Vol. 15, No. 2
INDUSTRIAL AND ENGINEERING CHEMISTRY
Bertram, S. H., Chern. Weekblad, 24, 226-9 (1927). Gardner, H. A., "Physical and Chemical Examination of Paints, Varnishes, Lacquers, and Colors", Washington, Institute of Paint and Varnish Research, 1939. Gay, P. J., J . SOC.Chem. Ind., 51, 126-9T (1932). Griffiths, H . N , and Hilditch, T. P., Ibid., 53, 75-81T (1934). Hilditch, T. P.. and Murti, K. S., Analyst, 65, 437-46 (1940). Jamieson,G. S.."Vegetable Fats and oils",New York, Chemical Catalog Co., 1932. Kass, J. P., Loeb, H. G., Norris, F. A., and Burr, G. O., Oil &
(18) Ibid.. 18. B. 231-9 (1940). (19) Matthews, N. L., Brode, W.R., and Brown, J. B., Oil & Soap, 18, 182-7 (1941). (20) Noonan, E., IND.ENG.CHEX.,AN.AL. ED.,10, 34 (1938). (21) Korris, F . A., Kasa, J. P., and Burr, G. O., Oil &. Soap, 17,
SOUP,17, 118-19 (1940). Kass, J. P., Lundberg, W. O., and Burr, G. O., Ibid., 17, 50-3
(24) Rollett, A., 2.physiol. Chem., 62, 422-31 (1909). (25) Rose, W. G., and Jamieson, G. S., Oil & S o a p , 18, 173-4 (1941). and Sando, C. E., J . (26) Whneler, D. H., Riemenschneider, R. W., B i d . Chem.. 132, 657-99 (1940).
(1940).
Kaufmann, H. P., Z. Untersuch. Lebensm., 51, 15-27 (1926). Kaufmann, H. P., and Baltes, J., Ber., 70B, 2545-9 (1937). Kaufmann, H. P., and Keller, bl., Z. angew. Chem., 42, 73-6 (1929).
Kaufmann, H P., and Mestern, H. E., Ber., 69B,2684-5 (1936). Kimura, W.,Fettchem. Umschau, 42, 78-80 (1935). Kimura, W.,J . Soc. Chem. I d . J a p a n , 32, 138-41B (1929). McCutcheon, J. W., C a n . J . Research, 16, B , 158-75 (1938).
123 (1940).
(22)
Riemenschneider, R. W., Swift, C. E., and Sando, C. E., Ibid.,
(23)
Riemenschneider, R. TT'., and Wheeler, D. H., Ibid.,
18. 203-6 (1941). -,. ~~
~
~
I
~
16, 219-21
(1939).
CoNTRIsurIoN from the Department of Agricultural Chemistry, North Dakota Agricultural Experiment Station. Progress report on Purnell Project No. 95 entitled "Tho Chemistry of Flaxseed". Published by the permission of the Director of the North Dakota Agricultural Experiment Station.
Cryoscopic Analysis of Styrene, Indene, and Dicyclopentadiene E. H. SMOKER AND P. E. BURCHFIELD Research Department, The United Gas Improvement Co., Philadelphia, Penna.
S
MALL quantities of certain impurities in monomeric
styrene, indene, and dicyclopentadiene markedly affect the properties of polymers produced therefrom. Limitations of analytical methods for the detection of small quantities of impurities based on bromine absorption or on the accurate measurement of several physical properties are well known to investigators in this field. A method for determining the amount of impurity in hydrocarbons from freezing and melting curves is described by Mair, Glasgom, and Rossini (1) but is not designed for control work where speed is essential. It was believed t h a t an analytical procedure based on the depression of the freezing point of the hydrocarbon mould provide a precise and sensitive method, dependent on the relative number of molecules of hydrocarbon impurities present and practically independent of the type of impurities. The development of the procedure entailed: (1) the preparation of "100 per cent" styrene, indene, and dicyclopentadiene, respectively, and (2) the design of a suitable apparatus for measuring freezing points, and the determination of the molal dispersions, Kj, of the compounds stated in (1).
Purification of Light Oil Hydrocarbons The light oil hydrocarbons mere purified by successive recrystallizations. The liquid phase was separated from the solid b y means of a centrifuge. The centrifuge cakes were not washed. Advantage of equilibrium melting was taken during centrifuging. The criterion of purity consisted of identity of freezing points of the solid and liquid phases of the final crystallization. The hydrocarbons thus processed were assumed to be pure. The purified compounds had the following physical properties. Compound Styrene Indene Dioyclopentadiene
vreesing point, (Cor.) -30.60 1.50 33.6
-
0
c,
Refractive Index, ;'n 1.5469 1.5764
....
I
Measurement of Freezing Points and Molal Depressions STYRENE.The cell in which the freezing points were determined was a glass tube 38 X 1.6 em. (inside diameter) surrounded by a tube 39 X 2.5 em. (inside diameter) nhichprovidedan air bath. A specially constructed mercury thermometer, 37.5 cm. long, graduated directly to 0.01"C. in the range of -30" to -38" C. was used to measure temperatures. The thermometer was calibrated by the Kational Bureau of Standards. A 15-cc. sample, which immersed the thermometer to the -37" mark, was found to be sufficiently large. The freezing cell and air bath !yere immersed to the -35' mark on the thermometer scale in a carbon dioxide-acetone bath contained in an unsilvered quart Dewar vessel. The bath was maintained at a temperature approximately 5" C. lower than the freezing point of the solution under examination. The solution, the freezing point of which was being determined, was stirred mechanically with a 12-gage Nichrome wire, provided with two loops surrounding the thermometer below the solution level. The highest temperature observed after crystallization began was taken as the freezing points. Solutions of varying composition were prepared from purified styrene and purified rn-xylene. The freezing points of the solutions were determined and the molal depression constant was calculated for each AT between consecutive solutions. Two observers independently checked each freezing point with the deviation shown, along with the complete data, in Table I. INDENE AXD DICYCLOPENTADIENE. Purified indene and purified dicyclopentadiene, respectively, nere diluted with p TABLEI. FREEZISG DATAOF STYRENE-XYLENE SOLUTIONS Styrene Mole % 100 99.77 99.58 99.22 99.0s 98.15 96.61 95.30
Freezing Point O C . (cor.) -30.60 -30.69 -30.78 -30.95 -31.01 -31.44 -32.15 -32.76
Supercooled OC.
