2 ADAMS AND NICHOLLS THE ANALYSIS OF MIXTURES CONTAINING The Analysis of Mixtures containing Acetone, Ethyl Alcohol and Isopropyl Alcohol. BY CHARLES AMBROSE ADAMS B.Sc. A.I.C. AND JOHN RALPH NICHOLLS BSc. F.I.C. (Read at the Meeting December 5 1928.) I. INTRODUCTION.-Both acetone and isopropyl alcohol find application as substitutes for ethyl alcohol and the analysis of a mixture of the three substances is occasionally required. The object of this paper is to give analytical methods which have proved satisfactory in this laboratory for the past three or four years. usual method for the determination of ethyl alcohol is based upon the specific gravity of its aqueous solution. Many substances which by their presence would invalidate this method of determination can be removed by a petroleum spirit and brine separation (Thorpe and Holmes J .Chem. Soc. 1903,83 314-7). Other substances commonly associated with ethyl alcohol in mixtures may however be only partly if at all removed by this preliminary treatment and in this category appear acetone methyl ethyl ketone methyl alcohol isopropyl alcohol rt-propyl alcohol benzyl alcohol etc. Hence the specific gravity of an aqueous solution of ethyl alcohol containing one or more of these ingredients affords no reliable criterion of the proportion of ethyl alcohol and the presence of an interfering sub-stance may not be anticipated. The purity of such aqueous alcohol solutions may be tested readily by means of the refractometer the reading obtained for the test solution being compared with that given by a pure ethyl alcohol solution of the same specific gravity.Tables showing the readings given by the Immersion Refractometer for aqueous solutions of methyl ethyl isopropyl and n-propyl alcohols obtained in this laboratory by J. Holmes in 1911 have been published (see “Alcohol,” by Simmonds; Macmillan & Co. 1919 pages 285 and 287). A similar table for acetone is given in the following paper. A comparative table interpolated from these results is given later in a form convenient for practical use. With the exception of methyl alcohol solutions of the other substances referred to give higher refractions than solutions of ethyl alcohol of the same specific gravity. Methyl alcohol can be tested for by the Den@ method (Simmonds ANALYST, 1912 37 16; Jones id.1915,40 221 etc.). It is to be noted that n-propyl amyl and benzyl alcohols when present in sufficient proportions give positive results in this test which is not therefore specific for methyl alcohol. The test is very valuable however for indicating the absence of methyl alcohol. Only in those cases therefore where the refractometer reading of a solution, purified if necessary by the Thorpe and Holmes method agrees with that given by a corresponding solution of ethyl alcohol and where the Deniges test has given a PRELIMINARY REMARKS ON THE DETERMINATION OF ETHYL ALcoHoL.-Th ACETONE ETHYL ALCOHOL AND ISOPROPYL ALCOHOL 3 negative result can the specific gravity of a solution be used with certainty for determining the ethyl alcohol. In our opinion the refractometer reading (or the determination of some other physical constant) should always be employed as a check upon a determination of ethyl alcohol made from the specific gravity of an aqueous solution unless the absence of interfering substances is known with certainty.SPECIFIC GRAVITIES AND REFRACTIONS OF AQUEOUS MIXTURES OF ACETONE AND THE LOWER ALcoHoLs.-h the case of an aqueous solution of acetone or of one of the lower alcohols neither the specific gravity nor the refraction is a linear function of the quantity in solution. When however the proportion present does not exceed about 10 per cent. by volume the function is very nearly linear. For purposes of calculation it is convenient to express the strengths of solutions in terms of a common unit and for this purpose each liquid is regarded as though it were the “proof spirit” of the Official Specific Gravity Tables (H.M.Stationery Office 1916). The following table shows the relation between percentages of apparent “ proof spirit ” and percentages by volume of some of the lower alcohols and of acetone. The ethyl alcohol figures are taken from the Official Specific Gravity Tables; the figures for methyl rt-propyl and iso-propyl alcohols are interpolated from data obtained by Holmes and those for acetone are interpolated from data obtained by one of us and given in the following paper : TABLE I. COMPARATIVE STRENGTHS (PERCENTAGE BY VOLUME). Apparent proof spirit. Per Cent. