V O L U M E 2 7 , NO. 12, D E C E M B E R 1 9 5 5
1895
Table 11. Analysis of Test Mixture
Benzene Styrene Anisole Benzaldehyde Acetopherione Phenol Benzoic acid Benzyl alcohol Total Standard deviation a
Found 26.3 5.02 6.73 4.52 5.76 13.8 10.3 27.8
100.23
Per Cent Known 22,95 5.12 6.65 4.43 5.79 13.59 10.41 28,03
Error' +1.35 -1.95 +1.20 f2.03
-0.52 +1.55 -1.34 -0.82
c,
M1. 23.1 96.2 138.6 173.2 231.0 323.4 423.5 565.9
100.00 &1,51%
Expressed a s per cent of amount present.
necessary that infrared absorption does not have the sensitivity of detection that ultraviolet absorption does. The principles described should be applicable to detection of materials separated by other means such as partition and ion exchange chromat,ography, liquid-liquid extraction, and fractional distillation, provided that the proper concentrations of absorbing substances in nonabsorbing media can be achieved. LITERATURE CITED
Blm, R. S., Acta Chem. Scand., 6, 1186 (1952). Alm, R. S., Williams, R. J. P., and Tiselius, A., Ibid., 6 , 826 (1952).
Bock, R. M., and Ling, Nan-Sing, ANAL. CHEM.,26, 1543 11954).
the spectrum of cut 3 (scans 6 and 7 ) , run a t scanning rate of 0.4 mfi per second on chart paper supplied by Applied Physics Corp. This procedure is time-consuming. A more desirable procedure would be t o combine all of the cuts containing a certain compound, measure the volume, and run a spectrum of the combined cuts. APPLICATION TO OTHER SYSTEMS
&,
The principles described in this paper have also been applied in a preliminary way to infrared scanning of chromatographic column effluents. An arrangement was devised whereby a Perkin Elmer Model 112 infrared spectrophotometer could be used in a similar manner. In this case also the limit switches were adjustable for any given wave-length interval. I n the preliminary investigations, however, only the C-H stretching region around 3.4 microns has been employed. Tp-o major difficulties are apparent: (1) Gradient elution is difficult, since no suitable polar solvent is available which does not have a C-H absorption itself; (2) because of the solvents available, such thin cells are
Br&vn:J. d.,Ibid., 25, 774 (1953). Busch, H., Hurlbert, R. B., and Potter, V. R., J . Biol. Chem., 196, 717 (1952).
Cherkin, 9.. Martinez, F. E., and Dunn, M.S.,J . Am. Cham. SOC..75. 1244 (1953).
Donaldson, K. O., Tulane, V. I., and Marshall, L. ll.,ANAL. CHEM.,24, 185 (1952). Hagdahl, L., Williams, R. 5. P., and Tiselius, A., Arkiv Kemi, 4, 193 (1952).
Hoyer, H., Kolloid-Z., 127, 166 (1952). Lakshmanan, T. K., and Lieberman, S., Arch. Biochem. Biophys., 45, 235 (1953). Marshall, L. M., Donaldson, K. O., and Friedberg, F., ANAL. CHGM., 24, 773 (1952). Paladini, A. C., and Leloir, L. F., Ibid., 24, 1024 (1952). Parke, T. V., and Davis, W. W., Ibid., 24, 2019 (1952). Parr, C. W., Proc. Biochem. SOC.,Biochem. J . (London), 56, xxvii (1954). Tenney, H. -M., and Sturgis, F. E., ANAL.CHEM.,26,946 ( I 954). Tiselius, A., Endeavour, 11, 5 (1952). Williams, R. J. P., Analyst, 77, 905 (1952). RECEIVED for review February 4, 1955. Accepted August 8, 1955. Presented before t h e Delaware Chemical Symposium, Newark, Del., February 19, 1955, a n d also before t h e Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, March 3, 1955.
Chromatographic Separation and Quantitative Estimation of Iodine-1 31labeled Derivatives of Sterols, Amines, Acids, and Aldehydes WILLIAM M. STOKES, WILLIAM A. FISH, and FREDERICK C. HICKEY Medical Research Laboratory, Providence College, Providence, R. 1.
Mixtures of compounds labeled with iodine-131 have been chromatographed. Apparatus has been developed for following the course of the development at all stages by automatically scanning the column and recording the distribution of the associated gamma activity. The method has been applied to the separation and quantitative estimation of the components of selected pairs of derivatives of the following types: sterols, amines, acids, and aldehydes.
