ANALYST, FEBRUARY 1986, VOL. 111 163 Determination of Vitamin C by Flow Injection Analysis Fernando Lazaro, Angel Rios, M . D. Luque de Castro and Miguel Valcarcel* Department of Analytical Chemistry, Faculty of Sciences, University of Cordoba, Cordoba, Spain A photometric procedure for the determination of vitamin C using chloramine T by two flow injection analysis (FIA) methods is described. The first is in the presence of potassium iodide - starch solution using the normal FIA technique (linear determination range: 15-150 pg ml-1, RSD +_0.97O/0 and sampling frequency 90 samples h-1) and the second is in the presence of potassium bromide - methyl red solution using an FIA titration (linear determination range: 0.5-1000 pg mi-’, RSD +0.82°/0 and a sampling frequency of 30 samples h-l).The two methods are compared with each other and with a conventional titration procedure. Keywords: Vitamin C determination; flow injection analysis; potassium iodide - starch; potassium bromide - methyl red Numerous conventional methods have been described for the determination of vitamin C-with electroanalytical detection. most frequently used type of detection. Methods based on the use of Ti(II1) ,1 Ce(1V) ,2 hexacyanoferrate(II1) ,3 bromine cyanide,4 vanadium,5 iodate,6 2,4-dinitrophenylhydrazine7J and a7a’-dipyridyl9-l1 have been described, each with their advantages and disadvantages. Flow injection analysis (FIA) has been applied twice to the determination of vitamin C with electro analytical detection. The first contribution was reported by Strohl and Curran,12 who used a reticulated glassy carbon electrode to carry out the coulometric and amperometric determination of the vitamin, achieving detection limits of a few nanograms.The major advantage of the coulometric method is that the analyte is completely electrolysed and can be determined without calibration graphs. However, the sampling frequency is decreased with respect to the amperometric method. The second contribution was an amperometric method using immobilised enzyme (ascorbic acid oxidase) reactors that allowed the determination of the vitamin in brain tissue, with good results. 13 In order to develop a photometric - FIA method for the determination of vitamin C we used the reagent chloramine T in the presence of starch - potassium iodide solution or methyl red - potassium bromide solution as an indicator,’4 by normal FIA in the first instance and by an FIA titration without a gradient chamber (expansion-scale technique)l5-18 in the second.Both methods are compared with each other and with the conventional titration procedures. Experimental and Results Apparatus An FIA-5020 Tecator Analyzer, an SP6-500 Pye Unicam UV - visible spectrophotometer and a 178.12 QS Hellma flow cell (inner volume 18 1.11) were used. Reagents Chloramine Tsolution, 5 x 10-4 M. Starch - potassium iodide solution. Containing 5.25 g of soluble starch and 0.50 g of potassium iodide in 500 ml of distilled water. Methyl red - potassium bromide solution. Containing 0.5 ml of methyl red, 0.1% in ethanol, made up to 100 ml with 0.5% KBr solution. Ascorbic acid solutions.Prepared by direct weighing. All solutions were stable for at least 24 h. * To whom correspondence should be addressed. Determination of Vitamin C with Chloramine T and Starch - Potassium Iodide Solution Although this determination is commonly carried out by titration14 this is unsuitable for use with FIA when starch is employed as an indicator, due to the fleeting appearance of the blue colour of the starch - iodine complex in the reactor. A plateau is not obtained when a high sample volume is injected because the excess of chloramine T degrades the starch - iodine complex. It is therefore essential that the blue colour should have developed by the time that the sample passes through the detector. Manifold and chemical system The manifold used is shown in Fig.1. The design is very simple. The sample, dissolved in an acidic medium, is injected into the chloramine T stream after merging with a starch - iodide stream. Iodide forms HI in an acidic medium, which