New Trends in Liquid Chromatography and Their Utilization in Analysis of Beer and Brewery Raw Materials. Part 3. Comparison of HPLC and UHPLC Determination of α- and β -Acids

The traditional method of High Performance Liquid Chromatography (HPLC) was compared with the currently increasingly used Ultra-High Performance Liquid Chromatography (UHPLC) on a representative group of 11 hop samples. The results of this work confirmed that the UHPLC technique ensures high throughput also in determination of α- and β-acids. Another benefit of this method is the decrease of result uncertainty which is determined by high separation properties of the UHPLC system. While original HPLC method originating from Analytica EBC 7.7 provides 4 separate peaks (co-elution of n-adhumulone and adhumulone and also of n-lupulone and adlupulone), all 6 analyzed bitter compounds were completely separated using UHPLC.

1 INTRODUCTION

The first step of all types of analyses is the extraction of α- and β-acids (soft resins) into a nonpolar solvent. At the beginning, the titration method used the extraction into toluene (Analytica-EBC, 2005a), but this solvent was later replaced by diethylether in acidic conditions (Analytica-EBC, 2005b). The extract is titrated with lead acetate leading to formation of complexes with α-acids and the conductivity of this system is measured at the same time. The content of α-acids in the hop is expressed as % (w/w) alias LCV (lead conductance value). This value summarizes the content of α-acids, however, the reaction is less specific and the determined value is less accurate in the presence of any compounds reacting with lead acetate. The proportion of β-acids is approximately the difference between the total content of soft resins and the LCV value. Ethereal extract can be analyzed by HPLC following its centrifugation and dilution in methanol. Chromatographic method enables either complete or partial separation of α-acid zones, cohumulone, n-humulone, and adhumulone. The efficiency of the separation is dependent on the chromatographic column and mobile phase used. A similar chromatographic behavior was recorded in case of β-acids, colupulone, n-lupulone, and adlupulone. In contrast to titration methods, the HPLC methods are more specific and provide results better corresponding to real soft resins quantity.

It is evident that accuracy and precision of determination of α-acids in hop and hop extract affect the marketing and trade relations. Hence, all subjects as breweries, growers, and traders, are interested in accurate and comparable results of analytical laboratories. Good reproducibility and accuracy of results obtained by HPLC methods is ensured by using unified international calibration standard ICE3, which is presently already the third in a series. European laboratories follow the methods of European Beer Convention (Analytica EBC, 1998) and they can regularly participate in international proficiency testing organized by MEBAK. RIBM takes part in proficiency testing organized by MEBAK regularly two times a year and in BAPS (Brewing Analytes Proficiency Testing Scheme) regularly six times a year. Additionally, methods for hop analysis are accredited by the Czech Accreditation Institute (CIA).

EBC methods for α- and β-acids are developed for the most commonly used HPLC systems. In parallel with the development of liquid chromatography instrumentation the UHPLC systems start replacing it in the analytical field. The benefit of this new system is not only reduction of analysis time and consumption of organic solvents, but also more repeatable results with considerably lower uncertainty (Olšovská et al., 2012b). Information about UHPLC was described in more detail in previous issues of this journal (Olšovská et al., 2012a,2012b). The total analysis time of the content of bitter acids in the hop could be eventually reduced by an effective pressure extraction procedure, which was developed also at RIBM (Čulík et al., 2009). The rapid extraction together with modern UHPLC technique could bring to laboratories the possibility of sample high-throughput, and, moreover it could ensure more accurate data on the content of α-acids for traders in the hop commodity.

2 MATERIAL AND METHODS

Hop samples with the content of α-acids ranging from 1.4 to 8.5 % (w/w) were used for comparison of UHPLC and HPLC methods. The samples were extracted and prepared for LC analysis according to EBC method 7.7 (Analytica – EBC, June 2005).

The international calibration standard ICE3 was used for the method calibration; the total content of α- and β-acids was 44.64 % (w/w) and 22.28 % (w/w), respectively. The calibration solution was prepared as follows. ICE3 (0.5 g) was weighed with 0.1 mg precision, dissolved in 100 ml methanol, and diluted ten times with methanol (Gradient Grade, Sigma Aldrich).

Chromatographic conditions HPLC

α- and β-acids were analyzed on chromatographic column with C-18 Hop reversed phase (250 x 4 mm, particle size 5 μm), Machery Nagel, Germany. The separation was in isocratic mode. The mobile phase was a mixture of 850 ml of methanol (for HPLC, Sigma Aldrich), 190 ml of water (Millipore, purity max. 5 ppb of organic compounds) and 5 ml of phosphoric acid (analytical grade, Merck). The column temperature was 35 °C, the flow rate was 0.8 ml/min, and the injection volume was 5 μl. α- and β-acids were detected in UV at 314 nm according to Analytica EBC.

Analyses were performed on SpectraSYSTEM liquid chromatograph (TSP, USA) with diode array detector (DAD). The data collec tion was carried out using chromatography software ChromQuest for Windows NT.

Chromatographic conditions UHPLC

UHPLC analyses were performed on BEH C18 Waters column (2.1 x 50 mm I.D, particle size 1.7 μm), column temperature was 25 °C, and mobile phase flow was 0.4 ml/min. The two-component mobile phase was formed by aqueous solution of TFA-water (0.1:99.9 v/v) (A) and acetonitrile (B). The separation of analytes was carried out by gradient elution from a starting point at 50 % (B) to 80 % (B) within 6 min. The analytical run was terminated with an equilibration step of 2 min. The injection volume was 2 μl. α- and β-acids were detected by UV at 314 nm.

