strong background color from either boron or zinc. The dye concentration is kept Ion- t o minimize background cdor. Improved color stability and better chchting properties are obtained by nitrating the Eriochrome Cyanine R. The nitrous arid formed in the reaction ('GUSCS instability of the dye solut'ion by reduction of the chromophore group ( 1 ) . Sitruns :icid is removed by the addition of urea or sulfamic acid. Dj.e solutions VI trt7ated have rcniained stable for 6 months or longer. RASGI: OF ALUMISC_RICOXCESTRAT I ( I S . ( )ptimuni aluminum concentration is aliout 10 y in the cdored coniplcs, :inti 20 y give a n absorbance of 1. O n a 5-1111. aliquot. this rcpresents 0.050 and 0.1005 aluniinum, respectively. K i t h Iiiglicr ronccntrations, R minller aliquot ShCJ11ld lie processed. DISCUSSION
M e t h o d B. This method is applicable to 1017- alloy and carbon steel. I n high alloy steels, chromium when present blocks t h e formation of the aluniinuiii complex, and vanadium contributtb n positive error of 0.12y0 aluminum for lYc of mnadium present. Chromium ni:i>-lw rcnioved n-ith fuming
perchloric acid aiitl tlrp hydrogen chloride, and the needed correction may be applied for vanadium when the concentration is known, thus enabling a determination of aluminum in high alloy steels by the direct method. The formation of the complex in perchloric acid is sloner than in nitric acid, and the dye must lie added before ascorbic acid t o utilize the calibration obtained with nitric acid. RESULTS
Table I Phons tlic results obtained by the caustic separation of Method A with high and lo^ alloy steels. Because there are no suitable standards to demonstrate the applicability of the method to alloys, knonn amounts of aluminum were added to these steels as shown. Good agreement nas obtained m-ith all types of steels. Table I1 shows the results from the direct photometric method on low and high alloy steels. Correction for vanadium was made and chromium was remoyed as described above in the Sational Bureau of Standards samples 5Ob, 123a, and 153. Table I11 shows the repeatability of the aluminum determination on a single sample carried out over a number of
Table 111.
Repeat Analyses of Electric Steel
Aluminum,
%
0 0 0 0 0 0 0 0 0 0
Dev. from >lean 0 000 0 000 -0 001 0 000
060 060 059
060 061 062 059 060
+0 001 $0 002 -0 001 0 000 -0 001 0 000
059
060 Mean 0 060 Std. dev.
0 0O09
days with Method A. Thr awrage deviation iq less than 0.001% aluminum. LITERATURE CITED A s . 4 ~ .CHEII. 28, 1419, (1956). (2) Ikenberry, L. C., Thomas, h b a , Ibid., 23, 1806 (1951). (3) Richter, F., 2. anal. Cheiii. 126, 426 (1944). (4)Seuthe, Adolf, Stahl it. Eisen 64, 493 (1944). RECEIVEDfor review 1Iarch 31, 1958. hccepted October 27, 1958. Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, Xarch 1,158.
(1) Hill, U. T.,
Polarographic Oxidation of Phenothiazine Tranquilizers PETER KABASAKALIAN and JAMES McGLOTTEN Chemical Research and Development Division, Schering Corp., Bloomfield,
)Tranquilizing drugs, derived from phenothiazine by substitution of alkyl tertiary amines a t the 10-position and containing various substituents a t the 2-position, give regular, reproducible, and well defined anodic voltammetric waves at a gold microelectrode. The voltammetric waves of 14 tranquilizers have been studied to determine their analytical applicability. The effects of pH, concentration, and temperature on the position and size of the waves as well as the effects of substituents in the 10- and 2-positions on half-wave potentials are discussed. The voltammetric oxidation of these compounds in aqueous sulfuric acid compares favorably with existing methods of analysis.
