Analyst, October, 1973, Vol. 98, PP. 745-748 745 Determination of Cyclamate in Soft Drinks by Reaction with Nitrous Acid . Part II.* Manual and Semi-automated Methods: Determination of Cyclohexyl Nitrite by Diazotisation and Coupling with Bratton - Marshall Reagent BY A. JEAN SHENTON (Lyons Central Laboratories, London, W14 O Q U ) AND R. M. JOHNSON (Food, Drink and Tobacco Industry Training Board, Leon House, High Street, Croydon, CR9 3NT) The determination of cyclamate a t a concentration of approximately 1 mg ml-l in soft drinks is described. Cyclohexyl nitrite, derived from cyclamate by reaction of the latter with nitrous acid, is determined in a non-aqueous system by initial diazotisation with sulphanilamide and subse- quent coupling with 2-aminoethyl-1-naphthylamine. Manual and semi- automated procedures have been evaluated and interference has been studied.THE initial products of the reaction of nitrous acid with cyclamate are sulphuric acid, nitrogen and carbonium ions- Subsequently, the unstable carbonium ions are rapidly converted into cyclohexene (with a 24 per cent. yield) and cyclohexyl nitrite (with a 76 per cent. yield, under strong nitrosating conditions) .1 Published methods for the determination of cyclamate depend on the deter- mination of sulphate, nitrite or cyclohexene. These include gas - liquid chromatographic,2-4 thin-layer chroma tog rap hi^,^,^ titrimetric,' gravimetric* and infrared spectrophotometri~~ procedures. Determinations based on cyclohexyl nitrite arouse particular interest because of the specificity of this compound and its formation in high yield.Previously, organonitrites have been determined by diazotisation and coupling in a two-phase system,1° but the one-phase system, as described in this paper, affords increased simplicity and precision of technique. EXPERIMENTAL REAGENTS FOR MANUAL METHOD- Standard sodium cyclamate solution, aqueous, 2.5 mg m1-I. Standard sodium nitrite solution, aqueous, 0.5 M. Sulphuric acid, 5 N. Tetrachloroethylene, redistilled, boiling range 120 to 121 "C. Ethanol, 95 per cent. Bratton - Marshall mixed reagentll-A 0.5 per cent. m/V solution of sulphanilamide in 1 + 1 hydrochloric acid - 95 per cent. ethanol (1 + 1) (this solution is stable for 4 weeks) is mixed with an equal volume of 0.1 per cent. m/V solution of 2-aminoethyl-1-naphthylamine dihydrochloride in 95 per cent.ethanol (this solution is stable for 4 weeks at 5 "C in the dark). PROCEDURE FOR MANUAL METHOD- Weigh accurately about 10 g of the sample (containing approximately 1 mg rnl-1 of cyclamate) and transfer it quantitatively into a 100-ml separating funnel. Adjust the volume to a total of 25 ml by the addition of distilled water and add, by pipette, 10 ml of 5 N sulphuric acid, 1 O m l of tetrachloroethylene and 15 ml of 0.5 M sodium nitrite solution. Shake the * For details of Part I of this series, see reference list on p. 748. 0 SAC and the authors. For Part 111, see p. 749.746 [Analyst, Vol. 98 mixture for 2 minutes, then separate the organic solvent layer and wash it with 1 O m l of distilled water.Discard the washings, filter the organic solvent layer through a Whatman No. 5 filter-paper and dilute an aliquot of the filtrate exactly ten times with 95 per cent. ethanol. Transfer an aliquot (1 ml) of the resulting solution into a dry 25-rnl calibrated flask; add, by pipette, 20 ml of 95 per cent. ethanol and 2 ml of Bratton - Marshall mixed reagent. Dilute to volume with 95 per cent. ethanol, mix the solution and, after 20 minutes, determine the optical density at Amax. (550 nm) by using l-cm cells with 95 per cent. ethanol in the reference cell. Read off the cyclamate content by reference to the calibration graph prepared as described below. CALIBRATION GRAPH FOR MANUAL METHOD- Range 0 to 12.5 