180
INDUSTRIAL ASD ENGINEERIKG CHERIISTRY
product resulted which could be removed only by the use of considerable amounts of hydrogen peroxide and boiling. Whereas the use of alcohol saturated with potassium chloroplatinate is rather general, this is a particularly necessary precaution with small amounts of precipitate. When alcohol of any strength was used without this preliminary saturation, low results invariably resulted. Eighty per cent alcohol, saturated with potassium chloroplatinate, was used because of the greater solubility of sodium salts in this strength than in stronger alcohol solutions (14). It should also be noted that any aldehyde or acetic acid, which are common contaminants of alcohol, may produce some reduction of the chloroplatinate, giving low results (IS). It is apparent that this procedure will be satisfactory in any case in which the gravimetric chloroplatinate method may be used. This volumetric modification is rapid, simple in operation, and appears to yield on small samples an accuracy which is superior to that which would be obtained gravimetrically, since the deviations are not greater than would be expected from ordinary weighing deviations on the microbalance. Further work is in progrees on the application of these methods to biological fluids.
VOL. 7 , KO. 3
Literature Cited Emich, F., and Schneider, F., “Microchemical Laboratory Manual,” p. 69, New York, John Wiley & Sons, 1932. Fajans, K., and Hassel, O., 2. Elektrochena., 29, 495 (1923). Fajans, K., and Wolff, H., 2. anorg. allgem. Chem., 137, 221 (1924). Hubbard, R. S., J. Biol. Chem., 100, 657 (1933). Kirk, P. L., IND.ENU.CHEY.,Anal. Ed., 7, 195 (1935). Kirk, P. L., Mikrochem., 14, 1 (1933). Kirk, P. L., Craig, R., and Rosenfels, R. S.,IND.EXQ.CHEN., Anal. Ed., 6, 154 (1934). Kirk, P. L., and Schmidt, C. L. A,, J . Biol. Chem., 83, 311 (1929). Kolthoff, I. M., Lauer, W. M., and Sunde, C. J., J. Am. Chem. Soc., 51, 3273 (1930). Maw, C. B., and Miller, K. R., Proc. Utah Acad. Sci., 8, 6L (1931). Miller, R. P., and Kirk, P. L., Mikrochem., 14,15 (1933). Schueler, J. E., and Thomas, R. P., IND.ENQ. CHEM.,Anal. Ed., 5, 163 (1933). Strebinger, H., and Holzer, H., 2. anal. Chm., 90, 81 (1932); Makris, C. G., J.Pharm. Chim., 13,569,570 (1931). Treadwell, F. P.,and Hall, ’cv. R., “Analytical Chemistry,” 6th ed. p. 59, New York, John Wiley & Sons, 1924. Van Rysselberge, P. J., IND.ENQ.CHEM.,Bnal. Ed., 3, 3 (1931). RECBIVED February 4, 1936. Aided by a grant from the Research Board of t h e University of California.
Analysis of n-Butanol, Acetone, and Ethanol in Aqueous Solution LEO M. CHRISTENSEN and ELLIS I. FULMER, Iowa State College, Ames, Iowa
I
K’ T H E butyl-acetonic fernientation of corn mash and
other suitable substrates, n-butanol, acetone, and ethanol are produced in the ratio of about 6 to 3 to 1. Hydrogen, carbon dioxide, acetic acid, butyric acid, and acetylmethylcarbinol are also formed, but the yield and proportion of the three solvents are of particular interest. It becomes necessary, therefore, in laboratory research on this process, to have available a rapid routine method for measuring these products. The method should also be applicable when only small amounts of the materials are available. Reilly and Ralph (6) have reported the densities of the aqueous solutions of t)hethree solvents. Bogin ( I ) developed a quantitative method based upon the determination of the total concentration of the three in the aqueous distillate, the measurement of the acetone by the usual iodometric method, and the determination of the n-butanol-ethanol ratio by titration of the anhydrous alcohols with water. Werkman and Osburn (7) developed a procedure for the analysis of solutions of n-butanol and ethanol, based upon their oxidation with potassium dichromate and phosphoric acid, the acetone having been previously removed with 2,4dinitrophenylhydrazine. The butanol yielded 90.7 per cent butyric acid and 9.3 per cent acetic