2800 LE SUEUR AND WITHERS: THE MECHANISMCCLXII1.-The Mechanism of the Action of FusedAlkalis. Part I. The Action of Fused PotassiumHydroxide o n Dihydroxystea& Acid andDihyclrozy behenic Acid.By HENRY RONDEL LE SUEUR and JOHN CHARLES WITHERS.ONE of the stages in a method f o r the determination of the consti-tution of fatty acids (Crossley and Le Sueur, T., 1899, 75, 161;1900, 76, 83) consists in the oxidation of the dihydroxy-derivativeof the acid under investigation by means of chromic acid, and itwas pointed out a t the time that when dealing with a higher fattyacid the insolubility of its hydroxylated derivative might renderthis oxidation a difficult operation. It was thought that fusion ofthe hydroxylatsd derivative with potassium hydroxide mightreplace the oxidation with chromic acid, and in order to investi-gate this point one of us (Le Sueur, T., 1901, 79, 1313) fuseddihydroxyst earic acid with potassium hydroxide in the hope t'hatoxidation would take place thus:CH,*[C&],*CH(OH)*CH(OH)*[CH,],*CO,H +The results obtained, however, were quite different, and it wasshown that the main product) of the action was a monohydroxy-dicasboxylic acid o i the formula CI8H3,O5.The present investiga-tion deals with the constitution of this acid and also of a similaracid obtained by the fusion of dihydroxybehenic acid with potassiumhydroxide.When the acid C,8H3405 is oxidised with potassium permanganatein acetone solution i t is quantitatively converted into a ketomono-carboxylic acid having the formula C17H3203; and when it is heatedCH3*[ CH& C0,H + C O,H* [ CH,],*CO,HOF THE ACTION OF FUSED ALKALIS.PART I. 2801aIone and above its melting point, it readily loses water and twoisomeric unsaturated dicarboxylic acids, C,,H,,O,, are produced.These facts led us t o suppose that the acid ClSH3,O5 was aP-hydroxy-acid having the constitutionCO,H*CH,* CH (OH). [CH,] 14*C02H.Such an acid would give on oxidation the ketonic acid,C02H*CH2*CO*[CH,],,*C02H (I),and the latter would readily lose carbon dioxide t o give the acid,CH3*CO*[CH,]14*C0,H. Although the free acid (I) possibly couldnot be isolated, one would expect that its ethyl ester would besufficiently stable t o admit of this being done. We accordinglytreated the ethyl ester of the acid, C,,HsO,, with oxidising agents,but in spite of a large number of experiments we were unable t oobtain the desired est'er; either no oxidation took place o r if it didthen carbon dioxide was lost simultaneously, and the acid C,,Ha03was produced.We then decided to determine the constitution ofthis ketonic a i d , and for this purpose it was treated with hydroxyl-amine, and the resulting oxime converted into the correspondingamides by heating with concentrated sulphuric acid, and theseamides were hydrolysed by heating with hydrochloric acid. Theproducts of the hydrolysis were identified as n-nonoic acid, octyl-amine, azelaic acid, and 7-amino-izroetoic acid. From this itfollows that the ketonic group is in the middle of the chain, andthat the above products are formed as follows:NH OH CH,*[CH,],*CO*[CH,],*CO,H --?+CH,*[CH,]7*C(NOH)*[CH,]7*C0,HThe' position of t2he carbonyl group and, of course, of the CH-OHgroup wliicli gives rises to it, having been established, it nowremained t o determine the position of the carboxyl group which iseliminated in the formation of the ketonic acid by the oxidationof the acid C18Ha05.It was natural to suppose hhat it was a t th2802 LE SUEUR AND WITHERS: THE MECHANISMend of the chain, and that the constitution of the acid CI8H3,O5was CO2H*~CH2],-CH(OR)*[CH2],-CO,H. The loss of water whensuch an acid is heated and of carbon dioxide when i t is oxidisedare difficult to explain, and this difficulty led us t o investigate itfurther.It has been stated already that when the acid C,,H,O, is heatedabove its melting point i t readily loses water t o give the unsaturatedacids C,,H3,04, and on submitting these to oxidation so as t o breakthe chain the following acids were obtained : n-octoic, n-nonoic,suberic, and azelaic.The occurrence of azelaic acid in the oxida-tion products shows that the group -C-[CH,],*C* is present in theparent acid, and the formation of 92-nonoic acid under like condi-tions is evidence that the group CH,*[CH,],*C* is also present.These results point to the conclusion that one of the carboxylgroups is near the middle of the chain, and this is supported by thefact that when the unsaturated acids formed by the loss of waterfrom the acid C,8H3405 are reduced with hydrogen in the presenceof platinum black a saturated dicarboxylic acid, C,,H,O,, meltinga t 71*5-'72*5O, is obtained, whereas the normal dicarboxylic acid,CO,H*[CH,],,-CO,H, melts a t 1 1 8 O .This and other considerationsled us t o the conclusion that one of the carboxyl groups and thehydroxyl group were attached to one and the same carbon atom,and that the acid C18H3405 was a-hydroxy-a-octylsebacic acid,CH3*[CH,]7*C(OH)(C0,H)*[CH,],*C0,H. This was proved to becorrect beyond all doubt by the actual synthesis of the acid bymeans of the hydrogen cyanide method thus :CH,I CH,I CH3 I[ICH,'I7 y 3 2 1 7 LC;'H,I,LOCf.'B& IHCN C(OH}.CN -3 HzO C(OH)*COzH----f I I I[ I1 21 7I["'2I7C0,H GO,H C02HThe syiitlietic acid had all t'he properties of the original acid ;thus when oxidised with potassium permanganate i t gave theketonic acid C17H3203, melting a t 78*5O, and when heated it readilylost water t o give a producb which readily reduced potassiumpermanganate, etc.The identification of n-octoic acid, n-nonoic acid, suberic acid,aiid azelaic acid as tlie products of the oxidation by potassiumpermanganate of the unsaturated acids