Original Papers. To be held in the Rooms of the Society Burlington House w.l. Birmingham Friday May loth at 4.30 p.m. Hugo Miiller Lecture “The Biogenesis of Phenolic Alkaloids,” by Professor D. H. R. Barton D.Sc. F.R.S. Joint Meeting with the University Chemical Society to be held in the Chemistry Department The University. Liverpool Thursday May 9th at 5 p.m. Lecture “Magnetism and Stereochemistry of First Row Transition Elements,” by Professor J. Lewis Ph.D. D.Sc. Joint Meeting with the University Chemical Society to be held in the Donnan Labora- tories Chemistry Department The University.Manchester Thursday May 2nd at 6.30 p.m. Hugo Miiller Lecture “The Eiogenesis of Phenolic Alkaloids,’’ by Professor D. H. R. Barton D.Sc. F.R.S. to be given in Room Fl Manchester College of Science and Technology. 120 Northern Ireland Tuesday May 21st at 4.30 p.m. Official Meeting and Lecture “Some Recent De- velopments in the Chemistry of Polyenes,” by Pro-fessor B. C. L. Weedon D.Sc. F.R.J.C. Joint Meeting with the Royal Institute of Chemistry and the Society of Chemical Industry to be held in the Department of Chemistry David Keir Building Queen’s University Belfast. North WaIes Thursday May 9th at 5.45 p.m. Lecture “Some Problems Experienced in the Manu- facture of Pure Beryllium,” by J. A. Dukes. Joint Meeting with the University College of North Wales Chemical Society to be held in the Chem- istry Department University College Bangor.PROCEEDINGS Oxford (Joint Meetings with the Alembic Club to be held in the Inorganic Chemistry Laboratory.) Monday May 13th at 8.30 p.m. Lecture “Some New Natural Products-Structural and Biosynthetic Studies,” by Dr. W. D. Ollis. Monday May 27th at 8.30 p.m. Lecture “Reactions of some Cyclobutadiene Com- plexes,” by Professor R. C. Cookson M.A. Ph.D. Reading Tuesday May 7th at 5.45 p.m. Lecture “Simple and Complex Metal Nitrates and Nitrites,” by Professor C. C. Addison D.Sc. F.R.I.C. Joint Meeting with the Royal Institute of Chemistry and University Chemical Society to be held in the Large Chemistry Lecture Theatre The University.OBITUARY NOTICES SIR IRVINE MASSON 1887-1962 SIRIRVINE M.B.E. formerly Vice-Chancel- MASSON lor of the University of Sheffield died in Edinburgh on October 22nd 1962 at the age of seventy-five. He was a grandson of David Masson Professor of English Literature at University College London 1853-65 and at Edinburgh University 1865-95 and the only son of Sir David Orme Masson Pro- fessor of Chemistry at the University of Melbourne during 1886-1923. James Irvine Orme Masson was born on September 3rd 1887 in Melbourne where he was educated at the Grammar School and University before coming to Great Britain in 1910 as an 1851 Exhibition Scholar. After a year with Sir James Walker at Edinburgh Masson moved to University College London where he worked with Sir William Ramsay with whom his father had been associated some thirty years earlier at University College Bristol.In 1912 Masson became assistant in chem- istry at University College London and during the First World War he served with the R.N.V.R. and later at the Research Department of the Royal Arsenal Woolwich. He was appointed Reader in inorganic chemistry at University College in 1920 and served the Chemical Society both as secretary (1921-24) and as an ordinary member of Council (1929-32). He became the first Professor of Chem- istry and Head of the Department of Pure Science of the University of Durham (Durham Division) in 1924 and in 1938 he was appointed Vice-Chancellor of Sheffield where he remained until retirement in 1952.At the outbreak of the Second World War he added to his duties as Vice-Chancellor the active direction of a Ministry of Supply team which was housed in the Chemistry Department and engaged in research on explosives. He was also a member and chairman of several committees of the Ministry. Irvine Masson was a man of varied talents. His chemical work which led to election to the Royal Society in 1939 was not voluminous since he never had a large team of helpers and much was done with his own hands. But his work is characterised by elegance and breadth; problems were reduced to the simplest possible terms and then approached by neat ingenious and well-considered methods.He be-longed to the group of chemists-now non-existent-who