Repository logo
 

Characterizing the impact of package type on different beer styles using advanced analytical tools

Date

2022

Authors

Fromuth, Kathryn Lenore, author
Prenni, Jessica, advisor
Sedin, Dana, committee member
Van Buiten, Charlene, committee member

Journal Title

Journal ISSN

Volume Title

Abstract

In 2020 there was over 9,000 breweries in the US, increasing the beer market competition and driving the importance of product stability under variable storage conditions. More breweries, specifically craft breweries, than ever before are choosing to package in cans due to ongoing effects of the current pandemic, the growing availability of smaller can line systems, and increased mobile canning options. Foundational beer stability research has focused on light lager styles packaged in bottles. Limited research has been conducted studying flavor stability in styles relevant to the American craft brewing industry, nor any comparisons of how package type (i.e., cans and bottles) affects flavor stability. Industry utilizes trained sensory panels to evaluate flavor stability; a resource that is both time consuming and expensive. Thus, this is a tool that is often inaccessible or inadequate for providing relevant and timely stability data. This research project, a collaboration between New Belgium Brewing and Colorado State University, aims to address the package-type knowledge gap and sensory panel restrictions by utilizing advanced analytical tools to characterize the changes in metabolite profiles over time between cans and bottles. A low-hopped amber ale (AA) and high-hopped India Pale Ale (IPA) were chosen for their distinct style and relevance to the American craft brewing industry. One batch of an IPA and AA was packaged into cans and bottles, then aged for a six-month period. The samples were stored under cold temperatures (4°C) for the first 30 days, and then at room temperature (20°C) for the subsequent time. Aliquots were collected biweekly for a total of 13 timepoints throughout the six-month aging period and stored at -80°C until chemical analysis. Chemical analysis was conducted by gas chromatography coupled to a mass spectrometer detector (GC-MS) and direct analysis in real time mass spectrometry (DART-MS) to address the research questions. Multivariate (MVA) and univariate (UVA) statistical analysis of the GC-MS data allowed for the characterization of the impacts of package container on the chemical profiles of AA and IPA over time. MVA of the DART-MS data explored the predictive power of the tool for streamlining beer flavor stability analysis. Partial least squares discriminant (PLS-DA) and Multiple Analysis of the Variance (MANOVA) statistical analyses were used to explore data produced by GC-MS and helped define a group of 17 detected metabolites important to explaining the data variation. PLS-DA models of AA samples demonstrated good model fit and package type predictability (R2 = 0.981, Q2 =0.964). This was not observed in IPA which indicates package effects are styles dependent. Differences in AA samples are due, in part, to can and bottle baseline differences in the detected amino acid and ester metabolites. Differences in the physical packaging process of cans, oxidations, and low hop polyphenol concentrations are proposed mechanisms for explaining the observed baseline differences. Analysis of variance (ANOVA) found ten metabolites in AA cans significantly (P ≥ 0.05) changing over time as compared to four metabolites in AA bottles. This indicates higher instability in cans for AA samples. Four detected hop volatiles (humulene, β-myrcene, α-calacorene, pinocarvone), identified by estimated marginal mean of linear models (95% confidence interval) had exhibited significant changes over time that were dependent on package type interactions, but to varying magnitudes and directions depending on the metabolite's polarity and susceptibility to packaging material interactions (e.g., scalping). PLS-DA models of data produced by DART-MS indicated a poor model fit and lack of beer storage time predictability in AA samples (R2 = 0.554, Q2 = -0.151) and IPA samples (R2 = 0.622, Q2 = -0.079). These results lack the evidence that DART-MS is a useful tool for streamlining beer stability analysis. However, results for package type predictability matched GC-MS analysis conclusions in that package type predictability is style dependent. The overall study results demonstrate there is much nuance in the effects of package type on beer flavor stability, and those effects depend on style, packaging material, and the individual metabolite. Targeted analysis is needed to fully understand the mechanisms driving the effects of package type on beer stability.

Description

Rights Access

Subject

Citation

Associated Publications