L. M . T O L M A N A N D I,. S. M U N S O N .
244
reputable firm in whose currant jelly we reported salicylic acid but which was present in no greater quantity than we have since found it in the fresh currants. A similar experience was lately had in one of the state laboratories for food control. In addition to the above work we are studying the distribution of benzoic acid in fruits and vegetables, and hope to be able to publish our results within the year. M O N T A N A EXPERIMEST STATION, BOZEM.4Ki, MONT.
s.
[CONTRIBUTIOX FROM T H E BUREAUO F CHEMISTRY-v. DEPARTMENT O F .4GRICl:LTURE, NO. 46. -SENT BY H. \%'TILEU.]
\v.
IODINE ABSORPTION OF OILS AND FATS. A COMPARISON OF METHODS. BY I,. M. TOLMAX ANI) I,.S.MUSSON. Received Dc'rnlber
?3.
IWZ.
THEHub1 method has been used almost entirely for this determination but it has several faults. First, the solution rapidly loses in strength so that it will change materially during a determination, and after standing a week or so becomes too weak for use. Second, it is so slow in its reaction with some of the oils, such as linseed, that a very serious error is brought about by the change in the strength of the solution during the time of reacting. Wijsl showed how considerable this error might be, and the following table taken from his work shows how much difference the time of titrating the blank makes in the iodine number. TABLEI.--IODINE NUMBZRS O F LINSEEDOIL Time of absorption. Hours. 2
7 24
Blank titrated a t beginning.
173.74 177.65 18r.Sg
BY Hi;.BI. METHOD.
Blank titrated at end.
.... .
170.39 163.16
These figures show a decrease in the iodine iiumber after seven hours, if the blank is titrated at the end of the determination. Wijs considers that the true iodine number in this case lies between 173.7 and 181.89. This variation is a very serious objection 1
Chew. RN.Felt.
u . H u r / z . Z?zd,,6 , 6 (1899).
I O D I N E ABSORPTION OF OILS A N D FATS.
245
to the Hiibl method, and, on such oils as have a high number, greatly reduces the value of the determination and probably accounts for the wide range of figures obtained by different analysts. A solution that would remain permanent would completely remove this source of error. Two new solutions have been proposed as substitutes for the Hiibl, and it is claimed for both that they change but little with age, and are also more rapid in their action than the Hub1 solution, yet give results agreeing very closely with those obtained by the Hiibl method. The Wijsl solution is iodine chloride dissolved in glacial acetic acid. The H a n u 9 solution is iodine bromide dissolved in glacial acetic acid. Wijs3 claims that a solution, made up with acetic acid (gg per cent.) showing no reduction with bichromate and sulphuric acid and containing a very slight excess of iodine in order to prevent the presence of any iodine trichloride, remained practically constant for eighteen months. Twenty-five cc. which were neutralized by 47.3 cc. N/IO thiosulphate at the beginning, at the end recluired 46.9 cc. N / I O thiosulphate. Lewkowitsch' found that after five months the strength of such a solution was practically unchanged. O u r experience has been the same. A solution remained practically without change for a month. Hanus5 showed that a solution of iodine bromide in glacial acetic acid changes only very slightly in three months, and our experience has corroborated this claim. The Hanus solution was prepared as suggested by Hunt;O 13.2 grams iodine were dissolved in 1000 cc. glacial acetic acid (99.5 per cent.), which gave no reduction with bichromate and sulphuric acid, and then enough bromine was added to double the halogen content of the solution. This takes approximately 3 cc. of bromine. The iodine solution should be cold when the bromine is added. Forty cc. of a solution prepared as directed above contained 1.0300 grams of iodine : a month later it contained 1.02791 grams. This difference might easily be due to the difference in temperature at the time the 40 cc. were measured. In order to get strictly accurate results a 1
2
3 4
6
Bey. d. chem. Ges., 31, 750 (1898). Zfschr. Nahr. u. Genus., 4, 913 (1901). fbid., 5 , 499 (1902). Analysf,24, 257 (1899). Zfschr. Nahr. u. Genus., 4,916 (1901). 3 . SOC. Chem. hi.. a i , 454 ( I ~ z ) .
