Lanthanoid analysis in seawater by seaFAST-SP3™ system in off-line mode and magnetic sector high-resolution inductively coupled plasma source mass spectrometer

Highlights • Lanthanoids as geochemical tracers in seawater.• A 4-step strict protocol and state-of-the-art technology for lanthanoids analyses in seawaters.• Sample pre-concentration system for matrix separation for the detection of ultra-low lanthanoids levels.

Instrumentation A commercially available seaFAST-SP3 TM system (Elemental Scientific Inc., Nebraska, USA) for matrix separation and analyte pre-concentration is used in the present method. The seaFAST-SP3 TM is an ultra-clean, automated, low-pressure ion chromatography system used for undiluted seawater pre-treatment. It consists of an autosampler, a sample loop with defined volume, one pre-packed cleaning column, one pre-concentration column filled with chelating resin (hydrophilic methacrylate polymer), three 12-port valves, and four syringes. The system is capable of single-digit picogram L − 1 detection limits. The 4 DX Autosampler with Dual Flowing Rinse is an ESI SC system autosampler (Elemental Scientific Inc., USA) for the complete automated sample introduction to HR-ICP-MS. This integrated autosampler increases productive instrument time by reducing sample uptake, stabilization, and rinse steps. Blank contamination at the rinse station is minimized using a gravity-fed or pressurized rinsing system, utilizing two metal-free valves to control the dual-flowing rinse solutions. Thermo Scientific TM Element XR TM High-Resolution ICP-MS (ThermoFisher Scientific Inc., Bremen, Germany) is an ultra-sensitive instrument. It is reliable multi-element analyses at trace-level concentrations, with an unequivocal separation of analyte ions from spectral interferences for the highest level of confidence. Element XR combine a dual-mode secondary electron multiplier (SEM) with a Faraday detector that increases the linear dynamic range by twelve orders of magnitude (from the of 0.2 cps to ∼5 × 10 12 cps equivalents to sub ppq to percentages). To reduce the risk of contamination, all work on the water samples, blanks, and calibration standards was carried out in a Class 100 HEPA fume hood and the ICP-MS instrument is in a clean room with 10 0 0 HEPA. Material and Reagents All lab materials are cleaned for trace metal analysis by submerging into 2 N HCl solution for 24 h, triple rinsing with MilliQ water, submerging into 2 N HNO 3 solution for 24 h, and then triple rinse with MilliQ water. All reagents used in the present study are for trace analysis grade. Ultrapure HCl and HNO 3 (J. T. Baker®, Avantor Performance Materials, LLC, Center Valley, PA., USA). High purity 21% NH 4 OH, 99.5% CH 3 COOH, and NaCl (Merck, Darmstadt, Germany). Linear calibration curves were obtained of certified standard solutions of 10 mg L − 1 lanthanoids in 2% HNO 3 (High-Purity TM Standards, Charleston, SC, USA). High-quality deionized water from the MilliQ Advantage A10 system (Millipore Corp., USA) was used in this work.

Method details
Step 1: Samples preparation Seawater samples (1 L) are filtered through a 0.45 μm filter (Whatman® membrane filters nylon). Filters are replaced with plastic tweezers before a new sample is taken. For each filter sample, 1.5 ml of HNO 3 are added to obtain pH < 2, then stored into acid-cleaned 1 L Teflon bottles within 24 h after To ensure the release of the organic species of the metal into the water and contribute to its oxidation and conversion to an inorganic form, the acidified samples and blanks are stored for a month.
Step 2: Samples pre-concentration Different pre-concentration and matrix elimination methods have been employed to determine lanthanoids in seawater [3 , 6 , 11] . The seaFAST-SP3 TM is a low-pressure ion chromatography system that removes the matrix and retains the lanthanoids released in low elution volumes with high concentration factors ( ∼20 fold). Aliquots of filtered and acidified samples are taken and transferred to 50 ml polypropylene vials for the pre-concentration step using the seaFAST-SP3 TM system for matrix separation by off-line configuration. The matrix is removed and lanthanoids chelated. A schematization and brief description of the seawater pre-concentration procedure by the seaFAST-SP3 TM system is provided in Fig. 1 . Samples and blanks were pre-concentrated by triplicate. Instrumental operation conditions for the seawater pre-treatment seaFAST-SP3 TM are shown in Table 1 .

