Nov., 1916
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 C H E M I S T R Y
cedure is not satisfactory for t h e estimation of free f a t t y acids in alkali soaps, owing t o t h e very decided hydrolysis of such materials in alcohol a n d ether, apparently no provision has been made for t h e occurrence of similar hydrolysis of t h e metallic soaps in t h e analysis of boiled oils a n d varnishes. Rag$ does mention t h e hydrolysis of lead a n d zinc linoleates b y water in discussing t h e perishability of paint films, but i t has not generally been ,known t h a t these soaps are similarly affected by alcohol. While carrying out a n investigation as t o t h e r81e of metallic soaps in t h e deterioration of mixed paints during storage t h e authors were struck by t h e very evident hydrolysis of t h e metallic linoleates and resinates during t h e estimation of t h e free f a t t y acids in the vehicle of t h e aged paints under observation. Since these salts of linseed oil are so easily hydrolyzed, it is apparent t h a t their presence in t h e dissolved state would cause a n oil t o show a n acid value which would include t h e amount of alkali necessary t o hydrolyze t h e soaps present as driers as well as t h a t required t o neutralize the free f a t t y acids. This is a point which we believe has never before been brought t o t h e attention of t h e oil chemists and is of great importance. I n order t o study this influence t h a t t h e metallic soaps exert on t h e determination of t h e acid value in oils, those soaps of linseed oil which are most likely t o occur in a boiled oil or varnish, i. e . , lead, manganese, cobalt, calcium and zinc, were prepared b y precipitation from a neutral solution of t h e sodium soap, and dried in a vacuum. Using 2 8 0 as t h e molecular weight of t h e acid radical, t h e theoretical acid numbers, assuming complete hydrolysis of KOH, were computed as follows: LEAD 146.2
MANGANESECOBALT 182.4 181.2
CALCIUM 186 8
ZINC 179.3
On titrating these soaps dissolved in a n alcoholether mixture, t h e actual acid numbers found varied only a few tenths from t h e above figures, except in the case of calcium soap. This showed a n acid number of b u t 7 9 . j, indicating, as might have been expected, only partial hydrolysis. I n these titrations t h e final end-point was taken when t h e pink of t h e phenolphthalein would remain for several hours without vanishing. However, t h e amount of alkali necessary after t h e first appearance of t h e fugitive pink color was in no case considerable. The rosin salts of t h e same metals mere prepared in a similar manner and dried in vacuum. On titrating these in a n alcohol-ether solution, they gave t h e following results. T h e calculated acid value was computed from t h e formula C20HaaOzfor rosin: APPARENT CALCULATED ACIDVALUE ACID VALUE Assuming Complete Hydrolysis 113.0 139.0 Lead salt.. . . . . Manganese s a l t . . 168.3 171.2 Cobalt s a l t . . . . . . 160.0 170.3 Calcium salt. . . . 8 6 . 0 175.3 Zinc s a l t . . . . . . . . 149 .O 168,i
. .
Percent Hydrolyzed 81.3 9 84 ., 30
.
-An oil having a n original acid value of 3 . 1
Oeslevv. Chem.-Zlr., [2] 12, 62-4.
