precipitate upon the addition of 8-quinolinol to a solution of silver nitrate. This material, on recrystallization from aqueous ammonia, gave a yellow material, Froin the analysis of these materials, the xuthors suggest that the white material may have the composition [.lg(CgH6SOH)2]S03.and the yellow ninterial obtained from ammoniacal solutions on recrystallization, AgC9H69S. CsH70X. The silver complex of 8-(2,5-dimethylbenzenesu1fonamido)quinoline is yelloiv while the 1 to 1 chelates of the previous two ligands discussed are white. Other instances where this extra mole of ligand s h o w up may be found in sei-era1 of the complexes of 8-mercaptoquinoline ( 5 ) . The complex formed by precipitation of silver in acid medium is believed to have the composition CgH,XiSlg. CgH6NSH. 3Hn0. It is difficult to understand the nature of the bonding between the normal chelate and the extra ligand molecule. It has been suggested that the estra molecule is bound merely by weak lattice forces. The metal chelates of 3,3’-bis(quinoline-8-sulfamoyl) biphenyl sulfone are formulated as fused ring systems consisting of two 5-membered rings and one 12-membered ring. It \vas hoped that combining such a bis(bidentate) ligand with a tetrLxoordinate metal ion would lead to a linear polymer. However, due to the flexibility of the biphenyl sulfone system, the four sites can assume close proximity and chelate one mole of metal leading to the formation of the structure shown here:
n- -43 F
SO?
SO2
7so2
It is evident that the proposed structures of complexes with chelate rings containing more than six atoms are not firmly established. Lack of x-ray and other conclusive data, the several possible linkages, and the possibility of polymerization, all tend to make the proposed structures highly speculative. Several examples of condensed ring chelates containing a large membered ring are found in the literature (3, 6, 7 ) . In the chelates of 3,3’-bis(quinoline-8sulfamoyl) biphenyl sulfone, the stability of the smaller rings and the flexibility of the biphenyl sulfone system probably account for the formation of these complexes. The structure assigned to the yellow silver chelate is:
Elemental analysis checks for a 1 to 1 mole ratio of ligand to metal atom and the infrared spectrum shows a sharp
band in the 3-micron region attributable to the presence of 3’-H in the chelate. The nature of the yellow compound formed in the reaction between lead and 3,3’ - bis(quino1ine - 8 - sulfamoyl) biphenyl sulfone is in doubt The results of two analyscs run on separate samples check fairly tvell but are about one half the calculated values. Two possibilities are polymer formation or a complm with an unusually large metal to ligand ratio. Further work T\ ill include the tlevelopment of quantitative applications and a study of the possibilities of further specificity within the six metal ions. With the proper choice of one or more masking agents, it is hoped that high specificity will be obtained. ACKNOWLEDGMENT
The authors thank C. E. Kaslow for his donation of 8-aminoquinoline. LITERATURE CITED
(1) Billman, J. H.,
Janetos, N. S., Chemin, R., ANAL.CHEM. 32, 1345
11960). ( 2 ) Block, B. P., Bailar, J. C., Pearce, D. W., J . Am. Chem. SOC.73, 4971 (1951). ( 3 ) Calvin, M., Barkelew, C. H., J . Am. Chem. SOC.68,2267 (1946). (4) Hein, F., Regler, H., Nature 23, 320, 11935). ( 5 ) Kuznetsov, V. I., Bankovskii, I. A., Ievenish, A. F., J . Anal. Chem., U.S.S.R. 13, 3 (1958). ( 6 ) Pfeiffer, P., Breith, E., Lubbe, E., Tsumaki, T., Ann. 84, 503 (1933). ( 7 ) Schlesinger, N., Chem. Ber. 58, 1877 (1925).
RECEIVEDfor review August 25, 1961. Accepted December 29, 1961. Contribution No. 1035 from the chemistry laboratory of Indiana University.
