INORGANIC CHEMISTRY.CONSIDERABLE interest has been sustained in the study of chelate and othercomplex compounds formed between metal cations and a variety of ligands.The growing interest in this branch of the subject has in part no doubt beenstimulated by the increasing importance of these complex compounds inseveral industrial processes, and great advances have been made in theirstudy in recent years. Although a complete specificity, such that a givenchelating agent would form a compound with only one type of ion, may yetbe out of reach, the different stabilities of many of these chelates providea valuable tool for the separation and purification of elements by solvent-extraction, ion-exchange, and other fractional techniques. For example,the use of ion-exchange techniques for the separation and purification ofthe lanthanons is now well established, several different eluants having beenused by individual groups of workers.A quantitative comparison has nowbeen made of the relative value as eluants in this connection, of seven carb-oxylic acids, viz., acetic, malic, tartaric, citric, aminoacetic, nitrilotriacetic,and e t h ylenediamine-NNN' N'-tetr a-acet ic, and the stability const ants havebeen determined for the last-named with several lanthanons.2The effects of the nature of the chelating agent and of the electronicstructure of the metal ion on the stability of the complex have been thesubjects of numerous investigations. Thus, the effect of a relatively minorchange in the structure of the chelating agent is shown by a comparison ofthe stabilities of the chelates of the anions of iminodiacetic acid and imino-dipropionic acid with copper(II), nickel(II), cobalt(II), zinc, and cadmiumions: replacement of the acetate groups by p-propionate groups in theligand results in a considerable decrease in the stability of the he late.^A study of the chelates of 2-hydroxyethylaminoacetic acid with a numberof bivalent cations'4 has shown, on the other hand, that the stabilities of thechelates of iminodiacetic acid are greatly increased when an amino-hydrogenatom is replaced by the hydroxyl group, though the stability constants arestill about 100-fold smaller than those of the corresponding nitrilotriaceticacid.The effect on chelate stability, of the structure of the chelating agent hasalso been investigated with the use of 8-hydroxyquinoline and analogousreagents.g The steric effect of 8-hydroxyquinaldine, believed to be re-sponsible for the non-reaction of this reagent with aluminium(III), has alsobeen encountered in the case of nickel(x1).In the course of this work it wasshown that there is an increase of chelate stability in a series of 8-hydroxy-quinoline chelates with bivalent ions of transition elements, as the transitionelectron shell becomes more completely filled.Evidence of the effect of steric considerations in a chelate compound onR. C. Vickery, J., 1952, 4357.S. Chaberek, Jr., and A. E. Martell, J . Amer. Chem. SOC., 1952, 74, 5062.S. Chaberek, Jr., R. C. Courtney, and A.E. Martell, ibid., p. 5057.G. Schwarzenbach, H. Ackermann, and P. Ruckstuhl, Helu. Chzm. Ada, 1949, 33,W. D. Johnston and H. Freiser, J. Amer. Chem. Suc., 1962, 74, 5239.2 Idem, J., 1952, 1895.1175; G. Schwarzenbach and E. Freitag, ibid., 1951, 34, 149282 INORGANIC CHEMISTRY.the reactivity of the central metal atom is also given by the radio-isotopicexchange of zinc between zinc acetate and a number of zinc chelate cokm-pound^.^ There is little or no exchange of the zinc in zinc phthalocyanine-a " fused-ring " type of chelate compound-whereas with others of the" non-fused-ring " type, exchange is rapid.A further study of the effect of steric hindrance has been made with thechelates of copper(x1) and nickel(xx) with a series of N-alkyldiamines,R*NHCH,CH,*NH, (R = H, Me, Et, Pr", Bun, and Pri).8 Except for then-butyl derivative, there is a general decrease in stability of the straight-chainalkyl complexes as the length of the chain increases, and the complexeswith the isopropyl derivative are less stable than those containing thestraight-chain alkyl groups.A rather different steric effect of the organic groupings attached to acentral metal atom is shown by the physicochemical properties of thealkyloxides of titanium and zirconium.It was established in the firstplace that the volatility of the amyloxides of these elements depended onthe structure of the particular amyl group in question, the branched-chaincompounds being in general much more volatile than the straight-chainamyloxides.This effect, it is believed, may be due to the screening by thebranched groups inhibiting intermolecular bonding between the centralmetal atom and oxygen. A number of new tetra-tert.-alkoxides of theseelements have been prepared. lo Ebullioscopic measurements show that theyall monomeric.Differences in the stabilities of complex compounds formed by opticallyactive organic stereoisomers with some complex inorganic compounds offera method of resolution of the racemates. For example, when a racemicmixture of an organic acid reacts with an equimolecular amount of a complexsuch as Zmo-carbonatobispropylenediaminecobalt ( XIX) two isomers areformed, one containing the dextro- and the other the Zmm-form of the acid.As these prove to have different stabilities, a partial resolution of the acidbecomes possible, and experimental results have been.presented for thepartial resolution in this way of tartaric, chloropropionic and lactic acid. l1A considerable interest has been shown recently in directive influences inthe reactions of inorganic co-ordination complexes. In this connection ,mention may be made of a review of the extensive experimental work andtheoretical considerations relating to the so-called tvans-effect, especially inrelation to the platinum complexes.12 The trans-effect stipulates that thebond holding a group trans to an electronegative or other labilising group isweakened thereby, so that the trans-group is the first to be removed in asubstitution reaction.A recent example of this effect is seen in the behaviour of dichlorodi-ethyleneplatinum(xx), (C,H,),PtCl,.In all except the olefin series of plati-num(I1) complexes, a t least one geometric isomer of the simple non-ioniccompound L,PtCl, (L = ligand) is known, and is usually more stable thanthe corresponding ionic complexes (LPtC1,)- and (L,Pt) + 2. The olefinD. C. Atkine, Jr., and C . S. Garner, J . Awzer. Chem. SOC., 1952, 74, 3627.F. Bas010 and R. K. Murmann, ibid., p. 5243.D. C. Bradley, R. C . Mehrotra, and Mi. Wardlaw, J . , 1952, 2027.lo Idem, J., 1952, 4204.l1 A. D. Gott and J . C. Bailar, Jr., J . Amer. Chem. SOC., 1952, 74, 4820.1* J. V. Quagliano and L. Schubert, Chem. Reviews, 1952, 60, 201FAIRBROTHER. 83complexes, however, were found to be an exception, and until recentlyall attempts to prepare (C2H4),PtCl2 had failed, although K(C,H,PtCl,) hasbeen recognised since 1830.14 (C2Hp)2PtC12, which may be obtained as aprecipitate by passing ethylene into dichlorodiethylene-vv'-dichlorodi-platinum(i1) in acetone a t -70°, dissociates a t room temperature with thereversal of the reaction and evolution of ethylene.The instability of thisnew diethylene compound is explained by supposing it to have a trans-configuration, the instability being due to the combined effects of the ratherweak co-ordinating affinity of ethylene and the strong trans-influence, orlabilising effect, of an ethylene molecule on the group in the transpositionto itself.15Further study has been made of the nature of the co-ordinate link incomplex platinum compounds by an examination of the equilibrium betweencis- and trans-bis(triethylphosphine)dichloroplatinum(II) through the meas-urement of the dielectric polarisation of their solutions as a function oftemperature.16 The highly polar cis-isomer is more stable than the trans-by about 10 kcal./mol., as are also the cis-(AsEt,),PtCl, and cis-(SbEt,),PtCl,than their corresponding trans-isomers.l7The influence of the configuration of an ion on its retention by an exchangeresin has been made use of for the separation of the cis- and trans-isomericdinitrotetramminocobalt (111) ions.18An ion-exchange technique has also been used to demonstrate theformation of some anionic complexes of cadmium and copper.lg Forexample, if an anion-exchange resin is pretreated with sodium perchlorate,then cadmium perchlorate is not retained by it and can easily be washed outby water ; on the other hand, if the resin is treated with sodium iodide thencadmium iodide is very strongly held, as would be expected from the well-known stability of the cadmium iodide complexes.In many metal-anion complex ion systems in aqueous solution a numberof different complexes exist simultaneously, the separation or identificationof which is often very difficult, especially if the equilibria are establishedrapidly. On the other hand, if the equilibria are established sufficientlyslowly and the different species have different ionic charges, then ion-exchange methods offer a convenient method of separation.This has nowbeen carried out in the chromic thiocyanate system from which the speciesCr(H,O),+++, Cr( H20) ,(SCN) ++, and Cr (H,O),(SCN),+ have been separated.2oThe prosthetic groups in cation-exchange resins have usually been sul-phonic, carboxylic, or phenolic groups or a combination of these. Resinshave now been developed in which the exchange groups are phosphonousand phosphonic. One interesting feature of these new resins is that theyappear to show a selectivity for sodium over potassium.21Clathrate compounds of oxygen and of nitric oxide in quinol have beenprepared in which some 40--50% of the available spaces are filled.22 Meas-l3 J. Chatt and R. G. Wilkins, Nature, 1950, 165, 859.l4 Zeise, Mag.Pharm., 1830, 35, 105.l6 Idem, J . , 1952, 273.l8 E. L. King and R. R. Walters, J . Amer. Chenz. Soc., 1952, 74, 4471.2o E. L. King and E. B. Dismukes, J . Amer. Chem. Soc., 1952, 74, 1674.21 J. I . Bregman and Y . Murata, ibid., p. 1868.z2 D. F. Evans and R. E. Richards, J . , 1952, 3295.1 5 T. Chatt and R. G. Wilkins, J., 1952,2622.Tdem, J . , 1952, 4300.I. Lenden, Svensk Kem. Tidskr., 1952, 64, 145; Chew Abs., 1952, 46, 856384 INORGANIC CHEMISTRY.urement of the magnetic susceptibility suggests that the paramagnetismof a gas is but little affected by its inclusion in the cage-like structure. It hasbeen pointed out that, in the case of nitric oxide, the clathrate compoundoffers a method of studying its properties a t low temperatures without thecomplications either of change of state or of dimerisation into diamagneticdinitrogen dioxide which occurs on liquefaction at 121" K.A remarkable and unique new iron compound, dicyclopentadienyliron,the first compound to be prepared which contains only carbon, hydrogen,and iron, has been described almost simultaneously by two groups ofworkers.23 I t is a stable orange diamagnetic compound, m.p. 1726-173",which vapourises above 100" without decomposition into a monomeric,undissociated vapour which obeys the perfect-gas laws evena t 400".24 A number of its chemical and physicochemicalproperties have been studied. On the one hand, it is readilyoxidised to a blue cation Fe(C,H,),+ 25 and, on the other, itundergoes reactions like those of aromatic hydrocarbons.26Both chemical evidence 25 and direct X-ray examination 27 ofsingle crystals indicate that it possesses a pentagonal anti-prismatic structure (Fig.l), there being no evidence ofrotation of the cyclopentadienyl groups.The corresponding ruthenium compound, (C5H5)2R~, hasFIG. 1. also been prepared.27u The names ferrocene and ruthenocenehave been suggested for these compounds in view of theirbehaviour as aromatic systems and " ferricinium " and " ruthenicinium "for their unipositive ions, the methods of preparation and properties ofthese compounds and their salts suggesting that in the neutral compound themetal is in the +2 oxidation state, and in the unipositive ion in the +3 state.In a similar manner, the reaction of cobaltic acetylacetonate with cyclo-pentadienylmagnesium bromide gives the unipositive ion [ (C,H5)&o]+ ofwhich several salts have been prepared.This ion, however, cannot bereduced to a neutral c~baltocene.