DENHAM THE METHYLATION OF CELLULOSE. PART m. 77 VIL-The Methylation of Cellulose. Part 111. Homogeneity of Product und Limit of Methylation. By WILLIAM SMITH DENHAM. IN previous'papers (Denham and Wooldhouse T. 1913 103 1735 ; 1914 105 2357; 1917 111 244) the methylation of cellulose and of starch was announced and some results were communicated of an investigation on the methyl ethers of cellulose and their hydrolytic proiducts. The latter include a well-defined crystalline trimethyl glucose to which a constitutional formula has been assigned. Immediate aims in the development of this research have been the preparation of a methyl ether of maximum methoxyl content in which the cellulose complex shall have undergone the minimum of degradation and the examination of the limiting and intermediate ethers and their degradation products.It is pro-posed to amplify and discuss in a subsequent communication the additional results now presented. For t8he preparation of a highly methylated cellulose the methood already described namely the treatment with methyl sulphate of cotton which has been soaked in a solution of sodium hydroxide, was found to be the most successful and repetitions of this process, variously modified have yielded a product containing 44.6 per cent. of methoxyl. This value approximates to the 45.6 per cent. od methoxyl required by theory for trimethyl cellulose. The slow rate of increase in the methoxyl content' during the final stages of the methylation itself an indication that the limit has been nearly approached has rendered advisable the additional observations now in progress before the limit can be definitely announced.That these ethers retain a high degree olf complexity is indicated by the almost complete absence of copper-reducing property in a product containing aboutl 40 per cent. of methoxyl in which as in more highly methylated preparations the fibrous structure ot cellulose persists. Further the displacement of hydrogen by methyl groups does not confer solubility in such solvents as alcohol, acetone or chloroform. The solubility of methylated cellulose in Schweitzer's reagent, on the other hand appears to decrease as the methoxyl content increases and products which contain about' 40 per cent. of methoxyl are insoluble. The question of the homogeneity of methylated cellulose is of great importance in relation to deduc-tions regarding the constitution of cellulose from the natur 78 DENHAM THE METHYLATION OF CELLULOSE.PART 111. of the hydrolytic products and some experiments which give information on this point may therefolre be referred to, although they were made primarily to ascertain whether by the restricted action of Schweitzer’s reagent especially on preparations of low average methoxyl content a highly methylated portion could be separated to servo as a convenient source of the crystalline trimethyl glucose. It was found that several preparations of methyl cellulose when treated with Schweitzer’s reagent so that only a part was dissolved yielded a residue with a methoxyl con-tent greater than that of the untreateid substance.The degree of heterogeneity thus revealed depends as might be expected on the manner of preparation of the methyl cellulose; f o r material of lolw average methoxyl content it was greatest where concen-trated alkali had been employed possibly owing to an unequal penetration of the cotton by the alkali. The most homogeneous product of those examined was one in the preparation of which an ethereal solution of methyl sulphate was employed and it may be nolted that the best yields apparent in an increased weight of material after methylation were obtained when etheir was present. A preparation containing about 40 per cent. of methoxyl gave indications of lack of uniformity explicable from its history; one containing 42 per cent.of methoxyl was quite insoluble in Schweitzer’s reagent and in alcohol acetone or chloroform and is therefore so far homogeneous. A comparison of the history of the sample of methylated cellu-lose which yielded the hydrolytic products already described and discussed (T. 1914 105 2357) with the histories of the prepar-ations made without ether which were examined by means of Schweitzer’s reagent indicates a superior uniformity in the earlier product as in