T
H.4XQUILIZISG drugs derived from phenothiazine by substitution of alkyl tertiary amines a t the 10-position may have electron-withdrawing groups at the 2-position. A common property
N. J.
is their susceptibility to oxidation, which is attributed to the phenothiazine nucleus, resulting in the formation of sulfoxides ( 3 ) . Generally, they are assaj ed by one of the following methods: ultraviolet ( 4 ) and infrared ( 5 ) absorption spectroscopy; colorimetry with sulfuric acid ( 2 ) or ferric chloride (7'); bromination (6) ; and aniperometric titration n ith tungstosilicic acid ( 1 ) . It was found in this laboratory that phenothiazine tranquilizers give oxidative voltammetric waves a t a solid microelectrode and these waves have been investigated to determine their analytical usefulness. The follon ing tranquilizers n ere examined: 10- [3-(4-p-hydroxyethyI-l -piperazinyl)propyl]phenothiazine (I); 10[3 - (4 - p - hydroxyethyl - 1 - piperazinyl)propyl]-2-chlorophenothiazine (11); 10 - [3 - (4- p - hydroxyethyl - 1 - piperazinyl)propyl]-2-acetylphenothiazine (111) ; 10- [3-(4-p-hydroxyethyl-l-piperazinyl)propyl] - 2 - (trifluoromethy1)-
phenothiazine (IT') ; 10-(3-dimethylaminopropy1)phenothiazine (T-) ; 10-[(lmethyl - 3 - piperidy1)rnethyljphenothiazine (VI) ; 10- [2-(1-pyrrolidiny1)ethyl ]phenothiazine (VII) ; 10-(2-dimethylaminopropyl) p h e n o t h i a z i n e 10-(2-diethylaminoprnpyl)(VIII) ; phenothiazine (IX); 10-(3-dimet hylaminopropyl) - 2 - chlorophenothiazine 10- [3- ('$-met hyl-1-piperaz inyl)(X) ; propyl]-2-chlorophenothiazine(XI) ; 10(3-dimethylaniinopropyl) - 2 - acetylphenothiazine (XII) ; 10- [3-(4-methyl1 - piperaziny1)propyll - 2 - (trifluoromethy1)phenothiazine IXIII) ; and 2diethylaminoethyl - 10 - phenothiazinecarboxylate (XIV).
R? = hydrogen or electronegative group. Rlo = tertiary alkyl amine. VOL. 31, NO. 3, MARCH 1959
431
Table I.
Half-Wave Potentials and Current Constants 0.1N Sulfuric Acid Aqueous Alcoholic Compound Eu2a id/Cb Eiiz id/C v 0.473 63 0.545 29 VI 0.503 60 0.600 28 s 0.541 66 0.636 29 XI 0.547 57 0.620 32 I1 0.550 56 0.619 28 VI1 0.567 58 0.635 31 XI1 0.583 47 0.684 24 VI11 0.619 63 0.696 31 IX 0.620 51 0,724 27 XIV No wave S o wave a Volt us. N.C.E. b Microamperes per millimole.
of Tranquilizers
0 0 0 0
550
585 641 588 0,607 0 626 0 675 0 639 0 585 S o nave
38 35 38 41 42
52
28 56 35
EXPERIMENTAL
Table II. Effect of 2-Substituents on Half-Wave Potential in Alcoholic Sulfuric Acid Solution
R2
Rio
H
Eiiz
H
c1
HEPPQ 0 619 COCHs HEPP" 0 660 CF, HEPPa 0 714 a 3 - (4 - p - Hj-droxyethyl - 1- piperazi-
ny1)propylphenothiazine.
Table 111. Effect of 10-Substituents on Half-Wave Potential in Aqueous Sulfuric Acid Solution ComR2 Ria E,,z pound v H -C-N- n- 47.1 -I H -C3-N0 482 VI H -C3-S0 503 1-11 H -C2-S0 567 VI11 H -Cv-X0 619 IX H -C2-h7-0 620 - Y
Table IV. Effect of pH on Half-Wave Potential in Alcoholic Solution
Compound VI11
a
pH 1 5 5.5 6 9 9 5 11 5 5.5 6 8 8 7" Rave splits at pH 9.5.
