about 0.6 of the original smoke condensate (water-free basis). Recoveries of BaP for the data in Table I, based on the perylene recoveries, ranged from 80 to 90 % with Method B and 70-75 with Method A. The amount of perylene recovered was determined by ultraviolet absorption spectrometry. Usually 7.0 kg of perylene were added to the methanol-water solution of the smoke condensate and the fractionation proceeded as described. BaP values obtained by the G C method are in good agreement with those obtained by the fluorometric method, as seen in Table I. A typical gas chromatogram of the BaP fraction isolated from cigarette smoke is shown in Figure 2. As little
as 1 ng of BaP was easily measured. Estimates of the test relative standard deviation calculated from these and other data are 8 % for the G C method and 5 % for the fluorometric method. N o statistically significant differences in BaP values between methods were found. ACKNOWLEDGMENT
The author gratefully acknowledges the capable technical assistance of Miss Marcia R. Thomas in carrying out the smoke condensate fractionations and fluorometric analyses. RECEIVED for review March 6, 1968. Accepted May 29,1968.
Separation of Iron(II)-I ron(III) Mixtures by Simultaneous Paper Chromatography and Electrophoresis W. E. Lingren, G . R. Reeck,’ and R. J. Marson2 Department of Chemistry, Seattle Pacific College, Seattle, Wash. 98119
ALTHOUGH the separation of the valencies of iron by paper chromatography and ion-exchange has been the subject of several investigations (1-9) in recent years, relatively little has been reported on the separation of this pair by electrophoretic means. Nagai and Kurata (IO) have studied the paper electrophoresis of hexacyanoiron(I1) and hexacyanoiron(II1) in 0.5M lactic acid solutions and Maslowska (11) reports the electrophoresis of these same anions in polyvinyl alcohol. In this study, descending chromatography with simultaneous low voltage paper electrophoresis was used to obtain separations of mixtures of Fe(I1) and Fe(II1) in 0.005N sulfuric acid. EXPERIMENTAL
Reagents. The stock solution of ferrous ion was made by weighing primary standard grade (Thorn Smith Co.) 1 Department of Biochemistry, University of Washington, Seattle, Wash. 98105. School of Education, Harvard University, Cambridge, Mass.
021 38. (1) S . Harasawa and T. Sakamoto,J. Chem. SOC.Japan, Pure Chem. Sec., 73,240 (1952). (2) K. A. Kraus and G. E. Moore, J. Am. Chem. Soc., 75, 1460 (1953). (3) J. R. Anderson and E. C. Martin, Anal. Chim. Acta, 13, 253 (1955). (4) C. Bighi, Ann. Chim. (Rome), 45,1087 (1955). (5) F. H. Pollard, J. F. W. McOmie, and A. J. Banister, Chem. Ind. (London), 1955, 1598. (6) H. M. Stevens, Anal. Chim. Acta, 15, 538 (1956). (7) F. H. Pollard, J. F. W. McOmie, A. J. Banister, and G. Nickless, Analyst, 82, 780 (1957). (8) F. H. Pollard, J. F. W. McOmie, A. J. Banister, and P. Hansen, J . Chromatogr., 4, 108 (1960). (9) Mohsin, Oureshi, Iqbal Akhtar, and K. G. Varsheny, ANAL. .CHEM., 38, 1415 (1966). (10) Hideo Nagai and Shiaehuro Kurata, N i b o n Kaaaku Zasshi, ’ 77,451, (1956); Chem. i b s t r . , 51, i5335f(ik7). (11) Joanna Maslowska, Chem. Anal. (Warsaw),10,1151 (1965).
Fe(NH4)2(S04)2. 6H20 and dissolving it in 0.005N HzSO4. The ferric ion stock solution was made from reagent grade Fe(NH4)(S04)2.12H20 and standardized by the 1,lO-phenanthroline method (12). Eaton-Dikeman electrophoresis paper No. 301-85 was used, without pretreatment, for all separations. It was cut into 23 x 28 rectangles and a row of holes punched along each 28-cm edge. Blanks run to determine the iron content of the paper revealed that the scissors and hole-puncher could easily contaminate the paper with minute quantities of iron (13). This was prevented by using new or freshly plated tools. Apparatus. The apparatus is similar to that described by Strain (14). The electrodes were made of 18-gauge platinum wire. The electrophoresis paper was supported, during a run, on a clean Styrofoam sheet held at an angle of about 50” with the horizontal. The electrolyte was placed in polyethylene troughs above the paper. The complete apparatus was enclosed in a large polyethylene bag. The voltage sources were Heathkit units, model Nos. EUW-15 or IP-32. Procedure. The mixture of iron valences was applied to the previously moistened paper at a point halfway between the electrodes and about 7 cm from the upper edge. The 6BFewas found by scanning with a Geiger tube. The nonradioactive species of iron were located using colorimetric reagents. The spots were torn out of the dried paper and wet-ashed with a mixture of 5 ml of 16M “01 and 7.5 ml of concentrated (70%) HC104. A vigorous but smooth reaction produces complete oxidation of the paper in about five minutes. The iron was determined with 2-nitroso-l-naphthol-4sulfonic acid using the colorimetric method of Pollard (7) or, in a few instances, by the 1,lO-phenanthroline technique.