~~~The discovery of these compounds opens up a new field in the borderlandhetween inorganic and organic chemistry.Group 1.-One of the major difficulties encountered in the laboratorypreparation of lithium aluminium hydride by the usual method has beenthe necessity for the troublesome pulverisation of the lithium hydride. Itis reported, however, that by using aluminium bromide instead of thechloride, coarsely powdered lithium hydride can be used.28What is stated to be the first example of a coloured double hydride,namely, AgAlH,, is obtained as a yellow-gold precipitate when etherealsolutions of lithium aluminium hydride and silver perchlorate are shakentogether at -80" : the compound decomposes a t -50".29,23 T.J. Kealy and P. L. Pauson, Nature, 1951, 168, 1039; S. A. Miller, J. A. Teb-z4 L. Ka lan, W. L. Kester, and J. J. Katz, J . Amer. Chem. SOC., 1952, 74, 5531.25 G. Wiykinson, M. Rosenblum, M. C. Whiting, and R. B. Woodward, ibid., p. 2125.26 R. B. Woodward, M. Rosenblum, and M. C. Whiting, ibid., p. 3459.z7 P. F. Eiland and R. Pepinsky, ibid., p. 4971.27a G. Wilkinson, ibid., p. 6146. 27b I d e m , ibid., p. 6148.28 E. Wiberg and M. Schmidt, 2. Naturforsch., 1952, 7, b, 59.29 E. Wiberg and W. Henle, ibid., p. 250.both, and J. F. Tremaine, J., 1952, 632FAIRBROTHEK.85A new method of preparation of lithium hydroperoxide monohydrate hasbeen described : dehydration of this compound over phosphoric oxide in adesiccator at 20 mm. gives almost pure lithium peroxide. X-Ray diffractionstudies indicate that neither the anhydrous hydroperoxide Li0,H nor theperoxycarbonate Li,CO, exists at room temperatures3*An experimental study of the ternary system NaBr-Nac1-6.6~-NaOHhas shown that, contrary to some theoretical predictions, these two halidesdo in fact form a continuous series of solid soIutions.31 It has also beenshown, by X-ray examination, that when RbCl and KBr, or RbBr and KC1,are melted together in any proportion, a single solid solution is obtainedwhich contains all the ions in the original mixture.32 Continuous series ofsolid solutions have also been shown to be formed in the systems (NH,),SO,-Cs,SO,-H,O, (NH,),SO,-K,SO,-H,O, and (NH,),S04-Rb,S0,.33 On theother hand, an examination of the system NH,F-NaF-H,O at 25" hasshown that the system is a simple one, with ammonium fluoride and sodiumfluoride as solid phases.34The various and widely used sodium metaphosphate polymers havecommonly been prepared by thermal dehydration of sodium dihydrogenorthophosphate.It has now been found that the hydration of a-phos-phorus(v) oxide a t a temperature of 15" or below produces chiefly tetrameta-phosphoric acid (H,P,O,,), which offers an alternative method of preparationof the sodium salt of this acid.35Czsium hexasulphide has been prepared by disproportionation ofCs,S,,H,O in aqueous ethyl alcohol into this and one or more lower poly-sulphide ions.The crystal structures of Cs,S, and of the dehydrated Cs,S,prove to be very similar, the polysulphide ions being in the form of non-branched, non-planar, sulphur chains.36A new method of preparation of copper hydride, CuH, has been de-scribed.37 When lithium aluminium hydride in pyridine-ether solution isadded to a solution of copper(1) iodide in pyridine, the following reactiontakes place : 4CuI + LiAlH, = LiI + AlH, + 4CuH, and a blood-red solu-tion is obtained from which the solid may be precipitated as bright red-brown needles on addition of more ether. The dry hydride is stable up toabout 60°, above which it decomposes (rapidly at 100"); in water it givesCuOH and hydrogen a t a somewhat lower temperature.The existence of the chlorocuprate ion CuC1,' in aqueous solution is wellknown, but until recently evidence of its existence in the solid state has beenlacking.X-Ray examination of the yellowish-orange Cs,CuCl, has nowdemonstrated the presence of this ion in the solid.38 In contrast, however,to the more usual planar configuration of quadricovalent copper(I1) ions,the chlorine atoms appear to be arranged about the copper in the form of aflattened tetrahedron : the absorption spectrum of the solid is also markedly30 A. J. Cohen, J . Amer. Chem. SOC., 1952, 74, 3762.31 E. L. Simons, C. A. Orlick, and P. A. Vaughan, ibid., p. 5264.32 L. J. Wood and L. J. Breithaupt, Jr., ibid., p.727.33 C. Calvo and E. L. Simons, ibid., p. 1202.34 H. M. Haendler and A. Clow, ibid., p. 1843.35 R. N. Bell, L. F. Audrieth, and 0. F. Hill, Ind. Eng. Chenz., 1952, 44, 568.36 S. C. Abrahams, E. Grison, and J . Kalnajs, J . Anzer. Chem. SOC., 1952, 74, 3701.37 E. Wiberg and W. Henle, 2. Natwrforsch., 1952, 7, b, 250.38 L. Helmholz and R. F. Kruh, J . Amer. C h e m SOC., 1952, 74, 117686 IKORGANIC CHEMISTRY.different from that of the aqueous solution. It would appear therefore thatthe structure of the aqueous chlorocuprate ion is different from that of theion in the crystalline state.Some new copper( I) complexes of methyldiphenylarsine have beenprepared. Four types, containing respectively 1, 2, 3, and 4 molecules ofthe tertiary arsine to each molecule of the copper(1) halide, have beenisolated : example of the last two have not been reported previously.Thosecomplexes with an empirical formula CuX,3AsMePh2 are non-electrolytesand contain quadricovalent copper. Those with an empirical formulaCuX,4AsMePh2 (X = I, CIO,, or NO,) are salts of the univalent anion X.3yA number of complex compounds of silver( I) with polydentate chelatingagents have been studied which indicate that the co-ordination number ofsilver is four in these complexes. When aqueous or alcoholic solutions ofsilver nitrate are allowed to react with triethylenetetramine (trien), severalcomplexes are formed : [Ag trien]+ and [Ag, trien]++ are present in bothaqueous and alcoholic solutions, and [Ag trien]NO, has been isolated fromthe latter.4*It has been shown that the compound of empirical formula (C7H7!,SAuBr,probably contains the gold in both univalent and tervalent states : it cannotbe formulated as a dimer, as a salt, or as a compound of bivalent gold.41An X-ray investigation of the corresponding chloride, obtained by theinteraction of solutions of " auric chloride " and benzyl sulphide, has shownthat the crystal consists of separate molecules of benzyl sulphide-mono-chlorogold(1) , (C,H,),S -+ AuCl, and benzyl sulphide-trichlorogold(m),(C,H,),S -+ AuCl,, so arranged that a highly disordered structure results.A study has been made of the preparation and properties of a number ofgold imides, of the type MTIAuX(imido),] (X = halogen) from succinimideand of the type M1[AuX,(imido),] from o-benzoicsulphonimide (saccharin)and ~hthalimide.~ZGroup 11.-It has been known for some time that the volatility of anumber of oxides at high temperatures is greatly increased by the presenceof water vapour, a result believed to be due to some reaction between theoxide and water or between the oxide and a decomposition product of water.A study of this increase in the case of beryllium oxide at 1200-1600" hasshown that it is due to the reaction BeO,,) + H,O,, -+ Be(OH),cgl.43The dialkyl derivatives of beryllium present some interesting structuraland valency problems arising from the fact that they are strongly electron-deficient compounds.A detailed study in this connection has been made ofdimethylberyllium and a number of its co-ordination compounds.In thevapour state it appears to polymerise to some extent to the dimer and tri-rner,& and it forms co-ordination compounds with trimethylamine, trimethyl-phosphine, and dimethyl and diethyl ether, but not with trimethylarsine ordimethyl sulphide.45 The properties of these compounds indicate that theR. S. Nyholm, J., 1952, 1257.40 H. B. Jonassen and P. C . Yates, J . Amer. Ckem. SOC., 1952, 74, 3388.41 F. H. Brain, (the late).C. S. Gibson, J. A. J. Jarvis, R. F. Phillips, H. M. Powell,4a A. M. Tyabji and (the late) C. S. Gibson, J., 1952, 450.43 L. Grossweiner and R. L. Seifert, J . Amer. Chem. SOC., 1952, 74, 2701.44 G. E. Coates and N. D. Huck, J., 1952, 4496.4 5 Idem, J ., 1952, 4501.and A. Tyabji, J . , -1952, 3686FAIRBROTHER. 87order of stability is N > I? > 0. I t also reacts with methylamine, di-methylamine, dimethylphosphine, methanol, methanethiol, or hydrogenchloride, but in these cases the products are methane and di-, tri-, or poly-meric products. Dimethylamine gives a trimeric compound (MeBe,NMe,),to which a cyclic structure has been assigned.46It is reported that pure magnesium cyanide can be prepared by thepassage of hydrogen cyanide over a specially prepared magnesium oxide at730" : the latter is obtained by ignition of magnesium oxalate at 600°,samples of magnesium oxide prepared by other reactions giving only impureproducts.47As part of a systematic investigation of isomorphous replacement inhydrated salts, the systems CdC1,-MClz-H,O (M = Mg, Mn, FeII, CuII,and Ca) have now been examined.The existence has been established of thedouble salts 2CdC1,,MgC12, 12H,O; CdC1,,2MgC12, 12H,O; 4CdC1,,MnCL2, 10H,O;CdC1,,CuCl2,4H,O ; and of a number of solid solutions.48Solubility isotherms are reported for the quaternary system Ba(ClO,),-BaBr2-Ba(NO,),-H,O at 10" and for the ternary systems Ba(C10,),-BaBr,-H,O at 10" and 25", Ba(ClO,),-Ba(NO,),-H,O a t lo", 2 5 O , and 45", andBaBr,-Ba(NO,),-H,O a t 10" and 25" : also for a number of other aqueousternary systems involving barium chlorate, bromate, and iodate.49The solubility of metallic cadmium in fused mixtures of cadmium chloridewith some other chlorides has been studied.50 As a result of this and previouswork it emerges that : (1) the solubility of Group I1 metals in their fusedchlorides increases as the cation radius increases, (2) the solubility (for Cdin CdCl,) is decreased by the addition of the chloride of an electropositiveelement, and (3) the solubility increases as the ratio of the number of anionsto cations increases.The last result, e.g., the effect of CeC1, as comparedwith KCI, has been interpreted as due to the process of dissolution consistingin the entry of the metal atoms into the octahedral holes of an almost close-packed fused chloride structure, these holes being clearly relatively moreabundant when less of them are already occupied by the metal cations.A re-investigation of the X-ray diffraction patterns of the compoundsformed in the reactions between mercurous chloride and ammonia hasdisproved the formation of several mercury(1) compounds which have beensuggested by previous workers and has shown that only the well-knownmercury(I1) compounds Hg(NH,),Cl,, HgNH,Cl, and Hg,NCl,H,O areformed in addition to mer~ury.~lGroup 111.-A study of the method of preparation of diborane by thereaction of lithium aluminium hydride with boron trifluoride in etherealsolution has shown that the reaction takes place by a t least two successivestages, involving the intermediate formation of lithium borohydride :(a) LiAlH, + BF, = LiBH, + AlF,; (b) SLiBH, + BF3 = 2B,H, + 3LiF.52Diborane can also be prepared by reaction of lithium hydride withthe boron trifluoride-ethyl ether complex.An investigation of this reaction4 6 G. E. Coates, F. Glockling, and N. D. Huck, J., 1952, 4512.4 7 H. Hartmann and H. Narten, 2. anovg. Chem., 1952, 267, 37.I s H. Bassett and R. N. C. Strain, J., 1952, 1795.49 J . E. Ricci and A. J . Freedman, J . Aunev. Chem. SOC., 1952, 74, 1765, 1769.6o D. Cubicciotti, ibid., p. 1198.51 L. Nijssen and W. N. Lipscomb, ibid., p. 2113.52 I. Shapiro, H. G. Weiss, M. Schmich, S. Skolnik, and G. B. L. Smith, ibid., p. 90188 INORGANIC CHEMISTRY.under various conditions has shown that it can proceed by two differentcourses : (a) 6LiH + 8BF3 = B,H, +. GLiBF, in the absence of promoters,and (b) 6LiH + 2BF3 = B,H, + 6L1F if small amounts of the ether-soluble active-hydrogen-containing promoters lithium borohydride andlithium trimethoxyborohydride are present.This offers an explanation ofthe experimental observation that higher yields of diborane by this reactionare obtained in the presence of methyl ortho-borate.53The preparation of a new type of substituted borohydride containinga quaternary ammonium cation has been described. Tetramethyl-,tetraethyl-, and benzyltrimethyl-ammonium borohydrides have beenprepared by metathetical reactions in aqueous solution between sodium orlithium borohydride and the respective quaternary salts or hydroxides.A number of deuterated boron compounds have been prepared andcharacterised, including deuterated diborane, borine carbonyl, and dimethyl-aminodiborane. 