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Mean value for 1 per cent.proof spirit Methyl alcohol. Per Cent. 0.60 1 *20 1431 2.42 3.01 3-61 4.19 4.76 5.31 5.85 6.42 6-99 7.57 8.13 8-68 9.33 9-88 0.585 Ethyl alcohol. Per Cent. 0.57 1.15 1-72, 2-29 2.86 3.43 4.0 1 4.58 5-17 5-73 6-30 6.87 7-45 8-01 8.58 9-16 9.74 0.573 n-Prop yl alcohol. Per Cent. 0.62 1 *26 1.90 2.53 3.17 3-84 4.50 5.16 5.80 6-50 7.19 7.87 8.55 9.20 9489 10.55 11-22 0.655 Isoprop yl alcohol. Per Cent. 0.60 1-20 1-80 2-40 3.00 3.60 4.21 4.83 5-44 6.08 6-69 7.33 7.97 8.59 9.20 9-83 10.47 0.612 Acetone. Per Cent. 0.76 1.53 2.32 3.11 3.89 4.66 5-41 6-15 6.91 7.69 8.42 9-20 9.94 10.65 11-36 12-05 12-75 0.76 4 ADAMS AND NICHOLLS THE ANALYSIS OF MIXTURES CONTAINING The use of this mean value will give a percentage by volume not differing by more than about 0.1 per cent.from the correct value. The following table gives the corresponding refractometer readings. All except the acetone figures which are obtained from the following paper are inter-polated from data obtained by Holmes in this Laboratory : TABLE 11. COMPARATIVE REFRACTOMETER READINGS. Immersion Refractometer reading a t 60" F. Apparent A \ proof Methyl Ethyl n-Propyl spirit. alcohol. alcohol. alcohol. P& Cent. 0 15.4 1 15.7 2 16.0 3 16.2 4 16.5 5 16.8 6 17.1 7 17.4 8 17.7 9 17.9 10 18.2 11 18.5 12 18.8 13 19.1 14 19.3 15 19.6 16 19.9 17 20.2 15.4 16.1 16.8 17.5 18.2 18.9 19.7 20.5 21.2 22.0 22.8 23.6 24.4 25.3 26.1 26.9 27.8 28.7 15.4 16.5 17-7 18-9 20- 1 21.3 22.6 23.9 25.2 26.6 28.0 29.4 30.8 32.2 33-6 35.0 36.4 37.8 Isoprop yl alcohol.15.4 16.5 17.6 18.6 19.6 20.6 21.7 22.8 23.9 25.0 26.2 27.4 28.7 29.9 31.2 32.4 33.7 34-9 Acetone. 15.4 16.6 17.9 19.1 20.3 21.6 22.8 24.0 25.3 26.5 27.8 29.0 30.2 31.4 32.6 33.7 34.9 36.0 We have repeatedly found that with aqueous mixtures of these substances, provided the total proportion does not exceed about 10 per cent. by volume the apparent proof strength and the refraction of the mixture are practically the sum of those due to each of the ingredients.Hence in mixtures of the above substances it may be stated that up to a strength of about 17 per cent. of apparent proof spirit the specific gravities and refractions are for practical purposes additive factors. With two known substances alone present the quantities of each can be calculated by proportion from two determinations vix. the specific gravity and the refracto-meter reading. With three known substances present if one can be determined by any independent process allowance can be made for it in the strength and re-fraction of the solution and the proportions of the other two can then be calculated. Similarly in an unknown mixture any ingredients which can be separately determined can be allowed for and the resulting figures give a measure of the undetermined ingredients.It is desirable always to have direct methods for the detection and determination of substances likely to be present and the following methods are applicable to mixtures containing acetone ethyl alcohol and isopropyl alcohol ACETONE ETHYL ALCOHOL AND ISOPROPYL ALCOHOL 6 11. THE DETECTION AND COLORIMETRIC DETERMINATION OF ACETONIL-The test used by Penzoldt for the qualitative detection of acetone (Dezct. Arch. klin. Med. 34 127; referred to in 2. anal. Chzem. 1885 24 149) appears to have escaped general notice. We have found this test satisfactory for the past four years and have modified it to furnish a rapid quantitative colorimetric method for the determination of acetone. When sodium hydroxide is added to a mixture of o-nitrobenzaldehyde and acetone indigo is formed as a condensation product: H ~ c > o q ~ CH(0H) .CHpCO.CHa 0 + H& NaOH -@;>C=C<;-Q CH(0H) .CH,.CO.CH, + 2CH,COOH + 2&0.The indigo separates almost at once as a flocculent precipitate and its appearance is conclusive evidence of the presence of acetone. With a very small concentration of acetone however precipitation of indigo may not at first be apparent but the solution will develop a colour ranging from yellow through