Regulatory control of energy drinks using 1H NMR spectroscopy

Abb. 5: Weindiferenzierung nach Jahrgang hauptsächlich 2008, 2009 konnte mit 97 % iger Wahrscheinlichkeit richtig zugeordnet werden ( Abb. 5 ). Weiterhin erscheint die quantitative Bestimmung von Weininhaltstofen über einen weiten Konzentrationsbereich möglich ( Zucker, organische Säuren, Aminosäuren, Phenole, etc. ) – Targeted Analyse. Durch Non-targeted-Analyse lassen sich Verfälschungen und Manipulationen sowie Weinfehler erkennen. 1. Godelmann R, Fang F, Humpfer E, Schütz B, Bansbach M, Schäfer H, Spraul M, in Vorbereitung Originalbeiträge Regulatory Control of Energy Drinks Using 1H NMR Spectroscopy K. Wegert1,2, Y. B. Monakhova1,3,4, T. Kuballa1, H. Reusch1, G. Winkler2, D. W. Lachenmeier1 1 Chemisches und Veterinäruntersuchungsamt ( CVUA ) Karlsruhe, 2Department of Life Sciences, Albstadt-Sigmaringen University, Sigmaringen, 3Bruker Biospin GmbH, 4Department of Chemistry, Saratov State University, Russia Abstract 400 MHz 1H nuclear magnetic resonance ( NMR ) spectroscopy was used in the context of food surveillance to control the regulatory limits of selected constituents of energy drinks. he preparation of the samples required only degassing, addition of 0.1 % of TSP in D2O for locking and referencing and phosphate bufer ( pH 4.0 ) for chemical shits adjustment. he quantiication of cafeine, glucuronolactone, taurine and inositol was simultaneously possible using external calibration curves and applying TSP as internal standard. he applicability of the method was proven by analysis of 73 samples of energy drinks. he average cafeine concentration was 0.30 g / L, which is below the maximum limit of 0.32 g / L. Introduction Alcohol-free beverages containing ingredients with presumed stimulant properties ( so-called energy drinks ) are increasingly ofered on the market [1], and, although they might be harmless, overdoses or combination of energy drinks with alcoholic beverages could be harmful to the health of some consumers under certain circumstances [2]. Energy drinks oten contain ingredients such as cafeine, the cafeine-containing plant guarana ( Paullinia cupana ), glucuronolactone, inositol and taurine. To protect the consumer from harm arising from these beverages, some regulations regarding composition and labelling of energy drinks exist. On a European level, the regulation ( EU ) No 1169 / 2011 on food information demands that beverages with high cafeine content ( i. e. in excess of 150 mg / l ) must be labelled with the following warning: ‘High cafeine content – not recLebensmittelchemie 66, 129–168 (2012) ommended for children or pregnant or breast-feeding women’ as well as with the actual quantity of cafeine. he EU regulation does not contain requirements for any other constituent of energy drinks. On a German national level, energy drinks are deined in the “Fruchtsat- und Erfrischungsgetränkeverordnung ( amendment of 21.05.2012 )” as cafeine-containing sot drinks, which may additionally contain taurine, inositol and / or glucuronolactone. he German regulation also deined maximum limits for these constituents in energy drinks, which are 320 mg / l for cafeine, 4000 mg / l for taurine, 200 mg / l for inositol and 2400 mg / l for glucuronolactone. To control the demands of EU and national legislation, analytical methods are required to determine the four regulated substances. Currently, several diferent methodologies have to be used, e. g. cafeine is typically determined using HPLC / DAD according to the German reference method ( § 64 LFGB L 18.0016 ) [3]. As taurine does not possess a chromophore, a diferent methodology is required and we have applied amino acid analysis with post-column ninhydrin derivatization or alternatively a screening analysis with FTIR [1]. Glucuronolactone may be determined as 1-phenyl-3-methyl-5-pyrazolone ( PMP ) derivative using HPLC [4]. For inositol, HPLC with post-column derivatization with ferric perchlorate has been suggested [5]. For eiciency and economic reasons, it would be desirable to have a single method suitable for simultaneous identiication and quantiication of all four substances. A possible methodological candidate certainly is NMR, which has been applied in analysis of alcoholcontaining [6–10] or alcohol-free beverages [11–16], most typically for qualitative analysis. We have recently shown that caffeine and several additives can be simultaneously quantiied in cola beverages using 1H NMR [17]. Based on this research, this study will be the irst to use 1H NMR for quantitative analysis of the regulated constituents in energy drinks. Materials and Methods