Separation and measurement of Pa, Th and U isotopes in marine sediments by microwave-assisted digestion and MC ICP-MS

Separation and measurement of Pa, Th and U isotopes in marine sediments by microwave-assisted digestion and MCICP-MS Journal: Analytical Chemistry Manuscript ID: ac-2008-02341y.R1 Manuscript Type: Article Date Submitted by the Author: n/a Complete List of Authors: Negre, Cesar; Universitat Autonoma de Barcelona, Institut de Ciencia i Tecnologia Ambientals Thomas, Alexander; Oxford University, Department of Earth Sciences Mas, Jose Luis; Universidad de Sevilla, Departamento de Fisica Aplicada I Garcia-Orellana, Jordi; Universitat Autònoma de Barcelona, Departament de Física Henderson, Gideon; Oxford University, Department of Earth Sciences Masqué, Pere; Universitat Autònoma de Barcelona, Departament de Física Zahn, Rainer; Universitat Autonoma de Barcelona, Institut de Ciencia i Tecnologia Ambientals; Institució Catalana de Recerca i Estudis Avançats (ICREA) ACS Paragon Plus Environment Submitted to Analytical Chemistry 1 Separation and measurement of Pa, Th and U isotopes in 1 marine sediments by microwave-assisted digestion and 2 MC-ICP-MS 3 César Negre*1, Alexander L. Thomas2, José Luis Mas3, Jordi Garcia-Orellana1,4, Gideon M. Henderson2, 4 Pere Masqué1,4, Rainer Zahn1,5,6 5 1 Institut de Ciència i Tecnologia Ambientals (ICTA), Universitat Autònoma de Barcelona, 08193 6 Bellaterra, Spain. 7 2 Department of Earth Sciences, Oxford University, Parks Road, Oxford, OX1 3PR, United Kingdom. 8 3 Departamento de Física Aplicada I. Escuela Universitaria Politécnica (Universidad de Sevilla), Virgen 9 de Africa 7, 41012 Sevilla, Spain. 10 4 Departament de Física, Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain. 11 5 Institució Catalana de Recerca i Estudis Avançats (ICREA), Lluís Companys 23, 08010 Barcelona, 12 Spain 13 6 Departament de Geologia, Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain 14 [email protected], [email protected], [email protected], jordi.gar[email protected]at, 15 [email protected], [email protected], [email protected]at 16 *Corresponding author: Tel +34-935811191. Fax +34-935813331. E-Mail: cesa[email protected] 17 ABSTRACT 18 Page 1 of 18 ACS Paragon Plus Environment Submitted to Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 2 This manuscript describes a new protocol for determination of Pa/Th/U in marine sediments. Based on 19 microwave-assisted digestion, it represents an important reduction of working time over conventional 20 hot-plate digestion methods, and the use of HClO4 is avoided. While Th and U are completely dissolved 21 with a first microwave step, around 40% of 231Pa remains un-dissolved and a short hot-plate step with 22 reverse aqua regia is required to achieve total digestion and spike equilibration. Next, the method 23 involves a separation of these elements and a further purification of the Pa fraction using Dowex AG1-24 X8 resin. Separation with Bio-Rad and Sigma-Aldrich resins was compared; while both perform 25 similarly for Th and U, Pa yields are higher with Bio-Rad. Finally, samples are measured using a Nu 26 instruments Multiple Collector Inductively Coupled Plasma Mass Spectrometer (MC-ICP-MS). Overall 27 chemical yields range around 50% for Pa, 60% for Th and 70% for U. 28 29 Keywords: protactinium, uranium series, sediment microwave digestion, anion exchange separation, 30 MC-ICP-MS 31 INTRODUCTION 32 The development of Pa, Th, U geochemistry in Earth Sciences research as tracers and chronometers of 33 natural processes, such as magma genesis [1], carbonate dating [2], ocean circulation and past export 34 productivity [3], has relied on the improvement of analytical techniques. Analysis of such radiogenic 35 elements in deep-sea sediments has been performed for many years [4], together with the development of 36 alpha and beta spectrometry, which have been the traditional analytical tools. The appearance of 37 Inductively Coupled Plasma Mass Spectrometry (ICP-MS) and more specifically of MC-ICP-MS in 38 recent years offers great advantages such as an improvement of detection limits and greater precision, 39 coupled with a rapid sample throughput [5,6]. 