N e w Fluorometric Micromethod for the Determination of Reserpine IVAN M. JAKOVLJEVIC, JOHN M. FOSE, and NORBERT R. KUZEL Analytical Development laboratory, Eli Lilly and Co., Indianapolis, Ind.
b
A new, rapid, highly sensitive, and reproducible micromethod for the fluorometric determination of reserpine has been developed. Concentrations of reserpine as low as 0.005 pg. per milliliter can be measured. The method does not measure the oxidative breakdown products of reserpine which occur frequently in pharmaceutical formulations. A solution of reserpine in glacial acetic acid is heated with a 1yo solution of p-toluenesulfonic acid in glacial acetic acid for 10 minutes, and the resulting fluorescence is measured.
410
ANALYTICAL CHEMISTRY
B
ECAUSE it
is an important therapeutic agent, many analytical procedures have been developed for the quantitative determination of reserpine. All of these procedures include many steps of quantitative separation such as: extraction, paper and column chromatography, paper ionophoresis, etc. Several methods are based upon the ultraviolet absorption properties of reserpine; among them is that of Sakal and Merrill (13) which combines the ultraviolet absorption method with paper ionophoresis. Banes ( 1 ) employs chromatography to isolate reserpine.
McMullen el al. (11) remove the decomposition products of reserpine by an extraction with dilute hydrochloric acid and sodium bicarbonate solutions. I n the same paper the authors give a procedure for the fluorometric determination of reserpine by heating with a mixture of sulfuric, hydrochloric, and acetic acids. This method is about 20 times less sensitive than the method described here. Poet and Kelly (12) determine reserpine fluorometrically by heating a sulfuric acid solution of reserpine in the presence of selenious acid. The sensitivity of this method appears
to be four times less than that found in the method here described. Bartelt and Hamlow (3) use a chromatographic technique to isolate reserpine from its decomposition products and employ ultraviolet absorption to measure the isolated drug. Several colorimetric methods have been reported. Banes ( 2 ) saponified reserpine and determined the resulting rtserpic acid through the use of a vanillin-hydrochloric acid reagent. The fluorescent greenish yellolv color produced by treating reserpine with dilute sulfuric acid and sodium nitrite was measured a t 390 mp by Szalkowski and hladtr (14) n ho extracted the decomposition products of reserpine with citric acid and sodium bicarbonate solutions. This method forms the basis for the XT’I reserpine present official U.S.P. procedure. Haycork and NIader (8) have investigated the specificity of the sodium nitrite reaction. The method of Booth (6) involves the formation of a chloroform soluble bromophenol blue compley. Wunderlich (15) used a colorimetric reineckate technique. Indenians, Jakovljevic, and Langerijt (9) developed an intense orange color (maximum a t 465 mp) by heating a reserpine solution in diluted acetic acid with sodium nitrite. The initial fluorescence produced in this reaction TI as destroyed by the heat. Bayer (4) determined reserpine in nonaqueous media by titrating with perchloric acid. For smaller amounts of reserpine he titrated with p-toluenesulfonic acid. Leemann and Weller ( I O ) used the same reagent to develop a yellow-orange color. Bayer ( 5 ) studied the decomposition products of reserpine by using paper chromatography. To eliminate time-consuming separation and extraction procedures, an investigation was made to develop a more specific reserpine assay. This assay should be useful in the analysis of pharmaceutical formulations and should also indieate the stability of these formulations. EXPERIMENTAL
Apparatus. A Turner Model 110 fluorometer equipped with a set of matched borosilicate glass test tubes (12 X 7 5 mm.) furnished by Turner was employed for all fluorometric measurements. Filter No. 7-60 was employed as a primary filter; this filter transmits light in the 360- to 400-mp region of the spectrum. Filter No. 8 was employed as a secondary filter; this filter transmits light a t 480 mp and above. To reduce the emission energy entering the multiplier phototube of the fluorometer, a No. 2 neutral density filter was used with the secondary filter. All filters are available through G. K. Turner Associates. An Aminco-Bowman spectrophotofluorometer equipped with silica cells
@
n
*I..:LzV:TH
r”
Figure 1 . Activation and fluorescent spectra of p-toluenesulfonic acid-reserpine fluorophor A.
Activation spectrum of fluorophor, peak at
380 m p 8.