Samples and chemicals: A total of 73 energy drinks were analyzed using NMR. he samples were randomly selected from supermarkets in Karlsruhe and Freiburg, Germany. Additional brands were purchased from an internet shop specialized in energy drinks. Reference substances ( cafeine, glucuronolactone, taurine and inositol ) as well as sodium azide, H3PO4 and HCl were purchased from Sigma Aldrich ( Steinheim, Germany ), KOH and KH2PO4 were purchased from Merck ( Darmstadt, Germany ). he NMR bufer was prepared by dissolving 10.21 g of KH2PO4 and 9.75 mg of sodium azide in 50 mL of pure water and then by adjusting the pH to 4.0 with H3PO4 or KOH. Sodium azide was used for the preservation of the samples because this addition was proved to be advantageous, so that even longer storage of samples at room temperature in the autosampler did not lead to signiicant changes in the spectrum [17]. Sample and standard preparation: he sample preparation is based on a previous method developed for measuring cola samples [17]. he energy drinks were degassed by ultrasonication for 10 min and 800 µL of the degassed solution were combined with 100 µL of an internal standard ( D2O containing 0.1 % of TSP ( sodium salt of 3-( trimethylsilyl )-propionate acid-d4 ) and 100 µL of NMR bufer ( see above ). We did not observe considerable shit changes in NMR spectra of our samples, and, therefore, more precise pH adjustment of the mixtures was not necessary [17]. 600 µL of the inal solution were poured into an NMR tube for direct measurement. Stock solutions with concentrations of 10 000 mg / L were prepared in deionized water from the pure substances. By diluting the stock solutions in pure water, about 10 calibration standards were prepared so that the calibration curves were accomplished 143 across the concentration ranges encountered in the sot drinks. he calibration standards were then prepared by the same procedure as it was used for the samples. 1 H NMR measurements at 400 MHz: All NMR measurements were performed on a Bruker Avance 400 Ultrashield spectrometer ( Bruker Biospin, Rheinstetten, Germany ) equipped with a 5-mm SEI probe with Z-gradient coils, using a Bruker Automatic Sample Changer ( B-ACS 120 ). 1H NMR spectra were acquired similar to our previous study of carbonated cola beverages [17]. However, to ensure increased throughput, only 64 scans were acquired, which decreases measurement time to only 15 min. In comparison, the spectra acquisition for cola beverages took 31 min. Additionally, 2D J-resolved NMR spectra were acquired using 4 scans and 16 prior dummy scans of 8.2k points with a spectral width of 16.6595 ppm ( F2 ) and 0.1952 ( F1 ) ppm, a receiver gain of 28.5, an acquisition time of 0.61 s. he data were acquired automatically under the control of ICON-NMR ( Bruker Biospin, Rheinstetten, Germany ). All NMR spectra were phased, baseline-corrected and calibrated by the TSP signal at 0.0 ppm. Quantiication and validation studies: he spectra of energy drink samples were compared to the spectra of the standards. Separated peaks ( i. e. peaks not overlapped or interfered by matrix ) corresponding to each substance were identiied and integrated. he singlet at δ 7.92–7.89 ppm for cafeine was selected in 1D spectra. All other compounds were quantiied using 2D J-resolved spectra, i. e. the doublet at δ 3.62–3.66 ppm for glucoronolactone, the triplet at δ 3.24–3.26 ppm for taurine as well as the triplet at δ 3.30–3.29 ppm for inositol. It was found that in all cases a linear relationship exists between the concentration of the analyte and the ratio between the peak area of the analyte and the internal standard ( TSP ) ( R > 0.99 ). For the method validation, energy drink samples and standard solutions were analyzed several times daily ( intraday, n = 5 ) and over several days ( interday, n = 4 ). he limit of detection ( LOD ) and the limit of quantiication ( LOQ ) were calculated according to the German Institute for Standardization standard DIN 32645 [18] using a separate calibration curve. he recovery rate was ascertained by adding standard solution at two diferent concentrations to a real sample. For all calculations statistical signiicance was assumed at below the 0.05 probability level. Results and Discussion Fig. 1 shows the complete 1H NMR spectrum of a typical energy drink as well as a magniication of the δ 10–6.0 ppm region, where the resonances of aromatic protons can be seen. In general, an extensive spectral overlap of chemical shits of several compounds exists especially in the mid-ield region. Moreover, Fig. 1: 1H NMR spectrum of an energy drink. he insert shows 1H NMR spectra in the 10–6 ppm region. 