Pixabay/Bernadette Wurzinger: Determination of Rebaudioside A in Mixed beer, Beer-based Beverages and Lemonades
The plant Stevia rebaudiana (Stevia sweet) is becoming a favourite sweetener in the food industry. Sweet taste is caused by diterpenic glycosides, so-called steviol glycosides, which lack calories while being 300 times sweeter than sucrose. While all the other glycosides are slightly or even disagreeably bitter, rebaudioside A possesses desirable sensorial properties. Because this sweetener is used to sweeten beverages including mixed beer, a fast routine analytical method was developed using HPLC with UV detection for determination of rebaudioside A in mixed beer and lemonade. The HILIC principle, which is suitable for separation of polar and hydrophilic analytes, was used for separation of steviol glycosides. The method is selective, reproducible and has a high recovery of 90-100 %.
Sweet stevia (Stevia rebaudiana) is native to South America, plentifully spread in Asia, especially in Japan and is used as a non-calorifi c sugar substitute.
French chemists isolated basic glycosides stevioside and rebaudioside A from plant leafs in 1931. They are diterpenic compounds with high sweetness, approximately 300 times sweeter than sucrose, and are contained in all parts of this plant with the exception of roots. Dry leaves contain about 10% of stevioside and 2–4% rebaudioside A. The stevia leaves contain altogether eight diterpenic steviol glycosides: stevioside, steviolbioside, rebaudioside (A, B, C, D, E) and dulcoside A; their concentration is minor. Diterpenic steviol glycosides have a slightly bitter taste except for rebaudioside A, which has minimal bitterness. The major produced compound, stevioside, exhibits appreciable unpleasant bitterness (Prakash et al., 2008) and rebaudioside A extract of is therefore the only one used in food and beverage industry. Rebaudioside A extracts can differ in the content and purity of the final product depending on plant origin and extraction technique. Apart from steviol glycosides, extracts of stevia can contain also other compounds without any sweetness.
The worldwide “stevia rediscovery” as a food supplement required the overcoming of numerous obstacles presented by lobbyist groups since it represented a significant competition to other artificial sweeteners then in use. The US Food and Drug Administration FDA recognized stevia as a food supplement only in 1995; before that sloustevia could be used only in cosmetic industry although its safe use was known more than 1500 years. The position of the FDA was influenced by the misleading studies by the Brazilian Prof. Mauro Alvarez in 1988, which described stevia as having contraceptive properties and even causing infertility of men and women. A number of studies had to be performed to disprove this myth. Nearly 40 years thus elapsed before FDA approved stevia as a common sweetener without any contraceptive effect. None of the current studies has confirmed its effect on human fertility (Kumar et al., 2008).
Also in European Union extensive discussions and often mutually contradictory studies preceded the recognition of stevia as food supplement with designation E960.
Steviol glycosides were registered by EU as permitted sweetener only on December 2, 2011. This approval concerns rebaudioside A with the content of at least 90% of pure substance and mixtures of sterol glycosides with a total content higher than 95%.
According to the Joint Committee of Experts for Food Additives (JEFCA) of World Health Organization (WHO) and of Food and Agriculture Organization of the United Nations (FAO) the acceptable daily intake of steviol glycosides is 4 mg of steviol per kg body mass a day (US Food and Drug Administration, 2008, JEFCA 2008). The sweetener rebaudioside A has found application in all branches of food industry as a non-calorific sweetener; it is used especially for sweetening of lemonades and syrups in beverage industry. Fruit beers sweetened with rebaudioside A have recently appeared increasingly often on the market. The slightly bitter taste of rebaudioside A harmonizes with the bitter beer character and it replaces more and more artifi cial sweeteners such as aspartam.
The widespread method for determination of steviol glycosides is HPLC, though there exist also some methods using thin layer chromatography, capillary electrophoresis or spectroscopic methods (NIR, VIS). The choice of HPLC separation conditions is dependent on the polar character of the glycosides, the choice of UV detection restricts the range of applicable absorbance to short wavelengths. An overview of earlier methods used for determination of steviol glycosides was written by Bovanová (1998). The HPLC methods describe a separation of steviol glycosides especially on columns with NH2 or C18 functional groups; the mobile phases include mostly acetonitril or methanol with water in various rations. Afandi et al (Afandi, 2013) described the optimization of conditions to obtain rebaudioside A from stevia leaves using Soxhlet extractor with subsequent HPLC determination on C18 column with mobile phase acetonitrile/water 80/20 (v/v). Bovanová et al. (Bovanová, 1998) published the preparation of liquid samples by SPE prior to HPLC analysis.
