380 NEWBERY : UNSTARLE COMPOUNDS OFXLI.- Unstable Compounds of Cholesterol withBarium Methoxide.By EDGAR NEWBERY.A METHOD for extracting cholesterol and cerebrone from brain byboiling with a methyl-alcoholic solution of barium hydroxide,evaporaking ta ,dryness, and extracting the dry material with anorganic solvent, has long been in use in the Chemical and Patho-logical Departments of the Manchester Un.iversity (compare Smithand Mair, J . Path. Bacf., 1910, 15, 122, and Lapworth, ibid.,1911, 15, 254).Professor Lapworth, using this method, found that with excessof baryta the cholesterol was extracted by ether in a Soxhletapparatue only with great difficulty; but if the excess of baryta *had been destroyed by a stream of carbon dioxide, or by additionof acetic acid before removal of methyl alcohol in the first instance,then the cholesterol was extracted from the dry residue with greatfacility.The work described in the present paper was undertaken witha view to ascertain, if possible, the cause of this retention of thecholest er 01.Cholesterol was carefully purified by repeated crystallisationfrom methyl alcohol, acetone, and ether, dried thoroughly, and asolution in ether made up containing 0.01 gram of dry cholesterolin 1 C.C.Definite volumes of this solution were measured out foreach experibent, and the ether distilled off, thus ensuring thepresence of a, pure substance free from the hydrated compound.Experiments were first performed t o determine the conditionsunder which cholesterol is taken up by baryta.Series I.-All the evidence collected pointed to the conclusionthat solid barium oxide, hydroxide, or methoxide added t o asolution of cholesterol in ether does not absorb any appreciablequantity of cholesterol, and consequently that the solid baryta andcholesterol must previously have been in solution together beforethe influence of the former on the solubility of the latter becomesappreciable.Subsequent investigations were therefore confined tocases where this condition was fulfilled. Since methyl alcohol isthe only available liquid capable of dissolving both solids, it wasused in all the operations described hereafter.Series IZ.-In these experiments 1-25 grams of barium hydr-* Throughout this paper, the word ‘ I bnryta ” is used where i t is doubtful if theoxide, hydroxide, methoxide, or cholesterol compound is the active agentCHOLESTEROL WITH BARIUM METHOXIDE.381oxide, or its equivalent of barium oxide, dissolved in methyl alcoholwas added to 0.25 gram of cholesterol, and the solvent subse-quently evaporated. Consistency in the weights of solids used wassecured by employing standard solutions of both, dry ether beingused as a solvent for the cholmterol, and evaporated off beforeaddition of baryta.(i) I n the first experiment of this series the dried cholesterol-baryta mixture was covered with 60 C.C. of ether, the whole boiled formeasure$ periods of time, and cooled, when 10 C.C. of the clear,supernatant solution were removed with a pipette and evaporatedto dryness on the steam-bath, with the following results:(a) The solution attained a definite concentration of 1-95 gramsof cholesterol per litre after ten minutes a t the boiling point, andfurther boiling for two hours did not appreciably affect this.(b) On pouring off the bulk of the residual ether, and replacingit by 30 C.C.of fresh ether, a steady state was again attained, butthe concentration of cholesterol was higher than before, namely,2-30 grams per litre.( c ) On repeating the last process three times with successivequantities of 30 C.C. of ether, the steady concentrations attainedwere successively 1.7, 1.0, and 0.4 grams per litre.The results of this experiment are represented by curve A *(ii) An attempt was made to repeat experiment I1 (i), usingboiling benzene as solvent, but after ten minutes' boiling 0.24 gram,or roughly 96 per cent., of the cholesterol had been dissolved fromthe solid.(iii) Experiment I1 (i) was repeated with benzene at 25O anda.t 3 5 O , but it was found that extraction was very slow; twelvehours a t 3 5 O and at least a week a t 2 5 O elapsed before a stcadystate was attained.The results of the successive extractions areshown in Fig. 3.(iv) I n continuation of 11 (ii) above, experiments were made t oascertain whether the cholesterol removed at 80° was re-absorbeda t 25O. The results showed that no appreciable re-absorption tookplace.Series ZII.-Conditions under which Equilibrium is A ttained.Re-absorption of Cholesterol by the Ezhausted Residue.