40 In line with ICP development, microwave-assisted digestion has been used as a sediment pre-41 treatment method for an increasing number of applications [7]. However, application for Pa/Th/U 42 analysis is limited [8,9] and microwave protocols have not been published. This manuscript provides full 43 Page 2 of 18 ACS Paragon Plus Environment Submitted to Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 3 details of a new microwave-assisted digestion protocol for Pa/Th/U determination in marine sediments. 44 Conventional digestion techniques use open vessels or Teflon bombs applying hot plate heating with 45 HCl, HNO3, HClO4, HF and/or H3BO3 in a sequence of multiple addition-heating steps [4,8,10]. 46 Microwave, in turn, enables to directly and rapidly heat the sample solvent mixture, and allows a precise 47 control of temperature and pressure, making total dissolution faster. Total dissolution is essential to 48 quantify and differentiate between detrital, authigenic and scavenged components of Pa, Th and U in 49 sediments, necessary for their use as tracers of natural processes. With microwave assistance, 50 temperature is not limited to the boiling point of the reagents, thus making it easier to achieve complete 51 digestion without HClO4, which requires specially adapted facilities (e.g. fume hood wash-down 52 equipment). In addition, volumes of solvents are typically reduced by a factor up to 10 contributing to 53 blank minimization and reducing the time spent for sample dry-down. 54 A good chemical separation of Pa, Th and U from one other and from other elements is essential to 55 ensure an accurate quantification with MC-ICP-MS. To avoid organic contamination, separation using 56 anion exchange resins has proved preferable over solvent extraction techniques [11]. Such methods are 57 based on the different and variable resin affinity of these elements under changing acid media. Pa 58 separation is especially complicated because of its tendency to be removed from solution by hydrolysis 59 and the necessity to completely eliminate any fluorides generated during digestion. Resins used with 60 success for Pa/Th/U separation include Dowex AG MP-1 [8], TRU-Spec [12], and Dowex AG1-X8 [13]. 61 In this article we describe a new protocol for the extraction, separation and measurement of Pa, Th and 62 U from marine sediments, performing microwave-assisted digestion with HNO3, HF and HCl, using 63 Dowex AG1-X8 anion exchange resin for element separation, and MC-ICP-MS for quantification. We 64 also show and discuss results obtained from experiments carried out to test the suitability of key parts of 65 the method: (1) microwave-assisted vs. hot-plate-based digestion methods (2) a reverse aqua regia step, 66 and (3) Pa monitoring through the elemental separation. 67 68 EXPERIMENTAL SECTION 69 Page 3 of 18 ACS Paragon Plus Environment Submitted to Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 4 Samples. Sediment reference materials exist for 230Th, 232Th, and 238U but not for 231Pa. Hence we 70 used a suite of marine sediment samples from cores MD02-2594 (34.71ºS; 17.34ºE; 2440 m water 71 depth) and MD02-2588 (41.12ºS; 25.30ºE; 2907 m water depth) from the South Atlantic. 76 samples 72 were analyzed, 28 of which were replicated 2 or 3 times with the aim to derive method precision and/or 73 test the different parts of the protocol, among them: (a) 2 samples for microwave digestion validity (1 74 and 2, Tab. 1), (b) 7 samples for the reverse aqua regia clean-up step (2 to 8, Fig. 2), and (c) 4 samples 75 for comparing Dowex AG1-X8 Bio-Rad and Sigma-Aldrich resins (9 to 12, Fig. 3). 76 Equipment. For sample dissolution we used a Microwave Sample Preparation System, Model MDS-77 2000 (CEM Corporation, Matthews, NC, USA). Maximum power output is 630 W at a frequency of 78 2450 MHz, programmable at 1% increments. The system is equipped with a Pressure Control System 79 for controlled pressure conditions up to 1.38 MPa (200 psi) inside a single closed vessel during heating. 80 Isotope abundances were measured with a Nu Instruments Multiple Collector Inductively Coupled 81 Plasma Mass Spectrometer (MC-ICP-MS) [14], coupled to a Nu Instruments DSN-100 desolvating 82 nebulizer with a typical uptake rate of 100 µL per min. Beam voltage is measured across 1011 Ω 83 resistors, and sensitivities were ∼100 V/ppm. A retardation filter on the central ion counter (Tab. 1) 84 reduces abundance sensitivity to 2×10-7 and 1×10-7 at 2 and 3 atomic mass units below 232Th, 85 respectively. 