Analyst, July, 1973, Vol. 98, PP. 485492 485 Ionic Polymerisation as a Means of End-point Indication in Non-aqueous Thermometric Titrimetry Part IV.* The Determination of Catecholamines BY E. J. GREENHOW AND L. E. SPENCER (Department of Chemistry, Chelsea College, University of London, Manresa Road, London, S . W.3) ( -)-Adrenaline, adrenaline hydrogen tartrate, L-noradrenaline, dopamine hydrochloride, L-dopa, DL-dopa, L-a-methyldopa, D-a-methyldopa and ( + ) - Corbasil have been determined in amounts down to 0.0001 mequiv by catalytic thermometric titration of their basic and acidic functions. Basic functions were determined by titration with 0.1, 0.01 and 0.001 M perchloric acid by using the ionic polymerisation of a-methylstyrene to indicate the end-point, while acidic functions were determined in a similar manner with tetra-n-butyl- ammonium hydroxide as the titrant and acrylonitrile as the end-point indicator.The L-dopa contents of tablets and capsules have been determined by using these techniques and the assay results have been compared with those obtained by alternative methods, namely, the recently described B. P. pro- cedure involving non-aqueous titration, and ultraviolet spectrophotometry. Magnesium stearate, which is used as a lubricant and flow promoter in tablet manufacture, is titrated as a base in the solvents used, but in titra- tions of the acidic function of catecholamines its effect is negligible. THE determination of catecholamines in biological specimens normally requires the use of trace analysis techniques such as fluorimetry and chromatography, as the concentrations usually encountered rarely exceed the microgram per gram level and are often in the sub- nanogram per gram range-l In contrast, catecholamines that form the active constituents in pharmaceutical preparations are present therein in relatively large amounts, and can be determined conveniently at the milligram per gram level or at even higher concentrations.For the routine assay of L-dopa [~-3-(3,4-dihydroxyphenyl) alanine], noradrenaline, adrenaline hydrogen tartrate and noradrenaline hydrogen tartrate, non-aqueous titration with perchloric acid is recommended.2J For the determination of catecholamines in formula- tions the more selective spectrophotometric methods are usually ~ s e d , ~ ~ ~ presumably in order to avoid interference from excipient material.However, in recent monographs,6 non-aqueous titration is the method prescribed for the assay of the L-dopa content of tablets and capsules. Thermometric procedures for the determination of organic acids and bases in amounts down to about 10 pg are described in Parts I7 and IL8 In these determinations a monomer capable of ionic polymerisation is added to a non-aqueous solution of the acid or base prior to titration. The end-point is indicated by a rise in temperature, which results from the ionic polymerisation of the monomer and which is initiated by the titrant following neutralisa- tion of the sample. The present paper describes the application of these thermometric procedures to the determination of some catecholamines of medical importance, namely (-)-adrenaline, adrenaline hydrogen tartrate, L-noradrenaline, dopamine hydrochloride, L- and DL-dopa, L- and D-a-methyldopa, and (+)-Corbasil.It is particularly concerned with the development of suitable solvent systems for these rather intractable compounds. The formulae of the compounds are shown in Fig. 1. Determinations have been carried out on the pure compounds and also on L-dopa tablets and capsules, and both the basic and acidic functions have been determined. As in the previous investigations, perchloric acid was used as the titrant and a-methylstyrene as the end-point indicator for the determination of basic functions, while tetra-n-butylammonium hydroxide was used as the titrant and acrylonitrile as the end-point indicator for the determination of acidic functions.The results obtained with L-dopa and * For details of Parts I, I1 and I11 of this series, see reference list on p. 492. 