T H E JOURNAL OF I N D U S T R I A L A N D ENGINEERING CHEMISTRY
50
emulsion. This was largely avoided b y using only a small amount of material and more water and ether, a n d a complete separation was usually obtained within 2 hrs. The extraction of larger amounts of material with the idea of running more t h a n one iodine number from one separation did not prove feasible because of t h e tendency of the material t o adhere t o glass. About 1.5 g. each of the saponifiable and unsaponifiable material were dissolved in chloroform and the volume made up t o 50 cc.; I O cc. aliquots were then used t o determine the iodine number. The checks obtained were no closer t h a n from individual separations and the results were much lower. A number of determinations of the per cent of ether extract and the iodine number of this material were made t o show the range of variation in different plants. The results obtained are shown in Table I. TABLE I w-
Plant &TO.
1 2 3 4 5
6
7 8 9
Ether Extract Per cent 5.59 4.04 3.92 4.20 4.72
4.56 6.08 3.71 3.84
Before Separation 110.4 120.9 123.8 125.7 109.7
116.4 119.0 106.2 126.2
~HUBL ~
NIJMB@RS-----. . ~ ~ _ ~ _ ~
_
-After SeparationUnsaponiSaponifiable fiable
.....
.. .. .. .. ..
148.4 124.71 122.31 126.71 142.2 .
.
I
.
.. .. .. .. ..
104.2 98.91 95.3' 101,lI 108.1
.
142.7 157.9 158.22
.....
103.4 122.3 122.02
.....
10 4.24 114.4 5.11 118.3 149.9 113.6 11 5.16 116.6 12 ..... ..... 1 These results are from using aliquots of a chloroform solution of the saponifiable and unsaponifiable matter from Plant 5 as mentioned above. 2 Duplicate of Plant 9.
From Table I i t can be seen t h a t the per cent of ether extract varies from about 3.7 per cent t o 6.0 per cent in the air-dried leaf, and the Hubl number of t h e crude ether extract from 105 t o 125. The Hubl number of the unsaponifi able material averages 149.7 a n d of the saponifiable I I 2.3, showing conclusively t h a t the former has the higher degree of unsaturation. T h e compounds which are obtained in the unsaponifiable material are probably largely cyclic alcohols (such as phytosterol and stigmasterol) and carotin, 'and perhaps some other highly unsaturated alcohols. Chlorophyll is not separated entirely from the f a t t y acids in the method used because upon boiling with alcoholic potash i t forms the tribasic salt, potassium chlorophyllide. The phytol group is split off and recovered in the unsaponifiable extract. Carotin is a neutral compound which does not react with bases. SUMMARY
The amount of ether-soluble extract from dry soy bean leaves is so small (3 t o 6 per cent) t h a t i t is doubtful whether they would be a profitable source of paint oil, even if i t were of good quality. The results obtained, however, show clearly t h a t the most highly unsaturated compounds in the leaves of the soy bean are not oils but probably belong t o the alcohol group of the waxes, which do not form a hard film on drying. For these reasons soy bean leaves do not form a n available source of oil for paint manufacture.
Vol.
