Table II.
Analysis of NBS Samples of High-Alloy Steel
Tungsten Molybdenum. PresPresent, Found, ent, Found, Sample % % % % 18.05 18 50b 0.401 0.40 6.29 6.2 132 7.07 7 . 2 6.3 7.1 6.20 6.1 132a 4.51 4.5 6.0 4.5 1.82 1.8 134 8.68 8 . 7 1.8 8.8 2.00 1.9 134a 8.35 8 . 3 2.0 8.3 2.0 8.4 1.4 1.58 153 8.38 8 . 3 1.5 8.5 1.4 8.3 155 0.039 0.035" 0.517 0.51" 0.50" 0,040" a Analyzed as a corrosion-resisting alloy but 5 ml. of niobium solution added in place of 1 ml. of HF. duction of the Biz03 occurs. I n attempting to overcome this difficulty i t was reasoned that if the Biz03 is added on top of the borax, the carbon in the bottom of the crucible will have been air-oxidized by the time the Biz03begins to fuse. This proved to be true but unfortunately the method had to be abandoned because a n appreciable amount of the Bi was lost by volatilization. It is evident that if Biz03 is t o be used it must be covered by the borax.
I n the proposed method BaO has been recommended since it cannot be reduced under the conditions of the borax fusion. The addition of the B a 0 t o the standard and sample disks markedly reduces the error due t o radiation absorption in the analyses but does not entirely eliminate it. However, the small negative error that remains can be largely compensated for by the use of composite calibration samples in which the refractory metal environment approximates that existing in the samples t o be analyzed. This makes it possible t o obtain good straight line calibration graphs and excellent accuracy in the analysis of the refractory metals (Tables I and 11). Other Analyses. Excellent chemical separation-borax disk methods have been developed for t h e determination of M o in low-alloy steel and ferromagnetic alloys and of W or Zr in cathode nickel. Moreover, in certain instances, where the refractory metal content is sufficiently high, where no line interference is encountered, and where the composition of the alloy does not vary too much, it has been possible t o determine Mo, W, Nb, Ta, Zr, or V by the borax disk method without resorting to chemical separations. To do this a 0.2-gram sample is dissolved in 2 ml. of "03 (1 1) plus a drop of HF in a covered Pt crucible. After evaporation, ignition a t 500" C., and
+
fusion with 10 grams of borax, the disk is analyzed. The standard disks required are prepared by fusing appropriate amounts of oxides of the matrix metals and of the metal in question with 10 grams of borax. ACKNOWLEDGMENT
The author expresses his appreciation t o Herbert Schreiber and Shirley AT.. Vincent for helpful discussions and suggestions in the x-ray work of the present investigation. LITERATURE CITED
1) Am. SOC. Testing Materials, Philadelphia, Pa., "ASTM Methods of Chemical Analysis of Metals," p. 137,
(1960).
(21 Ibid., pa 176. (3) Bandi, W. R:, United States Steel Corp., Monroeville, Pa., private communication, August 1962.
(4) Bandi, W.R., Buyok, E. G., Lewis, L. L Melnick, L. AI., ANAL. CHEM. 33, 1275 (1961). (5) Bruch, J., Arch. Eisenhiittenw. 33, 5 (1962). (6) Claisse, F., Norelco Reporter 4, 3 (1957). (7) Rothmann, H., Schneider, H., S i e buhr, J., Pothmann, C., Arch. Eisenhtdfenw. 33, 17 (1962). (8) Tomkins, bl. L., Borun, G. A , , Fahlbusch, W. A., ANAL.CHEM.34, 1260 (1962). (9) Yagoda, H., Fales, A. A , , J.Ani. Cheni. SOC.60, 640 (1938). RECEIVEDfor review June 12, 1962. Accepted October 31, 1962.
The Amperometric Determination of Amines Titration of Sympathomimetic Amines with Sodium Tetra phenylborate EDWARD SMITH,' LEE F. WORRELLt2 and JOSEPH E. SINSHEIMER College of Pharmacy, University o f Michigan, Ann Arbor, Mich. In order to investigate the amperometric titration of amines with sodium tetraphenylborate (TPB) a series of sympathomimetic amines was analyzed. Two general techniques were used. The first was based upon anodic depolarization a t a dropping mercury electrode by excess TPB titrant, while the second procedure was based upon electrochemical oxidation of the TPB ion a t a graphite electrode. The ability to determine sympathomimetic amines b y these techniques can b e correlated to the structure of the amines and solubilities of amine-TPB salts. Amperometric titrations can b e applied directly to amines even in complex mixtures provided their TPB salts have a solubility in the order of
58
ANALYTICAL CHEMISTRY
6 X 10-4M or less. These titrations were found to b e more convenient, rapid, and precise than other procedures currently in use.