produced by the removal oftlie elements of water by lieat from a-hydroxy-a-octylsebacic acidsliows that this loss takes place in two directions thusOF THE ACTION OF FUSED ALKALIS.PART I. 2803--+The product of the fusion of dihydroxystearic acid with potass-ium hydroxide also contains small quantities of a second acid,which we find to be identical in all respects with the ketonic acid,C17H3203, melting a t 78*5O, and produced by the oxidation of theacid C18H3,05 with potassium permanganate.That the occurrenceof this ketonic acid amongst the products of the fusion is due t othe further oxidation of the acid C18H,05 is proved by the factthat it is formed when the acid C18H305 itself is fused withpotassium hydroxide.I n the previous communication (T., 1901, 79, 1313) dealing withthe fusion of dihydroxystearic acid with potassium hydroxide theformula C18H3,03 was erroneously assigned to this second acidfound amongst the products of the fusion. This error is easilyaccounted for when one remembers that the amount of this acidobtained a t that time was small, and that there is only a smalldifference between the percentages of carbon and hydrogen requiredby the formulae Cl,H%O, and C17H3203.I n order to see how far this remarkable reaction was general wecarried out a similar investigation of the action of fused potassiumhydroxide on dihydroxybehenic acid, and find that the reaction inthis case is exactly analogous to that with dihydroxystearic acid.Dihydroxybehenic acid, obt'ained by the oxidation of erucic acidwith potassium permanganate in alkaline solution, when fusedwith potassium hydroxide is converted into the monohydroxy-dicarboxylic acid, C22H4205, which on oxidation with potassiumpermanganate loses carbon dioxide, and yields the ketonic acid,C,,H,,,03. The constitution of this ketonic acid has been deter2804 LE SUEUR AND WITHERS: THE MECHANISMmined, and the parent acid synthesised from it by the hydrogencyanide method.These various changes are tabulated below :CsH17*CH:CH[ CH2]11*C02H -4 RMnOFusion C*H,,*CH(OH)*CH(OH)*[CH,],,*CO,H zt=2KMnOA.HCN + hydrolysis,C (OH) (COZH) [ CH,], 1 CO2H t-- -+NH2OH.C,H,,*CO*[CH,I,,*CO,H ~ +C,H,;.C(NOH)*[CH2],*C02H 1 H280~The product of the fusion of dihydroxybehenic acid with potass-ium hydruxidel also contains small quantities of the, ketonic acidC,H4,0,, which is undoubtedly formed by the oxidation of theacid C22H42O5 by the fused potassium hydroxide.The above intramolecular changes which take place during thefusion with potassium hydroxide are of particular interest as theyinvolve the migration of the heavy group C8HI7* from one carbona.tom to another adjacent to it.It is probable that this changeprecedes oxidation, and that an acid containing a primary alcoholgroup is formed as an intermediate product. I f this assumption isgranted, then the formation of the dicarboxylic acid is easilyexplained, for it has beeln shown that. primary alcohols readilyvield acids when they are fused with potassium hydroxide (Guerbet,Compt. rend., 1911, 153, 1487; 1912, 154, 222, 713) OF THE ACTION OF FUSED ALKALIS. PART 1. 2805With the object of obtaining direct evidence as to the correctnessof the above supposition we fused dihydroxystearic acid withpotassium hydroxide in an atmosphere of nitrogen, and in thisway hoped to limit the reaction to the intramolecular change.I nthis, however, we were not successful, as the product under theseconditions is also the acid CI8HMO5, the yield of acid being thesame in both cases. This result is nevertheless interesting as itshows that the oxidation' occurring during the fusion is notdependent on tlhe presence of atmospheric oxygen.We have been unable to find any record of the move'ment of sucha heavy group within the molecule of an aliphatic compound. Asomewhat analogous case in the aromatic series is the formatioiiof benzilic acid by the action of fused potassium hydroxide onb e n d and by the action of aqueous potassium hydroxide and airon benzoin.EXPERIMENTAL.Prepration of a-Hpdroxy-a-octylse back A cia?,CO,H*C(OH) ( C8HI7) [CH,],* C0,H.This acid was obtained by the fusion of dihydroxystearic acidwith potassium hydroxide, and the method employed did not diffelrin any essential detail from that described in a previous communi-cation (Le Sueur, T., 1901, 79, 1317), except that the acidobtained by acidifying the aqueous solution of the fusion wasdissolved in ether, the ethereal solution dried and evaporated, andthe residue crystallised from a mixture of benzene and chloroform.The yield of pure acid melting a t 1 1 1 - 1 1 2 O was very constant,and amounted t o 65 per cent.of the theoretical.The b e w o y l derivative, CO,H~C(OBZ)(C,H,~)~[CH,]~*CO,H, wasprepared by the Schotten-Baumann method, in the hope of obtain-ing a solid derivative which would serve to characterise the parentacid.Unfortunately, i t proved t o be a glassy mass, which couldnot be purified:0.2309 gave 0.5776 CO, and 0-1838 H,O. C=68'22; H=8*91.0.3439 required 15.9 C.C. O*lN-NaOH. M.W. = 432.6.C25HB06 requires C = 69.08 ; H = 8-82 per cent.The m-nitrobenzoyl derivative,M.W. = 434.3.C02H*C(O*CO* C,H,-NO,) (C,H,,) [CH2I7* CO,H,was prepared from the acid and nz-nitrobenzoyl chloride by theSchotten-Baumann method. The product was a viscid mass, which,after treatnient with dilute hydrochloric acid, was freed fromnitrobenzoic acid by boiling with water. The insoluble residue wascrystallised, first from dilute formic acid and then from a mixtureof chloroform and light petroleum, from which i t separatee as 2806 T,E SUEUR AND WITHERS: THE MECHANISMwhit,e powder melting a t 80-81O.eum, and soluble in chloroform or glacial