acid, while the ethanol yielded only acetic acid. The acid ratio was determined by partitioning with isopropyl ether, Johnson (5) used a similar method, except that he measured the acid ratio by a modified Duclaux procedure. Fang (2) developed a method based upon measurement of the refractive index, determination of acetone by the usual iodometric method, and a second measurement of refractive index after partition with carbon tetrachloride, which removed n-butanol to a considerably greater extent than it did the acetone or ethanol. Several methods have been used in these laboratories for the analysis of the two alcohols and, as a result, procedures
have been developed for their determination which have met all requirements for speed, simplicity, and accuracy. They have been thoroughly checked by a number of individuals and found to be well suited for routine use. The fermentation mixture, representing a known weight of substrate, is neutralized to pH 7.0 to 7.5, and about one-third is carefully distilled into a receiver submerged in cold water. Usually 100 cc. of distillate are thus collected. All calculations are then based upon this volume of distillate. If it is of interest to estimate only the total solvent yield, the specific gravity, 25”/25’, is measured and the solvent concentration calculated by means of Equation 1. This equation holds for a 6 to 3 to 1 ratio, but is inaccurate when the ratio varies appreciably from the above proportion. The total concentration should not exceed about 10 grams per 100 cc. and is usually kept below 5 grams per 100 cc . Total solvents, grams/100 CC. = (1.0000 - sp. gr. 25”/25’) 698 (1) The dipping refractometer may also be used to measure the total solvent concentration. The reading, 25’/25’, is expressed by Equation 2, in which R is the dipping refractonieter reading and B , A , and E are the concentration of nbutanol, acetone, and ethanol, respectively, in grams per 100 cc. With a ratio of 6 to 3 to 1, the total solvent concentration is given by Equation 3.
+
R = 13.25 2.646B 4-1.800A 4- 1.610B Total solvents, grams/100 CC. = 0.437 ( R - 13.25)
(2) (3)
When it is desired to separate and identify the products, sufficient distillate to yield about 200 cc. of anhydrous solvents is saturated with sodium chloride, about one-fourth distilled into a cold receiver, and saturated with anhydrous potassium carbonate. The solvents separating are further
ANALYTICAL EDITION
MAY 15, 1935
The extraction of butanol is not a linear function of concentration, and the correction represents the deviation from the h e a r relation assumed in ap lying the constant, 2.75. The terms have the dimensions usei above. Instead of extracting with carbon tetrachloride, a second equation may be obtained by carrying out the oxidation under different conditions. Ten cubic centimeters of original distillate, containing not more than 2 grams of alcohols per 100 cc., are diluted to 100 cc., and 5 cc. of this dilution are added to a cold mixture of 25 cc. of concentrated sulfuric acid and 10 cc. of 0.4 N potassium dichromate in a 2.5 by 25 cm. test tube and the procedure for M1 is followed. A blank is run in the same way. The difference in titration is multiplied by two to give the value, N , shown in Equation 6, in which the dimensions are as above.
dried with anhydrous potassium carbonate and carefully fractionated. The usual derivatives can then be prepared. If no other compounds than n-butanol, acetone, and ethanol are pnesent, and this is ordinarily the case, the procedures described below may be applied. The data presented were obtained by the use of aqueous systems of the highly purified solvents. The acetone is determined by the Goodwin (4)modification of the Messenger method. The interference from ethanol is negligible, unless it be present in much larger amounts than is the acetone. One gram of acetone consumes as much iodine as 100 grams of ethanol under the conditions employed in the analysis.