I
of this investigation t o study the quantitative analytical application of sodium tetraphenylborate (TPB) to the determination of pharmacologically active amines. A general method was sought for the determination of amines in aqueous solutions and in complex mixtures, without tedious prior separations. Sympathomimetic amines were chosen for study because of their widespread use per se and in combination with T WAS THE PURPOSE
other pharmaceutical agents and because of the range of structures available. A summary of the extensive literature on the application of sodium T P B for chemical analysis has been included in a comprehensive review by Flaschka and Barnard (6) and in the bibliographies of Barnard and Buechl ( 3 ) . Precipitation of amines by T P B has been described by Crane (2) and by Howick and Pflaum (IO). Work in these laboratories (16) indicated ap1 Present address, Division of Pharmaceutical Chemistry, Food and Drug Administration, Washington, D. C. * Present address, College of Pharmacy, University of Texas, Austin, Tex.
plicability of sodium T P B to some sympathomimetic amines. Amperometric titrations of potassium a t the dropping mercury electrode involving T P B have been developed by Kemula and Kornacki (11), Findeis and DeVries (6), and Heyrovsky (7, 8). The first direct titration of potassium with sodium T P B by Amos and Sympson (1) was based upon anodic depolarization of a dropping mercury electrode by excess TPB. A modification of this procedure initiated our investigation of amperometric titrations of amines. This determination of amines was also advanced by the discovery of the electrochemical oxidation of TPB a t the graphite electrode as reported by Smith et al. (13). EXPERIMENTAL
Apparatus. Potentials were applied and currents measured during the titration with a Fisher Elecdropode manual polarograph. Either a conventional dropping mercury or a waximpregnated graphite electrode (4, 9, 12) was used as the indicating electrode. A saturated sodium chloridecalomel electrode was used as an external reference. Titrations were performed in a temperature-controlled cylindrical vessel as illustrated in Figure 1. A Beckman Model G pH meter was used for all pH measurements. Ultraviolet measurements were made on a Beckman DU spectrophotometer. Reagents. The titrant solution was prepared by dissolving 17.11 grams of sodium tetraphenylborate reagent (J. T. Baker) in 500 nil. of distilled water. The cloudy solution was filtered twice through M'hatman KO,5 filter paper to obtain a clear 0.1-V solution which had a pH of approvimately 9 and was stable over a period of three months. The sodium TPB solution was standardized by amperometric titration Tvith aliquots of O.1Jf solutions of reagent grade potassium salts. A liter of the p t € 4.6 buffer solution was prepared from 0.8 mole each of reagent grade acetic acid and sodium acetate. Standard and test solutions were prepared by dircct weighings of reagent grade chemicals and sympathomimetic amines dried to constant weight. Dropping Mercury Electrode (D. M.E.) Procedure. X sample of a dried, weighed salt or an aliquot of R 0.1M solution equivalent to 1 x 10-4 to 6 x l o w 4mole of amine n-as added to 50 ml. of buffer solution in the titration v w e l and mixed with aid of a magnetic st,irrer. The dropping mercury electrode and saturated sodium chloridecalomel electrode tips were immersed in the solution through a rubber stopper that covered the vessel. A potential of 0.08 volt was applied, with the dropping mercury electrode as the positive electrode. The solution was kept a t a temperature of 2.5' f 0.1' C. Deaeration of the titrate solution was not required because oxygen does not interfere a t this applied potential. Sodium TPR titrant was added in
small volumes from a 10-ml. buret (about 0.25- to 0.50-ml. increments before the end point and 0.05- to 0.10ml. increments after the end point). The solution and precipitate were stirred for 30 seconds after each addition of titrant. Current readings were recorded 1 minute after the stirrer was stopped. The midpoint of the galvanometer deflection due to the mercury drop was taken as the current reading. Currents were corrected for dilution, the corrected values were plotted uersus milliliters of titrant, and the end point was located by the usual extrapolation procedure. Direct Graphite Electrode Procedure. This procedure was identical to t h a t described for the dropping mercury electrode except t h a t a graphite electrode, the ends of which had been freshly lathed and wetted. was used as the indicator electrode. A potential of 0.55 volt was applied with the graphite being the positive electrode. Also, in contrast t o the D.M.E. procedure, it was possible to stir the solution and precipitate continuously for an interval of 1.5 minutes between the addition of titrant and current readings. Residual Titration. This procedure differed from the direct procedure in t h a t standard sodium T P B solution was added slowly t o the sample in buffer