Occurrence of furan in commercial samples of roasted coffee in Brazil Adriana P. Arisseto a, Eduardo Vicente a, Mariana S. Ueno a, Maria Cecília F. Toledo a a Institute of Food Technology, Campinas, Brazil (adriana.arisseto@ital.sp.gov.br) ABSTRACT Furan is a food processing contaminant which occurs in several heat-treated foods, such as canned and jarred foods, coffee and cereal products. Furan is classified as a possible human carcinogen and recent risk evaluations have indicated that the exposure to furan by commonly consumed foods in the diet is a human health concern. Previous studies indicate that roasted coffee contains the highest furan levels in comparison to other products, with mean and maximum values of 1807 and 6900 μg/kg, respectively. So far, no data on the levels of furan in roasted coffee samples from Brazil is available in the literature. Therefore, the objective of this study was to validate a method based on gas chromatography coupled to mass spectrometry preceded by headspace solid phase microextraction (HS-SPME-GC/MS) for furan determination and evaluate the levels of the contaminant in roasted coffees available on the Brazilian market. Results of the validation showed good linearity over the range 0-9600 μg/kg. A comparison between curves set on aqueous standard solutions and on matrix revealed a non-significant matrix effect. Limits of detection and quantitation were 3 and 10 μg/kg, respectively. Recovery, repeatability and within-laboratory reproducibility were in satisfactory ranges. The content of furan in the analyzed samples varied from 250 to 5332 μg/kg. The lowest mean level was found in instant coffee (449 ± 271 µg/kg) whereas the highest mean concentration was observed in strong ground coffee packed under vacuum (4247 ± 1090 µg/kg). It is expected that these results will contribute to data accumulation for worldwide health risk assessment and be helpful in establishing approaches to lower the exposure of the population to furan from the consumption of coffee. Keywords: Furan, coffee, SPME-GC/MS, processing contaminants. INTRODUCTION Furan and its derivatives have long been known as flavor volatiles of foods such as coffee, canned meat, baked bread and cooked chicken [1, 2]. Although furan had previously been identified in some thermally processed foods, a report published in May 2004 by the US Food and Drug Administration (US FDA) showing the presence of furan in commonly consumed foods that undergo heat treatment, such as canned and jarred products including baby-foods, raised for the first time a concern on the potential risks of furan to human health [3]. Furan is considered a possible human carcinogen (group 2B) by the International Agency for Research on Cancer [4]. Furan is clearly carcinogenic to rats and mice, showing a dose-dependent increase in hepatocellular adenomas and carcinomas in both sexes. It has also been demonstrated that furan is cytotoxic to the liver. No data are available on reproductive and developmental toxicity and there are also no human studies. The occurrence of furan in a large variety of foods suggests that there are probably multiple routes for its formation rather than a single mechanism. Experiments using model system have shown that furan is produced by the thermal degradation and rearrangement of precursors such as sugars, amino acids, ascorbic acid and polyunsaturated fatty acids [5, 6]. However, the exact mechanism of furan formation in foods is not completely understood. Due to its high volatility (boiling point 31ºC), gas chromatography coupled to mass spectrometry (GC-MS) has been suggested as a technique to analyze furan content in foods, preceded by headspace sampling (HS) or solid phase microextraction (SPME) [7-9]. Both HS and SPME approaches are very simple and convenient for volatiles analyses, demand no expensive equipment for sample extraction and give satisfactory and comparable results [10]. SPME seems to be more advantageous since allows sample concentration and affords higher sensitivity. Several authors have investigated the furan content in foods and data available in the literature indicate that roasted coffee contains higher furan levels than other products [11-13]. The highest concentration was reported in roasted coffee beans with mean and maximum values of 3611 and 6407 μg/kg, respectively, while
lower levels were found in ground coffee, with an average concentration of 1807 μg/kg [14]. This may be probably due to the roasting process where the high temperatures exceed most other food processing procedures. So far, no data on the level of furan in roasted coffee samples from Brazil is available in the literature. Therefore, the objective of this study was to validate a method based on HS-SPME-GC/MS for furan determination and evaluate the levels of the contaminant in roasted coffees available on the Brazilian market. MATERIALS & METHODS Standards and chemicals. Furan and [ 2 H 4 ] furan (furan-d 4 ) were obtained from Sigma-Aldrich at purity higher than 98%. Methanol was of HPLC-grade and water was purified by reverse osmosis (Gehaka). Individual stock solutions of both standards at ca 2 mg/ml were prepared by dissolving in methanol. Intermediate and work solutions at ca 20 µg/ml and 2 µg/ml, respectively, were prepared in water. Samples. A total of 41 samples were purchased at supermarkets in the city of Campinas, SP, Brazil, including traditional ground coffee of different intensities (n=27), instant (n=8), decaffeinated (n=2) and premium coffee samples (n=4). The samples were stored at 4ºC for at least 4 hours before homogenization in order to avoid losses of furan due to volatilization. All products were analyzed as bought. Determination of furan. Furan content was determined by using a method based on HS-SPME-GC/MS. A portion of 0.25 g of homogeneous sample was weighed in a chilled 40 ml screw-cap glass vial fitted with silicone-ptfe septum containing a 15 mm x 5 mm PTFE-coated stir bar. A volume of 150 µl of furan-d 4 working standard solution 2 µg/ml and 1 ml of water were added and the vial immediately closed. The SPME was carried out in a 75 µm carboxen-polydimethylsiloxane (CAR-PDMS) fiber (Supelco) at 35ºC during 30 min, under a constant magnetic agitation rate of 1200 rpm, approximately. Thermal desorption was carried out into a HP 6890 gas chromatography equipped with a MSD 5973 mass spectrometer (Agilent Technologies). Helium was used as the carrier gas at a flow rate of 0.7 ml/min. The Programmable Temperature Vaporizing (PTV) injector was operated in the splitless mode under the following temperature program: 40ºC (held for 0.1 min), 700ºC/min to 230ºC (held until the end of the run). The separation was performed on a 60 m x 0.25 mm, d f 0.25 µm HP-INNOWAX capillary column (Agilent Technologies) and the oven temperature program was: 30ºC (held for 0.1 min), 2ºC/min to 40ºC (held for 3 min), 12ºC/min to 200ºC (held for 2 min). The mass spectrometer was operated in positive electron impact ionization mode (+EI) with 70 ev of electron energy. Selected ion monitoring (SIM) was used for the detection of furan and furan-d 4, using m/z 68/39/69 for furan and m/z 72/42 for furan-d 4. Identification and quantification. Identification of furan was based on the relative retention time (RRT) and the presence of diagnostic ions. For confirmatory purposes, a comparison with a standard solution was performed using an acceptable deviation of ± 0.5% for RRT, ± 10% for ionic relative abundance considering m/z 39/68, and ± 50% for ionic relative abundance considering m/z 69/68, according to the acceptance criteria as stipulated in European Commission Decision 2002/657 [15]. The quantification of furan in samples proceeded by extrapolation from a linear analytical curve, using furan-d 4 as internal standard. Validation of the method. The method was validated in terms of linearity, selectivity, limit of detection (LOD), limit of quantitation (LOQ), trueness (recovery) and precision (repeatability and within-laboratory reproducibility) according to the guidelines laid down by the Brazilian Institute of Metrology, Standardization and Industrial Quality [16]. Linearity was evaluated over the range 0-9600 µg/kg (six calibration points). Selectivity was evaluated by comparison between curves set on standard solutions and on matrix by applying the F-test (Snedecor) and t-test (Student). LOD and LOQ were determined by seven replicates of the matrix and calculated as 3 and 10-fold standard deviation, respectively. Recovery, repeatability and within-laboratory reproducibility were evaluated by spiking the matrix with furan at 480, 1200 and 3600 µg/kg (seven replicates for each concentration level). As no blank matrix is available for coffee, most of the experiments were carried out with a sample of roasted ground coffee containing 1501 µg/kg. A sample of coffee brew containing 17 µg/kg was used to evaluate LOD and LOQ.
RESULTS & DISCUSSION The first objective of this study was to obtain a reliable and efficient method for the determination of furan in coffee by using SPME. Initial tests were carried out by using the same chromatography conditions established in our previous work with baby-foods [17]. A typical chromatogram of a sample of roasted ground coffee is illustrated in Figure 1, showing that a good separation of furan from co-extractives was achieve under the conditions previously established. Abundance Furan Time (min) Figure 1. Chromatogram of a sample of roasted ground coffee (carrier gas: helium; flow rate: 0.7 ml/min; Programmable Temperature Vaporizing (PTV) injector: 40ºC (held for 0.1 min), 700ºC/min to 230ºC (held until the end of the run); mode: splitless; column: 60 m x 0.25 mm, d f 0.25 µm HP-INNOWAX; oven: 30ºC (held for 0.1 min), 2ºC/min to 40ºC (held for 3 min), 12ºC/min to 200ºC (held for 2 min); mass spectrometer: positive electron impact ionization (70 ev)). The method was in-house validated in terms of linearity, selectivity, LOD, LOQ, trueness (recovery) and precision (repeatability and within-laboratory reproducibility) and the data obtained are shown in Table 1. Good linearity over the range 0-9600 μg/kg was obtained. A comparison between curves set on aqueous standard solutions and on matrix revealed a non-significant matrix effect. LOD and LOQ were 3 and 10 μg/kg, respectively. Recovery, repeatability and within-laboratory reproducibility were in satisfactory ranges. These results demonstrated the applicability of SPME for furan analysis in coffee. Table 1. Validation results. Parameters Linearity 0-9600 µg/kg (r 2 = 0.992) Selectivity No matrix effects (F calc = 1.17 < F tab = 5.05; t calc = 0.90 < t tab = 2.23) LOD (n=7) 3 µg/kg LOQ (n=7) 10 µg/kg Spike level (µg/kg) n % 480 7 101 Recovery (mean) 1200 7 76 3600 7 86 480 14 7.1 Repeatability (CV) 1200 14 3.5 3600 14 1.7 480 14 10.6 Within-laboratory reproducibility (CV) 1200 14 13.8 3600 14 6.2 CV = coefficient of variation.