could tackle equally well problems in the in- organic physical and organic sections of chemistry. His first publication in the Journal (1910 851) dealt with the action of calcium carbide as a dehydrating agent and from Australia also came work in con- junction with his father on the decomposition of metallic cyanates by water (2.phys. Chem. 1910,70 290). Subsequent publications were concerned with the theory of solutions and with the experimental investigation of van der Waals forces by the measure ment of pressure-volume-temperature relations in pure and mixed gases up to pressures of 125 at- APRIL1963 mospheres. His use of the inert gases was clearly an outcome of association with Ramsay for whom he had unbounded admiration; but the choice was nevertheless not haphazard.One of the initial objec- tives was to discover whether physical or chemical forces were responsible for volume changes on mix-ture; and by finding similarities between mixtures of ethylene with oxygen and with argon Masson pro- duced in the first paper of the series prima facie evidence in favour of the physical explanation. This investigation of gases was continued for a period of nearly 10 years (Proc. Roy. SOC.,1923 A 103 524; 1929 A 122,283; 1930 A 126 268) and was prob- ably the first systematic study of Dalton’s law of partial pressures. To the organic chemist his most significant con- tributions were his studies of nitration and of the organic compounds of iodine.The nitration work (in particular J. 1933 105) is a classic in which he de- limited precisely the conditions requisite for the nitration of nitrobenzene and in a fashion typical of his liking for graphical demonstration exhibited his results with great clarity in the form of a triangular diagram. He did not arrive at the modern mechanism of nitration but undoubtedly paved the way for those later researches which have had an important bearing on the development of theoretical organic chemistry. The study of organic iodine compounds had been neglected since the work of Victor Meyer and Willgerodt some forty years earlier; Masson developed this work in an important series of re- searches (J.,1935 1669; Nature 1937 139 150; J.1937 1718; J. 1938 1699 1702 1708) in which he demonstrated once more his remarkable ability to apply physical chemistry to problems in both organic and inorganic fields. Masson was interested in the teaching of science frequently serving as external examiner at other uni- versities and his book “Three Centuries of Chem- istry,” published in 1925 shows a profound know- ledge of the historical and philosophical aspects of his subject. In addition it is enlivened and made eminently readable by unexpected touches of dry humour. For instance in comment on the renowned misogynist “It must have been a lasting shame to Cavendish that once on Clapham Common he was so far overtaken by irrational impulse as to rescue a female from the menaces of a cow.” As has been already indicated Masson’s work as a chemist did not cease with his appointment as Vice- Chancellor.However the problems now under investigation such as the viscosity of amatols were mainly of immediate practical importance and not likely to bear fruit in the shape of post-war publica- tions. He turned for relaxation to other matters; for whilst science was his profession classics and the arts were hobbies in which he reached professional standing. He was a great lover of books and an expert in the early history of printing. His major contribution in this field-the fruit of research during the latter years of his vice-chancellorship-is an out- standing monograph on the Maim Psalters and Canon Missae.Masson was tireless in his efforts on behalf of the University of Sheffield and his flair for the lucid presentation of complex matters of finance and statistics proved invaluable. No one knew better than he what education was or what a university should and could do. As Vice-Chancellor he steered the University through its war-time problems and through the equally difficult period of development which followed. He made plans to guide the expan- sion from 850 students to a well-balanced and effec- tive University of some 2500 students (a target now increased to 4750); temporary structural alterations were made and building programmes initiated ; Stephenson Hall and the new Chemistry and En- gineering Departments were almost completed before his retirement in 1952.Though invariably courteous and dignified Masson was nevertheless a determined fighter often outspoken in formal debate. In contrast his natural shyness in personal relationships was a handicap frequently mistaken for aloofness. Rarely at his best in large gatherings he was an excellent host of a small party with a turn of wit which would have sur- prised his detractors. Those who knew him well found him a lively companion and a true friend. Masson became M.B.E. in 1918 was knighted in 1950 and received the honorary degree of LL.D. of the Universities of Edinburgh and Sheffield. His re- tirement spent in bibliographical work in Edinburgh and in advisory posts to the Carnegie Trust and the University of St.Andrews was saddened by the long illness of his wife Flora Lovell daughter of Professor G. Lovell Gulland whom he married in 1913 and who died in 1960. He leaves one son who is sub- librarian in charge of the Brotherton Collection at the University of Leeds. R. D. HAWORTH A. H. LAMBERTON. (Reproduced in part with permission from Nature.) PROCEEDINGS HAMILTON McCOMBIE 1880-1962 HAMILTON was born in London on May MCCOMBIE 7th 1880. He started his academic career as a student at Aberdeen where he took the M.A. degree in 1900 and then moved to the Royal College of Science (now part of the Imperial College of Science and Technology) where in 1903 he qualified as A.R.C.S. and also obtained the London B.Sc. degree. He com- pleted his student days by working at Strasbourg under Thiele and took the Ph.D.degree (magna cum lauda) there in 1905. He was then appointed initially to an assistant lectureship at Mason College (now the University of Birmingham) and attended the Australia meeting of the British Association for the Advancement of Science in the summer of 1914. The First World War was declared during this meeting which was at once abandoned. Overseas members had to find their own way home at a time of great confusion. McCombie hastened back as rapidly as possible and in January 1915 was commissioned in the Worcestershire Regi- ment. He proceeded to France and in June 19 15 was seconded to the newly-formed Special (Gas) Brigade of the Royal Engineers.The rapid development of gas warfare made the provision of expert chemical advice essential to the higher staffs of the armies and McCombie was soon appointed Chemical Adviser to the First Army with finally the rank of Major. His work at the Headquarters of the First Army was clearly excellent and he received the M.C. and the D.S.O. and was twice mentioned in dispatches. After the Armistice he went for a short spell to Germany with the Army of Occupation again primarily in a chemical advisory capacity. During the war he had met J. Barcroft (later Sir Joseph Barcroft and Professor of Physiology at Cambridge) who was a member of the Chemical Warfare Committee; he had also met W. J. Pope (later Sir William Pope) the Professor of Chemistry at Cambridge who had taken a prominent part in many chemical matters affecting the army in France.It was undoubtedly these scientists who suggested that McCombie should come to Cambridge when his army service ended. McCombie came to Cambridge in October 1919 as a Fellow of King’s College. The number of estab-lished University teaching posts was comparatively small at that time H. J. H. Fenton had the post of Demonstrator F. W. Dootson that of Additional Demonstrator and therefore McCombie became Second Additional Demonstrator. The new statutes in the middle twenties considerably increased the number of University teaching posts and enabled McCombie to shed his odd rank on election to a lectureship. During the period between the death of Sir William Pope (1940) and the election of A.R. Todd (now Lord Todd) to the Chair of Organic Chemistry (1944) McCombie was Acting Head of the Department of Organic Chemistry. He was elected to a personal Readership in Chemistry shortly before his retirement in 1945. Later he moved to Woking where he spent the rest of his life. McCombie’s academic abilities clearly lay more strongly in administrative than in research matters. At Birmingham his first paper (with Miss E. Parry J. 1909,584) described the preparation and