solution until there was about 2 ml. in excess. T h e solution was stirred for ten minutes after all the sodium T P B had been added: then electrodes were immersed, and a potential of 0.55 volt was applied. Standard potassium chloride solution was added in small volumes from a 10ml. buret (about 0.05- to 0.10-ml. increments before the end point and 0.25to 0.50-ml. increments after the end point), until a current plateau was reached (usually 5 ml. after the end point). The solution and precipitate were again stirred continuously and, after each addition of titrant, current readings were taken a t intervals of one minute. The equivalence point was then determined by the usual eutrapolation procedure. A large e w e s of sodium T P B is unnecessarT- since current values obtained more than 1.5 ml. before the end point showed deviation from a linear plot. Solubility Measurements of Amine TPB Salts. Fifty milliliters of O.lAT sodium T P B solution were added slowly, with constant stirring, to 5 x lo-' equivalent of amine or its acid salt dissolved in 50 ml. of acetate buffer. The solution and precipitate were stirred for about 30 minutes after the sodium T P B was added, The sloxer reacting phenol derivatives required one to several hours to achieve precipitation. Because catecholamines are easily oxidized, the buffer solution was saturated with nitrogen and precipitation was performed under an atmosphere of nitrogen. Precipitates were filtered on a Buchner funnel, washed free of excess T P B with several portions of distilled water, and dried t o constant weight over calcium chloride in a vacuum oven a t 30" C.
Approximately I-gram samples of these tetraphenylborate salts together with 50 ml. of acetate buffer were placed in a 2-ounce amber bottles with poly-seal lined tops. The bottles were placed in a water bath maintained a t 25' f 0.5" C. and agitated periodically. Samples were removed, filtered, and absorbances of the filtrates read a t 266 and 274 m,u. Solubilities were then calculated from the molar absorptivities a t both of these wave lengths. RESULTS AND DISCUSSION
Development of Procedures. Precipitation of amine T P B salts depend upon the presence of protonated amines. As strong acid solutions will decompose the T P B ion rapidly (14, buffers of pH 4 to 6 are in general the most satisfactory for the protonation and precipitation of amines as TPB salts. I n this investigation, the equimolar acetic acid-sodium acetate buffer used by Amos and Sympson (1) was retained for its simplicity, p H of 4.6, high buffer capacity, and good supporting electrolyte property. Amos and Sympson recognized that a major disadvantage of their method was the time required for a constant current value after each addition of titrant and suggested t h a t readings a t fixed intervals after the addition of titrant may be of value. Equal time intervals varying from 1 to 8 minutes were investigated in our titration of amines. A 30-second stirring period with a 1-minute waiting period was then chosen to shorten titration time. These conditions were compared t o the constant current conditions of Amos and Sympson and no appreciable changes in plotted equivalence points were observed. When the graphite electrode was used, a comparison of results indicated that it was unnecessary to stop stirring after the addition of reagents. Current values were much higher with constant stirring because of convection as me11 as diffusion current, but excellent results were still obtained 7%-henequal time intervals of 1 or 1.5 minutes after each addition of titrant were used before the current TVSS recorded. J?'ith some compounds, all the current values before the end point did not occur on a plateau. This anomaly appeared to be a solubility phenomenon, as the current resumed its normal value after the supernatant solutions became clear and precipitates coagulated. If the extraneous points on the current plateau were disregarded, the equivalence point could be determined or residual titrations eliminated the extraneous current readings, and gave very satisfactory results. Residual titrations were also used in those instances were precipitation was not instantaneous and no current VOL. 35, NO. 1, JANUARY 1963
59
Table 1.
Determination of Sympathomimetic Amine Salts
Compound Aliphatic amine salts Cyclopentamine hydrochloride Isometheptene mucate Methylhexaneamine hydrochloride Propylhexedrine hydrochloride Tuaminoheptane sulfate Phenylethylamine salts Amphetamine sulfate Amphetamine sulfateb Amphetamine sulfatee Dextro-amphetamine sulfate Dextro-amphetamine sulfate* Diethylpropion hydrochloride Mephentermine sulfate Methamphetamine hydrochloride Methoxyphenamine hydrochloride Phenmetrazine hydrochloride 8-Phenylethylamine hydrochloride Phenylpropylmethylamine hydrochloride Imidazole salts Naphazoline hydrochloride Tetrahydrozoline hydrochloride Benzylic hydroxyl amine salts Ephedrine sulfate EDhedrine sulfated Jfethoxamine hydrochloride Phenylpropanolamine hydrochloride
Sample range, - , mg.