L. M . T O L M A N A N D L.
246
S. 3IUNSON
number of 20 cc. portions were measured into stoppered flasks and titrated from day to day. The following table shows that the solution is constant: TABLE11. N
Time of standing Days.
I O thiosulphate 20 cc. iodine.
=
cc.
45.50 45.50 45.55 45.50 15.55 45.60
This is enough to show that either of the solutions is satisfactory and very much better than the Hub1 in retaining its strength. The second objection to the Hub1 solution is slowness of reaction. The time recommended to be allowed in order to complete the reaction is variously given by different authors from two to twenty-four hours. Allen' recommends two hours. T h e official method of the Association of Official Agricultural Chemists requires three hours. Lewkowitsch* recommends four hours. Wijs3 showed that with linseed oil the absorption was not complete in seven hours and perhaps longer. These different methods will give varying results. Wijs recommends, when his solution is used, fifteen minutes for non-drying fats and oils, thirty minutes for semidrying oils, and one hour for drying oils. T h e following table showing results obtained with both Hanus' and Wij s' solutions substantiates these claims. TABLEIIL-TIME NECESSARYFOR
COMPLETE .4BSORPTION O F IODINE. Iodine numbers 5y Iodine numbers by Hanus' method. Wijs' method.
, . c _ -
Lab. No.
I
Kind of oil.
,s
min.
4 5 9
Cocoanut ...... 8.7 Koniit. ........ 6.4 Butter 55.2 I' ......... 35.1 Oleo oil ....... 43.2 Oleomargarine. 52.4 .51.g . 64.6
I
52.2
773 22,077
1,170 1,168
.........
.
7
30
min.
I
min.
....
....
5.6 6.4 35.4 35.5
.... ....
43.4
.... .... ....
51.8 52.1
65.1 52.4
....
.... ....
"Commercial Organic Analysis," Voi. 11. Part I , p. 64. 2 '*ChemicalAnalysis of Oils, Fats and Waxes," 18yY.p . I;?. a Chem Rev. Fell. u . Harlz. I n d , 6 , i (1699). 1
--
1,s
hr.
.... ....
35,s
---. .I?
min.
I
hr.
.... ....
.*..
....
.... ....
35.9
36.0
....
.... ....
.... ....
....
52.9
.... ....
53.0
.... ....
....
....
....
IODINE
Iodine numbers by Wijs' method.
Iodine numbers by Hanus' method.
----
Lah.
Kindofoil.
KO.
T.5
miu.
Oleomargarine . 80.4 Lard oil ....... 70.0 Oliveoil... 90.1 I' 82.1 86.5 '' Peanut 96.0 125.0 Mustard
23,606 j98 960 833 772 777 771 776 775 770 774 1,162
247
ABSORPTION O F OILS A N D FATS.
.... ....... ....... ........ ....... ............. .............
....
.............. .............
Rape Corn. Poppy ......... 136.8 Linseed . . . . . . . . . . .
?p
min.
80.5 69.7 90.4 81.7 86.5 97.4 126.0 119.4 105.0
107.4 116.8 137.4 186.3
----
I
hr.
1.5 min.
30 min.
I
hr.
....
....
.... ....
.... ....
....
90.3
90.0
82.2
82.6 86.5 98.8 126.5
91.2 82.6 86.7 99.0 126.5 118.0 104.3 105.8 116.1 138.9 190.2
90.9 82.5
86.5 97.7 126.0 120.0
105.0 107.5 117.6 138.4 186.2
. . e .
....
.... 117.1
138.8 183.S
.... 87.2 99.0 128.0
118.2 104.6
105.7 118.5 139.1 188.7
For butter, lard, oleomargarine, and olive oil, the reaction is practically complete in fifteen minutes. With cottonseed, sesame, and mustard, thirty minutes are necessary, while with poppy and linseed one hour gives the best results. There seems to be practically no difference between the two solutions in respect to speed of action. Tables IV and V give the results obtained by ourselves on a number of oils by all three methods. Table VI is a compilation of the data available on the three methods. TABLEIV.-IODINE Lab No.
N U M B E R S O F OILS AND
Hubl's number, 3 hrs.
Oils.
Wijs's number 30 min.
Hanus's number. 30 min.