N
+ 0.691A + 8.78 E
= 2.75 B
calcu-
(7) (8)
(9)
Solvents in Original BuOH Found, EtOH Found, Distillate, G./lOO Cc. G./lOO Cc. G./lOO Cc. BuOH MeaCO EtOH MI Mz Calc. Error Calc. Error By Oxidation-Extraotion Method 3.81 0.87 0.88 34.0 18.0 0.81 -0.07 3.81 10.00 3.52 0.87 1.07 34.3 19.8 -0.02 1.06 -0.01 3.50 3.52 1.72 0.29 26.7 12.1 0.22 -0.07 3.53 $0.01 0.22 -0.07 4.32 1.08 0.29 32.4 14.0 4.27 -0.05 0.59 0.21 0.10 10.00 0.08 -0.02 4.9 2.7 0.59 0.18 -0.01 1.17 0.43 0.19 9.9 5.4 1.18 $0.01 8.1 -0.05 0.27 -0.02 1.76 0.64 0.29 15.0 1.71 2.35 0.86 0.39 19.7 10.6 2.33 -0.02 0.32 -0.07 2.93 1.07 0.48 25.0 13.2 $0.03 0.46 -0.02 2.96 1.07 10.00 3.09 1.29 1.07 31.7 18.9 3.11 + 0 . 0 2 4.32 1.29 0.00 30.1 11.6 -0.06 4.29 -0.03 -0.06 1.15 -0.02 -0.02 0.00 1.72 1.17 11.2 10.2 -0.02 B y Double-Oxidation Method 1.50 0.75 0.30 13.4 52.2 1.49 -0.01 0.29 -0.01 0.81 $0.01 0.37 0.80 1.00 0.40 9.5 40.6 -0.03 0.37 0.40 1.00 0.40 0.42 +o.oz -0.03 6.7 30.9 1.50 0.90 0.30 13.5 54.7 1.48 -0.02 -0.01 0.29 1.48 -0.02 1.60 0.60 0.30 13.4 49.6 0.30 *o .00 0.99 -0.01 1.00 0.50 1.00 13.2 39.0 0.99 -0.01 1.00 10.00 0.49 1.00 0.50 0.50 11.6 37.6 -0.01 1.02 1 0 . 0 2 0.39 -0.01 1.00 0.50 0.40 10.7 37.0 1.00 f O 00 1.00 1.00 1.00 16.2 50.7 0.97 -0.03 0.22 -0.02 0.24 0.11 0.04 7.8 2.1 0.04 10.00 1.42 +O.Ol 0.41 0.15 0.00 10.2 37.6 0.05 10.05 1.25 0.00 0.30 11.6 33.9 1.27 $0.02 0.33 +0.03
.
(4)
Summary Methods suitable for the rapid routine analysis of aqueous solutions of n-butanol, acetone, and ethanol are described. Acetone is measured by a n iodometric method and the al-
TABLE11. STANDARD CORRECTIONB FOR EQUATION 4 0.1
+o.oz
0.2 +0.04
+0.21
+0.22
$0.32 1-0.30 +o. 10 -0.16 -0.42
4-0.32
+o.zs
$0.07 -0.18 -0.44
0.3 +0.06 $0.24 $0.33 +0.27 +0.05 -0.21 -0.46
0.4 +0.08
$0.25 +0.33 +0.25 +0.02
-0.23 -0.49
of the distillate are placed in a test tube with 40 CC. of carbon tetrachloride, shaken, and allowed t o stand 2 hours at 25" C. It is essential that the temperature be closely controlled. Ten cubic centimetersof the aqueous solution are removed, diluted t o 100 cc. and 10 cc. of this dilution oxidized as above, to obtain n value, Mz, which is represented by Equation 5. M s
(6)
TABLEI. ESTIMATION OF BUTYLAND ETHYLALCOHOLIN MIXTURESWITH ACETONE.
A second titration is made to obtain another equation, and one of two procedures may be used. In one, a second oxidation, exactly like the above, is applied to a sample of distillate extracted with carbon tetrachloride, Twenty cubic centimeters
BuO" Uncorrected G./100Cc. 0 0 0 1 +o. 19 +0.31 2 3 +0.31 $0.12 4 -0.14 5 -0.39 6
+ 17.68A + 8.84E
I n Table I are given data showing the application of the above equations to various mixtures of the pure solvents. The use of these relations is not confined to mixtures, for they are equally applicable to the solvents alone.