E112
0 696
0.639
0 0 0 0 0 0
Table V.
a
590
510 545
550 540 536
Apparatus. T h e Sargent (E. H. Sargent & Co.) polarograph Model XXI was used throughout this work. T h e polarographic cell was a small H-cell with a mercury-mercurous sulfate reference electrode. The oxidations were carried out a t a rotating gold mire electrode (4 nim. in length and 1 mm. in diameter, sealed in a vertical, off-center position a t t h e end of a glass rod) using a constant speed Sargent synchronous rotator (600 r.p.m.). Platinum electrodes did not give well defined S-shaped oxidative waves. All pH measurements were made F\-ith the Beckman Model G p H meter. Cell resistances were measured with an Industrial Instruments, Inc., conductance bridge IIodel R C 16. The electrolysis cell in the controlled potential electrolysis work was similar in design to that described by Swann (8). However, a gold wire 2 feet in length and 1 mni. in diameter was used as the anode and the porous cup served as the cathode. All ultraviolet spectra m-ere obtained in methanol with the Cary recording spectrophotometer, Model 11. Electrode Preparations. Bt t h e end of a day's use, t h e gold electrode was immersed in hot concentrated nitric acid for 15 minutes. After washing with distilled water, i t was left in distilled water overnight. Millicoulometry. Approximately 5 pmoles of the desired compounds mere oxidized in 0.1-br aqueous sulfuric acid a t the rotating gold electrode until 30 to 40% of the material was oxidized. Polarographic waves were obtained before and after the electrolysis to determine the percentage of material
Effect of Substituents on Ultraviolet Absorption Spectrum in Methanol R2 R~Q xIn*xa log e H H 252 4 66 H HEPP 254 4 51 c1 H 256 4.70 c1 HEPP 257 4 56 COCHI H 244 4 38 HEPP 243 4 28 COCHI H 259 4 63 CF3 HEPP 259 4 53 CF, hIillimicrons.
432
ANALYTICAL CHEMISTRY
oxidized. The millicoulombs consumed during the reaction were determined by integrating the area under the currenttime curve produced by the recorder of the polarographic unit. Procedure. Sample solutions )\-ere prepared by weighing t h e desired material into a 10-ml. volumetric flask and adding t h e appropriate solvent a n d electrolyte. T h e solutions mere deaerated, prepolarized a t -0.5 volt for 5 minutes, and then held at t h e starting potential for 3 minutes before recording t h e polarogram. This was necessary t o obtain reproducible electrode surfaces during the day. All polarograms were recorded automatically a t a polarization rate of 74 mv. per minute (although this polarization rate is not critical, reproducible rates should be used). Halfwave potentials and diffusion currents were determined directly from the full curves by the line intersection method. All half-wave potentials are corrected for iR drop and are reported us. the normal calomel electrode (N.C.E.). Preparative Electrolysis. One gram of 10-[3-(4-p-hydroxyethyl-l-piperazinyl)propyl] - 2 - chlorophenothiazine was electrolyzed in 300 ml. of 0 . 2 s sulfuric acid a t a n anode potential of 0.8 volt us. t h e normal calomel electrode. After electrolysis, the solution was made basic and extracted with 300 ml. of chloroform. T h e extract, dried over anhydrous magnesium sulfate, was concentrated to 40 ml., then diluted with 400 ml. of ethyl acetate. The resulting solution was concentrated on the steam bath to a volume of 60 ml., cooled, and filtered. The filter cake was washed with a cold ethyl acetate-ether solution and dried, yielding 0.855 gram. This material was identified as the sulfoxide of the starting material by comparing its ultraviolet and infrared absorption spectra and mixed melting point with those of an authentic sample. RESULTS AND DISCUSSION