0 SAC and the authors.486 [Analyst, Vol. 98 with formulations that contain it are compared with those obtained by using the B.P. non- aqueous titration method6 and ultraviolet spectrophotometry. GREENHOW AND SPENCER: IONIC POLYMERISATION FOR END-POINT H R3 H Fig. 1. Structural formulae of the catecholamines (-)-Adrenaline OH H 2 CH, (+) -Corbasil OH CH, H H DL-Dopa H COOH H H Dopamine H H H H a-Methyldopa H COOH CH, H L-Noradrenaline OH H H H R, R2 R4 EXPERIMENTAL REAGENTS- Glacial acetic acid and propan-2-01 were of analytical-reagent grade and acrylonitrile, 1,2-dichloroethane, dimethylformamide and a-methylstyrene were laboratory-reagent grade materials.Oxalic acid (analytical-reagent grade) was dried at 110 "C for 3 hours before use. Other solvents and reagents were laboratory-reagent grade materials and were used as received. Perchloric acid, 0.1 M solution in acetic acid-Prepare this solution and standardise it by the method described in Part I.' Prepare 0-01 and 0.001 M solutions by diluting the 0.1 M titrant with 1,2-dichloroethane. Tetra-n-butylammonium hydroxide, 0.1 M in toluene - methanol-Laboratory-reagent grade material was used as received. Prepare 0.01, 0.002 and 0.001 M solutions by adding appro- priate volumes of toluene - propan-2-01 mixture (3 + 1) to the 0.1 M reagent. Standardise the solutions against benzoic acid (analytical-reagent grade) in dimethylformamide by the thermometric method.CATECHOLAMINES- (-)-Adrenaline, L-noradrenaline, L-dopa, DL-dopa, dopamine hydrochloride and adren- aline hydrogen tartrate were laboratory-reagent grade materials. All other catecholamines were gifts: L-dopa tablets (normal and sustained release) and capsules from Brocades (Great Britain) Ltd., L-a-methyldopa and D-cc-methyldopa from Merck, Sharp and Dohme, and (+)-Corbasil from Dr. S. Jones (Department of Pharmacy, Chelsea College). APPARATUS- described in Part 11.8 flask is adequate €or the present work. A. TITRATION OF THE BASIC FUNCTION- Manual method-Weigh about 0.1 mequiv of the sample into the titration vessel, add 0.5 ml of formic acid and mix well, then add 2-0 ml of acetic acid and 3 ml of a-methylstyrene.Add the titrant (0.1 M perchloric acid) at the rate of about 0.4 ml min-1 to within 0.3 ml of the end-point, noting the temperature at 15-s intervals, and complete the titration with addition of titrant at a rate not exceeding 0.2 ml min-1. Automatic method-Add the titrant at a constant rate (not exceeding 0.2 ml min-l) to a mixture of the sample, solvent and monomer in the 8-ml titration flask that is appropriate to the titrant concentration (see Table I). Record the temperature and titrant volume on a millivolt chart recorder (20-mV scale) at a chart speed of 600 mm h-l. All were dried over molecular sieve 4A before use. Manual method-Use the apparatus described in Part 1119 and the 15-ml titration flask Automatic method-Use the motor-driven syringe described in Part 111.An 8-ml titration PROCEDUREJuly, 19731 INDICATION IN NON-AQUEOUS THERMOMETRIC TITRIMETRY. PART IV Suitable amounts of sample, solvent and monomer are given in Table I. 487 TABLE I AMOUNTS OF SAMPLE, SOLVENT AND MONOMER RECOMMENDED FOR USE WITH THE DIFFERENT CONCENTRATIONS OF TITRANT Ti trant-Perchlor ic acid I A \ 0.1 M 0.01 M 0.001 M Samplelmequiv .. . . 0.1 t o 0-02 0.01 to 0.002 0.001 to 0~0001 Formic acid/ml . . f . 0.5 0.2 0.05f Acetic acid/ml . . .. 2.0 0.2 0.2 1,2-Dichloroethane/ml . . - Propylene carbonatelm1 . . - 1-1 - a-Methylstyrenelml .. 