144 resonances of some major compounds ( for example, sugars ) have much higher intensity than our compounds of interest. Nevertheless, we were able to develop a method aimed at simultaneous quantiication of cafeine, glucuronolactone, taurine and inositol in energy drinks. he singlet of cafeine is located quite distinctly in the NMR spectra of energy drinks and we encountered no diiculties to quantify this compound in the aromatic region. he singlet peak in the δ 7.90–7.80 ppm range ( derived from the imidazole ring of the molecule ) can easily be distinguished from other compounds even in 1D spectra. As an example, Fig. 2 shows the signal of cafeine in standard solutions and an energy drink sample. Fig. 2: 400 MHz 1 H NMR spectra of cafeine in standard solutions and energy drink sample he other compounds have signals only in the mid-ield region, which are too complicated for interpretation in 1D ( Fig. 3A ). herefore, J-resolved spectra have to be used for quantiication. Using a triplet at δ 3.25 ppm for taurine, a triplet at δ 3.29 ppm for inositol and a doublet at δ 3.66 ppm for glucuronolactone in 2D spectra, a reliable identiication of these compounds is possible in most of our samples ( Fig. 3B ). However, in some cases it was not possible to identify target resonances even using 2D NMR. hus, more advanced techniques, such as multivariate regression or curve deconvolution, would be required for these samples. he maximum limits for the investigated compounds are comparably high and well within the linear range of the NMR method. he limits of detection and quantiication were 1.3 mg / L and 3.8 mg / L for cafeine, 3 mg / L and 9 mg / L for taurine, 5 mg / L and 15 mg / L for glucuronolactone, and 4 mg / L and 12 mg / L for inositol. he intraday and interday precisions ( expressed as rela- Fig. 3: 1H NMR spectra of an energy drink sample containing 548 mg / L inositol, 3746 mg / L taurine and 3213 mg / L glucuronolactone measured in 1D ( A ) and 2D Jresolved ( B ). tive standard deviation ) were 0.3 % and 0.4 % for cafeine, 4 % and 11 % for taurine, 7 % and 8 % for glucuronolactone, and 6 % and 8 % for inositol. he recovery was 97 % for cafeine, 102 % for taurine, 99 % for glucuronolactone, and 123 % for inositol. Lebensmittelchemie 66, 129–168 (2012) Tab. 1: Cafeine concentrations ( g / L ) in energy drinks Samples Energy drinks 2010-2012 (HPLC) Energy drinks 2012 (NMR) Combined results n 115 73 188 Min 0.10 0.08 0.08 P05 0.20 0.23 0.21 P25 0.28 0.28 0.28 To demonstrate the applicability of the method, 73 diferent energy drink samples were measured. As mentioned above, we were able to quantify cafeine in all samples and the distribution of cafeine contents is shown in Table 1. Additionally, Table 1 shows the distribution of cafeine results previously determined using a HPLC method. No signiicant diference was found between NMR and HPLC distributions ( ANOVA p = 0.46 ), so that both distributions can be combined, e.g., for the aim of exposure assessment ( Table 1 ). he average concentrations of cafeine, taurine, inositol and glucuronolactone were 0.3 g / L, 2.8 g / L, 0.2 g / L and 1.5 g / L. Only few samples exceeded the above mentioned limits of the German regulation for energy drinks ( however, it must be noted that the regulation was not yet in force during time of our sampling in March 2012 ). Median 0.30 0.30 0.30 Mean 0.29 0.30 0.29 P75 0.31 0.34 0.31 P95 0.33 0.38 0.37 Max 0.49 0.39 0.49 16. Spraul M, Schütz B, Rinke P, Koswig S, Humpfer E, Schäfer H, Mörtter M, Fang F, Marx UC, Minoja A ( 2009 ) Nutrients 1: 148–155. 17. Maes P, Monakhova YB, Kuballa T, Reusch H, Lachenmeier DW ( 2012 ) J Agric Food Chem 60: 2778–2784. 18. DIN 32645 ( 2008 ) Chemische Analytik: Nachweis-, Erfassungs- und Bestimmungsgrenze, Ermittlung unter Wiederholbedingungen. Begrife, Verfahren, Auswertung. Beuth Verlag, Berlin, Germany. Rapid determination of 4-methylimidazole in caramel colours using 1H NMR spectroscopy Y. B. Monakhova1,2,3, C. Schlee1, T. Kuballa1, R. Schneider1, D. W. Lachenmeier1 1 Chemisches und Veterinäruntersuchungsamt ( CVUA ) Karlsruhe, 2Bruker Biospin GmbH, Rheinstetten, 3Department of Chemistry, Saratov State University, Russia Conclusion he method was found to be suitable for the purpose of simultaneous quantiication of all regulated compounds in most energy drink samples ( approx. 