86 Pa yields were determined by gamma detection of 233Pa with a CANBERRA model GCW3522 87 germanium semiconductor, connected to a preamplifier 2002CSL and an analogical-digital converter 88 CANBERRA 8701. Total detection efficiency is 22.7% at 312 keV with 1.52 keV resolution (FWHM), 89 assessed with a 237Np standard solution. The detector is confined inside a 15 cm thick iron shielding to 90 minimize background. 91 Reagents and materials. HCl 32% m/v PA Merck, HNO3 65% PA Merck, HF 40% PA-ISO Panreac 92 and 18 MΩcm-1 water (Milli-Q) were used in order to minimize sample contamination. 93 Page 4 of 18 ACS Paragon Plus Environment Submitted to Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 5 Dowex AG1-X8 anion exchange resin, chloride form, 100-200 mesh (80-150 µm particle size; Bio-94 Rad and Sigma-Aldrich Laboratories) was used for separation of Th, Pa and U, and further purification 95 of Pa fractions. Silica gel 40-63 µm particle size (230-400 mesh ASTM; Sigma-Aldrich) was used for 96 the preparation of the 233Pa spike. As resin support, conical polypropylene columns with 2 mL bed 97 volume (8 mm internal diameter) plus 10 mL reservoir (Bio-Rad Laboratories) were employed. 98 Internal tracers used were 233Pa, 229Th and 236U. 233Pa spikes were prepared by milking a 237Np 99 solution of 100.2 kBq, (Eckert & Ziegler Isotope Products, Valencia, CA, USA) adapting the protocol 100 described in Regelous et al. [15]. 231Pa impurities in the 233Pa spike, stemming from the 237Np solution, 101 were quantified during a preliminary calibration soon after milking and therefore before decay of 233Pa 102 to 233U. After a 6 months waiting period to allow nearly total decay to 233U, 233Pa spikes were calibrated 103 by MC-ICP-MS against a well known 236U standard [16]. This approach allows the 233Pa concentration to 104 be measured from a 233U/236U ratio and the 231Pa from a 231Pa/233Pa ratio, therefore eliminating the effect 105 of elemental fractionation within the mass spectrometer. 229Th and 236U sources were prepared at the 106 Universitat Autònoma de Barcelona by dilution of certified reference solutions (229Th, 73473-759, 107 Analytics, Atlanta, GA, USA; and 236U, 1168-70, Isotope Products Europe, CA, USA). For MC-ICP-MS 108 measurement corrections, uranium (CRM-145, New Brunswick Laboratory, NJ, USA) and thorium (Cat 109 # 100059-1, High-Purity Standards, Charleston, SC, USA) standards were used. 110 To avoid deposition of Pa on the walls of glassware [17], polypropylene (PP) or polytetrafluoroethylene 111 (PTFE) material was used. 112 Method. The analytical procedure for the measurement of Pa, Th, and U in sediments consists of three 113 steps: sample digestion, element separation with anion exchange resin and MC-ICP-MS measurement 114 (Fig. 1). Prior to digestion, samples were freeze dried and ground to a fine powder and homogenized 115 with a ceramic mortar. Immediately before weighing, samples were heated at 60 ºC to a constant weight 116 to completely remove moisture. 117 Aliquots of about 0.20 g of sediment were placed in 100 mL microwave PTFE liners and weighed 118 with a microbalance (to 1×10-4 g precision). Samples were covered with 2 mL of Milli-Q water and 119 Page 5 of 18 ACS Paragon Plus Environment Submitted to Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 6 traced with 236U (36 mBq), 229Th (0.32 Bq) and 233Pa (~4 Bq) spikes. Next, a mixture of concentrated 120 HNO3:HCl:HF (10:4:6 mL) was added to dissolve CaCO3 and break down aluminosilicates. After 12 121 hours for tracer homogenization, samples were run in the microwave at 0.69 MPa for 5 min and at 0.90 122 MPa for an additional 30 min [18]. Samples were left to cool and transferred to PTFE beakers. To 123 minimize the loss of sample by adsorption onto the microwave liner, the interior of each vial was rinsed 124 with about 2 mL of HNO3. Beakers were heated at 140 ºC to reduce solution volume to incipient 125 desiccation. As an additional clean-up step, 9 mL of reverse aqua regia (6 mL HNO3 + 3 mL HCl) were 126 added to remove remaining organic matter from the solution. Samples were dried once more at 140 ºC 127 and redissolved in 6 mL of HNO3. This operation was repeated twice in order to remove fluorides. 