pre-equilibrated by 50 ml. of such a with f3-U L i N 0 3 ~ barium was eluted first, then ytterbium, follo~ved aft'er a discontinuous elution with 6 to 5.2M LiN03 (gradient, -0.2~11 per column volume); neodymium and lanthamum were fin d l y eluted with 4 to 1.4M LiNO3 @adient, -0.2M per column volume), Figure 4 shows the elution curve of this separation. The fractions (4 ml.) were analyzed quantitatively by complexometry, while the elements were identified b y HDEHP chromatography ( 6 ) , comparing the RF's of the peaks with those of the pure elements.
Table IV.
Chromatographic Separation of Rare Earths on Paper Treated with 0 . 2 5 M TNOA
Rare earths separated Pr-Sin-D y Th-La-Y Pr-Dy-Yb Th-U-La-Sc Ce-Sm-Ho Th-La-Pr-Eu La-Sd-Er Th-Yb-La Th-Ce-Sd-Gd Ce-Dy-Yb
Lis03 eluent, dl 3 1
6 1
Length, cin.
17 15 16 17
4
17
2 2 G
16
3 5
17 15
16
15
RF values P r 0 . 1 3 ; S m 0 . 5 2 ; DyO.S9 Th 0.01; La 0.62; Y 0.93 Pr 0.04; Dy 0.22; Y b 0.68 Th 0 . 0 0 ; U 0.14; La 0.54; Sc 0.80 Ce 0.08; Sm 0.66; Ho 0.83 Th 0.01; La 0.24; Pr 0.57; Eu 0.88 La 0.18; S d 0 . 5 8 ; Er 0.92 Th 0.00; Yb 0.58; La 0.73 Th 0.01; Ce 0.25; S d 0.63; C k I 0.86 Ce 0.04; Dy 0.44; Y b 0.77
i*.CKNOWLEDGMENT
The author thanks E. Cerrai for valuable suggestions and A. illbini for useful laboratory collaboration. LITERATURE CITED
(1) Blake, C. A,, et al., Proc. 2nd Intern. Conf. Peaceful Uses At. Energy 28, 289 (1959). (2) Cerrai, E., Testa, C., Energia Xucl. ( M z l a n )8, 510 (1961). (3) Cerrai, Cerrai, E., Testa, C., J . Chromatog. 5, 442 (1961). ( 4 ) Zbid., (4) Zbzd., 6, 443 (1961). (5) Ibad., 7, 112 (1962). ( 6 ) Zbid., 8, 232 (1962). (7) Cerrai, E., Testa, C., Triulzi, C., Eneryia h'ucl. ( M i l a n ) 9, 193 (1962).
(8) Ibid., p. 377. (9) Coleman, C. F., Brown, K. B , U.S. .4t. Energy Comm., TID-7555,43-56 (1957). (10) Coleman, C. F., et al., Proc. 2nd Intern. Conf. Peaceful Cses At. Energy 28, 278 (1959). (11) Rlann, C. K., White, J. C., AXAL. CHEM.30, 989 (1958). (12) Marcus, Y., Israel At. Energy Comm., Rept. R/20 (1959). (13) RIarcus, Y., Abrahamer, I., Zbid., IA-608 (1961). (14) Marcus, Y., Nelson, F., J . Phys. Chem. 63, 77 (1959). (15) Martin, A. J. P., Synge, R. L. N., Biochem. J . 35, 1358 (1941).
(16) IIoore, F. L., ASAL. CHI;>?. 32, 1075 (1960). (17) Moore, F. L., U. S. At. Energy Comm., NAS-NS 3101 (1960). (18) Peppard, D. J., et al., J . Znorg. h'ucl. Chem. 4,334 (1957). (19) Ross, W. J., White, J. c., k A L . CHEJI.31, 1847 (1957). (20) Testa, C., J . Chromatoy. 5, 236 (1961). (21) White, J. C., rlSTJ1 Spec. Tech. Publ. N238,27-34 (1958). (22) White, J. C., ROSS,W.J., U. S. A4t. Energy Comm., NAS-NS 3102 (1961). RECEIVEDfor review April 23, 1962. Accepted July 23, 1962.
Use of Countercurrent Distribution in the Quantitative Determination of the Alkaloids of Commercial Veratrine GLENN R. SVOBODA' Department of Pharmaceufical Chemistry, School of Pharmacy, University o f Wisconsin, Madison, Wis.
b The partition characteristics of the Sabadilla alkaloids have been applied to a countercurrent separation of cevadine and veratridine from the other alkaloids in commercial veratrine. The mathematics of countercurrent distribution have been utilized to develop a method for the determination of the equivalent percentage composition of the known alkaloids and analyses for two samples of veratrine are reported.