The levels of furan in the analyzed samples are shown in Table 2. The furan content varied from 250 to 5332 μg/kg. The lowest mean level was found in instant coffee (449 ± 271 µg/kg) whereas the highest mean concentration was observed in strong ground coffee packed under vacuum (4247 ± 1090 µg/kg). These results are comparable to data reported in the literature by other countries [11-14]. There was no correlation between furan levels and the coffee intensity as indicated on the label (classic, strong and extra-strong). However, mean furan levels found in the samples packed under vacuum were higher than those packed in normal atmosphere. Table 2. Furan levels in roasted coffee. Product n Furan (µg/kg) Mean Min-Max Packaging in normal atmosphere Classic 7 1670 1129-2026 Extra-strong 6 1556 1247-1861 Instant 8 449 250-1012 Packaging under vacuum Classic 6 3472 2534-5021 Strong 2 4247 3340-5332 Extra-strong 6 2445 1556-5056 Decaffeinated 2 4082 3274-4778 Premium 4 1789 1273-2494 It has been reported that coffee brew is the most important source of furan in diet and one of the factors affecting the furan content in the beverages is the furan levels in roasted ground coffee [11, 14]. In this way, the results of the present study could be taken into account in order to reduce the daily exposure to furan from coffee brew intake. CONCLUSION The present study showed the applicability of the SPME for furan analysis in coffee with an adequate degree of confidence. The content of furan in Brazilian samples of roasted ground coffee varied from 250 to 5332 μg/kg, which is in accordance with data reported by other countries. It is expected that these results will contribute to data accumulation for worldwide health risk assessment and be helpful in establishing approaches to lower the exposure of the population to furan from the consumption of coffee. REFERENCES [1] Merritt C., Bazinet M.L., Sullivan J.H. & Robertson, D. H. 1963. Mass spectrometric determination of the volatile components from ground coffee. Journal of Agricultural and Food Chemistry, 11, 152-155. [2] Maga J.A. 1979. Furans in foods. Critical Reviews in Food Science and Nutrition, 11, 355-400. [3] US FDA. Exploratory data on furan in food: individual food products; United States Food and Drug Administration. 2004. [4] IARC. Furan. In: IARC Monographs on the evaluation of carcinogenic risks of chemicals to humans; International Agency for Research on Cancer. Lyon, v. 63, 1995. [5] Locas C.P. & Yaylayan V.A. 2004. Origin and mechanistic pathways of formation of the parent furan - a food toxicant. Journal of Agricultural and Food Chemistry, 52, 6830-6836. [6] Becalski A. & Seaman S. 2005. Furan precursors in food: a model study and development of a simple headspace method for determination of furan. Journal of AOAC International, 88, 102-106. [7] Goldman P., Périsset A., Scanlan F. & Stadler R.H. 2005. Rapid determination of furan in heated foodstuffs by isotope dilution solid phase micro-extraction-gas chromatography-mass spectrometry (SPME-GC-MS). Analyst, 130, 878-883. [8] Bianchi F., Careri M., Mangia A. & Musci M. 2006. Development and validation of a solid phase micro-extractiongas chromatography-mass spectrometry method for the determination of furan in baby-food. Journal of Chromatography A, 1102, 268-272.
[9] La Pera L., Liberatore A., Avellone G., Fanara S., Dugo G. & Agozzino P. 2009. Analysis of furan in coffee of different provenance by head-space solid phase microextraction gas chromatography-mass spectrometry: effect of brewing procedures. Food Additive and Contaminants, 26, 786-792. [10] Wenzl T. 2008. Methods for the determination of furan in food. Outcome of a survey conducted among EU food control laboratories. JRC Scientific and Technical Reports, EUR 23269 EN. [11] Zoller O., Sager F. & Reinhard, H. 2007. Furan in food: headspace method and product survey. Food Additive and Contaminants, 24, 91-107. [12] Morehouse K.M., Nyman P.J., Mc Neal T.P., Dinovi M.J. & Perfetti G.A. 2008. Survey of furan in heat processed foods by headspace gas chromatography/mass spectrometry and estimated adult exposure. Food Additives and Contaminants, 25, 259-264. [13] Kim T.-K., Lee Y.-K., Kim S., Park Y.S. & Lee K.G. 2009. Furan in commercially processed foods: four-year field monitoring and risk assessment study in Korea. Journal of Toxicology and Environmental Health, 72, 1304-1310. [14] EFSA. Update of results on the monitoring of furan levels in food; European Food Safety Authority. 2010. [15] EC. Commission Decision 2002/657/EC concerning the performance of analytical methods and the interpretation of results. Official Journal of European Communities, L 221:8-36, 2002. [16] INMETRO. Orientação sobre validação de métodos de ensaios químicos DOQ-CGCRE-08. Instituto Nacional de Metrologia, Normalização e Qualidade Industrial. Revisão 2, Junho 2007. [17] Arisseto A.P., Vicente E. & Toledo M.C.F. 2010. Determination of furan levels in commercial samples of baby-food from Brazil and estimated exposure. Food Additives and Contaminants, 27, 1051-1059.