1362 CAVEN : THE MOLECULAR CONFIGURATION OF CXXXVIL-The Moleculay ConJiyuration of Yhosphoryl Chloride and its Derivatives. By ROBERT MARTIN CAVEN, D.Sc., F.I.C. THE constitution of phosphoryl chloride may be represented in two ways; either as ClOP:Cl, or OPCI,. I n support of the formula representing an unsymmetrical con- stitution in which the phosphorus atom would be tervalent, Thorpe (Trans., 1880, 37, 388) has adduced evidence based on specific volume determinations under the assumption that the atomic volume of phosphorus in all compounds is constant. Ramaay and Masson, on the other hand, have shown (Trans,, 1881, 39, 52) that this assump- tion is unwarranted, evidence deduced from other elements pointing to an alteration of atomic volume with change from single to double linking, Consequently no conclusion as to the constitution of phos- phoryl chloride can be drawn from the evidence brought forward by Thorpe.The chemical evidence, however, is altogether in favour of the symmetrical formula, and this is accepted in what follows. The problem OF the spacial configuration of the molecule of which the constitution is accepted must now be considered, and the question is whether or not the individual chlorine atoms, considered singly, are similarly situated to the rest of the molecule regarded as a whole. Whilst the chlorine atoms themselves are all alike, i t is conceivable that differences of situation, if these existed, would confer upon them different properties. identified as in the scheme Let us suppose that the chlorine atoms can be /GI, \GI, OP-CIp 0 Then any existing difference will be made manifest by substitution, for /Cia /Cia OP-Cl, , \R OP-GI, /R 9 OP--K , \c1, \C1, will all be different if all three chlorine atoms are differently situated within the molecule.Several monosubstituted derivatives have been prepared and examined with the view of discovering such isomerism if it should exist. For instance, monaniiinophosphoryl chloride, OP(NH*C,H,):C12, wasPHOSPHORYL CHLORIDE AND ITS DERIVATIVES. 1363 obtained in a yield amounting to 90 per cent. of the theoretical. This substance was fractionally recrystallised from benzene, and the melting points and crystalline characters of the successive fractions were com- pared with one another and with those of the original substance.No physical differences were detected, and so far as this evidence goes the compound may be pronounced homogeneous. The same was found to be true of mono-p-toluidinophosphoryl chloride, OP(NH*C6H,*CH,):C1,. Hence monosubstituted derivatives of phos- phoryl chloride only exist in one form. From this fact, one of two conclusions may be drawn : either (i) The single chlorine atoms are substituted indiscriminately by the reacting base, so that from different molecules either CI,, Clp, or C, may be displaced, and the result being the same, the three chlorine atoms must be similarly situated within the molecule; or (ii) A difference being supposed to exist between the chlorine atoms, the reacting base displaces from each molecule only that chlorine atom which occupies a particular position.Whilst it is probable that the first conclusion is true, it will be necessary, in view of proving the similarity of position of the three chlorine atoms, to accept the second conclusion as a working hypothesis. Thus if all the chlorine atoms were differently situated, and one of them reacted more readily with bases than the other two, the monosubstituted product would be homogeneous. Experiments were next performed with disubstituted products, the two bases aniline and p-toluidine being introduced in an order which in a second case was reversed. Thus anilino-p-toluidinophosphoryl chloride was prepared and its properties examined, and then p-toluidino-anilinophoaphoryl chloride was similarly obtained, and its physical characteristics compared with those of the first product.These two substances were thereby found t o be similar in appearance and crystalline structure j they melted a t the same temperature, and when mixed intimately the melting point of the mixture was the same as that of the constituents. Unless, therefore, differences exist which