A new technique based on hydrophilic interactions in liquid chromatography (HILIC) is currently used to separate very polar and hydrophilic compounds. Alpert (Alpert, 1990) suggested term “chromatography of hydrophilic interactions” (HILIC) for chromatography on columns typical for systems with normal phases, which however are using water-organic mobile phases like the RP systems. The word “hydrophilic” characterizes the affinity to water added to mobile phases in HILIC separations on polar columns. The HILIC technique was first used for separation of saccharides, amino acids and peptides (Strege,1998). However, the number of newly developed columns for HILIC and applications for separations in the area of environment, foods, natural compounds, synthetic drugs and nonionic surfactants has grown very quickly in the last years. An important benefit of HILIC separations is higher sensitivity of LC-MS analyses because of improved electrospray ionization (ESI) in mobile phases with high content of acetonitrile (Brown at al., 2002; Jandera, 2011).
This principle can be used for HILIC separation of steviol glycosides with UV detection (Thermo, 2012). Compared to UV detection, the use of ELS detector (evaporative light scattering) provides improvement of the limits of quantification and detection. The disadvantage of this detector is the necessity to use volatile mobile phase.
Though the relevant literature contains many methods for determination of rebaudioside A and other steviol glycosides, only a few food laboratories use this method routinely. This is reflected in the cost of this analysis. This study therefore aimed at setting up an analytical method for the determination of rebaudioside A (Fig. 1) in mixed beer and lemonade using an available HPLC technique with UV detection and HILIC separation principle.
The structure of rebaudioside A with prevailing –OH groups determines its high polarity. The probable value of logarithm of distribution coefficient between water and octanol (log P) is – 3.12. Values lower than 1 characterize the polar compounds (author’s comment, see the Dictionary of Food Compounds on CD-ROM, 2004).
Fig. 1 Structure of rebaudioside A
Samples of lemonades are only pre-purified before analysis in view of the higher concentration of rebaudioside A. Samples of mixed beer, with much lower content of the sweetener, must be pre-concentrated so that the measured values will fi t into the range of the calibration curve of rebaudioside A. The extraction columns C18, E 1000 mg/6 ml (Strata, Phenomenex, USA) were used for purification and pre-concentration. The extraction column was first conditioned step by step with10 ml of methanol, 10 ml of deionized water and after that a sample of lemonade (5 ml) or mixed beer (25 ml) was applied on the column. The sample captured on the adsorbent was washed (10 ml of 20% acetonitrile in deionized water) and eluted into 5 ml of the mixture of acetonitrile – methanol in a 1:1 (v/v) ratio. The purified sample was loaded on the column after filtration on cellulose filter.
The range of the calibration curve was chosen to comply with determination of rebaudioside A in both lemonades and in mixed beers with concentration lower than 1 mg/l (these samples were preconcentrated). The calibration solutions were prepared by dissolving a standard in mobile phase. The stock solution of rebaudioside A was prepared by weighing 100 mg of rebaudioside A standard with accuracy ± 1 mg and its dissolution in 100 ml in graduated flask.
The other calibration levels of analyte, i.e. 1, 5, 10, 20 and 50 mg/100 ml were prepared by dissolving stock solution in appropriate volumes of mobile phase.
HPLC with UV detection at 210 nm on Dionex Ultimate 3000 liquid chromatograph (Thermo Scientific) with Chromeleon 7 control and processing software was used for determination of rebaudioside A. The column Acclaim Mixed – Mode WAX 1 (2.1 x 150 mm; 5 μm; Thermo Scientific) in HILIC mode was used for separation of polar rebaudioside A from other compounds present in beer and lemonade matrices. The mobile phase consisted of acetonitrile and 10 mM ammonium formate at pH 3.00 ± 0.05 in a ratio of 80/20 (v/v) in isocratic mode. The flow rate of mobile phase was 0.5 ml/min, the column was temperature-controlled at 40 °C, and injection volume was 5 μl.
