lysed, there was a very small wave a t a potential of approximately -0.98 volt (S.C.E.). Because subsequent runs with pure EtaPbCl resulted in well defined waves a t -0.98 volt, it is concluded that EtrPbCl was a decomposition product of HED. To establish the approximate rate of decomposition of HED, samples were removed periodically from a volumetric flask containing 6.7 X lO-4M H E D and polarographed. The flask was maintained a t 25” C. and exposed to normal laboratory lighting. The decomposition reached 2.9% in 30 minutes, 4.4% in 60 minutes, and 20.8y0 after 3 days. As EtaPbCl is a common impurity in H E D samples, several calibration runs were made on this material. The results (Table I) indicate that samples containing H E D can be analyzed simultaneously for EtaPbC1. Oxygen is a
common contaminant and is reduced in the voltage range corresponding to the El,* of EtaPbC1. Care must be exercised to remove oxygen completely for this determination. The second wave for EtaPbCl exhibited a maximum which could be s u p pressed by Triton X-100. No use was made of a suppressor in the measurements for EtsPbCl as the prewave a t - 0.98 volt could be measured without interference from the maximum. Various supporting electrolytes were tested in the H E D procedure, including lithium nitrate, bromide, sulfate, iodide, and sodium sulfate. Solubility limitations prevented consideration of certain other common salts. All of the electrolytes tested with added H E D resulted in the usual anodic wave for H E D , but none gave the ideal wave obtained for the lithium chloride electrolyte.
ACKNOWLEDGMENT
The authors are indebted to the Ethyl Corp., Detroit, Mich., for support of this work. LITERATURE CITED
(1) Calingaert, G., Dykstra, F. J., Shapiro, H., J . A m . Chem. SOC.67, 190 11945). (2j-Calhgaert, G., Soroos, H., J. Org. Chem. 2, 535 (1938). (3) Closson, R. D., Ethyl Corp., Ferndale 20, Detroit, Mich.,. private communica-
tion. (4) Kolthoff, I. hl., Lingane, J. J., “Polarography,” 2nd ed., Interscience, New York, 1952. (5) Page, J. A., Wilkinson, G., J . A m . Chem. SOC.74, 6149 (1952). (6) Waazonek, S., ANAL.CHEM.24, 32 (1952); 26, 65 (1954); 28, 638 (1956); 30,661 (1958). RECEIVEDfor review May 19, 1959. Accepted August 4, 1959.
Paper Chromatography of 2,4-Dinitrophenylhydrazones Estimation of 2-Alkanone, n-AlkanaI, Alk-2-ena1, and AI k-2,4-d ienaI Derivatives REX ELLIS and A. M. GADDIS Meat laboratory, Eastern Utilization Research a d Development Division, Agriculfural Research Service, U. S. Deparfment of Agriculture, Belfsville, Md.
b The application of paper chromatographic methods of separating 2alkanone, n-alkanal, alk-2-enal, and a I k- 2 , 4 d i en a I 2,4-dinitrophenylhydrazones into .individual compounds has been examined. In experiments with four mixtures the mean recoveries were consistent and showed small variations. In over-all recovery, average deviation from the mean was &3.4%. Recovery decreased from the 2-alkanones to the alk-2,4-dienals because of differences in stability of the classes. However, this was reflected in but small error in proportions of classes and the ratio of individual compounds found.