12,
No. r
PURIFIED CRESOL (CRESYLIC ACID) By Herbert C. Hamilton RESEARCH LABORATORY, PARKE,DAVIS& COMPANY, DETROIT, MICHIGAN Received July 12, 1919
One of the minor problems arising from the war because of interference with importations from Europe was t h a t of obtaining a substitute or equivalent for trikresol, a proprietary article imported from Germany and extensively used as a preservative and disinfectant. As has been shown t o be true in many other instances, it is equally true in this case t h a t there is no lack in America either of crude material or of ability t o purify it. There was required only the incentive. Triltresol is so named because it is a mixture of the three isomeric cresols naturally occurring in coal t a r . These three cresols are identical in composition, but have different physical and bactericidal properties. These differences, however, are unimportant and nothing of practical value results from their separa~ tion.. Trikresol, while useful as a general antiseptic and germicide, with a phenol coefficient of 21/2 t o 3 , found its most extensive application as a preservative for serums, vaccines, and similar biologic substances. Careful research has proved t h a t for this purposk, with one exception,' no other antiseptic has been found entirely suitable, either because of its efficiency or the toxic or irritating action when absorbed from a hypodermic injection. The cresols have practically the same toxicity as pure phenol, as shown in the accompanying table, but the corrosive action is so low and the germicidal value so high in comparison, t h a t the use of phenol as a germicide is no longer logical. To illustrate: cresol with a coefficient of 3, when diluted I to 60, is equal in every respect t o a j per cent solution of phenol, while the toxicity of the solution is only onethird as great because of the degree of dilution, and the corrosive action, while not measurable with accuracy, is less t h a n one-third as great. Superficially trikresol is identical with the cresols of coal tar, since a n average sample of the latter contains not over j per cent of constituents other t h a n the cresols. But i t was very promptly observed t h a t cresol, as i t appears on the market under various names, is inapplicable for use as a serum preservative because of three specific reasons, all of which are interrelzted, namely: 1-Incomplete solubility 2-Disagreeable odor 3-COlOr
Incomplete solubility is due t o the presence of one or more of three substances, naphthalene, colored compounds formed apparently a t the expense of t h e cresols, and phenols of higher boiling point and less solubility t h a n the cresols. The disagreeable odor is largely due t o pyridine and partly t o the naphthalene which have been incompletely separated in preparing the crude phenol. The origin of the color which appears in cresol and phenol is more or less uncertain. It is probably n o t 1
Carl Voegtlin, Hygienic Laboratory, Bulletzn 96.
Jan., 1920
T H E J O U R N A L OF I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y
always due t o the same cause, but may in some cases be due t o impurities in t h e cresol, and in others, t o incidental conditions, such as the effect of light or air or the action of alkali from the glass container. It is said t h a t the germicidal value of a highly colored lot is greater t h a n t h a t of a clear straw-colored sample. This, however, is probably a hastily drawn conclusion from insufficient evidence, since different lots are found t o differ much more than a water-white and a colored sample from the same lot. Redistillation corrects the color and can improve the solubility and odor, but not t o a sufficient extent. The development of color appears t o be a property not only of cresols but also of pure phenol and no method has been devised by which such a change can be entirely prevented. The coloring matter appears t o be a new constituent and t o have properties entirely different from those of the original cresols. It remains behind on redistillation, but further quantities form so t h a t only the freshly distilled material is entirely colorless. Gibbsl ascribes the development of color by the action of sunlight t o a labile hydrogen atom and describes experiments with the three cresols in which coloration occurred in varying times with the different ones but all were affected in the same way. This, however, does not explain the immediate cause of this coloration. A sample of a freshly redistilled lot was set in the sunlight and another was kept in a n amber bottle in the dark. The first, after three months, was very slightly tinged with pink, the other was decidedly reddened. This is not an isolated case but was a careful demonstration of what frequently occurs in practice with large lots. Sufficient observations have not been made t o arrive a t a theory as t o the cause or causes of the change and no method has consistently prevented its recurrence. The disagreeable odor is due t o pyridine and naphthalene which one would naturally think were eliminated in the process of separating the cresols from the creosote oil. I n this process the acid constituents of the oil are combined with an alkali such as caustic soda or lime, and in this form should be readily separated from the neutral and basic substances. I n fact only a very small percentage of these bodies remain with the cresols and are dissolved with i t when its alkaline combination is broken up, as i t is in practice, with carbonic or sulfuric acid. The odor of crude carbolic acid is not distinctly like either impurity, since the natural odor of the cresols masks the others until they are unrecognizable. Distillation carries these impurities over and little improvement can be accomplished by this step as the sole purifying process. Nevin and Mann2 depended on separation by redistillation t o obtain the proper fraction but this, as noted above, while removing the colored and some of the insoluble impurities, fails t o remove pyridine and naphthalene, which are responsible for the odor. While this odor is perhaps negligible, it is easy t o detect the difference between two lots of cresol, one purified 1 2
J . A m . Chem. S O L ,34 (1912), 1190. Ibzd., 1917, 2752.
