1024
.
T H E J O U R N A L OF IlVDUSTRIAL A N D E N G I N E E R I N G C H E M I S T R Y
method of precipitating the curd. I t is t o be noted t h a t the grain-curd caseins were all made by one creamery, which is employing the most modern methods. I n Table VI-B the values for nitrogen are considerably higher with the grain-curd, ejector method of natural-sour, and cooked-curd caseins. This is probably due t o the fact t h a t in these cases the character of the curd is such as t o per,mit of more thorough elimination of impurities during washing. I n Table VI11 i t will be seen t h a t washing the curd thoroughly after precipitation raises the nitrogen content, indicating a removal of such organic impurities as lactose. The acidity of the milk does not influence the nitrogen content. ACIDITY-AS is t o be expected, there is a fairly close relation between the acidity and the ash contest of the casein-the higher the ash, the lower the acidity, but this value is influenced also by other factors, such as the presence of some of the acid used for the precipitation. The acidity, therefore, is influenced first b y the same factors as govern the ash content, namely, the acidity of the milk, and second, by the thoroughness of washing the curd. I n this connection the change in acidity with ash content of the skim-milk caseins of' Table VI-B is t o be noted. The caseins of Table VI- A are not so well washed and hence this relation is more obscure. I n Table VI11 i t will be seen t h a t washing the curd thoroughly after precipitation tends t o lower the acidity. A comparison of A and B, Table VI, shows t h a t the present commercial methods of making technical casein can be improved t o a considerable extent. A t present it does not seem t o be t h e practice t o wash the curd sufficiently after precipitation. The ejector method for natural-sour and the grain-curd method for acid casein seem t o give the best results. S U A I MA R Y
r--A system for the proximate analysis of casein consisting of determinations of color, odor, moisture, f a t , ash, nitrogen, and acidity is described. 2-The results of the analysis of some 2 0 0 samples of caseins of different types and methods of manufacture are reported. 3-The fat content of caseins depends upon the efficiency of the cream separation (for skim-milk casein) a!nd is independent of the method of manufacture: 4-The ash content varies slightly with t h e thoroughness of the washing of the curd, but chiefly with the method of precipitation, the important factors being the amount of acid used and the length of time and intimacy of contact between curd a n d acid. The kind of acid seems t o be immaterial, provided t h a t different acids are used in equivalent amounts. 5-The nitrogen content depends chiefly on the thoroughness of the washing of the curd. It indicates the presence of 3 t o 1 2 per cent of organic impurities (lactose, etc.). 6-The acidity in general increases as t h e ash content decreases, but is also largely influenced by the
Vol.
11,
NO.1 1
amount of acid impurities present, which in turn varies with the thoroughness with which the curd is washed. ACKNOWLEDGMENT
Experts of the Dairy Division, Bureau of Animal Industry, U. S. Department of Agriculture, the New P o r k Agricultural Experiment Station, and the University of Wisconsin have given generously of their time and information t o help in carrying out this work. FORESTPRODCCTS LABORATORY MADISON, WISCONSIN
A MODIFICATION OF THE PHENYLHYDRAZINE METHOD OF DETERMINING PENTOSANS By PAULMENAULAND C.
T. DOWELL.
Received April 24, 1919
Everyone who has made determinations of pentosans by the phloroglucinol method realizes t h a t a shorter and less expensive method is very much needed. This is especially true a t this time when it is' almost, if not impossible, t o buy phloroglucinol even a t the high price asked €or it. Our purpose in this short investigation was t o t r y t o overcome this difficulty by modifying the phenylhydrazine method so as t o make it shorter and at the same time obtain results by t h i s method which would agree with the provisional phloroglucinol method. Recent investigations make it more and more evident t h a t many compounds other t h a n the pentoses and pentosans give furfurol when they are distilled with hydrochloric or sulfuric acid, and hence neither the phloroglucinol method nor any other method of determining the furfurol coming from such a distillation will enable one t o say definitely t h a t the original substance contained a certain per cent of pent os ans. We thought i t should be possible t o precipitate t h e furfurol with phenylhydrazine and determine the excess of phenylhydrazine in the filtrate by the use of some compound which would oxidize the phenylhydrazine. Tests were made using solutions of phenylhydrazine sulfate with iodine, potassium dichromate, ferric sulfate, sodium hypobromite, and Fehling's solution. With the iodine almost theoretical results were obtained when a large excess of it was used. The amount of ferrous salt found was always less t h a n the theoretical. N o better results were obtained with the other oxidizing agents. These results are i n harmony with what Chattawayl found in his work on the hydrazines. R. AdanZfound t h a t the reaction between phenylhydrazine, zinc, and copper sulfate always gave less than the theoretical amount of nitrogen. He stated t h a t secondary products such as chlorophenylhydrazine and diazo compounds were formed when hydrochloric acid was present. E. Ebler3 found t h a t a quantitative yield of nitrogen was 1 2
J . Chem. S O C9, 1 (1907), 1323; 95 (1909), 1065. Bull. SOC. chim. Belg., 2 1 (1907), 211; abstracted in J . Chem. SOC.,
1111 9 1 (1907),
657.
* Z. anorg.
90 (1906). 53.
