STUDIES IN STEAM DISTILLATION-PART IV. 125 STUDIES IN STEAM DISTILLATION- PART IV. : PROPIONIC, BUTYRIC, VALERIC, AND CAPROIC ACIDS. BY if. DROOP RICHMOND, F.I.C. (Read at the Meeting, March 7, 1917.) THE sample of propionic acid worked with was obtained from Kahlbaum; as a preliminary, 1.5 c.c. were diluted to 100 C.C. and 10 C.C. of this distilled. This was diluted to 100 c.c., and callled fraction " A "; a further 70 C.C.were distilled and rejected, and the remaining 20 C.C. were washed out of the flask, made up to 100 c.c., and called fraction " B." A and B were each distilled in fractions of 10 C.C. as previously described (ANALYST 1908, 33, 305), with results as below. Note.-The tables in this paper are condensed, and only give x, y, and a. water ddstilled per cent. 100 acid distilled per cent.1TO y = l - x = l - X 04985 0.7985 0-6985 0.5985 0.498 0.3975 0.2965 0.198 0.096 First 10 C.C. Y 0-873 0.7495 0.636 0*522 0-415 0-316 0.222 0.1345 0.058 n 1.269 1.282 1.263 1 -268 1 *258 1 -249 1-239 1 *240 1.215 2 0.8995 0400 0.700 0.600 0-4995 0.3985 0.2975 0-1975 0 *O96 Last 20 C.C. Y 0.8795 0.764 0.6525 0.541 0.4365 0.3355 0-2405 0.149 0.068 a '1.214 1.216 1-198 1 -204 1.196 1-189 1-177 1.174 1.149 As the mean factor for A was 1.256 and B 1.193, it was evident that the acid was not pure, and it was estimated to contain about 2 per cent.of acetic acid and 2-6 per cent. of buty-ric acid. The whole of the acid was fractionated; in the fist two fractionations the portion boiling between 139.7" .C. and 141.7" C. was cpllected; in the third that boiling between 140-2" C.and 141.2" C.; and in the fourth and fifth the portion boiling between 140.5O C. and 140.9" C. was taken, but little being re- moved below the lower temperature, and no higher fraction than 140.9" C. being collected in the Hth fraction; 1.5 C.C. of this was diluted to 100 C.C. and distilled as in the previous experiment, with results as follows:'126 RICHMOND : STUDIES I N STEAM DISTILLATION-PART IV.- 5 0.900 04300 0.300 0.600 0.500 0.400 0.300 0.200 0.100 0.904 0.809 0.714 First 10 C.C. Y 0.8775 0.757 0.6425 0.530 0-425 0-324 0.2295 0,1395 0.063 a 1-24 1-25 1-24 1 *24 1-23 1.23 1 -22 1-22 1.20 I Last 20 C.C. 2 0.900 0.800 0.700 0.600 0.500 0 -400 0.300 0.200 0-100 Y 0.875 0.759 0.642 0-534 0-425 0.3255 0.230 0.1415 0.061 a 1.25 1-24 1-24 1-23 1-23 1.22 1-22 1-22 1.22 It is seen that the two series 0, results are practica,-y identical, and do not indicate any appreciable amount of impurity. As before, in calculating the ratio the first fraction has been neglected, and the other results were weighted from a con- sideration of the effect of 0.1 per cent.error in the percentage combined with a 0.05 per cent.error in the volume considered as a cumulative error; for propionic acid the relative weights of the fractions are : Fraction . . 2 3 4 5 6 7 8 9 Weight .. 2 5 6 6 6 5 4 2 Seven other series from N/5 to N/lOO were distilled, and the results need not he given in full, as they were of the same nature as those just given. The factor deducted from the experiments is expressed by the formula 1.224+ 0-0222 log.100 S. and the table below gives the results found and those calculated: S. 0*0084 0.0165 0.02 13 0.033 0.083 0.166 - Fokd. 1.221 1.234 1.225 1.238 1.243 1 -252 FACTOR. --.c----, Calculated. 