51
t o remove the odorous impurities, as will be described later, and the other purified by redistillation only. I n my experiments an attempt was made t o study the sodium cresylate compound when prepared in molecular proportions. I n concentrated solution no observable separation takes place t o indicate t h a t purification by this means is feasible but on further dilution the naphthalene crystallizes out in a wellrecognized form and often in considerable quantities. This can be filtered out, but on again recovering t h e cresols no material improvement in odor results, because the naphthalene is the less objectionable of t h e two. My next experiments were carried out having in mind the examination of sodium cresylate in solid form t o see if impurities could be detected, identified, and removed by tests applied t o the dried or crystallized material. I n the process of evaporating the solution it was observed t h a t the vapors smelled distinctly of pyridine, and further, t h a t after a certain time no odor of this character could be detected. Carrying this experiment t o its conclusion and recovering the cresols, they were found t o be practically free from the objectionable odor of the crude cresols and on redistillation a waterwhite soluble product was obtained with no odor b u t t h a t of the pure cresols. The practical working out of this process is as follows: Dissolve the crude cresol in a solution of caustic soda molecularly equivalent, using sufficient water t o dilute the sodium cresylate t o a 2 5 per cent solution. Then boil, or drive live steam through the solution until the odorous impurities have passed off with the steam. If the solution is boiled over a n open flame, care is necessary t o avoid concentrating the solution too much as the cresylate breaks up and free cresol is volatilized and may take fire. I t is important t o add water t o replace t h a t lost by evaporation. The time necessary t o vaporize the impurities varies with the amount present, and can be determined by smelling. “The nose knows” when the pyridine is gone. The solution should be allowed t o become cold and then observed t o see if naphthalene or other neutral oils are present. Any floating oil can be skimmed off, while the naphthalene, if any remains unvolatilized, can be removed by filtration or centrifuging. Treatment with sulfuric acid equivalent t o t h e alkali originally used will break up the cresylate and set free cresol which can be recovered as a supernatant layer over the sodium sulfate solution. Separation should be very complete, as the water otherwise present causes trouble in distilling. The removal of these two impurities, which rarely amount t o more than 5 per cent of the cresols, is therefore equivalent t o a complete purification of the substance, since the color is automatically removed by redistillation, and a careful observation of the temperature of distillation a t the end of this step insures the removal of the higher boiling phenols which are
T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING CHEMISTRY
52
less soluble than the cresols and may for t h a t reason be considered as impurities. TOXICITY ASSAY
.. .. . . . . , . , , . . . . . . . . ,Purifiedpigs cresols . . .. .. ... .. .. .. .. .. .. .. .. .. ,. .. .. .. .. .. .Guinea .Subcutaneously
Sample., , , Animal.. , . . Method. , , WT OF ANIMAL 0.572 0.611 0.557 0.640 0.572
CRESOL DOSE PER KILO 0.6 0.6 0.7 0.7 0.8
RESULB Recovered Recovered Died Died Died
PHENOL
Recovered 0.5 0.437 Recovered 0.5 0.480 Died 0.6 0.446 Died 0.6 0.480 Died 0.6 0.570 0.340 Died 0.7 . .. Toxicity about 90 per cent of t h a t of phenol: Worth Hale, Hyg. Lab., Bulletin 8 8 ; James Leake and Hugh B. Corbin, Hygienic Laboratory,
Bwlletin 110.
GERMICIDAL ASSAY Sample, , , .Purified cresols Method.. , , . A . P. H. A. phenol coefficient method' Organism.. B. typhosus DILUTIONS ---TIME A N D RESULTS-SAMPLE 5 10 15 20 1-300 1-350 1-400 1-450 1-500 c PHENOL 1-100 f 1-1 10 1-120 1-130 1-140 Coefficient 3 6 1 Committee Report, Am. J . Pub. Health, 8 (1918), 506.
. .. ... .... .. .. . .. . .