General Survey. All compounds studied (Table I) gave a well defined S-shaped oxidative wave except 2diethylaminoethyl - 10 - phenothiazinecarboxylate. Their half-wave potentials were dependent on t h e substituents in t h e 2- and 10-positions. Effect of 2-Substituents. Table I1 shows t h e effect of substitution a t position 2 for 10- [3-(4-p-hydroxyethyl-lpiperaziny1)propyl]phenothiazine, 10[3 - (4 - p - hydroxyethyl - 1 - piperazinyl)propyl] - 2 - chlorophenothiazine, 10 - [3- (4 - p - hydroxyethyl- 1 - piperazinyl)propyl] - 2 - acetylphenothiazine, and 10- [3- (4-p-h ydroxyeth yl- 1 -piperazinyl)propyl] - 2 - (trifluoromethy1)phenothiazine. The oxidation hccomes more difficult with stronger electron withdrawing groups (CF3 > COCHs > C1 > H). This effect is also evident in the 2-substituted parent phenothiazines (Table 11). Effect of 10-Substituents. The
greatest increase in half-wave potential of t h e parent phenothiazine is Table VI. Effect of pH on Half-Wave due t o substitution in t h e 10-position Potential and Current Constant of as a comparison in Table I1 reveals. Compound II" in 50% Ethyl Alcohol Table 111 s h o w t h e effect of subEljz id/C pH Buffer Component stituents in t h e 10-position. Three HzSOd 0 622 28 1 5 carbon atoms effectively isolate t h e 3 3 I\lalonic acid-SaOH 0 621 28 10-substituted nitrogen iron1 t h e 4 4 Llalonic acid-KaOH 0 607 32 5 5 H0-4c-XaOH 0 604 42 phenothiazine nucleus as indicated 6 8 hlalonic acid-KaOH 0 619 45 by tlie lack of pH dependence of the 8 3 SH,OH-HzSOI 0 592 57 half-wave potent'ial of such compounds, 8 8 XH~OH-HZSO~ 0 599 50b and the fact that 10-(3-chloropropyI)9 2 XH,OH-H,SO, 0 596 5tib 9 8 SH,OH-HzSO4 0 609 69* 2-chlorophenothiazine has essentially Concentration, 1.0 mM. t'he same half-wave potential as these b . Erratic current values in this pH conipounds. Two carbon atonis offer a region . poor shield to the influence of the e l e c t r o n - ~ ~ i t h d r a ~ protonated ~ing nitrogen (in acid solution) atom. The oxidavalue) over t h e temperature range tion of these compounds becomes easier from 11' t o 46" C. i n t h e aqueous as the pH increases, and is accompanied sulfuric acid solutlons. This indiby a rcduction of the electron-withcates t h a t the electrode process is drawing pon-cr of the nitrogen atom probably diffusion controlled. T h e (Table IV). half-rr ave potentials were unchanged The composit,e effects of the 2- and (within experimental error) with changes 10-substituents on the half-wave potenin temperature. tial can be separated with the aid of Reproducibility. The diffusion curultraviolet absorption data which are rent and half-n a v e potentials were primarily sensitive to the 2-position repioducible on a n y given d a y as in neutral solution, thus determining shown in Table J X I . il standard the nature of the %substituent (Table deviation of 0.6% was obtained for V). The effcct of tlie IO-substituent on the current constant and 2.4 niv, for the lialf-wave potential can be calcut h e half-wave potential. T h e daylated aft'er appropriate allowance is to-day variation in current constants made for the influence of the 2-substitand half-n ave potentials was 3.5% uent. A qualitative identification of (1 u ) and 3 my.. respectively, when the the 10-substituent can then be made. nitric acid cleaning procedure was emEffect of p H on Half-Wave Poployed. ,4 substantially larger variatentials and Diffusion Currents. tion in both current conqtants and halfConijiound 11, 10-I3-(4-p-hydroxynave potentials (77, and 12 niv., cthyl - 1 - piperaziny1)propyll - 2respectively) n a s encountered from chlorophenotliiazin~,was exanlined in day to day when no special electrode50% ethyl alcohol solutions (to keep cleaning process was used. it' in sc~lutionin t h e neutral and basic Reaction Mechanism. Because dibuffers) buffered from pH 1.5 to 9.8 phenyl sulfide was not oxidizable (Table 1-1). The half-nave potential in t h e polarographic system used for was independent of pH. T h e limiting the tranquilizers T\ hile S-methylcurrent, however, varied considerably diphenylamine v a s (0.690 volt), t h e (Table VI) increasing froin 28 p a , per oxidative wave of t h e tranquilizers nimole at pH 1.5 to 57.0 pa. per nimole vias believed due t o a n attack on t h e a t pH 8.3. Because tlie currents obnitrogen atom or on t h e phenyl ring taincd in the inore hasic solution were rather t h a n sulfur. ?