T
H E present authors' use of radioactivity as a means of following the development of a chromatographic separation and estimating the quantities of the various components, as well as their purity, was found to offer so many advantages in the separation of certain sterol mixtures (6, 7 ) that they were interested in extending the method to systems characterized by other functional groups and improving the convenience and accuracy of the method. p-Iodobenzoic acid-iodine-131 derivatives of sterols and amines, p-iodoaniline-iodine-131 derivatives of acids, and p-iodophenylhydrazine-iodine-131 derivatives of aldehydes have been separated. The degree of purity
of certain of the separated fractions has been measured by the use of carbon-14-labeled components. The first separations were followed by manual spot scanning, then automatic scanning and photographic recording were used, and finally simultaneous automatic scanning and graphic recording were developed. METHOD AND APPARATUS
Preparation of Column. Chromatographic tubes from 1.2 to 4.5 cm. in diameter and from 60 to 120 em. in length were prepared from standard borosilicate glass tubing with a constriction a t the lower end. The tubes were coated with General Electric SR-53 silicone resin, and a plug of glass m-001 and a layer of clean sand were introduced to support the adsorbent. The adsorbent, two-thirds silicic acid and one-third Celite, was prepared as previously described (6) with the following changes. Celite 503 was found to increase the flow of developer and was therefore substituted for the 535 previously used. The Celite was calcined a t 500" C. to remove any traces of organic matter. The components were mixed in a ball mill and heated for 12 hours a t 200' C. To eliminate inhomogeneities due to different rates of fall of the components down the length of the column during packing, the adsorbent was mixed within the tube by revolving a notched rubber stopper, attached to a glass rod, through the adsorbent before the application of vacuum.
1896
ANALYTICAL CHEMISTRY
Apparatus. I n earl) runs the tube v’its attached to a Table I. Experimental Results filter flask and the pressure Snalysis from reduced t o 20 cm. of mercury to accelerate the flow of solridded t o Chromatographic Radioactl\ e Melting Points of radloactlvity assaY of Recovered Cornpound* vent. Recently a turret Column trace, eluates Literaa r r a n g e m e n t has been emComponent Ng, V T a 3Ig. c; Found ture ployed to allow two columns Cholestanyl p-iodobenzoate > 4 7 . 2 >50 9 5 2 . 1 186 186 (6) to develop simultaneously yet Cholesteryl p-iodobenzoate 49.5 54.2 184.5 184.5 ( 6 ) out the loss of vacuum. The N-p-Iodobenzoyl-m-toluidine 26.1 34,7 33.2 148.0 first survey method consisted N-p-Iodobenzoyl-u-toluidine 49.1 63.3 66.8 .I 177.4 in moving the counter (sliding B 177.4 on a steel shaft) manually. Acet-p-iodoanilide 79.0 57.7 54.1 73.4 51.1 185.0 184 13) taking 1- to 5-minute counts Prnprion-p-iodoanilide 57.8 42.3 39.3 60 8 4 2 . 4 168.0 Diacet-p iodoanilide 6.6 9.4 6 5 161-165 a t each centimeter down the length of the column. Subse194-195 (dec.) 196 ( 3 ) quently an electric drive ivds added. 146 ideo.) 148 (3) The apparatus, Figure 1, c o n s i s t s of t h e c h r o m a t o graphic assembly made up of the developer reservoir, A , The counter, shield, and yoke slide on the vertical steel shaft, E, chromatographic tube, B , and filter flask, C, all supported b y clamps from a vertical rod (not shown in the diagram), which is mounted on a firm steel base. The sliding assembly is counterbalanced by a lead weight attached t o a cable running over n firmly attached to the main supporting shaft, E. pulley at the top of the column. A spring-retracting steel centimeter tape, D: mounted on top of the shield, indicates the distance the counter has traveled downward from the top of the column. Parallel to the supporting column and offset to the side is the driving screw, F , which is driven by a synchronous motor, AI, through a double set of worms and gears, L. In the diagram, the gear on the driving screw, F , has been displaced upward for clarity. The driving screw passes throueh a bronze bearing attached t o the sliding counter assembly. A spring-loaded, retractable ratchet, J , engages the groove in the driving screw and draws the yoke, shield, and counter downward a t approximately 1 em. per minute. Withdrawal of the ratchet pin permits the manual return of the counter assembly to the top of the column. .1 microsFvitch, K , in series with the driving motor: limits the motion of the counter. V7hen the counter was moved manually, a scaler was used to record the activity. Along with the addition of an electric drive, a motion picture camera was used to record the total scaler count a t each