The Acquity UPLCTM (Waters) chromatograph was used for the analyses together with 2996 PDA detector operating at wavelengths ranging from 194 to 600 nm. Data were processed by Empower 2 (Waters) software.

3 RESULTS AND DISCUSSION

HPLC chromatogram obtained according to EBC 7.7 method is depicted in Fig. 1. α-Acids elute in two zones; cohumulone elutes in a discrete zone whilst n-humulone and adhumulone co-elute in a common zone. A very similar elution profile was obtained when β-acids were separated. When the UHPLC method was optimized, a complete separation of all three zones was obtained for α- and β-acids (see Fig. 2). An absolutely different type of column was used in UHPLC mode compared to HPLC. This column is based on the principle of “sub-2 μm” and we described this special column type in our previous article (Olšovská et al., 2012a). In the literature one can usually encounter the phenomenon of multiple increase of performance of UHPLC columns as compared with HPLC columns. Using of a new UHPLC method with modified mobile phase (with a higher content of more polar component) lead to a total separation of peaks in the case of n-humulone and adhumulone (and analogously n-lupulone and adlupulone), which were not at all separated in HPLC. Furthermore, analysis time significantly decreased from 25 min in HPLC to 8 min; this facilitates high-throughput of samples and cost reduction of solvent consumption. A very simple mobile phase, 0.1 % TFA in water (v/v) and acetonitrile, was used in the UHPLC mode. The consumption of organic modifier was minimal due to the speed of analysis (approximately 2 ml per one analysis including equilibration step), which leads to the price reduction of the method. When bitter acids were analyzed by HPLC, 17 ml of organic modifier (methanol) was used.

Fig. 1 HPLC chromatogram of hop extract (Sladek variety) Chromatographic conditions: Column C18-Hop (250 x 4 mm, particle size 5 μm), Macherey Nagel; column temperature 35 °C; mobile phase flow 0,8 ml/min; mobile phase composition 850 ml of methanol, 190 ml of water and 5 ml of phosphoric acid; isocratic elution; injection volume 5 μl; UV detection at 314 nm

Fig. 2 UHPLC chromatogram of hop extract (Sladek variety) Chromatographic conditions: Column BEH C18 (2.1 x 50 mm I.D, particle size 1.7 μm), Waters; column temperature 25 °C; mobile phase flow 0,4 ml/min; mobile phase composition (A) TFA–water (0.1:99.9, v/v), (B) acetonitrile; gradient elution from 50% (B) to 80% (B) within 6 min, equilibration 2 min; injection volume 2 μl; UV detection at 314 nm

The UV spectrum of α-acids shows three maxima at 237, 285, and 322 nm (see Fig. 3). The wavelength of 314 nm and one-point calibration was used according to EBC. 7.7. The content of α- and β-acids was measured by HPLC and UHPLC under these convention conditions in 11 hop samples. Tab. 1 gives an overview of analyzed hop samples together with LCV values. The values of total α- and β-acids content obtained by the two chromatographic methods together with their difference Δ are summarized in Tab. 2. The theoretical value of repeatability r95 calculated for individual levels follows from EBC 7.7 and expresses the permitted difference between two parallel results in one laboratory. The values of real and theoretical difference derived from method repeatability were compared. As shown in Tab. 2, UHPLC gives moderately higher results in case of α-acids (on average 0.13 % higher). Nevertheless, for all samples (except for one case) these differences are lower than the permitted repeatability. Both positive and negative differences were found with α-acids; except for one case they were again below theoretical repeatability.

Fig. 3 UV spectrum of cohumulone The UV spectrum was measured using UHPLC-DAD detector during the separation of α-acids under conditions mentioned in Materials and Methods. The other α-acids had the same UV maxima at 237, 285 a 322 nm

Tab. 1 An overview of tested hop samples and their lead conductance value (LCV)

Tab. 2 Comparison of the content of α- and β-acids in tested hop samples obtained by HPLC and UHPLC

The two methods can thus be considered comparable. Two hop samples (Sladek and Saaz) with different concentration levels of α- and β-acids were chosen for comparison of validation parameters. Both hop samples were measured repetitively (n=8) chromaand the repeatability and relative uncertainty estimate (k=2) were calculated. The Sladek variety (KH 10.18) contained 8.57 ± 0.1 % (w/w) and 3.81 ± 0.04 % (w/w) of α- and β-acids, respectively. The Saaz variety contained 2.81 ± 0.03 %(w/w) and 5.57 ± 0.06 % (w/w) of α- and β-acids, respectively. As follows from Tab. 3, the repeatability and relative uncertainty estimate for α- and β-acids are significantly (more than ten times) lower in all cases when the UHPLC is used.

Tab. 3 The comparison of HPLC and UHPLC validation parameters

4 CONCLUSIONS

This study is the last of a series of three studies, which dealt with a new generation of chromatographic columns and related liquid chromatography instrumentation known as Ultra High Performance Liquid Chromatography (UHPLC). In the first part, we introduced this new technology and described its enormous advantages over the classical HPLC method, which result from the basic principle of UHPLC instrumentation, and which will undoubtedly positively reflect in economical costs per analysis. Two useful studies were performed in the area of beer analysis to verify this claim.

The usefulness of the new method was confirmed on the example of iso-α-acid determination in beer and further on determination of α- and β-acids in hop. Besides an economic benefit (reduction of time, energy and organic solvents as a result of significant analysis time reduction) this method affords measurement with lower variance of results. The current methods of beer analysis are based on some conventions that presume the use of old HPLC instrument types. Based on both studies it is possible to state that the use of UHPLC provides comparable results with HPLC, when defined measure parameters (the way of detection, calibration and quantification) are maintained. Therefore, both types of LC can be used without compromising the accuracy of the results.