reactions of cyanhydrins of aromatic aldehydes. Then with A. E. Everest (J. 1911 1744) and subsequent students he studied the formation of imidazole com- pounds from aromatic a-amino-ketones. At Birm- ingham he also started two research projects to which he later returned in his Cambridge days.With H. A. Scarborough who worked with him at Birmingham during the period 1912-1914 and re- joined him at Cambridge from 1920-1935 he investigated the kinetics of ester saponification and of quaternary salt formation. These measurements were of significant value in the early development of quantitative theories of “induced polarity” which ultimately led to unification by the Hammett pa equation of 1935. Thus W. Blakey McCombie and Scarborough(J.,1926,2867) showed that the relative rates of saponification of ethyl disubstituted benzo- ates were approximately the product of the relative rates of saponification of the corresponding two monosubs t i tu ted esters .The second project started at Birmingham was the study of the stepwise chlorination of substituted phenols and with S. A. Brazier (J. 1912 968) he found that p-iodophenol gave first an iododichloride which on storage underwent a rearrangement to 2-chloro-4-iodopheno1. This was the first known example (and remains one of the very few known today) of the migration of a group in a side-chain to a nuclear meta-position. The fact that this rearrange- ment involved dissociation giving molecular chlorine which then gave nuclear substitution under the in- fluence of the phenolic group was clearly shown in wider studies by S. Buchan and McCombie (J.,1931 137; 1932 2857). In the meantime (1925-1929) it led McCombie and Scarborough with a succession of students (W.A. Waters W. Blakey W. Burns) to examine the possibility of transfer of a halogen atom from one ring to another by the rearrangement of a N-halogeno-amide of biphenyl or azobenzene. This work was extended to a general study of substitution in linked ring systems and showed that in compounds APRIL1963 such as biphenyl benzophenone azobenzene and diphenyl ether the reactivities of the two benzene nuclei were independent of each other apart from the inductive influence of each aryl substituent on the other ring (see W. A. Waters Chern. Rev. 1930 7 407). As the current importance of these research projects decreased McCombie apparently lost his interest in research for after about 1925 he initiated no new research projects.It should be added that in his early Cambridge days he also directed work on the destruction of mustard gas by chlorine and on the properties of related compounds such as divinyl sulphoxide and he continued to act as a member of the Chemical Defence Research Committee up to the end of the Second World War. He received the Degree of Doctor of Science at Birmingham and Cambridge Universities. His administrative activities in Cambridge were shown at King’s College where he became Assistant Tutor for Natural Sciences and was a member of the College Council and the Estates Committee for many years and in the University where he was for various periods a member of the Council of the Senate the General Board the Buildings Syndicate and the Faculty Board of Physics and Chemistry.In the wider sphere of activity he was a member of the Council of the Chemical Society for the periods 1923-6 and 193841 of the Council of the Royal Institute of Chemistry for 1923-6 1929-32 and 1937-8 and of the Council of the Society of Chemical Industry for 1942-5 when he became a Vice-president for the period 1945-8. McCombie was a friendly and very hospitable man. At the time of his arrival at King’s College C. T. Heycock-who was a Fellow of the College and Goldsmiths’ Reader in Metallurgy-used to in- vite the chemistry examiners in the Natural Sciences Tripos each year to a private dinner in the College. For many years McCombie and Dr. D. Stockdale continued this practice and there must be very many past examiners internal and external who have the happiest memories of these dinner parties both for the pleasure of the occasion itself and for the mental relaxation which it afforded during a trying period of intensive script-marking.McCombie was deeply interested in rugby football and it is unlikely that he ever missed attending the University match at Twickenham