ordinary physical methods fail to detect, these two preparations are identical. Moreover, anilino-p-toluidinophosphoric acid and p-toluidinoanilino- phosphoric acid were obtained and their identity proved by the melt- ing point test. The question of the mobility of the groups within the compound, and the production of identity by intramolecular rearrangement arises here. But in bhe absence of evidence to show that substituents of such large mass as are present in the above compounds possess this mobility, the point nezd not be Further discussed.1364 CAVEN : THE MOLECULAR CONFIGURATION OF The two following types are therefore identical : and it may be concluded that the two chlorine atoms replaced from the a- and P-positions in the above reactions are similarly situated with regard to the rest of the molecule.The next inquiry has reference to the remaining chlorine atom, C1,. The method adopted consist8 in first replacing one chlorine atom, which may be supposed to be CI,, by an alcoholic residue, leaving the other two atoms to react with bases ; consequently ethoxyphosphoryl chloride was made to react with aniline, producing ethoxyanilinophosphoryl chloride, and this substance then yielded with p-toluidine a compound which is the ethyl ester of anilino-p-toluidinophosphoric acid.Thus : /O*CZW4 /0*C2H5 /0*C2H5 OP-Cl (p) + OP--NH*C6H5 + OP-NH*C,H5 (Y) \Cl \NH- C ~ H ~ . CH, In the reaction in the reverse way, ethoxy-ptoluidinophosphoryl chlor- ide and the ethyl ester of p-toluidinoanilinophosphoric acid were suc- cessively produced. Thus : /0'C2H5(a) /O0CZH5 /O* C, H, OP-C1 (p) --+ OP--NH*C,H,~CH, -+ OP-NH*C,H,*CH,. \Cl (Y) \Cl \NH-C,H, The last compound was proved by tbe melting point test to be identical with the ethyl ester of anilinop-toluidinophosphoric acid ob- tained previously. So : /RI (4 /R' (4 \RIII(Y) \El1 ( y ) OP--R11 (P) and OP-R,,,(P) are shown to be identical, and Clp and Cl, are therefore similarIy situated within the molecule POC1,. Now C1, and C1, have previously been shown to be similarly situated.Therefore, subject to the limitations referred to in the course of the above argument, the three chlorine atoms in phosphoryl chloride have been proved t o occupy similar positions in the spacial configuration of the molecule. From this fact the important conclusion follows that the five atoms of the molecule of phosphoryl chloride cannot lie in one plane; for if this were the case the central chlorine atom, in whatever angular con-PHOSPHORYL CHLORIDE AND ITS DERIVATIVES. 1365 figuration might be adopted, would necessarily occupy a unique position with reference to the oxygen atom. Thus me arrive a t the following conception for the configuration of the molecule of phosphoryl chloride : The centres of gravity of the three chlorine utorns lie at the angles of ccn eqailateral triangle; and if an imaginary line is drawn through the centre of this triangle and at right angles to i t s plane, the centres of gravity both of the phosphorus atom and of the oxygen atom are situated in this line.The question whether the phosphorus atom lies in the same pIane as the three chlorine atoms remains undecided. The configuration of the molecule of phosphoryl chloride is there- fore tet’rahedral, and differs from that of a molecule of which an atom of carbon is the centre principally in the existence of the doubly- linked oxygen atom. Compounds of the type 0 = PL-R;, , built up according t o the model ‘RIII described above, do not contain a plane of symmetry; they are 4 asymmetric ’ although containing a double linking.On theoretical grounds, it should therefore be possible to prepare optically active derivatives of phosphoryl chloride, although the presence of the oxygen atom renders it unlikely that basic derivatives of this type will be obtained. Phosphonium derivatives of the type PR,R,IR,I,R,,X, how- ever, have already been prepared by Michaelis (Annalen, 1901,315, SS), and an attempt made to resolve them, but without success. The question of the resolution of acids of the type OPR,RII*OH therefore remains to be considered. Experiments were conducted with two distinct types of compound, namely, anilino-p-toluidinophosphoric