C
distribution is considered primarily as an effective sqiaration technique. However. a knodedge of the partition characteristics of the individual components of an alkaloidal mixture also allons a quantitative analysis of the total alkaloids without chemical modification, provided chemical artifacts do not appear under the conditions of separation. Veratrine is a commercial preparation OCKTERCURREKT
of the total alkaloids froin the seeds of Schoenocaulen oficinale (Sabadilla) available from S. B. Penick and Co. Several methods of analysis h a r e been reported in the literature (1, 5 ) . These procedures do not take into account the comple.;ity of the alkaloidal mixture present. iiuterhoff ( 1 ) assumed that veratrine was composed solely of a mixture of the alkaloidal esters, reratridine and cevadine, the two major components. On this basis an analysis of the steam-volatile veratric and angelic acids available upon ester hydrolysis would lead to erroneous composition values. Xacek et al. (5) have pointed out that such a procedure (1) does not take into account the presence of other ester alkaloids such as cevacine ( 4 ) , which liberates acetic acid upon hydrolysis. illso, duterhoff ( 1 ) does not take into account the nonester alkanolamines such as sabine (S),veragenine (Q), and veracerine (4,or the vanillic acid ester of veracevine, vanilloylveracevine (8).
The paper chroniatographic methods suggested b y Macek ( 5 , 6) are useful for qualitative and quantitative investigations of the major components of S a b adilla but are unsatisfactory in the identification of minor components. Rfacek et al. (6) have noted the difficulty of placing large quantities of veratrine on paper and then identifying a component which was less than 1% of the alkaloidal mixture. The partition characteristics of many of the alkaloids found in Sabadilla had been previously studied in this laboratory ( 7 ) . On the basis of this work it seemed feasible t o utilize a series of countercurrent distributions t o separate the major alkaloidal components of the mixture to facilitate a quantitative estimation of the minor constituents by the application of the mathematics of countercurrent separations. 1
Present addrese, Freeman Chemical
Corp., Port Washington, Wis.
VOL. 34, NO. 12, NOVEMBER 1962
1559
c
-.aE
iL0f N U V E E P
Y"VEE"
Figure 1. Countercurrent distribution of 1 00-mg. portion of P 6 , 7 , 8 from sample
B
80 transfers between chloroform and pH 4.85, 0.5M phosphate
82 transfers between chloroform and p H 4.22, 0.5M citrate buffer A. Veratridine
B. Cevadine C.
buffer
A.
E. C.
Cevadine Vanilloylveracevine Cevacine
Vanilloylveracevine
EXPERIMENTAL
Apparatus and Procedure. Large scale nine-plate countercurrent distributions were performed with separatory funnels of suitable volume. Equal volumes of buffer solution of t h e desired pH were placed in each of nine separatory funnels. T h e sample of veratrine was dissolved i n a volume of chloroform equal to t h e volume of buffer phase and shaken with the buffer phase in t h e first of t h e separatory funnels. A standard countercurrent distribution technique was utilized, moving t h e chloroform phase for eight transfers. Countercurrent distributions for more than eight transfers were made on a 200 - plate, robot - driven instrument, manufactured by H. 0. Post, Scientific Instruments Co., ivaspeth, S.Y. The apparatus was adjusted for 10-nil. volumes of both upper and lower phases. The 100-mg. and 200-mg. alkaloid samples were dissolved in 10 ml. of chloroform and placed in the first tube of the machine. Five-gram
Table I. Countercurrent Distributions Utilized for Determination of Alkaloidal Constituents of Veratrine
Material Sample A Mg. p6.7.8 200 Pq 200 ".3 . i" PO,1,2 100 Po.1,2 100 100 PO.l,Z Sample B 100 p6,7.8 200 PMS
Buffer
6.00
No. of transfers 80 85 _. 80
8.1
90
pH
4.4 4.85 6.80
i9
4.22 82 4.85 80 200 5g. 7.00 Po,1,2 100 8.00 90 PO,l2 a Contents of tubes 160-195 from distribution of PO.,,^ of sample B combined and isolated to give 1.476 grams of P0.l.p.