1-222 1.229 1.231 1 -236 1 ~244 1.251 A mixture of acetic and propionic acids was made containing 0.488 molecules of propionic to 0.512 of acetic acid, and the mean results indicated 0493 molecules of propionic acid.1 1 2 P ! Ratio __ Y Calc. , A+ P' i j I I I 1 1 Calc. for [ CaIc. for Acetic. Propionic . 0.900 0-7985 0.698 0.5985 0-4985 0.3985 0.2995 0.0995 0.9025 0.807 0.7115 0.619 0-523 0.429 0.2375 0.132 0-931 043585 0.782 0.7045 1 0.6215 1 0-533 j 0.333 ' 0.206 0.8755 0.7565 0.641 0.5295 ~ 0.4215 I 0.320 1 0.1365 I 0.0575 0.513 0.505 0.500 0.488 0-492 0.488 0-486 0.498RICHMOND : STUDIES IN STEAM DISTILLATION-PAR,T IV I27 Considerable difficulty was experienced in obtaining butyric acid in a state of purity.Kahlbaum's acid. was found to be of only approximate purity, as the follow- ing results show: X 0.950 0.899 0.800 0.700 0.5985 0.499 0-3995 0.2995 0.1995 0.0975 Y 04398 04085 0.747 0.5045 0.373 0.268 0.1795 0*1085 0.056 0.0235 a 2.10 2.00 1.955 1 a92 1.925 1.895 1-875 1 *84 1.79 1-61 And another specimen which I had fractionated from a commercial butyric acid was used.After about ten fractionations, the portions boiling from 16045" C. to 16145" C. being collected in each case, oomparakively small fractions being rejected a t each fractionation, an acid was prepared, which gave the following results on distillation : X 0.902 0.803 0.704 0.591 0497 0.406 0.308 0.209 0.105 Y 0.644 0.4955 0.3525 0.252 0.1705 001025 0.050 0.01 1 Weighted mean = 1-99.a.8045 a 2.00 1-96 2.05 2.07 1.95 1a90 1.89 1 *98 - As this was not consid.ered satisfactory owing to the continued fall in the factor, this was dissolved in water, and one-third distilled. Of the distillate a further third was distilled, and of this, again, one-third, and this distillate was distilled in fractions.X 0-902 04302 0.701 0.600 0.5005 0 -400 0.300 0.200 0.098 Y 0.800 0.634 0-482 0.354 0.246 0.1595 0.090 0.0415 0-012 a 2.06 2-00 2.00 2-01 2.06 1.99 1.99 1.89 - Weighted mean = 2-01 The weights for butyric acid for fractions 2 to 9 are- 0.7 1 *o 1.0 0.9 0 *7 0.5 0.3 0.1 This indicated that an acid approaching purity wits being obtained, and a, con- siderable quantity of the acid was again distilled thus, and the first fractions con- verted into barium salts which were crystallised from alcohol.128 RICHMOND : STUDIES IN STEAM DISTILLATION-PART IV- Several series of results were obtained, all similar to that given above and leading to a mean factor of 2.00 for butyric acid; and it was noticed that the first fraction always led to a higher ratio, and a series distilled in an unjacketed flask with ample condensation gave a factor of 2.12.z 0.900 0.799 0899 0*507 0.497 0-395 0.295 0.199 0-094 jl 0,798 0-618 0.462 0.332 0.231 0.142 0.078 0.032 0-009 a 2.14 2.14 2-16 2-14 2-10 2-10 2.09 2.13 1 -99 Kahlbaum'a iso-valeric acid was fractionated, and the fraction for 175"-176" C.collected; this was refractionated, and two fractions 175"-175.7" C. and 176=7"-176" C. collected. These polarised respectively - 6-9" and - 7.0" in a 200 mm. tube, and it was evident that further fractionation would not affect any appreciable separation of the iso- and active acids. AS valeric acid is not miscible with water in all proportions, the solubility at 100" C.was determined on the first fraction by shaking 1 C.C. with 8 C.C. of water in a test-tube immersed in boiling water for fifteen minutes and allowing the undissolved acid to separate; 16184 grms. of the solution contained 0-1362 grm. of valeric acid estimated by titration