Analytical Communications, January 1996, Vol33 (9-1 0) 9 Improved Method for the Determination of Ascorbic Acid in Beer by Using High-performance Liquid Chromatography With Electrochemical Detection David Madigan", Ian McMurrough" and Malcolm R. Smythb a Guinness Brewing Worldwide Research Centre, St. James's Gate, Dublin 8, Ireland a School of Chemical Sciences, Dublin City University, Dublin 9, Ireland A method for the determination of ascorbic acid in beer by using HPLC with dual electrode electrochemical detection is described.Ascorbic acid was separated by using ion suppression chromatography on a C18 column and detected by a dual-electrode system consisting of a porous graphite high efficiency electrode followed in series by a glassy carbon amperometric electrode. It was easily detectable at sub-ppm levels in beer.The use of dual electrode detection provided a constant measurement of both peak purity and electrode efficiency. Ascorbic acid (AA) is a powerful antioxidant and unlike many other antioxidants it has the ability to reduce molecular oxygen.' It is widely used as an antioxidant in the food and beverage industry, and some brewers add AA at 30-50 mg 1-l to beer in order to diminish residual oxygen after packaging. This is thought to delay the development of stale flavours in beer (believed to be a result of oxidative chemical reactions.2) The sensitivity of AA to oxygen, however, makes the analysis of the reduced form very difficult, particularly if the sample must undergo an extraction technique that permits oxidation. Therefore it is desirable to measure AA by rapid, sensitive methods which minimize contact of the sample with atmo- spheric oxygen.The standard AOAC method for AA determi- nation3 uses a titration based on the reduction of 2,6-di- chlorophenol indophenol. This method, however, is subject to numerous interferences in samples such as beers or fruit juices, which may contain substantial amounts of reducing substances such as flavanols, melanoidins or reductones.Added to this, the end-point of the titration is poorly defined, and this increases the possibilities for error in this method. The use of HPLC with UV absorbance detection is applicable to AA determination$" but is insufficiently sensitive for the determination of very low levels of AA.Sample preconcentration steps may be necessary, which may lead to the unwanted risk of absorption of oxygen by the sample. A suitable alternative to these methods is the use of electrochemical detection with HPLC. AA is very easily oxidized at a glassy carbon working electrode at low operating potentials, thereby providing a sensitive detection system. The ease of oxidation of AA means that potentially interfering analytes with higher oxidation potentials are not detected, so selectivity is high.While HPLC with electrochemical detection for AA determination has been widely documented,7-' 1 many reported methods suffer from either a lack of reproducibility or from losses of sensitivity due to fouling of the electrode with sample components. This has necessitated bracketing of samples and standards, which, as well as being troublesome to perform, greatly reduces the number of samples which may be analysed in a given time.Here, we present a method for the determination of AA in beer which provides excellent repeata- bility and long-term electrode stability, and which may also be applied to other beverages. Furthermore, the detection system used in this study possessed additional advantages over the more commonly used single-cell amperometric devices.The high efficiency (coulometric) electrode is claimed to be capable of oxidizing all of the sample which passes through it, due to the large electrode surface area. This should therefore provide added sensitivity and repeatability when compared to ampero- metric detectors, which, due to diffusion limitation, typically only oxidize 10% of the analyte.In addition, the response of the downstream amperometric electrode, which is operated at a higher potential, can be used as a valid indicator of peak purity and/or electrode poisoning. The ratio of current responses at the two electrodes is dependent on the shape of the hydrodynamic wave of the analyte, and also on the operating efficiency of the electrodes. Experimental Reagents and Instrumentation All reagents were of analytical-reagent grade.De-ionized water was prepared using an Elga Primamaxima purification System (Elgar, High Wycombe, Buckinghamshire). HPLC was per- formed on a Perkin-Elmer Integral 4000 HPLC system equipped with a photodiode-array detector (Perkin-Elmer, Beaconsfield, Buckinghamshire, UK).An ESA Analytical Coulochem I1 electrochemical detector (ESA Analytical, Hun- tingdon, Cambridgeshire, UK) fitted with a Model 5011 high sensitivity analytical cell was placed downstream of the photodiode-array detector. The cell contained a porous graphite high efficiency electrode placed upstream of a glassy carbon amperometric working electrode.The detector was operated under the following conditions for routine analysis, Channel 1 (high efficiency electrode): potential, +40 mV; output range, 10 pA; offset, +5%; and filter, 1 s. Channel 2 (amperometric electrode): potential, +350 mV; output range, 10 PA; offset, +5%; and filter, 1 s. The column used was a Waters Radial Compression Module (RCM) containing a cartridge ( 10 cm X 8 mm id) packed with 5 pm Resolve octadecyl stationary phase (Waters UK, Watford, Hertforshire, UK).This stationary phase was recommended for use at low pH in preference to other octadecyl phases available from Waters. The mobile phase was 50 mmol l-I potassium dihydrogen orthophosphate, 500 mg 1-1 Na2EDTA, adjusted to pH 3.0 with 10 moll-' orthophosphoric acid and supplied at a flow rate of 1 .O ml min- 1.The injection volume was 10 pl. Calibration Standard solutions of AA were prepared at 0.5-20 mg 1-1 in the mobile phase described above. These solutions were stored cold under N2 when not in use, and were prepared fresh daily. Before calibration of the HPLC system, 2 X 100 pl injections of a10 Analytical Communications, January 1996, Vol33 solution