55A cryoscopic study of mixtures of boron trifluoride and 100 yo nitricacid has revealed the existence of the solid compound HN03,2BF3, m.p.53'. If boron trifluoride is passed into 100% nitric acid considerable heatis evolved and a water-clear viscous liquid is formed from which largeprismatic crystals separate if the boron trifluoride content is sufficientlyhigh: if the composition of the mixture approaches that of the doublecompound, the whole mass eventually solidifies56When boron trichloride is passed into 1 : 4-dioxan at 20°, in an otherwiseevacuated apparatus, a stable, crystalline, 1 : l-addition compoundC4H,02,BC1, is formed. It is noteworthy that such co-ordination with oneof the oxygen atoms of dioxan apparently inhibits the other oxygen fromacting in like manner, since excess of boron trichloride does not give the2 : l-compound 2BC1,,C,H,02 either with dioxan itself or with the 1 : 1-complex.57Methyl orthoborate forms 1 : l-addition products with mono-, di-, andtri-methylamine, B(OMe),,NH,Me,,. 58A further confirmation of the analogy between benzene and borazole(B3N3H,) is given by the X-ray examination of single crystals of p-trichloro-borazole which have been shown to possess a molecular configuration closelyresembling that of 1 : 3 : 5-trichloroben~ene.~~The sodium salt of a new boron base, Na,HBMe,, has been prepared bythe action of sodium in liquid ammonia a t -78" on tetramethyldiboraneB,H2Me, which is split equally into Me,BH,NH3 and this new sodiumsalt. 6oAn additional contribution to the chemistry of covalent boron-nitrogencompounds has been made by the study of the chemical and physical pro-perties of the monomeric and dimeric dimethylaminoboron dichloride 6153 J.R. Elliott, E. M. Boldebuck, and G. F. Roedel, J . Arne.,. Chem. Soc., 1953, 74,5047.65 A. B. Burg, ibid., p. 1340.5 6 €3. Gerding, P. M. Heertjes, L. J. Revallier, and J. W. M. Steeman, Rec. Trav. chim.,67 A. K. Holliday and J. Sowler, J . , 1952, 11.~5r1 D. L. Coursen and J. L. Hoard, J . Amer. Chem. Soc., 1952, 74, 1742.61 C. A. Brown and R. C. Osthoff, ibid., p. 2340.64 M. D. Banus, R. W. Bragdon, and T. R. P. Gibb, Jr., ibid., p. 2346.1952, 71, 501.J. Goubeau and R. Link, 2. anorg. Chem., 1952, 267, 27.A. B. Burg and G. W. Campbell, Jr., ibid., p.3744FAIRB ROTHE R. 89and of diethylaminoboron dichloride which, in contrast to the dimethylMe,N-BCl, compound, does not dimerise. 62 When dimethylaminoborondichloride is kept at room temperature for several days, it isC1zB-NMe2 converted into the crystalline dimer [Me,NBCl,],, the smalldipole moment of which confirms the cyclic structure (inset) suggested byprevious workers.Improved methods of preparation of dimethylaminoboron dichloride andof the new difluoride (which also dimerises on storage) have been described :good yields of the dichloride are obtained by the dehydrogenation of thedimethylamine-boron trichloride complex by triethylamine in benzenesolution. The difluoride is prepared by refluxing the dimethylamine-borontrifluoride complex at 240-290", whereby both the dimer of dimethyl-amine-boron difluoride and the disproportionation products, trimethylamine-boron trifluoride and tetramethylammonium fluaroborate, are obtained.63The structures of dimethylaminodiborane Me,NB,H5 and aminodiboraneH,NB,H5 have been studied by electron diffraction and give furtherevidence for the bridged structure of the parent diborane.The results ofthis investigation are in accord with symmetrical structures which have4-fold co-ordination about the nitrogen atom and may be regarded asderived from diborane by the replacement of one of the bridge hydrogensby NMe, and NH, re~pectively.~~It is reported that when an electrodeless discharge is passed throughaluminium tri-iodide vapour at low pressures, in a vessel kept a t 50°, thetri-iodide undergoes decomposition with formation of a buff-colouredrn~noiodide.~~ It has also been claimed that potentiometric titration ofliquid-ammonia solutions of aluminium tri-iodide with solutions of potassiumin the same solvent gives evidence of the existence of Al++ and Al'ions.66It had been recorded 67 that when aluminium and other members of thisgroup are anodically oxidised in an electrolyte of ammonium or sodiumacetate in anhydrous acetic acid the mean oxidation state of the cationsformed was always appreciably lower than 3.In contrast to this observationit has now been shown that when the electrolytic oxidation is carried out inliquid ammonia, in a variety of electrolytes, initial valency numbers lowerthan 3 are only observed when the solution contains nitrate ions8It has long been known that hydrochloric acid will dissolve more alumin-ium than corresponds to the equation : 2A1 + 6HC1 = Al,Cl, + 3H,.When a large excess of aluminium is boiled with dilute hydrochloric acid,almost 6 equivalents of aluminium are dissolved to give a clear solutionfrom which the 5/6 basic chloride Al,(OH),Cl can be precipitated by additionof sodium or calcium chloride.69A clarification of the various solid phases formed in equilibrium with anaqueous or sulphuric acid solution of aluminium sulphate has been carried+ +62 R.C. Osthoff and C. A. Brown, J . Amer. Chem. SOC., 1952, 74, 2378.63 J. F. Brown, Jr., ibid., p. 1219.64 K.Hedberg and A. J. Stosick, ibid., p. 954.6 5 W. C. Schumb and H. H. Rogers, ibid., 1951, 73, 5806.6 6 G. W. Watt, J. L. Hall, and G. R. Choppin, ibid., 1952, 74, 5920.6 7 A. W. Dsvidson and F. Jirik, ibid., 1950, 73, 1700.6M W. E. Bennett, A. W. Davidson, and J. Kleinberg, ibid., 1952, 74, 732.6B G. Denk and L. Bauer, 2. anorg. Chew., 1952, 267, 8990 INORGANIC CHEMISTRY.out by a re-examination, at temperatures from 25" to 60°, of the systemAluminium hypophosphite, Al(H,PO,),, has been obtained as an anhydrouscrystalline precipitate by heating aluminium hydroxide, or a solution of analuminium salt, with 50% hypophosphorous acid at 80-90" for one hour,the precipitation being rather s ~ o w . ~ ~At -80" anhydrous aluminium chloride neither reacts with toluene nordissolves appreciably in it.On the addition of excess of anhydrous hydrogenchloride, however, it dissolves reversibly to give a clear, brilliantly greensolution in which one mole of hydrogen chloride is apparently taken upper mole of AlC13 which goes into solution. At -45.4" the amount ofhydrogen chloride taken up corresponds to one mole per mole of Al,Cl,.The reaction is believed to involve the formation of a carbonium cation[CH3*C,H6]+ and AlC14- or Al,C17- anions7,An X-ray and electron-diffraction study has been made of the poly-morphism of Ga203 and of the structure of gallia gels.73An examination of the In-In$, system by thermal metallographic andX-ray analysis has given evidence of the existence of four definite compounds,In2S,, (In3!&), (In&,), and 'Ins : the parentheses indicate some uncertaintywith regard to composition. No evidence was found for the existence ofthe sulphide of univalent indium In,S.74Insoluble bistripyridylindium chloride, bromide, and thiocyanate areformed by treating the corresponding indium salts with 2 : 2' : 2"-tripyridyl[2 : 6-di-(2-pyridyl)pyridine] in dilute aqueous ethanol solution.In asimilar manner, 5-nitro-1 : 10-phenanthroline gives the complexes tri-(5-nitro-1 : 10-phenanthro1ino)indium chloride, bromide, iodide, and thio-cyanate-all relatively insoluble. 75Thallium(Ir1) ions react with alkali alkyl xanthates to form the yellowxanthates (RO*CS,),Tl, which are insoluble in water but soluble in ethanoland other organic solvents.Aqueous or dilute ethanol solutions of 1 : 10-phenanthroline and 2 : 2'-dipyridyl react with thallium (111) ions to form the in-soluble chlorides, bromides, and thiocyanates, Tl(phenan)X, and Tl(dipy)X,,respectively. Thallium(m) iodide, on the other hand, co-ordinates with twomolecules of the chelating agent to form Tl(phenan),I, and Tl(dipy),I,. Itis reported that the latter compounds are so insoluble that they permit thedetection of thallium(II1) ions in a dilution of 1 in lo6 in the presence ofiron. 76A new and convenient method of preparation of thallium(II1) iodide hasbeen described which consists in the dissolution of thallous iodide in a solutionof iodine in concentrated hydriodic acid followed by evaporation to constantweight a t room temperature in vacuo over silica gel.It may be noted thatthe solution yields TlI, and not the anhydrous hypothetical acid HTlI,.The only intermediate phase in the thermal decomposition of TlI, to TI1and I, is Tl,I,.77A study has been made of a number of aqueous ternary and quaternary70 D. Taylor and H. Bassett, J., 1952, 4431. 71 D. A. Everest, J., 1952, 2945.72 H. C. Brown and H. W. Pearsall, J . Amer. Chem. Soc., 1952, 74, 191.73 R. Roy, V. G. Hill, and E. F. Osborn, ibid., p. 719.74 M. F. Stubbs, J. A. Schufle, A. J. Thompson, and J. M. Duncan, ibid., p. 1441.7 5 G. J. Sutton, AustraE. J . Sci. Res., 1951, 4, A , 651.76 Idem, zbid., p. 654.Al,(SO4),-H,SO~-H,O. 70.77 A. G. Sharpe, J., 1952, 2165FAIRBROTHER. 91systems involving thallous, ammonium, potassium, and cupric sulphatesat 25" in order to compare the behaviour of thallous salts with those of thecorresponding silver and alkali-metal salts.Despite certain well-knownsimilarities between argentous and thallous salts, it is found that Tl,SO,forms a continuous series of solid solutions a t room temperature with(NH,),SO, and K2S04, whereas Ag,SO, does not form solid solutions witheither of these salts. Thallous sulphate also forms the double saltT1,S0,,CuS04,6H,0, which is isomorphous with the corresponding doublesalts (NH,),S04,CuS04,6H,0 and K,S04,CuS04,6H,0. On the other hand,thallous sulphate does not form a solid solution with sodium sulphate at 25"or 46", whereas argentous salts form solid solutions with a variety of corre-sponding sodium salts, including the sulphate.78It is reported 79 that thallous sulphide Tl,S, free from the oxidationproducts which usually accompany it when it is precipitated in aqueoussolution, may be obtained by the action of dry hydrogen sulphide on analcoholic solution of thallous ethoxide.The well-known fact that metals usually exhibit theirhighest oxidation states as fluorides has suggested that higher oxidationstates of praseodymium and neodymium might be obtained by the use ofchlorine trifluoride or bromine trifluoride as fluorinating agents. WithClF, both Pr,O,, and Pro,, as well as the trioxides Pr,O, and Nd2O3, how-ever, yielded only the trifluorides : BrF, proved to be relatively inactivetowards the freshly ignited oxides.*OThe occurrence of the lanthanon elements as uranium-fission productshas stimulated interest in the properties of the pure metals themselves.Kilogram quantities of lanthanum and cerium, and somewhat less of praseo-dymium and neodymium, have been prepared by the reduction of theanhydrous chlorides by calcium in refractory-oxide lined crucibles, thereaction being initiated by the exothermic reaction between the calciumand a trace of added iodine.81 A technique for the preparation of smallerquantities (40 g.) of the highly pure metals has also been developed whichalso consists in the reduction of the anhydrous chlorides by calcium, thereaction being carried out, however, in tantalum crucibles and the resultingmetal vacuum-cast in tantalum containers.Pure lanthanum, cerium,praseodymium, neodymium, and gadolinium have been prepared by thismethod. It is noteworthy that samarium, europium, and ytterbium, thoselanthanons which exhibit stable bivalent oxidation states, are only reducedto the bivalent chlorides by this reaction, which can therefore be used fortheir rernoval.g2A good yield (85%) of the mixed anhydrous lanthanon chlorides has beenobtained by the direct chlorination of a mixture of monazite sand andcarbon a t 900". At this temperature the majority of impurities form volatileproducts, and the mixed chlorides can be drained away from the reactionmixture.83Lanthanons.'Is J. E. Ricci and J. Fischer, J . Amer. Chem. SOC., 1952, 74, 1443, 1607.'