1560
Figure 2. Countercurrent distribution of 200-mg. from sample B portion of P3,4,5
ANALYTICAL CHEMISTRY
samples were dissolved in 20 ml. of chloroform and placed in the first two tubes. The alkaloidal content of the tubes a a s analyzed by nonaqueous titration using perchloric acid in glacial acetic acid. Quinaldine red was used as the indicator. Large Scale Countercurrent Distributions. S i n e 500-ml. separatory funnels were prepared for a n eighttransfer countercurrent distribution. -4200-ml. volume of 0 . 5 M , pH 5.00 phosphate buffer was used as t h e stationary phase in each funnel. T h e first lower phase contained 15 grams. Upon completion of eight transfers, 25 ml. of 5N potassium hydroxide was added to each separatory funnel to concentrate the alkaloidal contents in the chloroform. The contents of each funnel were extracted until a positive Mayer alkaloidal spot test could not be obtained. The combined extracts and washings were dried over anhydrous sodium sulfate and evaporated to dryness under reduced pressure. The contents of each funnel, consecutively, weighed 1.3, 0.9, 0.5, 0.1, 0.1, 0.4, 1.5, 3.3, and 6.7 grams. On the basis of the ultraviolet spectra and the weight distribution, the alkaloids were combined into three fractions: plates 0, I, and 2 ( P o , l , 2 ) ; plates 3, 4. and 5 (P3,4,5);and plates 6, 7, and 8 (P6,7,8). These were labeled sample A. The procedure was repeated with six 15.0-gram samples from a different batch of material. The extracts from these six distributions were combined to yield three fractions of sample B as follows: Po,l,z= 15.4 grams; P3,4,5 = 4.8 grams; P6.7,8 = 66.2 grams. Countercurrent Distribution Conditions. A number of countercurrent distributions were needed t o separate P3,4,5, and the alkaloids in the P6.7.8 mixtures obtained from veratrine. The organic phase was chloroform for every distribution. Phosphate buffer solutions (0.5M) were used above pH
4.5 and 0.5-11 citrate buffers were used below pH 4.5. Data for the various distributions are shown in Table I. RESULTS AND DISCUSSION
Determination of Percentage Composition of Veratrine. When more than 20 transfers are run in a Craigtype countercurrent distribution. the fraction of material in a n y tube, r , is given b y the expression:
where n = number of transfers; a = number of tubes from the peak tube; X = K/K 1, where K is the observed partition coefficient and Y = 1/K 1. For the fraction of material in the peak tube. a = 0 and Equation 1 reduces to
+
+
1 =
d2nnXY
I n the distributions used, the partition coefficients could be readily calculated from the distribution data. The position of the peak tube of a distribution is given b y r,,, = nX ( 2 ) . The constituent could be identified by its distribution characteristics. I n the analysis of the countercurrent distributions, all of the alkaloidal material was moved into the chloroform phase by the addition of strong alkali to the aqueous layer. An aliquot of the chloroform phase was titrated directly in the chloroform. The typical distributions as shown in Figures 1 through 4 represent the volume of acid necessary to titrate the aliquot portions xgainst the tube number. In the peak tube, the volume of acid necessary t o titrate the total alkaloids in that tube would simply be a factor times the aliquot volume used. As i t is known from Equation 2 that the peak tube
represents a definite fraction of the entire peak, the volume of acid to titrate the whole peak can readily be calculated. Accurately weighed samples of the alkaloidal mixtures to be separated were titrated to supply the total volume of acid required by the mixture. A typical calculation is: Calculation of the veracevine content rmaX= 53 (see Figure 4 ) ; n = 90
=
Equivalent Percentage Composition of Known Alkaloids in Two Samples of Commercial Veratrine
Fraction 1investigated
Alkaloid Veratridine Cevadine
P017*8 P6,1*8 p3.4.6
-4 31 .O 45.8 1 .o
0.588;
X 0.473 X 0.178 X 100 = 5.4
Allicluot (2 nil.) of r,,,,, titrated b y 0.178 nil. Total neak tube titrated bv 0.178 X 5 = 0.890I d . Entire peak n-ould be titrated by 0.890, 0.0854 = 10.42 mi.
