CONDEXSATIONS OF SODIUM D E R I V A ~ S m. 2759 CCCLXX1X.-Condensations of the Sodium Derivatives of Trimethylene Glycol and Glycerol. By ARTHUR FAIRBOURNE and GRAHAM EDWARD FOSTER. IN extension of the work on the condensation of l-chlor0-2:4-dinitrobenzene with glycerol (J. 1921 119 1035) and ethylene glycol (ibid. p. 2077) the similar condensation in the cam of trimethylene glycol has now been investigated. A solution of the sodium derivative is readily obtained by dissolving sodium in excess of trimethylene glycol and may be used for effecting the condensation. In spite of the large excesa of glycol present and although monosodium glyceroxide gives rise to onl 2760 FAIRBOURNE AND FOSTER CONDENSATIONS OF THE one product in the correspondmg reaction (Zoc. cit. p.1037) both mono- and bis-dinitrophenyl ethers of trimethylene glycol are produced even when the chlorodinitrobenzene is added slowly. 5!H,mOH 9H2*OH ~H2'0*C6H,(N02)2 CH -+ (1.) YH + $ 3 3 2 (11.1 bH,*OH CH2- O*C,H,( NO,) CH,*O*C,H,(NO,), The relative yields of these ethers vary with the concentration of the sodium derivative and therefore may be regarded as dependent on the tendency to set up the equilibrium ~H,-O*C,H,(NO,) FH,*OH ~H,*O*C,H,(NO,) F;H,*OH CH,*OH CH2-ONa C€€,*ONa CH,*OH it being assumed as in the two previous papers that the primary condensation involves merely the replacement of the sodium by the substituent group : On this assumption depend the arguments previously put forward in support of the structure of glycerides (Zoc. cit. p. 1035).Doubt was cast on its validity however by the accidental preparation of the bisdinitrophenyl ether above mentioned and by the further discovery that tri-substituted derivatives of glycerol can be obtained by treating monosodium glyceroxide suspended in an inert solvent, with one equivalent of certain acid chlorides a reaction not con-templated in the previous papers. QH,*OE vH,*S03Ph F]H,*OH CH,*ONa CH,*SO,Ph CH,*OH Moreover such trisubstituted products were obtained from disodium glyceroside under identical conditions : p 2 -1- YH2 <t vH2 + y"2 R*O-Na + C1*C6H3(No2) -+ R*O*C,H3(N02),. 3VH-OH + 3Ph*S02C1 + yH*SO,Ph + 2YH*OH FH,*CO,Ph 7H2*OH bH,-CO,Ph CH,*OH 3C,H,O&a + 6PhCOC1 * 2 'H*CO,Ph + VH*OH It w a ~ rmlised that the formation of these substances might be due to direct interaction of labile chlorine atoms and hydroxyl groups, yH,*ONa 5!H200 *C6H3(N02) CH + 2C1*C,H3(N0,) -+ CH, &**OH 6H2*O*C,H3(NO,) ' the excess of the sodium derivative merely serving to destroy the hydrogen chloride thus formed.Glycerol ethylene glycol an SODIUM DERIVATIVES OF TRIMETHYLENE GLYCOL ETC. 2761 trimethylene glycol were therefore heated separately with l-chloro-2 4-dinitrobenzene in presence of calcium carbonate and tri-methylene glycol mono-2 4-dinitrophenyl ether was treated similarly in presence of sdEcient inert solvent to dissolve all organic matter but in no case was there a trace of ionisable halogen in the product. It appears clear therefore that no condensation between halogen atoms and hydroxyl groups occurs during the reactions with chlorodinitrobenzene and that the chlorine removes the sodium in the may previously assumed.The production from sodium glyceroxide of the tri-ester mentioned above might be explicable in the light of Fischer’s observation (Ber. 1920 53 1621 1634) that mono-esters of polyhydric alcohols in ethereal solution in presence of a catdyst such as potassium carbonate or sodium ethoxide are moderately rapidly transformed into di-esters and free alcohols. A similar explanation is untenable, however in the cam of trimethylene glycol bisdinitrophenyl ether, for when