49.1 88.3
was
997
heated with a quantity of lead linoleate t o a slightly cloudy solution. This boiled oil had a n ash content of I . I j and a n acid value of I O . 2 0 . A sample of t h e same oil run parallel to it b u t without t h e drier showed a final ash content of 0 . 1 8per cent and a n acid value of 2 . 3 1 , t h e decrease in t h e latter value being due t o t h e volatilization of some of t h e f a t t y acid originally present. The difference between t h e ash content of t h e oil containing drier and t h e oil without it indicated t h e presence of approximately one per cent of lead oxide in solution as t h e linoleate. This should cause a n apparent acid number of 7 . I for t h e boiled oil, assuming hydrolysis t o t h e same extent as in t h e previous experiments. The value of I O . 2 0 actually found might indicate t h a t t h e lead exerted some slight saponifying action on t h e oil a t t h e high temperature employed in effecting t h e solution. This error in t h e determination of the acid value of oils is in many cases not serious since t h e amount of soluble metallic soaps is limited. But in varnishes. where t h e hydrolyzable substance in solution is a metallic soap of rosin or gum, which may be present in considerable quantities, t h e apparent acid value is but a poor criterion of t h e conditions actually existing. UNIVERSITYOF MICHIGAN,ANN ARBOR
STORAGE CHANGES IN VEGETABLE AND ANIMAL OILS By HENRYA. GARDNER Received October 5, 1916
The writer’s attention has recently been called t o instances where linseed oil t h a t has been separated from paints has shown low chemical constants. I n one instance, several barrels of paint made of pure raw linseed oil stood in a factory yard for a few weeks, A temperature exposed in the daytime t o t h e sun. of 1 1 5 ’ F. was probably reached in t h e middle of t h e day. Upon analysis, t h e consumer found t h e oil t h a t was separated from t h e paint t o have a lower iodine number t h a n t h a t called for by t h e specifications. The shipment of paint was accordingly rejected. Since pure linseed oil of t h e proper iodine number was used in t h e paint, i t is apparent t h a t hydrolysis occurred on storage. T h a t such changes are possible should, therefore, be considered b y t h e testing engineer when examining t h e vehicles of specification paints t h a t have not been freshly used or t h a t have been exposed to high temperatures during storage. The writer has previously pointed out the degree of change which may be expected t o take place in linseed oil when ground with various pigments,l a n d he has also indicated the contributing effects of impurities in oils.2 Some more recent work has shown t h a t nearly all oils, even when not in contact with pigments, will show changes in their chemical constants when allowed t o stand for a period of time. The drop in iodine number is generally accompanied by a rise in specific gravity and acid value. Quantities of a number of oils mere obtained by 1 T h e “Effect of Pigments upon the Constants of Linseed Oil.” H. A . Gardner, J. Fvank. Inst., Oct., 1912, P P , 415-423. * “Changes Occurring in Oils and Paste Paints. Due t o Aut.ohydrolysis of the Glycerides,” H . A . Gardner, J F r o n k . I n s t . , M a y , 1914, pp. 533-540.
998
T H E JOL'RNAL OF I N D U S T R I A L A S D E-VGIJEERING C H E M I S T R Y TABLE 1-.4NALYSES
OF
OILS