with N/10 alkali = 9.0 per cent.; the vapour pressure is given by Kahlbaum as 49 mm.at loo", and these data allow of the calculation of the rate of volatility thus : 49 91 102 a=--- 760 9 18 - 8 figure which should agree approximately with that deduced for dilute solutions, but probably not agree exactly. One C.C. of each fraction was made up to 100 c.c., and 5 C.C. distilled from each, then a further 40 c.c., which was rejected, and the remaining 50 C.C.collected. The four fractions were then distilled in fractions of 10 C.C. each, with results as below: A Fraction 1 7 5 O - 1 7 5 - 7 O C. X 0401 0-801 0-7005 0.600 0.500 0-399 0.298 04200 0.098 Y 0.6'795 0 4 7 5 0.2805 0.1625 0*086 0.0415 0.015 0.003 First 5 C.C. a - 3-55 3-52 3-52 3-52 344 3.44 3.60 B. Fraction 175"-175*7" C. X 0.9005 0-7995 0.6995 0.598 0.498 0.398 0.2995 0.199 0.098 Y 0.688 0.456 0.287 0.168 0.093 0.0455 0.01 7 0.0065 0.001 Last 50 C.C.a - 3.38 3.43 3 *40 3-36 3-32 3.35 3 *08 2-94RICHMOND : STUDIES m STEAM DISTILLATION-PART IV. 129 C. Fraction 175*7"-176" C. X Y 0-900 0.6785 0.800 0.448 0.700 0.280 0.600 0.1595 0.500 0-0855 0.400 0.0405 0.300 0.016 0.1995 0-0035 0.099 - First 5 C.C. a - 3.55 3.53 3.58 3-52 3.48 3.41 3.50 - D.Fraction 175-7"-176" C. 2 Y 0.900 0.6845 04300 0-454 0.7005 0.2945 0.6005 0.1745 0.4995 0.0945 0.3995 0.050 0.2995 0*022 0.1995 0.0065 0.0995 - Last 50 C.C. 3-49 3.37 3.38 3.36 3.22 3.10 3.10 a - I It is seen that there is 110 appreciable difference between the two fractions boiling at 175"-175*7" C. and 175.7"-176" C., but there was evidence of a small amount of acid being present which was leas volatile in the last.50 C.C. ; the mean values of a deduced from the &st 5 C.C. (which. represented 17 per cent. of the total acid) were 3.51 and 3.515 for the two fractions, and for the last 50 C.C. (representing about 9 per cent.) were 3.37 and 3-35, The latter values diminished gradually as the distillation pro- ceeded, and it appeared from the results of the distillation th&t there was a small amount of an acid of considerably less volatility, which was reduced to an infinitesi- mal amount in the first 5 C.C.fractions; it appeared probable that the two valeric acids did not greatly differ in volatility. Twenty-five C.C. of the fraction boiling at 175-7"-176" C. were dissolved in 600 C.C. of water and sufficient soda solution to give a solution neutral to phenolphthalein, and 28.5 grms.silver nitrate dissolved in 400 C.C. of water added; both solutions being hot, and cooled oves-night. The precipitate was filtered over the pump and three times recrystallised from water, and a yield of 12.7 grms. of silver salts= 6.3 grms. of valeric acid (== 25 per cent.) was obtained.Four C.C. of nitric acid were added, and the valeric acid separated; this polarised - 0.37" in a 200 mm. tube. One grm.+3 C.C. N/10 strontia solution was made up to 100 c.c., and 5-2, 5.2-40 c.c., and 40-90 C.C. distilled; these three fractions were distilled in psrtions of 10 C.C. each. A. First 6.2 C.C. X Y a 0.9005 0,6865 - 04301 0.454 3.53 0*7015 0.289 3-46 0.6005 0.166 3-50 0.499 0-0905 3.43 0.400 0.044 3-39 0-299 0.0165 3.38 0.1995 0-0045 3-34 0.0995 - - B.5.2 c.c.