minute while the counter scanned the length of the column at constant speed. The standard 16-mrn. camera was provided with a solenoid-operated plunger, ,which depressed the trigger button with just sufficient force and duration to allo\y one frame t o be exposed a t a time. The solenoid was actuated at minute intervals by the discharge of a condenser through a timing motor witch. At t,he expiration of a run the exposed film was cut from the roll, developed, and read in a microfilm reader. The film record gave, for each minute interval, the total count accumulated from the start. Each count was subtracted from the succeeding one in order to obtain t’he counts per minute for that particular time interval and its corresponding section of the column length. While this method was satisfactory for slow-developing systems, such as the sterols studied, it imposed serious delays in more rapid systems as well as tedious reading, recording, and computation. Finally, a count-rate-meter incorporating a puke amplifier and discriminator, as rye11 as a highly stabilized power supply, was constructed. The output of the count-rate-meter was fed into a Brown recorder. Thus, a continuoiis plot of the activity at each level of the column was produced simultaneously with the scanning of the column. Since the descent of the counter head along the length of the column and the rate of travel of the recording pen n w e constant, the abscissa of the plot was proportional t o the distance from the starting point on the column, whereas the ordinate was proportional to the ganinia ray activity Figure 1. Apparatus for automatic radioactive surat each point. vey of chromatographic column by scintillation counter Cnlike conventional chromatography, where only the surface coloration is apparent during development, in radiochromatography the counter responds to the total activity across the column section subject to the geometry of the counting arrangement and The lead shield, H , for the scintillation counter, I , is supported the minor limitations for self-absorption, scattering, decay, and by a semicilcular yoke, G, which permits the counter and shield background fluctuation; hence, quantitative interpretation is to be rotated to the vertical position when i t is desired t o make use of the well-type crystal t o count active samples contained in feasible. I n practice the shielding surrounding the counter test tubes. The radiation reaching the crystal is collimated by permits the penetration of a small fraction of the radiation origiremovable lead plugs with and slit widths. The nating in sections of the column adjacent to the portion in front smaller slit width was used in the early stages of column develop of the slit, and the packing and glass scatter some of the rays into ment when the activity tTas concentrated a t the top of the column. ~~
d
V O L U M E 27, NO. 12, D E C E M B E R 1 9 5 5 the slit. Preliminary experiments showed these efiects to be small but the background does rise somewhat as a region of high activity in the column is approached. F17ithin this limitation and that of the precision of the counting instrumentation, the areas under the curves are proportional to the activity contained in the corresponding zones of the column as indicated by the correspondence within an average of 2.5% between the percentage cornposition of the original mixture and the results of analysis. This conclusion was confirmed by total recovery, i n one instance, within 5 % (Table I). For unifunctional compounds these 3re also proportional to the mole fraction present. The some\\-hat arbitrary choice of a background is the most szrious lirnitation to the quantitative eptimation of the various c0mponent.s of the system.
1897 (Figure 2) it was calculated that the cholesterol fraction comprised 47.9% of the total. A'-Cholestenyl and Cholesteryl p-Iodobenzoates-Iodine-131. The esters were worked up and chromatographed as described, except that a column 90 cm. long was employed. During develop ment, numerous surveys were made by the photographic recording technique. When development was complete, the column was allowed to run dry, and a final survey was made. When the data were plotted (Figure 3 ) an average background was chosen.
? nsoor
'" "c
2400
S Y STEAIS A\ ALYZED
Cholestanyl and Cholesteryl p-Iodobenzoates-Iodine-131. These were prepared as described by a published method ( 6 ) . The cholestanol-carhon-14 u as prepared by the biological reduction of cholestenone-carbon-14 ( 7 ) by rat liver
800C
I 3OOr 18
28
37
47
57
66
76
85
CENTIMETERS
Figure 3. Plot of photographic recording of automatic survey on developed chromatographic column Zone 4 , A7-cholestenyl p-iodobenzoate-iodine-131 Zone B, cholesteryl p-iodobknzoate-iodine-131
CENTIMETERS
Figure 2.