Average Range recovered,= recovered, Standard deviation 7% 70
17.8-89.1 49.2-132.1 22.9-76.2 19.2-96.0 16.4-98.2
100.8 101.1 99.4 102.1 99.9
0.8
1.1
11.38 10.71 10.34 10.59 10.35
18.4-92.4 18.5-111.5 18.5-92.4 18.5-92.6 18.5-147.5 24.2-121.0 23.1-115.3 37.4-92.8 43.3-108.2 35.5-88, 8 15.8-79.0
100.0 99.6 100.3 100.1 99.8 100.8 100.4 99.9 99.7 100.1 100.2
2.3 1.7 2.0 2.6 1.4 2.7 1.5 0.9 2.2 2.0 2.3
10.71 10.52 10.61 10.79 10.47 10.90 1 0 .75 10.27 10.74 10.74 10.75
18.7-93.3
100.2
2.0
hO.71
24.7-123.4 23.7-118.5
99.2 100.1
4.0 3.9
11.24 11.11
42.9-107.2 42.9-107.2 24.8-i21 1 37.5-93.9
100.1
2.3 2.1 3.0 2.8
1 0 .84 1 0 .67 1 0 .82 hO.89
99.7 100.6 100.1
4.6 1. 7
1.9
ing substances could be ions or ligands that form stable complexes or precipitates with mercurous or mercuric ions. For example, halides, cyanide, thiocyanate, and sulfide could give anodic currents which would obscure the anodic current due to the TPB. Amos and Sympson (1) found that there was no interference in the determination of potassium when chloride concentration was 0.24M or less. Homever, greater than 1% error was reported with bromide or iodine in l O - 3 M concentration. The ions which react with mercury ions do not interfere with the direct electrochemical oxidation of T P B ions at the graphite electrode (13). Thus, in addition t o the general advantage of the use of a solid electrode, the graphite electrode procedure is particularly useful when these ions are present, as for example, in the form of an amine salt. With amperometric titration procedures, i t is possible to determine sympathomimetic amines in combination with other pharmacological agents without tedious separation techniques. The presence of aspirin, acetophenetidin, and phenobarbital which are
Average of ten determinations. With the dropping mercury electrode. In the presence of aspirin and acetophenetidin. In t h e presence of phenobarbital. Table
II.
Unsuccessful Determinations
of Sympathomimetic Amine Salts
plateau could be obtained by direct titration. Applicability of the Amperometric Table I summarizes Procedures. compounds successfully analyzed by these procedures. I n Table I1 are listed compounds where stoichiometric results could not be obtained. T h e ability to determine sympathomimetic amines by these amperometric titration techniques was markedly affected by changes in structure and could be correlated t o the solubility of sympathomimetic amine T P B salts. I n general, the order of solubility of T P B salts as well as the ability t o determine sympathomimetic amines is as follows: aliphatic amines = phenylethyl amines > compounds n-ith a benzylic hydroxyl > phenolic derivatives > catechol derivatives. From solubility data listed in Table I11 can be seen the definite effect of an hydroxyl group upon the solubility of T P B salts of sympathomimetic amines. The benzylic hydroxyl group approximately doubles solubility while addition of phenolic hydroxyls increases solubility 5 t o 10 times t h a t of the parent compound. Methamoctol, a n aliphatic amine with a hydroxyl, has a solubility about four times that of other aliphatic amines. 60
ANALYTICAL CHEMISTRY
Amperometric titrations of these sympathomimetic amines were successful with samples of 1 x 1 0 - ~to 6 X lo-* equivalent of amine in 50 ml. of acetate buffer if the amine formed a precipitate with a solubility of 6 X 10-4M or less. Greater solubilities for the T P B salts of phenolic amines, catecholamines and for the T P B salt of methamoctol can be related to the low or negative results obtained in the titration of these amines. Also no T P B salt could be obtained from the phenolic amine, nylidrin. Presence of the bulky phenyl isobutyl group on the nitrogen would serve as an explanation of this behavior. Interfering substances which must be absent t o obtain stoichiometric results have been placed into three categories ( 1 ) . The first group is inorganic cations such as ammonium, potassium, cesium, rubidium, thallium (I), silver, and mercury ions which react with sodium T P B to form insoluble salts. The second group would be ions which could be reduced or oxidized at the potential used in the titration. Use of the D.M.E. procedure, which is effective at a lower potential, may eliminate this interference. When a dropping mercury electrode is used with TPB, the third group of interfer-
Compound
Results
Aliphatic amine salt Nethamoctol Soend mucate point Phenolic amine salts Hydrosyamrthetamine hydrochloride Low Rletarminol bitartrate Low K ylidrin Soend hydropoint chloride Phenylephrine Low hydrochloride Synephrine Low tartrate Low Tyramine hydrochloride Catecholamines Epinephrine Isoproterenol sulfate
Soend point Soend point
Type Of precipitation Amorphous"
Delayedh Delayed Sone Delayed Delayed Delayed
Sone Sone
A gummy precipitate coated the electrodes and titration vessel. b Delayed precipitation indicates that a precipitate formed only after about 1 hour n4th residual titrations and direct titrations gave no end point. 0
~
Table 111. Concentrations of Saturated Solutions of Tetraphenylborate Salts in Acetate Buffer a t 25” C.
