Analysis of acrylamide in food products by in-line preconcentration capillary zone electrophoresis

Journal of Chromatography A, 1129 (2006) 129–134

Analysis of acrylamide in food products by in-line preconcentration capillary zone electrophoresis Elisabet Bermudo, Oscar N´un˜ ez, Luis Puignou ∗ , Maria Teresa Galceran Department of Analytical Chemistry, University of Barcelona, Mart´ı i Franqu`es, 1-11, E-08028, Barcelona, Spain Received 3 May 2006; received in revised form 21 June 2006; accepted 23 June 2006 Available online 14 July 2006

Abstract Two in-line preconcentration capillary zone electrophoresis (CZE) methods (field amplified sample injection (FASI) and stacking with sample matrix removal (LVSS)) have been evaluated for the analysis of acrylamide (AA) in foodstuffs. To allow the determination of AA by CZE, it was derivatized using 2-mercaptobenzoic acid. For FASI, the optimum conditions were water at pH ≥ 10 adjusted with NH3 as sample solvent, 35 s hydrodynamic injection (0.5 psi) of a water plug, 35 s of electrokinetic injection (−10 kV) of the sample, and 6 s hydrodynamic injection (0.5 psi) of another water plug to prevent AA removal by EOF. In stacking with sample matrix removal, the reversal time was found to be around 3.3 min. A 40 mM phosphate buffer (pH 8.5) was used as carrier electrolyte for CZE separation in both cases. For both FASI and LVSS methods, linear calibration curves over the range studied (10–1000 ␮g L−1 and 25–1000 ␮g L−1 , respectively), limit of detection (LOD) on standards (1 ␮g L−1 for FASI and 7 ␮g L−1 for LVSS), limit of detection on samples (3 ng g−1 for FASI and 20 ng g−1 for LVSS) and both run-to-run (up to 14% for concentration and 0.8% for time values) and day-to-day precisions (up to 16% and 5% for concentration and time values, respectively) were established. Due to the lower detection limits obtained with the FASI–CZE this method was applied to the analysis of AA in different foodstuffs such as biscuits, cereals, crisp bread, snacks and coffee, and the results were compared with those obtained by LC–MS/MS. © 2006 Elsevier B.V. All rights reserved. Keywords: Acrylamide; Capillary electrophoresis; FASI; Stacking; Foodstuffs

1. Introduction The ancient practice of cooking improves microbial food safety, nutrient accessibility, and also palatability. However, in the last 20–25 years, modern science has shown that the heating of food can generate several potentially hazardous compounds, some of which are genotoxic and carcinogenic. In April 2002, Swedish researchers shocked the food safety world when they presented preliminary findings of acrylamide (AA) in some carbohydrate-rich foods, especially in fried and baked products [1]. These findings attracted considerable interest and wide attention all over the world due to the classification of AA as a “potential carcinogen to humans” [2]. Moreover, the Swedish findings on the high level of acrylamide in heat-treated foods

∗

Corresponding author. Tel.: +34 93 402 19 15; fax: +34 93 402 12 33. E-mail address: [email protected] (L. Puignou).

0021-9673/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.chroma.2006.06.076

were quickly confirmed by a series of government agencies through official website notifications [3]. Regarding the formation of AA in heated food, several of theoretical mechanisms have been proposed but the Maillard reaction between amino acids (mainly asparagine) and reducing sugars such as glucose or fructose is accepted as the most probable formation mechanism [4–7]. For the analysis of acrylamide two main approaches, based either on gas chromatography with mass spectrometric detection (GC–MS) or liquid chromatography coupled to tandem mass spectrometry (LC–MS/MS) have been proposed [8–11]. The main advantage of LC–MS over GC–MS is that acrylamide can be analyzed without prior bromination. However, acrylamide is a very polar molecule [12] with poor retention on conventional LC reverse-phase columns and despite the use of tandem mass spectrometry, this technique presents some limitations mainly due to the coelution of interferent compounds. A number of different stationary-phases such as graphitic carbon, hydrophilic