Fluorescent spectrum of fluorophor, peak a t
480 m p C. Fluorescent spectrum of 170 p-toluenesulfonic acid reagent
was used to check all activation and emission wavelengths. Figure 1 illustrates the activation and fluorescent spectra of the fluorophor. Reagents. Glacial acetic acid, analytical reagent, J. T. Baker ChemiGo. p-Toluenesulfonic acid, cal monohydrate, analytical reagent, Matheson Coleman & Bell. Reserpine, U.S.P. reference standard. A 1% solution of p-toluenesulfonic acid in glacial acetic acid was prepared. Because of the very Ion- concentration of the active material, careful cleaning of glassware must be maintained throughout the procedure. Procedure. Standard Curve. Using a microbalance, 1 mg. of U.S.P. reserpine was accurately weighed and quantitatively transferred to a 100ml. volumetric flask and diluted t o volume with glacial acetic acid. A 1-ml. aliquot of this solution was transferred to a second 100-ml. volumetric flask and diluted t o volume with glacial acetic acid. This last solution contained 0.1 pg. of reserpine per milliliter. One-, 2-, 3-, 4-, and 5-ml. aliquots of this solution corresponding to 0.1, 0.2, 0.3, 0.4, and 0.5 pg. of reserpine were pipetted into 5 test tubes (1.8 X 12 cm.). The total volume was made up to 5 ml. with glacial acetic acid, and then 5 ml. of the 1% p-toluenesulfonic acid reagent were pipetted into each tube. A reagent blank consisting of 5 ml. of glacial acetic acid and 5 ml. of the reagent was also prepared. All of the tubes were placed in the steam or boiling water bath for 10 minutes, At the end of this time the tubes were cooled to room temperature in a stream of t a p water, and the fluorescence was measured in the fluorometer. Applications. DETERMINATION IN TABLETS.Because of the very high sensitivity of the reagent, it was not necessary to use a n extraction technique except in certain specific cases
as mentioned later in the section dealing with interferences. Tablets containing 1 mg. of reserpine were pulverized, and a sample corresponding to the weight of one tablet was quantitatively transferred into a 100-ml. volumetric flask. Approximately 50 ml. of glacial acetic acid n ere added, and the resulting mixture A B S shaken for about 2 minutes, after bT-hich the flask was diluted to volume with glacial acetic acid. After allouing the mixture to settle for a few minutes, a 1-mL. aliquot of the supernatant liquid was. transferred to a second 100-ml. volumetric flask which in turn was diluted to volume \Tith glacial acetic acid. A 3-ml. aliquot of the last solution was transferred to a test tube, and the total volume was made up to 5 ml. with glacial acetic acid. Five milliliters of t h e reagent were added to the tube, and; from this point on the procedure as outlined in the preparation of the standard curve was foilon-ed. Tablets containing varying amounts of reserpine u-ere assayed in the same manner, taking care to adjust the concentration of the last dilution in the volumetric flask to about 0.1 pg. per milliliter. DETERMINATION IX AMPOULES, ELIXIRS, AND ORALDROPS. Appropriate dilutions were made in glacial acetic acid so that the final solution contained between 0.05 t o 0.5 pg of reserpine per milliliter. An aliquot of this solution was then transferred to a test tube and treated in the usual manner. DISCUSSION
Oxidation Products of Reserpine. This method does not measure the oxidative breakdocvn products of reserpine. Three solutions of reserpine, each a t the same concentration, were made up and subjected t o different light conditions. One solution was placed on a windon ledge in the sunlight. The other was kept out in the open laboratory under artificial light conditions, and the third was kept in the dark. From time to time samples were withdrawn from these solutions and assayed according to the above procedure. The results of these csperiments are shown in Table I. The solution exposed to sunlight did not exhibit any new increase in fluores-
Table 1.