subsequently reacts with chloramine T: NaCl + I2 The iodine formed oxidises vitamin C: Once vitamin C has been fully consumed, iodine binds to starch, which consitutes the indicator reaction. Sulphuric acid is added directly to the sample as insertion into the chloramine T stream results in the precipitation of the reagent. When the sample only contains sulphuric acid (blank) no iodine is consumed because of the absence of vitamin C and thus the maximum signal is obtained. The presence of vitamin C causes a weakening of the analytical signal proportional to its concentration.Optimisation of variables The optimum values for the FIA variables (Fig. 1) are as follows: flow-rate (4) = 2.04 ml min-1; injected volume (Vi) = 96.2 p1; reactor length ( L ) = 55 cm; inner diameter of the reactor (@) = 0.5 mm. These values allow the blue colour of the starch - iodine complex to be formed as the sample plug passes through the detector, where it is monitored at 650 nm. The optimisation of the variables ensures the maximum difference between the reference and the sample peaks. The influence of the chloramine T concentration is one of the most important variables in this determination. Its effect can be observed in Fig. 2. For high concentrations of chloramine T the presence of vitamin C has no effect on the164 I Starch - K I Sample (H+) Chlorarnine T i ANALYST, FEBRUARY 1986, VOL.111 650 nrn Photometer Methyl red - KBr Sample (H+) Chlorarnine T I Blank - 1 n q = 1.19 rnl rnin-1 1 W 1 Time - I Sample I Time - Fig. 1. (for details see, text) Manifolds used in the determination of vitamin C and recordings obtained by: (a) normal FIA mode and ( b ) FIA titration analytical signal. The maximum difference between the blank and the sample signal is obtained with 5 x 10-4 M chloramine T solutions. For lower concentrations of the amine the difference in absorbance decreases owing to the drastic weakening of the blank signal. The pH of the sample is adjusted by the addition of H2S04. The reaction medium must be acidic to allow the formation of HI. The difference in absorbance between the blank and the sample is high at low pH, but above a sulphuric acid concentration of 0.75 M the increase in this difference is small and some peak broadening is observed. A 0.90 M sulphuric acid concentration was chosen as the optimum.The most suitable concentrations of KI and starch are indicated under Experimental. It is well known that an increase in temperature is a negative factor for the development of the blue colour of the starch - iodine complex; therefore, the experiments were performed at room temperature. Determination of vitamin C This method allows the determination of vitamin C according to the equation: AA = 0.0035 [vitamin C]-O.O053; r = 0.999 where AA is the difference in the absorbance between the blank and the sample containing vitamin C (whose concentra- tion is expressed in pg ml-1).The equation is satisfied between 15 and 150 pg ml-1, the RSD (P = 0.05) for the determination of 100 pg ml-1 of vitamin C being &0.97%. The sampling frequency is 90 samples h-1. Determination of Vitamin C with Chloramine T and Methyl Red - Potassium Bromide Solution This method, based on the measurement of the peak width at a pre-determined height from the base line, has been termed “high-speed titrations” by Ramsing et al. 15 and “scale- expansion techniques” by Stewart and Rosenfeldl6 following the suggestions from Pardue and Fields.17?1* The chemical system is similar to the one described above but the indicator is methyl red - potassium bromide solution. I, B 1.200 al f (0 2 0.800 v) 2 0.400 0.1 1 .o 5.0 [Chlorarnine TI x 1 0 - 2 / ~ Fig.2. Influence of the chloramine T concentration in the stream. Absorbance values obtained with: A, sample with 5 X M vitamin C; B, blank; and C , difference between sample and blank This system may be used for FIA titrations because the red colour of methyl red in an acidic medium allows a wide peak to be obtained when a large volume of this indicator is injected. Manifold and chemical system The manifold used is similar to the one described