It required 0.163 nil. of acid t o titrate each iniliigrani of Po or 16.30 ml. would haye been required to titrate the 100-mg. fraction subjeded to countercurrent distribution. =
10.42/16.30 X 100
=
63.9
On a neight basis, it was known t h a t the ratio of Po1.2a recovered from the distribution of Po,l2 as compared t o the total weight of all alkaloids recovered in the separation was 1.476/3.117 = 0.473. Similarly, the total content of veratrine as compared t o the total alkaloids recovered in the separation was 15.4h6.4 = 0.178. 08
28.4 45.5 1.6
R
B 31.0
28.4
46.8
47.1
The percentage compositions of the alkaloids in the two samples of veratrine are shown in Table 11. Interpretation of Results. T h e individual alkaloidal substances from natural sources vary in their percentage of t h e total alkaloidal mixture even as t h e total percentage of alkaloids varies from batch t o batch. Sometimes a minor alkaloidal constituent may appear in one sample of a mixtpre and not in a second. These variations complicate the studies of alkaloidal mixtures. The percentage composition as outlined in this paper is actually a n equivalent percentage composition. All of the known alkaloidal constituents of Sabadilla have been shown to be monobasic in nature. It has thus been convenient to refer t o the ratio of the volume of acid to titrate a given weight of
alkaloidal miyture as the equivalent percentage of the constituent. I t is difficult t o determinate the true weight percentage. Where the molecular weight of a constituent is knon-n. the true weight of a substance can be calculated. However, the alkaloidq of Sabadilla have shown tendencies to crystallize with solvent of crystallization, and this, coupled with the difficulty of obtaining complete drying without decomposition, made the determination of the true alkaloidal weight of a mixture a difficult problem. I n the case of those Sabadilla alkaloid.; d i i c h do not possess a characteristic ultraviolet spectrum, a titration procpdure had the advantage of being affected by only the basic constituents presentnamely, the ester alkaloids and the nonester alkanolamines. The utilization of a large-scale eighttransfer countercurrent distribution separated most of the two primary alkaloidal constitumts of Sabadilla,
A J IE
I 56
B
~ _ in_veratrine _ _ _ _ _
Unknown XI Sabatine S'eracevine Sabine
yo veracevine in veratrine would be 0.639
Yc of veracevine in P0,lp
Total % of constituent
70in veratrine
Vanil1:ylveracevine Cevadine
Of-Po 12"
K f l
Table II.
1[
0 TUBE NUMBER
Figure 3. Countercurrent distribution of 5-gram portion of Po,1.2 from sample B 2 0 0 transfers between chloroform and p H 7.0, 0.5M phosphate buffer A. Mixture (large part cevacine) 6. Unreported component C. UnknownXI D. Sabatine E. Mixture (hydrophilic components)
TUSE HUMBER
Figure 4. Countercurrent distribution of 1 00-mg. portion of peak E, hydrophilic components from distribution as shown in Figure 3 A. B. C.
Sabotine Veracevine Sabine
VOL. 34, NO. 12, NOVEMBER 1 9 6 2
e
1561
cevadine and veratridine, into a single fraction, P6,i,& slight overlap of vanilloylveracevine into P6.7,8 occurred, as well as some of the cevadine into PA4,s. However, this preliminary separation facilitated determination of the minor constituents. Preliminary separations vould not be required if all constituents occurred a t the same approximate concentration or if one were interested in only the major constituents. The distributions obtained with samples of the plate combinations from sample B after the eight transfer preliminary separation are shown in Figures 1 to 4. As noted by inspection, the components of P3,4.5 and P6,7,8 could readily be separated in a single distribution. With sample A, countercurrent distributions \yere made a t three p H values, whereas with sample B it was possible t o separate
x
all but the most hydrophilic constituents of P0,1,,with one distribution, using a larger number of transfers. The hydrophilic fraction was collected and redistributed a t a second pH, as shown in Figures 3 and 4. The difference betn-een the analyses shown in Table I1 and unity undoubtedly depends upon the evistence of a number of additional alkaloidal components. Determination and separation of these T o d d involve concentrates from large amounts of veratrine and more exacting countercurrent resolution. LITERATURE CITED
(1) Suterhoff, H., d r c h . Pharm. 286, 69
(1953). (2) Craig, L. C., Craig, D., “Technique of Organic Chemistry,” Vol. 111, Part 1, 2nd ed., A. Weissberger, Ed., Interscience,. Ken- York, 1956. (3) Hennig, 4. I., Higuchi, T., Parks,
I,. &I., J . Am. Pharin. AMOC., S a .