130

E. Bermudo et al. / J. Chromatogr. A 1129 (2006) 129–134

end-capped C18 and ion-exchanger, have been used to solve the LC separation of this “difficult-to-analyze”compound from other co-extractives. Due to the high polarity of acrylamide, it is difficult to choose an appropiate mobile phase to achieve a reasonable retention time. Then, generally, high content water mobile phases are used which suppose poor ionization efficiencies when ESI is employed as ionization source in LC–MS. In order to solve these problems, Jezussek and Schieberle [13] developed a method based on the derivatization of acrylamide with 2-mercaptobenzoic acid that provides higher selectivity and sensitivity. Due to the relatively high cost of mass spectrometric instrumentation, alternative analytical approaches mainly using UV detection have been proposed. For the determination of acrylamide in foodstuffs, liquid chromatography [14–16] procedures without AA derivatization have been applied. Regarding capillary electrophoresis techniques, microemulsion electrokinetic chromatography (MEEKC) was firstly proposed for the analysis of acrylamide without derivatization [17] but this method provides relatively high detection limits. Better results were obtained for capillary zone electrophoresis (CZE), after derivatization with 2-mercaptobenzoic acid to obtain an ionic compound [18]. Using this procedure, acceptable results can be achieved although detection limits higher than those with LC–MS/MS or GC–MS were obtained which is mainly due to the typical low sample injection volume, some nL, for this technique. To enhance sensitivity in CE, several electrophoreticbased techniques of in-column preconcentration such as isotachophoresis [19,20], field amplified sample injection (FASI) [21,22], stacking [22–24] and sweeping [22,25,26] can be used. These techniques take advantage of differences in mobility and conductivity between samples and buffers to preconcentrate the analytes, and are widely employed because no special devices are required. Among these techniques, FASI is very popular since it is quite simple only requiring the electrokinetical injection of the sample after the introduction of a short plug of a high-resistivity solvent (mainly water). Stacking procedures that allow the focusing of the charged analytes on a sharp sample band during the application of the separation voltage, are also frequently used for in-column preconcentration in CE. Normal stacking requires a sample solvent of lower conductivity than that used for CE separation and in general, to increase the amount of sample introduced stacking with sample matrix removal is currently used. In this work, two in-line CZE enrichment procedures, FASI and stacking with sample matrix removal, also known as large volume sample stacking (LVSS), have been tested to improve acrylamide detection by CZE. Several parameters which can affect the performance of in-line preconcentration, such as buffer concentration and pH, injection time, and reversal time (in stacking with sample matrix removal), among others, were optimized. Limit of detection (LOD) and precision were established. Finally, the best in-line preconcentration method (FASI) was used for the determination of acrylamide by CZE in several foodstuffs. The results were compared with those obtained using a LC–MS/MS method previously established [27].

