January, 1924
INDUSTRIAL A N D ENGINEERING CHEMIXTRY
23
The Action of Catalysts with Nondrying Oils' By Augustus H. Gill MASSACHUSETTS INSTITUTEOF TECHNOLOGY, CAMBRIDGE. MASS.
(6
R , E D OIL" (commer-
Small percentages of iron salts cause a degree of heating with olive
PRELIMINARY PROCEDURE
cial oleic acid), and red oil which may be dangerous in textile mills. The iron can Seven grams of ordinary from the ease ~ i t h be easily removed with acid. Similar eflects are, curiously enough, produced with olive oil by bleached cotton wadding which it is washed out, is sodium, magnesium, and zinc salts, the most energetic action being are -weighed out in a porteone of the most desirable produced by those metals in which the lower oxide is more stable than lain dish O r on a .watch WOO1 Oils. Its use is, howglass, and 14 grams of the the higher. ever, attained with danger Oil to be tested are poured The heating test with o/ive oil may be used as a means of determinfrom spontaneous heating. ing if a metal is capable of existing in two states of oxidation. upon the cotton and thorA case Of this came to the writer'i. notice. A quantity oughly worked into it, care being taken to replace any of whi1,e woolen yarn had oil that is lost. The cotton is then placed in the cylinder been spun and was stored on the bobbins "in the grease," tightly nailed up in large cases, in a basement. On opening packed about the thermometer so that it occupies the upper them some months later heating had taken place, changing the 4.5 inches of the cylinder, and put into the boiling water white yarn to a full brown or brownish black, rotting and corn- bath. After the expiration of an hour, the bath having been pletely spoiling it; the oil used was red oil. Moreover, until kept in active ebullition, the temperature is read. Any recently this could not be predicted, for samples from the oil which shows a temperature exceeding 100" C. in 1hour or same manufacturer behave differently. Swett and Hughes2 200" C. in 2 hours should be regarded as a dangerous oil, in 1917 showed that the cause of this trouble was a small or likely to produce spontaneous combustion. percentage of iron (0.1 per cent as Fez03) which acted as a EXPERIMEKTS WITH REDOIL TVITHAN IRON CATALYST catalyst. They, however, were not successful in removing A distilled red oil of iodine number 89, containing iron the iron by dilute hydrochloric acid; this problem seemed an interesting one, and alPo the study of the influence of oleate corresponding to 0.27 per cent iron, gave a rise of temperature to 265" C. when heated 1 hour in the Mackey other elements and the behavior of other oils. It has long been known that certain elements, such as lead, apparatus. It was therefore a very dangerous oil. The manganese, and cobalt, acted as driers with the drying oils; iron was nearly washed out of it by hydrochloric acid, when iron salts were known to have a slight influence, but it is it gave the following temperatures when heated in the Mackey remarkable that theg should have so much with the non- apparatus: Per cent Temperature drying oils, as red and olive, even though these are chiefly Iron REMARKS oleic acid or oleates. That iron does act as a catalyst is Safe. Iron nearly washed out Safe brought to the attention of the chemist when he overhauls 0.01 Safe Aged by standing 2 weeks his fishing gear, and finds nets and lines rotted by rusty hooks. I n the textile mills numerous cases of spontaneous comIron oxide, as Venetian red, was added to this oil by grindbustion have occurred, some of which were undoubtedly due, ing them together and heating a t 110" C.' for 2 hours. The not to iron, but to a chrome mordant in connection with mixture was tested, with the following results: olive oil used as a wool oil. Ingle and &lackey3 reported Per cent Temoperature Iron C. REMARKS such a case and investigated the action of other metals. 