118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order. The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure.This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order. The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point.These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order. The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility.The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order.The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order.The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure.This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order. The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point.These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13. 18 THE MAGNETISM OF FREE RADICALS THE MAGNETISM OF FREE RADICALS. BY SAMUEL SUGDEN. Received 14th August, 1933. Free radicals have been defined as substances whose molecules contain an odd number of e1ectrons.l This definition is not quite comprehensive since, for example the CH, radical found by positive ray methods is an even molecule; these exceptions are, however, rare and nearly all the free radicals which interest the organic chemist are odd molecules. Most substances whose molecules contain an even number of electrons are feebly repelled by a magnet and were described by Faraday as 1 Schlenk, 4th Cons.Chim. Solvay, 503, 1931.S. SUGDEN I9 diamagnetic. Paramagnetic substances which are attracted by a magnet include compounds of the transition elements and a few other substances. In I924 G. N. Lewis deduced that all odd molecules should be paramagnetic.2 but the quantitative discussion of magnetic phenomena presented great difficulty until the advent of the modern quantum theory. The quantum theory of magnetism has developed very rapidly in recent years and is capable of accounting for a wide range of ex- perimental data ; i t is therefore possible to predict with fair certainty the kind of magnetic behaviour which molecules of a particular type should exhibit.In terms of the modern theory the molecular mass susceptibility of a substance which is not ferromagnetic may be regarded as the sum of three terms. XM = X d + X P + x r . * (1) The first term is the diamagnetic contribution and is due to the disturbance of the electronic orbits by the applied field. I t is always negative and does not vary with temperature. In the usual units xn ranges from - 1-88 for helium to a few hundred units for a complex organic molecule. For most organic compounds xp is zero and xr is probably very small so that X d is approximately equal to xM. Some ex- amples are given in Table I. TABLE ~.-DIAMA(PNETIC AND PARAMAGNETIC SUSCEPTIBILITIES. No. Substance. He xbr XP. 0 0 0 0 1292 I303 0 0 Xr - 0 ? ? ? ? ? 44 64 If the molecule possesses a permanent magnetic moment p this gives rise to the large positive term xp which is responsible for para- magnetism.The magnetic moment is determined by the resultant angular momentum of all the electrons and this is restricted by the quantum theory to certain definite values. For a polyatomic molecule i t has been shown4 that x,+ is determined almost entirely by the nett spin S and that, to a close approximation 5 4 w + 1) T ' x,, = 1.242 x 10 In an odd molecule with one unbalanced electron S = 4 and at 20' xp = + 1270. This is much larger than xd and of opposite sign ; from ( 2 ) x,, should vary inversely as the absolute temperature (Curie's Law). Many solutions of paramagnetic ions and some solid salts obey a For a comprehensive account of this theory see van Vleck, The Theory Valence and the Structure of Atoms and Molecules, p.148. Van Vleck, ref. 3, p. 274. of Electric and Magnetic Susceptibilities. Oxford University Press, 1932.20 THE MAGNETISM OF FREE RADICALS C T - A this law over a wide range ; for others the Weiss Law, xM = - holds with considerable accuracy. x,, is usually evaluated by subtracting xd from the experimentally measured value of x ~ . Pascal has shown that in diamagnetics the susceptibility is nearly an additive function ; since the diamagnetic correction is usually small compared with xy it can be computed with suficient accuracy by adding the appropriate constants. The data for bivalent copper and silver salts in Table I. show the relative magnitudes of x,, and xa.Even molecules have nearly always zero magnetic moment so that xp is zero. The normal state of the molecule is ST: with S = I ; this gives from equation (2) xp = 3390 and the observed values range from 3310 to 3480. Recent work on sulphur vapour indicates that S, is also 3,Z and has a strong paramagneti~m.