0.5 0.9 2.0 Solveats- 0-6 0.75 * With dopamine hydrochloride use the solvent mixture prepared by mixing 0.1 ml of formic acid, 0.5 ml of acetic acid and 0.4 ml of 1,2-dichloroethane.The end-point of the titration is measured, as in Part I,’ at the “upturn” temperature. This is the point where the titration curve leaves the tangent drawn to the “horizontal” portion of the curve in the vicinity of the temperature rise. B. TITRATION OF ACIDIC FUNCTIONS- Use the same general procedure as that used for titrations of the basic function but with the recorder set at 100 mV full scale. Prepare titration solutions by adding 2 ml of acrylonitrile to a solution of the sample in either 1 ml of dimethylformamide (suitable for dopamine hydro- chloride, adrenaline hydrogen tartrate and cc-methyldopa) or 1 ml of dimethylformamide containing an approximately equivalent and accurately known amount of P-toluenesulphonic acid [suitable for adrenaline, L-noradrenaline, L- and DL-dopa and (+)-Corbasil] .Appropriate amounts of sample for 0.1 M titrant (tetra-n-butylammonium hydroxide) are: 0.1 to 0.02 mmol of (-)-adrenaline, (+)-Corbasil and L-noradrenaline; 0.05 to 0.01 mmol of L- and DL-dopa, dopamine hydrochloride and cc-methyldopa; and 0.04 to 0.007 mmol of adrenaline hydrogen tartrate. With 0.01, 0.002 and 0.001 M titrant, use correspondingly smaller amounts of sample. The end-point of the titration is measured, as in Part 11,8 by the method of Vaughan and Swithenbank,lo in which it is located as the point where the tangent to the main heat rise leaves the curve at its lower temperature end. All determinations were carried out by using the automatic method. B.P. ASSAY METHOD- This method, which involves titration of a solution of the sample (about 500mg of active constituent) in a formic acid - acetic acid mixture with 0.1 M perchloric acid, with Oracet blue as the indicator, is described in detail in the British Pharmacopoeia.6 ULTRAVIOLET SPECTROPHOTOMETRY- Extract an amount of sample containing about 50 mg of L-dopa with 40 ml of 0.1 M hydrochloric acid.Filter and, by using 0.1 M hydrochloric acid as the diluent, make the volume up to 100 ml, then dilute 10 ml of this solution to 100 ml. Measure the absorbance of a l-cm layer of the solution at a wavelength of 280 nm and determine the L-dopa content by comparing the result obtained with values obtained with a “pure” sample of L-dopa. TITRATION OF MAGNESIUM STEARATE- Basic function-Add magnesium stearate (0.03, 0-06 or 0.1 mequiv) to a mixture of 0.5 ml of formic acid and 2 ml of acetic acid.Stir the mixture for 5 minutes, add 0.5 ml of a-methylstyrene and titrate the solution according to procedure A (automatic method). Basic function after precipitation of magnesium as the oxalate-Mix magnesium stearate (0.3, 0.6 or 1.0 mequiv) with 10 ml of formic acid, filter, wash the insoluble material with formic acid and make the volume of filtrate up to 20 ml with formic acid. Add 0 . 5 g of488 [Analyst, Vol. 98 dry oxalic acid, stir and allow the solution to stand for 4 hours. Filter and, by using 0-5 ml of the filtrate, determine the basic function as above. Acidic function-Add magnesium stearate (0-03, 0.06 or 0.1 mequiv) to 1 ml of a 0.1 M solution of benzoic acid in dimethylformamide and titrate the total acidity by using pro- cedure B.RESULTS AND DISCUSSION GREENHOW AND SPENCER: IONIC POLYMERISATION FOR END-POINT The sample must be partially soluble, at least, if consistent results are to be obtained by the thermometric method. Except for D- and L-a-methyldopa, which are soluble in dimethyl- formamide, the catecholamines examined in the present investigation were virtually insoluble in neutral and weakly basic organic solvents. For the determination of the basic function adrenaline and L-noradrenaline can be dissolved in acetic acid, but with L- and DL-dopa and (+)-Corbasil it is necessary first to dissolve the sample in formic acid before adding acetic acid. To avoid separation of the phases when a-methylstyrene is added to the sample solution there must be present a sufficient excess of acetic acid over the a-methylstyrene.A homogeneous solution can also be achieved by addition of propylene carbonate to mixtures of formic and acetic acids and a-methylstyrene. Formic acid and, to a lesser extent, acetic acid reduce end-point sharpness, particularly in titrations with 0.01 and 0.001 M titrants. With the weaker titrants the proportion of formic and acetic acids is kept to a minimum and 1,2-dichloroethane is used as a diluent. Details of suitable solvent systems are given in Table I. For titration of the acidic functions, dimethylformamide is a satisfactory solvent for adrenaline hydrogen tartrate and dopamine hydrochloride , as well as for the cc-methyldopas. The other catecholamines can be dissolved conveniently in dimethylformamide containing about 1 equiv of $-toluenesulphonic acid.It is, of course, necessary to subtract the volume of titrant consumed by the latter from the final titration volume. 