the mono-ether was treated with a solution of sodium in trimethylene glycol under the conditiom existing during the condemtion with chlorodinitrobenzene no bis-ether was found in the product the very different solubilities of the mono- and bie-ethers in acetic acid rendering the absence of the latter easily provable.The a-structure of monosodium glyceroxide (loc. cit. p. 1036) has been codrmed by the following series of reactions using isopropyl-idene glycerol (Irvine Soutar and Macdonald J. 1915 107 337; %her Ber. 1920 53 1589) X denoting p-C6H4*N0 or 3 5-C,H,(NO,), YE[,=OH YH-OH (111.) CH,*CO,X SCOCl YH,-OR YH-OH -GG+ YH,*OH ( CH,*ONa VH-OH h-/?-HCl-f E X P E R I M E N T A L. Trimethylene Glycd Mono-2 4-dinitrqhenyl Ether (I).-Sodium (1.14 g . ) WBB dissolved in small portiom to prevent charring in 40 C.C. of trimethylene glycol and the solution waa stirred and heated a t 100” while a suspension of 10-2 g.of 1-chloro-2 4-dinitrobenzene in 30 C.C. of the glycol waa gradually added. Heating and atirring were continued until t,he chlorodinitrobenzene dissolved (4 hour). The product wm poured into ZOO0 C.C. of 20% acetic acid an oi 2762 FAEELBOURNE AND FOSTER CONDENSATIONS OF THE being precipitated. The whole was boiled for 30 minutes; most of the oil had then redissolved and the remainder had crystallised (A). The solution waa filtered and cooled to 0" after 24 hours. The colourle~ needles deposited were dried in a vacuum over sulphuric acid and recrystallised from benzeneligroin (b. p. 4-0-60"). The ether thus obtained (yield 6 g.) m. p. 52" was soluble in ether, benzene alcohol acetone or acetic acid and sparingly soluble in water or ligroin (Found N 116.C&€,,,O,N requires N 11.55%). The ace@ derivative prepared by heating the ether for a few minutes with 2 vols. of acetyl chloride and acetic anhydride (equal parts) and boiling the product with much water separated from the filtered solution in long colourless needles m. p. 85" the yield being nearly quantitative (Found C 46.1 ; H 4.45 ; N 9.9. Cl1H1,O,N, requires C 46-4 ; H 4.25 ; N 9.9%). The benzoyl derivative was prepared by shaking a warm mixture of the ether and benzoyl chloride with successive small quantities of 20% sodium hydroxide solution added in sufficient amount fo prevent the orange colour from being discharged; water was then added. The benzoyl derivative obtained as a semi-solid mass on cooling the mixture separated from alcohol in colourless crystals, m.p. 95" soluble in most of the common organic solvenfs (Found : N 8.2. Cl6HI40,N2 requires N 8011%). Trimethylene Glycol Bis-2 4-dinitrophenyl Ether (=).-The substance (A) mentioned above was sparingly soluble in alcohol, ether or acetone but by recrystallisation from boiling glacial acetic acid the bis-etk was obtained in colourless needles (yield 0.5 g.) m. p. 180" (Found C 44.0; H 3.2 ; N 13.4. C15H12010N4 requires C 44.1 ; H 3.0; N 13.6%). Glyceryl Tri-p-toluenesulphonate C,H5(0*S02*C7H7),.-Disodium glyceroxide was prepared by the method of Lobish and Loos (Mon&h. 1881 2 842) (Found Na 33.0. A mixture of equimolecular proportions of either mono- or di-sodium glyceroxide and p-toluenesulphonyl chloride in dry ether or benzene was kept for 24 hours then boiled under reflux for hour, and filtered.The glyceride obtained on distilling off the solvent slowly solidified and crystallised from alcohol in colourless needles, m. p. 103" (Found C 51.6 51-8; H 4.8 4.8; S 174; M cryo-ecopic in benzene 564. CeaH,609S requires C 52.0; H 4-7; S 17.3%; M 554). Glyceryl tribenzenesulphonate obtained in a similar way from benzenesulphonyl chloride crystallised from alcohol in needles, m. p. 80" (Found C 49.0; H 3.9. C2,H,0gS requires C 49.2; H 3.9%). Action of Benzoyl CAloride upon Sodium G1yceroxides.