during his Cambridge years. He is survived by his widow and three daughters. I am indebted to many friends for information and particularly to Dr. W. A. Waters F.R.S. who kindly provided a summary of McCornbie’s research work. F. G. MANN. ADDITIONS TO THE LIBRARY Russian-English chemical and polytechnical diction- ary. L. I. Callaham. Pp.892.2nd edn.J. Wiley and Sons. New York. 1962. A survey of non-aqueous conductance data. G. J. Janz F. J. Kelly and H. V. Venkatasetty. Pp. 51. Rensselaer Polytechnic Institute. Troy New York. 1962. (Presented.) Consolidated index of selected property values physical chemistry and thermodynamics ;prepared by the Office of Critical Tables. (NRC Publication 976.) Pp.274. National Research Council. Washington. 1962. Structupl inorganic chemistry. A. F. Wells. 3rd edn. Pp. 1055. Clarendon Press. Oxford. 1962. Inorganic isotopic syntheses. Edited by R. H. Herber. Pp. 249. Benjamin. New York. 1962. Inorganic polymers. Edited by F. G. A. Stone and W. A. G. Graham. Pp.631. Academic Press. New York. 1962. Developments in inorganic polymer chemistry.Edited by M. F. Lappert and G. J. Leigh. Pp. 305. Elsevier. Amsterdam. 1962. Advances in physical organic chemistry. Edited by V. Gold. Vol. 1. Pp. 443. Academic Press. London. 1963. Stereochemical correlations. W. Klyne. (R.T.C. Lecture series 1962 No. 4.) Pp. 26. Royal Institute of Chemistry. London. 1962. (Presented by the publisher.) Reactions of organic compounds a textbook for the advanced student. R. C. Fuson. Pp. 765. J. Wiley and Sons. New York. 1962. Ions in hydrocarbpns. A. Gemant. Pp. 261. Inter-science Publishers Inc. New York. 1962. Die atherischen Ole. N. E. Gildemeister and F. Hoffmann. 4th edn. W. Treibs et al. Vol. 3b. Pp. 436. Akademie Verlag. Berlin. 1962. Pyridine and its derivatives. Edited by E. Klingsberg. (The chemistry of heterocyclic compounds.Edited by A. Weissberger. Vol. 14 (3).) Pp.915. Interscience Publishers Inc. New York. 1962. Quantitative chemical techniques of histo- and cyto- chemistry. D. Glick. Pp. 470. Vol. 1. Interscience Pub- lishers Inc. New York. 1961. Peptide synthesis. H. N. Rydon. (R.T.C. Lecture series No. 5.) Royal Institute of Chemistry. London. 1962. (Presented by the publisher.) Enzymatic synthesis of DNA. A. Kornberg. Pp. 103. J. Wiley and Sons. New York. 1961. The nature of biochemistry. E. Baldwin. Pp. 111. University Press. Cambridge. 1962. The principles and practice of modem cosmetics. Vol. 1. Modem cosmeticology. R. G. Harry. Revised by J. B. Wilkinson. 5th edn. Pp.683. Leonard Hill. London. 1962. Practical neurochemistry.H. McIlwain and R. Rodnight. Pp. 296. Churchill. London. 1962. Official standardised and recommended methods of analysis compiled and edited for the Analytical Methods Committee of the Society for Analytical Chemistry by S. C. Jolly. Pp.577. Heffer. Cambridge. 1963. (Presented by the S.A.C.) 124 The analytical chemistry of indium. A. I. Busev. (Translated from the Russian by J. T. Greaves.) Pp. 288. Pergamon Press. Oxford. 1962. The inorganic analysis of petroleum. J. W. McCoy. Pp. 271. Chemical Publishing Co. New York. 1962. Trattato di chimica industriale. Edited by M. Giua. Vol. 4 (2). Pp. 975. Unione Tipografico Editrice Torinese. Turin. 1963. (Presented by the publisher.) Laboratory planning. J. F. Munce.Pp. 360. Butter-worths Scientific Publications. London. 1962. Design and construction of laboratories. R. R. Young and P. J. Harrington. (R.I.C. Lecture series 1962 No. 3.) Pp. 14. Royal Institute of Chemistry. London. 1962. (Presented by the publisher.) Residue Reviews residues of pesticides and other foreign chemicals in foods and feeds. Edited by F. A. Gunther. Vol. 1. Pp. 162. Springer-Verlag. Berlin. 1962. Progress in infrared spectroscopy. Vol. 1 :Proceedings of the fifth annual Infrared Spectroscopy Institute held at Canisius College Buffalo New York August 14-1 8, 1961. Edited by Herman A. Szymanski. Pp. 446. Plenum Press. New York. 1962. NEW JOURNALS Acta Biochimica Iranica from 1962 1. Advances in Physical Organic Chemistry from 1963,l. Boletin de la Sociedad Chileiia de Quimica from 1961,ll. Electrochemical Technology from 1963 1. Textile Institute and Industry from 1963 1. Berichte der Bunsen Gesellschaft fur Physikalische Chemie (former& Zeitschrift fur Elektrochemie) from 1963 67.