acid and methoxy-p-toluidinophosphoric acid. Alkaloidal salts of these acids were not found to be crystallisable, but bornylamine and menthyl- arnine salts were obtained and fractionally crystallised.I n no case were the different fractions found to possess different specific rotatory powers, and when the acids mere liberated from their salts they were always found to be inactive. It is intended t o pursue this investigation further, using other types of acids. Meanwhile, it may be suggested that non-activity is due to racemisation caused by the wandering of the acidic hydrogen atom in the following manner : yo\ . \OH This might take place even in aqueous solution of n salt by hydro- VOL. LXXXI, 4 Y1366 CAVEN : THE MOLECULAR CONFIGURATION OF lytic dissociation. some support to this idea of tautomeric change. substituted derivatives such as OP(NH*C,H,):Cl, and OP(NH*C6H,*C El3) :GI, may be dissolved in cold dilute potash and recovered unchanged by the addition of hydrochloric acid.This fact may be best explained by assuming that the following tautomeric change has taken place : One fact must here be referred t o which lends It is that mono- /NHR //NR \Cl ' \Cl HO*P--Cl 0:P-Cl -+- a feeble monobasic acid being thus produced which forms a salt with alkalis. If this is the true explanation of the behaviour of these compounds with alkalis, then it is the more easy to believe that a wandering of the acidic hydrogen atom may take place with disubsti- tuted acids, These considerations render it less likely that the acids of the type OPR2R,,*OH will be obtained in optically active forms. I n view of this, an attempt was made to combine menthylamine directly with methoxy-p-toluidinophosphoryl chloride, and so obtain a compound incapable of undergoing such tautomeric change. This attempt was not successful, for although reaction took place with a partial sepma- tion of menthylamine hydrochloride, no crystalline product could be isolated.E X P E R I ME NTAL, I.--Compozcnds containing Arylamino-groups. AniEinophosphoryZ Chloride, OP(NH*C6H,):C1,. This substance was obtained by A. Michaelis and Schulze (Ber., 1893, 26, 2939) by heating aniline hydrochloride on the water-bath with the theoretical quantity of phosphoryl chloride. After the evolu- tion of hydrogen chloride had ceased, a thick liquid remained from which the compound was crystallised. The method adopted in the present work', which possesses the advantage of being carried out a t atmospheric temperature, consists in slowly adding two equivalents of aniline contained in dilute benz- ene solution to one equivalent of freshly distilled phosphoryl chloride, also considerably diluted with benzene.Aniline hydrochloride at once separates in a pure white condition, and the filtrate from this, after the excess of benzene has been distilled off, yields brilliant, colourless crystals of the required derivative. The melting point of this compound is considerably affected by traces of impurity. I n the first experiments, ether was used as the solvent instead of benzene, and the product melted fairly sharply a tPEOSPHORYL CHLORIDE AND ITS DERlVATIVES. 1367 79*. This product, however, fumed in the air and was not quite pure.Michaelis and Schalze employed hot benzene and light petroleum as solvent, and obtained a product which did not fume in the air and melted a t 84'. The substance, prepared according to the method described above, was practically stable in the air and could be kept for weeks in an ordinary corked bottle without undergoing any considerable decom- position, It melted sharply a t 89- 90' after recrystallisation from benzene containing a little light petroleum. Analysis proved the compound t o be anilinophosphoryl chloride. The specimen melting at 89-90' was submitted to fractional re- crystallisation, and four successive fractions melted a t the following temperatures : 89-90', 88', 88-89', 89'. The highest melting point observed with this substance was 93-94', and this was in the case of a specimen obtained some months subsequently and crystallised from dilute solution in benzene after the addition of an equal volume of light petroleum, Under