2. Experimental 2.1. Chemicals and consumables All the reagents used in this work were of analytical grade. Acrylamide (>99%) was provided by Fluka (Buchs SG, Switzerland). Anhydrous sodium dihydrogen phosphate and sodium monohydrogen phosphate were also obtained from Fluka. Methanol, hexane, sodium hydroxide, dichloromethane, ammonia (25%) and hydrochloric acid (25%) were purchased from Merck (Darmstadt, Germany). 2-Mercaptobenzoic acid was used as derivatizing reagent and was purchased from Sigma–Aldrich (Steinheim, Germany). A stock solution of acrylamide (1 mg mL−1 ) was prepared in Milli-Q water and was stored at 4 ◦ C for a maximum of 4 weeks. Solid phase extraction (SPE) cartridges Strata-X-C (200 mg, 6 mL) and Isolute ENV+ (200 mg, 3 mL) were provided by Phenomenex (Torrance, USA) and IST (Hengoed, Mid-Glarmorgan, UK), respectively. Syringe filters 0.45 ␮m of both nylon and nitrocellulose were obtained from Teknokroma (Barcelona, Spain). Water was purified with an Elix/Milli-Q water purification system (Millipore, Bedford, MA, USA). 2.2. Instrumentation Experiments were performed on a Beckman P/ACE MDQ capillary electrophoresis instrument (Fullerton, CA, USA) equipped with a diode array detection system. Electrophoretic separations were carried out using uncoated fused-silica capillaries (Beckman) with a total length of 60 cm (50 cm effective length) and 75 ␮m I.D. A 40 mM phosphate buffer (pH 8.5) was used as carrier electrolyte. The buffer was filtered through a 0.45 ␮m membrane filter, and degassed by sonication before use. New capillaries were pre-treated by rinsing with 1 M sodium hydroxide for 30 min and water for 30 min, before being conditioned with the carrier electrolyte. At the beginning of each session, the capillary was rinsed with 1 M sodium hydroxide for 15 min, with water for 15 min, and finally with the carrier electrolyte for 30 min. Between runs, the capillary was conditioned for 5 min with carrier electrolyte. Field amplified sample injection was performed as follows. The capillary was first filled with the carrier electrolyte and then a water plug (35 s, 3.5 kPa) was introduced. Samples were then introduced into the capillary by electrokinetic injection at −10 kV during 35 s. After sample injection, a small water plug (6 s, 3.5 kPa) was introduced to prevent AA removal by EOF. The electrophoretic separation was then performed by applying +25 kV through the capillary. Direct detection was carried out at 210 nm. Stacking with sample matrix removal involved several steps. The capillary was first filled with the carrier electrolyte and then a long plug of sample was introduced hydrodynamically by pressure (140 kPa) for 20 s. Under these conditions, close to 100% of capillary length was filled with the sample. A high voltage (−25 kV) was then applied and the electric current was monitored to indicate when the sample matrix was almost removed from the capillary. When the current was 95% of the original car-

E. Bermudo et al. / J. Chromatogr. A 1129 (2006) 129–134

rier electrolyte current value, the voltage was turned off and the electrodes were switched to the separation configuration. Electrophoretic separation was then carried out by applying +25 kV. Detection was performed at 210 nm.

131

water at pH ≥ 10 (adjusted with ammonia), filtered through a 0.2 ␮m nylon filter and injected into the CE system using FASI or LVSS procedures. 3. Results and discussion