formic acid :It is insoluble in light petrol-0-2078 gave 5.4 C.C.N, a t 22O and 755 mm. N=2*91.C,,H,,O,N requires N = 2-92 per cent.Prepmation of 8-Xetomargaric Acid, C,H,,*CO*[CH,],*C02H.Ten grams of a-hydroxy-a-octylsebacic acid were mixed with160 C.C. of acetone (previously purified by treatment with potassiumperrnanganate) and 20 C.C. of water, and the mixture was gentlywarmed with powdered potassium permanganate, which was gradu-ally added. After t'he addition of 8 grams of permanganate, theoxidation proceeded very slowly. The acetone was then removed bydistillation, the residue treated with water, and the manganesedioxide dissolved by dilute sulphuric acid and sulphur dioxide.The precipitated acid was dissolved in ether, the ethereal solutionwashed, dried, evaporated, and the residue crystallised from amixture of chloroform and light petroleum.The yield was 79.5 percent. of the theoretical:0.1560 gave 0.4098 CO, and 0.1578 H20. C=71*64; H=11*32.*0.3766 required 13.6 C.C. 0-W-NaOH. M.W. =277.C,,H,,O, requires C = 71.77 ; H = 11.35 per cent. M.W. = 284.3.8-Xetomargaric acid, C,H,7*COo[CH,],*C0,H, is readily solublein cold chloroform and moderately so in ether, alcohol, ethylacetate, or benzene in the cold. It is insoluble in light petroleum,and crystallises from a mixture of this solvent and chloroform inglistening plates melting a t 78'5O.It is not oxidised by an alkalinesolution of potassium permanganate in the cold. The acid is alsoproduced by the fusion with potassium hydroxide of a-hydroxy-a-octylsebacic acid. The white silver salt was obtained on addinga warm alcoholic solution of the sodium salt of the acid t o a warmalcoholic solution of silver nitrate :0.2784 gave 0.0784 Ag. Ag=28*16.C,,H,,O,Ag requires Ag = 27.60 per cent.Hethy1 8-ke tomargara t e , C,H,7*CO*[CH,],*C02Me, was obtainedby heating for four and a-half hours on the water-bath a mixtureof 8.5 grams of the acid, 90 C.C. of methyl alcohol, and 45 C.C. ofconcentrat'ed sulphuric acid. The product was poured into water,the est'er dissolved in ether, the ethereal solution dried, evaporated,and the residue crystallised from methyl alcohol, from which i tseparates in large, pearly flakes, melting a t 45.5O:0.1518 gave 0.4030 CO, and 0.1585 H,O.C=72*40; H=lI:68.C,,H,O, requires C = 72.42 ; H = 11.49 per cent.* A second analysis gave C: = $1 $9 ; H = 11.40OF THE ACTION OF FUSED ALKALIS. PART I. 2807h't h yl 6-ke to margar ate, C8H,,*CO* [ CH2],*C0,Et, was preparedby the method used for the preparation of the methyl ester, andwas purified by crystallisation from dilute alcohol, from which itseparates in pearly flakes melting a t 38O. It is readily soluble inalcohol, benzene, chloroform, or acetone in the cold :0*1109 gave 0.2964 CO, and 0.1172 H20. C=72.89; H=11.83.Cl,H3,O3 requires C = 73-01 ; H= 11.62 per cent.8-Ketomargaramide, C,H,,.CO*[CH,],*CO*NH,, was preparedthus: 1.2 Grams of the ketonic acid and 0.8 gram of thionylchloride were heated together on the water-bath for half-an-hour,and the cold product was poured into excess of ammonia. Theprecipitated amide was dissolved in ether, the solution washed,dried, and evaporated, and the residue crystallised from a mixtureof chloroform and light petroleum, when i t was obtained in irides-cent needles melting a t 119O.It is insoluble in light petroleum,and dissolves in alcohol, chloroform, ether, benzene, or hot water :0*1100 gave 4.6 C.C. N, a t 21° and 762 mm. N=4.75.CI7H,,O,N requires N = 4.95 per cent.8-lietu?ncrr~~raniZ~de, C8H,,*CO*[CH,],*CO-NHPh, was obtaiiieclby the action of the acid chloride (prepared by the interaction ofthe acid and thionyl chloride) on aniline.It was purified bycrystallisation, first from acetic acid and then from formic acid,and was obtained as a white, crystalline powder, melting a t 96.5O.It dissolves freely in alcohol, acetone, chloroform, or benzene, issparingly soluble in ether, acetic acid, or formic acid in the cold,and insoluble in light petroleum :0-2925 gave 9.9 C.C. N, a t 19O and 770 mm. N=3*95.C,,H,,O,N requires N = 3.90 per cent.The semicarbasone of 6-ketomargaric acid was obtained by heat-ing for five hours on the water-bath an aqueous-alcoholic solutionof 0.6 gram of the acid, 0.5 gram of semicarbazide hydrochloride,and 0.5 grain of dry potassium acetate, and was purified by crystal-lisation from acetone, from which i t separates in stellar aggregatesof prisms melting a t 1 1 1 O .It is sparingly soluble in alcohol,acetone, or ethyl acetate in the cold, and is insoluble in ether orlight petroleum :0.1206 gave 13.1 C.C. N2 a t 19O and 774 mm. N=12.55.The oxime of 0-ketomargaric acid was prepared by boiling on thewater-bath for three hours a solution of 5 grams of the acid,2 grams of liydroxylainine hydrochloride, and 3 grams of sodiumhydroxide in a mixture of 40 C.C. of alcohol and 20 C.C. of water.The alcohol was then evaporated, and the product poured intoC,,H,,O,N, requires N = 12.31 per cent2808 LE SUEUR AND WlTHERS: THE MECHANISMdilute sulphuric acid, when the precipitated oil was extracted withether. The ethereal solution was washed with water, dried, andevaporated.The residue was a colourless oil, which did not solidifyeven on leaving in a vacuum desiccator for three months:0.2618 gave 9.8 C.C. N, a t 15O and 750 rnm. N=4*30.C,,H,,O,N requires N =4.68 per centt.Conversion of the Oxime of 0-Ketomaryaric Acid into the AmidesCsH,7*CO*NH*[CH2]7*C0,H am2 CsH,,*NH*CO*[CH,]7-C0,H.The oxime (8 grams) was slowly poured into concentrated sul-phuric acid (40 c.c.) and the solution heated for one and a-halfhours on the water-bath. The resulting dark brown liquid wascooled and slowly poured on crushed ice, and the precipitatedamides were collected and crystallised from dilube acetic acid, fromwhich they separated in aggregates of thin plates melting a t68.5-69.5O. Although the substance thus isolated is undoubtedlya mixture of two amides, its melting pointl is nevertheless quitesharp :0.2450 gave 10.4 C.C.N, a t 17.5O an1 758 mm. N=4*90.C,,H,O,N requires N = 4-68 per cent.Hydrolysis of the above Amides.