A Comparison of Alternative High-Performance Liquid Chromatographic Systems for Measuring Bitterness in Beer

Abstract

Alternative systems for measurement by high-performance liquid chromatography of iso-α-acids in highly bittered stout beers were evaluated in an examination of contributions to error in the assay. The objective was to devise rapid and accurate systems for measuring iso-α-acids in pre-isomerized hop extracts and in stout beers bittered either partially or wholly with pre-isomerized extract. Special attention was paid to the preparation of standards, pretreatment of beer samples, calibration protocols, calibration frequency, chromatographic mobile phases and measurement wavelength. Variability between assays was smallest (coefficient of variation <3%) when a simplified sample pretreatment was used, in which high-molecular-weight interferences were removed from stout beers by precipitation with methanol. The main cause of assay variability was changes in the shape, area, and resolution of peaks, a tendency that was greater during chromatography with ion-pairing mobile phases than with ion-suppressing mobile phases. This change was not compensated for by internal standardization but was countered by increasing calibration frequency with “spiked blanks.” Accordingly, chromatographic variability with a single calibration cycle was minimized (cv 0.4%). Samples of stout beers containing 35–100 mg/L of iso-α-acids before bottling were shown to contain only about 70–95% of their iso-α-acid contents in solution after storage in bottles for three months at 18°C. Losses in soluble iso-α-acids were greatest in those beers containing high addition rates of pre-isomerized hop extracts.