0.10 100 0.09 per cent iron added and heated a t l l O o They found that even zinc and sodium would give a dangerC. for 2 hours 0.10 123 Same as preceding but previously aerated for ous rise of temperature on testing with olive oil in the Mackey 8 hours. Unsafe, perhaps on account of spontaneous combustion apparatus, when heated 2 hoursaeration Unsafe 0.14 102 i. e., double the lefigth of time usually required. Heated at 110' C. for 2 hours 0.14 116 Aerated for 8 hours 0.14 105 Zinc is spoken of as a drier with paints, but it is not very energetic. Ingle and Mackey state the results with sodium Similar results were obtained with two other pressed red are rather anomalous and unexpected; they even found that oils. It is to be noticed that a content of 0.10 per cent of aluminium gave a rise to 100" C. in 130 minutes. The sub- iron present as oleate is about the limit; 0.14 per cent, even ject seemed worthy of further study, which was undertaken in suspension, makes the oil dangerous. If the oil stands as follows: exposed to the air or is aerated, the liability to spontaneous heating is increased, owing probably to better contact of APPARATUS USEDBY MACKEY oxygen with the oil. The iron can be removed from the oil A cylindrical copper water bath 7 inches high and 4 inches in by vigorous agitation with muriatic acid diluted with twice diameter (inside measurements) is surrounded with a 0.5-inch its volume of water, centrifuging, and washing with water. water jacket. The cover is packed with asbestos and carries draft tubes 0.5 inch in diameter and 6 inches long which cause a EXPERIMENTS WITH OLIVE OIL current of air t o be sucked down one and up the other, thus As the salts of the higher fatty acids (soaps) were troubleinsuring a circulation of air in the apparatus. A cylinder made of 24-mesh wire gauze 6 inches high and 1.5 inches in diameter some to prepare, the acetates of the different metals were is supported upon a projection from the bottom of the bath. A usually employed. The olive oil was an edible oil, with an thermometer projects down into the center of the cylinder. If a metal condenser is connected to the water bath it can be used iodine value of 79 and free acid content of 1.15 per cent indefinitely without refilling and without danger of burning out. figured as oleic acid. Two per cent of the acetate of the metal chosen were ground up with the oil and gentle heat applied 'Received August 3, 1923. t o facilitate solution. The oil was tested in the apparatus in a THISJOURNAL, 9, 623 (1917). the usual way. The results obtained are shown in Table I 8 J . SOC.Cham. I n d . , 36, 454 (1916); 86,317 (1917).
@
INDUSTRIAL A N D ENGINEERING CHEMISTRY
24
TABLE I-ACTION
OB
Vol. 16, No. 1
DRIERSIN MACKEY TESTER
(2 per cent salt in olive oil) 20
...... ..
Cobalt acetate Manganese acetate.. Lead acetate. Zinc acetate.. Chromium acetate, Sodium acetate.. Sodium carbonate.. Sodium tetraborate..
TEMPERATURE REACHED--OC.
25
30
35
40
45
50
100
146
205 113
246 179
250 196 93
a 196
Min. Min. Min. Min. Min. Min. Min. 91 175 207 211 211 Stopped
........ ........ ... ..... ...
100
55
Min.
117
90
.
..... ...... Nickel borate.. ....... Nickel carbonate.. .... Ferrous oxalate.. ..... Ferrous carbonate.. ... Magnesium acetate, . . Barium acetate..
100
158 93 90 91 90
260 116 95 93 91
a 250 101 98 93
a 106 109 96
116 169 98
163 192 101
Stopped
103
99 106
100 108
101
101
111
101 114
101 118
96
99
104
250 in 155 min. a 260 in 134 min. (I 0.73
16
110
265 in 160 min. a 0.84
9
110 120 Min. Min. Min. Min. Min. Min. Min.
93.5
100
101
...
.... ..... .. ... .. .........
