Determination of acrylamide in foodstuff

Determination of acrylamide in foodstuff, Determination of acrylamide in foodstuff, Determination of acrylamide in foodstuff, Determination of acrylamide in foodstuff

Analytica Chimica Acta 559 (2006) 207–214 Determination of acrylamide in foodstuffs by liquid chromatography ion-trap tandem mass-spectrometry using an improved clean-up procedure E Bermudo, E Moyano, L Puignou ∗ , M.T Galceran Departamento de Qu´ımica Anal´ıtica, Universitat de Barcelona, Mart´ı i Franqu`es, 1-11, 08028 Barcelona, Spain Received 13 September 2005; received in revised form 30 November 2005; accepted December 2005 Available online 17 January 2006 Abstract The present paper describes an analytical method based on liquid chromatography coupled to tandem mass-spectrometry (LC–MS/MS) for the determination of acrylamide in foodstuffs Atmospheric pressure chemical ionization (APCI) as ionization source and an ion-trap (LCQ) analyzer were used, and to the best of our knowledge, this hyphenated technique has not ever been employed to this purpose In order to obtain clean extracts an improved purification procedure based on the coupling of two highly cross-linked polystyrene–divinylbenzene polymeric sorbents: Strata-X-C and ENV+, was also developed High recoveries (85%) and good reproducibility (relative standard deviation of 12%) were obtained using the two solid-phase extraction cartridges in combination One hundred percent water as mobile phase was used for the LC separation The obtained figures of merit showed detection limits of 250 pg for standards and 45 ng/g for samples, and run-to-run and day-to-day precisions of 3.3 and 8%, respectively Acrylamide (AA) was determined in several of the most frequently eaten carbohydrate-rich foodstuffs commercialized in Spain such as potato crisps and chips, biscuits, crisp breads, pastry, dried fruits, chocolates and coffee For the first time, a typical pastry product called as “churros” and highly consumed in Spain was also studied Some of the products tested such as french fries frozen or “churros” were household cooked Acrylamide was determined in the selected food commodities using the developed analytical methodology A commonly used method by liquid chromatography coupled to tandem mass-spectrometry with a triple quadrupole as mass analyzer (LC–QqQ–MS/MS) was also applied in order to validate the analytical results Different levels of acrylamide were obtained and pastry and dried fruits showed the lower levels (99%) was purchased from Fluka (Buchs SG, Switzerland) 2,3,3-D3 -acrylamide (98%) was obtained from Cambridge Isotope Laboratories (Andover, MA, USA) Methanol was analytical grade and provided by Merck (Darmstadt, Germany) Stock solutions of acrylamide (1 mg/mL) and [D3 ]-acrylamide (1 mg/mL) were each prepared by dissolving the compounds in Milli-Q water The stock solutions were stored at ◦ C for a maximum of weeks The solid-phase extraction cartridges Oasis HLB (200 mg, mL) and Strata-X-C (200 mg, mL) were provided by Waters (Milford, MA, USA) and Phenomenex (Torrance, USA), respectively Isolute Multimode (300 mg, mL) and ENV+ (200 mg, mL) were obtained from IST (Hengoed, Mid-Glamorgan, UK) Syringe filters 0.45 ␮m of nylon and nitrocellulose were purchased from Tracer Tecnolog´ıas Anal´ıticas (Madrid, Spain) Water used was purified with an Elix/Milli-Q water purification system (Millipore, Bedford, MA, USA) 2.2 Instrumentation Electrospray is the ionization source most commonly proposed in the literature for the analysis of acrylamide in food products by LC–MS/MS [9,10] However, it is well-known that ESI–MS is problematic when highly aqueous solutions, such as those required for the reversed-phase LC separation of acrylamide, are used In contrast, the gas phase ionization mechanism of APCI is less influenced by the mobile phase composition In the present work, the development of a LC–MS/MS method using this ionization source and an ion trap (LCQ) as mass analyzer was performed The validation of the method was carried out using a triple quadrupole (API 3000) 2.2.1 LCQ A quaternary pump system from Waters model Alliance 2690 was coupled to a LCQ (Finnigan MAT, San Jose, CA, USA) equipped with an APCI source and ion trap as analyzer Data acquisition was carried out by XcaliburTM 1.3 software Ionization source working parameters were optimized by flow injection, using an acrylamide standard (10 ␮g/mL), and by infusion at 10 ␮L/min into the mobile phase at their corresponding elution conditions The optimal source working parameters for monitoring positive ions were as follows: discharge current, ␮A; sheath gas, 50 