-2-5 Grams of the amides wereheated with 12 C.C. of concentrated hydrochloric acid in a sealedtube a t 180° for four hours. The dark brown products resultingfrom several such quantities were mixed with water and distilledin a current of steam (distillate=A). The residue was then madealkaline1 and again distilled in steam (distillate = B). The residuewas now concentrated t'o a sman bulk, and poured into excess ofconcentrated hydrochloric acid, and the mixture left for some time,when a crystalline solid separated (solid = C), which was relmovedby filtration (filtrate = D)IDistillate B was extracted several times with much ether, theethereal solution washed tlwice with small quantities of water,dried, and evaporated.The residue obtained from 21 grams oftho mixed amides weighed 4.8 grams, and all distilled bebween248" and 255O; it was purified by fractional distillation, and theportion boiling at 251-253O collected for analysis. This fractionreadily solidified on cooling in ice, and the resulting solid meltedat 1 2 - 1 3 O :0.3420 required 21.6 C.C. 0-1N-NaOH. M.W. = 158.4.n-Nonoic acid, C9H180.,, lias M.W. = 158, b. p. 253.4O (corr.), andni. p. 12*Go. The identity of the acid with n-nonoic acid wasfurther proved by converting a small quantity of it inta its zinOF T E ~ E ACTION OF FUSED ALKALTS.PART T. 2809salt, wliicli, after crystallisation froni absolute alcohol, melted a t1 3 4 O , which is the melting point of zinc 1,-nonoate.Distillate B.-This alkaline distillate was acidified with hydro-chloric acid, the resulting solution mixed with much ether in aseparating funnel, and then made strongly alkaline with potassiumhydroxide, and the whole quickly shaken. This procedure wasadopted in order t o enable tlie ether t o dissolve the base before thelatter had had time t o combine with the water t o form a hydrate.The dried ethereal solution on evaporation gave a residue whichdistilled between 173O and 177O. A small quantity of this wasconverted into the platinichloride, which wits analysed (Found,Pt= 29.15.[C8Hl7*NH2,HCI],Ptc1, requires Pt = 29.31 percent.).The free base gave tlie carbylamine reaction, and readilyabsorbed carbon dioxide from the air. It is therefore identicalwith octylamine, the boiling point of which is 175-177O.s-PI~enyZoct?lZcarbamide, C,H,,*NH*CO*NHPh, was prepared byadding 0.4 gram (1 mol.) of phenylcarbimide to 0.4 gram (1 mol.)of the above base, dissolved in 10 C.C. of dry benzene, and allowingthe mixture to remain overnight. The solution was then concen-trated t o half its bulk, and mixed with light, petroleum. Thecarbamide which separated was dissolved in a mixture of benzeneand light petroleum, from which it crystallises in flakes, meltingat 80°. It is readily soluble in alcohol, chloroform, or benzene,and is insoluble in light petroleum :0.1808 gave 17.4 C.C.N, a t 19O and 776 mm. N=11.32.C,,H,,QN2 requires N = 11.29 per cent.Solid C.-This consisted of long, flat needles, melting at105*5-106*5°, which, after crystallisation from water, were quitecolourless and melted a t 106.5O. The acid had the characteristicappearance of azelaic acid (m. p. 106O), with which it is un-doubtedly identical (Found, C =57'40; H =8*75. M.W. = 189.6.CnHIGOi requires C = 57.40 ; H = 8-58 per cent. M.W. = 188).Filtrate D.-This, which contained much free hydrochloric acid,was e'vaporated t o dryness, the residue partly dried, and thenrepeatedly extracted with acetone. This acetone solution on eva-poration left a residue which was dissolved in hot alcohol, and theresulting solution was mixed with an equal volume of ether.Fromthe mixture the1 hydrochloride of the amino-acid was depositedgradually in glistening needles melting a t 145-146O. (The hydro-chloride of q-amino-n-octoic acid melts a t 147O). (Found, N = 7.10.C8HI70,N,HC1 requires N=7*16 per cent.) A small quantity ofthe) hydrochloride was converted into the platinichloride (Found,Pt = 26-60. Calc., Pt = 26-80 per cent.)2810 I,E SUEUR AND WITHERS : THE MECEIANISMThe amino-acid resulting from the hyclrolysis of t h e mixed miidesis, therefore, q-amino-n-octoic acid.The yields of the' four substances obtained by the hydrolysis ofthe mixed amides were as follows:12-Nonoic acid .................................... S5.4 per cent.of the theoreticalOctylamine ....................................... 86.5 ,, 5 9 9 ,Azelaic acid ....................................... 83.0 ,, 9 9 3 7Hydrochloride of 7-amino-x-octoic acid ... 