THE ESTIMATION OW THE METEOXYL GROUP. 193 of the MethoxyE Group. By JOHN THEODORE HEWITT and WILLIAM JACOB JONES. SINCE the introduction of the Zeisel method for the estimation of methoxyl groups (Moizutsh. 1885 6 989; 1886 7 406; Benedikt and Griissner Cken2. Zce't. 1889 13 872) numerous attempts have been made to simplify the apparatus and shorten t'he operation. The msthod of rectifying the methyl iodide proposed by Zeisel is effective but troublesome; the inclined condenser has t o be of con-siderable lengkh and consequently occupies some bench space whilst the constant supply of water t o the condenser a t about 50° necessi-tates supervision. A further inconvenience of the Zeisel method is the trouble experieneed. in working up the precipitate of double iodide and nitrate of silver time being lost in evaporating the alcoho'l and obtaining the silver iodide in a pure condition.The use of a rectifying column provided witqh a thermometer in place of the inclined condenser fed with water a t an approximately d e h i t e temperature was recommended by Hewitt and Moore (T., 1902 81 318) and impurities in the hydriodic acid were removed by passing carbon dioxide through the acid a t 130° before-intro-ducing the substance under examination. Considerable saving in bench space was effected and during the operation it was only necessary to control th0 stream of carbon dioxide and the flame under the glycerol bath so as to give the necessary temperatuies in the reaction flask and a t the top of the rectifying colamn. Several modifications of the Zeisel method have been suggested in which ordinary rectification has been relied on in place of inclined condensers in which the temperatare is controlled by running water (Perkin T.1903 83 1367; Zeisel and Fanto t e i t s c l t . anal Chem. 1903 42 549; Stritar ibid. 579; fIewe Ber. 1906 39, 1142). Shortening t'he process by estimation of the methyl iodide ia a more rapid manaer was left. untouched for years. The necessity of esti&iag methyl iodide obtlained from methyl alcoho 194 HEWITT AND JONES: mixtures rapidly and accurately caused the present authors to search €or a quick process. Since combination of alkyl iodides with bases of the pyridine series takes place very rapidly there seemed to be a promising way of obtaining the iodide in an ionisable form and then estimating it volumetrically.After working out a satisfactory process it was found that the same fundamental idea of combining the methyl iodide with a tertiary base had already been utilised (Kirpal and Buhn Ber. 1914 47 1084) but the subsequent volumetric estimation of iodide by standard silver nitrate solution may be con-siderably shortened. Instead of rejecting the excess of pyridine by evaporation and estimating the iodide with standard silver nitrate using a chromat?e as indicator the pyridine and its meth-iodide may be directly diluted with water acidified with nitric acid a known amount of silver nitrate added and the excess of the latter determined by thiocyanate according to Volhard's method. Adoption of this procedure reduces the time of experi-ment considerably.The applicability of Volhard's method to the estimation of methyl iodide after reaction with pyridine was controlled i n a separate experiment. 3.10 Grams of freshly distilled methyl iodide were diluted to a volume of 100 C.C. with pyridine which had been saturated with carbon dioxide. By dilution with water addition of silver nitrate and determination of the excess of silver wit'h thiocyanate 3.13 grams of methyl iodide per 100 C.C. were found. Method. Hydriodic A cid.-The hydriodic acid is prepared by saturating an aqueous suspension of iodine with hydrogen sulphide distilling the resulting solution and collecting the fraction boiling between 123O and 127O (D 1.7) for use. Residues from analyses are redistilled and used again.Pyridime .-Complete separation from picoline is unnecessary ; it is however advisable to remove substances of high boiling point, Pyridine bases as obtained from tar distillers are mixed with two-thirds of their weight of water and fractionated. The fraction distilling betkeen 9 3 O and 97' containing the mixture of constant boiling point is shaken with one-third of its weight of solid sodium hydroxide the upper layer is separated fractionally distilled and the bases distilling between 114' and 1 1 7 O are collected for use, Residues containing pyriding may be accumulated rendered alkaline with onefiftieth of their weight of solid sodium hydroxide distilled, the distillate a t 93-97O collected and worked up THE ESTIMATION OF THE METHOXYL GROUP.195 The Estimation.