V O L U M E 27, N O . 2, F E B R U A R Y 1 9 5 5 apparatus. Only two ruptures of the port have occurred during a period of burning more than 350 samples. No harm to the operator or damage to t,he equipment was experienced. In the author's opinion, rupture of the port will occur only if the vacuum pump fails to operate or if the flame is extinguished and the chimney fills with combustible gases before the sample assemhly is lowered. For these reasons it is recommended that the operator attend the apparatus throughout the time needed to burn tlie sample. Operating the apparatus under the conditions cited in the procedure, the rubber gasket \rill function properly for 7 to 8 days. If the rate of sample flow is increased appreciably the gasket may need to be replaced dad!.. Under more severe burning conditions a g;isket per sample may he required. Examination of the per cent deviation colunin in Tatilcs I and I1 shows that the majority of the deviations are positive with respect to the amount of sulfur added. This is attributed to the presence of sulfur in the rubber gauket. Gravimetric analysis of sulfur in the rubber gasket indicated niorr than 2y0present. Since the conclusion of t,hr work reported here the apparatus has been modified by sealing the burner in a 4 6 / b ~ f outer glass joint and connecting it to the chimney through a ""/o f inner joint. This modificat,ion not only improved the repeatability of the results but eliminated the possibility of sulfur contaniination from a ruhlier ~ r a l . The sample of drip oil containing
269 0.058% sulfur (Table 11)showed deviations from 0.000 to 0.004% with the rubber gasket and 0.000 to 0.001OjO with the f glass joint. The extension of the method to tlie determination of halogens in petroleum liquids is obvious. Limited work has been done on the determination of organically hound chlorine in drip oils. Results are comparable to those reported for sulfur. ACKNOWLEDG4ZENT
The author wishes to express his thanks to J. A. Grant for suggesting the problem, to R. L. Aiken for designing the air inlet chamber and offering helpful suggestions throughout the course of the investigation, to K. C. Sichols and Mildred Gibson for operating the apparatus, and to Pan ilmerican Refining Corp. for permission to publish this work. LITERATURE CITED
Am. Soc. Testing Materials, Philadelphia, Pa., "Standards on Petroleum Products and Lubricants," p. 648, 1948. (2) Ibid., p. 15, 1952. (3) Ibid., p. 77, 1952. (4) Institute of Petroleum, London, England. "Sulfur by the Lamp Method (Using Fast Burning Lamp)," I.P. 107/45 (tentative). ( 5 ) Lane, W. H., ANAL.CHEM.,20, 1045 (1948). (6) Wear, G. E. C., and Quiram, E. R., Ibid., 21, 721 (1949). (1)
RECEIVED for review August 16, 1954. Accepted November 8, 1954.
Indirect Polarographic Determination of Propylenediamine A. D. HORTON, P. F. THOMASON, and M. 1. KELLEY Analytical Chemistry Division, O a k Ridge National Laboratory, O a k Ridge, lenn.
Propylenediamine has been determined indirectly by polarographic estimation of the propylenediaminecopper complex in a supporting electrolyte of 0.1M potassium chloride. Propylenediamine in concentrations of 26 to 97 y per ml. can be determined with an over-all relative standard deviation of 0.70%. The method is recommended for the estimation of microgram amounts of propylenediamine or other chelating amines in solutions that do not contain other substances that will form complexes with copper or that will reduce at half-w-ave potentials within 0.2 volt of the half-wave potential of the propylenediamine-copper complex, which varies slightly with variation in propylenediamine concentration.