67.0 ,, 9 9 9 9Formation of the Ui~saturated Acids, C,,H,,O,, from a-llydroxg-a-octglsebacic Acid.Eight grams of a-hydroxy-a-octylsebacic acid were graduallyheated under a pressure of 25 mm. in a distilling flask immersedin a metal-bath. Effervescence, due to the evolution of water-vapour, began a t about 220°, and rapidly increased as the tempera-ture rose. The vapour evolved contained only a trace of carbondioxide. The temperature was maintained a t about 280° until alleffervescence had ceased, and then raised gradually, when theliquid distilled between 290° and 310° and gave 6.8 grams ofdistillate.This was redistilled under 30 mm. pressure, and aportion boiling a t 294-295O collected for analysis :0.1624 gave 0.4152 CO, and 0.1520 Q O . C= 69.72 ; H = 10.34.A specinien from another preparation was converted into the0.2168 gave 0.0870 Ag. Ag=40*13.C,8H300,Ag, requires Ag = 41.02 per cent.The above figures agree sufficiently with the calculated values toshow that the substance is produced from a-hydroxy-a-octylsebacicacid by the loss of water only. It was kept for several months,but could not be obtained as a definite solid; also, redistillationunder diminished pressure produced slight decomposition. Theelimination of w a t a from a-hydroxy-a-octylsebacic acid by Iieat-big it with sulphuric acid was also tried.F o r this purpose, theacid was heated a t 200-220° for four and a-half hours with 60 percent. sulpliuric acid in a sealed tube. Much sulphur dioxide wasevolved when the tube was opened, and the contents were foundto consist for the most part of 8-ketomargaric acid, melting at78.5O. I n other words, the sulphuric acid had acted as an ordinaryoxidising agent. I n another experiment, the acid was boiled withdilute sulphuric acid for two hours, but$ in this case, the acidwas recovered unchanged.That the substance produced by heating a-hydroxy-a-octyl-C,,H,O, -requires @= 69.23 ; H = 10.258 per cent.silver sal* and analysedOh' THE ACTION OF FUSED ALT<AT,IS. PART I. 2811sebacic acid is unsaturated is confirmed by the facts that i t readilyreduces potassium permanganate in alkaline solution in the cold,and, under the influence of platinum black, absorbs hydrogen togive the saturated dicarboxylic acid, C18H304, described onpage 2812.Its constitution was determined by investigating theproducts obtained when it is oxidised by potassium permanganate,as described below. Two quantities of 11 grams each of the acidwere dissolved in a mixture of 170 C.C. of acetone and 20 C.C. ofwater, and to this, finely powdered potassium permanganate wasadded. Reduction took place immediately in the cold, and becamevery energetic on warming. Small quantities of permanganatewere added from time to time to the boiling solution until themwas no further reduction.The acetone was then distilled off, themanganese dioxide dissolved by means of dilute sulphuric acidand sulphur dioxide, and the whole distilled in a current of steam(distillate=X). The mother liquor in the flask deposited, on cool-ing, a solid (9 grams), which was collected and dissolved in 250 C.C.of boiling wat'er. This solution deposited, on cooling, a solid(2 grains), which was removed by filtration (filtrate= P), and thenredissolved in hot water. The solution was neutralised with excessof magnesium carbonate and filtered while hot. The filtrate wasconcentrated to a syrup, allowed to cool, and the magnesium salt,which had separated was collected. It was again crystallised fromwater, and finally decomposed by heating with dilute hydrochloricacid.The acid thus obtairied was crystallised from water, fromwhich i t separated in large, thin, glistening plates, melting a t105-106O. A mixture of equal parts of this acid and azelaicacid (m. p. 106O) melted a t 106O. (Found: M,W.=187*8.CgHl,O4 requires M.W. 188.)Filtrate Y was evaporated to dryness, and the dried residue,melting a t 90--126O, was exhracted four times with ether, using12 C.C. for each extraction. The undissolved residue melted a t137-140°, and, after crystallisation from water, melted a t 138.5O.(Found : C = 55.02 ; H = 8.20. M.W. = 174.6. C8H1,0, requiresC=55*14; H=8*10 per cent. M.W.=174.1.)The substance was thus proved t o be suberic acid.Distillate X was extracted with much ether, the ethereal solu-tion washed once with water, dried, and evaporated, and the resi-due was fractionally distilled.The fraction boiling at 234-242Owas identified by analysis as woctoic acid (b. p. 236-237O).(Found : M.W. = 142 ; Ag =43*35. C,H,,02 requires M.W. = 144,and f o r silver salt Ag=42.98 per cent.)The fraction boiling at above 245O (Found: M.W.=158.C9HI80, requires M.W. = 158) was converted, through th281 2 LE STTEUR A S T ) WITHERS : THE MECHAKISMsodium salt,, into the zinc salt, wliicli was shown to be zincu-nonoate. The amount of substance dealt with did not admit ofa complete separation and isolation in a pure state of the u-octoicand n-nonoic acids, but the analysis showed that these two acidswere undoubtedly present in the mixture.Prepamtion of a-Octylsehncic Acid, C0,H*CH(C,H,7)fCH2]7~C0,H,by Reduction of the Unsaturated Acids, C18H3204.This reduction was effected by the direct addition of hydrogenin the presence of spongy platinum.The catalyst was prepared bythe reduction of platinic chloride by means of formaldehyde inthe presence of sodium hydroxide (Loew, Ber., 1890, 23, 289;Willstiitter, Ber., 1912, 45, 1472; IIess, Ber., 1913, 46, 3120),and was washed first with water and then with glacial acetic acid.Five grams of the unsaturated acids CI8H3,O4, dissolved in 10 C.C.of glacial acetic acid, were reduced. The absorption of thehydrogen was never very rapid, the quickest rate being about50 C.C. in twenty minutes. Sometimes the hydrogen was under apressure of about 20 cm. of :mercury above the atmosphericpressure, but this did not appear to increase the rate of absorption.When no further reduction took place, the acetic acid solutionwas warmed, filtered, and the filtrate poured into water.The pre-cipitated acid was collected and crystallised from a mixture ofchloroform and light petroleum until its melting point was con-stant. The yield was 70 per cent. of the theoretical:0.1350 gave 0.3405 CO, and 0.1350 H,O. C = 68.79 ; H = 11-19.0.2054 required 13'2 C.C. 0-1N-NaOH.a-Octylsebacic acid, C02H*CH(C8H,7)*[CH2]7*C0,H, cryshllisesfrom a mixture of chloroform and light petroleum in aggregatesof