Catalysis-Based Fluorometric Method for Semiquantifying Trace Palladium in Sulfur-Containing Compounds and Ibuprofen

Trace palladium in synthetic materials can be rapidly and inexpensively semiquantified by a catalysis-based fluorometric method that converts resorufin allyl ether to resorufin. However, whether sulfur compounds would interfere with this method has not been systematically studied. Herein, we show that although thiourea in solution interferes with quantification, sulfide, thiol, and thiocarbamate do not. The fluorometric method can also detect palladium bound to sulfur-based scavenger resin and outperform inductively coupled plasma mass spectrometry for detecting trace palladium in ibuprofen.

A solution of 11.65 M trace metal grade HCl (16.1 mL) was diluted in water (283.9mL) in an amber bottle, and the resulting solution was stored at 24 °C.

Preparation of 500 mM HCl in 1:4 v/v DMSO/water (Solution A)
A solution of 625 mM aqueous HCl (100 mL) was diluted in DMSO (25 mL) in an amber bottle, and the resulting solution was stored at 24 °C.Preparation of 2.0 mM palladium standard solution in 10% w/v HCl A solution of 34-37% trace metal grade HCl (281.7 μL) was diluted to 10% w/v HCl in water (718.3μL).A solution of 1000 ppm (9.4 mM) palladium standard solution in 10% w/v HCl (425 µL) was diluted with 10% w/v HCl (1.57mL) to the final concentration of 2.0 mM palladium.The resulting solution was stored at 24 °C.

General procedure for calibration curve
Generating calibration curve using palladium standard solution A 2.0 mM palladium standard solution in 10% trace metal grade HCl (100 L) was diluted with Solution A (900 L) to prepare 200 and 20 M palladium.The 20 M palladium solution (1.00 mL) was then transferred to a deep 96-well plate.Two-fold serial dilutions using Solution A produced 10000, 5000, 2500, 1250, 625, 313, 156, 78.1, 39.1, and 19.6 nM palladium solutions.The final concentrations of palladium in the assay wells were 62.5, 31.3,15.6, 7.81, 3.91, 1.96, and 0 nM.
Procedure for detecting resin-bound palladium (Figure 3) The palladium-loaded thiol-based resin (6.8 mg) or palladium-loaded thiourea-based resin (4.3 mg) was suspended in EtOH (2.00 mL) in a 5-mL conical tube.This solution was treated with the solution of 240 μM RAE (1.00 mL) and the solution of 80 mM NaBH 4 and 720 μM TFP (1 mL) or the solution of 0 mM NaBH 4 and 720 μM TFP (1 mL).Fractions (200 μL) of these resulting solutions were transferred to a black 96-well plate in one replicate to measure the fluorescence values immediately after the addition of the NaBH 4 -TFP solution (0 min), after incubating at room temperature for 0.5, 1, and 18 h.
Procedure for comparing catalytic activity of tBuXPhos-Pd-G3, PEPPSI TM -IPr, and palladium standard solution (Figure 4) Preparation of solutions with palladium pre-catalysts PEPPSI TM -IPr catalyst (18.7 mg) and tBuXPhos-Pd-G3 (12.4 mg) were each dissolved in DMSO (13.8 mL and 7.81 mL, respectively) to the final concentration of 2.00 mM palladium.A 2-fold serial dilution was performed on each 2.00 mM palladium solution in 1:1 v/v DMSO/water to obtain 1000 and 500 M palladium solutions.Ten-fold serial dilutions on the 500 M palladium solutions produced 50 M, 5 M, 500 nM, and 50 nM palladium solutions.The solution of 1:1 v/v DMSO/water was used as 0 nM palladium solution in the assay.

Preparation of solutions with palladium standard solution
A 2-fold serial dilution was performed on 625 mM aqueous HCl in 1:1 v/v DMSO/water to obtain 313 mM HCl.This solution was used to dilute 2.0 mM palladium standard solution to 1000, 500, and 5 M, 500 and 50 nM palladium.The solution of 313 mM HCl was used as a 0 nM palladium standard solution in the assay.Fluorescence values were measured on the microplate reader immediately after the addition of the NaBH 4 -TFP solution (0 min) and after incubating at room temperature away from light for 15 min.