Ed. 40,168 (1951). (4) Kupchan, 11.K., Lavie, D., Delin-ala, C. B.,bndoh, B. T.rl., J . -4m.Chem. SOC.75 5519 (1953). ( 5 ) Pllacek, K., T’anecek, S., Pelcova, V., Vejdelek, 2. J., Chenz. Lzsty 50, 598
(1956). (6) Macek, K., 1-anecek, S., Vejdelek, Z. J., Zbzd., 49, 539 (1955). ( 7 ) PlIitchner. H.. Ph.D. thesis. rniversitr of Wisconsin School of Pharmacy, 1956. (8) Stuart, D. AI., Parks, L. M.,J . Am. Pharm. Assoc., Sei. Ed. 45, 252 (1956). (9) Vejdelek, Z. J., Macek, Ii., Iiakac, B., Chem. Listy 49, 1538 (1955). RECEIVEDfor reriev June 6, 1962. Accepted September 10, 196%. Kork supported in part by the Research Committee of the University of Wisconsin from funds supplied by the Kisconsin .ZbAlumni Research Foundation. stracted in part from a thesis submitted by G. R.Svoboda to the Graduate School of the University of Kisconsin in partial fulfillment of the requirements for the degree of dortor of philosophy. \
z
Separation of Metals by Cation Exchange in Acetone-Wa ter-Hyd rochloric Acid JAMES S. FRITZ and THOMAS A. RETTIG, Institute for Atomic Research and Department of Chemistry, Iowa State University, Ames, Iowa
b Distribution coefficients have been measured for the partition of metal ions between cation exchange resin and acetone-water-hydrochloric acid solutions. The differences in distribution coefficients of metal ions are greater in acetone-water media than in aqueous media of the same hydrochloric acid concentration. By using disiribution coefficient data, conditions for column separations of mixtures can be effected b y eluting with acetone-water-hydrochloric acid sohtions of different compositions. Successful separations of a number of mixtures are reported.
T
HE SEPARATION of metal ions as halogen complexes by elution from a cation exchange column with a n aqueous hydrohalic acid solution is now an established analytical technique. Fritz, Garralda, and Karraker (8) separated many metal ions using 0.1JI or 1JI hydrofluoric acid as the eluting agent. Yoshimo and Kojima ($2) and Strelow (17)separated cadmium(I1) from zinc(I1) and other metal ions b y elution with 0.5M hydrochloric acid. Fritz and Garralda (7) separated mercury(II), bismuth(III), cadmium(II), and lead(I1) from each other
1562
ANALYTICAL CHEMISTRY
and from other metal ions using 0.1 to 0.6M hydrobromic acid as the eluting agent. The extensive distribution coefficients of metal cations a t various concentrations of hydrochloric acid measured by Strelow (18) and b y Mann and Swanson (14) are a useful guide t o possible cation exchange separations in aqueous solution. It has been shown that metal ions are taken up more strongly and a t lower hydrochloric acid concentrations by an anion exchange column if an appreciable amount of a water-miscible organic solvent is added to the aqueous hydrochloric acid (10, 21). This behavior indicated the possibility of using a nonaqueous solvent to promote metal-halide complex formation for the selective elution of metal ions from a cation exchange column. Preliminary work by Pietrzyk (15) indicated that acetone is the most effective of the solvents tested. Buznea, Constantinescu, and Topar (8) and Ionescu, Segoescu, and Gainin (11) have effected the separation of copper(I1) and zinc(l1) on phenol formaldehyde-type cation exchange resin using acetone-water-hydrochloric acid solutions as eluting agents. Kember, Macdonald, and Kells have studied the behavior of several metals on Zeo
Karb 225, cation exchange resin using acetone-mater-hydrochloric acid eluants, but were able to separate only copper(I1) and nickel(I1) successfully (12). Van Erkelens has studied the ion exchange separation of complex mixtures of metal cations and anions in radiochemical amounts using acetonewater-hydrochloric acid solutions as eluting agents. He observed that certain separations. such as cobalt(I1)nianganese(I1) and iron(II1)-copper(I1)-zinc(I1) could not be effected, and that the addition of potassium iodide is necessary to effect the separation of copper(I1)-cobalt(I1) mixtures (90). I n the present aork. these mixtures are separated quite eacily, and many other separations either have been accomplished or are suggested hy the di-tribution coefficient data. EXPERIMENTAL
Reagents, Solutions, and Apparatus. R E ~ I N .DowexCATIONEXCHAXGE 50Q7 X8 analyzed reagent resin. 100- to 200-niesh, is used in t h e measurement of distribution coefficients and column separations. Before use, t h e rebin must be purified. Place the resin in a large column and backwash with distilled water to remove the fine particles. Then wash the resin