~
-
-
-
-
-
-
+++
-+ +
-
++ +
-+
-
++ +
-+
-
+ ++
The cost of the process is inconsiderable, since no complicated chemical or mechanical steps are necessary. It is evident from observation of the steps in the process t h a t no unusual equipment is needed and only the commonest chemicals are employed. It is evident, therefore, t h a t here again the German chemists profited a t our expense for many years while the crude materials waited only for proper development. The logical place for the economical production of the refined cresols is where the crude material is first separated from the oils distilled from coal tar. These crude phenols, necessarily dissolved in alkali t o separate them from the neutral oils, can, a t t h a t point, by suitable means, be freed completely from their impurities, and after fractional removal of the phenol proper, the cresols could then be recovered in pure form with one operation. The production of purified cresols is, therefore, a logical opening for American enterprise, as well as American resources, for here, as in Europe, are immense supplies of coal t a r on which t o draw for crude materials. A NEW HEXABROMIDE METHOD FOR LINSEED O n 1 By Lawrence L. Steele and Frederick M. Washburn U.S. BUREAUOP STANDARDS, WASHINGTON, D. C. Received September 17, 1919 I-INTRODUCTION
I t has been recognized for many years t h a t the hexabromide, insoluble in ether, which is derived by the addition of bromine t o linseed oil or its fatty acid, is a characteristic compound, the quantitative determina1 Published
by permission of the Director of the Bureau of Standards.
Vol.
12,
No. I
tion of which should be of the greatest importance in the examination of this oil for adulteration. Several methods for the determination of the hexabromide yield of linseed oil and its f a t t y acid have been published. I n every case, the author obtained concordant results, but when the method was used by other analysts, the results reported were unsatisfactory. I n some cases different workers using the same linseed oil obtained hexabromide yields as widely divergent as 30 and 50 per cent, The authors of this paper first made a study of the published hexabromide methods and from the experience gained have developed a new hexabromide method by which i t is believed concordant results can be obtained. A preliminary study of the hexabromide compounds prepared from linseed oil as compared with the fatty acid hexabromide led t h e authors, for two main reasons, t o concentrate their attention on those methods which involve the preparation of the latter derivative. I n the first place, the hexabromide made from the fatty acids of linseed oil is more stable, has a sharper melting point and a better crystalline structure t h a n the corresponding hexabromide derived from the glyceride. I n the second place, our present knowledge of linseed oil does not tell us whether or not the linolenic acid is present quantitatively as a simple triglyceride. It is obvious t h a t if mixed glycerides are present in linseed oil the hexabromide derived from t h a t oil might be a mixture, instead of one definite compound. I n this article, only those hexabromide methods in which bromine is added to the fatty acids of linseed oil will be considered. 11-DISCUSSION
O F HEXABROMIDE M E T H O D S
I n 1909, a representative committee of linseed oil chemists1 in this country made a study of linseed oil in order to prepare specifications for its purity. One of the tests which was studied was the hexabromide yield of linseed oil glyceride b y Tolman's method.2 The results which were obtained were unsatisfactory. I n 1911 the committee tried for the first time a hexabromide test on the fatty acids of linseed oil, employing the method of Hehner and Mitchell as described by Lewkowitsch. I n brief, the method was as follows: The fatty acids of linseed oil were dissolved in glacial acetic acid and the mixture cooled to 5' C. and bromine to excess slowly added. The mixture was allowed to stand for 3 hrs., filtered on a Gooch and washed successively with 5 cc. each of chilled glacial acetic acid, alcohol, and ether. The precipitate was dried in a steam oven and weighed. The results obtained by the committee with this method were uniformly low, the hexabromide yield running from 28 t o 3 2 per cent. I n the light of the higher results obtained by the Eibner method, t o be described later, i t is evident t h a t this method was defective in the addition of bromine. I n other words, the linolenic acid was not quantitatively converted t o hexabromide under t h e conditions followed, or 1 1 8
A. S. T. M. Committee Reports, 1909. THISJOUXNAL, 1 (1909), 340. "Chemical Technology of Oils, Fats and Waxes," 1 (1904), 365.
