Acrylamide quantitation in a diverse range of food matrices


Sabarinathan1, Sashank Pillai1 and Craig M. Butt2
1SCIEX, India; 2SCIEX, USA

Abstract


This technical note describes the LC-MS/MS analysis of acrylamide in solid food matrices (baby food, raw potato, potato chips, bread, biscuits and rusk), with quantitative performance achieving the AOAC Standard Method Performance Requirements (SMPR)1 criteria. Using the SCIEX QTRAP 4500 system, matrix spikes (n=6) achieved mean absolute recoveries of 77%–100%, demonstrating good acrylamide recovery for the diverse range of matrices. Matrix spiking levels varied based on the background acrylamide levels in the food. Internal standard-normalized recoveries ranged from 96%–121% with %CVs ranging from 2.9%–15% (XICs for raw potato shown in Figure 1). Method robustness was demonstrated by 20 continuous injections of a potato chip quality control sample, containing 1150 µg/kg of acrylamide, which showed consistent peak areas with a %CV of 4.1%. Further, the method applicability was shown through the analysis of 3 commercial food items.

Figure 1. Representative extracted ion chromatograms (XICs) of acrylamide from the quantifier ion (m/z: 72.3 / 55.0). The quantifier transition is shown for the diluent blank, 1 ng/mL solvent standard, unspiked raw potato and 20 µg/kg raw potato spike. Acrylamide was not detected in the diluent and unspiked raw potato.

Introduction


Acrylamide forms naturally as a byproduct during the high-temperature cooking of starch-rich foods and is elevated in foods such as french fries, potato chips, baked foods, cereals, chocolate and roasted coffee. 2 Specifically, acrylamide has been shown to form at high heat (>100°C) during the Maillard reaction between certain amino acids (for example, asparagine) and reducing sugars. 3,4 The International Agency for Research on Cancer (IARC) has classified acrylamide as a probable human carcinogen5 and the European Food Safety Authority (EFSA) determined that acrylamide in food potentially increases cancer risk. 6 Currently, there are no regulations concerning acrylamide in food. However, sensitive, accurate and robust food analysis methods are necessary for food safety and the protection of human health. This technical note presents a robust method for analyzing acrylamide in food samples with focus on achieving the performance criteria outlined in the AOAC SMPR 2022.006.

Key benefits of the analysis of acrylamide in food using the SCIEX QTRAP 4500 system
 

  • Method accuracy across a diverse range of food matrices: Mean absolute recoveries ranged from 77%–100%, demonstrating good method recovery and achieving the AOAC recovery criteria

  • 1 ng/mL limit of quantitation (LOQ) in solvent standards: The SCIEX QTRAP 4500 system showed good sensitivity for acrylamide with a LOQ of 1 ng/mL in neat solution

  • Good analyte retention and void volume separation: The Phenomenex Luna™ Omega Sugar column showed a retention factor (k’) of 0.61 (retention time ~3.5 min) within the 9-min runtime

  • Method robustness during 20 continuous injections: Raw acrylamide area counts showed a %CV of 4.1%, demonstrating no loss in sensitivity

Methods


Standard preparation: The acrylamide standard was acquired from Sigma-Aldrich and the acrylamide-d3 internal standard (ISD) was purchased from Toronto Research Chemicals. Analyte and internal standard stock solutions were prepared in LC-MS grade water. Calibration standards were prepared in acetonitrile and ranged from 1–500 ng/mL. The final internal standard concentration was 20 ng/mL in the calibration standards.

Procedural recoveries in spiked matrices: Method recovery was determined by performing matrix spike experiments (n=3) prior to sample preparation (“pre-spike”) and, in separate samples, after sample preparation (“post-spike”). The absolute recovery was calculated as the ratio of the pre- and post-spike area counts. Samples were initially screened to determine the spiking levels. The AOAC SMPR document specifies that matrices with low levels of acrylamide be spiked at the LOQ (20 µg/kg for baby food and bread, 50 µg/kg for other matrices). Acrylamide was not detected in the raw potato and was spiked at 20 and 50 µg/kg. Acrylamide levels of 16 µg/kg were detected in the rusk and therefore acrylamide was spiked at 50 µg/kg. High background levels of acrylamide were detected in the baby food, potato chips, biscuits and bread. Therefore, these matrices were spiked at 3000 µg/kg, approximately 3-5x the background levels, as specified by the AOAC SMPR. The acrylamide-d3 internal standard was spiked at 400 µg/kg in all the samples.