(12) E. B. Sandell, “Colorimetric Determination of Traces of Metals,” 3rd ed., Interscience, New York, 1959. (13) W. E. Lingren, G. Reeck, and R. Marson, J. Chromatogr., 24, 269 (1966). (14) Harold H. Strain, ANAL.CHEM., 30, 228 (1958). VOL. 40, NO. 10, AUGUST 1968
1585
Table I. Iron(I9 Migration Distancea Applied voltage, V/cm cm 2 4 6 8 12 16 a
Angle,” degrees 1 2 16 20 31 39
5.9 6.0 5.9 6.3
7.0 8.6
Range = 0.5 cm in worst case (16 V/cm). = 5 x in worst case (16 V/cm).
6 Range
Table 11. Ratio of Moles of Fe(1I) to Moles of Fe(II1) after Separation“
Applied voltage, V/cm 2 4 6 8
12 16
Fe(II)/Fe( 111)
Relative error,
0.970 0.974 1.08 1.11 1.12 1.12
-3.9 -3.6 $6.9 f9.9 $11 +11
Applied Fe(II)/Fe(IIT) ratio was 1.01 in each case.
RESULTS AND DISCUSSION The Fe(II1) spot trailed the Fe(I1) spot and tended to move straight down the paper showing slight deviations from this path only at the highest voltages. At 16 V/cm and above, a slight bulge of the spot appears on the negative electrode side. The migration of the Fe(I1) spot, which was always toward the negative electrode, varied with increasing voltage as shown in Table I. In Table I the distance of migration is measured from the center of the initial spot to the center of the final spot and the angle reported is that between a line parallel to the electrodes and a line drawn through the centers of the initial and final spots. The results are the average of at least four runs and were taken from experiments in which the time of voltage application was one hour. These data indicate that the electric field does not appreciably affect the separation until about 6 V/cm is applied. Separations above 16 V/cm were obtained; however, solvent evaporation due to the heating effect of the current complicated the results. Good separations were obtained in about 30 minutes at 12 V/cm with one-to-one mixtures containing one pmole of each species. One-to-one mixtures containing four pmoles of each species took about 50 minutes. Separations of mixtures in the range of molar ratios 40: 1 to 1:40 were also obtained. The one-to-one mixtures were studied quantitatively and the results placed in Table 11. These results have been cor-
1586
ANALYTICAL CHEMISTRY
rected for the iron content of the paper which was obtained from blanks, The total amount of iron extracted in each case agrees well with the total amount placed on the paper. The results of Table I1 were taken from experiments in which the time of voltage application was one hour and are the averages of at least three runs. There is evidence of slight tailing in the Fe(I1) spot which would make the observed ratios low. The migrating species of Fe(I1) is positively charged. The ligand S042- would not be expected to form a positive species with Fe(I1); thus it is likely that the migrating form is a hydrated ion such as Fe(H20)e2+. The main migrating species of Fe(II1) appears to be either a mixture of neutral and positive particles or a positive ion which requires higher voltages to deflect than those applied in this study. The tendency of Fe(II1) to hydrolyze, even at the pH of this study is well known, thus Fe(0H) (H20)52+,Fe (OH)z(H20)4+, and Fe(OH)3(H20)3, along with Fe(S04)+, could singly or in some combination, exist in this spot. Although Fe(I1) might be expected to be oxidized to Fe (111) by molecular oxygen, the tendency decreases with increasing acidity and in these experiments could not be detected. The increase of about 10% in the measured iron ratios of Table I1 at 6 V/cm and above indicates that either some Fe (111) was converted to Fe(I1) during the separation or that an equilibrium species of Fe(II1) migrates under these conditions with approximately the same response as the major Fe(I1) species. Reasonable care was taken to prevent simple chemical reduction of Fe(II1) by impurities in the solution and by electrolysis products. Earlier work suggested that a direct electrochemical reduction might be occurring ( 1 3 , but further investigation of that possibility did not support the idea. Tracer 59Fein the + 3 oxidation state was placed in the oneto-one mixtures and studied at 2 V/cm and 12 V/cm. Its behavior paralleled the data of Table I1 with about 9 % of the total activity found in the Fe(I1) spot after separation at 12 V/cm. The oxidation state of the iron activity in the Fe(I1) spot could not be determined unequivocally. Attempts to detect Fe(II1) in the Fe(I1) spot by spectrophotometric means were also inconclusive. This study shows that the iron valencies can be separated by this method but the loss of Fe(II1) to the Fe(I1) spot during separation detracts from the usefulness of the method as a quantitative tool. The method is slightly faster than the reported techniques using simple chromatography. RECEIVED for review February 28, 1968. Accepted May 14, 1968. Work partially supported by the US.Naval Radiological Defense Laboratory, San Francisco, Calif. (15) W. E. Lingren, G . Reeck, and R. Marson, USNRDL-TRC-84, San Francisco, Calif., September 1966; Chem. Abstr., 6 7 , 6 0 4 3 ~ (1967).