Date 6/22/61 6/23/61 6/25/61 6/28/61 7/5/61 7/14/61 7/25/61
Stability Data of Reserpine Solutions Galvanometer Deflections Solution Solution Solution in in artif. in dark light sunlight 26.0 26.0 24.5 23.8 21.5 20.0 19.0
26.0 26.0 22.5 20.0 18.5 16.0 14.8
VOL. 34, NO. 3, MARCH 1962
26.0 8.8 3.2 2.8 2.5 1.8 0
411
cence after reaching the zero point. This m o d d indicate that neither the first-step oxidative products of the oxidative degeneration of reserpine, nor later steps are measured by this procedure. Dechene ( 7 ) reported in his method that 3 tenfold increase in fluorescence was noted when a solution containing reserpine was heated with hydrogen pero.;ide. Even with this modification the method is still about eight times less sensitive than the method proposed here. The addition of hydrogen peroxide t o the fluorophor formed by the procedure presented here completely destroyed the fluorescence. Standard Curve. T h e standard curve is linear in concentrations from 0.005 to 5 pg. per milliliter. The fluorescence product in such high concentrations of reserpine as 5 pg. per nil. is visible and must be dilated 100 times after heating to fall rTithin the range of the instrument. One standard curve was constructed and ma3 checked from time to time. KOdiscrepancies in reproducibility were found. Reagent. When the concentration of p-toluenesulfonic acid in glacial acetic acid was varied between 0.5 and 20%, only small variations in the galvanometer readings n-ere noted. After aging the 1% reagent for over a month, no change in its effectiveness was observed. Acetic Acid. The glacial acetic acid which was used throughout the experiments exhibited no discrepancies in reproducibility from lot t o lot. The influence of water content is discussed under the section on interferences. The concentration of chloroform in the combination with acetic acid is not critical, and no differences are noted in the fluorescence when chloroform is present in amounts up to 20%. This enables one to extract the reserpine selectively from mixtures of water soluble salts with chloroform, and then, to react the chloroform solution directly with the reagent rather than going through a n evaporation of the chloroform and re-solution of the reserpine in acetic acid. This avoids the possibility
of decomposing the reserpine during the evaporation. Heating Time. No differences in fluorescence were observed upon varying the heating period of the reaction between 5 and 30 minutes. Stability of Fluorophor. The fluorophor produced in the reaction was stable for several hours after formation. The extreme ease and simplicity of the method arises from the unique stability of the fluorophor. Other factors such as the length of the heating period, the small differences noted in varying the concentration of the reagent, and the minor effect of the concentration of acetic acid on the formation of the fluorophor constitute four major elements nhich are not critical in the reaction. Thus the ease and siniplicity of the method cannot be overemphasized. Precision and Accuracy. At a level of 0.005 pg. per ml. based on four independent determinations, the accuracy was 98.8% of theory with one standard deviation equal to zt 1%. Comparison of Methods. T h e data in Table I1 compare results obtained by assaying several different reserpine formulations by three different methods, namely: t h e assay method which is currently official in the U.S.P. XVI, the colorimetric method of Indemans, Jakovljevic, and Langerijt (9), and the fluorometric method as outlined in this paper. TKO of the formulations, reserpine tablets containing methyl testosterone and diethylstilbestrol, and reserpine tablets containing pyrrobutamine, gave exceedingly low results when assayed by the U.S.P. XVI method, which obviously cannot be used for these preparations because of the interference of other ingredients in the formulation. When our fluorometric method was applied to the extracts obtained from the extraction procedure as outlined in the U.S.P. XVI colorimetric procedure, results were in agreement with those obtained by our technique. Interferences. T h e presence of water in amounts less t h a n 2% does not affect t h e reaction, b u t as the amount of water in the system is increased the fluorescence is decreased.
Table II. Comparison of Data Theory, U.S.P. Mg. XVI Colorimetric 0.261 0.260 0.250 Reserpine tablet 0.970 1.05 Reserpine tablet aged 1.OO 0.104 0.062 Reserpine tableta 0.100 0.260 0.160 0.250 Reserpine tabletb 2.53 2.57 2.50 Reserpine ampoule 2.05 2.04 Reserpine oral drops 2.00 0.247 0.256 0.250 Reserpine elixir a Tablets contained methyl testosterone and diethylstilbestrol. b Tablets contained pyrrobutamine.