0.5 0.6 0.5 0.4 0.6 0.4 0.6 0.7
.Maximum Deviation between Observers
c.
Kf
0.00 0.00 0.00 0.00 0.00 0.00
4:i
5.3 4.9 4.9 5.1 4.9 5.0
0.01 0.01
Av.
4.95
February 15, 1943
ANALYTICAL EDITION
TABLE 11. FREEZIXG DAT~ Indene M o l e 7o
Freezing Point OC.
Kf
Supercooled
(COY.)
O C .
Indene-p-Xylene Solutions 100
-1.50
0.3 0.4 0.3
-1.91 -2.20 -2.46 -2.78
99.34 98.85 98.47 97.93
Dicyzlopcntadiene M u l e 70
Av.
6:86 6.90 6.88 6.88 6.89
Av.
50.7
0'.'4
Freezing Point
C. Dicyclopentadiene-p-Xylene Solutions
129
carbon examined. Compounds higher in molecular weight than xylene produce a proportionally lower freezing point depression per unit weight, and impurities of lower molecular \\,eight produce a depression proportionally higher. High molecular weight polymers in small amounts n-ould not be detected b y this method. However, impurities of boiling points diverging appreciably from the boiling point of the light oil hydrocarbons in question can be separated by fractional distillation. Chemically reactive compounds, such as organic acids which produce a n abnormally high K,, should he determined and removed b y chemical means. The cryoscopic procedure i s used in research and control work only between concentrations of 92 t o 100 per cent. The true freezing points of t h e solutions of purified compounds and xylene are slightly higher than those herein reported, for reasons discussed by &lair, Glasgow, and Rossini (1).
xylene (freezing point 13.3" C.) and the freezing points were measured by means of a thermometer, graduated directly to 0.01" C., calibrated by the Kational Bureau of Standards. The freezing point technique was in general idrntical with that ,described under styrene. Results are summarized in Table 11.
The high K, and convenient freezing point of dicyclopentadiene suggest a possible use in the determination of molecular weights as a substitute for camphor. The reactivity of the compound, especially with respect t o oxidation, would necessitate some precautions not required in the caie of camphor .
Discussion
Literature Cited
The data presented indicate a precise analytical method for the detection of small q u a n h i e s of impurities in the hydro-
(1) Mair, Glasgow, and Rossini, J . Research Natl R w . Stnndnids. 26, 591-620 (1941)
Determination of Sulfur Dioxide in Beer A Modification of Monier-Williams Method B. H. NISSEN AND R. B. PETERSEN Anheuser-Busch, Inc., St. Louis, Mo.
U R I N G the determination of sulfur dioxide in beer b y the Monier-Killiamsmethod ( I ) , i t was noticed that aprecipitate of barium sulfate was formed in the oxidizing medium (3 per cent hydrogen peroxide). This was found to be due to a n excess of barium hydroxide added in removing the sulfate ion from the peroxide b y the regular procedure. The presence of this precipitate indicated that any sulfur dioxide determinations run using this hydrogen peroxide were in error by the amount of sulfate precipitated, as sulfate in this form is not titratable. An attempt was made to improve the accuracy of the procedure as well as increase the rapidity with which the hydrogen peroxide could be prepared b y eliminating the effect of the sulfate b y another method not involving the use of barium hydroxide, as it is difficult to avoid adding a n excess of this reagent. This was accomplished b y neutralizing the acid in the peroxide with 0.1 N sodium hydroxide to p H 4.0, using a glass electrode (the neutralization could be made to the bromophenol blue end point). p H 4.0 is the neutral point in the titration of the sulfuric acid by this method. The sulfuric acid in the peroxide is not removed as i t is when barium hydroxide is used, but the effect of any acid present in the peroxide is eliminated b y neutralization, thus avoiding any error in the volumetric result due t o sulfuric acid in the peroxide. Should it be necessary t o check the volumetric result gravi-
metrically, a blank on the hydrogen peroxide must be determined by precipitating the sulfate present with barium chloride and this blank deducted from the total barium sulfate.
Experimental EXCESS BARIUM IONIN HYDROGEN PEROXIDE KEUTRALIZED WITH BARIUM HYDROXIDE. Two separate portions of 3 per cent hydrogen peroxide were prepared in regular manner by diluting Super0x01 (Merck 30 per cent) and precipitating the sulfate present with 10 per cent barium hydroxide solution to the bromophenol blue end point. The precipitate of barium sulfate was allowed t o settle for 5 days at 5" C., when the solution was filtered, standardized with 0.1 N potassium permanganate, and diluted to exactly 3 per cent. Gravimetric blanks (Table I) were determined on the peroxide solutions to estimate the excess barium ion and the error due to its presence. This was done by adding 1 ml. of concentrated sulfuric acid to 20 ml. of the peroxide, and calculating the weight of
TABLE I. DETERMINATION OF BLAXKS Sample
(20-ml. sample) SO, Equivalent t o Excess Barium Ion P. p . m. 0.84
1 2
1.02 Av.
0.93