-
R
publications have described rapid paper chromatographic methods of separating mixtures of 2-alkanone, n-alkanal, alk-2-ena1, and alk2,4-dienal 2,4-dinitrophenylhydrazones into classes ( 2 , 4 , 6 ) and each class into individual compounds (1-4). These methods have been applied to the determination of changes in the proportions of steam-volatile monocarbonyl classes with the autoxidation of pork fat (6), ECENT
and to the identification of such compounds volatilized from a rancid pork fat (3). In the latter study, tentative quantitative data were reported for the classes and individual compounds. Aside from the inevitable mechanical losses in the manipulations, it was recognized that the most serious obstacle to quantitative application was the variation in stability of the different classes (2). The alk-2,4-dienal derivatives are particularly sensitive to light and air (8,6). This work was undertaken to determine the quantitative capabilities of the methods. EXPERIMENTAL
Solvents, reagents, materials, and equipment were the same as used in similar operations described in earlier papers (1-6). Authentic monocarbonyl 2,4-dinitrophenylhydrazones employed have been described (1, 8). Stock solutions of each hydrazone in carbon tetrachloride were prepared containing the equivalent of approximately 25 mg. per liter (25 y per ml.). Suitable volumes were taken from each stock solution to make up 100ml. solutions of 30 kmoles per liter con-
centration. Aliquots of the various solutions were used to prepare the mixtures used in the experiments. Separation into classes (8, 6) and resolution of the classes into individual compounds (1-5) were performed as described in earlier papers. Proportions of classes (2, 6) could be determined from spots extracted from three paper strips and measured spectrophotometrically in 3.00 ml. of carbon tetrachloride. However, it was necessary to accumulate a number of paper strips from the class separation to provide sufficient material for estimation of the individual compounds (3). Spectrophotometric measurements of absorbance were usually made a t the wave length of maximum absorption in 3.00 ml. of carbon tetrachloride. However, it was necessary to measure the spectra of compounds separated on vaselineimpregnated paper in alcoholic alkali (S), because of the presence of a persistent impurity (1, 3) that absorbed at the lower wave lengths. Every practical effort was made to protect the carbonyl derivatives from the effects of light and air. Spotted paper strips were chromatographed immediately, and separated spots extracted and measured as quickly as possible. Solvent was removed from VOL. 31, N O . 12, DECEMBER 1959
1997
extracts on the steam bath with a jet of nitrogen.
values at'X,.,. of the four classes was 0.778.) I n the class separation, 2.15 ml. of the mixture (2,s)were used to
RESULTS AND DISCUSSION
spot each paper strip. I n the separa,tion into individual compounds, 54 class separation paper strips were used for the 2-alkanones and n-alkanals, 42 strips for the alk-2-enals, and 36 strips for the alk-2,4dienal class. Results were calculated on the basis of 100 ml. of mixture and are shown in Table I. In the class separation] the proportion of alk-2,4-dienals was somewhat low and those of the other classes were a little high, because of the much lower recovery of the alk-2,4-dienal class I n the separation into individual compounds, recovery was higher and more uniform. Possibly the sensitive compounds were protected by the impregnating agents. Proportions were not greatly changed by the separation into individual compounds. However, unsaturated classes were low and saturated classes high. This was reflected in over-all recovery which decreased with degree of un-
The standard solutions containing 3 pmoles of 2,4dinitrophenylhydrazones per 100 ml. were found to have the following average absorbance values at their wave length of maximum absorption: Zalkanones 0.615, n-alkanals 0.615, alk-2-enals 0.828, and alk-2,4 dienals 1.12 These values were used as the basis for calculation of results. Mixture 1. Equal volumes of 2-alkanones Cd to CO,C1,, and C13, n-alkanals Ca to (214, alk-Zenals C3 to CIZ, and alk-2,4dienals C6 to Clz were combined. This mixture of 38 compounds contained 0.0789 pmole of each compound in 100 ml. of solution. It included only those members of the four homologous series that separate in their proper classes. The solution had an absorbance value a t of 0.70. (The sum of calculated absorbance
Table 1.