-~O C.C. X Y a 0.899 0.682 - 0.800 0.453 3-50 0.700 0.286 3.47 0.6005 0.166 3-50 06005 0-089 3-47 0.401 0.415 3.47 0.301 0-0155 3-48 0.201 0.0035 3.52 0.0995 - - C. Last 40-90 C.C. X Y a 0.900 0,684 3.48 0.800 0.454 3-47 0.700 0.286 3.46 0.600 0.168 3-44 0.499 0.090 3.39 0.399 0.0435 3.30 0.299 0*01S 3.24 0.198 0.005 - 0.098 - - Finally, the mother liquors, which would .contain a large proportion of active acid, were, after adding a quantity of nitric acid nearly sufficient to liberate the130 RICHMOND : STUDIES IN STEAM DISTILLATION--PART IV.Valerie acid, distilled to half their volume, and a portion of the distillate distilled in fractions as below: X 0.900 0.8015 0.7005 0.6005 0.500 0-400 0.299 0.199 0.0985 Y 0.685 0.457 0.284 0.163 0.089 0.041 0.01 7 0.004 a - 3-495 3-52! 3.55 3.48 3.48 3.36 3-40 The weights for valeric acid for the 2nd to 8th fractions a r e l .0 , 1.0, 1.0, 0.7, 0.4, 0.2, 0.06. It is seen that the last 40-90 fract.ion gives a very slight indication of some less volatile acid ( 2 nitric), and that the others give a value of a of 3-49 for the iso- acid and 3.505 for the acid containing a large proportion of active acid; also there is no indication of a change with concentration.The value 3-50 may be adopted for both iso- and active valeric acids or a mixture of the two without appreciable error. A solution distilled in a flask without a steam-jacket gave a value for a of 3-69. 5 0.901 04301 0.702 0.602 0.501 0.401 0.301 0.201 0.101 Y 0-676 0.438 0.268 0.153 0.079 0.035 0.014 0.005 0.002 a 3 -72 3-12 3-10 3-65 3-66 3-52 - Normal caproic acid from Kahlbaum was fractionally distiIled ; boiling com- menced at 190" C., and a few drops of milky liquid passed over before 204" C., to which the temperature rapidly rose, and fractions from 204-5' to 205.5" C., and 205-5" to 206-5" C.were collected, there being a fairly large fraction up to 210" C. The fraction 204.5" to 205.5" C. was refractionated, and the portion below 2045" (3. rejected (a slight milky drop coming over first) ; a fraction 204.5" t o 205.5" C. was collected, and the fraction 205*5" to 206-5" C. then added, and a small amount more distilled at 204.5" to 205.5" C.; there was an appreciable fraction above 205.5" C., and this was slightly dark.The middle fraction was refraetionated, and the portion 20445" to 205.5" C. collected; it was judged that the lower boiling fraction (quite small) contained traces of water, and thatl there was no serioura amount of sub- stance boiling below 204.5" C., but there appeared to be an appreciable fraction above 205.5" C., which did not appear to diminish in the last fraction.One C.C. of this fraction was shaken with 500 C.C. of water, and all dissolved and 150 C.C. of this were distilled in fractions. The value of a fell steadily from 3.6 to 1.7, and it was evident that there was some acid present which was less volatile.MCHMOND : STUDIES IN STEAM DISTILLATION-PAET IV. 131 To remove this, 50 C.C.were distilled from another 100 c.c., made up to 100 c.c., and distilled in fractions, giving a f&ly concordant set of values for a rather over 3.5 but falling away to 3.0 at the end. It was thought from the results of the fractionation and the distillation of the solution that there might be a formation of ltktones; 0-2215 grm. took 18-85 C.C. N/10 alkali for direct titration, and 19.24 C.C.after boiling with excess of alkali and titrating back. The first result indicates a molecular weight of 117.3, and the latter of 115-0; and as it was probable that the acid contained it little water, it appears that the sample was nearly pure caproic acid, mixed with its lactone. A quantity of N/lO sulphuric acid slightly in excess of the alkali was added, and 50 per cent.was distilled, made up to 100 c.c., and distilled in fractions. 