these circumstances, the compound presented the appearance of very fine, silky needles.p-Toluidinophosphoryl Chloridd, OP( NH*C,H,* CH,) : Cl,. According to Michaelis and Schulze (Zoc. cit.), this substance, prepared from p-toluidine hydrochloride and phosphoryl chloride, melts at 104'. It may be obtained in the same way as auiIinophosphory1 chloride, and when crystallised from benzene melted a t 107-108' ; fractional crystallisation from this solvent did not reveal any difference of melt- ing point. Subsequently, however, the observed melting point mas raised to 110-1 1 1' by recrystallisation from dilute solution in benzene after addition of much light petroleum.Action of Alkulis on the Monosubstituted Derivutives of Phosphoryl Chloride. Action of Ammoniu.-Michaelis and Schulze (Zoc. cit.) have observed that anilinophosphoryl chloride dissolves readily in aqueous ammonia, and they surmise that this solution contains the ammonium salt of anilinophosphoric acid. This, however, is not the case. When the ammoniacal solution is acidified with dilute hydrochloric acid, a white, scaly precipitate separates, which is shown to be anilinophosphamic acid, OP(NH*C6H,)(NH,)*OH. A quantity of this substance was obtained, washed thoroughly until free from chloride, by which means its bulk was considerably reduced, and dried, first on a tile, and then in a vacuum over sulphuric acid until its weight was constant.Analysis gave the following results : 4 ~ 21368 CAVEN : THE MOLECULAR CONFIGURATION OF 0.3878 gave 0.2500 Mg,P207. 0.1029 ,, P = 17.97. 14.3 C.C. moist nitrogen a t 11" and 746 mm. N = 16.29. Ontitration, 0.3600 neutralised20~80c.~.N/lO alkali. Mol. wt. = 173.1. C,H,O,N,P requires P = 18.02 ; N = 16.28 per cent. Mol. wt. of C,H90,N2P = 172. Anilinophosphamic acid crystallises in rhomboidal plates which are sparingly soluble in cold water and melt a t 157-158". p-~oZuidinophospi2anzic acid, OP(NH*C,H,*CH,)(NH,)~OH, may be obtained in the same way as anilinophosphamic acid by the action of aqueous ammonia on p-toluidinophosphoryl chloride and decomposition of the salt by hydrochloric acid.It is obtained in scales, or, if separated slowly by acidifying a dilute solution of its ammonium salt, in coarse prisms, and melts a t 159" : 0.4531 gave 0.2747 Mg2P207. 0.1166 ,, 15.1 C.C. moist nitrogen at 10" and 749 mm. N = 15-30. P = 16.89. C7H1,02N,P requires P = 16.67 ; N = 15.05 per cent. On titration : (i) 0.4624 neutralised 24.75 C.C. XI10 alkali. Mol. wt, = 186.8. (ii) 0,3888 ,, 20.85 C.C. N/lO ,, Mol. wt. =186*5. Action of Potash.-When either anilino- or p-toluidino-phosphoryl chloride is added little by little to an ordinary dilute solutinn of potash, the compound rapidly dissolves. If the liquid is cooled under the tap during the addition of the monosubstituted phosphoryl chloride and dilute hydrochloric acid added in excess when solution is complete, an oil is precipitated which quickly solidifies.This solid in each case was separated and crystallised from benzene with the addition of light petroleum. The product recovered from monoanilinophosphoryl chloride me1 ted at 93" ; that from monotoluidinophosphoryl chloride at 11 1-1 12'. Both substances contained chlorine, and an analysis of the former product gave the following result : C1= 33.83. C,H60NCl,P requires C1= 33.75 per cent. 3101. wt. of C7H,,02N,P =186.0. 0.2635 gave 0.3609 AgCl. It is evident therefore that the above monosubstituted phosphoryl chlorides can be dissolved in dilute potash solution and recovered unchanged by the addition of hydrochloric acid. If no precautions are taken t o cool the potash solution during the addition of the monosubstituted chloride, considerable heat is developed and the solution turns yellow.In this case, the liquid contains traces of the free base, as was proved in the case of aniline by the bleachingPHOSPHORPL CHLORIDE ASD ITS DEKlVATIVES. 