-(i) !Thefirst evidence of reverse action was obtained in the following way:A cholesterol-baryta mixture, prepared as before, was extractedfrom 60 C.C. of boiling ether, followed by five successive lots of* I n these curves, Figs. 1, 2, and 3, the abscise represent the combinedvolunies of the solvent used ; and the ordinates, the concentration of thecholesterol solution in grams per litre.Fig. 4 illustrates the influence of tempera-ture on the qllantity extracted hy 60 C.C. of ether382 NEWBERY: UNSTABLE COMPOUNDS 01'30 C.C. each, fifteen minutes being allowed for each extraction. Bythis means 0'226 gram out of 0.25 gram was extracted at constanttemperature 35O. This weight, 0-226 gram of cholesterol, wasthen dissolved in 60 C.C.of ether and the solution added to theresidue in the flask, thus giving a more concentrated solution thanhad previously been in contact with the baryta. After remaininga t 3 5 O for twenty-four hours, boiling twenty minutes, remaininganother eighteen hours, and again boiling twenty minutes, cooling,B2.01 *o0.0Volunze of ether w e d i ? ~ C.C.0 60 120 180 240 300 360I 2.0 j0 60 120 180 240 300 3609.3 *O'2.01 '00 60 120 180 240 300 360Yoltbmc of ether used in C.C.c u s 0.0 1 I0" 10" 20' 30" 40Volume of beiizene used in O.C. Temperature.and decanting, it was found that the 0.226 gram of cholesterol insolution had been reduced t o 0.205 gram. Boiling again underthe same conditions for twenty-five hours further reduced this to0.186 gram, indicating that 0.040 gram had been re-absorbed bythe solid.Further boiling for twenty-four hours produced noappreciable change of concentration.To test the accuracy of the observations, the solid residue wastreated with diluLe hydrochloric acid, and the liberated cholesteroCHOLES fEROL WITH BARIUM METHOXIDE. 383was recovered with ether and weighed. 0.064 Gram was present,showing an increase of 0*040 gram on that present a t the beginningof the reverse ” experiment.From this it appears that re-absorption does, in fact, take place,but very much more slowly than does the direct action.(ii) A similar experiment was conducted with benzene a t 3 5 Oin place of ether. Out. of a total of 0.162 gram of cholesterolextracted and then retizriied t o the mixture, with 60 C.C.ofbenzene, after remaining with frequent shaking f o r two days,0.024 gram was re-absorbed by the solid; after five days, 0.032graln; and after six weeks, 0.036 gram.Series IV.-Effect of Extent of Desiccation. of Solid.-In carry-ing out the above experiments, a solution of cholesterol with barytain methyl alcohol was in each case evaporated t o dryness beforeextracting with ether or benzene.This operation of drying the mixture was extremely difficult tocarry out efficiently without decomposing the cholesterol, andfurther experiment showed that the curves obtained were greatlymodified by more thorough drying.(i) When the methyl alcohol was boiled off in a flask on asteam-bath, and pure dry air passed over for five minutes, curve Awas obtained.(ii) When the mixture, after treating as in (i), was further driedin a vacuum for half-an-hour, curve B was obtained.(iii) On drying further a t looo in a stream of pure dry air,exhausting, and filling with pure dry air thirty times, and keepingin a vacuum for three hours, curve C was obtained.(iv) Still more careful drying led to curve D.It was also found that the more careful the drying the greaterwas the time necessary to attain constant concentration in theether solution, an hour’s boiling being needed in extreme casesin place of the ten minutes in the first experiments.From this i t is evident that the presence of water or methylalcohol profoundly affects the state of equilibrium attained in thesupernatant solution.(v) To determine the influence of water a solution was made ofanhydrous barium oxide in dry methyl alcohol, sodium-dried etherbeing used for extracting, and the same precautions taken to drythe baryta-cholesterol mixture as f o r experiment IV (iii).Theresults are illustrated by curve E.(vi) This was so different from what, had been expected that theexperiment was repeated, ab iizitio ; still more stringent precautionswere taken in drying and removing adherent alcohol, the dryingprocess being continued until no change in weight occurred, whe384 NEWBERY : C7SSTAHI.E COMPOUSDS OFthe mixture was kept in an exhausted desiccator €or twenty-fourhours. Curve F illustrates the result of this experiment.Morethan two-thirds of the cholesterol was extracted in the firstoperation.The weighings in this experiment, taken primarily with theobject of proving the mixture incapable of further drying, showedalso that 0.248 gram of methyl alcohol was firmly retained by1.111 gram of barium oxide present. Since 0.232 gram of methylalcohol is required t o form the compound HO*Ba-O*CH, with1.111 grams of barium oxide, only 0.016 gram of free methylalcohol could possibly have been present, and even this was possiblynot free, but combined in the form of barium dimethoxide.