Supporting Information High-Sensitivity Elemental of Fluorine by Ionization in Afterglow

Joseph E. Lesniewski†, Kunyu Zheng†, Paolo Lecchi‡, David Dain‡, and Kaveh Jorabchi†* † Department of , Georgetown University, Washington, DC ‡ DSM Nutritional Products, Columbia, MD * Corresponding author: [email protected]

Table of contents

 Experimental details o Operating parameters for PARCI-MS and chromatography-PARCI-MS o Sample preparation and data analysis  Details for ab initio calculations  Background spectrum of PARCI-MS with aqueous sodium acetate  Flow injections of NaF and p-F-phenylalanine  Flow injections for F LOD calculations  Calibration curve for fluoride detection with IC-PARCI-MS

Experimental Details

Sample introduction and PARCI. Aqueous analytes were introduced by flow injection using a 20 μL injection loop at 250 μL/min water supplied by an HPLC pump (LC10AD, Shimadzu, Columbia, MD). The analyte stream was then mixed with 80 μL/min 10 mM aqueous NaOH solution supplied by a syringe pump (Model 100, KD Scientific, Holliston, MA). The mixed stream was then guided into a nebulizer (HEN-120, Meinhard, Golden, CO) operated at a constant argon flow rate of 1.25 L/min. The aerosols passed through a cyclonic spray chamber (Twister Cyclonic, Glass Expansion, Pocasset, MA) and were mixed with a makeup gas flow of 0.55-1.25 L/min using a tangential mixer (ML151008N, Meinhard, Golden, CO) to provide total aerosol gas flow rates of 1.8-2.5 L/min. The total aerosol gas flow was injected via a 1.5-mm injector into an Ar ICP sustained at 1100 W (14 L/min outer gas flow, 1.2 L/min auxiliary flow) using a stand-alone RF generator (Nexion 2000, PerkinElmer Inc., Waltham, MA) with the exhaust flow adjusted to produce 3.5 m/s air flow at the bottom intake of the torch box. Note that exhaust flow has a significant impact in PARCI unlike conventional ICP- MS.1 Analytes underwent vaporization and breakdown within the plasma. The plasma products were then guided through a cooled aluminum sampler (2.5 mm orifice) into a stainless steel atmospheric-pressure reaction tube (5 cm long, ¼” o.d., 4.8 mm i.d.) where ion-molecular reactions led to ionization of plasma products.

S-1

Mass Spectrometer. The mass spectrometer used for detection of was a single quadrupole MS (SQ 300, PerkinElmer Inc., Waltham, MA). The electrospray was removed from the instrument and the PARCI source was placed so that the reaction tube was aligned with the inlet, 5-10 mm from the MS end plate. Neutrals were restricted from entering the mass spectrometer by a 5.5 L/min N2 counterflow gas controlled by an external mass flow controller (Model 246B, MKS, Andover, MA). Positive ions were pulled in to the MS inlet by applying potentials of -200 V to the end plate and -600 V to the glass capillary entrance. Background scans were collected using 1000 μs pulse counting time per point, 10 points per mass, and were averaged for 30 seconds. A nozzle-skimmer declustering potential of 50 V was used for all + acquisitions (+70 V capillary exit, +20 V skimmer). Na2F sensitivity measurements were conducted in selected ion monitoring (SIM) mode at an m/z value of 65 with an integration time of 500 ms. For measurements of LOD, an integration time of 1000 ms was utilized. To minimize premature aging of the detector due to the high background ion flux, an intentional ion deflection was applied to reduce the ion current impinging on the detector. This was accomplished by setting the Q1 offset parameter to zero (normally it is set to -2 V). The signal suppression factor was measured using flow injections with sequential suppressed and unsuppressed ion flux settings. The ion intensities measured at suppressed flux conditions were corrected using the measured factor for all of the values reported in this work.

Sample Preparation and Data Analysis. All solvents were LC-MS grade (Chromasolv, Sigma- Aldrich, St. Louis, MO). p-fluoro-phenylalanine (p-F-phe) was purchased from Sigma-Aldrich, St. Louis, MO and sodium fluoride was purchased from the JT Baker Chemical Company, Phillipsburg, NJ. Analyte solutions were prepared by dissolving solids in 2-mM NaOH in water at concentrations of ~100 mM for NaF and ~15 mM for p-F-phe followed by dilution in water to final concentrations of 6-500 μM fluorine in water. ACS grade sodium hydroxide (VWR, Radnor, PA) and sodium acetate (Fisher, Waltham, MA) were used for post-column sodium addition. Sodium salt solutions were prepared by dissolving the solids in water at a concentration of ~2 M, followed by dilution to 10 mM in water. All selected ion chromatograms from flow injections were exported as text files, and then manually integrated using the trapezoid rule by selecting the beginning and end points of a baseline for each peak. Peak heights were calculated as the difference between the top of the peak and the baseline.