rlnalyses were made when oils were first obtained and 44 mos. later, a t t h e time they were used in repainting tests. T h e oils were again analyzed 22 mos. later in September, 1916. Chemical changes occurring in the oils are denoted by the constants. Ref. Iodine Sapon. Sp. G r NO. NO. Acid N o Index R A W LINSEED OIL 186 188 2.0 March, 1911 . . . . . . . . . . 0.931 1 ,4867 185.4 189.6 2.8 November, 1914.. ...... 0.933 176.9 1,4798 190,2 3.3 September, 1916.. . . . . 0.936 SOYA BEAN OIL 129 .... March, 191 1 . . . . . . . . . . 0.924 189 2.3 193.1 4.7 1.4813 130.2 November, 1914. . . . . . . 0.925 1 ,4733 122.0 192.1 7.0 September. 1916.. . . . 0.937 MENHADENOIL .... 158 187 March, 1911.. . . . . . . . . 0.932 3,9 1.4850 November,.1914., , , . , 0.934 156.3 193.7 16.1 1 ,4768 144.9 191.4 19.2 September, 1916. . . . . . 0.938 R A W TULCOIL 166 183 3.8 .... March, 1911.. . . . . . . . 0 . 9 4 4 1 ,5050 161.5 190,3 5.7 November. 1914 ....... 0.946 5.6 1.5138 158.6 188. 7 September; 1916. . . . . . 0.944 PERILLA OIL 180 188 .... March, 1911.. . . . . . . . . 0 . 9 4 2 7.4 1 ,4874 172 195.4 November, 1914 ....... 0.94 14.8 1 ,4767 160.9 193.3 September, 1916.. .... 0.939 PERILLA OIL (SPECIAL) (a) 192 189 3.2 .... March, 1911 . . . . . . . . . . 0.94 1.4978 123.8 219.4 20.8 November, 1914.. ..... 0.981 1.4840 122.4 220.9 31.2 September, 1916.. . . . . 1.000 HEAVYBODIEDLINSEEDOIL 133 189 2.8 . March, 191 1 . . . . . . . . . . 0.968 1 ,4966 130.5 200 6.3 November, 1914....... 0.992 1 ,4876 124.4 206.3 9.0 SeDtember, 1916.. . . . . . . . . . LITHOCRAP~IC LINSEED OIL 102 199 2,i .... March, 1911. . . . . . . . . . 0.97 1.4978 103,4 150.9 13.4 November, 1914.. ..... 0 . 9 6 1.4890 108.5 15.2 137.7 0.974 September, 1916.. , . WHALE OIL 148 191 9.2 .... March, 1911.. . . . . . . . 0 . 9 2 4 1 ,4820 138.2 191.2 17.4 November, 1914 ....... 0.926 BOILED LIHSEEDOIL March, 1911.. , , , , , , , , 0 , 9 4 1 172 187 2.7 170 November, 1914.. . . . . . 0.943 1 ,4895 188 3.1 CORN OIL 124.8 November, 1914 ....... 0.921 1.4800 190.1 4.1 121.3 191.1 1.4707 September, 1916. . . . . . 0.924 4.6 COTTONSEED OIL 1.4781 111.7 194.3 November, 1914. . . . . . . 0.920 0.9 1.4681 110.6 192.9 September, 1916. . . . . . 0.924 1.4 ROSIN OIL 35.5 November, 1914 . . . . . . . 0,964 .... 68.9 32.4 66.0 36.6 .... September, 1916.. . . . . 0.964 31.6 TREATED T u x OIL(^) 101.3 1,4764 56.4 7.7 November. 1914.. . . . . . 0 . 8 8 2 103.2 1.4660 53 2 September, 1916.. . . . . 0.884 8.0 LGMBANG OIL 189 November, 1914., ..... 0.927 162 I ,4789 1.0 188.9 September, 1916.. . . . . 0.926 164.0 1 ,4748 1.9 SCNFLOWER OIL 1 ,4796 7.5 124.6 November, 1914.. . . . . . 0.924 189.3 1.4712 190.2 122.2 9.0 September, 1916.. . . . . 0.923 HEMPSEEDOIL November, 1914.. . . . . . 0.927 3.9 1.4822 191.1 149.4 146.1 5.0 September, 1916. . . . . 0.930 1 ,4745 191.0 SHARKOIL November, 1914.. ".., , 0.910 5.2 1.4815 158.9 132.8 September, 1916.. . . . 0.915 6.2 1.4722 163,3 127.4 SARDIXE OIL November, 1914. . . . . . . 0.919 10.4 1.4800 1J7.3 134.6 31.1 September, 1916. . . . . . 0.962 1.4755 180.2 91.4 PETROLGUM OIL I ,4773 November, 1914 ....... 0.851 52.9 28.2 1.1 1.0 September, 1916. . . . . . 0.850 1,4669 48.6 28.0 (a) H a s become highly viscous. ( b ) Heat treated with driers and thinned with mineral spirits.
....
.
. I
. / . .
TABLE2-RFFECT O F S T E R I L I Z A T I O N O N OILS Original oil analyzed on Kovember, 1914. Individual portions placed in separate bottles. One set of bottles sealed and heated t o 110' C. Both sets again analyzed during September, 1916. Sp. Gr. RAW LIXSEED011. Sovember, 1914 . . . . . . . . . 0 . 9 3 3 September, 1916. . . . . . . . 0.936 Sterilized-Sept,, 1916.... 0.934 SUNFLOWER OIL November, 1914 . . . . . . 0.924 September 1916.. . . . . . . 0.923 SterilizeddSept., 1916. . . . 0,925 LMENHADEN OIL November, 1914 . . . . . . . . . 0 . 9 3 4 September, 1916.. . . . . . . 0.938 Sterilized-Sept., 1916.. . 0.937 PERILLA011. November. 1914 . . . . . . . . . 0.94 September, 1916. . . . . . . . 0.939 Sterilized-Sept., 1916.. 0.941
Iodine No.
Sapon. No. Acid No.
