BACONF. CHOW
804 [CONTRIBUTION FROM
THE
VOl. 56
CONVERSE MEMORIAL LABORATORY, HARVARD UNIVERSITY ]
The Absorption of Oxygen by Linseed Oil in the Presence of Catalysts of Different Oxidation-Reduction Potentials BY BACONF. CHOW
A. Introduction Chow a.nd Kamerlingl recently suggested that in the absence of other catalysts the rate of oxygen absorption of oleic acid catalyzed by ferricyanide is apparently independent of the chemical structures of the catalyst but dependent on its oxidation-reduction potential Eo. In this paper an attempt was made to study quantitatively the various factors governing the oxygen absorption of linseed oil as catalyzed by ferri-, tungsti- and molybdicyanides. It was found that the r:zte is dependent not only on concentrations of hydroxyl ion and the oxidized form of the catalyst but is inversely proportional to the reduced form. An empirical equation was formulated to account for the results.
C. The Empirical Formula For the sake of the convenience of presentation, the empirical equation is given a t this point. It was found that for one catalyst, the rate of oxygen absorption was given by
where k' is the rate constant for a particular catalyst, (A) and (A-) are the concentrations of the catalyst in the oxidized and reduced states, respectively. Thus a t constant (A)2/(A-) ratio, the rate is given by -dOn/dt
(2)
= k"[(OH-)]'/3
On the other hand, if the hydroxyl-ion concentration is kept constant by a buffer, we have (3)
B. Experimental Procedure D. Experimental Results The absorption of oxygen was measured with Edect of Hydrogen-Ion Concentration.-Measthe conventional Warburg2 absorption apparatus, following the usual procedure, a t a temperature urements of the rate of oxygen absorption were of 25.0 =t 0.1'. The charge consisted of 0.1 made in buffers of different PH's,keeping the term cc. of raw linseed oil and 1.0 cc. of catalysts (A)a/(A-) constant. This was done by adding of specified concentrations in buffers, and 1.2 cc. solid reduced catalyst equivalent to 5% of the of the same buffer. TABLEI1 The composition of the buffers, made up with EFFECTOF HYDROGEN-ION CONCENTRATION water to 50 cc., and the PH'Sof the resulting solu[(A)z/(A-) J1" = 0.56 --doz/di,CC. x 10-3 tions were as follows. per hour PH k" (calcd.) X 10: TABLE I
IGFe(CN6
Mixture 2 5 . 0 ~of~ 0. . 8 2 4 M KsHPOh 2 1 . 6 cc. of 1 . 6 5 N NaOH 12.25 cc. of 1 . 6 5 N NaOH 2 5 . 0 cc. of 0 , 8 2 4 M KzHPOl 16.7 cc. of 0 , 3 3 M KOH 3 3 . 3 cc. of 0 . 3 3 M KzHPOd 30 cc. of borate solutiona 20 cc. 0.1 N NaOH 0 . 1 M Naz13407 2 4 . 2 CC. 1 N NaOH 2 5 . 8 CC. 1 M KHzPOc
+ + + +
+
PH
12.0 11.6 10.8 10.0 9.2 8.0
+ 100 cc. N NaOH.
8.0 4.0 2.1 1.3 0.57 .29
12.4 11.6 10.8 10.0 9.1 8.0
Borate solution: 12.4 g. H3B03
Only one sample of raw linseed oil has been used in this report because the purpose of this paper is to study the relative and not the absolute rates of oxygen absorption under different experimental conditions. Furthermore, the concentration of linseed oil was kept constant by using a saturated solution of linseed oil and the partial pressure of oxygen was maintained a t the atmospheric condition. (1) Chow and Kamerling, J . Biol. Che7rr., 104, 69 (1934). ('2) Warburg, Aimhem. Z.,161. 31 (1924).