9.05 6.43 36.2 35.9 43.5 53.5 53.7 52.9 66.0
8.60 6.40 35.3 35.4 43.3 52.3
No n -drying.
....
Cocoanut.. Konut I , I i o Butter 1,168 Butter Oleo oil I Oleomargarine 4 " 5
........ ....... ........ .......
9
4
...... ........... .. .. .. .. ..
23,606 Lard oil. 487 1,181 Magnolia oil 1,182 " 772 Peanut " 1,149 '* " 492 I' / '
1 2
8.93 6.09 35.3 34.8 42.6 53.6 52.8 52.5 66.3 69.3 73.7 81.7 76. I 96.3 94.5 107.7
Commercial oils. Adulterated with cottonseed
70.5 74.5 79.4 75.6 99.0 95.2 109.5
52.2 52.0
64.8 69.8 73.9 78.9 74.0 97.4 94. I 107.7
FATS. Difference between Wijs a n d Hubl.
Difference between Hanus and Hubl.
+- 0.12
- 0.33
t
+ + 0.60
0.34 f 0.90
+ 1.10
+ 0.31'
0.00
+- 0.90 +- 0.70 1.30 - 0.60 ++ 0.90 0.40 0.10
0.50~
- 0.30
- 1.50
++ +- 0.70 2.30 +
1.20
- 0.50
+ 3.00 t
+
70 1.80
0
0.50
0.20'
2.80
- 2.10
+ 1.10
-
+
0.10
0.002
I,.
248
Ivf. TOLMAN A N D L. S. MUNSON. Hiibl's number, 3 hrs.
Oils.
Lab. No.
770 Mustard oil . . 771 " ( ' .. 776 " .. 486 " .. " " 495 . 775 Rape oil . 490 " Semi-dgyiiq aitd Sunflower. . 1,159 Cottonseed oil. 1,160 'I " . I' 1,161 489 Sesame oil . . . 444 Corn " .. .. .... 49' ( ' .... 777 493 Poppy . ... .... 774 1,162 Linseed " . . . . 1,188 I ' ..
110.4 113.0 96.4 103.5 106.4
' I
' I
.. . .. . ......
I '
- ..
. .
' I
* (
'I
Hanus's number, 30 min.
118.5
115.5 116.8 103.8
118.2
Difference between Wijs a n d Hiibl. .L8.10
-!-
Difference between Hanus a n d Hiibl.
5.10 5.80 .t- 5.30 6.70' 8.40' -4- 3.80 -1-
5.20
7
101.3
104.3 112.5 117.3 105.7
I0j.2
100.2
104. I
102.8
f10.90 --- 4.40 3.90
109.2
107.2
+
2.so
-r 0.80
105.3
105.2 107.8 106.7 106.j
'
1.50
7-1.40
1.10
- - 1.60
1.40
drying 106.4 103.8 106.2
I C
Wijs's
numba 30 min
104.8 ro6.4
107.3 106.2 107.0
119.0 119.0 123.3
133.4 134.9 169.8 179.5'
110.2
114.8
:- 5.90
-:-
-;- 9.00
1 1-
4 -
-;-2 . 6 0 1
+
1.90
f 0.60
!
0.101
-*-
,
1.201
123.2 122.2 129.2
120.2
4
4.20
119.6 126.0
-.-
3.00
.- 0.40~
j.s0
-- 2 . 7 0
735.2
132.9
-- 1.80
-
'39. I 1S6.j I 88. 7
138.4
... 4.20
'.
1 s4. j
-:-16.70 - 9.20
183.7
~ . -
o.jo1 3.50 14.70' 4.20
It will be seen from these results that for oils and fats with an iodine number under IOO there is little difference between the three methods. For practical purposes they are the same. But this is not so true for the oils with higher iodine numbers. With mustard and rape oil there is a marked difference, 8.4 iiumbers higher with Hanus's and 10.9numbers higher with Wijs's method. It is with these oils and linseed oil that there is the widest variation, but there can be little doubt that the higher numbers obtained by the Hanus and Wijs methods are the more correct. TABLEv.-IODINE Lab. No.
795 796 797 798 960 833 1
1
NUMBERS O F OLIVE OILS.
Hiibl's number,
Wijs's number, 30 min.
Hanus's number, 30 mrn.