-The Calc. Na 3343%) SODIUM D E R X V A ~ S OF TRIMETHYLENE GLYCOL ETC. 2763 producta obtained by the action of an ethereal solution of benzoyl chloride on mono- and di-sodium glyceroxides were not identical.Monosodium glyceroxide yielded a-monobenzoyl glycerol (compare loc. cit. p. 1340). Disodium glyceroxide yielded tribenzoyl glycerol which even after repeated crystallisation from ligroin melted at 7 1 7 2 ' (Found C 71.25 ; H 4-95. Chlc. C 71.3 ; H 4.95%). The m. p. is given by Skraup (Nonatdi. 1889 10 393) as 75-76*5" by Balbiano (Ber. 1903 36 1574) as 71-72" and by &a& (itid., p. 4341) as 76". A comparison of these references and others showed that the substance melted at 76" or 72" according as it had been crysfallised from alcohol or ligroin. The above product was there-fore c r y s t a b d from alcohol. It then melted at 76". Subsequent recrys-tion from ligroin (b. p. W o o ) or slow solidification of the fused material reduced it to 72" again.Action qf p- A'itrobenzuyl CMOnde upon Monosodium G1yceroxide.-Monosodium glyceroxide was treated with an equivalent quantity of p-nitrobenzoyl chloride in ethereal solution and the reaction carried out in the usual manner. The product obtained by filtration and evaporation of the solvent was an oil which solidified on cooling and c r y a w from chloroform in needles m. p. 107". A mixture of this with a specimen of a-mono-p-nitrobenzoyl glycerol (m. p. 107") prepared by Fischer's method (Be. 1920 53 1596) also melted at 107'. In the preparation of isopropylidene glycerol as an inter-mediate product in the second method the resulting liquid boiled at 78-79"/10 mm. isoPropylidene Glycerol 3 5-Dinitrobenzocate (IV).-A mixture of 11.6 g.of 3 5-dinitrobenzoyl chloride and 8-2 g. of dry quholine WM dissolved in 15.6 C.C. of chloroform 6.7 g. of isopropylidene glycerol were slowly added with cooling and the whole was kept for 2 days. The product was shaken successively with water, dilute sdphuric acid sodium bicarbonate solution and water then dried over ignited sodium sulphate and the solvent removed. The residual oil solidified on cooling and crystallised from hgroin (b. p. M O O ) -benzene in colourless needles (yield SO%) m. p. 85", eoluble in benzene ether or chloroform and nearly insoluble in ligroin (Found N 8.6. Glycerol a-Mono-3 5-dinitrobenzacte (ID).-(a) A mixture of equivalent quantities of monosodium glyceroxide and 3 5-di.nitro-benzoyl chloride in ether was kept for 2 days and then heated under reflux.The suspended matter which had little or no a)kdine reaction was filtered off and the solvent evaporated. The reaidurtl oil which solid.ified crystallised from chloroform in colourleas needles m. p. 118" soluble in alcohol ether or benzene; it also C1&€140$2 requires N 8.6%) 27M MACDONALD AND HINSHELWOOD TBE FORMATION AND crystallised easily from water (Found N 9-75. C,oH,oO,N, requires N 9.79%). (b) A mixture of 5 g. of isopropylidene glycerol 3 5-dinifro-benzoate and 100 C.C. of N/2-hydrochloric acid was stirred and heated at 70-80" for 1 hour. The clear liquid was decanted and deposited glycerol a-dinitrobenzoate on cooling. This after being dried in a vacuum over sulphuric acid and crystallised from chloro-form (yield 700/,) melted at 118" alone or mixed with a specimen prepared by method (a). The authors wish to express their indebtedness to the Chemical Society for a grant to one of them (A. F.) which has partly defrayed expenses incurred in this work. Kx~a's COLLEQE LONDON W . C . 2. [Received A?tg?ist 7th 1925.