49 45 44 50 45 52
Av. k" = 48
K,W (CN)B 6.0 1 43
0.56
10.8 9.1 8.0
125 111 101
-
Av. k' = 112 K,Mo( CN)s 24.0 5.8 2.3
10.8 9.1 8.0
500
450 410
-
Av. K" = 453
CATALYTIC ABSORPTIONOF OXYGENBY LINSEED OIL
April, 1934
oxidized form. Since in general much less than 5% of (A-;I is produced, the ratio is but slightly changed and its cube root still less. In this way a straight line is obtained during the first part of the experiment Qhose slope gives the rate of absorption. ’The results for potassium ferri-, tungsti- and molybdicyanide in different buffers are collected in Table 11. Typical data for the rate of oxygen absorption by linseed oil catalyzed by potassium ferricyanide in different buffers are shown in Fig. 1 for purposes of illustration. Effect of the Concentration of Catalyst.A series oi experiments using ferricyanide as catalyst whose concentration varied from 0.91 to 0.00091 ,I1was performed. In each experiment, 5% of ferrocyanide was present. The results are given i n Table 111. TABLE I11 DETERMINE T H E E F F E C T OF VARYING CONCENTRATIONS OF THE CATALYST, FERRICYANIDE, I N P H 10 8 The rate of oxygen absorption is in mml. X lo3per hour.
TU
- dOz/dl 10.2 8 9 4.9 2 1 0.72
k”’
(A)
0 1 x 10-1 4.6 X lo-’ 9.1 x 10-2 3 1 x 10-3 9 . 1 x 10-4
x
io?
3.9 4.3 4.1 3.8 2.8 Av. k“’ = 3 8
The average value of k”’ is 3.8 X from which k’ can be calculated by the equations (1) and (3). This gives k’ = 4.4 X lo-*. Effect of the Ratios of Ferricyanide to Ferrocyanide.-A series of experiments was carried out by adding varying amounts of ferrocyanide to 0,0091 M ferricyanide, in the phosphate buffer PH 10.8. By virtue of equation (3), k”’ can be calculated. The results are given in Table IV.
895
Effect of the Variation both in the Ratio of ( A ) 2 / ( A-) and in their Concentrations.-A series of experiments was carried out with different
2
6 8 10 12 14 Hours. Fig. 1.-Showing the rate of oxygen absorption by linseed oil catalyzed by KaFe(CN)e in different buffers: Curve 1, PH = 12.4; Curve 2, PH = 11.6; Curve 3, PH = 10.8; Curve 4, PH = 10.0; Curve5, PH = 9.1; Curve 6, PH = 8.0. 4
concentrations of both ferricyanide and ferrocyanide. The results are recorded in Table V. The validity of the empirical equation ( I ) can best be verified by the constancy of k’. In TABLE V TO DETERMINE T H E EFFECT O F VARYING T H E CUNCENTRATIONS OF K3Fe(CN)6 AND K4Fe(CN)6 The rate of oxygen absorption is in mml. X los per hour. K4Fe(CN)e (a) Total concentration of K3Fe(CN)e 0.0091 M
+
-dOn/df
1.7 1.08 0.76 27
TABLE I\’
DETERMIKE T H E E F F E C T O F T H E ADDITIONO F F E R R O TO 0.0091 M FERRICYANIDE AT PH 10.8 The rate of oxygen absorption is in mml. X lo3per hour.