hard prisms, melting a t 71'5-72'5°. It is readily soluble inalcohol, chloroform, o r ether in the cold, sparingly so in coldbenzene, and insoluble in light petroleum.It does not reduce analkaline solution of potassium permanganate even when heated,nor does it decolorise a solutJon of bromine in chloroform. Thewhite silver salt was obtained on adding a solution of the sodiumsalt of the acid t o a solution of silver nitrate:M.W. = 311.2.C,,H,O, requires C = 68.75 ; H= 10.90 per cent. M.W. = 314.25.0.1540 gave 0.0622 Ag. Ag=40*39.C18H3,0,Ag2 requires Ag = 40.86 per cent.Uet 12. yl a-oct ylsebacat e, CO,Me*CH ( C8H17) *[CH,],*CO,Me, wasprepared by the interaction of the dry silver salt of the acid andmethyl iodide in solution in dry benzene. It is an oily liquidOF THE ACTION OF FUSED ALKALIS. PART I. 2813boiling a t 230-233O/25 mm., and dissolves readily in the commonorganic solvents :0.1580 gave 0.4054 CO, and 0.1636 H,O.0.3056 in 15.7840 benzene gave At - 0.307O.CmH,04 requires C=70.11; H=11*19 per cent.The diamide , NH,* CO*CH( C8EI7) *[ CH,], CO*NH,.-One gramof the acid and 0.5 C.C.of thionyl chloride were warmed togetheron the water-bath for fifteen minutes, the product was dissolvedin ether, and the ethereal solution poured into a concentrated,aqueous solution of ammonia. The precipitated amide was col-lected, washed with water, and crystallised from dilute alcohol,from which it separates in slender needle,s, melting a t 167-168O.It is insoluble in water, ether, benzene, or chloroform in the cold,and dissolves re'adily in boiling alcohol :C=69.98; H=11*59.M.W. = 322.M.W.=342.0.1990 gave 15.6 C.C.N, a t 20° and 754 mm. N=8.92.C,,H,O,N, requires N = 8.97 per cent.Synthesis of a-Hydroxy-a-octylsebacic Acid from 8-HetomargaricA cid.Ten grams of pure, dry hydrogen cyanide and six drops oftriethylamine were added to a solution of 10 grams of 6-keto-margaric acid in 50 C.C. of dry chloroform. After remaining fora fortnight in a well-stoppered bottle, kept in a cool place, thechloroform and excess of hydrogen cyanide were evaporated bymeans of a current of air, and the solid residue was dissolved inether, the ethereal solutlion washed with dilute sulphuric acid,then with water, dried, and evaporated. The residue was hydro-lysed by dissolving i t in SO C.C. of alcohol, previously saturatedwith dry hydrogen chloride.The mixture was left a t the ordinarytemperature f o r three and a-half days, and then boiled on thewater-bath for eight hours. The resulting solution, on cooling,deposited 24 grams of crystalline solid, melting a t 38-39O, whichwas found to be the ethyl ester of the unchanged ketomargaricacid (see page 2807). This was removed by filtration, the filtrateevaporated to dryness, and the residue boiled for forty hourswith 24 grams of potassium hydroxide dissolved in a mixture of50 C.C. of alcohol and 10 C.C. of water. The complete hydrolysisof the hydroxy-cyanidel was difficult to effect, and, even after boil-ing for the period mentioned, traces of ammonia were still beingevolved. The alcohol was evaporated, the, residue diluted withwater, and extracted with ether.The aqueous mother liquor wasthen poured into dilute hydrochloric acid, the precipitated aciddissolved in ether, the ethereal solution washed, dried, andVOL, cv. 8 2814 LE SUEUR AND WITHERS: THE MECHANISMevaporated. The residue weighed 7.5 grams, and, after three re-crystallisations from chloroform, melted a t 11 1-1 12O. A mixtureof equal parts of this acid and a-hydroxy-a-octylsebacic acid melteda t exactly the same, temperature:0.1371 gave 0.3270 CO, and 0.1286 H,O.0.2254 required 13.5 C.C. O'lN-NaOH.The synthetic acid, like the original a-hydroxy-a-octylsebacicacid, readily loses water when heated, to give a liquid which veryreadily reduces potassium permanganate. When oxidised withpotassium permanganate in acetone solution, the synthetic acidgives a monobasic acid melting a t 78*79O, identical with the8-ketomargaric acid obtained by the oxidation of a-hydroxy-a-octyl-sebacic acid.This identity was fully confirmed by the fact thata mixture of equal parts of the two acids melted a t 78-79O, andfurther, the methyl ester of the ketonic acid obtained from thesynthetic acid had all the properties of methyl 6-ketomargarate.C = 65.05 ; H= 10.50.M.W. = 330.3.M.W. = 333.9.C,,H,,O, requires C= 65.40; H,=10*38 pelr cent.Preparation of Dihydroxybehenic Acid.The method employed for the oxidation of erucic acid differedsomewhat from that used by Hazura and Grussner (Monatsh.,1888, 9, 948), and since t'hew observers obtained a yield of only25 per cent.of the theoretical, whereas ours amounted to 75 percent., we give the essential details of the method we used. Thirty-five grams of crucic acid (m. p. 33O) and 14 grams of potassiumhydroxide were dissolved i n 2 litres of water, the solution cooledto Oo, and maintained a t this temperature during the whole ofthe oxidation. A solution of 32 grams of potassium permanganatein 1500 C.C. of water was then added drop by drop to the solu-tion, which was kept well stirred by means of a turbine. Themixture 'was allowed to remain over night, the manganese dioxidedissolved by means of dilute sulphuric acid and sulphur dioxide,and the precipitated acid filtered and crystallised from alcohol.Preparation of a-Hydroxy-a-oc tyldodecanedicar boxylic A cid,C02H*C(OH)(C,H,7)*[CH2]11*COeH.Twenty grains of dihydroxybehenic acid (m.p. 127-128O), 100grams of potassium hydroxide, and 40 C.C. of water were mixedtogether in p nickel crucibles, and the whole' gradually heated ina metal-bath to 240-245O, and maintained a t this temperature forone hour, the fused mass