Difference between Wijs and Hiibl.
89.7 39.7 89.8 89.7 80.9 84.S
90.9 90.6 91.4
90.4
$-I.2
7.0. 7
90.0
I-0.3
91.1 s2.5
90.4 S1.7
0.9 -1.6 tI.4 :-1.6
56.7
86.5
7-0.8 +I.7
3 hrs.
Commercial oils. Four hours for Hub1 determination.
90.0
L1.9
Difference between Hanus a n d Hubl.
0.2 '-0.7
-7
I O D I N E ABSORPTION OF OILS A N D FATS.
Lab. No.
831 832 834 936 933 835 962 955 95 2 840 935 953 958 959 954 934 956 932 961
836 93 I. 838 839 837 673 841 84 2
843 844 1,ogI
Hiibl's number. 3 hrs.
Wijs's number,
85.2 84.9 84.5 81.8 80.6 82.7 81.2 82.6 81.3 86.3 80.5 79.2 81.8 81.4 86. I 81.1 84.5 80.5 80.8 86.0 80.7 86. I 89.0 84.0 83.3 86.9 87.2 85.1 84.2 81.9
86.7 86.5 86. I 82.8 81.6 83.1 81.j 83.6 82.2 87.8 80.9 79.9 83. I 83.4 87.7 82.4 85.9
30
min.
82.0
82.3 87.7 82.2
87.9 97.3
...
84. I 88.0 88.I 86.6 85.0 83.4
Hanus's number, 30 min.
Difference between Wijs and Hiibl.
85.9 85.9
85.1 82.8 81.1 82.6 80.9 83. I 81.8 86.7 81.6 80.6 81.8 81.9 80.3 82.3 85.6 81.4
Difference between Hanus and Hiibl.
+I4
$0.7
+1.6 +1.6
5i.0
+I.O
+I.O
$0.6
+1.0
f0.5
+0.4 $0.3
-0. I
+I.O
+0.5
$0.9 SI.5 $0.4 to.7 iI.3
50.5
t2.o
+0.5 $0.2
+1.6 $1.3 $1.4 $1.5
81.1
$1.5
86.6 81.5 87.1 89.9 84.9 82.6 87.3 87. I 85.6 84.2 82.0
tI.7 SI.5 +1.8 +2.3
Average, 36 samples,
249
-0.3
$0.4 +I.I
t1.4 0.0
tr.1 +I.I
$0.9 $0.3 $0.6 $0.8 +I.O
...
+0.9 $0.9
+0.8
4 . 7
+I.I
+0.4 +o. I $0.5
$0.9
SI.5 +0.8 +I6 +I. 2
0.0 +I.IO
-$0.59
The iodine numbers of the thirty-six samples of olive oils of known purity obtained by the three methods are given in this table. The maximum difference between Wijs and Hub1 is 2, the Wijs method always giving higher results. On the thirty-six samples the Wijs method gave an average of 1.2numbers higher. The Hanus method gave a maximum of 1.4numbers higher than Hiibl, with an average of 0.59 number higher. This gives the Hanus solution a very slight advantage over the Wijs and this difference seems to hold good on all the oils-the Hanus being slightly nearer the Hubl than is the Wijs.
I O D I N E A B S O R P T I O S O F 011,s A N D FATS.
2.50
TABLEYI,-IODIXE SCJIBEKS BY DIFFERENTMETHODS Differe lice
between
Variety of oil.
Cocoanut.. Butter .... Lard P e a n u t . . ..
......
'( I'
.... ....
Rape ......
......
I'
......
z t
Sunflower. Sesame ... '( ... Corn ...... Poppy ....
....
....
Linseed
...
. . . . ... I
Hiibl'i number
Wijs's number
.... .... ....
9.03 30.7 56.4
SS.3
....
s7.2
Qj.2
91.8
93.4
Hanus'a number.
9.03 ~30.6
56.9
ss.4 9: .6 9s.s
102.9
'03.3
....
103.0
102. I
101.9
7.8 I10 3
i
II
.... ....
19.0
I I I.;
107.1
....
1q.S
128.1
....
....
.... ....
IO7.j
....
Difference between Hub1 and Hanus.
hnaly.it.
0.0
I-Ianus.