P
OL.4ROGRAPHIC studies of the propylenediamine-copper complex have been made by Calvin and Bailes ( 1 ) to determine the stability of chelate compounds and by Carlson and coworkers ( 2 ) to determine the equilibrium constant of amine complexes with copper ions. In these studies copper was added in small aniounts to aniounts of propylenediamine that were in stoichiometric. excess of the copper. In such cases, a single polarographic wave vas obtained and the diffusion current was proportional to the amount of coppel present in the solution. I t was proposed to reverse these conditions-that is, to add samples of propylenediamine to copper chloride solutions in which the total amount of copper \vas fixed a t a constant amount that would always he in stoichiometric excess of the propylenediamine to be determined. A noncompkx-forming supporting electrolyte was needed to make the conditions suitable for the polarographic determination of the propylenediamine-copper complex. The polarogram of such a solution was expected to show a double wave analogous to that obtained when a copper-
ammonia complex is determilied polarographically ( 4 ) . Subsequent investigations have shown that the double wave obtained was not a split wave that resulted from the stepwise reduction of copper, but instead was an initial wave for excess copper and a second wave for the propylenediamine-copper complex. REAGENTS AYD APPARATUS
All the reagents used, except propylenediamine, were reagent grade. Copper chloride solution, 500 y of copper per ml. Prepare by placing 0.1342 gram of cupric chloride dihydrate in a 100-ml. volumetric flask and diluting to volume with double-distilled water. Potassium chloride solution, 1M. Prepare In double-distilled water. Propylenediamine, 880 y per nil. purified by azcotropic distillation with toluene. A dilution of the azeotrope was made with double-distilled water and the solution was standardized by titration with standard hydrochloric acid solution to the methyl red end point. Polarograph, ORNL high sensitivity ( 3 ) . The dropping mercury electrode (D.M.E.)was used as the indicating electrode and the saturated calomel electrode (S.C.E.) as the reference electrode. EXPERI3IENTAL
All polarographic analyses werc carried out in an aii-conditioned laboratory where the temperature was controlled to within 1 2 ' F. In the evaluation of this indirect polarographic method for the determination of propylenediamine, it n as desired to find a supporting electrolyte that would not complex either propvlenediamine or copper. A sodium borate buffer of pH 9.0 was tried first. The wave of the propylenediamine-copper complex (second wave) was well defined in this medium, but the diffusion current decreased after the solution had remained in the polaro-
,
ANALYTICAL CHEMISTRY
270 graphic cell for a short time. The diffusion current decrease was due to the gradual change in the equilibrium of the reaction between copper and propylenediamine. Carlson and coworkers ( 2 ) have found that copper forms a stable complex with propylene-
! 020-
0 9600 9000 8400780
-
0 720
-
0 660-
N
9
0 6000540-
2 0480’
2
04200 360 -
0.300-
i
0.240-
0.480-
P o l a r o p r a p h , O.R.N.L. High Sensitivity Supporting Electrolyte, 0.48 K C i indicating Electrode, dropping mercury Reference E l e c t r o d e , r a t u r a l e d calomel Temperoture.room l c a 23.C)
/e I
,I
0.420-
I
I
diamine, ethylenediamine, and other chelating amines. They have suggested that the formula for the propylenediamine-copper complex is CupnzClz. However, this complex prevails only when propylenediamine is in stoichiometric excess of the copper. When copper is in excess, the complex CupnClzprevails. In either case, Cupn2Clnis formed first. If copper is in excess, the CupnzC12 decomposes into CupnCIz. Calvin and Bailes ( 1 ) used both potassium chloride and potassium nitrate as noncomplexing electrolytes to obtain polarograms of solutions of the order of l O - 3 M in propylenediamine that contained copper in great excess of the propylenediamine. They ’ mixture of pyridine and water rather than used a 50 volume % pure water as a solvent for the complex because the dissolution of mercury that occurs a t potentials below the order of -0.1 volt is less in the pyridine-water mixture than in water. However, the dissolution of mercury did not interfere either with the formation of the complex wave or with the formation of the copper wave when no propylenediamine was present, so water was chosen as the solvent for the sample. Polarograms were obtained of a series of solutions of propylenediamine-copper complexes by use of potassium chloride as the supporting electrolyte. The diffusion currents of these waves were related to that of the copper reduction wave by obtaining the ratio of the diffusion current of the wave for the complex t o that of the wave for pure copper. The maximum concentration that could be determined by the method is limited by the concentration of propylenediamine that would be equivalent to 50 y of copper per ml The range could have been extended t o the limit of sensitivity of the polarographic by increasing the copper concentration. These experiments were repeated with potassium nitrate as the supporting electrolyte, but the diffusion currents of the complex waves were not as reproducible, as were those with potassium chloride as the supporting electrolyte.