a.u.; auxiliary gas, 10 a.u.; heat capillary temperature, 150 ◦ C; vaporizer temperature, 250 ◦ C; capillary voltage, 11 V; tube lens offset, 11 V Tandem mass-spectrometry provides a high degree of selectivity, leading to chromatograms that are almost free of interfering peaks Moreover, false peak identification is prevented when working in product ion scan E Bermudo et al / Analytica Chimica Acta 559 (2006) 207–214 mode The collision induced dissociation (CID) fragmentation of protonated acrylamide (m/z 72) revealed a tandem mass spectrum with a base peak at m/z 55 due to the loss of NH3 Ions at m/z 44 and 54 assigned as [CONH2 ]+ and [C3 H2 NH2 ]+ , respectively, were also observed but with low relative abundances (∼10%) The same fragmentation pattern shifted in mass according to D3 was observed for D3 -AA, internal standard used for isotopic dilution quantification Product ion scan was used scanning m/z from 35 to 100 and the m/z 55 product ion was selected for quantitative purposes The MS/MS parameters were: normalized collision energy, 33%; isolation width m/z 1.5; activation time 30 ms; activation Q, 0.45 Chromatograms were acquired in centroid mode using an injection rf value of 0.35, a maximum injection time of 50 ms and microscans for each data point 2.2.2 API 3000 A quaternary pump system from Agilent Technologies (USA) model Series 1100 was coupled to a PE Sciex API 3000 (Applied Biosystems, Foster City, CA, USA) equipped with an APCI as ionization source and a triple quadrupole as analyzer Data acquisition was carried out by Analyst 1.4 software The optimal ionization source working parameters were: nebulizer gas, 10 a.u.; curtain gas, 12 a.u.; vaporizer temperature, 250 ◦ C; nebulizer current, ␮A; declustering potential, 30 V The data acquisition was performed using selected reaction monitoring (SRM), using as precursor ion the protonated molecular ion [M + H]+ ; the collision cell offset voltage applied was 25 V with a collision gas pressure of a.u.; the m/z corresponding to the most abundant product ion (m/z 55) was monitored for quantitative purposes and m/z 54 for confirmation 2.3 Chromatographic conditions The chromatographic separation of acrylamide was carried out by reverse-phase liquid chromatography using a Semi-micro ODS-80 TS column (5 ␮m, 150 mm × 2.1 mm i.d.) (Toso Haas, Stuttgart, Germany) Optimum separation was achieved with an isocratic elution using 100% water as mobile phase at a flow rate of 0.3 mL/min The sample volume injected was 10 ␮L 2.4 Sample preparation Prior to extraction and purification stages, 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 Subsamples of g were weighed into 15 mL centrifuge tubes, and 40 ␮L of 2,3,3-D3 -acrylamide (10 ␮g/mL) and 10 mL water were added Each tube was shaken for h on a rotating shaker Rotary Mixer 34526 (Breda Scientific, Breda, The Netherlands) Then, the tubes were centrifuged at 4000 rpm for 30 with a Selecta Centronic centrifuge (Selecta, Barcelona, Spain) In the case of french fries samples, centrifugation at 20,000 rpm was performed using an Allegra 64R (Beckman, Fullerton, CA, USA) A mL aliquot of aqueous solution was filtered through a syringe filter 0.45 ␮m of nitrocellulose (Tracer) 209 2.5 Clean-up procedure In this paper, a new purification procedure (method A) that includes a pre-concentration step is proposed for the analysis of acrylamide The procedure consists of an aqueous extraction, centrifugation and clean-up using two SPE cartridges consisting of a cation exchange polymer (Strata-X-C) in combination with a cross-linked polystyrene-based polymer column (ENV+) This method was compared with a previously published purification procedure [14] (method B) to establish the best conditions for the determination of acrylamide in food products For method A, both Strata-X-C and ENV+ SPE cartridges were conditioned with mL of methanol followed by mL of water using a Supelco Visiprep® vacuum manifold (Supelco, Gland, Switzerland) Each Strata-X-C was loaded with mL of filtered extract The extract was allowed to pass through the sorbent material Then, the column was eluted with mL of water and the eluant was collected for ENV+ SPE clean-up All of the eluant collected from the Strata-X-C was loaded, and eluted with mL of MeOH:H2 O (60:40) Finally, the extract was evaporated to 400 ␮L under a stream of nitrogen and filtered through a filter 0.45 ␮m of nylon and transferred into amber glass vials for LC–MS analysis For method B, in contrast with method A, subsamples of g were weighed into 15 mL centrifuge tubes and 30 ␮L of 2,3,3-D3 -acrylamide (100 ␮g/mL) was added Oasis HLB SPE cartridges were conditioned