yellowish-green to greenish-blue depending on the proportion of acetone present. By allowing the solution to stand for about 1 hour and filtering through a small white paper a distinct indication of precipitated indigo can be obtained from 10 ml.of a 0-05 per cent. solution of acetone. The yellow and yellowish-green colours produced in solutions containing not more than 0.2 per cent. of acetone are suitable for quantitative colorimetric deter-minations. Under the conditions described the colour reaches its maximum in 10 to 15 minutes and a faint but perceptible colour is given by 1 mgrm. of acetone in 10 ml. (0.01 per cent. of acetone). The o-nitrobenzaldehyde solution used tends to darken in colour on keeping. A recently-prepared solution should be used for quantitative work and the blank should be colourless. The quantitative method is as follows:-To an aliquot portion of the distillate to be tested (containing not more than 0.02 grm. acetone and diluted with water to 10 ml.) is added 1 ml.of a 1 per cent. solution of o-nitrobenzaldehyde in 50 per cent. ethyl alcohol (not methylated spirit as this contains acetone). After mixing 0-5 ml. of a 30 per cent. solution of sodium hydroxide is added and the test solution allowed to stand for about 16 minutes avoiding strong daylight. At the end of this time the colour is compared with the colour developed in a set of standard acetone solutions containing from 0 to 20 mgrms. of acetone in 10 rnl. which have been treated similarly at the same time. The range of the colours produced is very marked and it is possible to have as many as twenty readily differentiated standards within the range suggested 6 ADAMS AND NICHOLLS THE ANALYSIS OF MIXTURES CONTAINING 111. DETECTION AND COLORIMETRIC DETERMINATION OF ISOPROPYL ALCOHOL.-1sopropyl alcohol can be oxidised readily to acetone which can then be deter-mined by the above method. We have found the most convenient oxidising agent for this purpose to be potassium dichromate and sulphuric acid under the con-ditions to be described later. For the purpose of the routine testing of alcoholic distillates however it was desired to avoid the distillation necessary in the dichromate oxidation and to use a method of oxidation which could be applied simultaneously to a large number of distillates. Various oxidising agents were tried. Dilute nitric acid hydrogen peroxide and sodium peroxide tested under various conditions gave poor results. Potassium permanganate either in acid or alkaline solution gave good results, but the use of this reagent necessitated the filtration of the test solutions.Strong nitric acid was too vigorous in its action. Bromine water gave excellent results, and the conditions for the use of this reagent were therefore worked out. For the detection of isopropyl alcohol in solutions containing ethyl alcohol the method is as follows:-The test solutions (after appropriate purification) are diluted to a strength of approximately 10 per cent. by vol. of alcohol. To 10 ml. of each solution in a test-tube 5 ml. of saturated bromine water are added. The tubes are lightly corked and allowed to stand for about 3 to 6 hours in the cold or even over-night in a cool dark cupboard. After the prescribed period of standing 1 ml. of a 1 per cent. solution of o-nitrobenzaldehyde in 50 per cent.ethyl alcohol is added and the solutions are mixed by gentle shaking. Finally 2 ml. of a 30 per cent. solution of sodium hydroxide are added and the solutions shaken once more. The tubes are then allowed to stand for about 15 minutes in diffused daylight and the colours developed are compared with those produced in standard isopropyl alcohol solutions similarly treated. A convenient standard solution is one containing 2.5 ml. of pure isopropyl alcohol in 100 ml. of 10 per cent. ethyl alcohol. The standard colours are those produced by using 1-0 1.5 2.0 ml. etc. of this solution made up to 10 ml. with 10 per cent. ethyl alcohol. Under the conditions described the test will detect with certainty the presence of 0.025 m1. of pure isopropyl alcohol in the 10 ml.of 10 per cent. alcoholic solution used. The method is not so delicate as when applied directly to acetone owing to the complex nature of the action of bromine on isopropyl alcohol but the pro-portion