Results of manual survey of developed chromatographic column
Sections 1 to 6. cholesteryl p-iodobenzoate-iodine-131 Sections 7 to 12, cholestanyl p-iodobenmate-iodine-131 Counts per minute at each centimeter are plotted against distance down the column
A chromatographic column 1.8 by 60 cm. was prepared as deicribed and pretreated with one holdup volume of l to 10 benzene-Skellpsolve C. Without allowing the column to run dry, the activity was applied in Skellysolve C and the same solvent {{as used for development. A4fterabout 12 hours the activity front had traveled 58 cm., and preliminary survey indicated that the zones were well separated. The column was allowed to run dry and counts were made a t each centimeter down the length of the column. Figure 2 shows the results. From a calculation of the areas under each curve, the column was sectioned a t such intervals that each component was divided into six approximatelv equal portions. Each portion was eluted with benzene-alcohol, the dry residues were hydrolyzed, the hydrolyzates extracted with ether, the dry ether extracts taken up in 90% alcohol, and the sterols precipitated with digitonin. The carbon-14 specific activity of sections 1 to 5 of the cholesterol zone was 5.1 d= 2.1 counts per minute per milligram; t h a t of sections 8 to 12 of the rholestanol zone was 363 i~6 counts per minute per milligram, indicating that the zones were practically homogeneous ( 7 ) and that the cholesterol was contaminated with less than 1.4% of cholestanol. From the composition of the original mixture, 49.1% was liver sterol, containing a t least 95% cholesterol, and ,iO.9% was added cholestanol. From the areas under the curves
The content of each zone was obtnined from the expression ( S t , - iYr,) - B(t2 - ti) where A$7Tt, and jYTtl are the scaler readings a t the limits of the particular curve, B is the average background in counts per minute, and t z and t, are the times, in minutes, when the counter slit passes the limits of the zone. The results are given in Table I. The low results for A'-cholestenol were to be expected, because the original 47-rholertenol was known to contain cholesterol as an impurity. The AT-cholestenyl p-iodobenzoate softened a t 179" C., and shrank t o a light brown mass between 180" and 185' C., but failed to give a clear melt up to 225' C. N-p-Iodobenzoyl-o-toluidine-Iodine-I31 and l\r-p-Iodobenzoyl-m-toluidine-Iodine-131. These compounds were prepared according to the method of Cheronis (4),using p-iodobenzoyl chloride-iodine-131 ( 6 ) . A chromatographic tulle, 1.4 mi. in tlianieter and 45 em. long, Tvas filled and packed in the us11:il \\-a?'. I t \\-as prewmhed with one holdup volume of lY0 :wetic acid ill benzene, and the mixed N-p-iodobenzopl-o-toluidine and .\~-p-iodobenzoyl-m-tol~~idine were added, as a suspension, in the same solvent, which was used also for development. In 5 hours the separation as practically complete, as shown in Figure 4. Since the zone corresponding to S-p-iodobenzoyl-o-toluidine exhibited an unusual convexity :ilorig t,he trailing edge, this zone was divided into two portions, A and B in Table I and Figure 4, as a test for homogeneity. The three sections \\-ere elnted with methanol and the resiiltiiig solutions evaporated to dryness. The solutes were taken up in Skellysolve C, filtered, evaporated to incipient turbidity a t the boiling point, and allowed to cryst,allize. The crystals were collected on small filter paper disks. indicated by the data in Table I the two portions of the ortho zone had the same melting points. p-Iodoaniline-Iodine-131. This was prepared according to the method of Brewster ( 1 ) on 1 to 6 % scales. The sodium iodide131, 1 to 10 mc. in aqueous solution, was added with the first portion of iodine-l2i. It was found that steam distillation yielded the purest product with the least difficiilty. Proprion- and Acet-p-iodoanilide-Iodine-13 1. These were prepared by a published method ( 5 ) . A chromatographic tube, 1.8 cm. in diameter and 75 em. long, was prepared in the usual way and prewashed with one holdup
ANALYTICAL CHEMISTRY
1898 volume of a 2.5% solution of 1-propanol in Skellysolve C. The mixed anilides were applied to the column in 5% 1-propanol in Skellysolve C. The same solvent was used as developer. After 18 hours the zones were well se arated, were symmetrical, and the radioactivity between them ]Rad fallen to background. The smoothed recorder trace is shown in Figure 5 . However, unexpectedly, a third zone appeared which traveled faster than the two main zones. Iodine analysis indicated that this component was diacet-p-iodoanilide-iodine-131. The areas under the curves for the three zones were determined and the percentage of each component was calculated (Table I ) .
