ADrL 1923
INDUSTRIAL A N D EhrGINEERIiVG CHEMISTRY
36.5 ~~
The Heat Treatment of China Wood and Linseed OiIs' By Charles F. Mabery C.4SE S C H O O L O F
T
H E n a t u r e , com-
A P P L I ~SCIENCE?, D CLEVELAND,
OHIO
oil and heated to 250" C., China wood oil polymerizes slowly at temperatures below 150" C. forms a clear solution which position, and changes to the condition of saturation in oleic acid, as shown by iodine becomes solid on cooling by heat of China numbers, more rapidly in a current of air, the oxygen taking part -a jelly-like solid, reprewood oil have received conin the change. I n a current of carbon dioxide, however, the change senting a solid colloidal siderable attention within proceeds as in absence of air, giving softer products. solution-from which he recent years, especially on Linseed oil is not affected,oz only slightly, in absence of air when inferred that by chemical account of its i n t e n s e heated below 150" C. Heated in a current of air, polymerization change an equivalent to coherent quality when conproceeds slowly at temperatures up to 150" C.,more rapidly at 10 or 20 per cent of the verted by heat into a par200' C., giving products from which acids may be separatelg reoriginal oil had been transticular stage of polymeriduced to the condition of saturation of oleic acid, as shown by iodine formed into some sort of zation which renders it numbers. a polymerized p r o d u c t adaptable for the formation When polymerizedslowly in a current of air at 150" to 200" C., which formed with the of strong flexible paint and both China wood oil and linseed oil are converted into their most remaining oil, as in his varnish films,together with cohesive form-a thick, sticky jelly. The solubility of both is experiments with linseed oil, similar changes of the greatly reduced when polymerization reaches the solid condition. a solid colloidal solution; longer known linseed oil. but he insisted that the Notablv. bv the work of Fahrioi,' an: Von Shapringer (working under the direction nature of the change was unknown. The object of this paper is to define with more precision of Dr. Ubbelohde a t Karlsruhe), the structure of its principal constituents has been identified. Two of the cleavage prod- than heretofore the changes that result in the heat treatment ucts, azelaic acid ( C ~ H M Oand ) ~ valeric acid (C~HIOO~),of China wood and linseed oils a t moderate temperatures having previously been obtained, Fahrion gives Von Shap- and carefully regulated conditions, and to ascertain the ringer the credit of identifying the third, succinic acid precise temperatures a t which the various stages in composi(C4H604),which definitely prows the structure of the prin- tion occur from the beginning through the t,hin and thick jelly consistency to the solid granular end-product. China cipal acid copstituent, eleostearic, as follows: wood oil is so extremely sensitive to heat-as every manuCHa--(CH2)s-CH = CH--(CH*)n-CH = CH-(CHz)7-COOH facturer of paints and varnishes knows-that it must be Fahrion calls attention to the incorrectness of Von Shap- treated with particular care. The possible changes by heat, ringer's statement that eleostearic acid is a stereoisomeride all of which have received attention, include polymerization, of linolic acid of linseed oil, which has been shown to have absorption of oxygen, and the formation of anhydrides, the formula: lactones, and loss of carboxyl. The last change requires a very high temperature-in fact, to the breakage point of CH~-(CHZ)~-CH =CH-CHz-CH = CH-(CH2)7-COOH the molecule-and the formation of lactones is uncertain. showing that the only difference between the two acids is The particular changes to receive attention here are polymthe position of the double linkage. Von Shapringer found erization alone and in presence of air. For the former, no oleic acid in China wood oil, although earlier work had the oil was heated in tightly closed flasks in an air bath, and seemed to show that it contained 25 per cent of this acid. also in a current of carbon dioxide, and in the latter, air was In referring to the uncertain evidence based on the acetyl drawn through the heated oil. derivative on which Von Shapringer relied, Fahrion stated Table I gives the constants of the two specimens of the that according to his experimental proof this oil contains