being stirred continuously. The coldmass was dissolved in water, acidified with dilute sulphuric acid,and the precipitated acid dissolved in ether. The ethereal 601~OF THE ACTION OF FUSED ALKALIS. PART I. 2815tion was washed, dried, evaporated, and the residue crystallisedonce from chloroform and then from acetone until its meltingpoint, was constant. The yield was 64 per cent. of the theoretical:0.1233 gave 0.3083 CO, and 0.1213 H,O.0.2178 required 11.3 C.C.O'lN-NaOH.a-Hydroxy-a-octyldodecamn&carb oxylic acid is sparingly solublein alcohol, acetone, or benzene in the cold, but dissolves readilyon heating; it is sparingly soluble in hot ether, and is insolublein light petroleum. It crystaIlises from acetone in needles, melt-ing a t 115-116O. The white silver salt was obtained on addinga warm solution of the sodium salt of the acid to a warm solutionof silver nitrate :C=68*19; H=11*01.M.W. = 386.34.M.W. = 385.6.CZ2H4,O5 requires C = 68.33 ; H = 10.96 per cent.0.2206 gave 0.0784 Ag. Ag=35-54.C,2H4,05Ag2 requires Ag = 35.96 per cent.When the acid is heated, it behaves exactly like a-hydroxy-a-octylsebacic acid, in that it readily loses the elements of water.The product is a viscid oil, which shows no signs of solidifyingwhen kept; for several weeks.It readily reduces an alkaline solu-tion of potassium permanganate, but does not decolorise a solutionof bromine in chloroform.Prepration of p-Xetohelzeicosoic A c i d ,CH,*[CH,],*CO.[CH,],,*CO,H.Eight grams of a-hydroxy-a-octyldodecanedicarboxylic acid weredissolved in a mixture of 120 C.C. of acetone and 20 C.C. of water,and finely powdered po€assium permanganate was added to thesolution. There was no action in the cold, but, on warming,oxidation immediately took place, and became very energetic a tthe boiling point of the acetone. The permanganate was addedin small quantities a t a time until no further oxidation ttook place.The acetone was evaporated, the manganese dioxide dissolved bymeans of dilute sulphuric acid and sulphur dioxide, and the pre-cipitatod acid collected and crystallised from acetone :M.W.= 339.4.0.1194 gave 0.3240 CO, and 0.1273 H,O.0.5260 required 15.5 C.C. 0.1iV-NaOH.p-KetolwTzeicosoic acid is readily soluble in cold chloroform,sparingly so in alcohol, benzene, or ether, and insoluble in lightpetroleum. It crystallises from acetone in glistening, micaceousplates, melting a t 89-90°. The white silver salt was obtainedby adding a warm solution of the sodium salt t o a warm solutionof silver nitrate:C=7.4.01; H=11*93.M.W. = 340.32. C21H4003 requires C= 74.05 ; H = 11.85 per cent.8 x 2516 LE SUEUR AND WITEERS: THE MECHANISM0,2224 gave 0.0538 Ag. Ag=24*19.C2,H3,O3Ag requires Ag = 24.16 per cent.Methyl p-ke toheneicosoa te, CH,* [ CH,],*CO*[ CH2]ll*C02Me, wasprepared by the method used for the preparation of methyl 8-keto-margarate (p.2806), and was purified by crystallisation frommethyl alcohol, from which it separates in glistening, thin plates,melting a t 59-60°. It is readily soluble in ether, chloroform, orbenzene in the cold, and dissolves sparingly in cold acetone oralcohol :0.1296 gave 0.3535 CO, and 0.1402 H,O.C,,H,,O, requires C = 74-50 ; H = 11.95 per cent.Et h y l p-keto heneicosoa t e, CH,=[CH,],* CO*LC?H,],,*C02Et~ wasprepared by the method used for the preparation of the methylester, and was crystallised from dilute alcohol, from which itseparates in glistening leaflets, melting a t 56O. It is sparinglysoluble' in cold alcohol, and dissolves readily in the other commonorganic solvents :C=74-39; H=12*10.0.1463 gave 0*4010 CO, and 0.1593 H,O.C~H4403 requires C = 74.90 ; H = 12.04 per cent.The' semicarbaame was prepared by the method employed forthe preparation of t'he semicarbazone of 8-ketomargaric acid, andwas purified by crystallisation from acetone, from which itaeparates in aggregates of slender needles, melting a t 104-105°.It is sparingly soluble in boiling acetone or ethyl acet'ate, and in-soluble in light petroleum or ether:C=74*75; H=12.18.0.1978 gave 18.6 C.C.N, a t 22O and 763 mm. N=10*74.C2,H,,0,N, requires N = 10.58 per cent.Syntkesis .of a-Hydroxy-a-octyldodecanedicarh oxylic A cid fromp-Xetoheneicosoic Acid.The' method used to effect this synthesis was practically identicalwith that employed for the preparation of a-hydroxy-a-octylsebacicacid from 8-ketomargaric acid.The addition of the hydrogencyanide in this case also was very slow, and the complete liydro-lysis of the hydroxy-cyanide difficult to effect. The crude acidwas crystallised from chloroform until its melting point was con-stant, at 115-116O. A mixture of equal parts of this acid andthe original a-hydroxy-a-octyldodecanedicarboxylic acid melted a texactly the same' temperature :0.1115 gave 0.2788 CO, and 011094 H,O.0.2590 required 13.0 C.C. 0-1N-NaOH.C,zH4z0, requirw C= 68.33; H= 10.96 per cent.I n order t o confirm further the identity of the synthetic acidC=68*20; H=10.98.M.W. = 398.M.W.