2.3. Sample preparation and clean-up procedure 3.1. Acrylamide derivatization A previously published purification method [27] was used to extract acrylamide from the different analyzed samples. Briefly, prior to the extraction and purification, the food samples were ground and homogenized using a supermixer blender system (Moulinex, Lyon, France) and an Ultraturrax T25 basic (IKA® WERKE GMBH&CO.KG, Staufen, Germany), respectively. Sub-samples of 2 g (1 g for some samples such as potato crisps) were weighed into 15 mL centrifuge tubes and 10 mL of water was added. Each tube was shaken for 1 h on a rotating shaker (Rotatory Mixer 34526; Breda Scientific, Breda, The Netherlands) and centrifuged at 4000 rpm for 30 min in a Selecta Centronic centrifuge (Selecta, Barcelona, Spain). For potato crisps samples, centrifugation at 20,000 rpm was necessary and then an Allegra 64R centrifuge (Beckman, Fullerton, CA, USA) was used. An aliquot (3 mL) of the supernatant was filtered through a syringe filter of nitrocellulose (0.45 ␮m) and de-fatting was performed by adding 2 mL of hexane. For clean-up, both StrataX-C and ENV+ SPE cartridges were conditioned with 4 mL of methanol followed by 4 mL of water using a Supelco Visiprep® vacuum manifold (Supelco, Gland, Switzerland). Each StrataX-C was loaded with 3 mL of filtered extract. The extract was passed through the sorbent material. Then, the column was eluted with 3 mL of water and the eluant was collected for ENV+ SPE clean-up. All the eluant collected from the Strata-X-C was loaded, and eluted with 1 mL of MeOH:H2 O (60:40). The eluent was filtered through a 0.45 ␮m nylon filter and transferred into a glass vial for derivatization. 2.4. Acrylamide derivatization For derivatization, the procedure described in a previous paper [18] was used. Briefly, the pH of the extract from the ENV+ cartridge (1 mL) was adjusted to 8.0 (5 ␮L of 0.01 M sodium hydroxide) and 100 ␮L of 2-mercaptobenzoic acid in aqueous sodium hydroxide (154 mg in 10 mL of 1 M NaOH) was added for derivatization. The solution was degassed with nitrogen, and the reaction mixture was stirred in the dark for 3 h. The excess of derivatizing reagent was removed by precipitation with lead(II) acetate (saturated solution in water), as suggested Jezussek and Schieberle [13], and after centrifugation at 4000 rpm for 3 min, the supernatant was transferred to another vial and acidified to pH < 1.5 by adding 1 mL 25% hydrochloric acid. The solution was centrifuged at 4000 rpm for 3 min in order to remove the excess of lead(II) as lead(II) chloride. The supernatant was extracted with dichloromethane (4 × 10 mL). The organic solvent was then evaporated to 1 mL with nitrogen using a TurboVap® II Concentration Workstation (Zymark Corporation, Hopkinton, Massachusetts, USA), and finally evaporated to dryness using a Visidry vacuum manifold (Supelco, Gland, Switzerland). The residue was finally reconstituted in 1 mL of

In the previously describing CZE method for the analysis of AA [18], the excess of derivatizing reagent (2-mercaptobenzoic acid) was not eliminated, and the solution was directly injected into the CE system. However, the high salinity of the extracts prevented the use of both FASI and LVSS preconcentration methods with an acceptable performance. So, in this work, liquid–liquid extraction was performed after AA derivatization as recommended by Jezussek and Schieberle [13], although dichloromethane was used instead of ethyl acetate. This procedure, liquid–liquid extraction, evaporation to dryness of the organic solvent and reconstitution of the extract with 1 mL of water at pH ≥ 10, allowed to reduce the salinity of the final extract and, as a consequence, the application of FASI and LVSS was possible. 3.2. FASI The derivatized AA, a stable thioether with a carboxylic acid group, was introduced into the capillary using reversed polarity. Although the conductivity of the solvent (water at pH ≥ 10) was lower than that of the carrier electrolyte (40 mM phosphate buffer pH 8.0), no significant increase in the response of acrylamide compared with hydrodynamic injection occurred. This fact can be explained because the electroosmotic flow probably prevents the entrance of the analyte which have a low electrophoretic mobility. In order to prevent this effect and improve the FASI procedure by stacking, a high-resistivity solvent plug was introduced in front of the sample. Two solvents, water and methanol were studied. Although a significant increase in the response of derivatized acrylamide was observed for both solvents, methanol caused the electrophoretic voltage to fail frequently, probably due to the formation of bubbles in the capillary. For this reason, a plug of water was selected. On the other hand, in order to assure the total deprotonation of the thioether acrylamide derivative, different compounds such as sodium hydroxide, ammonia and dimethylamine were used to adjust sample solvent at pH 10. Although the effectiveness of the preconcentration was higher when dimethylamine was employed, probably due to its lower conductivity, ammonia was selected as the best option because of the poor stability of the AA derivative standards and samples on dimethylamine (the signal decreased two times after 24 h). Moreover, an additional plug of water at the end of the injection plug was introduced to prevent the elimination of the analyte by the electroosmotic flow. Injection times for both the plug of water (hydrodynamic mode) and the sample (electrokinetic mode) were simultaneously optimized. Hydrodynamic injection (0.5 psi) of a water plug from 5 s to 40 s, and electrokinetic sample injection (−10 kV) from 5 s to 40 s were tested. Moreover, a fixed water