Ab initio calculations. Gaussian 162 was used to calculate reaction thermochemistries at a theory level of ωB97xD/aug-cc-pVTZ at 600 K (the approximate temperature at the end of the reaction tube). This level of theory has previously been used to describe the thermochemical properties of halide-neutral adducts.3 Structures were optimized using the MMFF94 force field in the Avogadro program4 to generate starting geometries for each species prior to optimization at the final theory level. For molecules where more than one conformer is possible, the lowest energy conformer was used.

Ion Chromatography. For measurements of F- in infant formula, an ion chromatography unit (ICS- 2000 pump, 300-uL injection loop, an IonPac AG11-HC 4x50 mm guard column, and IonPac AS-11HC 4x250 mm anion exchange column, Thermo Fisher Scientific, Waltham, MA) with NaOH gradient elution was utilized. Separations occurred at a flow rate of 1.1 mL/min and the eluent was split after the conductivity detector with 250 uL/min mixed with 80 uL/min 10 mM NaOH prior to introduction into PARCI-MS. This arrangement allowed simultaneous conductivity and PARCI-MS detection. Use of NaOH rather than other sodium salts was preferred to increase the pH of the solution after the column and to prevent reactions and loss of F- in transfer to the plasma. PARCI-MS was operated at a total aerosol carrier gas flow rate of 2.2 L/min (1.25 L/min nebulizer, 0.95 L/min auxiliary) while other parameters were kept constant as described above. Fluoride calibration curve was constructed using injections of NaF standard in 18 MΩ water from 6.7 to 122 µM of fluoride using a 5 mM NaOH isocratic elution. For infant formula analysis, 1.5 g of powder sample (Enfamil, Mead Johnson & Company, LLC, Evansville, IN) was dissolved in 12 ml of 18 MΩ water. The sample was subjected to centrifugal filtration using 3000 Da filters (Amicon Ultra-15, Millipore Sigma, Burlington, MA) to separate the proteins and

S-2 large fats. Spiked infant formula samples were prepared by adding 15 and 52 µM fluoride to 10 ml of infant formula sample using a NaF standard solution prior to filtration. The filtrate was directly injected into ion chromatography without any further sample treatment. Separation of fluoride in infant formula was achieved via a multi-step NaOH gradient starting at 0.2 mM for 4 min followed by 5 mM for 16 min. The eluent from the IC was connected to PARCI only between elution times of 5 and 10 min for infant formula samples to avoid introduction of unretained and strongly retained matrix into PARCI. The column was cleaned using 80 mM NaOH for 10 min followed by reconditioning at 0.2 mM NaOH for 10 min prior to each injection.

Figure S1. Background spectrum of PARCI-MS using aqueous 10 mM sodium acetate for post-flow injection (post-column) sodium addition.

S-3

Figure S2. Flow injections of 50 µM NaF (first triplicate) and 50 µM p-F-phe (second triplicate) lead to + formation of Na2F in PARCI-MS. Note that difference in peak heights for the two compounds is a result of differences in flow injection peak shapes.

Figure S3. Ten replicate flow injections of 10 µM NaF used for estimation of limit of detection.

S-4

Figure S4. Calibration curve for IC-PARCI-MS using NaF standards in water with a linear least-squares fit.

References (1) Lesniewski, J. E.; McMahon, W. P.; Jorabchi, K. J Anal Atom Spectrom 2018, 33, 1981-1992. (2) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Petersson, G. A.; Nakatsuji, H.; Li, X.; Caricato, M.; Marenich, A. V.; Bloino, J.; Janesko, B. G.; Gomperts, R.; Mennucci, B.; Hratchian, H. P.; Ortiz, J. V.; Izmaylov, A. F.; Sonnenberg, J. L.; Williams; Ding, F.; Lipparini, F.; Egidi, F.; Goings, J.; Peng, B.; Petrone, A.; Henderson, T.; Ranasinghe, D.; Zakrzewski, V. G.; Gao, J.; Rega, N.; Zheng, G.; Liang, W.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Throssell, K.; Montgomery Jr., J. A.; Peralta, J. E.; Ogliaro, F.; Bearpark, M. J.; Heyd, J. J.; Brothers, E. N.; Kudin, K. N.; Staroverov, V. N.; Keith, T. A.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A. P.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Millam, J. M.; Klene, M.; Adamo, C.; Cammi, R.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Farkas, O.; Foresman, J. B.; Fox, D. J.: Wallingford, CT, 2016. (3) Hyttinen, N.; Otkjaer, R. V.; Iyer, S.; Kjaergaard, H. G.; Rissanen, M. P.; Wennberg, P. O.; Kurten, T. J Phys Chem A 2018, 122, 269-279. (4) Hanwell, M. D.; Curtis, D. E.; Lonie, D. C.; Vandermeersch, T.; Zurek, E.; Hutchison, G. R. J Cheminformatics 2012, 4.

S-5