Analysis of 2,4-Dinitrophenylhydrazone Mixture 1
Class Separation
% Based on Absorbance pmole PresPresyo ent Found ent Found ,ecovery at
xm,,
ZAlkanones, C4-Cs,Cli, Ct3 16.65 17.68 0.632 n-klkanals, C3-C*4 24.98 25.43 0.947 Alk-2-enals, c3-c12 28.03 29.04 0.789 Alk-2,4dienals, 30.31 27.85 0.632 Ca-C,2 Total 3.000 Average
saturation. Average deviation from the mean in the over-all recovery was an acceptable f3.0%. Mixture 2. This mixture was made u p of equal volumes of 2-alkanones CI to CS,CllJ and CI3, n-alkanals C1 to CI4, alk-2-enals Cs t o C12,and alk2,4-dienals C6 t o Clz. The resulting solution contained 41 compounds of equal concentration (3.00 pmoles per 100 ml.). This mixture differed from Mixture 1 because 2-alkanone Cs separates in the n-alkanal class, and n-alkanals C1 and Cz separate Kith the alk2-enal class (8). In separation into individual compounds, this results in inseparable mixtures of Zalkanone CJ with n-alkanal Caand n-alkanal CZwith alk-2-enal C3 It was necessary to study this separation to determine the influence of the inseparable groups on quantitative determinations. The resulting data are shown in Table I1 for 100 ml. of mixture. Results were very similar to those obtained with Mixture 1 for the separable compounds.
Proportions, pmole 70 Present Found
Individual Separation pmole @%!, u
Av. propordev. 5; tions Found frommean recovery found
Over-all Recovery Av. dev. from % mean
0.53
84.0
21.0
22.0
0.47
fO.0018
87.9
22.7
73 9
12.3
0.77
81.3
31.6
31.9
0.67
fO.0017
87.5
32.4
71.2
f 3 2
0.65
82.4
26.4
27.0
0.53
f0.0026
81.1
25.6
66.8
1 3 3
0.46 -
72.9
21.0
19 1
0.40 - 10.002G 2.07 f0.0022
86.5
19.3
63.1
f8 2
68.8
f3.0
2.41
Table II.
80.2
85.6
Analysis of 2,4-Dinitrophenylhydrazone Mixture 2
Class Separation
Proportions -4s Separation Actual Goes pmole yo pres- PresPresent Found recovery ent ent Found ~
~
Individual Separation pmole Over-all pmole .4v. % av. dev. dev. 70 from Total from so propor- % found mean recovery tions recovery mean
2-Alkanones,
c4-cs,
C11,
0.586 0.51
C13
2-Alkalione C3 0.0'732) n-Alkanal C8 n-Alkanals C4-C14
0.07321 0.951 0.77 0.805
1
87'0\
21.9
80.9
34.]
0.877 0.69
1998
0 658
0.43
f0.003
84.3
21.2
87.0
33.2j
I
31.6
1
78.61
Alk-2-enal C3 Alk-2-ennls C4-c12 Alk-2,4dienals C5-C12 Total .4vernge
21.5 32.5
0.58
n-Alkannl C, n-Alkanal C2
19.6
1
f4.7
61.5 72.0
320.002
f2.8
54.6
0.04
29.2
73.3
82.6
29.1 1 0 . 1 0
28 2
168.1
24.41
1
0.43 0.586 0.40 3.00 2.37
ANALYTICAL CHEMISTRY
68.3 78.7
19.6
19.6
16.9
3=0.003)
0.35 =!=0.003 2.02
f0.003
87.5 85.4
17.4
65.4
f4.6
59.7
*4.3
64.9
f1.l
Analysis of 2,4-Dinitrophenylhydrazone Mixture 3
Table 111.
StandComposition Mixture, ard Solupmoles tions IndiMax. vidual Absorb- comAs ance pounds classes
n-Alkanals
G
CS
C, c 1
1.23 0.060' 34.29
0.030,
co
1.04 0.045 1.35 0.049 0.83 0.030,
CII c 7
CO
ClO
GI
79.2
.
II
77.2
0.216
9.8
0.185
0.162 0.009
0.20
0.14
75.7
8.4
11.8.
8.4
11.6
8.1
-
Total Average
2.20
1.73
77.4
Over-all pmole
Individual Separation yo Proportions yo Indirevidual cov- pres70 pmoles ery ent foulid 0.039'
79.8, 80.3
I
0.23 0.006 0.28 0.008 6.04 0.35
1.39
0.043
'
1.18 0.043
Alk-2,4-dienals
Found
81.81
1.35 0.049'
CS ClO
%
Recovery