(vii) To determine the effect of excess of water on the curves,1 C.C. of water was added to a carefully dried baryta-cholesterolmixture, prepared as before, and the mixture again well dried.The whole of the cholesterol was extracted a t once by the first lotof ether.The flask containing the mixture was weighed before and afterthe addition of water, and it was found that a loss of weight hadoccurred, doubtless due t o the decomposition of methoxide bywater.From this it is evident that the barium methoxide or dimeth-oxide is in some way responsible for the retention of the chole5tero1,and the decomposition of this barium compound by water, orcarbon dioxide, as in Lapworth's experiments, sets free all thecholesterol.It would thus appear that the absen,a of water andthe presence of methyl alcohol are the two most important condi-tions for preventing extraction of the cholesterol.This conclusion was fully justiiied by the following experi-ments :Serzes Tr.-(i) A baryta-cholesterol mixture was prepared asbefore, using anhydrous barium oxide in dry methyl alcohol, butthis time was left visibly moist with methyl alcohol.The extraction curve G was obtained.On adding ether (60 c.c.)and boiling for five minutes, the concentration of the solutionbecame constant a t 0.4 gram per litre, and did not rise abovethis when boiled for three hours further, although the solubilityof cholesterol in ether a t this temperature is five'hundred times asgreat as this concentration suggests.I n the earlier stages of the work an unsuccessful attempt wasmade to absorb cholesterol from an ethereal solution by meansof baryta which had been diMolved in methyl alcohol and evapor-ated to dryness.I n the light of the last two experiments, thCHOLESTEROI, WITH BARIUM METHOXIDE. 385attempt would have succeeded i f water had been excluded andexcess of methyl alcohol had been present.This conclusion again was justified by experiment as follows:(ii) A solution of anhydrous barium oxide in dry methyl alcoholwas evaporated nearly to dryness, but left visibly moist with thealcohol. Sixty C.C. of ether containing 0.25 gram of cholesterolwere then added, and the solution was boiled gently for twelvehours. After remaining a t 1 5 O for a further twelve hours, it wasfound that more than half (0.138 gram) of the cholesterol had beenremoved from the solution by the solid. With further boiling theconcentration of the solution was again slightly diminished, show-ing that equilibrium was not fully attained in twenty-four hoursunder theae favourable conditions.From the foregoing data i t may confidently be inferred thatbarium methoxide forms a more or less definite solid compoundwith cholesterol, which is quite stable only in presence of excessof methyl alcohol.The horizontal sections of curves C, D, and Gsuggest the presence of a solid two-phase system, which probablyconsists of the solid compound, with its dissociation product,barium methoxide, above which there is a definite concentration ofcholesterol.The right-hand end of the horizontal section would thus corre-spond with the disappearance of the solid compound, whilst thegradually descending section of the curve is probably characteristicof an adsorption of cholesterol by barium oxide or methoxide orboth.The initial rising sections of €he cnrves are probably associatedwith the removal of the stabilising excess of methyl alcohol; theform of this section also suggests that adsorption of methyl alcoholoccurs, and may correspond with a three-phase solid system of solidcompound, barium methoxide, and a solid or solid solution whichcontains adsorbed or dissolved methyl alcohol.Summary and Conclusions.( a ) Under certain conditions, Smith and Mair’s original methodfor extracting cholesterol from brain matter may either fail toyield all the cholesterol or require an excessive time and quantityof solvent to remove it.The conditions are:(1) Presence of methyl alcohol in excess.(2) Low temperature of extraction (see Fig. 4).(3) Absence of water and other substances capable of decom-The reason for the difficulty of extraction appears t o be theposing barium methoxide386 DAWSON AND MARSHALL: THE REACTIONformation of a compound or possibly a series of compounds ofbarium methoxide with cholesterol.( b ) Cholesterol is not singular in the above respects, as hexadecylalcohol has been found to behave in a similar way.I n conclusion, thO author desires to express his thanks toProfessor Lapworth fcr suggesting this work, and for the interesthe has taken in its progress.THE CHEMICAL TARORATORIES,UNIVERSITY OF MANCHESTER