mg of cyclamate-To a series of 100-ml separating funnels add 0, 1,2,3,4 and 5-ml portions of standard sodium cyclamate solution (2.5 mg ml-l).Adjust the volume in each instance to 25ml and continue as described under Procedure for manual method as far as “. . . by using l-cm cells with 95 per cent. ethanol in the reference cell.” Construct a calibration graph relating cyclamate concentration to optical density. (Under our conditions the calibration graph obeyed the equation y = 0.55x, where x is the amount of cyclamate in milligrams and y is the optical density.) The calculated E value for cyclamate based on this graph was 3.07 x lo4 at Amax. (550 nm). PROCEDURE FOR AUTOMATED METHOD- By using Technicon AutoAnalyzer equipment and the manifold shown in Fig. 1, set the base-line by continuously sampling air through the sample line A, aqueous sodium cyclamate SHENTON AND JOHNSON : DETERMINATION OF CYCLAMATE I N ~ N-nitrosat ion Solvent extraction F I ow-ra te / mI min-’ Sample 0-4 HzSO4, 2 N 3-4 Sodium cyc I a mate, 0-4 0.8 mg ml-I (wash) NaN02, 2.5 1.5 mg.rn1-I Re-sample.1 -2 Solvent (tetrachloro- 2.0 ethylene) Waste n * J - n U Water Waste ~ A * - 1 *6 2.0 2.5 1.1 ml of solvent metered Bratton - Marsha I I Diazotisation reagent -f 3.9 15-mm flow cell Air 1-2 Double coi I 605 Waste * 3.9 mixing - Colour measurement filter Fig. 1. t Reagent specified in manual method but diluted twenty times with water Manifold for determination of cyclamate in the concentration range 0.80 to 1-30 mg ml-l . All Tygon tubing except where marked *, when yellow Tygon tubing was used.October, 19731 SOFT DRINKS BY REACTION WITH NITROUS ACID. PART 11 747 solution (0*8mgml-l) through the wash line B and other reagents as specified in Fig.1. Construct a calibration graph by sampling standard sodium cyclamate solutions , covering the range 0.80 to 1-30 mg ml-l, through the sample line A at the rate of 40 per hour. (In our work, a change of 0.10 mg ml-l in the cyclamate concentration produced a change in optical density of about 0.1 unit.) Sample the test solutions and determine the cyclamate content of the samples by reference to the calibration graph. RESULTS AND DISCUSSION Results of analyses are given in Table I. TABLE I DETERMINATION OF CYCLAMATE IN SOFT DRINKS WITH 1 mg ml-l OF ADDED CYCLAMATE Manual method Automated method - - Type of soft drink . . .. .... Orange Lemon Orange Lemon Number of samples analysed . . .. .. 10 10 20 20 Mean apparent recovery, per cent. . . . . 106.4 106-9 106-1 106.6 Apparent recovery range, per cent. . . . . 104 to 109 104 to 110 103 to 109 103 to 110 Mean deviation, (&) per cent. . . .. . . 1.6 2.0 1.5 2.1 The manual procedure is simple and fairly rapid. Maximum colour is developed within 20 minutes and the colour remains stable for at least 1 hour. The automated procedure has proved somewhat difficult to perform over long periods of time, because of the problem of maintaining equilibrium in each of the two-phase separators. This problem would probably be overcome by using the techniques described by Carter and Nickless.12 Assessment of the interference caused by various classes of substances has been made by applying the conditions of nitrosation, extraction and colorimetric determination, as specified above in the Procedure for manual method, to several test substances (see Table 11). TABLE I1 INTERFERENCE OF FUNCTIONAL GROUPS Class of substance Alcohol Terpene Aldehyde Amine Amino-acid Phenol Reductant Test substance Propan-2-01 D-Linalol D-Limonene Formaldehyde Methylamine DL-Serine L-Cysteine Phenol Phl oroglucinol Ascorbic acid Sulphur dioxide Interference, per cent.* 100 38 0.5 0.5 <0*1 <0*1 (0.1 < 0.1 (0.1 (0.1 <0*1 Molar absorbance of test substance Molar absorbance of sodium cyclamate * Interference, per cent.