412
ANALYTICAL CHEMISTRY
Fluorometric 0.253 0.930 0.101 0.250 2.42 2.00 0.247
A standard which gave a fluorescence reading of 26.0 when reacted in t h e usual manner, gave a reading of 24.5 when the system in which it was reacted contained 7% water. Amphetamine hydrochloride, theophylline, phenobarbital sodium, caffeine, acetylsalicylic acid, ascorbic acid, methyltestosterone, diethylstilbestrol, and pyrrobutamine do not interfere with the fluorescence when present in large amounts. Calcium chloride. barium chloride, potassium chloride and ferrous chloride give no interference when present in concentrations fifty times greater than that of reserpine. Potassium ion causes negative inhibition of the fluorescence when present in large amounts. When ferric ion is present in concentrations equal to or greater than that of reserpine, it causes negative inhibition of the fluorescence also, and it is recommended that a simple chloroform extraction be used to free reserpine from potassium and ferric ions when they are present. Rutin and acetophenetidin cause negative inhibition of the fluorescence when present in concentrations fifty times greater than that of reserpine but give no interference when present in amounts equal t o reserpine. When reacted with the reagent, DL-tryptophan exhibited the same activation and emission maxima as reserpine, but its fluorescence was only 12% that of reserpine. If indole is treated with the same reagent, a red fluorescence is obtained. X few of the indole alkaloids showed the following characteristics when treated with the reagent. Alstonine hydrochloride exhibited a major activation maximum a t 310 nip with emission maximum a t 440 nip. I t had a minor activation maximum a t 380 nip with an emission maximum a t 460 mu. The emission maximum which occurs at 450 nip overlaps that of reserpine (480 mp) and gives a fluorescence about 2801, that of reserpine when measured using the Turner filter combination. Pseudoreserpine had the same activation and emission spectrum as reserpine and gave approximately the same amount of fluorescence. Ajmalicine exhibited the same activation and emission spectrum, but the fluorescence was about 24% that of reserpine. Ibogamine and rauneseine had the same activation and emission spectra, but the fluorescence n as about 6% that of reserpine. Deserpidine, rescinnamine, sarpagine, and rauwolfine gave no interferences !Then treated with the reagent. ACKNOWLEDGMENT
The authors are indebted to Norbert Neuss and Marvin Gorman for offering their counsel and providing various samples.
LITERATURE CITED
( 1 ) Banes, D., J . Am. Pharm. Assoc. 44,408 (1956). ( 2 ) Ibid., 46, 601 (1957). ( 3 ) Bartelt, W. F., Hamlon-, E. E., Ibid., 44, 660 (1955). (4) Bayer, J., Magyar Ktm. Folydirat 62, 355 (1956). ( 5 ) Bayer, J., Phaunazae 8,468 (1958). ( 6 ) Booth, R. E., J . Am. Pharm. Assoc. 44,568 (1955).
Dechene. E. B.. Ibid.. 44. 657 (1955). (8) Haycock, R. P:, Mader, ‘W. J.; Ibid., (71
of Biological Chemistry 126th Meeting,
I. M., Langerijt, J. J. A. M., Pharm.
ACS, pll’& York, September 1954. (13) Sakal, E. H., Merrill, E. J., J . Ana. Pharm. Assoc. 43, 709 (1954). (14) .Sxalkowski, C . R., Mader, ST. J., Ibad., 45,613 (1956). i 15) Wunderlich, H., Pharm. Zentralhalle 96, (2), 68 (1957).
hlissan, S. R., et al., J. Am. Pharm. Assoc. 44,446 (1955). (12) Poet, R. B., Kelly, J. hf., Division
RECEIVED for review October 12, 1961. Accepted January 8, 1962.
46,744 (1957). (9) Indemans, A. W. M., Jakovljevic,
Veekblad 94, 1 (1959). (10) Leemann, H. G., Weller, H., Helv. Chzna. Acta 43, (5), 1359 (1960). (11) blcillullen, W. H., Pazdera, H. J.,
Determination of Ultimate Capacity of Weakly Acidic and Mixed Acid Cation Exchange Resins JOHN UNGAR Technical Services (Iabs.), Ionac Chemical Co., A Division of Pfaudler-Permufit Inc., Birmingham, N. J .
b A method is described for the determination of both strongly and weakly acidic functional groups, in organic cation exchange materials. The resin is converted fully to the hydrogen form by an excess of HCI; strong acid groups are determined by titrating the acidity liberated by a 5% sodium chloride solution (methyl orange end point), and weak acid groups are determined by titrating the acidity liberated by a 1% disodium hydrogen phosphate solution (phenolphthalein end point). Typical results are quoted, and the reproducibility of the method is illustrated.