TO
(A) i n molarity
8.6 X loW3 6.8 X 4.6 X 2.3 X
(A -) i n molanty
4.5 2.3 4.6 6.8
k“’ X 103 3.1
X lo-‘ X low3
4 .0 4.5 3.0
X X
CYANIDE
-dOzjdt
1.90 1.34 1.21 0.67 .55
(A3 in molarity
4.6 X 2.3 x ~1.6 x (5.8x (3 7 x
Av. k”’
k”’ X 10’
lo-’ 10-3
10-3 10-3 10-3
Av. k”‘
The average value of k’” is 3.5 X which we get k’ = 4.1 X
=
(b)
3.3 4.1 4.6 2.9 2 6 3.5
from
=
3.7
Therefore k‘ = 4 . 3 X
+
Total concentration of K3Fe(CN)~ 0.0455 M 4.1 x 1 0 - 2 4.6 x 10-3 1.8 3.4 X 1 . 1 X lo-$ 1.5 2.3 X 2.3 X 1 1 1.1 X 3 . 4 X lo-’ 0.58 .27 2.3 x 10-3 4 . 3 x 10-2
KdFe(CN)e
2.5 3.3 3.9 3 .9
5.3 -
Av. k”’ = 3 8 k’ = 4.4 X lo-’
BACONF. CHOW
896 TABLE V (Concluded) (c) Total concentration of &Fe(CN)e 0.218 M
- dOl/df 3.4 3.7 2.2 0.72 .3t
( A 7 in
(A)
10-1 10-1 10-1 10-2
KIFe(CN)' k"'
molarity
molarity
2.1 x 1.13 X 1.1 x 5.5 X 1.1 x
+
1 . 1 x 10-2 5 . 5 X lo-* 1 . 1 x 10-1 1 . 6 X 10-1 2 . 1 x 10-1
X 10s
2.2 3.4 4.6 2.7 3.8
-
Av. k"' = 3 . 4 k' = 3 . 9 X loF2 (d)
Total concentration of K8Fe(CN)a 0.455 M 8 9 1 . 3 x 10-1 2 . 2 x 10-2 5.0 3 . 9 X lo-' 6.8 X 1 . 1 x lo-' 3.8 3 . 4 x 10-1 3.1 2 . 3 X lo-* 2 . 3 X lo-'
+
KdFe(CN)e 4.4 3.9 3.7 5.0
-
Av. k"' = 4 . 3 k' = 5 . 0 X lo-*
Table VI: the average k' values obtained by different methods are tabulated in the second column. TABLE VI Method
By the variittion of PHusing KaFe(CN)sas catalyst By the variation of concentrations of &Fe(CN)$ By the variation of the ratio of KsFe(CN)&/K1Fe(CN)5 By the variation of both the ratio and concentration of KsFe(CN)6 and ICFe(CN)e
4.8 4.4 4.1 4.4
Equation (1) shows that the rate of oxygen absorption is proportional to the product of concentration of the catalyst and the ratio of the oxidized form to the reduced. The appearance of the concentration of the reducing agent in the denominator of the rate expression makes it plausible that a preliminary oxidation-reduction equi1ibriu.m is reached. If we assume that the rate is proportional to e(nF'RT)Ea, then equation (3) can be transformed into - doz = k [ e ( n F / R T ) E o (A)eWA)/(A-)]'/z (4) dt
where k"' = k[e(nF'RT)Bo]i/g, n, F, R, T have their usual meanings. Collecting the exponents of e and substituting the .value of E for Eo
-
dO2dt
+ RT 1(' we get nF In (A-)'
- ko [ ( A ) & F / R T ) E J ' / ~
referred. Thus to test the assumption, it is best to compare the experimental ratios of rates of two catalysts with the calculated value. Thus a t constant hydroxyl ion concentration and (A)2/(A-) ratio, we can rewrite equation (4) as follows
where subscripts T and F denote tungsticyanide and ferricyanide, respectively. Similarly, the ratios of the rates of oxygen absorption catalyzed by molybdicyanide and ferricyanide can be calculated. The cases of methylene blue and indophenol are complicated by the fact that the concentrations of the reduced forms are not known. However, the experimental results of both dyes are of the right order of magnitude and direction (see Fig. 2). TABLE VI1 THERELATION BETWEEN THE RATE AND EO The rate of oxygen absorption is in mml. X lo3pet hour. Catalyst