132

E. Bermudo et al. / J. Chromatogr. A 1129 (2006) 129–134

plug (6 s, hydrodinamic mode) at the end of electrokinetic injection was used in all the experiments. The best results were obtained with an injection time of 35 s for both the water plug and the sample. Obviously, when increasing injection time an enhancement of the response was observed; however, peak broadening occurred at sample injection times higher than 35 s. On the other hand, a reduction of the water plug produced a significant decrease in the signal. Although in a previous work [18], the pH and concentration of the buffer were optimized for the CZE-UV method, in this paper, these values were reoptimized in order to study the effect of the introduction of additional steps in acrylamide derivatization. For this purpose, were obtained the electropherograms at pH 7, 8 and 8.5. The lower migration time was achieved at pH 8.5 and moreover no interferences were observed at acrylamide migration time at this pH. In summary, the proposed preconcentration method based on the FASI technique is: sample solvent water pH≥10 (adjusted with ammonia), 35 s hydrodynamic injection (0.5 psi) of water, 35 s of electrokinetic injection (−10 kV) of the sample and 6 s hydrodynamic injection (0.5 psi) of water. 3.3. Stacking with sample matrix removal In order to improve the AA signal and reduce the analysis time, the concentration and pH of the phosphate buffer were studied using a standard acrylamide solution of 1 mg L−1 . Values comprised between 20 mM and 60 mM were tested, and the electropherograms obtained are given in Fig. 1. As can be seen, the AA derivative signal improved with buffer concentration, although an increase in the analysis time, especially in the 20 mM and 40 mM concentrations, was also observed. Since a

considerable rise in intensity was observed at 60 mM (200 ␮A) and, as a consequence most probably electrophoretic voltage failures can occur, 40 mM was chosen as the best value. On the other hand, different pHs ranging from 6.5 and 8.5 were studied but no variation in the AA relative signal was observed, although a slight decrease in migration time was achieved at the highest pH. So a 40 mM phosphate buffer (pH 8.5) was selected as carrier electrolyte for LVSS preconcentration. In order to achieve an efficient preconcentration when using large volume sample stacking the field polarity reversal time was established. First, the injection time needed to fill the capillary with the sample was determined. Different hydrodynamic injection times (140 kPa) comprised between 5 s and 25 s were studied. An important increase in the signal (>500 times) was obtained when injection times rose to 20 s and any improvement was observed at higher injection times. Under these conditions 3.3 min were needed to push out the matrix of the capillary. 3.4. Quality parameters Instrumental quality parameters using both FASI and LVSS procedures under optimal conditions were calculated and the values are given in Table 1. Calibration curves based on the peak area at concentrations between 10 ␮g L−1 and 1000 ␮g L−1 for FASI and 25 ␮g L−1 and 1000 ␮g L−1 for LVSS were established and good linearity was obtained with a correlation coef-

Table 1 Instrumental quality parameters Parameter

CZE [18]

LODs Concentration standard (␮g L−1 ) Concentration samplea (ng g−1 )

70 200

1 3

7 20

LOQsb Concentration standard (␮g L−1 ) Concentration samplea (ng g−1 )

210 600

3 10

20 60

Run-to-run precision (%RSD, n = 5) Concentration Low level (3× LOD) Medium level Migration time Low level (3× LOD) Medium level

6 2c 0.8 0.3c

Day-to-day presision (%RSD, n = 3 × 5) Concentration Low level (3× LOD) 11.2 Medium level 7c Migration time Low level (3× LOD) Medium level Working range (mg L−1 ) Correlation coefficient, r Fig. 1. Effect of buffer concentration on the separation of AA by LVSS. Electrolyte: monohydrogen phosphate-dihydrogen phosphate buffer. Acquisition wavelength, 210 nm. Sample: acrylamide standard of 1 mg L−1 derivatized with 2-mercaptobenzoic. Hydrodynamic injection, 20 s (140 kPa). Applied potential, −25 kV (sample matrix removal), +25 kV (electrophoretic separation).

a b c d e

4 3c 0.3–100 0.999

FASI

14 4d 0.6 0.2d

16 10d 4 2d 0.01–1 0.999

Crisp bread sample. LOQs, limits of quantification (signal-to-noise ratio 10:1). 1000 ␮g L−1 . 200 ␮g L−1 . 400 ␮g L−1 .