@ B. Reuter and A. Goebel, 2. anorg. Chem.., 1952, 268, 101.A. I. Popov and G. Glockler, J . Amer. Chem. SOC., 1952, 74, 1357.F. H. Spedding, H. A. Wilhelm, W. H. Keller, D. H. Ahmann, A. H. Daane,C. C. Hach, and R. P. Ericson, Ind. Eng. Chem., 1952, 44, 553.B* F. H. Spedding and A. H. Daane, J . Amer. Chem. SOC., 1952, 74, 2783.88 F. R. Hartley, J . Appl. Chem., 1952, 2, 2492 INORGANIC CHEMISTRY.The preparations of cerous ammonium acetylacetonate, m. p. 143-144"(decomp.), and of praseodymium ammonium acetylacetonate, m. p. 145",have been described. These compounds are only very slightly soluble(-1 mg. /ml.) in carbon tetrachloride, acetone, and benzene, and are insolublein light petroleum, hexane, and isooctane.84Group 1V.-The preparation and crystal structures of a further numberof disilicides of the lanthanon elements have been reported, namely, LaSi,,CeSi,, PrSi,, NdSi,, SmSi,, and of YSi,.85The reactions between carbon, silicon, and germanium tetrafluoridesand Al,Cl,, A1,Br6, A1216, MgCl,, CaCl,, and BaC1, have been investigated.86At high temperatures the reaction between SiF, and Al,Cl, goes to com-pletion, but a t lower temperatures a mixture of SiCl,, SiFCl,, SiF,Cl,, andSiF,C1 is obtained: similar results are obtained when Na,SiF, is sub-stituted for SiF,, and it is possible to prepare SiCl,, SiBr,, and SiI, by thereaction of SiF, or Na,SiF, and aluminium in the presence of the appropriatehalogen.Interest has revived, largely in connection with its technical applications,in the chemistry of the so-called silicon oxyhydride, the solid, highly cross-linked, polymeric hydrolysis product, with an empirical formula SiHO,,which is obtained as a precipitate when a benzene solution of SiHCl, ispoured into cold water or when the trichlorosilane vapour is hydrolysed at450" with steam.As might be inferred from its formula, it has very strongreducing properties, and when heated it gives hydrogen and silicon sesqui-oxide Si203 : ZSiHOa = Si,O, + H,.B7Evidence has been presented 88 to show that the hydrated sodiumsilicates, commonly formulated as Na,SiO,,S, 6, and 5H30, andNa,Si,O,,llH,O are in fact acid salts of orthosilicic acid and their hydrates,viz., N+[H,SiO,], 8, 5, and 4H,O ; Na3[HSi0,],5H,0.Convenient methods have been described for the preparation, as a con-tinuous process, of monosilane by the reduction of silicon tetrachloride withlithium aluminium hydride and the subsequent alkylation and alkoxylationof the silane by reaction of the monosilane with the appropriate organo-lithium compound : tetraphenyl-, tetraethyl-, triethyl-, diethyl-, triiso-propyl-, tetra-Z-naphthyl-, and tri-l-naphthyl-silane have been prepared inthis manner.Phenylsodium reacts with silane to give tetraphenylsilane andsodium hydride. Silane has been found to react with various alcohols inthe presence of alkoxide ions to give tetra-alkoxysilanes, (RO),Si, andhydrogen. 89A number of new alkylgermanium and alkylsilicon compounds havebeen prepared and their reactions studied,g0 and further work has beencarried out on the synthesis of dialkylamino-germanes and -~ilanes.~lTrichlorodimethylaminosilane and chlorobisdimethylaminosilane havebeen prepared from dimethylamine and silicon tetrachloride.Both com-a* J . R. Seehof, J. Amer. Chem. SOC., 1952, 74, 3960, 3961.8 5 G. Brauer and H. Haag, 2. anorg. Chem., 1952, 267, 198.8 6 M. Schmeisser and H. Jenker, 2. Naturforsch., 1952, 7, b, 191; W. C. Schumb87 G. H. Wagner and A. N. Pines, Ind. Eng. Chem., 1952, 44, 321.8 6 E. Thilo and W. Miedreich, 2. anorg. Chem., 1952, 267, 76.90 H. H. Anderson, J . Amer. Chem. SOC., 1951, 73, 5798, 5800, 5802, 5804.O1 Idem, ibid., 1952, 74, 1421..and D. W. Breck, J. Amer. Chem. SOC., 1952, 74, 1754.J. S. Peake, W. H.Nebergall, and Y. T. Chen, J. Amer. Chem. SOC., 1952, 74, 1526FAIKBROTHER. 93pounds give silane instead of the expected dimethylaminosilane on reductionwith lithium aluminium hydride. They are sufficiently basic to form severalhydrochlorides but do not form quaternary salts with methyl iodide.92Several methods have been described for the preparation of the newcompounds TiCl,,CH,CO,Et ; TiC13*OEt,CH,*C0,Et ; andTiC1,*OPri,CH,*C02Et ; these can be distilled under reduced pressurewithout change in composition, which suggests the possibility of a quinque-covalent titanium. A new series of titanium trichloride monoalkoxidesTiC1,mOR (R = Me, Et, Pri, and Bun) has also been obtained by the rapidradical-interchange reaction between titanium tetrachloride and the appro-priate tetra-alk~xide.~,The preparation of a number of new alkoxides of zirconium and hafniumand the effect of molecular complexity on their properties has already beenmentioned.loWhen a solution of zirconyl chloride, ZrOC1,,6H20, in alcoholic hydrogenchloride is treated with pyridine, a quantitative yield of pyridinium chloro-zirconate (C,H,N),ZrCI, is obtained. This compound forrns a very con-venient starting material for the preparation of the zirconium alkoxidesZr(OR), (where R = Et, Pri, Bus, and Bun), which have been obtained in apure state by passing ammonia into a suspension of (C5H,N),ZrC1, in amixture of benzene and the appropriate alcohol.94A study of the thermal decomposition of ammonium heptafluorozirconatehas shown that the decomposition takes place in three distinct stages, viz.,(NH4),ZrF, 4 (NH,),ZrF, -+ NH,ZrF6 -+ ZrF,.The decompositiontemperatures for the three successive decompositions are a function of thepressure : at 760 mm. they are 297", 357", and 410°, respecti~ely.~~Details have been given of a method of purification of zirconium fromcommon impurities, and especially iron, by the alternate precipitation ofZr(S0,),,4H20 by the addition of concentrated sulphuric acid to an aqueouszirconium sulphate solution in the presence of hydrochloric acid (essentialfor the removal of iron) and the dissolution of the precipitate in wateragain .96The complexes formed by zirconium( ~ v ) with 2-nitroso-l-naphthol havebeen investigated in aqueous ethanol and aqueous dioxan solutions, in theabsence and in the presence of HC10,.In the absence of HC10, a 1 : 1complex is formed, whereas if the solution is 3~ or stronger with respect toHClO, and 3 x ~O-,M in Zr(Iv), a 1 : 4 complex is formed.g7A detailed investigation has been carried out on the chemistry of zir-conium in nitric, hydrochloric, perchloric, and sulphuric acid solutions by acombination of ion-exchange, radiochemical, and other techniques. Theresults indicate the presence at lower acid concentrations' of a variety ofpolynuclear hydrolysis products, and at higher acidities of such complexes(in nitric acid) as [Zr(NO,),H,O),]++, [Zr(OH),(NO,) (H20),]+,[Zr(OH) ,(NO,) 2(H,O) 2]++, and [zr( OH) ,(NO,) 41 * -989a R. Cass and G.E. hates, J., 1952, 2347.O3 D. C. Bradley, D. C. Hancock, and W. Wardlaw, J., 1952, 2773.94 D. C. Bradley, F. M. Abd-el-Halim, E. A. Sadek, and W. Wardlaw, J., 1952, 2032.96 H. M. Haendler, C. M. Wheeler, Jr., and D. W. Robinson, J . Anzer. Chem. SOC.,97 H. B. Jonassen and W. R. de Monsabert, ibid., p. 5298.1952, 74, 2352. 9 6 W. S. Clabaugh and R. Gilchrist, ibid., p. 2104.B. A. J. Lister and (Miss) L. A. McDonald, J . , 1952, 431594 INORGANIC CHEMISTRY.A process has been developed for the concentration of hafnium from amixture of zirconium and hafnium in which it occurs only to the extentof 2% by weight up to about 90% in two cycles. Each cycle consists of theadsorption of the mixed zirconium and hafnium chlorides in methanolsolution on activated silica gel, followed by extraction with 1.W-anhydroushydrogen chloride in methanol (which preferentially removes the zirconium),followed by 7~-sulphuric acid which removes the remaining ad~orbate.~gThis follows the earlier experiments reported lo0 that silica gel adsorbshafnium in strong preference to zirconium from a methanol solution of thetetrachlorides.It has long been known l01 that zirconium and hafnium tetrachloridesform addition products with phosphorus pentachloride and phosphorusoxychloride which, since they can be distilled, are of interest in connectionwith the separation of the metals. From this point of view a wider studyhas now been made of the reactions of these tetrahalides with phosphorusoxychloride and the oxyfluorides POFCl,, POF,CI, and POF, ; lo2 withPOC1, and POFC1, the tetrahalides give addition products with the com-position 2POX3,MC14, which decompose when heated under reduced pressure(< 0-1 mm.) to give the 1 : 1 addition products POX,,MCl,. With POF,Cland POF,, however, the metal tetrahalides give only POX,,MCl, at roomtemperature.The introduction of fluorine not only decreases the thermalstability of the addition products but also introduces -the possibility ofhalogen exchange. If the tetrahalides are kept sufficiently long in contactwith excess of POFCl, or POF,Cl, complete halogen exchange takes placewithin the phosphoryl molecule and 2POC13,MC14 crystallises from solution,whilst i f the 1 : 1 addition products with POF,Cl or POF, are heated, somefluorination of the metal halide takes place.Whilst an f electron in thorium is not essential to the general actinidehypothesis, the existence of a tervalent fluoride isostructural with uranium(Ir1)fluoride might be expected on the basis of an electron in the 5f orbitalparalleling cerium(zI1) fluoride with an electron in the 4f orbital.In con-trast, however, to the preparation of the other lower halides lo3 by reductionof the tetrahalides, especially the iodide, by the metal, a number of attempts,by a variety of methods, to reduce ThF? by the metal a t temperatures upto 1600" have failed. There was some indication, however, that thoriumdissolves slightly in its tetrafluoride a t high temperatures as do a number ofother metals in their molten halides.lMThermal, metallographic, X-ray, and chemical analysis have demon-strated the existence of four definite phases in the binary Th-Se system,viz., ThSe, Th,Se,, Th,Se,,, and ThSe2.lo5Distribution measurements of thorium ion between aqueous solutionsand a solution of thenoyltrifluoroacetone in benzene have confirmed thatG.H. Beyer, A. Jacobs, and R. D. Masteller, J . Amer. Chem. SOC., 1952, 74, 825.loo R. S. Hansen, K. Gunnar, A. Jacobs, and C . R. Simmons, ibid., 1950, 72, 5043;Ann. Reports, 1950, 47, 108, ref. 91.lol A. E. Van Arkel and J. H. de Boer, 2. anorg. Chem., 1924, 141, 289.loe E. M. Larsen, J. Howatson, A. M. Gammill, and L. Wittenberg, J . Amer. Chem.SOC., 1952, 74, 3489.lo3 J. S. Anderson and R. W. M. D'Eye, J., 1949, S244; E.Hayek and Th. Rehner,Experientia, 1949, 5, 114; E. Hayek, Th. Rehner, and A. Frank, Monalsh., 1951,83, 575.106 R. W. M. D'Eye, P.G. Sellman, and (Miss) J. R. Murray, J., 1952, 2555.104 J. C . Warf, J . Amer Chem. SOC., 1952, 74, 1864FAIRBROTHER. 95thorium(1v) is present a s a simple, hydrated, tetrapositive ion in perchloratesolutions of an acidity greater than about 0 . 0 8 ~ . The same technique hasbeen used to measure the association constants of some complexes formedbetween thorium ion and fluoride, chloride, nitrate, sulphate, and phosphateions, severally.lo6Some investigations have been reported in connection with bivalentgermanium. Freshly prepared hydrous germanous oxide, obtained byprecipitation with alkali in the cold, is yellow and retains its colour if storedunder water at room temperature.It is, however, thermodynamicallyunstable and on boiling or treatment with aqueous hydrochloric acid changesto dark brown. Potential measurements give GeO(y,llow) = GeO(broq, AGO =-7.2 kcal./mole and GeO, + 2Hf + 2e- = GeO(brown), E" = -0.118 If 0.010volt at 250.1°7 Measurements have also been made of the heat of oxidationof GeI, to germanic acid lo* and of the equilibrium Ge(s) + GeO,(s) =2GeO(g).lo9Several additional new complex compounds of bivalent germaniumhave been prepared : GeI2,2NH,Me ; Ge(H,PO,),,GeCl, ; 3Ge(H2P0,),,GeBr,.These resemble the corresponding compounds of bivalent tin. If ger-manium dioxide is heated alone with hypophosphorous acid it goes intosolution and is reduced to bivalent germanium.This is in contrast tothe behaviour of stannic hydroxide which, although soluble in hypo-phosphorous acid, is not reduced in the absence of hydrochloric acid.l1°It is well known that ordinary tin is converted into the low-temperaturegrey (a) modification at temperatures around 0" if a few particles of pre-viously prepared grey tin are also present. The difficulty usually met withis to obtain some of the latter to initiate the transition. It has now beenshown that if a small cylinder of pure white tin is surrounded by solid carbondioxide, and cold-worked a t this temperature (e.g., submitted to a pressureof several tons), and then kept at -78", it is almost completely transformedinto the grey variety in 24 hours.lllAn examination of the reaction between stannic chloride and hydrazinemono- or di-hydrochloride in varying amounts, has shown that mono-hydrazinium chlorostannate, N2H4,H,SnC1,, or N,H,SnCl,, although re-ported in the literature, does not appear to exist.The only product isolatedfrom these reactions is dihydrazinium chlorostannate, (N,H,),SnC16. 