-The apparatus consists of the usual carbon dioxide generator decomposition flask heated in a glycerol bath to 130° and rectifying column (four-pear or other suitable form). The carbon dioxide carrying the methyl iodide vapour is passed through two tesbtubes in series each containing 10 C.C. of pyridine. For the estimation a suitable weight of the substance i b taken, and 20 C.C. of hydriodic acid (D 1.7) are added. After the experiment has been in progress for one hour the con-tents of the tesbtubes are completely washed into a graduated flask, when they are diluted with water and the iodide is estimated by the process indicated above. In all cases early in the course of t8he experiment a yellow coloration develops in the pyridine.This vanishes on diluting the pyridine a t the end of the experiment. Its appearance therefore, need cause no apprehension to the analyst that free iodine is finding its way into the pyridiiie. Moreover carefully purified samples of metqhyl iodide and pyridine on admixture' develop colour. The behaviour of a mixture on dilution with water was compared with that of an iodine solut.ion of equal depth of colour. On diluting a 3 per cent. solution of methyl iodide in pyridine with a quarter of its volume of water the colour became very pale and with its own volume of water it almost vanished. The iodine solution on being similarly diluted still retained its colour. It would thus appear that the coloration is not due to free iodine. Should how-ever the colour persist on dilution with water it is then advisable to discharge it with thiosulphate solution.Analyses. Substance. Brucine hydrate .................. Methyl oxalate .................. Methyl alcohol ..................... 9 ....................... Y )) ..................... 99 , ..................... Methyl salicylate .................. Percentage of methovyI group /-Found. Calculated 13.5 13.3 50.1 52.5 95.2 96.9 95.2 96.9 96.0 96.9 96.3 96.9 19.7 204 A methylat'ed cellulose which gave 39.1 per cent. of rriethoxyl by the gravimetric Zeisel method was found to contain 39.2 per cent. by the present method. Unsatisfactory results were obtained with a sample of methyl benzoate and with one of hydrated quinine sulphate. This comparatively rapid method for the estimation of methoxy 196 HPWIT!C AND JONES: groups may be applied conveniently to the products of wood dis-tillation or other mixtures containing methyl alcohol.Methyl dcohol has asually been estimated in these products by conversion into methyl iodide and measurement of the volume of the latter compound (Krell Ber. 1873 6 1310; Grodzky and Kriimer ibid. 1492). Zeisel and Stritar's process of weighing volatilised iodine as silver iodide obviates the inexactness due t o determination of the volume of the methyl iodide but time may be saved by combining the methyl iodide with a tertiary base and estimating the iodine volumetrically . A suitable amount (see below) of the liquid to be analysed is heated with 20 C.C. of hydriodic acid (D 1.7) for one hour.The contents of the test-tubes are then completely washed into a graduated flask and made up with water t o 100 C.C. An aliquot portion (see below) of the diluted solution is introduced into a glass stoppered bottle of 250'c.c. capacity 70 C.C. of water are added and tihen in order 25 C.C. of NIIO-silver nitrate solution and 30 C.C. of approximately IOlV-nitric acid. The bottle is well shaken by hand for five minutes and 5 C.C. of concentrated ferric alum indicator are added. N / 10-Thiocyanate solution is now run in until further addition of one drop imparts a permanent orange cdour to the liquid. Suitable amounts of liquids to be taken for analysis are given below. a represents the volume of material to be operated on and its dilution when necessary.b gives the volume of the diluted aqueous pyridine solution, obtained as described above to be actually used in a titration. c is the formula to be used giving the weight in grams of methyl alcohol in 100 C.C. of the liquor analysed where t is t'he number of C.C. of N/lO-thiocyanate solution used in the titration. (Note.