1369 powder reaction. The addition of hydrochloric acid after a time causes no immediate precipitate ; it appears therefore that the monosubstituted phosphoric acids are readily soluble in water. Attempts to isolate these acids proved unsuccessful, and since partial. decomposition by potash was proved, the investigation was not carried further, Anilirzo-p-tolzlidinop~osphoryt ChZoride, OP( NH*C,H,) (NH* C6H,*CH3)C1. This substance was prepared, together with the similar substance obtained by allowing the substituents to react in reverse order, for the purpose of proving the identity of the a- and P-positions of the chlorine atoms in phosphoryl chloride.It was obtained in very fine needles which, after two recrystallisations from benzene, melted sharply at 133-134'. 0.7042 gave 0.3644 AgCl. The yield of the substance was very poor, not more than about 2 grams being obtained from 10 grams of phosphoryl chloride. After the first crop of crystals had been removed from the benzene solution, further standing produced a small crop; but soon a gummy mass separated which appeared to contain most of the material, but from which nothing was isolated at this stage. p-ToZ~id~~occnilino~~osp~o~~2/2 Chloride, On analysis : C1= 12.78. C,,H1,ON,CIP requires C1= 12.65 per cent.OP( NH* C6H4* CH3)(NH* C,H,) C1. -This compound was prepared in a similar manner to the above, and an intimate mixture of the disubstituted chlorides prepared in the two ways melted a t 133-134'. These two substances are therefore shown t o be identical. AniZino-p-toZzLi~ino~~~os~~~oiiic Acid, OP( NH. C,H,)( NH* C,H,*CH,) OH. When anilino-ptoluidinophosphoryl chloride is digested for a short time with dilute sodium carbonate solution, it is completely dissolved. No odour of aniline or p-toluidine can be detected, and after cooling and allowing to stand for some time the solution remains perfectly clear. When this clear solution is acidified with dilute hydrochloric acid, a curdy, white precipitate of anilino-p-toluidinophosphoric acid at once separates. The method of recrystallisation employed was similar to that sug- gested by Autenrieth and Rudolph (Ber., 1899,32, 2099) €or dianilino- phosphoric acid, and was as follows.The crude precipitated acid was digested on the water-bath with acetone, the hot solution filtered, and diluted with about one-fifth of its volume of water to which a few drops1370 CAVEN : THE MOLECULAR CONFIGURATION OF of strong hydrochloric acid had been added. On stirring the solution and allowing it to cool, a, good yield of crystals was obtained in the course of a few minutes. It softens at 134O, and if not quite pure melts completely, but quickly becomes solid again, and melts finally at 195-196', turning somewhat brown. Anilino-p-toluidinophosphoric acid crystallises in shining scales.On analysis of the air-dried crystallised acid : 0-2456 gave 0.1028 Mg2P207. On titration : (i) 0.4674 neutralised 18.05 C.C. 3/10 alkali. (ii) 0.3595 ,, 13-85 C.C. N/10 ,, Mol. wt. =259*6. I n connection with the possession of a dual melting point by this acid, it is interesting to notice that, according to Rudert (Ber., 1893, 26, 565) ditoluidinophosphoric acid melts at 124', but according to Autenrieth and Rudolph (Zoc. cit.) a t 195'. These later authorities attribute the low melting point observed by Rudert to his acid not being crystallised or free from water. I n order to clear up this discrepancy, a specimen of ditoluidinophos- phoric acid was prepared and crystallised from dilute acetone. It then softened at 148' and melted a t 193-194", turning brown.Dianilinophosphoric acid, on the other hand, melts, according to Michaelis and Schulze (Ber,, 1894, 2'7, 2574) a t 313", and according to Autenrieth and Rudolph (Zoc. cit.) at 214-216'. A specimen prepared and crystallised as above melted a t 213O with- out previous softening. It appears therefore that the possession of a dual melting point by two of the above disifibstitnted acids is due in some way t o the presence of a p-toluidine residue within the molecule. p-Toluidinoanilinophos~horic Acid. -This acid was prepared from phosphoryl chloride by the successive action of p-toluidine and aniline in the manner above described. The product was identical in physical properties with anilino-p-toluidinophosphoric acid, softening a t 134" and melting at 195-196'. An intimate mixture of the two substances softened a t 134" and melted at 195-1969 These two products are therefore identical.P = 11-67. C,,H,,O,N,P requires P = 1 1 *83 per cent. Mol. wt.