Procedure for quantifying palladium in ibuprofen by standard addition (Table 2)
Preparation of 80, 20, 5, and 0.2 ppb palladium in ibuprofen In a 2-dram vial, ibuprofen (200 mg) was dissolved in MeOH (500 µL).The API sample was spiked to contain 25 ppm palladium by addition of 500 g/mL tBuXPhos-Pd-G3 solution in DMSO (74.6 L) to the ibuprofen in MeOH (500 L).After a 1 h incubation, the spiked ibuprofen solution was placed on a rotary evaporator to remove MeOH and then lyophilized overnight.This sample (80 mg) was dissolved in DMSO (2 mL) using a volumetric flask to obtain 9400 nM palladium in DMSO (1 g/mL palladium).The diluent was prepared by dissolving ibuprofen (1.000 g) in DMSO (25 mL) to obtain 40 mg/mL ibuprofen.The 9400 nM palladium solution (1.00 mL) was then diluted in 40 mg/mL ibuprofen (1.50 mL) to prepare 3760 nM palladium.This solution was then diluted five-fold three times to obtain 30.1 nM palladium.This solution was then diluted four-fold twice to obtain 7.525 and 1.88 nM palladium.The 1.88 nM solution was diluted five-fold twice to obtain 0.0752 nM palladium.These solutions correspond to 80, 20, 5, and 0.2 ppb (ng palladium/g ibuprofen) in solid state respectively.

Synthesis of 2 nm palladium nanoparticles (Pd NPs)
2-nm Pd NPs were synthesized using a procedure described by Zou et al. 2 Briefly, K 2 PdCl 4 (0.4 mmol) was dissolved in water (12 mL) in a 20-mL scintillation vial.In a separate vial, tetraoctylammonium bromide (TOAB) (2 mmol) was dissolved in toluene (25 mL).The clear TOAB solution was then layered on top of the aqueous K 2 PdCl 4 solution.The biphasic mixture was vortexed until the organic layer became red and the aqueous layer became clear.The organic phase containing the (TOA)PdCl 4 complex was removed and transferred to a 250-mL round-bottom flask equipped with a stir bar, followed by the addition of oleylamine (OAm) (1.264 mL).NaBH 4 (8 mmol) dissolved in water (2 mL) was then added with vigorous stirring and allowed to sit for 1 h before the particles were precipitated with EtOH and isolated via centrifugation, followed by resuspension in CHCl 3 .

Synthesis of 5 nm Pd NPs
5-nm Pd NPs were synthesized using a procedure described by Mazumder et al. 3 Briefly, Pd(acac) 2 (0.1 mmol) was dissolved in OAm (5 mL) and heated to 70 °C under argon.At this temperature, a separate, room temperature solution of borane tert-butylamine complex (100 mg) in OAm (2.5 mL) was injected into the heated reaction flask.The reaction was then heated to 90 °C and held at 90 °C for 1 h.After cooling to room temperature, particles were precipitated with EtOH and isolated via centrifugation, followed by resuspension in CHCl 3 .

Synthesis of 25nm Pd NPs
25-nm Pd NPs were synthesized using a method developed by the Millstone group.The synthetic procedure was completed using standard air-free techniques.Pd(acac) 2 (0.1 mmol) was dissolved in OAm (1 mL).
Separately, OAm (9 mL) and HDD (1 mmol) were added to a round bottom flask equipped with a stir bar and condenser.This mixture was then degassed at 100 °C for 1 h.The solution was then heated to 250 °C under argon, at which point the solution of Pd(acac) 2 was injected and held at 250 °C for 3 h.After cooling to room temperature, particles were precipitated with EtOH and isolated via centrifugation, followed by resuspension in CHCl 3 .

Ligand Exchange and Aqueous Phase Transfer of Pd NPs with 1kDa PEGSH
Pd NPs of each size were phase transferred into water by ligand exchange with 1kDa PEGSH.In all cases, purified OAm capped Pd NPs were dispersed in CHCl 3 (5 mL).A solution of 20 mM 1kDa PEGSH in CHCl 3 (5 mL) was then added to the NP solutions, and the mixture was left stirring for 24 h.After incubation, the PEGSH-functionalized Pd NPs were precipitated with the addition of hexanes and isolated via centrifugation.The supernatant was removed, and the resulting pellet was resuspended in water (5 mL).The Pd content of these solutions was then measured by inductively coupled plasma-optical emission spectrometry (ICP-OES) and diluted with water to obtain 1 mM stock solutions.