102
together with a comparison of those obtained by the Livache (oxygen absorption) test, in the last two columns. The best catalysts are those in which the lower oxide is more stable than the higher. The results of both the spontaneous combustion test and the Livache test seem to prove that aluminium does not appreciably catalyze the oxidation of the olive oil used. The known driers of linseed oil-cobalt, manganese, chromium, lead, nickel, and iron-are also driers for olive oil. Table I1 shows the relative drying powers of the metals on olive oil as compiled from the results obtained in this investigation. METALSALTSON OLIVEOIL Time t o Reach 200' C. in Tester STABILITY OF OXIDES Minutes 33 Lower oxide more stable Lower oxide more stable 34 Lower oxide more stable (196O C.) 50 Lower oxide more stable 64 Intermediate 77 Higher oxide more stable (196" C.) 120 Lower oxide more stable 130 Higher oxide more stable Lower oxide more stable (110OC. 150 Lower oxide more stable 155 (110" C. Lower oxide more stable (107OC. 130 ( 990 c . 120 Higher oxide more stable ( 900 c . 156 35 Higher oxide more stable One state of oxidation No rise OB
1
From the results here shown it would seem to be unnecessary to prepare the metallic soaps, linoleates, resinates, etc., for use as driers, but the acetate, oxalate, or even carbonate, would serve equally well. A most surprising feature of the results is the comparatively marked activity of the sodium salts (except sodium chloride). The sodium acetate and sodium carbonate acted about the same, reaching about the same maximum a t approximately the same rate. Borax reached a much higher maximum in a little longer time. However, sodium chloride showed only a slight rise in temperature. There are probably two important factors that influence the catalytic action of the salt on the oil-namely, solubility of the salt in the oil, and the ease with which the salt tends to break up and form the oxide on being heated. The three salts of sodium used here differ somewhat in their stability; sodium chloride is the most stable, sodium carbonate the next, and sodium acetate the least. But all three salts are very stable at temperatures 80 much higher than those possible to attain in the tester, that it is probable that dissociation by heat plays a small part, if any, in the catalytic action of these
105
--
102
90 104
92 105
105
106
107
102
103
104
105
99
100
101
102
103
104
96
196
Below 90
C-
100
Stannous acetate.. Copper acetate.. 90 Dropped Aluminium acetate. Aluminium oleate. Aluminium oleate.. Pure olive oil Glacial acetic acid, 2% a Thermometer removed, cotton fired in air blast.
METAL Cobalt Manganese Lead Zinc Chromium Sodium Nickel Iron Magnesium Barium Strontium Tin Copper Aluminium
80
91.5
.....
TABLE11-ACTION
90
LIVACHETEST Per cent Gain in Time Weight Days 4.35 15
70
Stopped
Sodium chloride. Nickel acetate..
Strontium acetate.,
60
107
99
99
99
99
99
96
97
97 90909090-
98
99
--Below -Below -Below --Below
3.10 1.38 -1.16 0.64 Stopped -0.16
16 15 10
15 12
221 in 146 min.
110 in 150 min. 110 in
155.min. 0.33
9
0.32 2.08
9 13
107 in 130 min.
Maximum
-0.23 -0.89
15 14
salts. The solubility of these salts in water is high and approximately the same for all the sodium salts used in the tests. Therefore, we can conclude in regard to the action of sodium salts that solubility and dissociation by heat have probably no appreciable effect on their catalytic action. I n the case of the nickel salts, it is noticed that nickel carbonate showed no rise in temperature, while the borate and acetate did show a rise and ended by firing the cotton when held in an air blast. The carbonate is the least stable of the three; therefore, decomposition probably plays no part in this case, for if it did, then the carbonate should have shown the greatest rise in temperature, whereas it showed no rise. The acetate is very soluble in water and showed the quickest rise. The borate is a little more soluble than the carbonate, which is very insoluble in water. These differences in solubility in water may be comparable to those in the olive oil. Therefore, these variations may account for the difference in catalytic effect on the oil. This may account also for the fact that ferrous oxalate showed a considerable rise in the tester and ended by firing the cotton in the air blast, while ferrous carbonate showed only a slight rise, reaching only 29" C. in 2 hours. The oxalate is more soluble in water than the carbonate. It is interesting to note the catalytic power of zinc. The results place it next to lead and above chromium in activity. The probable existence of a peroxide of zinc may explain this fact. Its lower oxide is much more stable than the peroxide, which may account for its rather high activity. I n the case of sodium, the peroxide is more stable than the lower oxide, as shown by the fact that when sodium burns in air i t forms the peroxide. This may account for its being a rather weak catalyzei. An important sequence of this work, which had not been expected, is that the heating test with olive oil or oleic acid can be used as a test to determine if an element can exist in two states of oxidation; of all the elements tested, it is to be noted that but one failed to act as a catalyst. This was aluminium, which forms but a single oxide. All the others form two or more, some of which, as zinc, have but comparatively recently been discovered. ACKNOWLEDGMENT Acknowledgments are due C. H. Leander, H. Rosenfield, and M. Hart, by whom the experimental work has been performed.