90 % of the analysed samples ). In the rest of the samples, some or all resonances of taurine, inositol and glucuronolactone were interfered by matrix compounds, so that integration was impossible even in 2D spectra. he samples in question were typically cloudy due to contents of fruit material. Future research may be directed to either improving the sample preparation to avoid the interferences, or alternatively using multivariate regression techniques to quantify in the overlapped regions. Currently, NMR can provide a rapid screening analysis of energy drinks especially for cafeine, but samples above limits should be conirmed by HPLC before the results are used in expert opinions about regulatory non-conformance. Acknowledgements: he authors are grateful to Margit Boehm and Bernd Siebler for excellent technical assistance. 1. Triebel S, Sproll C, Reusch H, Godelmann R, Lachenmeier DW (2007 ) Amino Acids 33: 451–457. 2. Santa-Maria A, Diaz MM, Lopez A, de Miguel MT, Fernandez MJ, Ortiz AI ( 2002 ) Ecotoxicol Environ Saf 53: 70–72. 3. Waizenegger J, Castriglia S, Winkler G, Schneider R, Ruge W, Kersting M, Alexy U, Lachenmeier DW ( 2011 ) J Cafeine Res 1: 200–205. 4. Suzuki S, Hayase S, Nakano M, Oda Y, Kakehi K ( 1998 ) J Chromatogr Sci 36: 357– 360. 5. Franz H, Maier HG ( 1993 ) Deut Lebensm Rundsch 89: 276–282. 6. Lachenmeier DW, Frank W, Humpfer E, Schäfer H, Keller S, Mörtter M, Spraul M ( 2005 ) Eur Food Res Technol 220: 215–221. 7. Duarte IF, Barros A, Almeida C, Spraul M, Gil AM ( 2004 ) J Agric Food Chem 52: 1031–1038. 8. Lopez-Rituerto E, Savorani F, Avenoza A, Busto JH, Peregrina JM, Engelsen SB ( 2012 ) J Agric Food Chem 60: 3452–3461. 9. Monakhova YB, Kuballa T, Lachenmeier DW ( 2012 ) Appl Magn Reson 42: 343– 352. 10. Monakhova YB, Schäfer H, Humpfer E, Spraul M, Kuballa T, Lachenmeier DW ( 2011 ) Magn Reson Chem 49: 734–739. 11. Belton PS, Colquhoun IJ, Kemsley EK, Delgadillo I, Roma P, Dennis MJ, Sharman M, Holmes E, Nicholson JK, Spraul M (1998 ) Food Chem 61: 207–213. 12. Le Gall G, Puaud M, Colquhoun IJ ( 2001 ) J Agric Food Chem 49: 580–588. 13. Wilson AK, Myers C, Crull G, Curtis M, Patterson PP ( 1999 ) J Chem Educ 76: 1414– 1416. 14. Charlton AJ, Robb P, Donarski JA, Godward J ( 2008 ) Anal Chim Acta 618: 196– 203. 15. Spraul M, Schütz B, Humpfer E, Mörtter M, Schäfer H, Koswig S, Rinke P ( 2009 ) Magn Reson Chem 47 Suppl 1:S130–S137. Lebensmittelchemie 66, 129–168 (2012) Abstract A method utilizing 1H NMR spectroscopy has been developed to measure the concentration of 4-methylimidazole in caramel colours used as food additive ( E150a-d ). For sample preparation, a very simple dilution with water-ethanol mixture and addition of internal standard is suicient. 4-Methylimidazole produces a distinct peak of the methyl group of the molecule in the δ 2.27– 2.25 ppm range, where it can be distinguished from other matrix compounds. he method was shown to be of adequate sensitivity with a limit of detection ( LOD ) of 5.0 mg / L adequate to control the 4-methylimidazole regulatory limit. he developed method was applied for the analysis of 7 authentic food colourings. Only two products ( E150d ) contained detectable concentrations of 4-methylimidazole ( 662 mg / kg and 401 mg / kg ). he NMR method may be used by governmental or industry laboratories for routine control of food colourings. Introduction According to Regulation ( EC ) No 1333 / 2008 on food additives, plain caramel ( E150a ), caustic sulphite caramel ( E150b ), ammonia caramel ( E150c ) and sulphite ammonia caramel ( E150d ) are authorised at quantum satis for the use as food additive for colouring in a wide range of food products but are especially applied in certain beverages including cola. he term caramel in the sense of the food additives regulation relates to products of a more or less intense brown colour which are intended for colouring. It does not correspond to the sugary aromatic product obtained from heating sugars and which is used for lavouring food ( e.g. confectionery, pastry, alcoholic drinks ). he Regulation ( EU ) No 231 / 2012 laying down speciications for food additives speciies that not more than 200 or 250 mg / kg of 4-methylimidazole ( 4-MI ) may occur in E150c and E150d, respectively. his purity criterion is expressed on equivalent colour basis i.e. it is expressed in terms of a product having a colour intensity of 0.1 absorbance units. 4-MI was found as being carcinogenic in animal experiments [1] and was recently evaluated by the WHO International Agency for Research on Cancer ( IARC ) as “possibly carcinogenic to humans” ( group 2B ) [2, 3]. Some consumer organizations took this evaluation to postulate a “cancer fear over cola colourings” and demanded the minimization of 4-MI in the cola beverages [4]. Based on an evaluation from the European Food Safety Authority ( EFSA ), which has concluded that the efect of 4-MI is 145