Analysis of Furan and Alkylfurans in Food Samples (Part 2): Optimizing Experimental Conditions for the Analysis of Food Matrices using the SPME Arrow

In our first blog of this series, we demonstrated the improved sensitivity offered by the SPME Arrow for the analysis of furan and alkylfurans. We also presented the GC-MS conditions that enabled us to have good separation among critical analytes of interest with run times under 15 min. In this blog, we want to discuss the importance of various experimental conditions in obtaining a robust SPME workflow.

Coating Type

First, let’s talk about the coating type. Depending on the analytes of interest, various SPME coating chemistries can be selected. Polydimethylsiloxane (PDMS) is perhaps the most known SPME coating as it was the first extraction phase used when SPME was originally introduced. Although PDMS has been broadly used for the analysis of several volatile analytes in various applications, it has a limited sensitivity in comparison to other coating chemistries as it extracts via absorption. For the analysis of highly volatile substances, such as furan and alkylfurans, coating materials with porous sorbents that enable adsorption processes are better suited. It is well-known that the type of pores available in carbon-based coatings offer improved responses when analyzing highly volatile analytes, such as furan and alkylfurans [1]. However, it has been also reported that the high affinity of carbon-based coatings towards furan and alkylfurans could lead to carryover issues [2]. As shown in Figure 1, the wide range carbon coating provided the best results among the evaluated coatings. An assessment of the carryover of the wide range carbon SPME Arrow after desorbing it at 280 °C for 1 min showed that analyte responses were negligible, and therefore effective desorption was attained at the selected conditions. Based on these results, the wide range carbon SPME Arrow was chosen for the analysis of furan and alkylfurans.

Figure 1. Comparison of analyte responses obtained with three different coating chemistries (n=3) after sampling from the headspace of 20 mL vials containing 10 mL of sodium chloride solution (30%) spiked with 20 ng of each target analyte. SPME Arrow conditions: 2 min extraction time, 5 min incubation at 40 °C, 1 min desorption at 280 °C, agitation at 250 rpm.

Extraction Temperature

Once the SPME coating was selected, the effect of three different extraction temperatures (40, 50 and 60 °C) on the response of the six target analytes was evaluated. It is worth emphasizing that to attain satisfactory SPME method precision and accuracy, it is critical to ensure that the extraction temperature is properly controlled.  For this assessment, baby formula blank samples were spiked with the target analytes. Briefly, samples were prepared by weighing 0.5 g of baby formula in 20 mL glass vials and then adding 10 mL of 30% sodium chloride solution [2]; extractions were conducted in an automated fashion from the headspace of the vials using a PAL CTC autosampler. Results demonstrated a decrease in the response of the most volatile analytes (furan, 2- and 3-methylfuran) with an increase of the extraction temperature. 2-ethylfuran and 2,5-dimethylfuran did not show a significant difference in the responses obtained at 40 and 50 °C, whereas 2-pentylfuran exhibited a slight increase in the response with an increase in the extraction temperature (Figure 2).  Based on this data, we considered 50 °C to be a good compromise for all the analytes. Finally, three sample incubation times were tested (5, 10, and 15 min). No significant differences in analyte responses were observed among the tested incubation times; considering that the PAL CTC autosampler allows for SPME sample prep to occur while the GC-MS is running, 10 min was selected as the sample incubation time.

Figure 2. Evaluation of the effect of the extraction temperature on the responses of furan and alkylfurans spiked in baby formula (n=3). SPME conditions: 10 min extraction time, 10 min incubation, 1 min desorption at 280 °C, agitation at 250 rpm.

Extraction Time

Selecting the right extraction time when developing a SPME method for several analytes is crucial to attain the desired method precision and sensitivity. It is worth emphasizing that SPME is a microextraction technique where an amount of analyte proportional to the analyte’s concentration in the sampling media is extracted. In other words, with SPME we are not aiming to have exhaustive extraction recoveries of the analytes of interest, but the small amount of analyte extracted representative of its concentration should be consistent and reproducible. If the SPME fiber is exposed to the sample matrix for a long enough time, an extraction plateau is reached and, we can say that we are extracting at equilibrium conditions. At equilibrium conditions, the best sensitivity and precision are attained; however, reaching equilibrium for all analytes of interest is not always feasible as some compounds may display very long equilibration times (the higher the affinity for the SPME coating, the longer the time needed to reach equilibrium).

Figure 3. Extraction time profiles corresponding to all target analytes spiked in baby formula (0.5 g of sample mixed with 10 mL of sodium chloride solution in a 20 mL vial) at a concentration of 40 µg/kg (n=3). SPME Arrow conditions: 10 min incubation time, 50 °C extraction temperature, 1 min desorption at 280 °C, agitation at 250 rpm.

To evaluate the effect of the extraction time on the furan and alkylfurans responses, an extraction time profile was constructed for each of the six analytes at the already optimized conditions. Extractions were carried out from the headspace of 20 mL vials containing baby formula and 10 mL of sodium chloride solution spiked with all the analytes at 40 µg/kg. Extractions were conducted for 1, 2, 5, 10 and 20 min (n=3). As can be seen in Figure 3, the most volatile analytes (furan, 2- and 3-methyl furan) reached equilibrium at 10 min, whereas the rest of the compounds did not reach a plateau at the evaluated extractions times. Considering these results, and as a compromise between method sensitivity and throughput, 10 min was chosen as the extraction time.

To summarize, we selected a wide range carbon coating for the analysis of furan and alkylfurans in food samples. By using baby formula as a model matrix, we selected a desorption temperature of 280 °C, a desorption time of 1 min, an extraction temperature of 50 °C, an incubation time of 10 min, and an extraction time of 10 min. In our next blog post, we will discuss how to construct calibration curves for the analysis of these food contaminants, so please stay tuned!

References

1. Pawliszyn, J., Handbook of Solid Phase Microextraction. Chemical Industry Press, Beijing 2009.
2. Frank, N., Dubois, M., Huertas Pérez, J. F., Detection of Furan and five Alkylfurans, including 2-Pentylfuran, in various Food Matrices. J. Chromatogr. A 2020, 1622, 461119.