128 Samples were finally dissolved in 4 mL of HNO3 7.5M, ready for anion exchange. 129 Pa, Th and U were separated adapting the procedure described in Edwards et al. [19]. This involves (1) 130 washing of 2 mL of Dowex AG1-X8 anion exchange resin per column with 10 mL Milli-Q, 10 mL HCl 131 6M, 10 mL Milli-Q and pre-conditioning with 10 mL HNO3 7.5 M; (2) loading samples onto the resin; 132 (3) washing columns with additional 4mL of HNO3 7.5 M and converting to chloride form with 1.5mL 133 HCl 6M; (4) eluting Th, Pa, and U with 6 mL of HCl 6 M, 6mL HCl 6 M + HF 0.05 M, and 6 mL Milli-134 Q water, respectively; (5) drying Pa fractions and subsequent recovery with 4 mL of HNO3 7.5M; (6) 135 purification of these Pa fractions in a second ion exchange separation, repeating steps 1-4. The second 136 Pa purification is required to improve Pa:Th separation so as to reduce the interference of 232Th hydrides 137 to the 233Pa beam during measurement. 138 The day prior to MC-ICP-MS measurement, Pa, Th and U fractions were reduced to a single drop and 139 recovered in 1 mL of distilled 2% HNO3. Small and diluted aliquots of the Th and U fractions were used 140 for a rapid concentration check of 238U and 232Th in the multi-collector to optimize their intensities in 141 the final measurements at ∼4V. U, Th and Pa isotope abundances were then measured by MC-ICP-MS 142 at Oxford University following standard protocols [16] (Tab. 2). Uranium was measured statically with 143 234U in an ion counter, correcting for machine biases by bracketing each sample with a measurement of 144 CRM-145 uranium standard. Thorium was measured in a two-step dynamic routine, correcting for 145 Page 6 of 18 ACS Paragon Plus Environment Submitted to Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 7 intensity variation by normalising to 235U measured in each step. In cases where the sample 235U signal 146 in the diluted Th fractions was below 5 mV (i.e. ∼50ppt solution concentrations of 235U), CRM-145 U 147 standard was added to the Th solutions to increase the beam size to ∼10 mV to facilitate this correction. 148 While the 230/229 measurement does not require correction for ion counter efficiency the 232/229 mass 149 ratio is corrected using the standardised 235U/234U ratio of CRM-145 measured with 234U in the ion 150 counter. Protactinium analyses were made using multiple ion counting channels. Typical 231Pa beam 151 sizes were 10-7-10-6 V. Ion-counter gain and mass bias were measured by preceding each sample 152 analysis with a CRM-145 U standard measurement. Abundance sensitivity and molecular interferences 153 were assessed by measuring a 232Th standard solution, as if it was a Pa sample, monitoring the 232Th 154 hydride and tailing contributions at the masses of interest. After corrections, 231Pa/232Th ratios in the Pa 155 fraction were in the order of 3×10-5. Abundance sensitivity originating from U isotopes is negligible 156 because of the good separation of U from Pa in the two-step anion exchange, typical 231Pa/238U ratios 157 measured in the Pa fraction were around 6×10-3. 158 159 RESULTS AND DISCUSSION 160 Microwave-assisted digestion with reverse aqua regia clean-up step. The suitability of the 161 additional clean-up step was assessed by processing 7 samples (2-8, Fig. 2) with and without reverse 162 aqua regia. For reference, sample 2 was also analysed following the hot-plate methodology described by 163 Thomas et al. [20]. While both techniques produced equivalent Th and U specific activities, Pa was 164 systematically lower by an average of 40% in samples not exposed to reverse aqua regia. Results 165 indicate that microwave-assisted digestion achieves a complete extraction of Th and U, but 231Pa 166 becomes only partially extracted and full spike equilibration is not achieved. Incomplete digestion leads 167 in some 231Pa either passing through the resin too quickly or being retained on the columns. This loss is 168 reflected in the low 231Pa/233Pa measured in Pa