the later ones there were fewer stages in the methyl-ation. It is thus probable that the trimethyl glucose was derived, not from a small proportion of highly niothylated substance but frolm a more general distribution of trimethyl glucose residues in the cellulose complex, As the limit is approached irrsgularitics in methylation will be smoothed out but the question of yield remains.The loss of material due to chemical action need not bei great. I n one series, 100 grams of cotton gave 104 grams of methyl cellulose containing 31.3 per cent. of methoxyl the yield being thus 91 per cent. of the theoretical ; the product had undergone four treatments that is it had been soaked in alkali treated with methyl sulphats and washed and dried folur times so that a t Isast a part of the loss was mechanical. Owing to losses due to mishaps the yields have not been followed continuously to the highest degree of methyl-ation attained; the loss is small however from stage to rrtage an DENHAM !CHE METHYLATION OF CELLULOSE. PART m. 79 is usually quite clearly mechanical loss sustained in manipulation.The conclusions seem to be justified1 that the limit of methyl-ation of cellulose lies in the neighbourhmd of that required for trimethyl cellulose and thatl a methyl cellulose of this limiting methoxyl content can be prepared which is representative of the whole of the original cellulose. The further investigation of these questions and of the fission products of methyl cellulose is in progress. EXPERIMENTAL. The following examples illustrate the yields obtained in several preparations of methyl cellulose and the behaviour of the products t olwar ds Sch w ei t z er ’s reagent . I. Methyl Cellulose condaining 20.4 per cent. of MethoxyJ.-The solution of sodium hydroxide employed contained 20 grams of the solid in 1OO.c.c. of solution.No solvent was employed for the methyl sulphate sIightly more of which was added than is represented by the ratio MI+~O,:N~OR. In each stage the temperature rose to1 70° during the reactioa and the mixture was afterwards warmed gently until acid throughoutl. 1st Stage.-l65 Grams of cotton wool 500 C.C. of sodium hydr-oxide solution 390 grams o€ methyl sulphate. Yield 135 grams. Methoxyl content not determined. 2nd Stage.-135 Grams of the product from the first stage, 400 C.C. of sodium hydroxide solution 236 grams of methyl sulphate. Yield 98 grains. OMe=20*4 per cent. Action of Schweitzer’s Reagent.-The prod-uct from the second stage after remaining for seven days with 4 litres of Schweitzer’s reagent leftl 42 grams of undissolved residue containing OMe=22*7 per cent.Three quantities of 3 grams each of the substance containing OMe= 22.7 per cent. were again treated each with 170 C.C. of Schweitzer’s reagent: (a) Residue after two days (1.5 grams) contained OMe= 23.1 per centl. ( 6 ) Residue after five days (1.3 grams) contained OMe=24.3 per centi. ( c ) Residue after seven days (1.3 grams) contained OMe=24.7 per cent. The mixed residues were completely soluble in Schweitzer’s reagent. ( d ) By the action of Schweitzer’s reagent on 28 grams of the substance containing OMe= 22.7 per cent. 4.4 grams of undis-solved residue were obtained which contained OM@= 23-4 per cent. The 4.4 grams were again treated with Schweitzer’s reagent so as to leave a small residue (weight not determined) which contained OMe=28 per cent 80 DENHAM THE METHYLATION OF CELLULOSE.PART 111. The Schwelitzer s reagent was prepared t y shaking freshly pre cipitated copper hydroxide with concentrated aqueous ammonia (D 0*88) and had D aboiut 0.93. It was not standardised as regards content of ammonia or copper but the reagent was tested to make sure that it dissolved cotton. The undissolved cellulosic residue was recovered by diluting the solution with aqueous ammonia filtration through nickel gauze and washing with aquelous ammonia and water. Small quantities were relcovered by allowing the filtered solution to settle. The methyl cellulose the hydrolytic prolducts froim which have beien described (T. 1914 105 2357) was prepared in six stages; in the first five the solution of sodium hydroxide cantaineld 17 grams of the solid in 100 c.c.and the product contained OMe=20.7 per cent. I n the final stage the sodium hydroxide solution contained 38.3 grams of the solid in 100 c.c. and the productl contained OMe = 23.6 per cent. 11. Methyl Cellulose con taming OMS = 39.5 peT cent .