LAUSR

Abstract The aim of the presented study was to evaluate the potential of a headspace fingerprint as a multivariate intrinsic indicator for monitoring temperature history variation that can occur, for example, during in-pack food processing at the product level. Using solid-phase micro-extraction gas chromatography mass spectrometry (SPME-GC–MS), we monitored the extracted volatile fraction of a series of 8 well-defined thermally processed broccoli purees, differing only in the maximal process temperature reached. Our results showed that the relative composition of the extracted volatile fraction clearly depended on the processing intensity (as a measure for thermal variation) applied. In addition, this headspace fingerprinting approach, including multivariate data analytical approaches, allowed a swift selection of specific intrinsic fingerprint markers. The evaluation of the concentration of each of these markers allows to discriminate between the different processing intensities. However, we suggest to perform a linear combination of the information from relevant identified intrinsic fingerprint markers given the increased reliability of a multiple response indicator. The presented approaches are a promising proof-of-principle that has potential to be exploited for monitoring other processing non-uniformities as well. Industrial relevance Thermal treatment is by far still the most commonly used preservation method for food with a high water content. In-pack retort treatment commonly results in some temperature non-uniformity occurring during the treatment at both the reactor and/or the product level. While the use of thermocouples for acquiring time-temperature history is often an excellent solution, the use of intrinsic food components as indicators can be of great advantage for process monitoring. So far usually pre-selected single component intrinsic indicators were suggested. In this work, we are presenting the utilization of the headspace fingerprint as a multivariate intrinsic indicator instead of using a single compound. From the industrial point of view, we present an innovative concept of monitoring product specific extracted volatile fraction, followed by selection of specific intrinsic fingerprint markers and linearly combining the information from those markers. While the advanced data analysis for selection of the fingerprint markers will have to be done per specific product, only the linear combination of the specific data from the markers can be used for routine analysis. Such concept can provide increased reliability due to the utilization of a non-pre-selected multiple response indicator with a relatively feasible monitoring of the markers using gas chromatography mass spectrometry (GC–MS).