The method was optimized for mixed beer and lemonade matrices, and its validation involved verification of linearity, repeatability and recovery of the method.
As follows from Fig. 2, the elution zone of rebaudioside A is distinctly separated from other polar analytes even for its very low concentration in the sample. The last minority component elutes at 7.04 min while the elution zone of rebaudioside A is at 7.95 min. The total separation of rebaudioside A from other components complies with the requirements for a good selectivity of the method. This is given by the use of HILIC mode which has a markedly better selectivity to polar compounds and thus to rebaudioside A than the usual non-polar columns. Another advantage of the HILIC mode is a higher sensitivity; it is 10 – 1000-fold higher for hydrophilic analytes than reverse mode techniques.
Fig. 2 Chromatogram of separation of beer sample containing rebaudioside A. Column Acclaim Mixed – Mode WAX 1 (2.1 x 150 mm; 5 μm), mobile phase acetonitrile and 10 mM ammonium formate 80/20 (v/v) in isocratic mode. Flow rate of mobile phase 0.5 ml/min, temperature of column 40 °C, injection volume 5 μl. The retention time of rebaudioside A is 9.75 min
The linearity was verified in a calibration range of 1-50 mg/l where the detector response of rebaudioside A has clearly a linear dependence. It is evident from table 1 that the concentrations of rebaudioside A in the original mixed beer appear in the close proximity of the lower point of the calibration curve.
Table 1 Repeatability of the method
A Cola-type lemonade and mixed beer flavored with fruit extract, both with the stated content of rebaudioside A, were used for determination of repeatability and recovery of the method. The concentration of the monitored analyte was determined for both samples. Each sample was extracted on a SPE column and subsequently analyzed on a chromatographic column. The determined content of rebaudioside A in each sample and repeatability (expressed as a limit of repeatability) are given in Table 1. The permissible difference between two parallel determinations (limit of repeatability) in both the beer and the lemonade is 0.42 mg/100 ml, which attests to the good precision of the method.
The repeatability of the method was determined using standard addition of rebaudioside A to samples of mixed beer and lemonade. Two distinct concentrations of the standard were prepared to eliminate mistakes in the dosage of different volumes of standard added into the beer and the lemonade. A solution having the concentration of 1000 mg/100 ml water was prepared for spiking of lemonade while a solution of 100 mg/100 ml water was prepared for spiking beer. The standard was added into 100 ml volumetric flasks to achieve two different concentrations. With both matrices, an adequate volume of the standard was filled up to the graduation mark.
The dosage of 1 ml or 2 ml of standard solution into lemonade caused an increase of the analyte concentration to about 10 resp. 20 mg/100 ml. The dosage of 2 ml resp. 4 ml of standard solution into beer signified the increasing the concentration of analyte by about 2 or 4 mg/100 ml. Every concentration level was determined twice. The results are given in Table 2. The recovery of the method for lemonades and mixed beer ranges in the interval from 101.9% to 102.8% and from 90.2% to 93.1%, respectively. This fact is apparently caused by larger matrix effect in beer and by pre-concentration of beer before analysis. Due to the solubility of the analyte in alcohol its partial leak occurs during of sorption and washing of SPE column.
Table 2 Recovery of the method
The HILIC technique permits the separation of rebaudioside A from other polar compounds also in the very complicated beer matrix, which would not be possible using conventional polar column due to the high polarity of rebaudioside A. The elimination of co-elution of rebaudioside A with other polar compounds assures the accuracy of determination without overrating of the results. All other analytical parameters are in the agreement with the results of other published methods with UV detection in the range of short wavelengths.
The newly developed method to determination of rebaudioside A shows satisfactory precision (repeatability 0.42 mg/100 ml) and accuracy (recovery 90–100%). The solubility of rebaudioside A in alcohol explains the equilibrium shift on extraction column, and thus also a lower recovery from beer in comparison to non-alcoholic lemonade. The concentration of rebaudioside A in mixed beer (or a beer-based beverage) and lemonade after its treatment on the SPE column complies with a linear calibration relationship in the range from 1 to 50 mg/100 ml of the analyte.