0 0.950 04351 0.751 0.651 0.550 0.450 0.248 Y 0.827 0 546 0.340 0-2045 0-1 10 0.055 0-007 a 3.70 3.75 3.77 3.70 3-69 3.64 3.56 Another portion of the acid was boiled with excess of soda, an equivalent amount of sulphuric acid added, the volume made up to 100 c.c., and fratgons of 11 c.c., 15 c.c., and 50 C.C. distilled; each of these, which contained approximately one-third of the acid, were made up to 100 C.C.and distilled in fractions, giving mean results for a of 3.71, 3.73, 3.70, agreeing within the limits of experimental error. Prom the results of several distillations the mean value of a of 3-71 was deduced, a value that was much lower than expected. Diethyl acetic acid, for which I am indebted to Mr.J. Marshall and which was made in the Sir Jesse Boot's Research Laboratories, was prepared by heating di-ethyl-malonic acid to 190" C. - The following table gives the results of the distillation : 5 0.902 0.8035 0.7035 0.6005 0.4995 0.3975 0.2975 0.195 0-100 Y 0.629 0.362 0.194 0.093 0.041 0-015 0.003 0-001 a - 4.76 4-78 4-70 4.65 4-76 4.75 The mean figure (4.75) was confirmed by another experiment.Taking the results for 0.01 N solution, the factors agree fairly well with those calculated by the formula: a = 0-4013 x 1*72", where rt, is the value iu the formula C,H,+ ,CQOH132 RICHMOND : STUDIES IN STEAM DISTILLATION-PART IV. a n O.U)13 x 1-12’ Formic .. .. 0.4013 0 0.4013 Acetic * . .. 0,677 1 0.69 Propionic . ... 1 a224 2 1.18 Valeric .. .. 3-50 4 3.51 Butyrk .. .. 2-00 3 2-04 The results for the caproic acids do not agree with the formula, and a study of the results obtained with the other isomers will form the subject of a later com- municat ion. D. C. Dyer ( J . Bid. Chem., 1917,=, 445, see p. 149) has lately published a steam distillation method in which the volume is kept constant at 150 c.c., and under these circumstances, as pointed out by me (ANALYST, 1908, %,209), the curve is a simple logarithmic one; the results give a straight line if plotted out on a logarith- mic chart for a single acid, but a curved one for a mixture, and the probable natures and proportions of the acids may be approximately read from the curves and tables given.Boekhout and de Vries (Centr. Bakt. Par., 1916,1l, 46, 505) have lately studied Duclaux’s method, and find that if the differences between each fraction be calculated as percentages of the acid present a t the commencement of the distillation, and these expressed as a ratio of the amount distilled to 110 C.C. (the original volume adopted by Duclaux), constant figures are obtained for each fraction, being- Formic acid .. . . 3.5 Acetic acid .. . . 5-9 Propionic acid . . . . 11-8 Butyric acid . . . . 19.1 Valeric acid .. . . 26.0 These figures, if multiplied by 0.11, should give the same values as I have found for a, and, indeed, show an approximate agreement. This mode of calculation is only a crude method of differentiation, and better results are obtained by the method that I have used. The bulk of this work wm done in the laboratories of ths Aylesbury Dairy Company, and it was completed in the laboratories of Messrs. Boots’ Pure Drug Co., Ltd., to both of whom I wish t o express my thanks, and I have to acknowledge the rtssiatance of Messrs. W. B. Walker and S. Darnell.