0.1 M Tetra-n-butylammonium hydroxide/ml (1 division= 1 ml) Fig. 2. Thermometric titration of the acidic functions of catecholamines in suspension in weakly basic media Compound/mg N, 21.5 C, 16-6 A, 21.4 D, 19.0 A, 18.8 Solvent/ml F,2 F , 2 P, 1 H, 1 M,2 a b C d e Acrylonitrile/ml 1 1 2 2 1 Compounds-N, L-noradrenaline; C, (+)-Corbasil; A, (-)-adren- aline; and D, DL-dopa Solvents-F, dimethylformamide; P, NN-diethyl-3-aminopro- pionitrile ; H, hexamethylphosphoramide ; and M, N-methyl- morpholine Arrows indicate theoretical end-pointsJuly, 19731 INDICATION IN NON-AQUEOUS THERMOMETRIC TITRIMETRY.PART IV 489 It is possible to titrate these latter catecholamines in dimethylformamide alone, and also in N-methylmorpholine, NN-dimethyl-3-aminopropionitrile and hexamethylphosphor- amide, although they are almost insoluble in these solvents, but the precision of the deter- mination is not good. In the course of the titration there is initially a sharp rise in temperature until the combination of the heat evolved and added titrant brings about dissolution of the sample. The temperature then slowly decreases as neutralisation of the sample continues ; when neutralisation is complete the usual sharp temperature rise occurs. The resultant S-shaped titration curve (Fig. 2) is characteristic of the particular catecholamine.The lack of precision of the determination may be caused by the temperature fluctuations. In determinations of both the basic and acidic groups, amounts of sample were chosen so as to give titration volumes in the range 0.1 to 2.0 ml. In this range, calibration graphs were linear for the 0-1,O-Ol and 0.001 M titrants and 0-002 M tetra-n-butylammonium hydro- xide. In determinations of the acidic functions only one of the hydroxyl groups of the catechol moiety could be titrated (see Table 111, Part 11). Thus the dopa and a-methyldopa isomers and dopamine hydrochloride were determined as dibasic acids and adrenaline hydrogen tartrate was determined as a tribasic acid. Typical titration curves obtained by using the manual method for the determination of the basic function are shown in Fig.3. Curves of similar shape were obtained with the auto- matic apparatus, which was used for most of the determinations including those carried out in order to establish the precision of the method, both for basic and acidic functions. The results of the precision measurements are summarised in Tables I1 and 111. The coefficients of variation are similar in order to those obtained with the simpler bases and acids examined in previous paper^,^,^ and lie in the range 0.2 to 1-76 per cent. 0.1 M Perchloric acid/ml (1 division = 1 ml) Fig. 3. Thermometric titration curves obtained in determina- tions of the basic functions by the manual method Compound/mg r I a b C d L-Dopa, 19.8 (-)-Adrenaline, 18.0 hydrochloride, 19.1 L-Noradrenaline, 16.9 Dopamine Arrows indicate theoretical end-points With the 0.001 M titrants catecholamines could be determined in amounts down to about 20 pg.by titration of the basic function and to less than 20 pg by titration of the poly- functional acids. A comparison of the results given in Table I1 with those in Table 111 indicate that titration of either the basic or the acidic groups should be a satisfactory procedure for the assay of the catecholamines although, with compounds containing two or more acidic functions, titration of the acidic function has the advantage of higher sensitivity. The L-dopa content of formulations, namely, ordinary