112Group V.-If hydrazine is added to an aqueous solution of a metal ion,such as Zn++ or Ni++, one usually obtains a precipitate in which the pro-Dortion of hvdrazine to metal ion is half that to be exDected from the usual I J co-ordinationH,N NH, I .'-.S,"' Inumber of the metal, e g . , Zn(N,HJ,X,, Ni(N,H,),X,,Ni(N,H,),X,. This has led to the supposition that in thesecompounds both nitrogen atoms of a given hydrazine moleculeare involved with the same metal ion in the formation of aH2N" NH2 three-membered chelate ring (inset).Evidence has now beenobtained which suggests that these metal hydrazine complexes may be inI *OLv'~~,.llo6 E. L. Zebroski, H. W. Alter, and F. K. Heumann, J. Amer. Chem. Soc., 1951,lo' W. L. Jolley and W. M. Latimer, ibid., 1952, 74, 5751.lo8 I d e m , ibid., p. 5752. 109 I d e m , ibid., p. 5754.110 D. A. Everest, J., 1952, 1670.ll1 E. S. Hedges and J . Y. Higgs, Nature, 1952, 160, 622.73, 5646.W. Pugh and A. M. Stephen, J . , 1952, 413896 INORGANIC CHEMISTRY,fact not chelate compounds, but endless networks in which each hydrazinemolecules i s bound to two metal cations :H N H N H4Nz H4N2 H N H N ,A a++ P .,,a++ ,' '.++ ,' ',,* ,A,:++ ,e4*8; .;M:' ;M' ';Ma' ;Ml ,m,' '\:2" H N 'HN 'HN H< H N 4 a 4 2 4 2 4 a 4 2 4 2In two cases no precipitate is obtained-those of the perchlorates and of thefluoroborates. The former are too explosive for investigation, but it hasbeen shown that in the fluoroborates, where the system remains homogeneous,Ni++ ion co-ordinates six molecules of hydrazine, and Zn++ four, the hydrazinefunctioning therefore as a monodentate group.l13Further progress has been made in the study of the liquid dinitrogentetroxide solvent system. The properties of a solution of diethylnitrosamine,which behaves as a " base " in this solvent, indicate that the double com-pound N,O,,SEt,N*NO, which is formed between solvent and solute, under-goes electrolytic dissociation according to the scheme :N204 + 2Et2N*N0 + N2O4,2Et2N-NO + (Et,N.NO),NO+ + NO,-Metallic zinc has been found to react rapidly with solutions of " bases " inliquid dinitrogen tetroxide.With a solution of diethylnitrosamine, nitricoxide is evolved and a red liquid is formed which is immiscible with thedinitrogen tetroxide. This red liquid is indistinguishable from the productobtained by dissolving the compound Zn(N0,),,2N20, in diethylnitrosamineor by adding dinitrogen tetroxide to a solution of anhydrous zinc nitrate indiethylnitrosamine. The reactions which take place are analogous to thereactions of metallic zinc and of zinc hydroxide with aqueous solutions ofalkali : 115Zn + 4N20, + 4Et,N*NO + [(Et2N-NO),NO+],[Zn(N0,),] + 2N0Zn(NO,), + 2N20,, + 4EtaN*N0 + [(Et2N~NO)2NO+]2[Zn(N0,),]Dinitrogen tetroxide also forms addition compounds with ethers, ofwhich a number have been prepared ; 116 with diethyl ether N,O4,2(C,H,),O,m.p. -74.8"; with tetrahydrofuran N,O,,C,H,O, m. p. -20.5", andN20,,2C4H,O (incongruent melting) ; with tetrahydropyran N20,,2C,Hlo0,m. p. -56.8"; and with dioxan N,O,,O[C,H,],O, m. p. 45.2". The struc-tures of these compounds have been studied by spectroscopic and magneticmethods. It is of interest that a suggested explanation of the relativelyhigh melting point of the dioxan compound is the use of both oxygens ofthe dioxan molecule to make possible the formation of an indefinitelyextended aggregation. It may be recalled 67 that co-ordination of dioxanwith one molecule of boron trichloride inhibits the donor character of theother oxygen atom.In connection with the study of liquid dinitrogen tetroxide systems,evidence has been obtained for the existence of a new series of nitrogenoxyacid compounds of the general formula Na,N,O,(x = 3-6).Il7 Thefirst member of this series Na,N,O,, the familiar sodium hyponitrite, is113 G.Schwarzenbach and A. Zobrist, Helv. Chim. A d a , 1952, 35, 1291.11* C. C. Addison and C. P. Conduit, J . , 1952, 1390.116 Idem, J., 1952, 1399.116 B. Rubin, H. H. Sisler, and H. Shechter, J. Amer. Chew. Soc., 1952, 74, 877.117 C. C. Addison, G. A, Gamlen, and R. Thornson, J., 1952, 338, 346FAIR3ROTHER. 97rapidly oxidised by liquid dinitrogen tetroxide to sodium P-oxyhyponitrite(p-Na,N,O,) which differs in chemical properties from the a-compound, ofthe same empirical formula, prepared from hydroxylamine and ethyl nitrate.The latter, sodium a-oxyhyponitrite, also undergoes rapid oxidation to thecompound Na,N,O, which again is a different compound from sodiumnitrite.N%N,04 undergoes a further slow oxidation in liquid dinitrogentetroxide to a compound of empirical formula NaNO, which may be thedimer Na,N,O,, and a silver salt believed to have the formula Ag,N20,has been prepared. Sodium p-oxyhyponitrite undergoes slow oxidation inliquid dinitrogen tetroxide to the compound Na,N,O, which is also formedrapidly by the action of nitrogen dioxide gas a t 100" on sodium hyponitrite,and is further oxidised slowly under these conditions to the NaNO, (orN+N,O,) stage.Ultra-violet absorption spectra of aqueous solutions ofthe hydrolysis products of these compounds have been examined and anattempt has been made to formulate structures for them.Hydrogen peroxide has been shown to react with nitrous acid, nitricoxide, and nitrogen dioxide with the formation of pernitrous acid, HNO,.From the behaviour of the products of these reactions in initiating the poly-merisation of methyl methacrylate and in hydroxylating and nitratingbenzene, it has been suggested that pernitrous acid undergoes homolyticfission to give OH radicals and nitrogen dioxide.' l8Further studies have been made of nitrosyl chloride as an acid-baseionising solvent and of the behaviour in solution of several nitrosyl doublecompounds : NOFeCl,, NOAlCl,, NOBF,, (NO),SnCl,.Nitrosyl com-pounds react in nitrosyl chloride solution with the slightly soluble (CH,),NC!in the sense of the acid-base neutralisation NO+ + C1-c NOCl as demon-strated by a conductometric titration of NOFeC1, by (CH,),NCl and by thepartial neutralisation equilibria of NOBF,, NOClO,, and (NO),S,O,, eachwith (CH,),NCl. The idea that NO+ represents " acid I' and C1- " base "is further justified by the electrolysis of NOFeC1, in NOC1, which gives NOat the cathode and C1, at the anode.llsSome work has been carried out on the constitution and reactions ofnitryl chloride, N0,Cl. On hydrolysis and alcoholysis this behaves as anitrosyl hypochlorite. 120 By the reaction of nitryl chloride with antimonypentachloride in liquid chlorine, one can obtain a compound NO,Cl,SbCl,which behaves as nitrosyl chloroantimonate [No,]+[sbc~,]- : it dissolves inliquid sulphur dioxide to give a conducting solution and undergoes ionicreactions with tetramethylammonium perchlorate and fluoroborate to givenitryl perchlorate and fluoroborate respectively : 121[NO,][SbCl,] + [NMe,]ClO, = [NO,]CIO, + [NMeJSbCl,[NO,][SbCI,] + [NMeJBF, = [N0,]BF4 + [NMe,]SbCI,An investigation of the behaviour of nitryl chloride towards ammoniaand a number of Lewis acids suggests that the nitrogen-chlorine bond isnot so strongly polarised as to act as a source of negative chlorine, excepttowards exceedingly strong electron acceptors such as sulphur trioxide,with which it forms N0,C1,2S03, presumably nitronium chlorodisulphate118 E.Halfpenny and P. L. Robinson, J., 1952, 928.ll9 A. B. Burg and D. E. McKenzie, J . Amer. Chem. SOC., 1952, 74, 3143.lZo F. Seel and J. N6grAdi, 2. anorg. Chew., 1952, 269, 188.leL I?. Seel, J. N6gr&di, and R. Posse, ibid., p. 197.REP.-VOL. XLIX. 98 INORGANIC CHEMISTRY.N0,(ClS206). Thus, on reaction with ammonia it yields chloroamine andammonium nitrite, rather than nitramide and ammonium chloride, and itdoes not react with SnCl, or BF,, which might be expected to co-ordinatea negative chlorine. It has been suggested that the addition of the secondoxygen atom in going from nitrosyl chloride to nitryl chloride either greatlyreduces the polarity of the N-C1 bond or actually reverses its direction, sothat the chlorine becomes positive.122Dinitrogen tetrasulphide, N,S,, which has been known for a long time,123has received renewed attention.This compound, which is prepared bythe action of sulphur in N,S, (e.g., in CS, solution in an autoclave a t 100-120"), is a dark red, fairly unstable solid, m. p. 23". It is soluble in manyorganic solvents and is diamagnetic. The latter excludes a dissociation intoNS, molecules, which would be paramagnetic, whilst its chemical behaviourtowards a variety of reagents suggests that its structural formula is quitedifferent from that of its oxygen analogue dinitrogen tetroxide.124The two new double compounds Li,AlP, and Li,AIAs, have been pre-~ared.1,~ In their chemical reactions these compounds behave more likeLi,P and Li,As than as A1P or A1S respectively.It is reported that compounds of the type PX,,IY, where X and Y aredifferent halogens, can be synthesised directly, either by fusion together ofthe components PX, and IY, or by dissolving these in CCI,. PBr,ICl hasbeen prepared by the latter method and forms cherry-red needle-shapedcrystals, m.p. 112.8". It is formulated by the authors as [PBr,+][BrICI-].PC1,IBr forms yellow crystals, m. p. 140" (decomp.).12,A spectrophometric study of phosphorus hexachloroiodide and hexa-bromoiodide has shown that these compounds dissociate in carbon tetra-chloride solution into their component molecules, PCl,I 2 PCl, + ICI;PBr,I -+ PBr, + IBr + Br,, whereas in polar solvents such as aceto-nitrile the dissociation is ionic : PCI,I -+ PC1,' + Ic12- and PBr,I 4PBr,+ + IBr2-.127Solubility measurements of antimonous oxide in water and in dilutesolutions of hydrochloric acid and of sodium hydroxide show that in bothacid and alkaline solutions the antimony is present as a univalent ion, SbO+and Sb02-, respectively.lZ8Antimony pentafluoride , a compound with some remarkable physicalproperties,l29 appears to possess a power of compound formation which ismuch greater than that of arsenic pentafluoride.Some unusual chemicalcompounds which include antimony pentafluoride have been reported. Itdissolves sulphur, selenium] and tellurium to give blue, yellow, and redsolutions, respectively, from which the stable crystalline compounds(SbF,),S, (SbF,),Se, and (SbF,),Te can be isolated.Sulphur dioxide formsSbF,SO,, and NO, forms SbF,N0,.13*122 H. H. Batey and H. H. Sisler, J . Amer. Chem. SOC., 1952, 74, 3408.134 M. Goehring, H. Herb, and H. Wissemeir, 2. anorg. Chem., 1952, 267, 238.lZ6 R. Juza and W. Schulz, ibid., 1952, 269, 1.126 I. D. Muzyka and Ya. A. Fialkov, Doklady Akad. Nauk. S.S.S.R., 1952, 83, 415;127 A. I. Popov and E. H. Schmorr, J . Amer. Chem. Soc., 1952, 74, 4672.128 K. H. Gayer and A. B. Garrett, ibid., p. 2353.lee A. A. Woolf and N. N. Greenwood, J., 1950, 2200.l30 E. E. Aynsley, R. D. Peacock, and P. L. Robinson, Chem. and I s d . , 1951, 1117.F. L. Usher, J., 1925, 730.Chem.Abs., 1952, 46, 6983FAIRBROTHER. 