Ten (cubicCentimeters of original distillate containing not more than 5 grams of alcohols are diluted to 100 cc. and 10 cc. of this dilution are added t o a cold mixture of 10 cc. of 0.4 N potassium dichromate and 10 cc. of concentrated sulfuric acid in a 2.5 by 25 cm. test tube. Two glass rods are dropped into the tube and the contents thoroughly mixed. A stopper carrying a 1-mm. capillary tube, the lower end of which is bent at right angles, is inserted and the tube placed in a vigorously boilin water hath. After 10 minutes' exposure, it is cooled and dilutes to 400 cc. in a 1-liter Erlenmeyer flask. Fifteen cubic centimeters of a 20 per cent potassium iodide solution are then added, the flafk is stoppered and allowed to stand 2 minutes, and the iodine released is titrated with 0.1 N sodium thiosulfate. A blank is run in the same way. The difference is designated as MI and has a value shown by Equation 4, in which M I is in cubic centimeters of 0.10o0 N dichromate consumed, B is equal t o the weight, in grams, of butanol in the original distillate, A , the acetone, and E, the ethanol, on the same basis. 6.92 B
24.62 B
+
When a mixture of n-butanol, acetone, and ethanol is oxidized with potassium dichromate and sulfuric acid, the reducing value of the solution depends upon the concentration of sulfuric acid, time, and temperature, as shown by Fauser (3). He found that 10 minutes' exposure of a 30- to 40-cc. mixture in a 2.5 X 25 om. test tube in a boiling water bath gave easily duplicated results if the sulfuric acid concentration was carefully adjusted. Two flat portions were observed in the curve when reducing value was plotted against acid concentration, one when 10 volumes of acid per 30 cc. of mixture were used, and the second when 40 cc. of mixture contained 25 cc. of concentrated sulfuric acid. The difference between the reducing values obtained under these different conditions was found to be due entirely to the n-butanol and acetonle. The procedure for the determination of the alcohols is as follows:
-
=
From these values, ethanol and n-butanol may be lated by means of Equations 7,8, and 9. B = 0.25 (Mi - M 2 ) - 0.01 M2 - 0.07 A correction (see Table 11) B = 0.057 ( N - Mi) - 0.96 A E =a 0.114 Mi - 0.788 B - 0.0787 A
Determination of Ethanol and n-Butanol
MI
181
+ 0.386A + 8.22 E plus correction
(5)
0.5 +0.10 +0.26
+0.33 +0.23 kO.00 -0.26 -0.52
0.6
0.7
0.8
+o. 12
+O. 14
+O. 16
-0.28
-0.06 -0.31 -0.57
+0.27 +0.33 f0.21 -0.03 -0.54
$0.28 +0.33 +o. 19
+0.29 +0.32 $0.17 -0.09 -0.33 -0;59
0.9 +0.18 +0.30 +0.31 $0.14 -0.11 -0.36 -0.61
cohols are measured by oxidation with dichromate and extraction with carbon tetrachloride, or by oxidation at two different concentrations of sulfuric acid in the oxidizing mixture. The methods are equally accurate and rapid. For purposes of checking, a combination of the methods may be employed. As a means of proving the identity of the prod-
INDUSTRIAL AND ENGINEERING CHEMISTRY
182
ucts, both the oxidation and the oxidation-extraction methods, taken in conjunction with measurement of specific gravity and the dipping refractometer reading, will prove of value.
Acknowledgment Many individuals contributed to the development of these methods. In particular, thanks are due Donald Starr, Gerrish Severson, Ernest Fauser, L. C. Bryner, and A. R. Kendall.
(1) (2) (3) (4) 15) (6)
VOL. 7, NO. 3
Literature Cited Bogin, C. D., ISD.Em. GRIM., 16, 380 (1924). Fang, H. C., Iowa State Coll. J. Sci., 6, 423 (1932). Fauser, E. E., thesis for M.Sc. Degree, Iowa State College, 1934. Goodwin, L. F., J. Am. Chem. Soc., 42, 39 (1920). Johnson. M. J.. IXD.ENQ.CHEW..Anal. Ed.. 4.20 (1932). Reilly, Js,and' Ralph, E. W., Sci. Proc. Royal Dublin Soo., 15,
597 (1919). (7) Werkman, C. H., and Osburn, 0. L., IND.ENG.CHIUM., Anal. Ed., 3, 387 (1931). RECEIVED January 25, 1935.
Estimation of Small Amounts of Lead in Copper BARTHOLOW PARK and E. J. LEWIS, Michigan College: of Mining and Technology, Houghton, Mich.
L
EAD occurs in most refined copper in about the same
amount as other metallic impurities, from a few to several hundred parts per million. Several methods have been proposed for its separation and estimation. Heath (1) and Jones (8) separate lead as dioxide by electrolyzing a nitric acid solution of the copper. Keffer (3) separates the lead by co-precipitation with calcium carbonate from a n ammoniacal solution before electrolysis. Electrolytic methods for lead are open to criticism because some of the lead is apt to deposit with the copper; other elements, such as bismuth and manganese, may deposit with the lead; the composition of the deposited lead peroxide varies; and prevention of mechanical loss while handling the deposit is difficult. In order to find a better method for the separation of lead from copper the authors have tried a number of schemes, all based on the formation of an insoluble lead salt and its collection by some occluding agent. The method adopted proved capable of separating and estimating, with an error of about 50 per cent, as little as one part of lead in one million parts of copper.