Inductively Coupled Plasma Optical Emission Spectrometry (ICP-OES)
ICP-OES analysis was performed using an argon flow with a PerkinElmer, Inc. Optima spectrometer (Department of Chemistry, University of Pittsburgh).An ultrapure aqua regia solution was prepared with a 3:1 ratio of hydrochloric acid (Sigma-Aldrich, > 99.999% trace metal basis) and nitric acid (Sigma-Aldrich, > 99.999% trace metal basis) and diluted with water for a 5% v/v aqua regia matrix.Unknown Pd concentrations were determined by comparison to a 7-point standard curve with a range of 0.10-10 ppm of Pd (0.10, 0.50, 1.0, 2.5, 5.0, 7.5, and 10 ppm) prepared by volume using a Pd standard for ICP (Fluka, TraceCERT 1000 ± 2 mg/L Pd in HCl), diluted in a 5% aqua regia matrix.All standards and unknown samples were measured 3 times and averaged.

High-Resolution Transmission Electron Microscopy (HRTEM)
Pd NP samples were prepared for TEM by drop-casting an aliquot of the purified solution in water onto carbon film-coated copper TEM grids (Ted Pella, Inc., Redding, CA) for bright field imaging.TEM characterization for all Pd NPs was performed on a Hitachi H9500 Environmental TEM with an accelerating voltage of 300 kV (Nanoscale Fabrication and Characterization Facility, Petersen Institute of Nanoscience and Engineering, Pittsburgh, PA).The size distributions of the NPs were determined by measuring 250 NPs from various areas of the grid using ImageJ 1.53k (National Institutes of Health, USA) (Figure S4) Procedure for investigating compatibility with palladium nanoparticles (Figure 5) Preparation of nanoparticle samples Palladium nanoparticles of 2 nm, 5 nm, and 25 nm in water (1 mM palladium determined by ICP-OES) were diluted to 500 µM palladium using water or 20% aqua regia.Aqua regia was prepared by mixing a 3:1 ratio of 34-37% trace metal grade HCl and 67-70% trace metal grade nitric acid and diluted with water to make 20% aqua regia.The 500 µM palladium solutions of Pd NPs were subsequently diluted using Solution A to make 50 µM, 5 µM, 1 µM, 500 nM, and 250 nM palladium solutions.As a negative control, water (1 mL) was also subjected to the addition of 5 or 20% aqua regia (1 mL) and subsequent dilution with Solution A. As a positive control, the 2.0 mM palladium standard solution in 10% trace metal grade HCl was diluted with water to make a 1.0 mM palladium standard solution.This solution was diluted with water, 5% aqua regia, or 20% aqua regia to make a 500 µM palladium solution.These standard solutions were subsequently diluted using Solution A to make 50 µM, 5 µM, and 500 nM palladium solutions (Figure S5).

Detecting palladium in spiked API solutions (Table 1)
Table S3.Raw fluorescence values used to generate calibration curves in Figure S1.Calibration curve solutions were made fresh for each day of testing.The remaining fluorescence intensity of the 0 nM palladium calibration solution was also subtracted from calibration solutions, forcing the calibration curve through the origin to simplify analysis and account for experimental variation.Table S7.Raw fluorescence values for penicillin tested 2 and 4 days post spike.Data processing is described in Table S3.Detecting resin-bound palladium (Figure 3) Table S11.Raw fluorescence values for detecting resin-bound palladium with or without the addition of NaBH4.

Figure S3 .Figure S4 .
Figure S3.Standard addition curves for palladium-spiked ibuprofen samples: 80 (a), 20, 5, and 0.2 (b) ppb.Each linearregression line was extrapolated to the x-intercept to calculate experimental [palladium].Reported measurements are expressed as the average of these values.

Table S1 .
Sample calculation of ppm in solid state.Calculated from API and palladium concentrations in solution state.