99It is reported131 that almost pure vanadium monoxide (99.68%) hasbeen prepared by heating a compressed intimate mixture of V203 andfinely divided vanadium in a vacuum a t 1750". The product is describedas having a speeific gravity of 5.55, a hardness of 8-9 on Mohs' scale, andto dissolve in acids to give the blue or violet solutions characteristic ofhypovanadous salts.The monoboride VB and the mononitride VN have also been examined,the former being prepared 132 as the product of the simultaneous reductionof V205 and B203 with carbon in a graphite crucible at 1650" in an atmosphereof hydrogen, and VN in a crystalline form by the action of nitrogen andhydrogen on the vapour of vanadium tetrachloride in the presence of aheated filament.The latter technique has also been used to prepare niobiummononitride NbN, though in the form of smaller crystals than the vanadiumnitride.133A preliminary report has been made of a new solvent extraction methodfor the separation of niobium and tantalum. It has been found, by usinga tracer technique, that niobium can be extracted almost quantitativelyfrom concentrated hydrochloric acid solution by a solution of methyldioctyl-amine in xylene : under these conditions the extraction of tantalum appearsto be almost negligible. The niobium can then be extracted from theorganic phase with nitric, sulphuric, or dilute hydrochloric a ~ i d . 1 ~ ~An inorganic chromatographic separation of these two elements fromone another and from other elements present in complex minerals, andsuitable for their quantitative determination, has been developed.Thismethod is based on the adsorption of the metals as fluorides on a cellulosecolumn in a Polythene tube, and selective extraction by ethyl methylketone containing hydrofiuoric acid. 135The elution by 7.0~-hydrochloric acid of carrier-free 95Nb, adsorbed froma 10.0M-hydrochloric acid solution on a Dowex 2 anion-exchange resin,instead of giving the usual symmetrical curve of activity of eluant versusvolume of eluant, gives a curve which shows several peaks. As these can beidentified by the characteristic disintegration rate of 95Nb, the presence ofany other species, or of isotopic separation, may be excluded.This resulthas been attributed to the slow establishment of equilibrium among thevarious ionic species present, which would not necessarily have differentcharges as in the case of the thiocyanate complexes of chromium, whichhave been separated by this kind of technique,20 but may contain differentnumbers of chloro-, oxy-, and hydroxy-groups. Elution with 6.0~-hydro-chloric acid gives the usual symmetrical elution curve.136It has been shown that niobium pentachloride and tantalum penta-chloride form a continuous series of mixed ~rysta1s.l~~When niobium pentachloride is heated at 350-400" with niobium metal,in a molar ratio greater than 4 : 1, niobium tetrachloride may be obtained131 M. Frandsen, J . Amer. Chem. Soc., 1952, 74, 5046.132 H.Blumenthal, ibid., p. 2942.133 F. H. Pollard and G. W. A. Fowles, J., 1952, 2444.la4 G. W. Leddicotte and F. L. Moore, J . Amer. Chem. Soc., 1952, 74, 1618.lS6 F. H. Burstall, P. Swain, A. F. Williams, and G. A. Wood, J., 1952, 1497;136 E. H. Huffman and G. M. Iddings, J . Amer. Ckem. Soc., 1952, 74, 4714.137 H. Schafer and C. Pietruck, 2. anorg. Chem., 1952, 267, 174.A. F. Williams, J., 1952, 3155; R. A. Mercer and A. F. Williams, J . , 1952, 3399100 INORGANIC CHEMISTRY.as large needles. With a greater proportion of metal, the trichloride isobtained, or some lower chloride which disproportionates near 600" into themetal and the trichloride.138Niobium tetrachloride also disproportionates on heating, and the equili-brium pressure of the decomposition 2NbC1, = NbCl, + NbC1, has beenmeasured.139Although quinquevalent tantalum is much less easily. reduced thanquinquevalent niobium, TaCl, being unaffected by hydrogen a t temperaturesup to 400°, there is evidence that at temperatures over 500" some reductionof TaCl, does occur. Similarly, if hydrogen chloride or hydrogen bromideis passed over the metal a t about 400' only the pentahalide is formed. Athigher temperatures (between 600" and SOO"), however, lower halides areformed which may undergo some disproportionation with the deposition offilms of metallic tantalum on the walls of the apparatus.140The reduction of niobium and tantalum pentoxides, and mixtures of 'thetwo, to the quadrivalent dioxides by moist hydrogen a t 900°, has beeninvestigated.141Group VI.-A comprehensive review has been published 142 of themethods used for the production and determination of abundance of isotopicoxygen and of its applications. Details have also been given 143 of theconstruction and operation of a number of fractionating columns for theenrichment of l 8 0 in water. One of these is reported as being capable ofproducing per day, 200 C.C. of 0.6% H,180, or 50 C.C. of 1.7% H2180 or20 C.C. of 3.2% H,I80 : another, operating as a second stage, gives 10 C.C.of 12% H,180 per week.The isotope l80 has been used as a tracer in the study of the mechanismof oxidation of hydrogen peroxide.lU The oxygen liberated when H,O,is oxidised in aqueous solution by Ce(Iv), Mn04-, Cl,, HC10, and Cr20,.p isderived wholly from the hydrogen peroxide and not from the water.Simi-larly, the oxygen liberafed in the catalytic decomposition of H,O, by Fe(m),I-, I,, Br-, Br,, MnO,, and Pt is derived only from the H,O,.A series of higher sulphur chlorides with the composition S&1, (where xcan have values up to about 100, depending upon the temperature of re-action), and consisting of long sulphur chains terminated a t the ends bythe two chlorine atoms, has been prepared by a " hot-cold '' tube methodfrom S,Cl, and h ~ d r 0 g e n . l ~ ~Some new trifluoromethyl derivatives of sulphur have been prepared.Trifluoroiodomethane, CFJ, reacts with sulphur to give bistrifluorornethyldisulphide (CF3),S,, carbon disulphide, thiocarbonyl fluoride, and poly-sulphides.Bistrifluoromethyl disulphide undergoes an unusual type ofhydrolysis in aqueous alkali, the first stage of which consists in the hydrolyticfission of the S-S bond to give trifluoromethanethiol and trifluoromethane-138 H. Schafer, C. Goser, and L. Bayer, 2. anorg. Chem., 1951, 265, 258; C. H.139 H. Schafer, L. Bayer, and H. Lehmann, 2. anovg. Chem., 1952, 268, 268.140 R. C. Young and C. H. Brubaker, Jr., J . Amer. Chem. SOC., 1952, 74, 4967.1 4 1 H. Schafer and B. Breil, 2. anorg. Chem., 1952, 267, 265.142 M. Dole, Chem. Reviews, 1952, 51, 263.143 I. Dostrovsky, D. R. Llewellyn, and B. H. Vromen, J . , 1952, 3509 : I. Dostrov-sky, J. Gillis, D. R. Llewellyn, and B. H. Vromen, J., 1952, 3517.144 A. E. Cahill and H. Taube, J .Amer. Chem. SOC., 1952, 74, 2312.145 F. FCher and M. Baudler, 2. an0p.g. Chem., 1952, 267, 293.Brubaker, Jr., and R. C. Young, J . Avner. Chem. SOC., 1952, 74, 3690FAIRBROTHEH. 101sulphenic acid which subsequently break up to give fluoride, carbonate; andsulphide ions :NaOHCF,*S*S*CF, -> CF,-SH + CF,*S*OHF-, CO,', S" f- F2C:S + HF F-,CO,', S"J. J.On irradiation in the presence of mercury, the disulphide yields bis(trifluor0-methy1thio)mercury (CF3S),Hg : in the absence of mercury it gives bistri-fluoromethyl sulphide which, in contrast to the disulphide, is quite stableto aqueous alkali. 146It is reported that the chloride of trichlorophosphazosulphuric acid,ClSO,N:PCl,, has been prepared by the reaction a t 100" between sulphamicacid and phosphorus pentachIoride.147New methods have been described for the preparation of seleniumoxyfluoride and selenium tetrafluoride. The latter is prepared by theaction of dilute fluorine on selenium a t 0". It is a liquid, m.p. -9-5', b. p.106", with some remarkable solvent properties. It dissolves the fluoridesof sodium, potassium, rubidium, and caesium to form complexes with acomposition approaching MSeF, which is different from that of any of theother complex selenium halides, viz., M,SeX,. 148Selenium, diselenium, and triselenium di(benzenesu1phinate) and di-(toluenesulphinate), Se(SO,R),, Se,(SO,R),, and Se,(SO,R),, have beenprepared, and represent a new class of selenium compounds of which thesulphur analogues have been known for a long time.In their reactionsthey behave as derivatives of Se++, Se2++, and Se,++, re~pectively.1~~ Selen-ium and tellurium di(benzenethiosu1phonate) and di(to1uene-p-thiosulphate)have also been prepared. 15*Conductivity measurements have shown that tellurium tetrachloridebehaves as an " acid " in arsenic trichloride solution and may be titratedconductimetrically in this solvent with (CH3),NC1. Compounds can beisolated from the resulting solution which are probably the " acid " and" normal " salts respectively, (NMe,) (AsCl,) (TeC1,) and (NMe4),TeC1,.151The system chromium(II1) oxide-water has been studied in the temper-ature range 145-560" by the hydrothermal method. A definite blue-greycompound, CrO(OH), d = 4-12, is formed below 419424'. This decom-poses sharply and endothermically a t about 430".Rhombohedra1 Cr,O,is the stable phase above 450°.152The products obtained when chromium(v1) trioxide is heated in a vacuumhave been investigated by X-ray and chemical ana1y~is.l~~ Evidence hasbeen obtained of the existence of three definite compounds intermediatebetween CrO, and C1-203, vix., Cr,O,, Cr,O,, and CrO,.Pure chromyl fluoride, CrO,F,, has been prepared for the first time, bythe reaction between CrO, and anhydrous hydrogen fluoride in an apparatusbuilt out of copper, silica-free glass, and Kel-F tubing. It forms violet-red1413 G. A. R. Brandt, H. J. Emelkus, and R. N. Haszeldine, J., 1952, 2198.14' A. V. Kirsanov, Zhur. Obshckei Khim., 1962, 22, 88; Chem. Abs., 1952, 46, 6984.148 E.E. Aynsley, R. D. Peacock, and P. L. Robinson, J., 1952, 1231.149 0. Foss, Acta Claem. Scand., 1952, 6, 508.150 Idem, ibid., p. 521.161 V. Gutmann, Monatsh., 1952, 83, 159.152 A.158 R. S . Schwartz, I. Fankuchen, and R. Ward, ibid., p. 1676.Laubengayer and H. W. McCune, J . Apner. Chem. SOC., 1952, 74, 2362102 INORGANIC CHEMISTRY.crystals which have a v. p. of 760 mm. at 29.6" and melt to an orange-redliquid at 31.6" under a pressure of 885 rnrn.15,An examination of the spectra of chromic acid and chromic acid-phos-phoric acid systems has indicated the formation of two chromate-phosphatecomplex ions, HCrPO," and H,CrPO,-. 155The alkali-metal compounds of chromium of the type MCr3.0, have beenprepared and examined.156 The oxidation states of chromium in thesecompounds, which are obtained by heating mixtures of the compositionM,Cr,O, + XrO, at 350" in air for two hours and extracting the excess ofthe dichromate with water, appear somewhat problematical. The blackmetallic appearance of, for example, the potassium compound suggests thatit is not a compound containing chromium in its two usual and independentstates of oxidation 3 and 6, K,Cr,(CrO,), or K2(CrO)2(Cr207)2, but that thereis some interaction between the chromium ions so that the effective overalloxidation state is +5.A compound of univalent chromium (dipy3Cr1)(C10,) (dipy = 2 : 2'-dipyridyl) is reported to be formed by reduction of the bivalent compound(dipy,Crrl) (ClO,), by magnesium in the presence of ammonium perchlorate,and with exclusion of air.157On treating chromium hexacarbonyl with alcoholic potassium hydroxide,a brilliantly yellow derivative is obtained, which on acidification yields awhite, volatile, unstable, crystalline substance, the properties of whichsuggest that it is C T ( C O ) ~ H , .