Ref. Index
185.4 116.9 187.1
189.6 190.2 190.4
2.8 3.3 3.1
,4867 ,4798 ,4776
124.6 122.2 124.6
189.3 190.2 189.9
7.5 9.0 6.3
,4796 ,4712 ,4696
156.3 144.9 155.6
193.7 191.4 192.1
16.1 19.2 16.3
,4850 ,4768 ,4782
172 160.9 174.3
195.4 193.3 194.2
7.4 14.8 12.6
,4874 ,4767 ,4788
1'01. 6 , S O .I I
TABLE.?-RAPID HYDROLYSIS OF 1,IKSEED OIL ~ M U L S I O K 5 Eoual Darts bv volume of water and oil emulsified with 1 Der cent neutral g i m 'Emu1s;ons kept ip incubator a t 20' C. for 4.5 days. A t this temperature, in the presence of water, and with a possibility of the accelerating action of enzymes, oils have rapidly developed free acid. April 1 .4pril 6 Bpril 18 M a y 4 May 16 No. ACID VALUB: 2.1 3.0 5.6 6.0 8.4 1 R a w. . . . . . . . . . . . . . . . . . . . . 2 Raw-Sterilized.. . . . . . . . . . 2 . 1 2.5 2.6 4.6 6.2 3 AlkaliRefined.. . . . . . . . . . . 1 . 6 2.4 2.4 6.6 15.6 15.8 4 Acid Refined (Washed) 24.8 35.8 44:8 5 Acid Refined. . . . . . . . . . . . . . 3 . 0 34.0 38.0 4.0 15.0 6 Boiled.. 4.6 1.4 13.6 22.0 38.8
....
.................
t h e writer several years ago for use in some practical exposure tests t o determine t h e wearing properties of such oils when used in paints. Analyses of these oils a t various periods are shown in t h e accompanying tables. INSTITUTE
RESEARCH WASHIKGTON
O F IBDUSTRIAL
THE NITRATION OF TOLUENE TO TRINITROTOLUENE By IRWIN W. HUMPHREY Received M a y 6, 1916
-4lthough t h e nitration of toluene t o trinitrotoluene has been carried out on a very large scale since Hausserman started its manufacture in t h e Chemischen Pabrik Griesheim in 1 8 9 1 , very ~ little has been published on the subject. I t s employment as an explosive for military manufacture for this purpose appears t o have been taken u p next b y Italy in 190; and still more recently b y other countries. Generally it is preferred t o make first crude niononitrotoluene (consisting chiefly of t h e ortho derivative with a small percentage of t h e paranitro compound) b y nitrating toluene with sulfuric acid (1.84) and nitric acid ( 1 . 4 2 ) , and then nitrating t h e crude mononitrotoluene thus obtained t o trinitrotoluene, b y separating from t h e dilute acid and then treating with a mixture of sulfuric acid (100 per cent H2S04) and nitric acid ( 9 2 t o 94 per cent H S O s ) . (Xitration in two steps, first t o dinitrotoluene and then t o t h e trinitrotoluene. T. X. T., is also carried out.) T h e method extensively employed industrially. involving t h e nitration of mononitrotoluene, mas described briefly in I 9 I 2 b y Langenscheidt. The following d a t a indicate clearly t h a t the yields of T. N. T., obtainable a t a given temperature, are b y no means a. function of t h e water concentration of t h e reaction mixture as has often been assumed from t h e equation:
It is known t h a t dilute nitric acid (sp. gr. 1 . 1 2 ) a t yields principally phenylnitromethane, C6HaCH2N02.3 On t h e other hand, a n acid mixture of too low m-ater concentration gives poor yields, owing t o oxidation, trinitrobenzoic acid being one of t h e products. This is well shown b y t h e results tabulated below. I n his recent book on t h e nitro explosive^,^ Escales recommends substantially t h e method described b y Langenscheidt. While t h e relative cost of toluol, acids, a n d t h e alcohol, Tyhich may be used 100'
12.angew. C h e w , 4 (1891), 508. 2.ges. Schiess Sprengsloffw., 1912, 425. 8 Ronowalow, Ber., 28 (1895), 1860, 1861. 5
4
"Die Nitrosprengstoffe." Leipzig, 1915, 149.