LVSS

8 2e 0.8 0.3e

12 7e 5 2e 0.025–1 0.999

E. Bermudo et al. / J. Chromatogr. A 1129 (2006) 129–134

ficient higher than 0.999 in both cases. The limit of detection of AA, based on a signal-to-noise ratio of 3:1, was 1 ␮g L−1 for FASI and 7 ␮g L−1 for LVSS. These values were 70 and 10 times lower than that obtained without in-line preconcentration [18]. So, both preconcentration methods allowed analyzing AA by CZE at low ppb levels. Detection limits in samples were estimated using a crisp bread sample with a low content of acrylamide. Values obtained were 3 ng g−1 for FASI and 20 ng g−1 for LVSS. FASI values were slightly better than those obtained by LC–UV [14–16]. Moreover, these values are similar to that reported for AA using LC–MS/MS with electrospray (ESI) as ionization source and a triple quadrupole [9] and lower than that obtained with an ion trap mass analyzer [27]. Run-to-run and day-to-day precisions for AA quantification were calculated at two concentration levels, low level (3× LOD) and medium level (200 ␮g L−1 for FASI and 400 ␮g L−1 for LVSS) for both preconcentration procedures. For run-to-run precision, five replicate determinations for each concentration level were carried out, using both preconcentration methods under optimum conditions. On the other hand, day-to-day precision was calculated by performing 15 replicate determinations of each concentration level for 3 non-consecutive days (five replicates each day). The relative standard deviations (RSDs) obtained for run-to-run and day-to-day precisions with FASI at medium concentration level (4% and 10%, respectively) were only slightly higher than those previously obtained by CZE without in-line preconcentration. However, when quantification was performed at the low concentration level (3× LOD level), RSD values increased (14% and 16%, respectively) which can be explained because of the poor reproducibility of electrokinetic injection [28] and the low concentration level quantified (70 times lower than for CZE method). For LVSS, the RSDs obtained for run-to-run and dayto-day precisions at medium concentration level (2% and 7%, respectively) were similar to those obtained by FASI. However, at low concentration level, better reproducibilities were obtained (8% and 12%, respectively). Finally, in terms of migration times good run-to-run and day-to-day precisions were obtained for both in-line preconcentration methods (Table 1). These values were similar to those obtained using CZE. 3.5. Application As previously mentioned, FASI provided better detection limits than LVSS for the analysis of AA. Moreover, LVSS requires checking the reversal time for each standard and sample to control differences in sample matrix. As a consequence, analysis time considerably increase because two runs are needed, one for the reversal time determination and another one for AA analysis preventing automation of the method. For these reasons, FASI–CZE methodology is proposed for the analysis of food samples. In order to evaluate the applicability of FASI–CZE for the determination of AA in different kinds of food products, eight foodstuffs (including biscuits, crisp bread, cereals, snacks, potatoes and coffee) were analyzed following the procedure described in Section 2. As an example, Fig. 2 shows the elec-

133

Fig. 2. Electropherograms of: (1) crisp bread Ortiz; (2) biscuits Flora; and (3) snacks Doritos obtained by CZE–FASI–UV after acrylamide derivatization with 2-mercaptobenzoic acid. Capillary voltage: +25 kV. Acquisition wavelength, 210 nm. Water plug loaded by hydrodynamic injection (35 s, 3.5 kPa), sample loaded by electrokinetic injection (35 s, −10 kV). Carrier electrolyte: 40 mM monohydrogen phosphate-dihydrogen phosphate buffer (pH 8.5).