crude China wood oil used in this experiment: 10 per cent of oleic acid, besides 2 to 3 per cent of saturated TABLE I acids, all of which Shapringer overlooked in his separation 1 2 3 of eleostearic acid. It is now known that the separation of Iodine numbers 178 171 907 1161 Molecular weight this acid from oleic acid and its other associates in the wood Index of refraction 1.5191 1.1.5191 77.88 77.50 oil depends on the insolubility in ether of the lead salt of Analysis 10.91 11.21 eleostearic acid. Another point of difference between the 0 (by difference) 11.21 11.29 work of these two experimenters is the action of heat, Von Shapringer asserting that, whether heated in presence of air As shown in Table 11, the action of heat in China wood oil or not, only polymerization took place, while Fahrion, de- was tried under a great variety of conditions to ascertain pending on the uncertain evidence of increase in specific just what was due to polymerization, and what to the action gravity when the oil was heated exposed in air, asserted that of air. With exclusion of air little change was observed oxidation took place, but that the experiments of Von Shap- below 100" C., but increasing rapidly above this point through ringer were carried on in test tubes that did not allow suffi- the jelly stage, 125" to 150", between 150" and 200" C., cient exposure to the air. it was quickly transformed into the solid, brittle condition. The last word has been said more recently by Wolff, who The close concordance in composition shown in Experiments has taken up the action of heat on China wood oil from the 1,2, 3, and G , indicates that polymerization is the only change, point of view of colloidal solutions. After heating the oil and that it is solely a function of the temperature. It is for 8 h s . a t 200" C., contrary to previous experience, he also apparent that the principal change takes place a t the found that a part of the product to the extent of over 70 per higher temperatures, and in Experiments 3 and 6 that agitacent is soluble in ether, benzene, and chloroform, and the tion by COz retards solid formation, even retaining the jelly insoluble part, mixed in varying proportions with linseed condition in Experiment G a t the higher temperature, and with the higher degree of saturation shown by its iodine 1 Received October 21, 1822.
I N D UXTRIAL A N D ENGINEERING CHEMIXTRY
366
Temp. NO.
a
1
Closed
2
Closed
3
coz
4
Air
5
Air
6
c.
TABLE11-HEAT TREATMENT OB CHINA WOOD OIL Iodine Index of C No. Change Refraction %
100 125 150 150 100 128 150 90 150 60
1.15135
Vol. 15, No. 4
H
0 a
%
%
CONDITION
75.54
10.70
11.76
Solid gum
77.49
10 87
11.64
Semisolid
77.28
10.90
11.82
Sticky
75.92
10.47
13.61
Solid g u m
134
44
130
41
130
48
113
65
103
66
1.4900
65
106
1.5040
77.97
10.78
11.25
Thick jelly
!"" $?
149
29
1.5150
76.07
10.80
13.13
Thin jelly
90
141
30
1.15140
Thick sirup
90
90 125
coz
1 r;n
-1-
200 Air Air By difference.
7 8
a
65
Jelly
number. In Experiments 4 and 7, where the oil is heated in a current of air, the effect of oxygen is plainly evident in the analysis by the increased percentage, 13.61 per cent. Since for the addition of one atom of oxygen to the molecule of China wood oil 13.39 per cent is required, it appears that one-bond linkage is affected wh'en polymerization is carried to the solid condition. This difference, 2 per cent, between the composition of this product and that of the oil polymerized with air excluded leaves no doubt as to the different effects on the oil of this variation in treatment. Further evidence also appears in the different solubility of the lead salts. For practical purposes these experiments indicate that the most coherent polymerized condition of China wood oil is obtained by slow heating in a current of air between 126" and 150' C., or at a higher temperature with exclusion of air, preferably by a current of carbon dioxide. The change in molecular weights by heating is very marked. Two specimens of the commercial oil gave in stearic acid molecular weights of 907 and 1151. The thick sirup in Experiment 5 gave as its molecular weight, 2337. On account of the insolubility in stearic acid of the solid, polymerized forms of the oil, their molecular weight could not be ascertained. The change in air, therefore, includes the absorption of oxygen and polymerization.