Sample preparation: Food samples were purchased from a local grocery store. A 1 g sample aliquot was weighed into a 50 mL centrifuge tube and 10 mL of water was added. The solution was vortexed for 10 minutes and centrifuged at 4500 rpm for 10 minutes. After centrifugation, the water layer was transferred to a 50 mL centrifuge tube and 10 mL of acetonitrile was added. The solution was vortexed for 10 minutes, 4 g of magnesium sulfate and 2 g of sodium chloride was added. The solution was vortexed for 10 minutes and then centrifuged at 4500 rpm for 10 minutes. After centrifugation, 1 mL of the acetonitrile layer was collected and filtered through a Phenomenex CLARIFY-PTFE 0.22 µm syringe filter (P/N: AF8-7702-12) prior to analysis. Figure 2 highlights the sample clean-up efficiency of the extraction protocol, as shown by potato chips and baby food.

Chromatography: An ExionLC AD system was used with the Phenomenex Luna Omega 3 µm SUGAR column for chromatographic separation (150 mm x 4.6 mm, P/N: 00F-4775- E0). The mobile phases were 5mM ammonium formate in water (pH 3.22) and acetonitrile. The flow rate was 0.800 mL/min and the gradient conditions used are shown in Table 1. The injection volume was 5 µL and the column oven temperature was 30°C. The autosampler temperature was set to 10°C and 0.5 mL of rinsing solution was used for the needle washing.

Table 1. LC gradient program for the analysis of acrylamide.

Mass spectrometry: Samples were analyzed on the SCIEX QTRAP 4500 system operated in multiple reaction monitoring (MRM) mode with positive electrospray ionization. Optimized source and compound-specific parameters are presented in Tables 2 and 3, respectively. Two selective MRM transitions were monitored (Table 3). Confirmation of the targeted analytes was based on the ion ratio.

Data processing: All data were processed using SCIEX OS software (2.1.6).

Table 2. Optimized source and gas parameters for the analysis of acrylamide using the SCIEX QTRAP 4500 system.

Table 3. MRM compound-dependent parameters for the analysis of acrylamide and acrylamide-d3 on the SCIEX QTRAP 4500 system.

Figure 2. Representative extracted samples of potato chips (left) and baby food (right) showing the extraction efficiency. A) Matrix samples before the clean-up procedure. B) Sample clean-up using MgSO4 and NaCl, layer separation between water (lower layer) and acetonitrile (upper layer). C) The final sample of the upper acetonitrile layer was filtered using PTFE 0.22 µm hydrophobic syringe filter prior to LC-MS/MS analysis.

Good chromatographic retention using the Phenomenex Luna Omega Sugar column


Acrylamide analysis is challenging due to its highly polar property resulting in poor retention on reverse-phase chromatography columns. Traditional HILIC columns have been employed but are often difficult to use and require long conditioning and equilibration times. However, the unique characteristics of the Phenomenex Luna Omega Sugar column and 9-min gradient provided good analyte retention after the void volume, as shown by a retention factor (k’) of 0.61 (retention time ~3.5 min). Figure 1 shows good peak shape and retention of acrylamide and acrylamide-d3. 

Sensitivity, precision and linear dynamic range in solvent-based calibration standards


The SCIEX QTRAP 4500 system showed good sensitivity for acrylamide and achieved an in-vial LOQ of 1 ng/mL in the solvent-based standards. The LOQ was chosen based on the 2 MRM transitions, both achieving a signal-to-noise (S/N) ratio of >10, accuracy ±10%, precision <10% and ion ratio tolerance ± 30%. Accuracy and precision of the 1 ng/mL LOQ standard was further evaluated with duplicate injections of the 1 ng/mL LOQ standards prepared in triplicate (n=6). The mean LOQ accuracy was 105% and mean LOQ %CV was 3.6% (Table 4). 

Table 4. Calibration curve correlation coefficient (r2 ) and accuracy range across the calibration curve and LOQ average accuracy and %CV for quantifier ion. 

Considering the full 1–500 ng/mL calibration range, good linear dynamic range was shown with an r value of 0.999. Further, accuracies ranged from 95%–107% for all other calibration levels. The internal standard-normalized calibration curve for acrylamide from the quantifier transition is shown in Figure 3. 

Figure 3. Representative calibration curve for acrylamide from the quantifier transition (m/z: 72.3/ 55.0). A linear curve across the range of 1–500 ng/mL with an r-value of >0.999 was achieved using the weighing factor 1/x2 .

Accuracy and precision in matrix spiked samples


Commercial food items for the matrix spike experiments were purchased from a local market and processed, as described. Sample extracts were analyzed against the solvent calibration curve. Prior to determining the matrix spiking levels, all food matrices were analyzed to determine the background acrylamide concentrations. Acrylamide was not detected in the raw potato and therefore, these samples were spiked at the AOAC LOQ target levels of 20 and 50 µg/kg. Low acrylamide levels were detected in the rusk samples and therefore samples were spiked at 50 µg/kg. In contrast, the potato chips, biscuits, baby food and bread contained relatively high acrylamide levels. According to the AOAC SMPR document, these food matrices were spiked at 3–5x higher than the background acrylamide levels, at a concentration of 3000 µg/kg (XICs shown in Figure 4). 