with mL of methanol followed by mL of water Each cartridge was loaded with mL of filtered extract The column was eluted with mL of water, and the eluant was collected for Isolute Multimode cartridge, previously conditioned with mL methanol followed by mL water Finally, the eluant was filtered through a filter 0.45 ␮m of nylon (Tracer) and transferred into amber glass vials for LC–MS/MS analysis 2.6 Quantification Acrylamide was quantified by the isotope dilution method This quantification methodology gives more precise data because a correction of both extraction efficiency and changes in instrument performance is achieved The addition of a known level of deuterated (D3 -AA) standard to the samples at the beginning of the clean-up process allowed the quantification of the analyte In all cases, 0.93 ␮g of labeled standard was added, which provided native AA/deuterated AA ratios comprised between 0.05 and 20 Results and discussion 3.1 Optimization of SPE procedure The combination of two highly cross-linked polystyrene– divinylbenzene-based polymer sorbents (Strata-X-C and ENV+) was evaluated to obtain clean hydro-organic extracts of acrylamide Both cartridges are functionalized with polar and cation exchange groups that exhibit numerous retention mechanisms including hydrophobic, hydrogen bonding, ␲–␲ and cation exchange, making these columns ideal for the extraction of polar 210 E Bermudo et al / Analytica Chimica Acta 559 (2006) 207–214 analytes [15,16] It is well-known that enamides can be protonated with hydrogen ion donors to form N-acyliminium cation species in solution Such behavior can be favorably used to retain acrylamide on a cation exchange sorbent (Strata-X-C) Then, the strongly acidic sulfonic acid moieties on this cartridge can furnish the hydronium ion for protonating acrylamide favoring its retention On the other hand, the functionalization with hydroxy groups of ENV+ allows the retention of polar ionizable and/or non-ionizable analytes from aqueous samples, so it could be a good choice of column for the clean-up of acrylamide Because of the high polarity of acrylamide, water was chosen as eluting solvent for Strata-X-C However, the combination of this extractant with other polar solvents such as methanol could offer some advantages such as easier solvent evaporation and consequent higher pre-concentration in ENV+ Nevertheless, a minimum elution volume was required in order to achieve a maximum enrichment factor without an evaporation step To minimize elution volume, the elution curve [recovery (%) versus volume of water] was established; from this curve, the minimum solvent volume needed to remove the retained acrylamide quantitatively was found to be mL for Strata-X-C cartridge On the other hand, water with different percentages of methanol comprised between 50 and 100% was added to the ENV+ column for eluting the retained acrylamide The study of the recoveries obtained with several methanol/water mixtures showed that the acrylamide elution decreased using high percentages of methanol and the maximum amount that provided the best recovery was 60% (Fig 1A) Higher amounts of water did not significantly improve the recoveries and, moreover, prevent fast evap- oration in the concentration step Then, the optimum elution volume using MeOH:H2 O (60:40) was found to be mL in this case Moreover, in SPE the breakthrough volume is an important feature to take into account since it determines the detection limit that can be reached [17] Different volumes of standard solutions of acrylamide at various concentration levels with the sample amount kept constant (1 ␮g) were used After extraction the samples were injected into the LC–MS/MS system (LCQ) Peak areas were measured and the recoveries were calculated by comparing the peak areas with those of a control sample (1 ␮g/mL), representing 100% recovery Fig 1B shows a considerable decrease in the recovery of acrylamide at mL in both cases For this reason, mL was taken as the optimum volume to obtain maximum recoveries Under these conditions the recovery was 90 and 95% for Strata-X-C and ENV+, respectively The recovery of the whole process was determined using different representative food samples Biscuits, crisp bread, breakfast cereals and potato chips fortified with labeled acrylamide (D3 -AA) at 0.3 ␮g/g were selected Values equal to or higher than 85% and a R.S.D (%) of 3–14% were obtained showing that this method can be properly used for the clean-up of acrylamide in food samples These values are in accordance to those previously presented using different clean-up processes [4,14,18] 3.2 Quality parameters To check performance