-This substance was prepared in nine stages. After three stages in each of which the material was impregnated with half its weight of sodium hydroxide added as a solution containing 21.4 grams of the sollid in 100 c.c. the OMe content was 24.8 per celnt. In all the stages somewhat less methyl sulphate was employed than is reipresented by the ratio Me2S0,:NaOH and the substance remainelcl alkaline after the reaction was over. From 160 grams of cotton wool were obtained 135 grams the yield being 7 9 per cent.of that obtainable of a product having OMe=24-8 per cent. In the later stages the1 sodium hydroxide solution contained 38.8 grams s f the solid in 100 C.C. Difficultiw were encountered in impregnating the1 material with the alkaline solution and the material was cut up into small pieces squeezed in a press pounded, etc. so’ that there was considerable loss. The percentage of metholxyl remained stationary from the eighth to the ninth stage (OMe = 39.7-39.5 per cent.). A c t i o n of Schweitzer’s Reagent.-2.8 Grams of the product con-taining OMe=39*5 per cent. after treatment with 50 C.C. of Schweitzer’s reagent for four days gave 2.6 grams of undissolved residue containing OMe =40 per cent. ; 2.4 grams of this residue, on further treatment with Schweitzer’s reagent were reduced to 2-2 grams which contained OMe=40*4 per ce!nt.Copper Reduction.-0*978 Gram of substance (OM@= 39.5 per cent.) was treated with Fehling’s solutioln as in the examination of cottoin. 0.15 C.C. olf iV/lO-sodium thiosulphate was required for the titratioin of the reduceld copper. 111. Preparation of M e t k p l Ce7liiloaP using Ether.-One Cu,O=Q~ll per cent DENHAM THE METHYLATION OF CELLULOSE. PART III. 81 hundred grams of cotton wool were soaked for one day in 1 litze od a solution of sodium hydroxide (20 grams in 100 grams of solution) and then drained and pressed so thatl 350 grams od‘ the solution (coataining 70 grams of sofdium hydroxide) were left. One hundred C.C. of methyl sulphate and 700 C.C.of ether were added the mixture was heated under reflux for a few hours and the ether was then removed by distillation. The weight of the washed and dried profduct was 105 grams. A repetition of this process gave 107 grams oif EL product containing OMe=24*5 per cent. In a third and fourth repetition of the process where the impregnation of the material by the solution ot alkali was assisted by exhausting the containing vessel less sodium hydroxide was retained. The product from the fourth treatment contained OMe=31-3 per cent. and weighed 104 grams the yield being thus 91 per cent. of the theoretical. The weights were determined under similar conditions after drying the material in a steam-oven. Action of Schweitzer’s Reagent.-From 2.88 grams of subst4ance (OMe=31.3 per cent.) after immersion in 100 C.C.of Schweitzer’s reagent for three hours 1.93 grams of undissolved substance (OMe=32.4 per cent.) were recovered. Eighty-eight grams of methyl cellulose containing OMe = 15.6 per cent. which had been made in one stage using ether were immersed in 600 C.C. of Schweitzer’s reagent for three days when 8.4 grams of undissolved material were recovered containing OMe=15.2 per cent. IV. Methyl Cellulose containing moTe than 40 per cent. of Metho&.-Highly methylated products were prepared by variants od the methods already described. For one of these the following analytical results were obtained (Found C =52-19 ; H = 7-63 ; OMe= 44.6 per cent. Trimdhyl cellulose C,H,,O,, requires C=52.9; H=7*84; OMq=45*6 per cent.). A ctim of Solvents.-The behaviour of a preparation containing OMe=42 per cent. was examined towards alcohol acetone and Schweitzer’s reagent. 1.8802 Grams after being extracted with alcohol for three hours in a modified Soxhlet apparatus weighed 1.8796 grams. The same sample was then extracted in the same way with acetone for three hours and afterwards weighed 1.8790 grams. 0.408 Gram of the same substance was immersed in Schweitzer’s reagent for twenty-four hours; the weight of the undissolved substance recovered was 0.407 gram. CHEMISTRY DEPARTMENT, ST. MARY’S HOSPITAL MEDICAL SCHOOL, PADDINGTON W. 2. [Received November 29th 1920.