Determination of Bisphenol A in Milk Powder using a Chromolith HighResolution RP-18 endcapped column

Transcription

1 Determination of Bisphenol A in Milk Powder using a Chromolith HighResolution RP-18 endcapped column Adelijiang Xiamuxiding and Patrik Appelblad EMD Millipore Introduction Bisphenol A (BPA), figure 1, is a synthetical organic compound, a diphenylmethane derivative. BPA is frequently used in plastics and epoxy resin production, where final consumer products are clear and durable. BPA is used in polycarbonate, a high performance transparent, rigid plastic (used for baby and recycling water bottles). Another common use of epoxy resins containing BPA is surface coating of food and beverage cans. BPA can migrate in small amounts into food and beverages stored in materials containing the substance. Canada became the first country to declare BPA as a toxic substance in BPA is permitted in the European Union (EU) under Regulation 10/2011/EU for use in food contact materials, i.e. plastic materials and articles intending to come into contact with foodstuffs. However, the EU and the United States have banned BPA use in baby bottles. The reason is that Bisphenol A is thought to act as an endocrine disruptor and may lead to negative health effects. (I) (II) Figure 1. The chemical structure of Bisphenol F (I) and Bisphenol A (II). This application note presents a rapid and inexpensive method for BPA screening in milk samples using a monolithic Chromolith HighResolution RP-18 endcapped column and fluorescence detection on a Chromaster HPLC system. The method allows quantitative analysis of BPA above 1 ppb (1 ng/ml). The detection limit is sufficient and compare well with other reported non-ms based methods and can be used for BPA screening purposes. However, data from non-ms methods for BPA in food and beverage samples should be confirmed by MS methods; more specifically isotope dilution MS or tandem mass spectrometry (MS/MS) of regulatory reasons. The sensitivity of the method can be further improved by increasing lamp energy (from medium to high) but in order to maintain maximum detector Xenon-lamp lifetime this was unwanted. The analytical data presented herein demonstrate the excellent long term performance attained with Chromolith columns.

2 Experimental Section Column: Chromolith High Resolution RP-18 endcapped, 100x4.6 mm (P/N ) Injection: 10 μl Detection: VWR Hitachi Chromaster-5440 Fluorescence Detector, λ ex =230; λ em =316 nm Mobile Phase: Acetonitrile and Water 25:75 (v/v) Acetonitrile (grade for liquid chromatography LiChrosolv ; P/N ) Water for chromatography LiChrosolv ( ) or directly from a Merck Millipore Milli-Q system Flow Rate: Flow Rate gradient ml/min, ml/min, @ ml/min Temperature: 35 C Diluent: Mobile phase Backpressure: Bar ( psi) Milk powder The analysed milk powder contains the following: Protein (11%), Carbohydrates (55%, of which 48% is lactose), Fat (27%), Fiber (2.3%) and Vitamins (A, D, E, K, C, Thiamine, Riboflavin, Niacin, Vitamin B6, Biotin), Panthothenic acid, Choline, Inositol, Taurine, L-Carnithine, Minerals (Na, K, Cl, Ca, P, Fe, Mg, Zn, I, Cu, Mn, F, Se) and Nucleotides (AMP, CMP, GMP, IMP, UMP) Preparation of 100 ml baby milk formulation Boil water and transfer 100 ml into plastic baby milk bottle. Add 3 spoons (12.1 g) of powder and mix thoroughly by shaking. Sample Preparation 100 ml milk sample was acidified with 100 μl of concentrated (37%) hydrochloric acid. 5mL of the acidified milk solution was transferred into a 20 ml plastic vial and mixed with 10 ml of mobile phase and shaken for 5 minutes. The samples were thereafter centrifuged for seven minutes at 6400 rpm and 15 C, followed by filtered using a Merck Millipore Samplicity sample preparation unit with 0.20 μm hydrophilic PTFE Millex Samplicity filters (but normal hydrophilic PTFE 0.20 μm syringe filters also work well but is manual and thus more laborious). The clear sample solution was finally transferred to 2.0 ml HPLC vials prior to analysis. Baby milk formulation spiked with BPA 12.1 g of milk powder dissolved in 100 ml hot water. Spike BPA to the solution, Add 0.1% of Hydrochloric acid and 10% acetonitrile. Add internal standard BPF 60 ppb. Centrifuge and filter the sample with Millex Simplicity filter prior analysis.

3 Results and Discussion 100 Intensity (mv) Retention Time (minutes) No. Compound Retention Time (min) Retention factor Asymmetry 1 Void volume Bisphenol F Bisphenol A Figure 2. Separation of a standard solution containing 150 ppb Bisphenol A and 60 ppb F dissolved in methanol on a 100x4.6 mm Chromolith High Resolution RP-18 endcapped column using a mobile phase based with acetonitrile and water 25:75 (v/v) at a flow rate of 2.0 ml/min for the first 4 minutes, thereafter it was increased to 2.0 ml/min for another 3.5 minutes. The injected sample volume was 10 µl and detection was carried out on a VWR Hitachi Chromaster HPLC system equipped with a 5440 Fluorescence Detector (λ ex =230; λ em =316 nm). Calibration Data Five replicate injections of BPA standard solution (n=5) were analyzed at five concentration levels ( ppb) while keeping the internal standard (BPF) concentration constant at 60 ppb. to determine repeatability, and within and between day reproducibility, see Table 1. The relative standard deviation, given in %, is presented in brackets for every calibration point. In Figure 3 the calibration graph from day 1 of analysis is presented and a table 2 summarizes the curve equations and regression analysis for three separate calibration curves on three different days of analysis over a three week period. Data compares very well and the method can be considered as robust and reliable.