containing 1 g 1- AA were made, followed by 2 X 100 p1 injections of mobile phase.This removed residual oxidizing species from the system and improved repeatability. Sample Preparation Samples (50 ml) were de-gassed by helium sparging, avoiding any contact with atmospheric oxygen, and were mixed with an equal volume of mobile phase. A sub-sample (10 pl) of the resulting solution was injected immediately.Results and Discussion Sample Preparation Significant lossses of AA occurred during sample preparation if oxygen was not totally excluded. Even when very gentle agitation was used to release dissolved CO2, losses of 10-20% of the measurable AA were not uncommon. When helium de- gassing was used, however, repeatable results and high recoveries were obtained. Although this severely limits the number of samples which may be processed on a given day, it is an absolute requirement for accuracy.Separation Conditions The method of Bode and Rose12 was slightly modified. The buffer strength was reduced from 200 to 50 mmoll-1, in order to improve column lifetimes, and Na2EDTA was added to the mobile phase at 500 mg 1-1 to inhibit metal-ion-catalysed autoxidation of AA.By operating at pH 3.0 it was possible to separate AA, which has pK, values of 4.0 and 11.3,l by ion suppression HPLC, and thereby dispense with the requirement for corrosive ion-pairing reagents. The column used was very stable at this pH. Although a Waters Nova-Pak 4 pm stationary phase gave almost equally good resolution of AA, the Resolve packing material was chosen because of its long-term durability at low pH.Optimization of Detection Conditions A hydrodynamic voltammogram is illustrated in Fig. 1, in which the potential of the upstream high efficiency electrode was varied from -300 to +400 mV, while maintaining the downstream amperometric electrode at a constant potential of +350 mV.At the chosen combination of analytical potentials (+40 mV and +350 mV at the high efficiency and amperometric electrodes, respectively) the ratio of peak areas from the two electrodes remained constant within the linear range of the method. Provided that the electrode response is reproducible, any gross departure from this ratio in samples would be indicative either of the presence of interfering compounds or of electrode fouling.This is of particular importance when 2.5 Q, Excessive Background Current J 0 Coulometric Electrode (variable potential) a" 1.0 0 Arnperometric Electrode (fixed potential of +350 mV) I 0.01 ' I ' I ' I ' I ' I " " -300 -200 -100 0 100 200 300 400 Applied PotentiaVrnV vs Pd reference Fig. 1 Hydrodynamic voltammetry of ascorbic acid.analysing beer, which contains a wide variety of reducing substances.13 Method Performance The method was linear in the range 0.5-20 mg 1-1 AA, which was suitable for the applications described here. Note that at higher concentrations of AA the high efficiency electrode began to exhibit a non-linear response, whereas the amperometric electrode response remained linear. This is important both from the point of view of calibration and also if peak area ratioing is used as a measure of peak purity.The limit of detection was 0.2 mg 1-1 AA in beer for the method as described, although higher sensitivity could be obtained by varying the electrode potentials and injection volume. Twenty consecutive injections of a beer sample containing 10 mg 1-1 added AA, diluted with an equal volume of mobile phase, gave s, values in peak area of 1.3 and 1.4% at the coulometric and amperometric detectors, re- spectively, and in peak height of 1.2 and 0.9%, respectively.This demonstrated not only the repeatability of the analysis, but also the resistance of the electrodes to fouling by sample components. The inclusion of Na2EDTA was vital to maintain repeatability of the analysis.Eight consecutive injections, made over an 80 min period, of a beer sample without Na2EDTA added, showed a gradual decrease to 50% of the starting AA peak area. A similar effect was observed for a standard solution of AA prepared in water without Na2EDTA addition. Sample stability was found to improve with Na2EDTA present, and maximum stability was obtained through the addition of 0.5 g 1-l Na2EDTA to the mobile phase, and by using this mobile phase to dilute samples before analysis.The method was also used for the analysis of a limited number of non-alcoholic vitamin drinks and fruit juices. Conclusions The described method was shown to be suitable for the analysis of AA in beer, and should be adaptable to the analysis of other beverages or solid foodstuffs. This method offers distinct advantages over previously reported methods for analysis of AA in beer in terms of accuracy, freedom from interferences, and repeatability.Furthermore, use of the electrochemical detector allows the determination of low residual AA levels in beers that have deteriorated during prolonged storage. References 1 2 3 4 5 6 7 8 9 10 11 12 13 Wong, D.S. W., Mechanism and Theory in Food Chemistry, Van Nostrand Reinhold, New York, 1989, ch. 10, pp. 359-365. Irwin, A. J., Barker, R. L., and Pipasts, P., J . Am. SOC. Brew. Chem., 1991, 49, 140. Official Methods of Analysis of the Association of Official Analytical Chemists, AOAC, Arlington, VA, USA, 14th edn., 1984, sec. 43.064. Finley, J. W., and Duang, E., J . Chromatogr., 1981, 207, 449. Seiffert, B., Swaczyna, H.. and Schaefer, I., Dtsch. Lebensm. Rundsch., 1992, 88, 38. Nisperos-Carriedo, M., Buslig, B. S., and Shaw, P. E., J. Agric. Food Chem., 1992,40, 1127. Knudson, E. J., and Siebert, K. J., J . Am. Soc. Brew. Chem., 1987,45, 33. Moll, N., and Joly, J. P., J . Chromatogr., 1987, 405, 347. Iwase, H., and Ono, I., J. Chromatogr. A . , 1993, 654, 215. Leubolt, R., and Klein, H., J . Chromatogr., 1993, 640, 271. Felton, S . P., Grace, R., and Halver, J. E., J . Liq. Chromatogr., 1994, 17, 123. Rose, R. C., and Bode, A. M., Biochern. J., 1995,306, 101. Chapon, L., Louis, C., and Chapon, S., European Brewing Conven- tion, Proceedings of the 13th Congress, Estoril, 1971, Elsevier, Amsterdam, 1972, p. 307. Paper 510 7096A Received October 27, 1995 Accepted December 5, 1995