of acetone produced by this method from a given quantity of isopropyl alcohol in a given time under similar conditions appears to be constant. In routine testing the value of the indications is not appreciably affected by the presence of small proportions of formaldehyde or methyl alcohol. Normal propyl alcohol in this test however gives a brownish colour on the addition of soda probably due to the resinification of propyl aldehyde produced by the treat-ment with bromine. Although this colour gradually fades it is liable to interfere with the shade of colour produced from isopropyl alcohol.In the presence of wpropyl alcohol a quantitative determination of isopropyl alcohol by this method should only be attempted when the proportion of fi-propyl alcohol is known s ACETONE ETHYL ALCOHOL AND ISOPROPYL ALCOHOL 7 that the equivalent quantity of n-propyl alcohol may be added to the standards before oxidation with bromine. In this Laboratory the chief value of the test has been found in its giving a ready qualitative method of examination applicable to a large number of alcoholic solutions at a time. Where a positive indication is obtained the test is repeated omitting oxidation with bromine in order to detect whether acetone is present. Quantitative deter-minations are then made as described above if necessary both before and after oxidation.Alternatively the dichromate oxidation process described later may be employed. IV. DETERMINATION OF ACETONE ETHYL ALCOHOL AND ISOPROPYL ALCOHOL MIXTURES.-FOr the determination of acetone the above-described colorimetric method is suitable or the well-known Messenger process can be employed. For ethyl alcohol and isopropyl alcohol it was thought that if conditions could be found whereby these could be oxidised completely to acetic acid and acetone respectively, then the determination of these products would enable the proportions of the alcohols to be calculated. Of the oxidising agents tried potassium dichromate and sulphuric acid gave the most promising results especially as it was found that the oxidation could be carried out at room temperature in a closed vessel thus obviating loss.The behaviour of each of the three substances under various conditions of oxidation was studied. Oxidation of Ethyl Alcohol.-The factor which is most pronounced in deter-mining the rate of oxidation at room temperature is the proportion of sulphuric acid. With small proportions of sulphuric acid oxidation to acetic acid takes many hours or even days whilst with fairly large proportions it is complete in a few minutes. Unless the quantity of acid is extremelylarge oxidation does not go beyond acetic acid. Under the conditions finally chosen the time required is less than 15 minutes. After reduction of the excess of dichromate the acetic acid can be steam-distilled and titrated. Oxidation of Isopropyl A lcokol.-The proportion of sulphuric acid similarly determines the rate of oxidation which proceeds regularly to acetone; with high proportions of acid the acetone formed is further oxidised.Under the selected conditions the oxidation is complete in 15 to 20 minutes. After reduction of the excess of dichromate the acetone produced can be distilled. Oxidation of Acetom.-With moderate proportions of sulphuric acid oxidation is extremely slow but the rate of oxidation increases with the proportion of acid. Under the selected conditions no appreciable action takes place in half-an-hour. Method of Oxidation.-The concentration of sulphuric acid in the oxidising mixture which gives the best results is 12.5 ml. of concentrated acid per 100 ml. The proportion of potassium dichromate present is not so important but about 5 grms.per 100 ml. is a suitable quantity. It is convenient to prepare a stock oxidising mixture of double strength to be mixed with an equal volume of the solution to be oxidised 8 ADAMS AND NICHOLLS THE ANALYSIS OF MIXTURES CONTAINING An aliquot part of the distillate to be tested containing not more than 1.5 grm. ethyl alcohol or 3 grms. of isopropyl alcohol is placed in a distillation flask of about 800 ml. capacity and diluted with water to approximately 100 ml. An equal volume of the stock oxidising mixture (10 grms. of potassium dichromate and 25 ml. of conc. sulphuric acid per 100 ml.) is