tropherograms of three representative food samples (biscuits, breakfast cereals and snacks). In a previously established CZE method without preconcentration [18] we demonstrated that both quantification methods standard addition and external calibration provided similar results for the analysis of AA in foodstuffs. In order to evaluate whether the FASI preconcentration has any effect on the results obtained depending on the quantification method used, two food products (biscuits Flora and snacks Doritos) were analyzed using both standard addition and external calibration. For standard addition, three un-spiked and four spiked samples at different concentration levels (100%, 200%, 300% and 500%) were prepared by addition of accurately measured amounts of an AA standard at the beginning of the clean-up process. The results obtained using both quantification methods were 193 ± 23 ␮g Kg−1 and 175 ± 20 ␮g Kg−1 (biscuits Flora) for standard addition and external calibration, respectively, and 215 ± 21 ␮g Kg−1 and 226 ± 18 ␮g Kg−1 (snacks Doritos). Since the results achieved were not significantly different, external calibration can be proposed as a fast and suitable method to quantify AA by FASI–CZE. The AA levels found in eight Spanish food products when performing the analysis with the proposed FASI–CZE methodology are given in Table 2. In order to validate the method, the samples were also analyzed using LC–MS/MS [27] and the results are also included in the table. A statistical paired-sample comparison analysis was performed with the results obtained with both FASI–CZE and LC–MS/MS. The significance values (P) obtained comparing

134

E. Bermudo et al. / J. Chromatogr. A 1129 (2006) 129–134

Table 2 Determination of acrylamide in different Spanish food products by FASI–CZE Food product

Biscuits Flora Biscuits Yayitas Crisp bread Ecological Crisp bread Ortiz Breakfast cereals Cheerios Potato crisps “al jamon” Snacks Doritos Coffee Soley a

FASI–CZE

Significance level (P-value)a

LC–MS/MS [27]

AA (␮g Kg−1 )

RSD (%)

AA (␮g Kg−1 )

RSD (%)

175 403 180 58 266 325 226 208

14 6 10 8 10 8 12 13

181 387 202 59 286 328 205 205

2 5 8 1 5 7 2 2

0.549 0.123 0.084 0.894 0.094 0.687 0.062 0.644

Significant differences between procedures for P < 0.05 (at the 95% confidence level).

the two procedures are given in the Table 2 for each sample. For a 95% confidence level the results achieved were not significantly different, showing that the FASI–CZE–UV procedure developed in this work is an economic and reliable method for the analysis of AA in foodstuffs. 4. Conclusions Two in-line preconcentration methods (FASI and LVSS) to enhance sensitivity in the determination of acrylamide by CZE after derivatization with 2-mercaptobenzoic acid were evaluated. Limits of detection of 1 ␮g L−1 and 7 ␮g L−1 was obtained for standards with FASI and LVSS preconcentration methods, respectively. This means a 70 and 10-fold enhancement on the response when compared with the CZE method without preconcentration. For samples, detection limits were 3 ng g−1 and 20 ng g−1 for FASI and LVSS, respectively. These values were slightly better than those presented in the literature by LC–UV and similar to that obtained by LC–MS/MS using triple quadrupole instruments. Good linearity (r2 > 0.999) and run-torun and day-to-day precisions (RSDs in the range 2–16%) were achieved. The addition of in-line preconcentration did not negatively affect the precision of the CZE method at medium concentration levels, although a slight increase in RSD values was observed with FASI at low concentration levels. Due to the lower detection limits, the FASI–CZE method was used for the analysis of AA in different commercial food products. The results were compared to those obtained by LC–MS/MS and similar results were achieved with both methods. So, FASI–CZE–UV can be proposed as an alternative to other more expensive methods for the analysis of AA in foods. Acknowledgements This work has been carried out with support from the European Commission, Priority 5 on Food Quality and Safety (contract no. FOOD-CT-2003-506820 Specific Targeted Project), “Heat-generated food toxicants-identification, characterization and risk minimization”. This publication reflects the author’s views and not necessarily those of the EC. The information in this document is provided as is and no guarantee or warranty is