above. In this instance the methyl red - potassium bromide stream is substituted for the starch - potassium iodide one. The HBr formed by merging this stream with that of the sample in an acidic medium (H2S04) is oxidised by chloramine T and the Br2 produced oxidises vitamin C to dehydroascorbic acid.An excess of bromine with respect to the stoicheiometric amount of vitamin C present reacts with methyl red, causing its degradation. The presence of the vitamin in the sample (resulting in a decrease in the degradation of the indicator) brings about a widening of the peak obtained at 535 nm. The measurement of the peak width (in seconds) at a pre- determined absorbance value is the basis of the FIA titrations.ANALYST, FEBRUARY 1986, VOL. 111 165 Table 1. Calibration graphs for the determination of vitamin C by FIA titration Absorbance * Range/M Equation 0.020 (0.3-1.7) x At = 8.248 log At = 18.17 log 0.050 (0.3-1.7) x 10-5 At = 4.67108 At = 15.18 log At = 20.60 log 0.100 (0.3-1.7) x 10-5 At= 4.46108 At = 15.88 log (0.1-2.8) x 10-3 (0.2-1.1) x 10-4 (0.2-5.7) x 10-3 (0.02-5.7) X 10-3 *Absorbance at which the peak width (At) is measured vitamin C] + 49.63 vitamin C] + 102.94 vitamin C] + 28.55 vitamin C] + 79.08 vitamin C] + 100.38 vitamin C] + 28.01 vitamin C] + 80.86 r 0.991 0.991 0.993 0.997 0.991 0.990 0.999 Table 2. Determination of vitamin C in synthetic samples [Vitamin C] by normal FIA*/M [Vitamin C] by FIA titration ?/M Added Found Error, YO 9.02 x 10-5 9.66 x 10-5 +6.8 1.70 x 10-4 1.77 x 10-4 +4.1 2.84 x 10-4 2.67 x 10-4 -5.9 5.66 x 10-4 5.56 x 10-4 -1.7 1.14 x 10-3 1.17 x 10-3 +2.6 1.42 x 10-3 1.43 x 10-3 -1.4 * Equation: AA = 615 [vitamin C] - 0.005 t Equation At = 15.882 log [vitamin C] + 80.86 Added Found Error, YO 2.08 x 10-5 2.17 x 10-5 +4.3 5.68 x 10-5 5.82 x +2.5 1.70 x 10-4 1.71 x 10-4 +0.6 4.26 x 10-4 4.23 x 10-4 -0.7 8.52 x 10-4 8.63 x 10-4 +1.3 1.42 x 10-3 1.39 x 10-3 -2.1 "t i 60 v) L Q 40 2o t "\ " Absorbance Fig.3. Variation of At with the value of the absorbance at which the measurement of the peak width is performed. A, Sample with 5 X 10-3 M vitamin C; B, blank; C, difference between sample and blank Optimisation of variables The optimum values for the FIA variables are as follows: q = 1.19 ml min-1; Vi = 923.0 pl; L = 70 cm and @ = 0.5 mm. The large sample volume injected ensures that wide peaks are obtained, the lower the vitamin C concentra- tion, the wider the peaks. This sample volume, which is unusually large for normal FIA, is the key to FIA titrations. The concentrations of some reagents are decisive.Thus, the optimum concentration of chloramine T is 5 X M, which allows maximum differences to be obtained between the sample and the blank. The samples contain 1% of concen- trated H2S04 (0.15 M). Owing to the short reactor length, the temperature does not exert a significant influence. Thus, the experiments were performed at room temperature (18-20 "C). The absorbance at which the peak width is measured influences the sensitivity of the determination. It is advisable to measure the peak width at the absorbance yielding the maximum difference between the sample and the blank. Fig. 3 shows this effect for a sample containing 100 yg ml-1 of vitamin C. The differences increase as the absorbance approaches zero, but owing to instrumental limitations, peak widths over 99.9 s cannot be measured.The increase in peak width observed above 0.100 A (curve C, Fig. 3) is not significant, and as the corresponding peak widths are very small, the errors introduced in these measurements are large. Therefore, the criteria adopted to run the calibration graphs at different absorbances according to the high or low concentra- tion of vitamin C in the samples are as follows: the peak width is measured at 0.100 A for high and at 0.020 or 0.050 for low analyte concentrations. Calibration graphs A plot of the peak width (At) versus logarithm of the unknown concentration, characteristic of these titrations, shows differ- ent linear ranges according to the value of the absorbance at which At is measured (as is shown in Table 1). In