fractions. Total Pa spike equilibration requires the more 169 aggressive chemical treatment with the additional reverse aqua regia step that removes organic matter 170 Page 7 of 18 ACS Paragon Plus Environment Submitted to Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 8 as a potential binder of non-extracted 231Pa. Using reverse aqua regia, 231Pa results reproduce those 171 achieved with hot-plate digestion (Tab. 1) suggesting that microwave-assisted digestion enables 172 adequate dissolution of all Pa, Th and U isotopes as long as the reverse aqua regia step is used. 173 Anion exchange separation. Very low concentrations of 231Pa in marine sediments require 174 optimizing Pa recoveries hence making ion exchange elemental separation particularly critical. Pa 175 behaviour was studied in 2 batches of 11 samples (4 replicates, shown in Fig. 3) and a procedural blank 176 each, that were processed with Dowex AG1-X8 provided by two different manufacturers, Bio-Rad and 177 Sigma-Aldrich Laboratories. 233Pa content in each fraction after the first ion exchange separation was 178 monitored by gamma spectrometry (Fig. 3a). Results show higher Pa yields using the Bio-Rad 179 Laboratories resin, with (61 ± 8)% of Pa being recovered in the Pa fraction (at a rate of 10% per mL) 180 when pouring HCl 6M with HF traces through the resin. With Sigma-Aldrich resin only a (12 ± 14)% of 181 Pa was recovered in the Pa fraction, likely due to some matrix effect affecting Sigma-Aldrich resin more 182 than Bio-Rad. This conclusion is supported by optimum Pa behaviour in (a) both procedural blanks, (b) 183 the second Pa purification step, and (c) the improvement of Pa recoveries up to 80% when smaller 184 amounts of sediment (∼0.1 g) were processed. Using Bio-Rad, more Pa was eluted with the Load and 185 Washing fraction than with Sigma-Aldrich (3.8 against 2.5% per mL), but very little was lost into the Th 186 fraction (0.6% per mL) compared to Sigma-Aldrich (10% per mL). Loss of Pa into the Load and Wash 187 fraction was likely a consequence of competition with other anions in the sample dissolution. In order to 188 maximize Pa recoveries resin volumes shall be increased, or less sediment processed. 189 Recoveries, validity, precision and uncertainties. Overall average recoveries at the time of 190 measurement were (53 ± 12)% with Bio-rad (n=66, 1σ) versus (24 ± 19)% with Sigma-Aldrich (n=44, 191 1σ). Despite the low recoveries achieved with the Sigma-Aldrich resin, Pa was nonetheless quantified 192 accurately (Fig. 3b) thanks to adequate spike equilibration during microwave digestion (including aqua 193 regia) and a detection limit of the MC-ICP-MS of about 3×10-8 V (∼3×10-4 ppt) in the central ion 194 counter. This is reflected in the 231Pa concentration results, which are similar between the two resins (to 195 Page 8 of 18 ACS Paragon Plus Environment Submitted to Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 15 Figure 2. Page 15 of 18 ACS Paragon Plus Environment Submitted to Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 16 299 Figure 3. 300 Page 16 of 18 ACS Paragon Plus Environment Submitted to Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 17 301 Microwave-assisted digestion method Hot plate digestion method Sample 231Pa 238U 230Th 232Th 231Pa 238U 230Th 232Th 2.57 ± 0.08 56 ± 2 65 ± 2 11.1 ± 0.4 2.73 ± 0.13 52.18 ± 0.10 67.1 ± 0.6 12.68 ± 0.16 1 2.57 ± 0.09 - 65 ± 3 11.0 ± 0.4 2.54 ± 0.13 53.18 ± 0.10 68.7 ± 0.7 11.09 ± 0.17 6.56 ± 0.15 59 ± 2 104 ± 3 19.2 ± 0.5 6.35 ± 0.18 55.21 ± 0.11 107.6 ± 0.9 19.9 ± 0.2 6.78 ± 0.11 59 ± 2 103 ± 5 18.4 ± 0.7 6.31 ± 0.17 55.02 ± 0.11 107.0 ± 1.3 19.9 ± 0.3 2 5.99 ± 0.11 58 ± 2 101 ± 4 18.7 ± 0.7 - - - - Table 1. Page 17 of 18 ACS Paragon Plus Environment Submitted to Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 18 302 STEP F (+7) F (+5) F (+3) F (+2) F (+1) AXIAL F (-1) F (-2) IC (-3) F (-4) IC (-5) F (-6) IC (-7) F (-8) 1 238 235 234 Pa 2 238 235 234 Bracketing CRM 145 standard Th 3 238 235 234 Uranium 4 238 236 235 234 5 235 232 230 229 Thorium 6 238 235 230 229 Protactinium 7 238 236 235 233 232 231 229 Table 2. 303 Page 18 of 18 ACS Paragon Plus Environment Submitted to Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60