4 Calibration Curve Data Bisphenol A (BPA) and internal Standard Bisphenol F (BPF) Mean Response (Arbitrary Area Units) Area Ratio N Day Concentration BPA BPF BPA/BPF (ppb) ± 3566 (±1.9%) ± (±3.6%) 0.134± (±4.2%) ± (±3.8%) ± (±4.2%) 0.440± (±3.8%) ± (±3.5%) ± (±3.4%) 0.856± (±1.4%) ± (±0.9%) ± (±1.1%) 1.372± (±1.0%) ± (±1.3%) ± (±1.0%) 1.667± (±0.8%) ± (±9.0%) ± (±3.9%) 0.139± (±8.6%) ± (±2.1%) ± (±3.0%) 0.486± (±2.2%) ± (±3.6%) ± (±0.9%) 0.940± (±4.3%) ± (±2.4%) ± (±1.5%) 1.506± (±1.1%) ± (±4.7%) (±47%) ±41307 (±31%) (±3.1%) 1837± 1.837± (±2.9%) (±29%) ± (±6.5%) ± (±3.0%) 0.154± (±8.7%) ± 9193 (±1.6%) ± (±2.4%) 0.445± (±3.3%) ± (±1.0%) ± (±3.0%) 1.069± (±3.8%) ± (±2.0%) ± (±3.0%) 1.460± (±3.2%) ± (±1.0%) ± 7119 (±0.5%) 1.948± (±1.1%) 5 Table 1. Calibration curve data (Day 1). Five replicate injections of BPA standard solution were analyzed at five concentration levels ( ppb) while keeping the internal standard (BPF) concentration constant at 60 ppb. BPA/BPF BPA Concentration (ppb) Figure 3. Calibration curve (Day 1). BPA/BPF represent the area ratio between the BPA and BPF measurement. Curve Equation and Regression Coefficient Day Curve Equation (R 2 ) 1 y = x y = x y = x Table 2. The curve equations and regression analysis for three calibration curves constructed at three different days of analysis over a three week period.

5 Analysis of Spiked Samples Baby milk formulation was spiked with BPA at four different concentration levels ( ppb) to determine total method recovery. Internal standard (BPF) at a concentration of 60 ppb was added to improve method robustness. Six individual samples were prepared and injected in triplicate (n=18) at each spiking level. The recovery of BPA was found to be ranging between 85-91%. The method performance greatly improved with use of internal standard, see table 3. The detection limit (LOD) was determined to1.4 ppb (ng/ml) based on analysis of low calibration standards and calibration curve data. Recovery Baby Milk formulation spiked with Bisphenol A (BPA) Spiked BPA Concentration (ppb) BPA Mean Response (Arbitrary Area Units) Area Ratio (BPA/BPF) Table 3. Determination of BPA recovery in spiked baby milk samples. Recovery (%) ± (±4.1%) 0.511± (±2.8%) 89.0± ± (±3.7%) 0.976± (±1.8%) 86.7± ± (±2.8%) 1.467± (±0.1%) 87.6± ± (±2.2%) 2.376± (±0.3%) 85.6± N Inten nsity (mv) Time Flow Rate (min) (ml/min) Chromatographic Data Retention Time (minutes) No. Compound Retention Time (min) Retention factor Asymmetry 1 Void volume Bisphenol F Bisphenol A Figure g of milk powder dissolved in 100 ml hot water. Spike 300 ppb of BPA to the solution. Add 0.1% of Hydrochloric acid and 10% acetonitrile. Add Internal standard BPF 60 ppb. Centrifuge and filter the sample with Millex Simplicity hydrophilic filter prior analysis.

6 Method robustness and column longevity were evaluated by performing long term analysis of extracted milk samples and evaluated by means of retention time and area measurement for BPA and BPF. Over 2700 injections were carried out over a period of three weeks. The relative standard deviation on retention time was better than 2.5 % for both compounds, see table 4. In total 65 L of mobile phase was flushed through the column (24 ml per injection) before peak shape started to deteriorate. Retention time reproducibility Bisphenol F (BPA) Bisphenol A (BPA) N Retention Time (min) S.D. (min) R.S.D (%) Retention Time (min) S.D. (min) R.S.D (%) * Table 4. Retention time reproducibility of a three week period of analysis. In total over 2700 samples were analyzed in this period and the column was flushed with 65 liter of mobile phase. Conclusion This method allows quantitative analysis of BPA above 1 ppb (1 ng/ml) in baby milk samples without any pre-concentration nor sample preparation procedures beside protein crash followed by filtration. The detection limit is sufficient for screening purposes and compare well with other reported non-ms based methods. The method sensitivity can be further improved by increasing lamp energy (from medium to high) but to maintain maximum detector Xenon-lamp lifetime (and lowest possible running cost) this was unwanted. The detection limit is sufficient and compare well with other reported non-ms based methods and can be used for BPA screening purposes. However, data from non-ms methods for BPA analysis in food and beverage samples should be confirmed by MS methods; and more specifically isotope dilution MS or tandem mass spectrometry.