O
10
5
15
20
25
30
35
The p-iodophenylhydrazine was stored as the tin chloride double salt, as it appeared to be more stable to light and air than is p-iodophenylhydrazine. The derivatives were prepared by a published method (3). A chromatographic column, 1.8 cm. in diameter and 60 cm. long, was prepared in the usual way and prewashed with one holdup volume of Skellysolve C. The mixed p-iodophenylhydrazones-iodine-131 were dissolved, with heating, in the smallest possible volume of 10% 1,4-dioxane in Skellysolve C and added, by means of a pipet, to the top of the column. Development was carried out with 5% 1,4-dioxane in Skellysolve C. In 5 hours the front of the ortho compound had traveled about 44 cm. As Figure 6 shows, the zones were symmetrical and the valley between them was almost a t background.
I
40
CENTIMETERS
Figure 4.
Recorder trace of developed chromatographic column
Radioactivity in arbitrary units is plotted against distance from top of column Zone A and B, N- iodobenzoyl-o-toluidine-iodine-131 Zonc C, N-p-iodo&nzoyl-m-toluidine-iodine-131
The column was sectioned a t appropriate points to isolate the three zones, and elution was carried out with 1 to 1 l-propanolSkellysolve C. Since considerable decomposition had been experienced, on previous runs, during the evaporation of the solvent, the three zones were analyzed by the following radioactivity method. The eluates were diluted to known volumes, in volumetric flasks, and 3-ml. samples were removed for radioactivity determination in a well-type, high sensitivity, gamma ray counter.
C
5
10
15
20
25
30
35
40
45
50
CENTIMETERS
Figure 6. Recorder trace of developed chromatographic column Zone A, o-nitrohenzaldehyde p-iodophenylhydrazonbiodins-131 Zone B, m-nitrobenzaldehyde p-iodophenylhydrazone-iodine-131
The column was then allowed to run dry and air was drawn through it until no more solvent dropped from the end. The tube was sectioned to isolate the two zones and the corresponding components were eluted with diethyl ether. The latter waa e v a p orated and the samples were dried in vacuo, recrystallized from methanol (considerable loss was experienced on attempted recrystallization from ethyl alcohol, especially of the ortho compound) and melting points taken as shown in Table I. ACKNOWLEDGMENT
ob
’
io
I1O ’ 20 ’ ’ 40 ’ 5 0 ’ 610 ’ CENTIMETERS
‘
Figure 5 . Smoothed recorder.trace of developed chromatographic column Zone A, acet-p-iodoanilide-iodine-131 Zone B, pfoprionlp-iodoanilde-iodine-131 Zone C, diacet-p-iodoanilide-iodine-131
The activity of these samples was compared to that of a highly purified sample of proprion-p-iodoanilide-iodine-131from the same original preparation which had, therefore, the same molar degree of labeling. From the total counts per minute of the samples, the dilution factors and the equivalent weights, the weight of each component was determined. The results are given in Table I. 0- and m-Nitrobenzaldehyde p-Iodophenylhydrazone-Iodine131. p-Iodophenylhydrazine-iodine-I31 was prepared from p-iodoaniline-iodine-131 by an adaptation of the method of Bulow ( 2 ) for the synthesis of the corresponding 2,Pdichloro compound.
This work was supported by grants-in-aid from the American Cancer Society, upon recommendation of the Committee on Growth of the National Research Council, from the American Heart Association and from the Damon Runyon Memorial Fund for Cancer Research. The authors wish to acknowledge their indebtedness to H. R. Seelen of the Radio Corp. of America for essential electronic components, to the Leach Machinery Co. of Providence for the donation of the driving screw, to the Atomic Instrument Co. of Cambridge, and to William Dandreta and Co. of Providence for assistance received in the design of the electronic circuits employed. The authors are also indebted to Leo DiGioia for the preparation of the illustrations. LITERATURE CITED (1) Brewster, R.
(2) (3) (4) (5)
Q.,“Organic Syntheses,” Coll. vol. 11, p. 347, Wiley, New York, 1943. Biilow, C., Ber. deut. chem. Ges., 51, 399 (1918). Chattaway, F. D., and Constable, A. B., J . Chem. Soc., 1914, 124. Cheronis, N. D., “Micro and Semimicro Methods” in “Techniques of Organic Chemistry,” A. Weissberger, ed., vol. 6, p. 538, Interscience, Kew York, 1954. Korner, hl. E. G., and Wendler, V., Gazz. chim. it& 17, 486
(1887). (6) Stokes, W. M., Fish, W. A.. and Hickey, F. C., J . Am. Chem. SOC., 76. 5174 119541. (7) Stokes, W. iI., Fish, W A . , and Hickey, F. C., J . Bwl. Chem., 213, 325 (1955).
RECEIVED for review June 15, 1955. Accepted dugust 27, 1955.