given that the information is fit for any particular purpose. Users thereof use the information at their sole risk and liability. References [1] Swedish National Food Administration, Information about Acrylamide in Food, 24 April 2002. http://www.slv.se. [2] International Agency for Research on Cancer (IARC), Some Industrial Chemicals, IARC Monographs on the Evaluation for Carcinogenic Risk for Chemicals to Humans, vol 60, IARC, lyon, 1994. p. 435. [3] Press Briefing of 17 May 2002. RSA Study of Acrylamide in Food, Background Information and Research Findings, Food Standards Agency, UK, http://www.food.gov.uk/wws/wwsarchive/65268. [4] A. Becalski, B.P.Y. Lau, D. Lewis, S.W. Seaman, J. Agric. Food Chem. 51 (2003) 802. [5] D.S. Mottram, B.L. Wedzicha, A.T. Dodson, Nature 419 (2002) 448. [6] R.H. Stadler, I. Blank, N. Varga, F. Robert, I. Han, P.A. Guy, M.C. Robert, S. Riediker, Nature 419 (2002) 449. [7] V.A. Yaylayan, R.H. Stadler, J. AOAC Int. 88 (2005) 262. [8] T. Wenzl, M.B. de la Calle, E. Anklam, Food Add. Contam. 20 (2003) 885. [9] L. Castle, S. Eriksson, J. AOAC Int. 88 (2005) 274. [10] R. Weisshaar, Eur. J. Lipid Sci. Technol. 106 (2004) 786. [11] Y. Zhang, G. Zhang, Y. Zhang, J. Chromatogr. A 1075 (2005) 1. [12] M. Friedman, J. Agric. Food Chem. 51 (2003) 4504. [13] M. Jezussek, P. Schieberle, J. Agric. Food Chem. 51 (2003) 7866. [14] H. Terada, Y. Tamura, J. Food Hyg. Soc. Jpn. 44 (2003) 303. [15] E.K. Paleologos, M.G. Kontominas, J. Chromatogr. A 1077 (2005) 128. [16] V. Gokmen, H.Z. Senyuva, J. Acar, K. Sarioglu, J. Chromatogr. A 1088 (2005) 193. [17] E. Bermudo, V. Ruiz-Calero, L. Puignou, M.T. Galceran, Electrophoresis 25 (2004) 3257. [18] E. Bermudo, O. N´un˜ ez, L. Puignou, M.T. Galceran, J. Chromatogr. A 1120 (2006) 199. [19] T.W. Chine, T.C. Wei, W.H. Chin, C.P. Teng, M.P. Tzu, T.W. Ren, Jpn. J. Toxicol. Environ. Health 40 (1994) 197. [20] P. Gebauer, P. Bocek, Electrophoresis 23 (2002) 3858. [21] S. Sentellas, E. Moyano, L. Puignou, M.T. Galceran, J. Chromatogr. A 1032 (2004) 193. [22] D.M. Osbourn, D.J. Weiss, C.E. Lunte, Electrophoresis 21 (2000) 2768. [23] W. Poinsot, M. Lacroix, D. Maury, G. Chataigne, B. Feurer, F. Couderc, Electrophoresis 27 (2006) 176. [24] A. Juan-Garcia, G. Font, Y. Pico, J. Sep. Sci. 28 (2005) 793. [25] J.-B. Kim, S. Terabe, J. Pharmaceut. Biomed. Anal. 30 (2003) 1625. [26] J.P. Quirino, J.-B. Kim, S. Terabe, J. Chromatogr. A 965 (2002) 357. [27] E. Bermudo, E. Moyano, L. Puignou, M.T. Galceran, Anal. Chim. Acta 559 (2006) 207. [28] J.P. Schaeper, M.J. Sepaniak, Electrophoresis 21 (2000) 1421.