= 386OF THE ACTION OF FUSED ALKALIS, PART I. 2817with the a-hydroxy-a-octyldodecanedicarboxylic acid obtained bythe fusion of dihydroxybehenic acid with potassium hydroxide, asmall quantity of it was oxidised with potassium permanganate inacetone solution, when an acid was obtained which melted a t89-90°, and had all the properties of the p-ketoheneicosoic acidproduced by the oxidation, under similar conditions, of a-hydroxy-a-octyldodecnnedicarboxylic acid.Constitution of the Ketorzic Acid, C21H4003.The oxime, of the acid, C,,H,,O,, was prepared by the methoddescribed on page 2807, and was converted into1 the correspondingamides by heating with five times its volume of concentratedsulphuric acid for two and a-half hours on the water-bath. Theproduct was poured on crushed ice, and the solid collected andcrystallised irom glacial acetic acid containing a little water, fromwhich it separates in aggregates of stout prisms, melting a t85*5-86*5° 10.2667 gave 9.3 C.C.N, a t ZOO and 764 mm.The above isomeric amides were hydrolysed by heating withhydrochloric acid in a sealed tube, as described on page 2808. Theproduct was diluted largely with water, and distilled in a currentof steam (didillate 1). The residue was then made strongly alka-line, and distilled in a current of steam (distillate 2).Distillate 1 was extracted with etlier, the ethereal solutionwashed, dried, evaporated, and the residue distilled, when it allpassed over a t 150--155O/31 mm.It was fractionally distilledunder the ordinary pressure, and the fraction boiling a t 248-249Owas collected for analysis. It readily solidified when cooled inice. (Found, M.W. = 157.9. C9HI8O2 requires M.W. = 158.1.)The zinc salt was prepared in the usual way, and, after crystallisa-tion from alcohol, melted a t 136O, proving the acid to be n-nonoicN=4*02.C,,H,,O,N requires N = 3-94 per cent.. - acid.Distillate 2.-The amine was isolated from this distillate asdescribed on page 2809, and was obtained as a clear, colourlessliquid, boiling a t 174--176O, which was identified as octylamineby analysis of the platinichloride (Found, Pt.= 29.07. Calc.,Pt=29*21 per cent.), and by the preparatdon of phenyloctyl-carbamide which melted at 79-80°.The mother liquor left after obtaining distillate 2 was con-centrated to about 350 c.c., and then poured into a slight excessof concentrated hydrochloric acid.The solid which separated oncooling was collected, dried, and boiled with 50 C.C. of alcohol.The solutioa was filtered, and the filtrate concentrated t o abou2818 THE MECHANISM OF THE ACTION OF FUSED ALKALIS.20 c.c., and then mixed with 50 C.C. of ether, when a voluminous,crystalline mass separated. This was collected (filtrate 3) andcrystallised from hot water, from which it separates in thinlamellae, melting a t 163O:0.1782 gave 8.8 C.C. N, a t 20° and 778 mm.A portion was converted into the pZati?aichloride, which melts0.2394 gave 0.0550 Pt. Pt=22*97.The' substance is theref ore the hydrochloride of A-aminohuricacid, CO2H-[CH2],,*NH2,HC1.It is insoluble in hot or cold ether,benzene, chloroform, or acetone, freely soluble in alcohol, andmoderately so in water.The free amino-acid was isolated by adding the calculatedvolume of a dilute sodium hydroxide solution to a warm, aqueoussolution of the above pure hydrochloride'. The white, crystallinemass which separated was collected, washed with cold water, anddissolved in boiling water, from which it crystallised on cooling inlong, thin plates with almost square ends:0.0942 gave 0.2306 CO, and 0.0982 H20.0.1574 ,, 9.0 C.C. N2 a6 18O and 770 mm. N=6.70.N=5.75.Cl2H2,O2N,HC1 requires N =5*57 per cent.and decomposes a t 209O:(C,2H2,0,N)2H2PtC1, requires Pt = 23.21 per cent.C=66*76; H=11.67.C,,H,O,N requires C = 66.91 ; H = 11.71 ; N = 6-51 per cent.A-Aminohuric acid, NH2*[CHJ,,*CO2H, melts a t 184O, and ispractically insoluble in the common organic solvents.I n order more completely to chazacterise the acid, its P-naph-thalenesulphonyl derivative was prepared.F o r this purpose, amixtare of the amino-acid (1 mol.), N-sodium hydroxide (1 mol.),and an ethereal solution of P-napht8halenesulphonyl chloride(2 mols.) was shaken in a stoppered bottle. At intervals of onehour, three more molecular proportions of sodium hydroxide wereadded, and the shaking was prolonged altogether f o r five hours.The mass was diluted with water and allowed t o flow into cold,dilute sulphuric acid, the ethereal layer being retained in aseparating funnel. The white solid which separated was collected,washed free from sulphuric acid, and dissolved in hot 25 per cent.alcohol, from which it crystallised in small flakes, melting at 115O:C22H3104NS requires N = 3-46 per cent.C10H7*S02-NH[CH2]ll*C02H,is readily soluble in hot alcohol, sparingly so in cold alcohol, hotether, acetone, or water, and insoluble in light petroleum.0.1238 gave 3.8 C.C. N2 a t 18O and 750 mm.A-P- Na ph t ha I en esulp h o n yla min olauric acid,N=3.50THE POLYSULPHIDES OF THE ALKALI METALS. PART 11. 2819The alcohol-ether filtrate 3 (p. 2818) was evaporated on thewater-bath, and ths residue diluted with water. On leaving thissolution, crystals soon separated ; these were collected, and crystal-lised first from water and then from benzene, from which the sub-stance was deposited as a crystalline powder, melting a t 113-5O.(Found, C = 63.89 ; H = 9.92. M.W. = 247.2. C,,H,,O, requiresC=63*88; H=9*91 per cent. M.W.=244*19.) A portion wasconverted into the silver salt (Found : Ag= 46.89. CI3H,0,Ag,requires Ag=47*11 per cent.) and another portion into the amide,which melted a t 174-175O. (Found : N= 11.66. C,,H,60,N,requires N = 11.57 per cent,.)The dicarboxylic acid which resulted from the hydrolysis of theabove mixture of isomeric amides was thus fully identified withbrassylic acid, CO2H[CH2],,*CO,€I. Fileti and Ponzio (J. pr.Chem., 1893, [ii], 48, 323) found as the, melting point of brassylicacid 114O, and of the amide 177O.The yields of the four substances obtained by the hydrolysis ofthe above mixed amides were as follows:n-Nonoic acid ................................. 90 per cent. of the theoretical.Octylamine .................................... 89 ,, 99 99Brassylic acid 57 ,, 27 ,Y Hydrochloride of A-aminolauric acid.. .... 55 ,, Y, I >.................................CHEMICAL LABORATORY,ST. TfIOMAS’S HOSPITAL,LONDON, S.E