Step 3: Off-line injection of the eluted sample
Sample's introduction into the Thermo Scientific TM Element XR TM High-Resolution ICP-MS is performed in off-line mode with an ESI SC autosampler integrated system (4 DX Autosampler with Dual Flowing Rinse). The eluted sample is introduced into the instrument through a micro-Flow PFA nebulizer (200 μL min −1 ) into the Peltier PC 3 (Elemental Scientific Inc., USA) cooled quartz impact bead spray chamber ( Fig. 1 ). This chamber acts as a collision-reaction cell to remove interferences that degrade the detection limits. Then, the sample is introduced into an argon plasma as aerosol droplets (nebulized).

Step 4: Lanthanoids analysis
The high-resolution inductively coupled plasma source mass spectrometer (HR-ICP-MS) is one of the best techniques to determine lanthanoids (La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Tb, Er, Tm, Yb, and Lu) in seawater [6] . In particular, the Thermo Scientific TM Element XR TM is a Nier-Johnson inverse geometry (e.g., focused on energy and mass/charge) dual-focus sector magnetic ICP-MS instrument that is widely used in multi-elemental determinations (e.g., lanthanoids) in complex sample matrices (e.g., seawater). A schematization and description of the lanthanoids analyses procedure by HR-ICP-MS is provided in Fig. 2 . All measurements were performed in off-line mode with Thermo Scientific TM Element XR TM with an ESI SC autosampler integrated system. The measurement was conducted as 'Triple' detector mode, automatic switching using the Element software. Detailed instrumental conditions and the operating parameters for optimizing and ion lens voltages to achieve high and     reproducible signals are provided in Tables 2 and 3 . The instrument conditions were checked daily and adjusted for optimum sensitivity using a tune solution of 1 μg L −1 ( 6 Li, 115 In, 238 U). External calibration solutions of lanthanoids were prepared daily by increasing additions of a multi-element stock standard solution (10 mg L −1 lanthanoids in 2% HNO 3 , High-Purity TM Standard). External standardization was applied using a six-point calibration (1, 5, 10, 50, 100, and 10 0 0 ng L −1 , Fig. 3 ). Calibration solutions were processed through the SeaFAST-SP3 TM , such as the samples. Blank solutions of 2% HNO 3 and NaCl matrix in 2% HNO 3 were used as an instrumental blank, measured before each sample and standard. An internal 115 In standard solution of 100 μg L −1 was prepared weekly by diluting single-element stock standard solutions. Calibration standards, reference seawater aliquots, and samples were doped with 115 In (0.1 μg L −1 ) to monitor and correct instrumental drift. 115 In correction was used when the instrumental drift was > 5%. A Careful rinse (with 2% HNO 3 ) followed by an instrumental background check was performed before every sample measurement to monitor sample to sample memory effects and correct for them, if necessary. Because of the low level of oxides formation, it was not necessary to apply mathematical correction accounting for oxide formation (e.g., UO < 2.5%). The software compares the intensities of the measured pulses to those from standards, which make up the external calibration curve, to determine the concentration of each element. Lanthanoid concentrations were calculated using the slopes of the standard external curves (R 2 > 0.999, Fig. 3 ).

Method validation
To check the validity of this method, we analyzed estuarine water reference samples (SLEW-3, National Research Council of Canada). Although certified values for lanthanoids are not available for these certified reference materials (CRM), we compared our results with other studies ( Table 4 ; [1 , 2 , 10 , 12] ). The recovery based on the SLEW-3, averaged from 78.6% for Pr to 106% for Ce. Because UV oxidation was not performed in the pre-treatment of the acidified samples, to oxidize natural organic ligands and to dissociate lanthanoids of ligands [5] , a low recovery for several elements can be explained by the presence of complex organic in seawater. Also, the external accuracy was assessed via repeat analyses of an artificial seawater sample (NaCl matrix, 35 g L −1 ). Column recovery was evaluated using 100 ng L −1 multi-element lanthanoid standard solutions, and the artificial sample means recoveries ranged from 95 to 100% ( Fig. 4 ). Exceptions were observed for Ce and Lu (recovery 105-110%) and Pr and Dy (recovery 87-90%). The coefficient of variation (CV, %) for every lanthanoid in samples was ≤10% (10 repeated analyses of the artificial seawater sample), except for Gd, Tb and Yb (11-13.75%) ( Table 3 ). The blanks varied between 0.01 and 0.07 ng L −1 . Blanks represent < 5%  of the SLEW-values and < 1% in the synthetic seawater. The procedural detection limit, which was calculated three times the standard deviation of 10 blank measurements, varied from 0.01 to 0.03 ng L −1 ( Table 4 ).

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have influenced the work reported in this paper.