as a pasty mass, and very little remains in solution. Both lead salts were well washed with ether while in solution, the. larger part leaving 25 to 35 per cent of the insoluble salt of eleostearic acid. After decomposition with sulfuric acid, the lead sulfate required long heating and stirring to separate the entangled acid. The washed and dried acids were examined for their iodine numbers, composition by analysis, index of refraction, and molecular weight. In the following table the numbers corrFspond to those of the oils in Table 11. TABLE 111-CONSTANTS
ACIDOF OIL POLYMERIZED UNDER DIFFERENT CONDITIONS Index Iodine of Refrac- Mol. C H 0 s CONDITION No. tion Wt. % % % 1 Closed 114 1.5040 645 2 Closed 127 1.5030 77.06 10.89 12.05 3 COB 84 1.5080 76.19 10.88 12.93 91 601 4 coz 5 Air 82 1.5050 76.32 10.86 12.76 75.88 10".92 13.10 138 6 Air Calcd. for CisHszOa 77.14 11.43 11.43 a By difference. OF
The separated acids also show evidence of oxygen addition, especially those from the oils heated with air. The higher percentage of oxygen in the acids from the oils heated in carbon dioxide is difficult to understand, for condensed carbon dioxide from a tank could hardly contain enough air to cause such a change. The molecular weights of the acids in Experiments 1 and 5, more than twice as large as the theoretical value of the acid of the untreated oil (280), are evidently the result of polymerization. In Experiments 5, 6, and 7 the iodine numbers are reduced to the equivalent of oleic acid, Iodine No. 91-that is, one set of double bonds was involved in these changes. All the treated acids separated were compact solids. POLYMERIZATION OF LINSEEDOIL
SEPARATION OF THE CHINAWOOD-OIL ACIDS The raw oil saponifies slowly when heated with sodium hydroxide, but the polymerized oils emulsify to such an extent that saponification is prohibited. But when dissolved in alcohol and heated with the solid alkali the oil slowly saponifies. To separate the eleostearic acid from its polymers and other associates in the polymerized oil, it must be converted into the lead salt, insoluble or nearly so in ether, while the lead salts of the oleic and other associated acids are readily soluble. This is best done by adding to the alcoholic solution of the sodium salt an alcoholic solution of lead acetate. Linseed oil, heated to 200' C., showed little or no change, On account of a difference in solubility the lead salts of the oil heated with exclusion of air must be treated differently either in closed flasks or in carbon dioxide, and the changes from those of the air-treated oil. The former salts remain in air were much slower than those of China wood oil. The in solution in the alcohol, and water does not precipitate iodine number of the specimen used was 173. A combustion them. But when the alcohol is distilled off the salts are left gave: C, 78.95; H, 11.14; 0 (by difference), 11.91: and as a sticky mass, which solidifies on stirring. The lead index of refraction a t 20, 1.4810. The raw oil heated in air salts of the oil polymerized in air precipitate immediately gave the values in the following table: TABLEIV-RAW OIL HEATEDIN AIR NO.
Time Hrs.
Temp. O C .
Iodine No. 110
Change 63
140
115
58 33
111 90
62 83
129
42 65
Index of Refraction
C
%
74.84
H
%
10.48
05
%
1.4900
108
By difference.
1.4950 1.4910
CONDITION
14.68)
75.17
10.78
14.05
75.17
10.71
14.12
i
Sirup Thick jelly Jelly Thick sirup
April, 1923
INDUSTRIAL AND ENGINEERING CHEiWISTRY
As in China wood oil, the addition of oxygen forms an important aid in the polymerization of linseed oil, although the change proceeds more slowly, but in prolonged heating to 200" C., as shown by iodine numbers, it is reduced in saturation to the conditions of oleic acid.
THE ACIDSFROM POLYMERIZED LINSEEDOIL For saponification the polymerized oil was dissolved in alcohol and heated with solid caustic soda. To form the lead salts, an alcohol solution of lead acetate threw down the salts as a pasty mass. The acids separated by sulfuric acid collected after long heating as an oily layer that did not solidify-a mixture of polymerized Enolic acid with oleic acid and-its other associates and isomers.
367 TABLE V
Bottle
Iodine No.
