2.1

Experimental resin

An experimental light-cured dental resin was prepared, consisting of 83% silanized silica (CAS Number 100402-78-6), 16% BisGMA (CAS Number 1565-94-2), <1% of Camphorquinone (CPQ, CAS Number 10373-78-1), and <1% ethyl 4 – (dimethylamino) benzoate (EDMAB, CAS Number 10287-53-3). The starting materials were the same as used in 3M commercial oral care resins, and were obtained from proprietary sources. Ingredients were processed in a planetary mixer and the mixed resin was transferred to circular stainless steel molds (15 mm ID × 1.7 mm high) for curing. Light-curing was carried out in a Triad 2000 curing unit (Dentsply, York PA, USA) for 10 min per side, using an unfiltered halogen bulb emitting a broad spectrum of visible light, with an irradiance of 150 mW/cm2 and a light source-sample distance of approximately 10 cm. We used the Dentsply unit rather than a hand-held curing light because for this study we wished to eliminate the oxygen inhibition layer as a variable in the extraction. During irradiation, poly films were used to cover the top and bottom surfaces of the samples. The cured discs were removed from the molds and lightly cleaned with a low-lint laboratory wipe to remove any dust generated during disassembly. Cured discs were stored at room temperature until the time of extraction (total of 7 weeks of storage).

2.2

Resin extraction

Resin extraction was carried out using guidance in ISO 10993-12 . Cured discs were placed in glass vials with polypropylene caps (Fisher Scientific, Pittsburg PA, USA) and extracted in HPLC grade methanol (VWR Scientific, Radnor PA, USA) at a ratio of 3.3 cm ² disc/mL methanol, equivalent to 0.42 g disc/mL. Methanol was used because it is a suggested extraction solvent for generation of medical device samples for testing in biological systems and because our experience indicates that methanol is a highly aggressive solvent for BisGMA-based resins.

The methanol was spiked with 30 ng/mL d16 BPA (Cambridge Isotope Laboratories, Tewksbury, MA, USA) as an internal standard (IS). The purity of d16 BPA was listed as 98% d16 by the manufacturer. Extraction was conducted at 37 °C with mild agitation (100 rpm) for 24 h. A method blank consisting of methanol alone was included with all extraction steps and subsequent analyses.

2.3

BPA quantitation

BPA separation and quantitation were carried out on an Agilent 1200 SL with UV diode ray detection (DAD) (Agilent Technologies, Santa Clara, CA, USA) coupled to an AB/Sciex Qtrap 3200 Liquid Chromatograph with tandem mass spectrometry capability (AB/Sciex LLC, Framingham, MA, USA). Separation was performed with an Agilent Eclipse XDB-C18 column (Agilent Technologies, Santa Clara, CA, USA) with 2.1 × 150 mm dimensions and 3.5 micron packing. All chromatography solvents were HPLC grade from EMD Millipore, Billerica, MA, USA. The mobile phase was a mixture of unbuffered water (channel A) and methanol (channel B). The gradient was applied as follows: 20% B (preinjection reequilibration, 5 min); 70% B (0−0.5 min); 75% B (0.5−6 min); 99% B (6.1 min followed by a 4 min hold). The flow rate was 0.5 mL/min and the injection volume was 5 μL. Column effluent was split between the diode array detector and the Qtrap mass spectrometer. A wavelength of 225 nm was used to plot the UV response, but full DAD spectra were acquired during all LC/MS experiments. The Qtrap was operated in standard triple quadrupole multiple reaction monitoring mode. The Turbo IonSpray source of the Qtrap was optimized for BPA detection with ESI voltage of -4500 V, nebulizer gas of 80 psi, desolvation gas pressure and temperature of 70 psi and 700 °C, respectively, and counter-current gas of 30 psi. A declustering potential of -50 V was optimized for generation of the deprotonated BPA molecular ion, and the entrance potential and collision cell entrance potential were optimized at -10 and -20 V, respectively. Two fragment ions, one involving loss of methyl radical (m/z 212) and one involving loss of phenol (m/z 133) were used to monitor BPA. The 227/212 transition was optimized using a collision energy of -25 V and collision exit potential of -2.0 V, while the 227/133 transition employed a collision energy of -36 V and a collision exit potential of -1.0 V. Each of the three transitions monitored during the entire chromatographic run was assigned a 100 ms dwell time.