~ ~ ~The term " tungsten bronze " is used to describe the non-stoicheiometriccompounds of general formula M,WO, (where M is an alkali metal and x isless than unity) obtained on reduction of the alkali tungstates. Thesehave been considered (a) as solutions of tungsten(v1) oxide in a hypotheticaltungsten(v) compound MWO,, and also ( b ) , on account of their high electricalconductivity, low magnetic moment, and general metallic appearance, assolid solutions of alkali metal in WO,.The essential difference betweenthese two models is that in (a) the electron from the alkali metal is con-sidered to be strongly associated with the tungsten(v) ion, which wouldtherefore be paramagnetic, whereas in (b) the electron from the alkalimetal is part of the electron gas that is associated with the whole lattice.To distinguish between these two models, the magnetic susceptibilities of avery wide series of lithium tungsten bronzes have been measured.159 Thevalues of the susceptibilities are very low and actually become diamagneticas the concentration of lithium is decreased. The results are in agreementwith those calculated for an electron-gas model.Studies have been made of the alkali-metal molybdate systems, K,MoO,-MOO,, Rb,MoO,-MoO,, and CS,MOO~-MOO,,~~~ and of the alkali fluoridernolybdenum(v) systems, LiF-MOO,, NaF-MOO,, KF-MOO,, RbF-Moo,,and CsMoO,.lG11 5 4 A.Engelbrecht and A. V. Grosse, J . Amer. Chem. Soc. 1952, 74, 5262.155 F. Holloway, ibid., p. 224.1 5 6 L. Suchow, I. Fankuchen, and R. Ward, ibid., p. 1678.1 5 7 Fr. Hein and S. Herzog, 2. anorg. Chem., 1952, 26'4, 337.1 5 8 M. G. Rhomberg and B. B. Owen, J . Amer. Chem. Soc., 1951, 73, 5904.1 5 9 L. E. Conroy and M. J. Sienko, ibid., 1952, 74, 3520.160 V. I. Spitsyn and I. M. Kuleshov, J . Gen. Chem. U.S.S.R., 1951, 1493; Chem.161 0. Schmitz-Dumont and I. Heckmann, 2. amrg. Chem., 1952, 26'4, 277..Abs., 1952, 46, 9006FAIRBROTHER. 103Uranium and the trans-uranic elements.Further investigation has beenmade of the sodium uranates. Chemical, X-ray diffraction, pH, and con-ductivity data indicate that when sufficient sodium hydroxide is added touranyl nitrate solutions, two compounds Na,U702, and Na6U,O2, (ormixtures of these) are precipitated. In the first place, when just insufficientsodium hydroxide to produce precipitation is added (which requires 1 mol.or more of NaOH per mol. of U02++), basic uranyl ions UO,UO,++ and(U03),U02++ are formed. Further addition of NaOH to a mole ratioNaOH/U of 2.29, causes the quantitahe precipitation of the uranium asNa,U,O,,. This in turn may react with excess of alkali to give Na6U7Oz4.l6,The diuranates of the alkaline-earth metals have been prepared by theignition of the corresponding metal uranyl acetates, a method of preparationwhich has been found to give a product free from an excess of either thealkaline-earth metal or uranium oxide.The thermal stabilities in vacuumand in oxygen up to 1100" have been investigated and the results have shownthat the metal diuranate-oxygen systems are reversible below thistemperature. 163The magnetic susceptibilities of UF4-ThF, solid solutions reportedearlier,164 which seemed to indicate the presence of two 5f electrons in UF,,have been corrected by the author. The amended values 165 show that,within experimental error, both the susceptibility at room temperature andthe moment extrapolated to infinite dilution agree with the values predictedfor two unpaired spins with the orbital contribution to the moment com-pletely quenched.These results therefore fall into line with those obtainedfor the U02-Th0, solid solutions 166 and imply a configuration of 6d2 forthe U4+ ion.The magnetic susceptibilities of plutonium dioxide and tetrafluoridehave also been measured, over the temperature range 90-450" K. Measure-ments made on the solid solutions of PuF, in the isomorphous ThF, lead toan extrapolated susceptibility at infinite dilution in agreement with a 5f4configuration. The dioxide has approximately the same susceptibility atinfinite dilution, but the behaviour with increasing concentration is morecomplex and some of the evidence may indicate that 6d levels are occupied.167The sexavalent plutonium ion should have two unpaired electrons, and inorder to determine whether these are in the 5f or the 6d level, the magneticsusceptibility of sodium plutonyl acetate has been measured over the tem-perature range 90-300" K.The results agree with the theoretical value forspin-only for two electrons, which therefore may be taken as evidence thatthe ground state of the plutonyl ion has a 6d2 configuration.168 + H(aq)+. = + 2.4 v, and Pqaq)++ + +H,,,) = Pr,,,,3+ + H(aql+ = + 2.9 v have been estim-ated from thermodynamical data and .measurements of the heats of reactionof the oxides with nitric acid and fluoroboric acid in a micro-calorimeter.169The potentials of the couples + +H,,, =162 C. A. Wamser, J . Belle, E. Bernsohn, and B. Williamson, J.Amer. Chem. SOL,H. R. Hoekstra and J . J . Katz, ibid., p. 1683.164 J. K. Dawson, J., 1951, 2889; Ann. Reports, 1951, 48, 102, ref. 146.166 I d e m , J., 1952, 1185.166 W. Trzebiatowski and P. W. Selwood, J. Amer. Chem. SOL, 1950, 72, 4504;Ann. Reports, 1950, 4'7, 116, ref. 164.1 6 7 J . K. Dawson, J., 1952, 1882.16s L. Eyring, H. R. Lohr, and B. B. Cunningham, J . Amel.. Chem. SOC., 1952,74,1186.1952, 74, 1020.lB8 Idem, J . , 1952, 2705104 IXOKGANIC CHEMISTRY.An interesting feature of this work was that the primary objective was theevaluation of the Am4+-Ama+ potential and that the chemically similar butmore abundant praseodyrninium was used-as the authors express it-as a" stand-in " for the perfection of the technique.A further study has been made of the chemistry of sexavalent americium,which has been shown to resemble that of U(VI), Np(vr), and Pu(v1).170A search has been made in aqueous solution for oxidation states ofcurium higher than Cm3+ by using macro-amounts (up to 238 pg.per experi-ment) of the element, and using 'americium(m), which does exhibit thehigher oxidation states +5 and +6 in solution, as an internal check. Noevidence was found for the existence of Cm4+, Cm5 I , or Cm6+ in the oxidationof Cm3+-Am3+ mixtures in either acid or alkaline solutions, under con-ditions where the Am3+ was oxidised quantitatively to Am5+ or Am6+.171Group VI1.-The conditions and products of reaction of elementaryfluorine with zinc, zinc oxide, zinc bromide, zinc sulphide, nickel, nickel(I1)oxide, nickel(m) oxide, and nickel(r1) sulphate have been investigated.There was no evidence of the production of a higher fluoride of nickel bythe fluorination of the so-called nickel(II1) oxide, Ni20, : in all cases thebivalent fluorides were the only non-volatile products observed.172Silver tetrafluoroborate can be simply prepared by the action of brominetrifluoride on dry silver borate, and undergoes rapid decomposition at 200"into silver fluoride and boron trifluoride.These reactions form a convenientmethod of preparation of small quantities of anhydrous silver fluoride, andcan be carried out in quartz apparatus.173The magnetic susceptibilities of a number of simple and complex fluoridesof transition metals have been measured, vzz., K2TiF, ; K,TiF,,H,O ;VF, ; K,VF, ; K,CrF5,H,0 ; KCrOF, ; K,MnF5,H20 ; Li,FeF, ; Na,FeF, ;K,FeF, ; CsFeF, ; CuF,,2H20 ; TaF, ; RhF, ; Na,RhF, ; PdF, ; PtF, ;K,PtF, ; AuF, ; AgAuF,.Compounds of the first transition series, exceptwhen magnetically concentrated, show the number of unpaired electronswhich would be expected for ionic binding. For the second and third tran-sition series, however, when the number of electrons in the metal ion iseven, the compounds are diamagnetic, and paramagnetic when the number isodd, but with a moment which corresponds to only one unpaired electron,which shows that for these two series, minimum multiplicity is the rule evenfor f l ~ 0 r i d e s . l ~ ~The commercial availability of chlorine trifluoride, and the ease withwhich it can be used, offer the possibility of its substitution for elementaryfluorine as a fluorinating agent in a number of reactions, and a number ofmetal fluorides have been prepared from the metals by its use.Detailshave now been given of the preparation of cobalt(m), nickel(II), and silver(I1)fluorides from chlorine trifluoride and cobalt(rI), nickel(n), and silver(1)chlorides, re~pective1y.l'~The existence of bromine monochloride in carbon tetrachloride solution170 L. B. Asprey, S. E. Stephanou, and R. A. Penneman, J . Amer. Chem. SOC.,171 S. E. Stephanou and R. A. Penneman, ibid., 1952, 14, 3701.172 H. M. Haendler, W. L. Patterson, Jr., and W. J . Bernard, ibid., p. 3167.173 A. G. Sharpe, J., 1952, 4538.174 R.S. Nyholm and A. G. Sharpe, J . , 1952, 3579.175 E. G. Rochow and I. Kukin, J . Amer. Chem. Soc., 1952, 74, 1615.1951, 73, 5715FAIRBHOTHE R. 105was clearly proved nearly 25 years ag0,1'~ and described as probably highlydissociated. The extent of this dissociation has now been re-investigatedspectrophotometrically and found to amount to 43.2&1% at 25". Theequilibrium constant for the reaction 2BrCl + Br, + C1, in carbon tetra-chloride has been calculated to be 0-145 & 0-006.177Alkali hypobromites are familiar in aqueous solution, but no solid hypo-bromites appear to have been isolated.178 It is reported, however, that at-5" it is possible to isolate the crystalline alkali hypobromites NaBrO,SH,Oand 7H20, and KBr0,3H20 from the products of the action of bromine onsolutions of the respective hydroxides at this temperat~re.~'~that molecular iodine formedbrown solutions, and was polarised, in electron-donor or " basic " solvents,which included aromatic and olefinic hydrocarbons.It was subsequentlydemonstrated that iodine formed 1 : 1 addition compounds with severalaromatic hydrocarbons. Ultra-violet absorption measurements havenow shown that iodine forms 1 : 1 addition compounds also with olefins atlow temperatures. 82The system MnS0,-H2S04-H,O has been investigated at O", 20", 25",45", 65", and 95.7", and the following solid phases identified : MnS0,,5H20,MnSO,,H,O, MnS0,,H2S0,,H,0, MnSO,,H,SO,, and MnS04,3H,S04.183The separation of technetium from uranium-fission product wastes hasmade this element available in weighable amounts and therefore madepossible the investigation of its chemistry by ordinary analytical methodsinstead of only by tracer techniques. Approximately 0.6 g.of the spectro-graphically pure metal has been prepared by the hydrogen reduction ofammonium pertechnetate lS4 and it has been established that the lightyellow crystalline oxide (m. p. 119.5" 0.1') obtained when technetium isheated in dry oxygen at 400-600" is Tc,O,. The pertechnetate ion possessesan intense ultra-violet absorption (molar extinction at 2470 = 4000)which can be used for its spectrophotometric detennination-as little asg. of technetium can be detected in this way.lS5 Technetium hepta-sulphide, previously assumed to be Tc2S7 on the basis of analogy with Re$,,has also been examined and its formula confirmed by chemical analysis.186The existence of rhenium in aqueous solution in the -1 oxidation stateis now accepted on the basis of the stoicheiometry of its reduction.A solidrhenide, however, has recently been reported for the first time. Whenpotassium perrhenate is reduced by potassium in ethylenediamine, a whitesolid is deposited which is stated to contain rhenium in the -1 oxidationstate, together with some potassium hydroxide.la7It was pointed out several years ago17& A. E. Gillam and R. A. Morton, Proc. Roy. Soc., 1929, A, 124, 604.177 A. I. Popov and 3. J. Mannion, J . Amer. Chem. Soc., 1952, 74, 222.1'9 R. Scholder and K. Krauss, 2. anarg. Chem., 1952, 288, 279.180 F.Fairbrother, J., 1948, 1051.181 H. A. Benesi and J. H. Hildebrand, J . Amer. Cham. Soc., 1949, 71, 2703.1*2 S. Freed and K. M. Sancier, ibid., 1952, 74, 1273.183 D. Taylor, J., 1952, 2370.184 J. W. Cobble, C. M. Nelson, G. W. Parker, W. T. Smith, Jr., and G. E. Boyd.J . Amer. Chem. SOC., 1952, 74, 1852.185 G. E. Boyd, J . W. Cobble, C. M. Nelson, and W. T. Smith, Jr., ibid., p. 556,186 C. L. Rulfs and W. W. Meinke, ibid., p. 235.187 E. Griswold, J . Kleinberg, and J. B. Bravo, Science, 1962, 115, 375.N. V. Sidgwick, " The Chemical Elements and Their Compounds," Oxford,1950, p. 1221106 INORGANIC CHEMISTRY.Group VII1.