Index of Refraction
C
H
0
Experiments 5 and 7 show that linseed oil a t high temperatures and in prolonged heating has its acids reduced t o the condition of saturation in oleic acid. ACKNOWLEDGMENT The author desires to express his appreciation of the faithful service of his assistant, Mr. George Grossman, during the progress of this work.
Control of the Gelling Point of Glue' With Specific Application to Drying Barrel Linings in Hot Weather By Robert E. Wilson and William B. Ross MASSACHUSETTS INSTITUTE OF TECHNOLOGY, CAMBRIDGE, MASS.
This article summarizes the results of an inoestigation designed to determine the cause and most suitable remedy for the dificulties found in drying the glue sizing in barrels during hot weafher. Contrary to the general impression, it was found that the cause was not primarily the high humidity of the drying air, nor fermentation of the glue, but lay in the fact fhat the outside temperature was above the gelling point of the glue (about 80" F. for the solutions used). As a result, instead of a thin film being held in place until dry, the sizing coat continually drained to the bottom of the barrel, and the resulting pool fermented before it had time to dry. After considering several methods of overcoming this dificulty, it was found that the addition of small amounts of chrome alum to
glue solutions was an entirely practicable method of raising their gelling point without injuring any of the other properties of the glue or fhe resultingfilms. Extensive data are presented on the gelling point and viscosity of solutions containing Qarying amounts of glue and chrome alum. Datq are also presented on the gradual decrease in the gelling point of various solutions when kept hot. Special precautions are necessary in adding chrome alum, so as to prevent Iocalizedhigh concentrations which cause irreversible setting, but with proper mechanical arrangements it is entirely possible to obtain satisfactory glue coats by the regular drying schedules with outside temperatures as high as 100 ' F.
I
of time and regulating the temperature or concentration of
N THE process of sizing barrels with glue, as practiced extensively in the oil industry, satisfactory drying of the glue film during hot weather has been found almost impossible, The glue, instead of drying on the sides of the barrel, drains down and forms pools in the bottom, and soon gives unmistakable evidence of fermentation. This difficulty has generally been attributed either to rapid fermentation of the glue during hot weather, or to the retardation of the drying caused by the high humidity of the air blown through the barrels. The brief study described in this paper was undertaken a t the request of the Vacuum Oil Company in an effort to determine the cause of these difficulties and the best method of remedying them a t the barrel factory in their Rochester plant. The customary process of applying the second heavy coat of glue sizing to a finished barrel is to pour in hot, concentrated glue solution (generally from 30 to 40 per cent glue a t temperatures from 110' to 140' C. depending on weather conditions), rotate the barrel so as to coat all the inside surface, and then drain out the excess glue through the bung for a predetermined length of time. The barrels are then taken to the drying room where a blast of atmospheric air is blown into them for several hours until the surfaces are substantially dry. The thickness of the glue coat is controlled by noting the average glue consumption per barrel over fairly short periods 1 Received October 23, 1922 Published as Contribution No 70 from the Research Laboratory of Applied Chemistry, M I. T.
the glue solution so as to leave in just the proper amount after draining-thus, in winter weaker solutions of glue are used than in summer, in order to obtain the same average thickness of glue coat. A preliminary investigation of the source of trouble with the drying process in hot weather brought to light the unsuspected fact that neither fermentation nor the high humidity of the air was the primary cause of the difficulty, but that the failure of the glue to gel before the end of the draining process was the real source of trouble. At ordinary temperatures the glue layer un the inside of the barrel cooled sufficiently to gel in a coat of the desired thickness before draining ceased, but during hot weather the film did not gel until after the drying had been in progress for some time and most of the glue had drained down into the bottom of the barrel. The resultant pools naturally dried but very slbwly and fermentation generally took place before the drying was completed, but this was entirely a secondary phenomenon. This explanation was further borne out by the fact that the gelling point of a 35 per cent solution of the glue used was found to be 82" F., just a t the point which would have been predicted by the observation that drying was never satisfactory when the temperature in the drying room was above 80" F. Throughout this paper the "gelling point" is defined as the temperature a t which the glue solution just ceases to flow when slowly cooled and stirred with a thermometer. It can