0.060I
/
/
I
l
l
1
1
1
PROCEDURE
1
Table I.
Polarographic Determination of Propylenediamine as the pn-Copper Complex (Supporting electrolyte, 0.1M XCl. Diffusion current of 50 y Cu per ml.
-
4.67 pa. (td:).)
Known Concentration pn, y/Ml. 26.4
35.2
61.6
79 , 2
96.8
105.6
114.4
Analysis of a Sample. Determine the diffusion current of a solution that contains 50 y of copper per ml. of 0.1M potassium chloride solution. Pipet 1.00 ml. of a solution that contains 500 y of copper, as cupric chloride, per ml. into a 10-ml. volumetric flask. Pipet into the flask a volume of sample (not more than 7 ml.) that will give a concentration of 26 to 97 y of propylenediamine per ml. when the solution is diluted to 10 ml. Add 1.0 ml. of 1 . O M potassium chloride solution to the flask.
Diffusion Current pn-Cu Complex (idz), p.4
idz tdl
p n Found,
?/MI.
Rel., Std. Dewation
0,850 0.865 0.850 0,860
0.182 0.185 0.182 0.184
26.26 26.68 26.26 26.54
0.80
1.210 1.225 1.225 1,220 2.350 2.365 2.350 2.335 3.04 3.02 3.05 3.03 3.06 3.84 3.84 3.87 3.83 4.10 4.11 4.04 4.04 4.36 4.40 4.40 4.39
0.259 0.262 0.262 0.261 0.503 0.506 0.503 0.500 0.631 0.647 0.693 0.649 0.655 0.822 0.822 0.829 0.820 0.878 0.880 0.865 0.865 0 934 0.942 0.942 0.940
34.93 35.33 35.33 35.20
0.54
61.60 61.96 61.60 61.23 79.20 78.71 79.44 78.96 79.68 96.68 96.68 97.51 96.45 106.1 106.3 104.5 104.5 113.7 114.6 114.6 114.4
T
t.0pa
i OMPLEX WAVE
0.47
0.48 POLAROGRAPH, 0 R N L HIGH SENSITIVITY SUPPORTING ELECTROLYTE, 0 4 KCI !NDICATING ELECTRODE, DROPPING MERCURY T E M P E R A T U R E , c a 23’C. COPPER IN S A M P L E , 507 p e r m l . P R O P Y L E N E D I A M I N E IN S A M P L E , 4 2 1 per mi.
M
0.48
0.94
0.37
CO.4
0.0
-0.4
-0.2
-0.3
-0.4
-0.5
-0.6
-0.7
-0.8
E d , e , ~ s . t h e S.C.E.
Figure 2.
Typical polarogram of the pn-Cu complex
271
V O L U M E 27, NO. 2, F E B R U A R Y 1 9 5 5 Add 0.1 ml. of 1.0% gelatin solution, dilute the solution to 10 ml., and mix it thoroughly. Pipet a suitable aliquot of the final solution into the polarographic cell, and determine the diffusion current of the wave of the propylenediamine-copper complex, which is the second section of the double wave. Calculate the ratio of the diffusion current of the complex wave to the diffusion current of the solution of 50 y of copper per ml. of 0.1M potasfiium chloride. From the calibration curve, determine the propylenediamine concentration in micrograms per milliliter that corresponds to the ratio calculated in the previous step. The 10-ml. sample volume given above is suggested as a convenient volume, but the range of the method may be extended by use of a much smaller volume, bearing in mind the recommended concentrations of copper, supporting electrolyte, and gelatin. Preparation of Calibration Curve. Prepare the calibration curve as indicated above, by use of a standardized propylenediamine solution instead of the sample, and repeat the procedure for a reasonable number of values within the concentration range indicated above.