--0. I
~.0.5 --0,
I
0.0
....
- 1.6
0.2
...
...
99.3
Hiihl and U'i j+
....
0.5
....
- 0.4
1.1
u.9 '
....
1.2
....
1.;
.... 3.6
0.4
2
....
....
:22.6
13-1.6
....
l j j 2
.... ....
lI9.h 174.8 1S0.9
1196 177.3
....
0.0
....
174.5
-2. j
a.3
122..1
170.2
....
IS2.1
....
Ij L
.o
--1.2
....
-.
li.2
0
Hailus.
\Vijs.I H an 11s .
6
.... '
\Vijs. Hunt.' Hanus. St'i j s, I H u 11t .? \Vi j 4 , I
'3.5
\i'ija.' Ill1 I l t , y
\Vijs, Ilanus.
These figures show practically the same results as the previous tables although not quite so wide a variation, which may be somewhat accounted for by the fact that four hours lvere met1 in the Hub1 determinations, which gives a little higher figure than the three hours on the higher-absorbing oils. These results shoiv vc.:.p little choice between the two proposed substitutes for the EJCiliI method. The Hanus gives results a little closer to the I-iiibl figures but the difference is of practically no consequence. Its chief advantage over the \T;ijs is in the ease of preparatioii. Iodine trichloride is a rather rare reagent and the preparation or' the iodine chloride by passing chlorine gas into the acetic acid soiution of iodine until the halogen content is doubled is a some\vhat tedious operation compared with the addition of bromine. On this account we favor the Hanus solution. \Vith this change of method, a new set of figures will have to be obtained for the semidrying and drying oils, and such oils as mustard and rape with high iodine absorption. F o r the butters. lard and olive oil the Hanus method compares satisfactorily with the Hiibl. The stability of the solution does not lessen the need 1 2
Seven minutes used in determination One hour w e d in determination.
DOES CHOLESTEROL OCCUR I N N A I Z E O I L
?
251
of blanks for each set of determinations. The high coefficient of expansion of glacial acetic acid, 0.00115 for I' C., makes an appreciable error if a slight change of temperature takes place, as will be seen from the following table. TABLEVII.
.
40 cc. iodine monobromide in Niio thiosulphate.
16.0
92.05
17.5
91.85
18.5
91.80 91.35
Temperature. OC
21.5
24.5
g1.10
27.0
go.80
I IO
1.25 cc. decrease.
change
A change of t I O C. gives a change of 0.I I cc. in the amount of N / I O thiosulphate necessary to neutralize 40 cc. of the iodine solution, so that a blank titration of the iodine solution could easily vary I cc. N/IO thiosulphate in a day, which would completely vitiate the results.
DOES CHOLESTEROL OCCUR IN n A I Z E OIL? Br AUGUSTUS H.GILLA N D CHARLES G.TUFTS. Received December 3'. 1902.
ACCORDING to Hoppe-Seylerl and to Hopkins,* cholesterol occurs in maize oil. As there is evidence of the Occurrence of this substance in no other vegetable oil, except olive oil, this is interesting, and of possible importance as a means of detecting maize oil in mixtures with other oils. The statement of Hoppe-Seyler was made before this group of bodies was clearly differentiated, however, and the melting-point ascribed by Hopkins to the alcohol found by him in maize oil is not the true melting-point of cholesterol. The difference in the melting-points is shown below. Cholesterol, 146"- 147' ;3 "cholesterol",from maize oil, 137'-1 37.5 .3 Hopkins used no other means of identification except the color BUN.SOC.Chim 121, 6,342 (1866); Medicin.-chem. Unlevsuch., I, 162. This Journal, ao, 948 (1898). a Wislicenus and Moldenhauer : A n n . Chcm. (Liebig), 146, 179 (m. p. 147'); Reinitzer: Monalsh. Chem., 9, 422 (m p. 147.5') ; Hesse : A n n . Chrm. (Liebig), 192,177 (1878)(m.p. 145~-146'); Salkowski : Zlschr. ana2. Chem.. a6, 567 (m. p. 146'); Bomer: Ztschr.. Unfrrsuch. Nuhr. 14. Genus. (IS@), p. 81 (averagem. p. of fifty-two samples, 146.4'-147.30). 1 2