Chem., 4'7 (1905), 371; abstracted in J . Chem. S O C , [II],
Nov., 1919
T H E JOURNAL OF I N D U S T R I A L A N D ENGINEERING CHEMISTRY
obtained in the reaction between an ammoniacal solution of copper sulfate and hydrazine. We thought t h a t this reaction might be quantitative for phenylhydrazine also if no chloride or other halogen ion was present in the solution t o cause t h e side reactions found hy Adan. The distillation was, therefore, made using sulfuric instead of hydrochloric acid and using sodium sulfate instead of sodium chloride, to lessen t h e solubility of t h e precipitate.' I n t h e distillation t h e volume in the flask was kept practically constant by adding water as t h e distillation proceeded. After the addition of the phenylhydrazine t o an aliquot of t h e distillate the solution was stirred for the required time by bubblin'g carbon dioxide into it. An aliquot ( j o cc.) of t h e filtrate from t h e hydrazone was p u t by means of a dropping funnel into a zjo-cc. Fresenius nitrogen bulbJ2 which had been previously filled with a I O per cent solution of ammoniacal copper sulfate and heated t o expel t h e air. A 100-cc. burette was connected t o t h e tube of t h e Fresenius bulb and a glass tube provided with a stopcock connected the neck of t h e bulb with a Schiff's nitrometer containing sulfuric acid. The aliquot of t h e filtrate was brought into the bulb through the dropping funnel by closing the cock leading t o the nitrometer and lowering t h e burette The bulb was then heated so as t o keep t h e reaction mixture near t h e boiling point until the reaction was complete. Results obtained using t h e above method t o determine t h e pentosans in three samples of grain sorghum which were cut a t different stages of growth, are given in t h e table below. Results obtained by Mr. Freidemann of this laboratory using t h e same samples and t h e phloroglucinol method are given for comparison. Neither Mr. Freidemann nor Mr. Menaul knew what results the other had obtained until the work had been finished. TABLESHOWINGWEIGHTS
PENTOSANS IN ONE GRAMOF SORGHUM Phloroglucinol Phenylhydrazine Method Method 1-0. Gram Gram 0.1788 I ..................... 0.1790 0.1526 I1.................... 0.1523 I11. . . . . . . . . . . . . . . . . . .0.2149 0.2150 OF
Other determinations showed t h a t t h e addition of sodium sulfate t o t h e distillate with t h e phenylhydrazine was not necessary. The same volume of nitrogen was obtained by taking two parts of the distillate t o one of which sodium sulfate had been added and not t o the other. The time required t o make a determination of the excess of phenylhydrazine in t h e filtrate is about 20 min. This method then makes it possible t o use a much cheaper substance t h a n phloroglucinol and t o make a determination of pentosans in a much shorter time. OKLAHOMA AGRICULTURAL EXPERIMENT STATION STILLWATER, OKLAHOMA
I Recommended in U. S. Department of Agriculture, Bureau of Chemistry, Bulletin 49. 2 See catalogue of Central Scientific Company, p. 371.
I025
STUDIES ON THE NITROTOLUENES. I-BINARY SYSTEMS OF A NITROTOLUENE AND SYMMETRICAL TRINITROMETOXYLENE1~2 By JAMESM. BELLAND JAMESP. SAWYER Received July 19, 1919
This paper records t h e freezing points of binary systems in which one component is ~,4,6-trinitroxyl&ne and t h e second component is a nitrotoluene. Three nitrotoluenes have been used: +mononitrotoluene, I ,2 ,4-dinitrotoluene, and I ,z,4,6-trinitrotoluene. We have also observed .the ternary eutectic points when two of these nitrotoluenes and trinitro-m-xylene are t h e components. PREPARATION
AKD
PURIFICATION
OF
TRIKITRO-m-
xYLENE-Xylene, whose boiling point ranged between 139' and I39.8', was nitrated with mixed nitric and sulfuric acids, t h e yield being about 80 per cent of t h e theoretical amount. Several crystallizations of t h i crude product from acetone gave crystals melting at I 8 2 O, and further crystallizations did not alter this temperature, which is t h e accepted melting. point of t h e compound. We have also found that benzene is a good solvent for the recrystallization of this nitroxylene. PURIFICATION OF THE NITROTOLUENES-A11o f t h e nitrotoluenes were purchased rather t h a n prepared. The trinitrotoluene was recrystallized several times from a 2 : I mixture of 95 per cent alcohol and ether. The final product melts sharply at 80. s o , the accepted melting point. The mono- and dinitrotoluene had already been purified for another investigation, which will be reported later. The melting points, 51. z o and 69.4', respectively, accord well with the accepted values. METHOD O F DETERMINING MELTING POINTS OF BINARY
mxTuRxs--We have used t h e cooling curve method for the determination of freezing points. It is common experience t h a t the true freezing point of mixtures is almost always difficult t o read owing t o supercooling. I n a subsequent paper there will be a critical discussion of t h e proper interpretation of a cooling curve. Fig. I shows t h e triplicate cooling curves with a mixture of 65 per cent trinitro-m-xylene and 3 j per cent trinitrotoluene. I t will be noted t h a t we have taken t h e intersection of t h e cooling curve for the liquid and of t h e cooling curve where trinitrom-xylene is crystallizing out regularly, after the irregularity due t o supercooling has passed. This point may be duplicated readily, whereas the temperature a t which crystals first appear may vary, depending upon the degree of supercooling. I n this instance three readings give a freezing point of 1 6 0 . 0 ~(uncorr.) or 162.8' (corr.). The cooling of t h e mixture for each curve lasted only about 5 or 6 min., the readings being taken a t half-minute intervals. The apparatus consisted of a test tube containing I O g. of the 1 This paper is the first of a series dealing with the freezing points and thermal properties of the nitrotoluenes, the investigation having been undertaken at the request of the Division of Chemistry and ChemiFaI Technology of the National Research Council. 2 Presented.to the faculty of the University of North Carolina by J. P. Sawyer in partial fulfillment of the requirements for the degree of Bachelor of Science in chemistry, June 1919.