Table S2 :
Demonstrating the ppm corresponding to nM palladium in solution.
A solution of 85.7 M RAE/286 mM NH 4 OAc in 1.7:98.3v/v DMSO/EtOH (140 L) was transferred to wells in a black 96-well plate.Solutions of 1, 0.33, and 0.11 mg/mL API with varying palladium concentrations (20 µL) were subsequently added.To generate a calibration curve, solutions of 625-0 nM palladium (20 µL) were added to separate wells to generate a calibration curve.Finally, the solution of 900 M TFP/100 mM NaBH 4 /375 mM NaOH in 3:22:25 v/v/v DMSO/EtOH/water (40 L) was added to all wells.Fluorescence values were measured immediately after the addition of the NaBH 4 -TFP solution (0 min) and after incubating at room temperature away from light for 15 min.Calibration curves were generated each day.Data for calibration curves shown in Tables AssayFinal conditions: 60 μM RAE, 180 μM TFP, 202 mM NH 4 OAc and 20 or 0 mM NaBH 4 in 2.4:78.6:10v/v/v DMSO/EtOH/water.
4:18:77.6 v/v/v DMSO/water/EtOH.The final well volume was 200 L tested in three replicates.A solution of 85.7 μM RAE/286 mM NH 4 OAc in 2.6/148.4v/v DMSO/EtOH (140 μL) was transferred to wells in a black, 96-well plate.Solutions of 2-nm, 5-nm, and 25-nm Pd NPs of 1 µM, 500 nM, and 250 nM palladium diluted with water or 20% aqua regia (20 µL) were transferred to wells.Standard solutions of 0 or 500 nM palladium diluted with water, 5% aqua regia, or 20% aqua regia (20 µL) were transferred to wells.A multi-channel pipette was used to transfer 900 μM TFP/375 mM NaOH in 3:22:25 v/v/v DMSO/EtOH/water (40 μL) to all wells.Fluorescence values were measured on the microplate reader immediately after the addition of the NaBH 4 -TFPsolution (0 min) and after incubating at room temperature away from light for 15 min.

Table S4 .
Calibration curves generated from TableS3.A simple linear regression was performed to generate respective equations.Equation was used to determine expected fluorescence values of APIs (biotin, N-acetyl-L-cysteine, thiourea, and thiocarbamate) 2 (a) and 4 (b) days after spiking.Raw fluorescence values used to generate calibration curves in FigureS2.Penicillin was tested 2 days and 4 days post spiking, and calibration curve solutions were made fresh for each day of testing.

Table S8 .
Sample signal recovery calculation used to generate values in Table 1, S9, and S10.

Table S9 .
Expected and experimental fluorescence values and % signal recovery for biotin, N-acetyl-L-cysteine, thiourea, and thiocarbamate tested 2 and 4 days post spiking.Values are mean ± SD, tested in three replicates.Fluorescence signals are arbitrary units (AU).FI: Fluorescence Intensity

Table S10 .
Expected and experimental fluorescence values and % signal recovery for penicillin tested 2 and 4 days post spiking.

Table S14a .
Raw fluorescence values for quantification of palladium by method of standard addition for trials 1-3 for 80 ppb sample and trails 1 and 2 for 20, 5, and 0.2 ppb samples of solid-state palladium in ibuprofen.Fluorescence intensities at 0 min following addition of NaBH4 were subtracted from fluorescence intensities following incubation for 30 min.

Table S14b .
Raw fluorescence values for quantification of palladium by method of standard addition for trials 4 and 5 for 80 ppb sample and trials 3 and 4 for 20, 5, and 0.2 ppb samples of solid-state palladium in ibuprofen.Fluorescence intensities of wells without NaBH4 (reaction blanks) were subtracted from fluorescence intensities of wells containing NaBH4 following incubation for 30 min.

Table S15 .
Linear-regression equations and extrapolated x-intercepts used to back-calculate palladium content in various spiked samples.Reported measurements are expressed as the average of these values.

Table S16 .
Raw fluorescence values for palladium standard solutions and 0 nM Pd with and without acid.Figure illustrating acid digestion with 5% and 20% aqua regia does not affect assay and fluorescence output.Fluorescence intensities measured immediately after NaBH4 addition (t0) were subtracted from intensities measured after 15 minutes.Remaining fluorescence of 0 nM palladium samples were averaged and subtracted from 50 nM palladium samples.

Table S17a .
Raw fluorescence values for palladium nanoparticles (2 nm, 5 nm, and 25 nm) without acid digestion.Palladium concentration is 50 nM based on dilution from 1.0 mM palladium in Pd NPs determined by ICP-OES.

Table S17b .
Raw fluorescence values for palladium nanoparticles (2 nm, 5 nm, and 25 nm) acid digested with 20% aqua regia.Palladium concentration is 50 nM based on dilution from 1.0 mM palladium in Pd NPs determined by ICP-OES.

Table S17c .
Raw fluorescence values for 0 nM palladium control.Remaining fluorescence of the 0 nM palladium solution was averaged and subtracted from values in Tables S17a and S17b so that intercept of Figure5is set at the origin and analysis is consistent with Table1and FigureS5.