= - x 100. Alcoholic constituents are a major cause of interference (Table 11) but attempts to remove these compounds from control samples that were free from cyclamate have not reduced sample blank values by more than 10 per cent.Preliminary treatments examined include removal of steam volatiles, oxidation with permanganate , bromination and extraction with diethyl ether of the sample saturated with sodium chloride. Interference associated with the basic constituents of soft drinks has been evaluated by determining the apparent cyclamate content of control samples that were free from added cyclamate (Table 111).748 SHENTON AND JOHNSON TABLE I11 INTERFERENCE DUE TO BASIC CONSTITUENTS OF DRINKS* Twenty replicate analyses Type of drink . . .. .. Orange Lemon Mean .. .. .. . . 6.9 6.2 .. 4 t o 8 4 to 10 Range .. .. .. Mean deviation (&) . . . . 0-9 1.4 * Expressed as a percentage for a nominal cyclamate content of 1 mg ml-l. The results show that the interference is reasonably constant between samples. By deducting the appropriate mean values (listed in Table 111) from the apparent recovery values (listed in Table I), recovery values of added cyclamate of approximately 100 to 101 per cent.are obtained. High reagent blanks can arise from impurities present in tetrachloroethylene, which readily undergo nitrosation, but, by limiting the volume of solvent used at the nitrosation - extraction stage, the solvent blank has been effectively reduced. Only trace amounts of inorganic nitrite remain entrained in the solvent phase and they contribute negligibly to the solvent blank. Investigations have shown that quantitative extraction of cyclohexyl nitrite is effected by the single-extraction stage specified in the manual method.Use of a pre-mixed diazotisation and coupling reagent results in increased precision and only a small reduction in sensitivity, compared with procedures that involve the separate addition of diazotisation and coupling reagents. The one-phase non-aqueous system for colour development described in this paper gives a quantitative reaction, rapid colour development and good reproducibility, and is preferred to a two-phase system described by earlier workers.1° The contribution of cyclohexyl nitrite and cyclohexene to the final colour measured has been determined; E values are 4.16 x lo4 for cyclohexyl nitrite and 0.09 x lo4 for cyclo- hexene (compared with 3.07 x lo4 for cyclamate).Under the conditions defined for the N-nitrosation of cyclamate, the yield of cyclohexyl nitrite is 76 per cent. and that of cyclo- hexene 23 per cent. The calculated E value for cyclamate based on the contribution of these products is 3-17 x 104; this result is in good agreement with the determined value of We thank J. Lyons and Co. Ltd. for the opportunity given (to A.J.S.) to carry out 3-07 x 104. this work, and colleagues for helpful advice. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. REFERENCES Shenton, A. J., and Johnson, R. M., Tetrahedron, 1971, 27, 1461. Dalziel, J . A. W., Johnson, R. M., and Shenton, J., Analyst, 1972, 97, 719. Rees, D. I., Ibid., 1965, 90, 568. Richardson, M. L., and Luton, P. E., Ibid., 1966, 91, 520. Kamp, W., Pharm. Weekbl. Ned., 1966, 101, 57. Dickes, G. J., J . Ass. Publ. Analysts, 1965, 3, 118. Richardson, M. L., and Luton, P. E., Analyst, 1966, 91, 522. Wilson, J . B., J . Ass. Off. Agric. Chem., 1955, 38, 559. Oi, N., and Inaba, E., J . Pharm. SOC. Japan, 1967, 87, 640. Hoare, D. E., Ogilivie, R. R., and Wellington, C. A., Analyt. Chew., 1966, 38, 1799. Bratton, A. C., Marshall, E. K., Babbitt, D., and Hendrickson, A. R., J . Biol. Chem., 1939, 128, 537 Carter, J. M., and Nickless, G., Analyst, 1970, 95, 148. NOTE-Reference 2 is to Part I of this series. Received October 24th, 1972 Accepted May 3 4 1973