tablets, sustained-release tablets and capsules, has been490 [Analyst, Vol. 98 determined by titration of the basic function and of the titratable acidic functions. The same procedures as those used for titration of the pure catecholamines were used.The results of these determinations have been compared with those obtained by two methods that are currently used for the assay of L-dopa and its formulations, namely, the recently described B.P. method6 and ultraviolet spectrophotometry of a solution of the L-dopa in aqueous hydrochloric acid. It can be seen from the summary of the comparative study, shown in Table IV, that the results obtained by the B.P. method, ultraviolet spectrophoto- metry and thermometric titration, with one exception, do not differ by more than 1-6 per cent., and can be considered to be comparable. The exception is the value obtained in the thermo- metric titration of the acidic functions of capsules which, a t 95-2 per cent., is significantly higher than those obtained by the other methods.This higher value may possibly be attrib- uted to acidic excipient material, e.g., stearic acid or citric acid. GREENHOW AND SPENCER: IONIC POLYMERISATION FOR END-POINT TABLE I1 RESULTS FOR PRECISION FROM THE THERMOMETRIC TITRATION OF CATECHOLAMINES WITH 0.1 TO 0.001 M SOLUTIONS OF PERCHLORIC ACID Amount/ Catecholamine mg (-)-Adrenaline . . . . 18.36 DL-Dopa .. .. . . 19-56 L-Dopa (tablets) . , . . 29-52 L-Dopa (sustained-release tablets) . . ,. . . 31-22 L-Dopa (capsules) . . . . 20.08 (-)-Adrenaline . . . . 1.84 L-Noradrenaline . . . . 0.17 Dopamine hydrochloride . . 0.19 (-)-Adrenaline . . . . 0.18 Titrant Mean molarity* n t titrelml 0.1 4 1-04 5 1.01 5 1.04 4 1.27 3 0.94 0.01 3 1-43 0.001 3 0.86 3 0.95 3 1.24 Standard deviation 0-006 0.009 0.005 0.011 0.016 0.006 0.015 0.014 0.007 Coefficient of variation, per cent.0.56 0.88 0.50 0.91 1-23 0.40 1-76 1.49 0.58 * Nominal value. t Number of determinations. When the B.P. method is used for the assay of L-dopa formulations, care must be taken in deciding when the end-point has been reached as the indicator, titrant and L-dopa can be adsorbed on to the surface of insoluble excipient material in the sample. If such adsorption is likely, it is advisable to stir the titration solution for about 3 minutes between additions of titrant near the end-point so as to ensure that all of the titrant is consumed. This operation will, of course, increase significantly the time required for each determination. TABLE I11 RESULTS FOR PRECISION FROM THE THERMOMETRIC TITRATION OF CATECHOLAMINES WITH 0.1 TO 0.001 M SOLUTIONS OF TETRA-n-BUTYLAMMONIUM HYDROXIDE Catecholamine Dopamine hydrochloride L-Dopa (tablets) .. (-)-Adrenaline .. (-)-Adrenaline . . DL-Dopa .. .. DL-Dopa .. .. Dopamine hydrochloride Dopamine hydrochloride Amount/ 15.63 9-48 15.07 0.31 0-97 0.94 0.095 0.029 mg* p-Toluene- sulphonic acid/mg 18.98 18-98 1.90 1.90 - - - 0-057 Titrant Mean molarityt nS titrelml 0.1 3 1.93 4 1.03 3 2-50 0.01 3 0.41 3 2.18 3 1.03 0.002 3 0.84 0.001 3 1.70 * In 1 ml of dimethylformamide. t Nominal value. Number of determinations. Standard deviation 0.017 0.006 0.005 0.005 0.025 0.006 0*01.2 0.027 Coefficient of variation, per cent.0.89 0.63 0.20 1.23 1.16 0.59 1.46 1.61 Some formulations contain small amounts of magnesium stearate, e.g., from 0.5 to 2 per cent., as a lubricating agent for ease of tabletting and as a flow promoter. It was found that magnesium stearate dissolved in formic acid is titrated as a base when the thermometric method is used. An addition of 2 per cent. of magnesium stearate increases the titre byJuly, 19731 INDICATION IN NON-AQUEOUS THERMOMETRIC TITRIMETRY. PART IV 491 0.66 per cent., which is not insignificant in the assay of catecholamines in formulations. Magnesium can be precipitated as a solvated oxalate from formic acid solutions by the addition of an excess of oxalic acid. However, trace amounts are difficult to remove by precipitation because the oxalate is slightly soluble in formic acid and, after filtration, the solution still contains about 0.01 mequiv ml-l.This amount is equivalent to 10 per cent. of the titre when one is determining catecholamines at the 0-1 mequiv ml-1 level. TABLE IV COMPARISON OF METHODS FOR THE DETERMINATION OF THE L-DOPA CONTENT O F FORMULATIONS Thermometric titration method of spectro- Basic Acidic B.P. Ultraviolet Sample assay photometry function functions Levodopa tablets (normal) . . .. 