LC method for the direct and simultaneous determination of four major furan derivatives in coffee grounds and brews.

The simultaneous quantification of two potential genotoxic hydroxymethyl furan derivatives in coffee (furfuryl alcohol and 5-hydroxymethylfurfural) alongside their carboxylic acid derivatives (2-furoic acid and 5-hydroxymethyl furoic acid, respectively) was carried out. Their extraction from ground roasted coffee using sonication, simple shaking or heat-assisted shaking lead to similar results. A minimum of 97.3% of the four furan derivatives were extracted during the first extraction cycle using water, whereas methanol showed considerably lower extraction efficiency. A simple high-performance liquid chromatography method coupled with diode array detection was developed for the simultaneous determination and was applied to roasted coffee extracts or brews. No sample pre-treatment except for centrifugation was needed. The diode array detector was used to assess the purity of the peaks of interest in analyzed samples against authentic standards. The linearity according to Mandel, accuracy (recovery ≥ 89.9%) and precision (inter- and intraday relative standard deviation ≤ 4.5%) were checked. The values for the limit of detection and quantification ranged within 0.11-0.76 and 0.35-2.55 µg/mL, respectively. Filtered and espresso brews were analyzed for the four furan derivatives where furfuryl alcohol showed double the concentration of 5-hydroxymethylfurfural and about ten times the concentrations of 2-furoic acid or 5-hydroxymethyl furoic acid.

Investigation on the mitigation effects of furfuryl alcohol and 5-hydroxymethylfurfural and their carboxylic acid derivatives in coffee and coffee-related model systems.

The mitigation of furfuryl alcohol, 5-hydroxymethylfurfural, 2-furoic acid, and 5-hydroxymethyl 2-furoic acid was conducted in two dry model systems mimicking coffee and an actual coffee system by incorporating 14 chemicals, that are categorized to phenolic acids, flavonoids, non-phenolic antioxidants, and non-antioxidant agents. Mitigation effects were determined as the decrease in the levels of the studied furan derivatives after the systems went through a controlled roasting process. Strong mitigation effects in the dry model systems were observed after the application of phenolic acids, quinic acid or EDTA. The mitigation effects of phenolic acids and flavonoids depended on the number and availability of phenolic hydroxyl groups. Certain agents exhibited a furan derivative-specific reducing effect while most of them showed a generalized effect. The mitigation efficacy decreased with the increasing complexity of the tested systems. In the coffee system, mitigation effects were almost completely lost in comparison with dry model systems. Still, taurine and sodium sulfite exerted the strongest mitigation effect in the coffee system.

Formation kinetics of furfuryl alcohol in a coffee model system.

The production of furfuryl alcohol from green coffee during roasting and the effect of multiple parameters on its formation were studied employing HPLC-DAD. Results show that coffee produces furfuryl alcohol in larger quantities (418µg/g) compared to other beans or seeds (up to 132µg/g) roasted under the same conditions. The kinetics of furfuryl alcohol production resemble those of other process contaminants (e.g., HMF, acrylamide) produced in coffee roasting, with temperature and time of roasting playing significant roles in quantities formed. Different coffee species yielded different amounts of furfuryl alcohol. The data point out that the amounts of furfuryl alcohol found in roasted coffee do not reflect the total amounts produced during roasting because great amounts of furfuryl alcohol (up to 57%) are evaporating and released to the atmosphere during roasting. Finally the effect of the moisture content on furfuryl alcohol formation was found to be of little impact.

Parameters affecting the exposure to furfuryl alcohol from coffee.

Recently, furfuryl alcohol (FFA) was labelled a human potential carcinogen (group 2B) by the International Agency for Research on Cancer. Its alimentary exposure is mostly from coffee since in any other foods the concentrations are significantly lower. The various storage conditions of roasted coffee, the different brewing techniques applied and the bioaccessibility after ingestion are potential parameters that might alter the exposure to FFA from coffee. An 8 weeks stability study at varying temperatures showed that FFA is stable in the ground coffee matrix. Moreover, different brewing techniques extracted different amounts of FFA and affected its final concentration. The evaluation of the relative exposure to four furans (FFA, 5-hydroxymethyl-furaldehyde, 2-furoic acid, and 5-hydroxymethyl-2-furoic acid) revealed that FFA amounts were at least 2-fold the amounts of other studied furans in the same brew. A 22-fold variation in the concentration of the four furans in brews prepared using different coffee grounds and brewing techniques could be observed. 90% of the four furans were extracted by the first 25-30% fraction of the filter brew. A significant decrease of FFA is observed after stressing with simulated gastric fluid. However, this decrease could not be reproduced when mimicking a regular coffee ingestion situation.