The same d16 BPA used in the resin extraction was employed as an IS in all cases to correct for any losses during sample preparation or variability injection volume or detection efficiency. The d16 IS was monitored using a 241/223 transition involving loss if d3 methyl radical was used and the same tuning employed for the analogous transition if d0 BPA was used. Tests of the d16 IS in our LC/MS/MS assay showed that it gave no signal at d0, and as such did not contribute to the BPA signal during the assay. The level of detection was 3 ppb BPA.

2.4

Identification of co-eluting chemical(s)

Identification of co-eluting chemicals was carried out on an Agilent 1260 HPLC coupled to an Agilent 6224 Time of Flight (TOF) LC/MS (Agilent Technologies, Santa Clara, CA, USA) with Dual ESI source operated in positive and negative modes. Chromatography was performed as for the initial BPA analyses except that some samples were also run using water and acetonitrile (both containing 6 mM ammonium acetate to aid ionization), to enhance detection of less polar analytes in positive mode for the TOF analyses. The dual ESI was set at nebulizer gas pressure of 45 psi, desolvation gas flow and temperature of 11 lpm and 325 °C respectively, capillary voltage of 400 V, skimmer voltage of 65 V, fragmenter voltage of 155 V, and octopole radiofrequency of 750 V. All scans were automatically recalibrated with the standard Agilent Reference Mixture using 121.0509 and 922.0098 ions for recalibration. The full scan mass range employed was 105−1700 atomic mass units (amu) at 1 scan per second.

2.1

Experimental resin

An experimental light-cured dental resin was prepared, consisting of 83% silanized silica (CAS Number 100402-78-6), 16% BisGMA (CAS Number 1565-94-2), <1% of Camphorquinone (CPQ, CAS Number 10373-78-1), and <1% ethyl 4 – (dimethylamino) benzoate (EDMAB, CAS Number 10287-53-3). The starting materials were the same as used in 3M commercial oral care resins, and were obtained from proprietary sources. Ingredients were processed in a planetary mixer and the mixed resin was transferred to circular stainless steel molds (15 mm ID × 1.7 mm high) for curing. Light-curing was carried out in a Triad 2000 curing unit (Dentsply, York PA, USA) for 10 min per side, using an unfiltered halogen bulb emitting a broad spectrum of visible light, with an irradiance of 150 mW/cm2 and a light source-sample distance of approximately 10 cm. We used the Dentsply unit rather than a hand-held curing light because for this study we wished to eliminate the oxygen inhibition layer as a variable in the extraction. During irradiation, poly films were used to cover the top and bottom surfaces of the samples. The cured discs were removed from the molds and lightly cleaned with a low-lint laboratory wipe to remove any dust generated during disassembly. Cured discs were stored at room temperature until the time of extraction (total of 7 weeks of storage).

2.2

Resin extraction

Resin extraction was carried out using guidance in ISO 10993-12 . Cured discs were placed in glass vials with polypropylene caps (Fisher Scientific, Pittsburg PA, USA) and extracted in HPLC grade methanol (VWR Scientific, Radnor PA, USA) at a ratio of 3.3 cm ² disc/mL methanol, equivalent to 0.42 g disc/mL. Methanol was used because it is a suggested extraction solvent for generation of medical device samples for testing in biological systems and because our experience indicates that methanol is a highly aggressive solvent for BisGMA-based resins.