k' X 102
(5)
Experimental verification of equation (4) is complicated by the arbitrary standard, the "normal" hydrogen electrode to which all Eo's are
Vol. 56
KdFe(CN)$ K&W(CN)a &Mo(CN)s
-dOt/dt
44 125 500
Ratio of rates Of
[,,(nF/RT)(Ep- EoF))1/,
2.8 11.4
1 2.5 38
EO,volt absorption 1 0.45
.53 .72
Table VI1 shows a rough correlation between Eo and the rate. The choice of EO given in Table VI1 needs a slight comment, since EOof the polyvalent ions is greatly affected by the salt concentration. According to Conant and Pratt's3 measurement the average values of Eo for molybdicyanide and tungsticyanide a t the range of PH between 8 and 11, are +0.72 v. and $0.53 v., respectively. Collenberg4 gave the normal potential of molybdicyanide and tungsticyanide as +OS39 v. and +0.569 v., when referred to the normal hydrogen electrode. Both values of Collenberg are higher than those of Conant and Pratt. However, the latter's values are used because their experimental conditions are more comparable with those reported in this paper. E. Inhibitors It is to be pointed out that inhibitors must not reduce the catalysts appreciably and the ones, such as hydroquinone, etc., which are very powerful in inhibiting the other chain reactions cannot be used in this reaction. I n spite of this limitation the following compounds show marked inhibiting effects. ( 8 ) Conant and Pratt, THIS JOURNAL, 48, 3239 (1926). (4) Collenberg, Z. physik. Chcm.. 109, 353 (1924).
CATALYTIC AE+SORPTION OF OXYGEN BY LINSEED OIL
April, 1924
TO
TABLE VIII DETERMINE THE EFFECT OF INHIBITORS SOLYJTION OF KsFe(CN),, Absorption of 0 2 in mml. X Names
None Acetoxime a-Aminoiso-
butyric acid dl-Serine Arginine Histidine Cinnamic acid
Concn. of inhibitors in mol.rrity
i
4.5 X 0'0° 4 . 5 >< 4.5 X 4.5 X 4.5 X 4 . 5 >: 4 5 >: 4 . 5 >: f 9 . 1 >: < I\ 9 . 1 >:
10-4 10-5 10-4 10-5 10-4 10-4 10-4 10-4 10-4 10-5
5
10
0.0046 &f
IN
2 hrs.
hrs.
hrs.
hrs.
4.4 1.1 1.7 2.4 2.8 1.6 1.8 1.3 2.2 2.2 3.6
9.1 2.9 4.2 5.6 6.3 3.9 4.5 3.8 5.5 4.9 8.5
17.0 6.3 7.6 10.7 12.5 7.8 9.4 8.1 11.2 9.8 18.5
23.8 12.1 16.1 21.4 23.0 16.5 18.8 15.2 21.4 19.2
20
-dOsidl in mml X101 per hr. 2.2 0.62 .81 1.1 1.3 0.78 .93 ,811 1.14 1.09 1.7
89 7
which is a measure of the free energy. The importance of the relationship, if true, is obvious. However, it must be pointed out that although equation (1) can be formulated from equation ( 5 ) , it must not be taken as a proof for the correctness of the assumption. In a chain reaction the concentration of an inhibitor' also appears in the denominator. I t is possible that the reduced form of the catalyst may function in the chain breaking process.