Compound Aliphatic amine salts Cyclopentamine Isometheptene If ethamoctol Methylhexaneamine Propylhexedrine Tuaminoheptane I’henylethylamine salts Amphetamine D iethylpr opion Mephentermine Methamphetamine Methoxyphenamine Phenmetrazine Phenylethylamine Phenylpropylmethylamine Imidazole salts Naphazoline Tetrahydrozoline Benzylic hydroxyl amine salts Ephedrine Methoxamine Phenylpropanolamine Phenolic amine salts Hydroxyamphetamine Metarminol Phenylephrine Synephrine Tyramine Catechol amine salts Epinephrine Isoproterenol
Concn. mole/l. 1 . 4 x 10-4 1.2 7.8 1.6
r
1.0 1.5
U
2.8 0.4 1.2 2.0 1.2
___.
0.5 1.3 1.2
E
0.2 0.3
€3 t,
6.0 5.7 5.9 12.5 14.0 12.0
12.0
12.4
G
Figure 1. A. 6. C.
D. P.
15.9 14.5
F. G.
H. 1.
J.
conimonly dispensed with these amines did not affect the stoichiometry. The usual tablet formulation ingredients also did not have adverse effects. These techniques can be applied to other basic nitrogen compounds if a TPR salt can be formed rapidly and has a solubility in the order of 6 X 10-4df or less. The amperometric titration procedures were more convenient, rapid, and precise than most procedures currently being used for the determination of amines in mixtures. LITERATURE CITED
W.R., Sympson, R. F., ANAL. CHEM.31, 133 (1959). ( 2 ) Crane, F. E., Ibid., 28, 1794 (1956). ( 3 ) Barnard, A. J., Jr., Buechl, H., (‘henLst-Analyst 48, 44 (1959). (1) Amos,
,.L!-.
r
Titration apparatus and thermostated vessel Working electrode (dropping mercury or graphite) Saturated sodium chloride-calomel electrode Sodium chloride-agar gel in tygon tubing Buret Sample and buffer solution Magnetic stir bar Magnetic stirrer Chamber with flowing constant temperature water Air space Rubber stopper
(4) Elving, P. J., Smith, D. L., ANAL. CHEM.32, 1849 (1960). (5) Findeis, A. F., DeVries, T., Zbid., 28,1899 (1956). (6) Flaschka, H., Barnard, A. J., Jr., “Tetraphenylboron as An Analytical Reagent” in “Advances in Analytical Chemistry and Instrumentation,” pp. 1-117, Vol. I, C. N. Reilley, ed., Interscience, Xew York, 1960. (7) Heyrovsky, A., Coll. Czechoslov. Chem. Comm. 24, 170 (1959). (8) Ibid., 26, 1305 (1961). (9) Horyna, J., Jehlickn, Ir.)Ibid., 25, 1769 (1960). (10) Howick, L. C., Pflaum, K. T., Anal. Chim. Acta 19, 342 (1958). (11) Kemula, W., Kornacki, J., Roczniki Chem. 28 (4), 635 (1954); C. -4.49, 8733 (1955).
(12) Morris. J. B., Schempf, J. M,, Anal. Chem.31,286(1959). (13) Smith, D. L., Jamieson, D. R., Elving, P. J., Ibzd.,. 32, 1253 (1960). (14) Wittig, G., Keicher, G., Rlickert, A., Raff, P., Ann. 563, 110 (1949). (15) Worrell, L., Ebert, W. R., Drug Std. 24, 153 (1956). RECEIVEDfor review August 14, 1962. Accepted Sovember 13, 1962. Division of Analytical Chemistry, 141st Meeting, ACS, Washington, D . C., March 1962. Abstracted in part from a thesis submitted to the Graduate School, The University of Michigan, by Edward Smith in partial fulfillment of the requirements of the Ph.D. degree, February 1962.
VOL. 35, NO. 1 , JANUARY 1963
61