of the developed LC–MS/MS method, quality parameters such as limit of detection, limit of quantification, repeatability or run-to-run precision, medium term or day-to-day precision and linearity range were established Limit of detection (LOD) and limit of quantification (LOQ) were determined as the amount of analyte that produced a signalto-noise ratio of 3:1 and 10:1, respectively LOD and LOQ were calculated using standard solutions at low concentration levels, and a blank sample (bread crumb) spiked with very low amounts of AA in the beginning of the extraction For standards, the LOD and LOQ values were 250 and 800 pg injected, respectively In bread crumb, LOD and LOQ were always higher (45 and 100 ng/g) than those for standards due to matrix complexity that affected both chemical noise and ionization efficiency Linearity range was studied between limit of quantification and 20 ␮g/mL and the response was linear up to 10 ␮g/mL To determine repeatability, 11 replicate injections of a AA standard solution at both low (3 × LOD) and medium (0.3 ␮g/mL) concentration levels were carried out obtaining R.S.D values lower than 2.5% For the determination of medium-term precision, five replicate injections along consecutive days of an AA standard solution at low and medium levels of concentration were carried out R.S.D were calculated from the 15 calculated concentration values, and results lower than 8% were obtained 3.3 Analysis of Spanish food samples Fig (A) Optimization of solvent elution for ENV+ cartridge (B) Effect of sample volume on the recovery of acrylamide for Strata-X-C ( ) and ENV+ ( ) Vertical lines represent the standard error of the mean value (n = 3) The proposed method was applied to the analysis of acrylamide in food products Thirty samples were collected from supermarkets in Barcelona (Spain) in the summer of 2004 E Bermudo et al / Analytica Chimica Acta 559 (2006) 207–214 211 Fig Chromatograms of acrylamide in two food samples: (A) crisp bread sticks, ion trap (product ion scan); (B) pastry product (sobaos Mart´ınez), triple quadrupole (MRM) Chromatographic conditions included in Section and the winter of 2005 All samples were analyzed with method A (Section 2.5) to provide information on acrylamide amounts in a set of food commodities consumed in Spain The analytical survey comprised potato chips, biscuits, cookies, crisp breads, crackers, dried fruits, pastry, breakfast cereals, chocolate, coffee and french fries prepared from frozen potatoes Quantification was performed by isotopic dilution method, and three replicates were determined for each sample Analyses were carried out using the LC–MS/MS developed method using an ion-trap mass analyzer (LCQ) However, when the quantification was not possible because the concentration was lower than the detection limits (45 ng/g) of this instrument, a triple quadrupole was employed [19] As an example, Fig shows the chromatograms corresponding to a crisp bread stick sample and a pastry product (sobaos Mart´ınez) obtained using the ion trap and triple quadrupole mass analyzers, respectively The chromatograms of acrylamide (m/z 55) and its labeled compound (m/z 58) are presented in Fig 2A and B The confirmation was performed by the tandem mass spectrum for the crisp bread sample in the ion-trap instrument (Fig 2A) and by the transition m/z 72 → 54 (Fig 2B) for API 3000 in the pastry product sample The amount of acrylamide present in the food extracts was calculated using the integrated area of the m/z 58 response 212 E Bermudo et al / Analytica Chimica Acta 559 (2006) 207–214 Table Food analysis using a LCQ and API 3000 as mass analyzers (n = 3) Sample Snacks Frit Ravich Potato chips Torres Potato chips Vidal Sticks Vicente Vidal Lays Light 40% Lays Mediterraneas Artesanas Potato chips Oleguer French fries Findus Potato Balls Findus “Churros” Crisp bread Bimbo Crisp bread sticks Panrico Crisp bread Recondo Crackers Cream Cuetara Biscuits Napolitanas Biscuits Oreo Biscuits Birba Biscuits Trias Milk bread La Bella Easo Sobaos Martinez Small sponge cakes La Bella Easo Spiral shaped pastry Dulcesol Milk and almonds Chocolate Valor Chocolate with biscuit Nestle Breakfast cereals Nestle Corn flakes Pascual Nescafe soluble coffee Roasted hazelnuts Borges Pistachios Frit Ravich Peanuts Eagle LCQ API 3000 Mean (ng/g) R.S.D (%) Mean (ng/g) R.S.D (%) 528 500 524 9250 403 753 799 512 940 430 – 175 100 350 301 544 300 120 – – – 110 117 100 558 350 188 – – 140 12.0 7.4 6.2 5.6 13.8 2.0 7.9 5.5 12.0 12.7 – 3.1 1.0 12.8 2.8 8.5 6.7 1.0 – – – 6.2 7.5 12.0 7.3 7.6 9.5 – – 5.1 502 548 544 9280 417 745 805 532 925 450 25 176 98 370 299 589 312 118