may be applied, include the condensation polymers, as well as materials of the methacrylic or acrylic acid type. I n addition, there are the carbonaceous materials, such as sulfonated coal, which may contain various proportions of weak acid groups in addition to the sulfonic acids. Both granular and bead type materials having sulfonic, as well as Carboxylic, groups in their structure have been reported, and for such materials, it is very important to determine the relative proportions, as well as the ultimate quantities, of strong and n-eak acid functional groups. PRINCIPLE
T
HE T o r h L CAPACITY of strongly acidic cation exchangers, such as the polystyrene sulfonic acid type, or the sulfonated condensation polymers, is usually determined by regenerating the material with a n excess of a mineral acid, rinsing out the excess acid, displacing the hydrogen ions by means of a strong solution of sodium chloride or sodium sulfate, and titrating the resultant acid. This method is widely known and used; it has been referred to in the literature, for esample, by Calmon and Kressman ( I ) , Kachod and Schubert (6), and Kunin ( 5 ) . The determination of the ultimate capacity of cation exchange materials containing weakly acidic groups has apparently been neglected in the literature, although Kunin (4, 6) does refer to the admixture of sodium hydroxide to the usual sodium chloride solution, in order to obtain complete conversion of cation exchange materials from the hydrogen to the sodium form, and Helfferich (2) does treat the matter from a purely theoretical standpoint. Commercially available resins to .which such a method of determination
OF
METHOD
A strong alkali, such as sodium hydroside, would be unsatisfactory in a general method of this kind, because it would cause chemical and physical damage to materials of the carbonaceous or of the phenol condensation type. Sodium bicarbonate is also unsatisfactory. Since carboxylic acid groups attached to this type of resin network tend to eschange only the cations of weak acid salts, it was necessary to find a convenient weak acid salt which would be eschanged with a reasonable efficiency and a t a reasonable reaction rate, so that a determination would not become too cumbersome. Disodium hydrogen phosphate solution a t a strength of 1% w./v. of the anhydrous salt was adequate. The method described combines the sodium chloride value determination with the phosphate value determination. PROCEDURE
Twenty to twenty-five milliliters of the ion exchanger in question are measured under water in a graduated cylinder, by tapping it gently against
a bench or by placing it in a high frequency vibrator, until minimum volume (let this be V ml.) is obtained. This measured quantity of material is then transferred quantitatively to the prepared chromatographic tube of approsimately 1.0 cm. i.d., and the resin bed is backwashed with a n upward flow of water, to remove extraneous dirt or any air bubbles trapped between the resin particles. After backwashing, the material is allowed t o settle, and five bed volumes of 2N hydrochloric acid, c.P., are passed through at such a rate, t h a t the last of this acid will drain to the level of the ion exchanger bed within 30 to 40 minutes. When the last of the acid has drained to bed level, 15 to 20 ml. of distilled water or demineralized water are passed through a t the same rate, followed by more distilled or demineralized water a t a rate approximately 3 times that of the acid passage, until the effluent from the tube is neutral to methyl orange indicator. Now, 250 ml. of 5% w./v. sodium chloride solution, c.P., are passed through, a t the same rate as the acid. The effluent is collected, mixed well, and 25 ml. are titrated against 0.1N standard sodium hydroxide solution to the methyl orange end point. A further 50 ml. of 595 sodium chloride is now passed through a t the same rate. If this effluent requires less than 0.5 ml. of 0.1N sodium hydroxide solution to the methyl orange end point for a 5-ml. aliquot, one may proceed to the next stage. If the last mentioned titration figure is greater than 0.5 ml., then further 50-ml. quantities are passed through until the titration figure for the sodium chloride effluent. is less than 0.5 ml. The sum of these titration figures is noted, let it be A ml. Then, 250 ml. of 1% w./v. disodium hydrogen phosphate solution, prepared from the C.P. anhydrous material or an equivalent quantity of C.P. crystallized disodium salt, is passed through in approximately 30 minutes. This effluent VOL. 34, NO. 3, MARCH 1962
413