-The figure 25 given in the formulze must be multiplied by f the factor for the silver nitrate solution if this is not exactly deci-normal. ) Pyroligncous Acid.-(a) Take 5 C.C. of the original liquor; ( 6 ) 40 c.c.; ( c ) 0.16 ( 2 5 - t ) . Crude Wood Naphtha.-(a) Take 10 c.c. dilute to 100 c;c. with water and use 5 C.C. of the diIuted solution for distillation with hydriodic acid; ( b ) 40 C . C . ; ( c ) 1.6 (25 - t ) .Methyl Alcohol and Mixtures o./ the Alcohol with Acetone.-(a) Take 10 c.c. dilute to 100 C.C. with water and use 5 C.C. of this diluted solution for the estlimation; ( b ) 20 c.c.; ( c ) 3.2 (25- t ) THE ESTIMATION OF TEE METHOXYL GROUP. 197 Amlyses. Art.ificia1 mixtures containing methyl alcohol and other products of wood distillation were made up and analysed by the mebhod described. Composition. No. of grams of methyl alcohol per 100 C.C. of liquor. Number. Actual. Found. Liquor 7-1 ........................ 2.39 2.35 2 ........................ 68.5 68.1 3 ........................ 2.39 2-35 4 ........................ 31-8 31.6 Anklyses of purified methyl alcohol by the present method have already been given. It will be seen that on an average the results are 1 per cent.too low. Stritar and Zeidler (Zeitsch. anal. Ckem. 1904 63, 387) found that the maximum amount of methyl iodide obtainable from pure methyl alcohol in a Zeisel estimation corresponded with a 99 per cent. yield. Liquors 1 and 2 were aqueous solutions of purified methyl alcohol. Liquor 3 contained per 100 c.c. 7-03 grams of acetic acid 0.80 gram of acetone and 2.39 grams of methyl alcohol the remainder being water. It represented a pyroligneous acid. Liquor 4 was an equilibrium mixture prepared from 10.04 grams of acetic acid 39.60 grams of acetone and 31.82 grams of methyl alcohol made up to 100 C.C. with water (1.27 grams). This mix-ture which of course contained methyl acetate represented the first runnings obtained in the rectification of crude wood naphtha.Compounds yielding Meth4yl Iodide other t h m Methyl AZcohd, present in Wood Distillates. The constituents of wood distillates have been examined by Stritar and Zeidler (Zoc. cit.) with a view to determine which yield methyl iodide on treatment with hydriodic acid. They found that acetone gave no methyl iodide whilst the yield from both form-aldehyde and acetaldehyde was negligible. Methyl acetate gave one equivalent. of methyl iodide and methylal and dimethylacetal each gave two. Ally1 alcohol yielded its equivalent of sec.-propyl iodide. Guaiacol and other methyl derivatives of the phenols yielded their equivalent of methyl iodide. Of these substances the quantities of acetd encountered are too VOL. axv. 198 FRANKLAND CHALLXNGZR AND NIOgOLLS : small to be of consequence.According to Grodzki and Kriimer (Zoc. cit.) the amounts of allyl alcohol and of methyl alcohol in raw wood spirit are in the ratio of 2 t o 1000. Stritar and Zeidler find that guaiacol may be .eliminated from aqueous solutions of crude wood spirit by shaking with animal charcoal. They state that on omitting this treatment the amount of methyl alcohol found is t,oo high by about 2 parts in 100 parts. It will be seen that the quantities of alkyl iodide yielded by the amounts of allyl alcohol and of guaiacol present in the aqueous wood distillates, relatively to that yielded by the methyl alcohol present border closely on the experimental error. It is the experience of the authors that where precautions are taken t o eliminate these sub-stances preliminary t o analysis the errors due to losses outweigh the error introduced through ignoring their presence. The present method gives the total methyl alcohol including both the free alcohol and that which is combined as methyl acetate. I f it is desired the amount of ester present may be determined by quantitative hydrolysis. The alcoholic silver nitrate of the Zeisel method of estimating methoxyl may be replaced by pyridine. The pyridinium methyl iodide formed can be determined by Volhard's thiocyanate method. Methyl alcohol in wood dist'illates may be determined by the method described above. [Received January 171h 19 19.