Determination of Furan in Food by Gas Chromatography-Mass Spectrometry and Headspace Sampling

Padmaja Prabhu, PerkinElmer Inc., Waltham, Massachusetts, USA.

Furan is naturally occurring at low levels in many foods and drinks.¹ Based on studies in laboratory animals, furan consumption has been classified by IARC as possibly carcinogenic to humans. The primary source of furan in food is considered to be thermal degradation of carbohydrates, such as glucose, lactose and fructose.

This application note demonstrates a rapid method for the identification and quantification of furan in food samples using gas chromatography (GC), headspace sampling (HS) and mass spectrometry (MS). In addition to method optimization and standard analysis we have analysed a number of food samples for furan.

The PerkinElmer Clarus 680 Gas Chromatograph, Clarus 600 C Mass Spectrometer and a TurboMatrix HS-40 system was used for this application. The separation was achieved on an Elite-624 column ramped from 40 °C to 250 °C over 20 min. The mass spectrometer was scanned from m/z 35–150 with 5 scans per s.

Headspace was chosen as the sample introduction technique because furan can easily be partitioned from liquid matrices and headspace facilitates automated extraction of furan without the instrument contacting the sample matrix. Caution must be taken when setting the vial oven temperature; a high temperature can result in furan formation in the sample during analysis. To reduce this risk, the method presented here uses a temperature of 60 °C.

The GC–MS was calibrated across the range of 1.0 to 40 ng/mL, each calibration point was run in triplicate to demonstrate the precision of the system, the average %RSD across the range was 4.3%. The coefficient of determination for a line of linear regression across the range of 1.0 to 40 ng/mL was 0.9997.

Sample Preparation

Samples were collected from the local food market. These samples included coffee, milk, canned foods, sauces, peanut butter and apple juice. 10 mL of sample was transferred into a headspace vial; 4 g of NaCl was added to it. Milk and other viscous samples were diluted with water (1:2 or 1:4). The semi-solid samples were ground and 5 g of sample was added to the headspace vials with 5 mL of saturated salt (NaCl) solution. Coffee powder was dissolved following directions on the package and then treated like a nonviscous liquid sample.

Figure 1: Chromatogram furan peak in coffee sample.

The recovery of the method was tested with the analysis of the brewed coffee sample spiked at three different levels: 2, 5 and 10 µg/L. The measured amount was 2.03, 5.44 and 9.54 µg/L demonstrating the technique to be quantitative.

Results

Eight samples of common beverages were analysed using the HS-GC–MS method developed here. Brewed coffee was demonstrated to have the highest levels of furan at 250 µg/L.

Conclusion

This application provides a method for the determination of furan in beverages using headspace sample introduction. Headspace-GC is fast, reliable and can be used for the quantification of furans in common beverages. The internal standard calibration of furan across 1–40 µg/L responded linearly. Beverages were analysed and the level of furan was determined. Furan was identified by both the retention time and the MS fragmentation pattern. This method was validated at several levels on coffee matrix recovery, with values between 95–101%.

PerkinElmer Inc.

940 Winter Street, Waltham, Massachusetts 02451, USA

tel. +1 800 762 4000 fax +1 203 944 4904

Website: www.perkinelmer.com