Purification of Reagents Nitric, hydrochloric, and acetic acids and ammonium hydroxide were tested for lead by distilling 500 cc. of each from a silica st ill, dissolving the residue in distilled nitric acid, concentrating to 5 cc., and then testing the concentrate spectrographically. These reagents were Iead-free. Solutions of manganese sulfate (25 per cent), ferrous ammonium sulfate (20 per cent), calcium chloride (20 per cent), and disodium hydrogen phosphate (50 per cent) were rendered leadfree by adding a few cubic centimeters of a saturated solution of lead-free copper sulfate, saturating with hydrogen sulfide, filtering off the precipitated sulfides, and repeating until a precipitate was obtained which showed no spectrographic trace of lead. Usually the third precipitate was lead-free. Ten er cent potassium dichromate solution was treated with a little oPthe lead-free ferrous ammonium sulfate solution, ammonium hydroxide was added and the mixed precipitate of iron and chromium hydroxides was filtered out and tested. This procedure was repeated until the precipitate gave no test for lead. Five er cent otassium permanganate solution was treated with a fttle of tEe lead-free dichromate solution and then repeatedly with ferrous ammonium sulfate and ammonium hydroxide until a lead-free precipitate was obtained. The authors were unable to remove the rather large amount of lead present in the c. P. reagent brand of ammonium carbonate. Lead-free ammonium carbonate solution was prepared by passing carbon dioxide into a solution of two parts of water to one part of strong ammonium hydroxide until a precipitate formed. Table I gives a summary of the various schemes of separation which were tried. In each case 50 grams of copper con-
taining about 0.5 mg. of lead were washed in hydrochloric acid to remove surface impurities and dissolved in 200 cc. of nitric acid. From this point the procedure varied.
Procedure 1. The sample was taken to fumes with sulfuric acid, diluted to 500 cc., and heated to boiling and then 2 CC. of manganese sulfate solution were added. Ten cubic centimeters of per-
manganate solution were added with constant stirring and the solution was allowed to stand overnight. The precipitate was dissolved and reprecipitated as PbS, which was then dissolved in nitric acid and tested spectroscopically for lead. 2. Ammonium hydroxide was added to neutralize the excess nitric acid, the solution was diluted to 500 cc., and 15 cc. of acetic acid were added. The pH of the solution at this point, determined by means of a quinhydrone electrode, was 2.0. Ten cubic centimeters of the dichromate solution were added and then the manganese dioxide recipitation was carried out as described under (1). Each of tiree consecutive precipitates showed a small amount of lead. 3. The excess nitric acid was boiled off, 15 grams of borax were added, and the solution was diluted to 500 cc. The pH was 2.8. Three manganese dioxide precipitates, after the addition of dichromate solution, failed to remove all of the lead. 4. The solution was prepared as in (3), but 25 cc. of ferrous ammonium sulfate were added instead of the manganese sulfate and permanganate. Three consecutive precipitations all showed lead. 5. Four grams of Norite were added to a solution prepared as in (3) and the mixture was stirred mechanically overnight. The Norite was filtered out and extracted with hot nitric acid. This procedure failed to remove any lead from the solution. 6. The nitric acid solution was diluted to 500 cc., and then 10 cc. of dichromate solution, 25 cc. of ferrous ammonium sulfate solution, and enough ammonium hydroxide to bring down a small amount of copper hydroxide were added. Next 200 cc. of concentrated ammonium hydroxide were added and the solution was heated to 80" C. and filtered hot. Three consecutive precipitates all showed a small amount of lead. 7. This procedure was the same as (6), except that 3 cc. of manganese sulfate and 10 cc. of permanganate solutions were substituted for the iron. A complete separation was not obtained. 8. To the hot ammoniacal solution containing dichromate 4 grams of Norite were added. After a few hours of stirring the Norite was filtered out and extracted with nitric acid. No lead was removed by this method. 9. T o the cold ammoniacal solution were added 50 cc. of ammonium carbonate solution, followed by 5 cc. of calcium chloride solution. It was found that if the calcium chloride was added before the ammonium carbonate (method of Keffer, 3) a much less efficient separation was made. The mixture was stirred well and allowed to stand overnight, and the precipitate filtered out and tested in the usual way. Although this procedure gave a better separation than any of those previously mentioned, it was unsatisfactory. Three consecutive precipitates all showed lead.