\lillicoulometric erratic, analytical deterniinations should determinations of n for t h e oxidation be liased on pH controlled acid solutions. of X-methyldiphenylamine gave erratic Effect of Concentration on Halfvalues near 1, n-hereas the same deterWave Potential and Diffusion Curmination for the oxidation of 10-[3-4-prent. Aqueous 0 . 1 s sulfuric acid solutions of 10-[3-(4-~-hydrosyethyl- hydroxyethyl - 1 - piperaziny1)propyll2-chlorophenothiazine gave very repro1 - piperazinyl)proppl] - 2 - chloroducible values of n equal to 2. Eviphenothiazine were run which varied dently the similarity in half-wave potenin concentration from 0 . l i t o 2.0 niM. tials for the amine and the tranquilizers d linear relationship was observed was coincidental because different reIietn-een current and concentration action mechanisms m e involved. as shown in Table VII. T h e halfLarge scale electrochemical oxidation n a v e potential vas independent of a t a gold nire electrode yielded the concentrat,ion. Linear concentrationsulfoxide of 10-[3-4-fi-hydroxyethyl-ldiffusion current relationships were obpiperazinyl)propyI] - 2 - ehlorophenothiaserved a t all pH's studied in alcoholic solutions, Effect of Temperature on Eliz and I d . T h e temperature coefficient for t h e diffusion current \vas 3% per degree (with respect t o tlie 2j0 C.
Table VII. Effect of Concentration on Half-Wave Potential and Current Constant of Compound I1 in Aqueous Sulfuric Acid C0ncn.a E112 id/C 0 17 0 545 55.3 0 25 0 545 57.5 0 49 0 520 53.5 1 16 0 517 56 3 2 00 0 516 55 0 a Millimolar. Table VIII. Reproducibility of Current and Half-Wave Potential for Cornpound I I in Aqueous Sulfuric Acid Solutions Within Days" Day-t o-Day Eijz idle E112 id/C 0 541 56.3 0.539 56 1 56.1 0.543 56.7 0.537 51.5 0.545 57.2 0.537 0.536 55.3 0.546 56.7 0.538 51.6 0.545 56.5 53.3 0.539 0.544 56.4 0 542 55 1 0 539 56 7 57 1 56 0, 0 542 0 543 55 6 0 547 56 I 0 545 0 543 57 2 0 547 56 7 0 545 56 2 0 538 53 4 0 545 57 0 0 544 56 0 56 5 0 542 Std. dev. Eliz = Std. dev. E l j z = 0 0024volt 0 003 volt i d / C = 0 6% i d / C = 3 570 a Separate samples run on same day. Each reading represents average of two samplee. Data collected on 13 different days. zine. The sulfoxide is formed in preference to the X-oxide probably because of sulfur's ability to expand its octet. For this reason resonance structures such as Structure 2 can be drawn where sulfur is electron-rich and nitrogen electron-poor, making sulfur the most likely point for oxidative attack. " -.__ LITERATURE CITED
(1) B>iek, J., ceseskoslov. f a r m . 5, 210 II R A ) . ( 2 ) Dubost, Paul, Pascal, Suzette, Ann. pharm.frang. 11, 615 (1953). \ - - - - I
(3) Gilman, Henry, Nelson, R. D., J. Am. Chem. SOC.75. 5422 (1953). (4) Rleyer, F.,' Brzneimitfel-Forsch. 7, 296 (1957). (5) Salversen, B., Domange, L., Guy, J., Ann. pharm. f r a n f . 13, 169 (1955). (6) . . Sandri, G. C.. Farmaco Ed. xi. 10. 444 (1955). ( 7 ) Sasaki, Daizo, Folia Pharmacol. Japon. 48,95 (1952). (8) Weissberger, Arn:,ld, "Technique of Organic Chemistry, 2nd ed., Yo]. 11, p. 400, Fig. 3 4 Interscience, New York, 1956. RECEIVEDfor review June 13, 1958. Accepted October 14, 1958. Division of Analytical Chemistry, 134th Meeting, ACS, Chicago, Ill., September 1958.
VOL. 31, NO. 3, M A R C H 1959
433