~ The last term of equation ( I ) can usually be neglected in discussing paramagnetics. It is due to " exchange " forces, or interactions between the electrons of adjacent atoms in the molecule and the quantum theory shows that it is small, positive in sign, and independent of temperature. A feeble paramagnetism which does not vary with temperature has long been known in chromates and permanganates and appears in more complex salts, e.g. the cobaltammines, when the observed susceptibility is corrected for the diamagnetism of the attached groups.Two examples of this type of feeble paramagnetism are quoted in Table I. I t is clear from these theoretical considerations that the magnetic susceptibility furnishes a good test for the existence of free radicles (odd molecules). At room temperature such molecules should give a value of x,, = xv 1 xa of the order of + 1300 units. If xp is negative, or has a small positive value, then the molecule most probably contains an even number of electrons. The available experimental data for inorganic odd molecules are collected in Table 11. The simplest odd molecule, nitric oxide gives a value of xp of the expected order of magnitude but this is really a co- incidence.Since NO is a diatomic molecule it is necessary to consider the orbital component of magnetic moment and not the spin component only as in equation (2). A complete theory has been given by van Vleck,6 which gives a very satisfactory explanation not only of the value of the susceptibility a t room temperature but also for its variation with temperature down to - 160' C. The low value found by Son6 for nitrogen peroxide is most probably erroneous. More recently Havens has measured the susceptibility of this gas over a wide range of pressures at 20' and allowing for the varying degree of dissociation computed the susceptibilities of NO, and N,O,. (The latter is known to be diamagnetic.) For NO, he finds a molecular mass susceptibility of about 1400 units in fairly good agreement with theory.ClO, measured in benzene solution by Taylor and Lewis also gives the value predicted for an odd molecule. As a contrast the hypophosphates (Nos. 12-15, Table 11.) are dia- magnetic and must therefore be derived from the acid H,P,O, and not A striking exception is furnished by oxygen. Naud6 and Christie, Physic. Rev., 37, 174, 1931 ; Shaw and Phipps, ibid., Ref. 3, p. 269. 38, 174 ; N6e1, Compt. vend., 2035, 1932.S. SUGDEN 21 TABLE II.-ODD MOLECULES, INORGANIC. NO. 9 I0 I1 I2 13 I4 15 16 Subs tame. NO NO, t. 20 I35 20 20 20 20 20 20 20 XM. I465 I53 I392 1341 - 78 - 128 - 148 - 164 I I00 xd * - I 0 - 15 - I5 - 25 - 40 - 118 - 192 - 40 - I00 XP- I475 158 (calc. 913) I407 I 366 - 38 -48 - I 0 I2 1140 from H2P03, which is an odd molecule.This result confirms the con- clusions of Arbusov and Arbusov,13 who have prepared the true ethyl ester and find that its molecular weight (obs. 250 - 260) indicates the formula Et4P20, (M = 274). Very recently Asmussen has shown that the yellow salt of the empirical formula (KSO,),NO obtained by the alkaline oxidation of hydroxylamine disulphonic acid has only a feeble paramagnetism in the solid state whilst its blue aqueous solutions give nearly the theoretical paramagnetism for an odd molecule. This is ascribed to the change (solid) (KS03),N,02 + z(KSO,),NO (solution). Taylor and Lewis lo have also examined thallium amalgum and a solution of sodium in liquid ammonia but find only a feeble para- magnetism of the order of + 20 units per gram atom of thallium or sodium.Many metals, e.g. aluminium, whose atoms contain an odd number of electrons exhibit only feeble paramagnetism in the solid state. These magnetic anomalies are probably due to the ease with which an electron is set free giving rise to “ metallic ” conduction. For a discussion of this type of magnetic behaviour see van Vleck, op. cit., P* 347- Table 111. contains the available magnetic data for odd organic molecules. The triaryl methyls are the gee radicals which havg the longest history; so far only one, No. 17, has been examined mag- netically. Taylor found a considerable paramagnetism in a 7-2 per cent. solution of this substance in benzene, but the value deduced for xp is markedly low for an odd molecule. Too much stress cannot be placed on this value for it is based on an assumed degree of dissociation which is extrapolated from observations on more dilute solutions by Gomberg,14 and may be considerably in error.The next substance (No. 18) was prepared by Kenyon and Banfield l5 by oxidising a substance of the ‘Son4 Sci. Rep. Tohztku, I I, 139, 1922 ; Bauer and Piccard, J . Physique. 