-The bicentenary of the discovery of nickel in 1751 byCronstedt has been marked by a symposium on recent developments in thechemistry and applications of nickel and its compounds.188A detailed X-ray examination of the compound Ni(CN),,NH,,C,H,,which is obtained as a precipitate when benzene is added to a solution ofnickel cyanide in ammonia, has shown that the benzene-which exhibits no'detectable vapour pressure a t room temperature-is not linked to thenickel complex by chemical bonds, but that the benzene molecules are heldin cavities formed by the solid structure of the complex.189A study of the reaction between nickel tetracarbonyl and o-phenylene-bisdimethylarsine has shown that two of the carbonyl groups are readilyreplaced by the chelate group to yield a stable, crystalline, diamagneticcompound of the formula Ni(CO),(diarsine) : attempts to replace all fourcarbonyl groups, however, were unsuccessful. Oxidation of this complexwith iodine liberates carbon monoxide, giving the compound NiI,(diarsine) ,the first of a previously unknown class of cis-planar bivalent nickel complexes.The corresponding chloride and bromide are much less stable.By oxidationof the bromide with excess of bromine, a tervalent nickel complex bromideof the formula NiBr,(diarsine) has been prepared.lgODetails have been given of a convenient laboratory method of preparationof nickel and cobalt carbonyls and of a number of their derivatives, which isbased on the reaction between carbon monoxide, at atmospheric pressure,and aqueous ammoniacal solutions of nickel and cobalt salts in the presenceof sodium dithionite.lglSeveral new reactions of sodium in liquid ammonia with the carbonylsof cobalt and iron have been reported.lg2Cobalt(I1) chloride is well known to form a mono-, a di-, and a hexa-hydrate, but there is less definite evidence of the existence of a tetrahydrate.An investigation of the system CoC1,-H,O-acetone has now shown that thetetrahydrate and a hitherto unreported trihydrate also exist over narrowranges of water a c t i ~ i t y .1 ~ ~A spectrophotometric study of the deep blue solution obtained when acobalt(I1) salt is dissolved in strongly alkaline solution has shown that thecolour is due to a trihydroxycobalt(I1) ion, C O ( O H ) ~ - . ~ ~ ~The products of the reduction of cobalt(I1) nitrate with two equivalentsof potassium in liquid ammonia are (i) insoluble cobalt(I1) amide and (ii) amixture of soluble nitrate and nitrite.If a large excess of potassium isused, however, almost all the nitrate is reduced to nitrite, and the insolubleproduct consists principally of potassium hydroxide and elemental cobalt.The cobalt which is obtained by the reduction of cobalt(II1) bromide inliquid ammonia in this way exhibits a marked activity as a catalyst for thehydrogenation of ally1 alcohol a t room temperature.lg5Some preliminary results have been published of an investigation of theSymposium, I n d . Eng. Chem., 1952, 44, 949.J . H. Rayner and H. M. Powell, J., 1952, 319.190 R. S. Nyholm, J., 1952, 2906.191 W. Hieber, E. 0. Fischer, and E. Bockley, 2. anorg. Chem., 1952, 269, 308.192 H.Behrens, 2. Nalurforsch., 1952, 7, b, 321.lgS L. I. Katzin and J. R. Ferraro, J . Amer. Chem. Soc., 1952, 74, 275.lg4 S. Gordon and J. M. Schreyer, ibid., p. 3169.195 G. W. Watt and C . W. Keenan, ibid., p. 2048FAIRBROTHER. 107preparation and properties of the curious compound known as " cobalticacetate "-the product obtained by anodic oxidation of cobalt(I1) acetateThe elucidation of thex4c0<~~>cOx4 structure of this compound presents some difficulties, butits properties seem to indicate a possibility that it may be a binuclearcomplex containing the p-dihydroxo-bridge grouping (inset). lg6Tracer experiments with H,180 have shown that in the aquation ofcarbonat open t amminocobalt (111) ionwhich takes place rapidly in water, and still more rapidly when the solutionis acidified, the cobalt-oxygen bond remains intact, the removal of carbonatefrom the complex ion taking place by the removal of carbon dioxide, i.e.,by the breaking of the carbon-oxygen bond.lg7 The mechanism is thereforesimilar to that observed in the hydrolysis of an ester.lg8A study of the type of bonding in a number of bidentate chelate cobaltcomplexes has been made by a radio-isotope exchange technique, using 6oCo.Where the bonding is mainly ionic, a rapid exchange occurs and as the co-valent character of the bonding increases so one can expect a diminution inthe rate of exchange of the cobalt.The same technique has been used tostudy the behaviour of bis(salicylideneani1ine) cobalt (11) and cobalt (11)acetate in pyridine solution on alumina and similar surfaces and on an ion-exchange resin, The break-up of the complex on these surfaces points tothe considerable ionic nature of the binding of the cobalt atom in suchcomplexes. lg9Further sexadentate cobalt (111) compounds have been prepared by theuse as ligands of condensation products of a series of sulphur-containingaw-diamines of the general formula : NH,*[CH,]Z*S*[CH,],*S*[CH2]z*NH2,where x, y , and x are 2 or 3, with salicylaldehyde or 2-hydroxy-l-naphth-aldehyde.Most of the complex salts prepared have been resolved intooptical isomers, some with extremely high rotations.200In connection with a study of the magnetic moments of octahedralcomplexes of Cr, Mn, Ni, Co, and Fe, with the chelate agents dipyridyl,and o-phenylenebisdimethylarsine, C,H,(AsMe,),, the following newcompounds have been prepared: [Co(diarsine),] (ClOJ,, [Co(diarsine),] (C1OJ3,[Co(diarsine),( OAc) 2] (C10,) , [Co( dipy),] ( C10,),,3H20, and[Cr(dipy),Cl,]Cl,ZH,O. Also the co-ordination of tervalent chromiumwith a tertiary arsine has been reported for the first time.In compoundsof the type MIr(dipy),X2 the bonds are ionic 4s4P34d2 when MI1 = Mn orCo and covalent 3d24s4P3 when the metal is Fe or Cr. The paramagnetismof the corresponding nickel compounds indicates that the bonds are ionicrather than covalent. In the dipyridyl complexes of tervalent iron orcobalt, the bonds are covalent, as are also the bonds in the stable octahedralcomplexes of the ditertiary arsine with Ni, Co, and Fe in their bi- and ter-valent states.2otetrahydrate in glacial solution.[Co(NH,),CO,]f + 2H+ = [Co(NH3),HZO]+++ + HZC03l9* J . A. Sharp and A. G. White, J., 1952, 110.197 J. P. Hunt, A. C. Rutenberg, and H. Taube, J . Amer. Chem. SOC., 1952, 74, 268.198 M. Polanyi and A. L. Szabo, Trans. Faraduy SOC., 1934, 30, 508.lg9 B. West, J., 1952, 3115, 3123.goo F. P. Dwyer, N. S. Gill, E. C. Gyarfas, and F. Lions, J . Amer. Chem. SOC., 1952,74, 4188. 201 F. H. Burstall and R. S. Nyholm, J., 1952, 3570108 INORGANIC CHEMISTRY.Potentiometric titrations in aqueous solution show that cobalt tetra-carbonyl hydride, CO(CO)~H, is a strong acid and that iron tetracarbonyldihydride, Fe(CO),H,, is a weak dibasicA new method has been described for the preparation of nitrosylcobalttricarbonyl Co(N0) (CO), and of dinitrosyliron dicarbonyl Fe(NO),(CO), ;it consists in the acidification of solutions of the alkali salts of the respectivecarbonyl hydrides in the presence of corresponding amounts of nitrite,=, e.g.Fe(CO),HNa + 2NaN0, + 3HOAc --+ Fe(NO),(CO), + 2CO + 3NaOAc + 2H20Fe(CO), + 3NaOH -+ Fe(CO),HNa + Na2C03 + H20Iron can be removed from solutions of ferric phosphate by either cation-or anion-exchange resins.Evidence has been obtained which indicates thatthe extraction of the iron as an anion is due to the formation of a complexion containing 3 phosphate groups to each iron atom, the phosphate groupspresumably behaving as bidentate groups, forming H3[Fe(HP0,)] orH,[Fe(P0,),].204What is claimed to be the first cationic iron(x1x) complex to be obtainedin enantiomorphic forms has been prepared bv the ceric ammonium nitrate-nitric acid oxidation of (+)- and (-)-tris-2 : 2'-dipyridyliron(n) perchlorates.The resolution of the latter compounds was carried out through the iodideantimony1 tartrate : (+)- and (-)-tris-1 : 10-phenanthrolineosmium(Ix1)perchlorates have also been prepared by oxidation of the correspondingosmium(xx) compounds by chlorine.205The stabilities of some 5-substituted 1 : 10-phenanthrolineiron(1x) com-plexes have also been determined.2o6A contribution has been made to the preparation of trinuclear rutheniumcompounds by that of the basic acetate, [Ru,(OAc),(OH),] (OAc),7H20,by the reduction of ruthenium tetroxide by acetaldehyde in anhydrousacetic acid-carbon tetrachloride solution.This compound dissolves inwater to give an intensely blue solution, and rapidly in pyridine to givea solution which becomes green when warmed. Addition of chloro-platinic acid to the aqueous solution of the pyridyl derivative precipitatesSexavalent ruthenium is commonly met with as an anion in the form ofalkali-metal ruthenates M,RuO,, which are stable in alkali but immediatelydisproportionate on treatment with acid, into the +4 and +S oxidationstates; no simple salts of the ruthenyl cation RuO,++ have hitherto beenisolated. The reduction of ruthenium tetroxide by a variety of reducingagents in sulphuric acid solution has given evidence that a sexavalent stateis capable of existence, as a green solution, in this medium, but decomposesin a few hours at room temperature, probably by the above disproportion-ation. The experiments suggest, however, that even in these solutions theruthenium is present as the anionic complex [RuO,(SO,),]' rather than asthe ruthenyl ion RUO~++.~O~A study has been made by spectrophotometric methods of the solution[ RU,(OAC),py,]ClPtC16.2072*2 W. Hieber and W. Hubel, 2. Naturforsch., 1952, 7, b, 322.203 F. Seel, 2. anorg. Chem., 1952, 269, 40.204 J. E. Salmon, J . , 1952, 2316.205 F. P. Dwyer and E. C. Gyarfas, J . Amer. Chem. SOC., 1952, 74, 4699.206 W. W. Brandt and D. L. Gullstrom, ibid., p. 3532.207 F. S. Martin, J . , 1952, 2682. 208 Idem, .I., 1952, 3055FAIRB ROTH E R. 109chemistry of ruthenium in the +6, +7, and +S oxidation states. The +7state, which is known in the solid compounds NaRuO,,H,O and KRuO,,has also been identified in aqueous solution, and its properties studied.20gSpectrophotometric techniques have also been used to study the form-ation of a number of complex compounds of ruthenium. The orange-coloured complex formed by the reaction of ruthenium(1v) perchlorate withthiosemicarbazide was identified as Ru[SC(:NH)*NH*NH,]+~, and the brightred complex with 4-phenylthiosemicarbazide as Ru[SC( :NPh)*NH*NH,] +2.Both the thiosemicarbazide and the 4-phenylthiosemicarbazide behaved asweak acids, liberating a hydrogen ion for each molecule of ligand whichentered the complex.210With a number of complexing ions, both ruthenium(1v) and ruthenium(II1)form the same coloured complex, the quadrivalent ruthenium being reducedin each case by the ligand before complexing occurs; i.e., with thiocyanate,the same deep blue [Ru(CNS)lf2, with thiourea the blue-green complexes,Ru[SC(:NH)*NH2lf2 and Ru[SC(:NH)*NH,],,~~~ and with dithio-oxamide, theblue-green complexes Ru[SC(:NH)*CS*NH,]+~ and Ru[SC( :NH)*CS*NH2]3.212A new series of a nitrosopentamminoruthenium(I1I) ion [Ru(NH,),*NO]+~has been prepared : the salts are diamagnetic.213A number of salts of tetracyanopalladic(11) acid have been studied : 214the acid, which is prepared by acidification of a solution of palladium(11)cyanide in excess of cyanide ions, forms insoluble salts with silver andcopper(I1) ions or with their ammines. It also forms normal salts withbenzidine, naphthaquinoline, and oxine, and the salts of nitron. Palladium(I1)cyanide itself also forms co-ordination compounds with a large number ofnitrogen-containing organic compounds.A number of fluoro-complexes of palladium and gold of the type M,{PdF,]and M[AuF,] (M = alkali metal) have been prepared, viz., Cs,PdF,, Rb,PdF,,K2PdF6, and KAuF,, by the fluorination of the corresponding chlorine com-pounds. All these compounds are bright orange to yellow and, in contrastto the corresponding chlorine or bromine compounds, are immediatelyhydrolysed by water. During the fluorination of the alkali-metal chloro-aurates red intermediate products-possibly mixed fluorochloroaurates-were obtained : these were more stable towards ~ a t e r . ~ l ~The reactions of NN-diethylglycine and N-ethyl-N-methylglycine withcobalt (111) and platinum(I1) have been studied, and dinitro-(N-ethyl-N-methylglycine)platinate(II) has been prepared ; this has been shown tocontain an asymmetric nitrogen atom through its resolution by fractionationwith (-)-quinine and treatment with optically active quartz powder.216F. FAIRBROTHER.20s R. E. Connick and C. R. Hurley, J . Amev. Chem. SOC., 1952, 74, 5012.210 R. P. Yaffe and A. F. Voigt, zbid., p. 5043.2 1 1 Idem, ibid., pp. 2500, 2503.212 Idem, ibid., p. 3163.213 K. Gleu and I. Buddecker, 2. anorg. Chem., 1952, 268, 202.214 F. Feigl and G. B. Heisig, J . Amev. Chem. Soc., 1951, 73, 5630.215 R. Hoppe and W. Klemm, 2. anorg. Chew,., 1952, $368, 364.216 J. R. Kuebler and J. C. Bailar, Jr., J . Amer. Chew Soc., 1952, 74, 3535