G. The Secondary Oxygen Absorption I t was observed in the experiments with potasEthanolamine ., sium molybdicyanide and tungsticyanide in the A series of the iderivatives of aniline and phenol alkaline buffer of PH9 or greater that the oxygen has been tried, e. g., #- or m-nitroaniline, #-amino- absorption would occur a t the normal rates benzoic or #-sulfanilic acid; #- or o-carboxyl- only to be followed by a tremendous increase. An experiment was performed to follow the phenol. The rate of oxygen absorption was cut down 40%, under the same experimental conditions. From Table VIII, acetoxime is a fairly powerful inhibitor. The presmole of acetoxime cuts % 20 ence of 5 x the rate to half. All the amino acids show definite inhibition, though there is no apparent relation between their structures and power of inhibition. Ethanolamine is not a very powerful o l 0 inhibitor. I t is interesting that cinnamic acid which possesses a conju- 2 gated double bond apparently reduced the rate considerably. Lastly, it is to be mentioned th.at dimethylaniline, 10 20 30 40 which is a good inhibitor for the oxygen Hours. absorption by oleic acid, is a catalyst for the oxygen absorption by linseed Fig. 2.-Showing the rate of axygen absorption by linseed oil with catalysts of different oxidation-reduction potentials: 1, molybdicyanide; Oil. This is prob'tbly due to the fact 2, tungsticyanide; 3, ferricyanide; 4, indophenol; 5, methylene blue that dimethylaniline being more soluble in linseed oil than in water goes into the linseed oil rates of oxygen absorption and of the reduction of phase. Francke5 has shown that dimethylaniline molybdicyanide. The results showed that the normal rate of oxygen absorption as required by is a good catalyst for the oxygen absorption. equation (1) continued until about 80% of the P. Discussion of Results molybdicyanide was used up. Then the rate of Equation (5) shows two peculiarities. First, oxygen absorption gradually increased to a maxithe cubic root in the kinetic equation is justified mum. An electrometric titration with ferroonly by the results. Second, the rate is pro- cyanide showed no presence of molybdicyanide at portional not only to the concentration of the this point. The oxygen absorption continued for catalyst but also to the potential of the system,6 several hours a t the maximum rate and gradually ( 5 ) Francke, A n n , ISIS, 129 (1932). decreased to zero in not less than thirty hours. (6) Studies on the relation between the rate and the potential of It is not clear just why the absorption took homogeneous sokutions .are in progress Preliminary experiments on the nitrogen evolution of hydrazine in the presence of an oxidizing place after molybdicyanide was used. Howagent show that the rate i s also related to the potential of the reagent y.
such as K3Fe(CN)s, KaU'(CN)e. etc
(7) Jeu and Alyea, THISJ O U X N A L , 55, 575 (1933).
898
ARTHURW. THOMAS AND
JAMES
ever, the following experiments point to the possibility of peroxide formation. (1) This secondary absorption occurred most readily only in the alkaline solutions where molybdicyanide is metastable. In PH 7.0 only the normal oxygen absorption was obtained within twenty hours. (2) Preliminary experiments showed that linseed oil absorbed oxygen a t a great rate in the presence of hydrogen peroxide and peroxidase. The “complex” is greatly deactivated by heat. A sample which showed a rate of absorption a t the maximum was heated in nitrogen for half an hour over the ‘steam-bath, whereupon it absorbed only about 1/300 as fast. The author is greatly indebted to Prof. James
[CONTRIBUTION FROM
THE
CLAUDE THOMSON
VOl. 56
B. Conant for invaluable suggestions and to Prof. G. B. Kistiakowsky for reading and criticizing the manuscript. H. Summary 1. An empirical equation relating the rate of oxygen absorption of linseed oil to the potential of the catalysts has been obtained. 2 . A series of inhibitors has been found. Acetoxime is the most powerful one. The presmole of acetoxime will lower the ence of 5 x rate to about one-half. 3. A “secondary” oxygen absorption has been found when molybdicyanide or tungsticyanide were used as catalysts in PH9 or greater. RECEIVED NOVEMBER 1, 1933
DEPARTMENT OF CHEMISTRY, COLUMBIA UNIVERSITY ]
The Preparation of Pure Eleostearic Acids from Chinese Wood Oil BY ARTHURW. THOMAS AND JAMES CLAUDE THOMSON The following prescriptions are proposed for the isolation of eleostearic acids from tung o i l , ~ lighted the ‘perations being performed in a room using a ‘loth to the glass apparatus. *Eleostearic Acid Siponify 25 g. of tung oil with 100 cc. of 10% solution of potassium hydroxide in alcohol by gentle boiling under a reflux condenser for one hour. Remove the condenser and insert in the flask a three-holed rubber stopper containing a tube for entry of carbon dioxide gas, a siphoning tub