over-all precision of O.iO%. was obtained by the formula
The value for over-all precision
Over-all precision =
tot?^^^
N
where N represents the number of sets of data, n represents the number of determinations in each set, and S represents the relative standard deviation for each set of data. .4n investigation of the suitability of this method for the determination of micro amounts of propylenediamine in solutions or soluble solids has shown the reason for a double wave The two waves, observed in a solution containing a concentration of cupric ion exceeding that of the propylenediamine (Figure 2), are due to the two-electron reduction of the uncomplexed cupric ion to metal (wave I) and the two-electron reduction of the cupric monoamine complex to the metal (wave 11). Figure 2 shows a typical polarogram of the propylenediamine-copper romplex in excess copper solution. The approximate half-wave potentials (E,/2 ) of the propylenediamine-copper complex wave vary from -0.4 to -0.6 volt us. S.C.E. with variations in propylenediamine concentration of the sample. The more positive half-wave potential is obtained with the minimum propylenediamine concentration in the range of this method, and the more negative value is obtained with the maximum concentration.
RESULTS AND DISCUSSIOY
Figure 1 sho-s a plot of the ratios of the diffusion currents of the propylenediamine-copper complex waves to the diffusion current of a solution of 50 y of copper per ml. on the ordinate us. concentration of propylenediamine on the abscissa. The minimum propylenediamine concentration that would produce a well-defined Ivave \vas 8.8 y per ml., and the maximum concentration that could be determined was 114.4 y per ml. The portion of the graph between the concentrations of 26.4 and 96.8 y per ml. is linear. In the lower concentration range, maxima tended to form, but they were eliminated by the addition of gelatin in the amount of O . O l ~ oby volume. Table I gives the results obtained for solutions of known concentration that were prepared from the standardized propylenediamine solution described above. The results obtained within this range have an
LITERATURE CITED (1) Calvin, M., and Bailes, R. H., J. Am. Chem. SOC.,68, 949 (1946). (2) Carlson, G. A., AlcReynolds, J. P., and Verhoek, F. H., Ibid., 67,1334 (1945). (3) Kelley, hl. T., and Miller, H. H., ANAL.CHEM.,24, 1895 (1952). (4) Kolthoff, I. M., and Lingane, J. J., “Polarography,” 2nd ed., Vol. I, pp. 223-7, Interscience Publishers, iiew York, 1952. RECEIVBD for review June 10, 1854. Accepted h’ovember 3, 1854.
Use of Sodium Borohydride for Determination of 3-Keto bisnor-4-cholene-22-al ERIK H. JENSEN and WILLIAM A. STRUCK Research Laboratorks, The Upjohn Co., Kalamazoo, M i c h .
This investigation was undertaken to devise a method for determining the purity of 3-ketobisnor-4-cholene22-al. This aldehyde is an important intermediate in a partial synthesis of progesterone from either ergosterol or stigmasterol. Accordingly, the determination of its purity is important. The essential reaction is a sodium borohydride reduction in alkaline medium, and the purity of 3-ketobisnor-4-cholene-22-alis based on the amount of borohydride used in the reduction of the aldehyde group. The accuracy of the method is excellent and the standard deviation is *0.34% for samples in the 80 to 100% purity range.
T
WO methods of obtaining progesterone from 4,22stigmastadien-3-one have been reported by Hey1 and Herr ( 2 , 3, 4). In both of these methods the key intermediate CH. is 3-ketobisnor - 4- cholene- 22-ali ’ HC-C=O (I). I 1 This aldehyde is prepared from 4,22-stigmastadien-3-one( 11) and is contaminated principally with (11) and 3-ketobisnor-4-cholenic 0 acid(III), the oxidation product (1)
of the aldehyde. present.
Other undefined contaminants may be
CH
l 3
HC--C=O
Similarly, Sheperd et al. (IO)have described the synthesis of progesterone from ergosterol. Again, the 3-ketobisnor-4cholene-22-a1 (I) is a key intermediate. In this instance, the aldehyde is prepared by ozonolysis of the side-chain double bond of 4,22-ergostadien-3-one (IV).