=259. Mol. wt. of C,,H,,O,N,P = 262. II.--Compozcnds containing an Etiioxygrouy. "he purpose for which these compounds were investigated was in order that the identity of the /3- and y-positions of the chlorine atomsPEOSPHORYL CHLORIDE AND ITS DERIVATIVES. 1371 might be demonstrated, the a-position being supposed to be occupied by the ethoxy-group. Ethoxyphosphoryl chloride, OP(O*C,H5):C1, (Jcchresber., 1876, 205) formed the starting point for the preparation of the compounds described below. Ethoxyadinophosphoryl Chloride, OP(0 C,H,) (NH C6H,)C1.-This compound was prepared from ethoxyphosphoryl chloride by reaction with aniline in ethereal solution ; it crystallises in triincated pyramids and melts a t 61-629 It is much more soluble in the ordinary sol- vents than anilinophosphoryl chloride, and cannot be so easily obtained in a state of purity.0.4495 gave 002873 AgCl. C1= 15-79. C,H,lO,NCIP requires C1= 16.17 per cent. Decompo&tiort, by Wccter.-The compound was dissolved in a little alcohol, which has no perceptible action upon it, and water added to the solution, The substance separated again in the solid state, but on standing slowly dissolved. The clear solution, when evaporated over sulphuric acid a t the ordinary temperature, gave leafy crystals which were proved to be aniline hydrochloride. The following probably represents the course of the reaction with water : On analysis : /O*C,H, H,O /O*C,H5 H20 /O*C,H, \Cl \OH \OH OP-NH*C6H, -+ OP-NH*C,H, j OP-OH + HC1 + C6H,*NH2,HCl, but ethoxyanilinophosphoric acid, the intermediate product of hydro- lysis, could not be isolated.Barizcm XaZt.-This and other barium salts of disubstituted phos- phoric acids were prepared in order to furnish material for experiments on resolution in conjunction with optically active bases. Although ethoxysnilinophcsphoric acid cannot be obtained by the bydrolgeis of its chloride by water, its barium salt was prepared by the use of baryta solution. This compound, when free from barium chloride, crystallised from ethyl alcohol containing a little water in minute, slender needles which, when dry, appeared as a chalky powder.The salt thus obtained was anhydrous, and underwent no change when heated to 130°, but at a somewhat higher temperature decomposed, giving off aniline. On analysis : 0.4777 gave 0.2086 BaSO,. Ba = 25.68. 0*3616 ,, 17-2 C.C. moist nitrogen at 16' and 758 mm. N=5-53. C,H,,O,NP,Ba requires Ba = 25.57 ; N = 5-21 per cent. The amide, OP(O.C,H,)(NH*C,H,)*NH,, wcis obtained by passing1372 CAVEN : THE MOLECULAR CONFIGURATION OF dry ammonia gas into an ethereal solution of the chloride, and crystal- lised from hot water in beautiful, shining prisms melting at 127". On analysis : 0,6653 gave 0.3768 Mg2P20,. C,H,,0,N2P2 requires P = 15.50 per cent. The aqueous solution of this amide was neutral in reaction to litmus. The compound dissolved easily in warm dilute acids, but suffered decomposition in the process, the solution containing aniline and ammonium salts.P = 16-78. Ethyl anilino-p-toluidinophosphate, OP( 0 C,H,) (NH* C,H,) N HwC,H,* C H,, was prepared from ethoxyanilinophosphoryl chloride and p-toluidine. It crystallisecl from dilute alcohol in colourless needles and melted a t 11 6-1 17". On analysis : 0.4318 gave 0.1678 Mg,P20r. Ethox?l-p-toZuidinophosphory I Chloride, P= 10.83. C,,H1,0,N2P requires P = 10.69 per cent. OP(O*~,H,)(NH0~,H,*~H,)~1. -This substance crystallises in oblique prisms and melts a t 74-75". On analysis : 0.2830 gave 0.1709 AgCI. CgHl,O,NOIP requires C1= 15.20 per cent. Barium &'&-The barium salt of this acid was obtained from its chloride in a manner similar to that employed in the previous case.The salt is, however, less soluble in water than its aniline homologue, and separates during evaporation of the solution in radiating, tuft-like masses. The compound may thus be separated from the mother liquor containing the excess of barium chloride, and after three crystallisations is pure. C1= 14.92. On analysis : 0.6063 gave 0.2427 BaSO,. 