The methanol was spiked with 30 ng/mL d16 BPA (Cambridge Isotope Laboratories, Tewksbury, MA, USA) as an internal standard (IS). The purity of d16 BPA was listed as 98% d16 by the manufacturer. Extraction was conducted at 37 °C with mild agitation (100 rpm) for 24 h. A method blank consisting of methanol alone was included with all extraction steps and subsequent analyses.

2.3

BPA quantitation

BPA separation and quantitation were carried out on an Agilent 1200 SL with UV diode ray detection (DAD) (Agilent Technologies, Santa Clara, CA, USA) coupled to an AB/Sciex Qtrap 3200 Liquid Chromatograph with tandem mass spectrometry capability (AB/Sciex LLC, Framingham, MA, USA). Separation was performed with an Agilent Eclipse XDB-C18 column (Agilent Technologies, Santa Clara, CA, USA) with 2.1 × 150 mm dimensions and 3.5 micron packing. All chromatography solvents were HPLC grade from EMD Millipore, Billerica, MA, USA. The mobile phase was a mixture of unbuffered water (channel A) and methanol (channel B). The gradient was applied as follows: 20% B (preinjection reequilibration, 5 min); 70% B (0−0.5 min); 75% B (0.5−6 min); 99% B (6.1 min followed by a 4 min hold). The flow rate was 0.5 mL/min and the injection volume was 5 μL. Column effluent was split between the diode array detector and the Qtrap mass spectrometer. A wavelength of 225 nm was used to plot the UV response, but full DAD spectra were acquired during all LC/MS experiments. The Qtrap was operated in standard triple quadrupole multiple reaction monitoring mode. The Turbo IonSpray source of the Qtrap was optimized for BPA detection with ESI voltage of -4500 V, nebulizer gas of 80 psi, desolvation gas pressure and temperature of 70 psi and 700 °C, respectively, and counter-current gas of 30 psi. A declustering potential of -50 V was optimized for generation of the deprotonated BPA molecular ion, and the entrance potential and collision cell entrance potential were optimized at -10 and -20 V, respectively. Two fragment ions, one involving loss of methyl radical (m/z 212) and one involving loss of phenol (m/z 133) were used to monitor BPA. The 227/212 transition was optimized using a collision energy of -25 V and collision exit potential of -2.0 V, while the 227/133 transition employed a collision energy of -36 V and a collision exit potential of -1.0 V. Each of the three transitions monitored during the entire chromatographic run was assigned a 100 ms dwell time.

The same d16 BPA used in the resin extraction was employed as an IS in all cases to correct for any losses during sample preparation or variability injection volume or detection efficiency. The d16 IS was monitored using a 241/223 transition involving loss if d3 methyl radical was used and the same tuning employed for the analogous transition if d0 BPA was used. Tests of the d16 IS in our LC/MS/MS assay showed that it gave no signal at d0, and as such did not contribute to the BPA signal during the assay. The level of detection was 3 ppb BPA.

2.4

Identification of co-eluting chemical(s)

Identification of co-eluting chemicals was carried out on an Agilent 1260 HPLC coupled to an Agilent 6224 Time of Flight (TOF) LC/MS (Agilent Technologies, Santa Clara, CA, USA) with Dual ESI source operated in positive and negative modes. Chromatography was performed as for the initial BPA analyses except that some samples were also run using water and acetonitrile (both containing 6 mM ammonium acetate to aid ionization), to enhance detection of less polar analytes in positive mode for the TOF analyses. The dual ESI was set at nebulizer gas pressure of 45 psi, desolvation gas flow and temperature of 11 lpm and 325 °C respectively, capillary voltage of 400 V, skimmer voltage of 65 V, fragmenter voltage of 155 V, and octopole radiofrequency of 750 V. All scans were automatically recalibrated with the standard Agilent Reference Mixture using 121.0509 and 922.0098 ions for recalibration. The full scan mass range employed was 105−1700 atomic mass units (amu) at 1 scan per second.