each instance the contribution of the blank has been subtracted.This represents a major improvement of the method proposed above; a wider range and lower limit of detection are attained, possibly owing to the instability of the starch - iodine complex in the presence of chloramine T, which was a limiting factor in the previous method. An RSD ( P = 0.05) of +0.82% and a sampling frequency of 30 h-1 were obtained in the determina- tion of 100 pg ml-1 of vitamin C (At measured at 0.050 A). Study of Interferents in Each Method In a study of the interferences in both FIA methods, Fe2+, Ca2+, Mg2+ , NH4+, P043-, C032-, oxalate, glucose, glycine, hystidine, urea, cysteine and uric acid can all be tolerated at a ten-fold excess over vitamin C, except for cysteine and uric acid, which interfere when present at the same concentration as the vitamin in the FIA titration method.It is worth noting the smaller influence of foreign species in the FIA technique compared with conventional methods. 14 This behaviour was observed in earlier studies19 and is attributable to the non-equilibrium state of the chemical system at the point of detection. The higher selectivity of the method employing iodine is a result of the lower oxidising strength of this halogen166 ANALYST, FEBRUARY 1986, VOL. 111 compared with bromine. Likewise, the higher acidity of the sample in the first method facilitates the precipitation and removal of interferents (e.g. , uric acid). Comparison of the Results Obtained by the Two Methods The results obtained by both methods for the determination of vitamin C in several synthetic samples are given in Table 2.A similar degree of accuracy is afforded by both methods. The determination of vitamin C with chloramine T by FIA titration offers the following advantages over the normal FIA technique: a wider determination range; a higher sensitivity; and a slightly higher accuracy. These advantages stem from the high instability of the starch - iodine complex in the presence of chloramine T, which is a limiting factor in the determination of vitamin C by the normal FIA technique. Normal FIA does however have some advantages, such as the need for only one calibration graph, smaller sample volume, higher selectivity, which makes it applicable to the determination of vitamin C in urine, and higher sampling rate.The determination of vitamin C by FIA titration presents the following advantages over the manual procedurel4: it is faster and simpler, as the colour change in the manual method is slow and difficult to detect visually; it uses reagent and sample more sparingly; the manipulation and intervention of the operator is a minimum; and it is more sensitive and accurate . References 1. Gupta, D., Sharma, P. D., and Gupta, Y. K., Talanta, 1975, 22, 913. 2. Rao, G. G., and Sastri, G. S . , Anal. Chim. A m , 1971,56,325, 3. Sastri, G. S., and Rao, G. G. Talanta, 1972, 19, 212. 4. Paul, R. C., Chauhan, R. K., and Prakash, J . , Indian J . Chem., 1971, 9, 879. 5. Eremina, Z. I., and Gurevich, V. G., Zh. Anal. Khim., 1964, 19, 519. 6. Murty, C. N., and Bapat, N. G., Fresenius Z. Anal. Chem., 1963, 199, 367. 7. Roe, J. H., Oesterling, M. B., and Llamron, C. M., J. Biol, Chem., 1948, 174,201. 8. Teruuchi, J., and Kitazato, K . , Arch. Exp. Med., 1951, 1, 55. 9. Zannoni, U., Lynch, M., Goldstein, S., and Sato, P., Biochem. Med., 1974, 11, 41. 10. Okamura, M., Clin. Chim. Acta, 1980, 108, 259. 11. Okamura, M., Bitamin, 1981, 55, 495. 12. Strohl, A. N., and Curran, D. J., Anal. Chem., 1979,51, 1045. 13. Bradberry, C. W., and Adams, R. N., Anal. Chem., 1983,55, 2439. 14. Verma, K. K., and Gulati, A. K.,Anal. Chem., 1980,52,2336. 15. Ramsing, A., RfiiiCka, J., and Hansen, E. H., Anal. Chim. Acfu, 1981, 129, 1. 16. Stewart, K. K., and Rosenfeld, A. G., Anal. Chem., 1982,54, 2368. 17. Pardue, H. L., andFields, B., Anal. Chim. Acta, 1981,124,39. 18. Pardue, H. L., and Fields, B.,Anal. Chim. Acta, 1981,124,65. 19. Linares, P., Luque de Castro, M. D., and ValcArcel, M., Anal. Paper A5134 Received January 29th 1985 Accepted August 28th, 1985 - Left., 1985, 18 B1, 67.