67.9 68.7 67.2 67.7 Levodopa capsules . . .. .. 92.3 93.2 91-3 95.2 Levodopa tablets (sustained release) . . 74.2 74.3 74-9 73-9 L-Dopa .. .. .. .. .. 98.9 (100) * 99.5 100.9 * Calibration standard. All determinations were carried out on the same bulk sample of each formulation type, prepared by grinding together twenty dosage forms.T’alues given are per cent. m/m, and are the average of three determinations. In contrast, the effect of magnesium stearate on the titration of acidic functions is small and can be attributed partly, but not wholly, to small amounts of free stearic acid. The addition of 2 per cent. of magnesium stearate to L-dopa would increase the titre by about 0.1 per cent., and for all practical purposes this effect can be ignored. The results of a study of the titration of the basic and acidic functions of relatively large amounts of magnesium stearate are summarised in Fig. 4. 0.05 0.1 Magnesium stearate/mequiv Fig. 4. Calibration graphs for the thermometric titration of magnesium stearate : a, titration of the basic function; b, titration of the basic function after removal of precipitated magnesium oxalate ; and c, titration of the acidic function + 1 mequiv of benzoic acid.Details are given in the procedure Tables I1 and I11 show that the precision that can be obtained in the determination of the L-dopa content of formulations is similar to that achieved with the pure catecholamines. When one considers also that the thermometric method could be considerably faster than492 GREENHOW AND SPENCER the B.P. method, particularly if difficulties arise with indicator adsorption in the latter, then the method reported in this paper might be worthy of consideration as a procedure for routine assays. With the technique at its present state of development, i.e., with a lower limit of deter- mination of 10 to ZOpg, it cannot yet be considered to be suitable for the determination of catecholamines in body fluids and tissue extracts, unless a concentration step can be introduced into the analytical procedure.We thank the donors of the test compounds and formulations as indicated above. Mr. M. B. Arthur of Brocades (Great Britain) Ltd. is thanked for helpful advice on assay procedures for L-dopa. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. REFERENCES Holzbauer, M., and Sharman, D. F., in Blashko, H., and Muscholl, E., Editors, 44Catecholamines,” Handbook of Experimental Pharmacology, Volume XXXII, Springer-Verlag, Berlin, Heidelberg and New York, 1972, p. 111. “The British Pharmacopoeia 1968,” The Pharmaceutical Press, London, 1968, pp. 19, 20 and 662; “The British Pharmacopoeia 1973,” H.M. Stationery Office, London, 1973, p. 262. “The Pharmacopoeia of the United States of America, 18th Revision, 1970,” United States Pharmacopoeial Convention Inc., Washington, D.C., 1970, pp. 362, 227 and 230. Clarke, E. G. C., “The Isolation and Identification of Drugs in Pharmaceuticals, Body-fluids and Post Mortem Material,” The Pharmaceutical Press, London, 1969, pp. 176 and 322. “The Pharmacopoeia of the United States of America, 18th Revision, 1970,” United States Pharmacopoeial Convention Inc., Washington, D.C., 1970, pp. 352 and 354. “The British Pharmacopoeia 1973,” €3. M. Stationery Office, London, 1973, pp. 263 and 264. Greenhow, E. J., and Spencer, L. E., Analyst, 1973, 98, 81. , Ibid., 1973, 98, 90. -- , Ibid., 1973, 98, 98. Vauihan, G. A., and Swithenbank, J. J., Ibid., 1970, 95, 890. NOTE-References 7, 8 and 9 are to Parts I, I1 and I11 of this series, respectively. * -- Received January 30th, 1973 Accepted March 2nd, 1973