0.3160 ,, 14.1 C.C. moist nitrogen a t 17" and 760 mm. N = 5 * 1 8 . The salt was also recrystallised from alcohol containing a little 0.6457 gave 0.2670 BaSO,. Ba = 24.32. Ba = 23.54. [CgH1,O,NP],Ba,H,O requires Ba = 23.55 ; N = 4.77 per cent. water, and obtained in the form of long, slender needles : [C911,30,NP],Ba requires Ba = 24.30 per cent. Thus when the least possible amount of water is used in its crystal- lisation the salt is obtained in an anhydrous state.The amide, OP(O* C,H,)(NH*C,H,*CH,)*NK,, was prepared in the same manner as the analogous aniline derivative. It was not foundPHOSPHORYL CHLORIDE AND 1TS DERIVATIVES. 1373 suitable to crystallise it from hot water, but when a mixture of ethyl acetate and light petroleum was employed, the substance separated in rhomboidal plates melting a t 125'. On analysis : 0.1726 gave 20.0 C.C. moist nitrogen at 16' and 751 mm. N=13.36. CSH150,N,P requires N = 13.09 per cent. This amide is sparingly soluble in cold and easily soluble in hot Ethyl p-toluidinoaniiinophosphate, water. Its solution is neutral in reaction. OP(O* C2H,)(NH*C,H,*CH,)*NH0C6H5, crystallised in needles from dilute alcohol. It presented the same appearance as the product described on p.1372 ; it melted a t 117'. An intimate mixture of ethyl anilino-p-toluidinophosphate and ethyl ptoluidinoanilinophosphate was made and melted a t 116-1 17'. These two products are therefore identical. II I.-C o mp o u n d s c o n t a i n i n g a M e t h ox y-g r o up. Methoxyphosphoryl Chloride, OP(O-CH,):Cl,. This simple substitution product of phosphoryl chloride appears never to have been previously prepared, although it may be obtained in the same way as the ethoxy-compound. It was found to distil at 62-64' under 15 mm. pressure. It is a colourless liquid which fumes in the air, but is not so vigorous in its reaction with water as phosphoryl chloride. For analysis, a weighed quantity was decomposed by dilute potash, the solution acidified with nitric acid was then precipitated with silver nitrate : 0.2812 gave 0.5400 AgCl.This compound cannot be distilled under atmospheric pressure. C1= 47.43. CH,02C12P requires C1= 47.65 per cent, Methoxyanilinopho~phor yl Chloride, oP( 0 cH,)(NH*C,H,)Cl. This compound was prepared by bringing together aniline and methoxyphosphoryl chloride in benzene solution. The product was found to be very soluble in benzene, and could only be obtained as a pasty mass by evaporating this solvent. A small quantity, however, was obtained in a pure condition by heating the mass with light petroleum of high boiling point and allowing the solution t o cool slowly. Methoxyanilinophosphoryl chloride cry s tallises in thick needles which melt at 82-83".It is slowly but completely dissolved1374 CAVEN : TEE MOT,ECULAR CONFIGURATION OF by dilute potash, and the solution does not give the aniline reaction with bleaching powder, N = 7.07. On analysis : 0.3258 gave 20.5 C.C. moist nitrogen at 21° and 749 mm. Barium Salt.-This salt was obtained by digesting the crude chloride with baryta solution and crystallising the product, The salt is con- siderably less soluble in water than barium chloride, from which it can easily be separated by crystallisation. It crystallises in beautiful, transparent needles quite different in appearance from either of the ethoxy-salts previously described, and the air-dried salt, when heated at looo, gives off much water. The salt contains seven mols. of water of crystallisation : C7H,02NC1P requires N = 6.81 per cent.0.3912 gave 0.1512 BaSO,. Ba=22*75. 0.2907 ,, 12.2 C.C. moist nitrogen at 23' and 763 mm. N = 4.75. [C7H90,NP],Ba,7H20 requires Ba = 23.62 ; N = 4.61 per cent. I n an attempt to estimate the water of crystallisation, 09466 gram of the salt lost 0.0418 gram at 130°, and on heating t o 155-160' for 15 mins. an additional 0,0169 gram; but in the latter case the salt turned somewhat yellow and underwent slight decomposition. If 7H20 are present, 0,2466 gram contains 0.0512 gram of water. Whilst the water cannot be directly estimated owing t o decomposition, these results confirm the previous analytical figures which indicate the presence of 7H,O in the salt. Methoxy-p-toluidi~op~osp~oryl Chloride, OP(O*CH,)( NH*C,H,*CH,)Cl.This substance, obtained