added and the flask is corked and allowed to stand for 25 to 30 minutes. An excess of ferrous sulphate is added and the solution is steam-distilled the contents of the flask being allowed to concentrate to about 100 ml.The concentration should not be carried too far or the distillate becomes yellowish and gives an acidity not due to acetic acid. This occurs when evaporation has proceeded so far that iron compounds separate in the distilling flask and cause bumping. If the volume of liquid is not reduced below about 100 ml. this condition is completely avoided. The whole of the acetic acid should be in the distillate when about 500 to 600 ml. have been collected but it is advisable to collect a further 100 ml. in a fresh receiver. The distillate is titrated with normal sodium hydroxide solid phenolphthalein being used as indicator. The addition of 10 to 20 grms. of common salt makes the end point rather sharper.1 ml. of N alkali=0@46 grm. of ethyl alcohol=0*058 ml. of ethyl alcohol=O.lOl ml. of proof spirit. The neutralised solution is then redistilled the distillate being made up to a known volume and the acetone present determined from the speciiic gravity of the solution (see following paper) or if small in amount colorimetrically. The acetone found represents that originally present together with that produced from the isopropyl alcohol. 1 ml. acetone = 1.043 ml. isopropyl alcohol. The following results were obtained by this method:-EthyZ AZcohoZ :-lo ml. of a solution of specific gravity 0.97840 (=lo761 ml. Acetic acid titration=30*35 ml. of N/1 alkali. Ethyl IsopropyZ AZcohoZ:-25 ml. of a solution of specific gravity 0.98770 (=2-43 ml.isopropyl alcohol) oxidised. Acetic acid titration=0*2 ml. of N/1 alkali (equivalent to 0.01 ml. ethyl alcohol). The recovered acetone diluted to 100 ml. had specific gravity 0*99748=2*28 ml. acetone =2*38 ml. isopropyl alcohol. Acetorte:-398 grms. (=5*00 ml.) pure acetone treated. Acetic acid titra-tion=O*l ml. of N/1 alkali. The re-distilled acetone diluted to 100 ml. had specific gravity 0*99467=4*96 ml. acetone. present the above oxidation method can be used to determine isopropyl alcohol, but not ethyl alcohol. rt-Propyl alcohol is oxidised by dichromate and sulphuric acid and as with ethyl alcohol the proportion of acid determines the rate of oxidation. The product however is not solely propionic acid but a mixture of that acid and acetic acid. The relative proportion of the two varies with the conditions the ratio of acetic acid to propionic acid increasing with the proportion of acid.No conditions were found which at room temperature would give solely ethyl alcohol) oxidised. dcoh01=30*35 X 0*058=1*760 d. V. NOTE WITH REGARD TO N-PROPYL ALcoHoL.-When %-prOpyl alcohol i ACETONE ETHYL ALCOHOL AND ISOPROPYL ALCOHOL 9 propionic or solely acetic acid. Under the above specified conditions of oxidation rt-propyl alcohol gives about 70 per cent. propionic acid and 30 per cent. acetic acid. Although the product is a mixture 1 mol. of .n-propyl alcohol gives 1 mol. of acid so that after steam distillation the mixed acid can be titrated; and with fin-propyl alone present 1 ml. of N/1 alkali=0-060 grm. rt-propyl alcoho1=0*0743 ml. rt-propyl alcohol. With mixtures of ethyl and n-propyl alcohols the titration of the acids produced by oxidation may form a useful check of the proportions of the two alcohols calculated from the strength and refractometer readings of their aqueous solution. VI. SUMMARY.-~. Comparative Tables are given of the strengths and immersion refractometer readings of aqueous solutions of acetone and of some of the lower alcohols. 2. Provided not more than 10 per cent. by volume of these substances are present in aqueous mixtures the specific gravities and refractometer readings are approximately additive factors. 3. A rapid method for the detection and colorimetric determination of acetone is given with an adaptation of the process to the determination of isopropyl alcohol. 4. Conditions are given for the complete oxidation of ethyl alcohol and isopropyl alcohol respectively to acetic acid and acetone with a method for the determination of these products. The authors desire to thank the Government Chemist for permission to publish this work. GOVERNMENT LABORATORY, w.c.2