8 Son& JOG. cit. 1 0 Taylor and Lewis, Proc. Nut. Acad. Sci., 1 1 , 456, 1925 ; Taylor, J . Amer. l1 Bell and Sugden, J.C.S., 48, 1933. laAsmussen, 2. an. Chem., 212, 317, 1933. 13J. p r . Chem., 130, 121, 1931. l4 Chem. Rev., I , 104, 1924. 1, 97, 1920. Havens, Physic. Rev., 41, 337, 1932. Chem. Soc., 48, 858, 1926. l5 J.C.S., 1612, 1926.22 THE MAGNETISM OF FREE RADICALS ** 20' 17" 17" 24" 24' structure I with silver oxide.One atom of hydrogen was removed quantitatively Me& . CH, . CMe P h . N . O H PhN+O Me,. C . CH, . CMe I I I I PhN:O PhN+O I I1 but the molecular weight remained practically unchanged. From these and other considerations Kenyon and Banfield assigned structure I1 to the red oxidation product which is therefore an odd molecule. In accordance with these structures the parent body (No. 4, Table I.) is diamagnetic whilst the red oxidation product is strongly paramagnetic and has a susceptibility of the right order of magnitude both in benzene solution and in the solid state. This substance is a remark- ably stable free radicle. After nearly two years' storage in contact with air it is unchanged in appearance and has the same susceptibility. A substance with a similar structure has been described by Wieland and Offenbacker,ls viz., diphenyl nitrogen oxide, Ph,N : 0, but this appears to be much more unstable.Substance 16, Table II., has ob- viously a similar structure. Finally Table 111. contains some preliminary results obtained by the writer in a study of the magnetism of the ketyls. The evidence for the existence of ketyls as free radicals is largely based on chemical reactions and varying views have been expressed as to the amount of free radical present in the coloured solutions obtained by the action of alkali metals on anhydrous solutions of ketones. The high reactivity of these solutions with oxygen, iodine, and alkyl iodides is strong evi- dence in favour of the existence of the free radical R,C-0-Na ; Schlenk XP' -- 570 I293 1109 1080 1050 TABLE III.-FREE RADICALS. ORGANIC.17 18 I g 20 No. I Subs t ancc CL Naphthyl diphenyl methyl Cl,H&N* Phenyl +-diphenylyl ketone potassium Benzophenone potassium -I Conditions. In 7 per cent. benzene soh. Solid In 20 per cent. benzene soh. In 17 per cent. dioxan soh. In 15 per cent. dioxan soh. and Weickel,20; on the other hand Bachrnanq2l have recently ob- tained 98 per cent. yields of the pinacols by pouring the coloured sol- utions into acetic acid and conclude that the essential constituent of the coloured solution is the sodium derivative of the pinacol R, . C-0 Na R, C-ONa I l8 Bey., 47, 2111, 1914. l7 Taylor, J . Amer. Chem. SOC., 48, 858, 1926. l8 Kenyon and Sugden, J.C.S., 170. 1932. lS Sugden, unpublished data.*l J . Amer. Chem. SOC., 55, 1179, 1933. aOBer., 44, 1182, 1911.S. SUGDEN 23 It is clear however that if a free radical is present it may react very differently in acid and in alkaline solutions. Physical evidence gives a much more definite answer to the problem. Schlenk and Thal,22 showed that the soluble potassium compound of phenyl p-diphenylyl ketone had a molecular weight in ether corresponding with the simple formula R,C-0-K ; hence the substance in solution is almost entirely in the form of the free radical. The new magnetic evidence leads to the same conclusion. By using dioxan as solvent high concentrations of the ketyls were obtained and their paramagnetism placed beyond doubt. In a typical experiment a 20 per cent. solution of phenyl p-diphenylyl ketone in dioxan was found to have a mass susceptibility of - 0.593. After shaking with potassium in a stream of nitrogen the susceptibility became + 0.052. The solution was poured into water and the liberated alkali titrated; this gave a concentration of 17.4 per cent. of ketyl in the clear dark-green solution. The value of xp deduced from these results is + 1080; hence a t least 85 per cent. of the product is present as the free radical. The benzophenone-potassium compound gave a similar result; with this substance part of the ketyl was in the form of a fine crystalline suspension. The method of analysis used may overestimate the amount of ketyl present and the coloured solutions probably contain very little of the pinacol. These experiments are being extended to other ketones and other solvents to test this point more completely. Birkbeck College, University of London, Fetter Lane, E.C. 4. Ber., 46, 2840, 1913.