by the interaction of methoxyphosphoryl chloride and p-toluidine in benzene solution, is characterised by a less degree of solubility in benzene and other organic solvents than the mixed es ter-amino-substit u t ed phosphoryl chlorides previously descri bed. It crystallises in prisms from its solution in hot benzene with the utmost ease, and can therefore be obtained pure in any desired quan- tity, and melts at 115-1169 05197 gave 0.3409 AgC1. This compound is easily soluble in dilute alkalis with the formation of the corresponding salts of methoxy-p-toluidinophosphoric acid. The barium salt is much less soluble in water than any of the barium salts previously described, and on cooling its concentrated solution separates in shining needles which, after one recrystallisation, are free from barium chloride.When the air-dried salt is gently heated, it gives off much water. On analysis : C1= 16.20. C,H,,O,NClP requires C1= 16.15 per cent.PHOSPHORYL CHLORIDE AND ITS DERIVATIVES. 1375 (i) 0.7463 gave 0.2711 BaSO,. Ba = 21.38. (ii) 1.0101 ,, 0.3644 BaSO,. Ba=21*23. (i) 0.3490 ,, 12.9 c.c.moist nitrogen at 21' and '762 mm. N =4 22. (ii) 0.41 12 ,, 15.6 C.C. ,, ,, 20.5' ,, 766 mm. N = 4.37. [C,H1,0,NP],Ba,7H,0 requires Ba = 21-62 ; N = 4.40 per cent. 0.9444 gram of the crystallised salt lost 0.1685 gram a t 130-140'. On raising the temperature further, the solid became yellow and began to decompose, This loss is 17-83 per cent., and represents a little more than 6 mols. of water. Thus the salt contains '7 mols.of water of cry stalli sa t ion, Potassium SuZt.-Some of the barium salt was decomposed in solu- tion by its equivalent of potassium sulphate. On evaporation to small bulk, the potassium salt was obtained as a mass of fine needles which are very soluble in water. Attempts to obtain the free acid from its salts were unsuccessful. IV -Attempts t o resolve Asymmetricully Xuhtitzcted Phos- p h o r i c Acids into Optically Active Constituents. Anilino-p-tolzcidinophosphoric Acid. A quantity of this acid was prepared and handed to Professor -Kipping, who tried to resolve it by forming salts with optically active bases. Bornylamine and menthylamine salts mere obtained in a crystalline form and fractionally crystallised. The acid was recovered from the various fractions by decomposing the salt with dilute mineral acid, and was examined polarimetrically. It was in all cases found to be inactive. Me thox y- p- t ol zcid inophosp h o ~ i c Acid. This acid was chosen as the most suitable of the above described ester acids because of the slight solubility of its barium salt in cold water and the comparative ease with which it can be prepared. Menthylamine Xu1t.-A sample of menthylamine, prepared from men- thone by reduction of its oxime, and which had been proved to be homo- geneous, was employed, A neutral solution of the sulphate was used to precipitate barium methoxy-p-toluidinophosphate. The precipitation was carried out a t about 6O0, and the solid which separated was very bulky in appearance, the barium sulphate having carried down with it some of the crystallised menthylamine salt, which was removed by washing with hot water, On evaporation of the neutral filtrate on the water- bath, the menthylamine salt soon began to separate in filmy t u f t s of fine needles on the surface of the liquid, and by successive removal of these by filtration, the salt could be separated into fractions. On analysis of the air-dried salt :1376 MORGAN: INFLUENCE OF SUBSTITUTION ON THE 0.5696 gave 0.1852 Mg,P,O,. P= 9.06. 0.2700 ,, 18.9 C.C. moist nitrogen at 22' and 765 urn. N = 7.99. Polarimetric examination of these fractions did not reveal any Moreover, a specimen of alkali salt recovered from a fraction of the Thus the attempt to resolve methoxy-p-toluidinophosphoric acid by C18H3303N2P requires P = 8-71 ; N = 7.87 per cent. difference of rotatory power, and their melting points were all alike. menthylamine salt was quite inactive. means of menthylamine was not successful. The author desires to thank Professor Kipping for valuable sugges- tions made especially during the earlier part of the research. UNIVERSITY COLLEGE, NOTTINGHAM.