Figure 4. Representative XICs from unspiked and 3000 µg/kg acrylamide spike in potato chips, baby food, biscuit and bread samples. The quantifier transition is shown for all XICs.

The method recovery was calculated as the ratio of the pre- and post-spike raw area counts and was not normalized to the acrylamide-d3 response. This recovery value is considered the absolute recovery because the internal standard should correct for any loss during sample preparation and matrix effects. Pre-and post-extraction spikes were prepared in triplicate and each sample was injected in duplicate (n=6). Concentrations were blank-corrected using the unspiked matrix levels. Absolute recoveries ranged from 77%–100%, achieving the SMPR recovery criteria of 75%–110%. These results demonstrated good method recovery for the diverse range of matrices tested (Tables 5 and 6).

The acrylamide-d3 internal standard, spiked at the beginning of the sample preparation, improved the overall data quality. ISD-normalized acrylamide recovery values ranged from 96%–121%. Although the absolute recoveries showed acceptable values, the stable isotope-labelled acrylamide internal standard is recommended for complex matrices.

Table 5. Average accuracy, %CV and recovery for acrylamide in potato chips, biscuits, bread and baby food spiked at 3000 µg/kg. All the measurements were performed based on pre-spiked and post-spiked samples (n=6).

Table 6. Average accuracy, %CV and recovery for raw potato and rusk samples spiked at 20 µg/kg and 50 µg/kg. All the measurements were performed based on pre-spiked and post-spiked samples (n=6). The rusk sample was not spiked at 20 µg/kg due to background levels detected.

Method robustness


Method robustness was demonstrated through 20 continuous injections of a single unspiked potato chip extract. Figure 5 shows the raw acrylamide area count for the 20 injections. The %CV was 4.1% with a mean area count of 1.9E6 (approximately 1150 µg/kg). These results demonstrate robust sample preparation and analysis methods with no observable sensitivity loss. 

Figure 5. Method robustness in extracted potato chip. Raw acrylamide area counts for 20 continuous injections.

Analysis of market samples


The applicability of the method was further demonstrated by analyzing commercially available potato chips, microwave-fried potato chips and biscuits. Samples were extracted in triplicate using the isotope dilution method described above. Mean concentrations were 446 µg/kg in the potato chips, 830 µg/kg in the microwave-fried potato chips and 1601 µg/kg in the biscuits (Table 7). These results show that the method is suitable for analyzing acrylamide in a variety of solid food matrices.

Table 7. Mean acrylamide concentrations (µg/kg) in different solid food matrices (n=3).

Conclusion
 

  • The mean absolute recovery ranged from 77%–100% across the diverse range of food matrices (baby food, raw potato, potato chips, bread, biscuits and rusk). These results satisfied the recovery criteria outlined for acrylamide in the AOAC SMPR. 
  • The matrix LOQ in raw potato was shown to be 20 µg/kg. This was the only matrix without background levels of acrylamide. 
  • The solvent standard LOQ was 1 ng/mL, demonstrating the sensitivity of the SCIEX QTRAP 4500 system 
  • The chromatographic gradient using the Phenomenex Luna Omega SUGAR column showed good analyte retention and void volume separation 
  • The method was robust in 20 continuous injection of a potato chip extract showing precision (%CV) of 4.1%

References
 

  1.  Standard Method Performance Requirements (SMPRs®) for Acrylamide in Potato Products, Baby Food, Bread, Other Cereal and Bakery Products, Cocoa Products, Coffee, Tea, Herbs and Spices (Including Their Extracts and Mixtures), Dry Pet Food, and Nuts. 2023 AOAC International. AOAC SMPR® 2022.006.

  2. Rosén, J.; Hellenäs, K.-E. 2002. Analysis of acrylamide in cooked foods by liquid chromatography tandem mass spectrometry. Analyst, 127(7), 880-882.

  3. Mottram, D.S.; Wedzicha, B.L.; Dodson, A.T. 2002. Acrylamide is formed in the Maillard reaction. Nature, 419, 448-449.

  4. Stadler, R.H.; Blank, I.; Varga, N.; Robert, F.; Hau, J. et al. 2002. Acrylamide from Maillard reaction products. Nature, 419, 449-450.

  5. World Health Organization, International Agency for Research on Cancer. 1994. Lyon, France. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. Volume 60: Some industrial chemicals.

  6. European Food Safety Authority (EFSA). 2015. Scientific opinion on acrylamide in food. EFA Journal, 13(6), 1-321.
Request information