Precise Determination of the Absolute Isotopic Abundance Ratio and

Oct 22, 2012 - The absolute R(37Cl/35Cl) ratios for NIST SRM 975, NIST 975a, and ISL 354 by the precise calibrated .... related to continental inflow ...
0 downloads 0 Views 927KB Size
Article pubs.acs.org/ac

Precise Determination of the Absolute Isotopic Abundance Ratio and the Atomic Weight of Chlorine in Three International Reference Materials by the Positive Thermal Ionization Mass SpectrometerCs2Cl+‑Graphite Method Hai-Zhen Wei,† Shao-Yong Jiang,*,† Ying-Kai Xiao,‡ Jun Wang,§ Hai Lu,§ Bin Wu,§ He-Pin Wu,† Qing Li,† and Chong-Guang Luo‡ †

State Key Laboratory for Mineral Deposits Research, Department of Earth Sciences, Nanjing University, Nanjing 210093, PR China Qinghai Institute of Salt Lakes, Chinese Academy of Sciences, Xining 810003, PR China § National Institute of Metrology, Beijing 100013, PR China ‡

ABSTRACT: Because the variation in chlorine isotopic abundances of naturally occurring chlorine bearing substances is significant, the IUPAC Inorganic Chemistry Division, Commission on Isotopic Abundances and Atomic Weights (CIAAW-IUPAC) decided that the uncertainty of atomic weight of chlorine (Ar(Cl)) should be increased so that the implied range was related to terrestrial variability in 1999 (Coplen, T. B. Atomic weights of the elements 1999 (IUPAC Technical Report), Pure Appl. Chem. 2001, 73(4), 667−683; and then, it emphasized that the standard atomic weights of ten elements including chlorine were not constants of nature but depend upon the physical, chemical, and nuclear history of the materials in 2009 (Wieser, M. E.; Coplen, T. B. Pure Appl. Chem. 2011, 83(2), 359−396). According to the agreement by CIAAW that an atomic weight could be defined for one specified sample of terrestrial origin (Wieser, M. E.; Coplen, T. B. Pure Appl. Chem. 2011, 83(2), 359−396), the absolute isotope ratios and atomic weight of chlorine in standard reference materials (NIST 975, NIST 975a, ISL 354) were accurately determined using the high-precision positive thermal ionization mass spectrometer (PTIMS)-Cs2Cl+-graphite method. After eliminating the weighing error caused from evaporation by designing a special weighing container and accurately determining the chlorine contents in two highly enriched Na37Cl and Na35Cl salts by the current constant coulometric titration, one series of gravimetric synthetic mixtures prepared from two highly enriched Na37Cl and Na35Cl salts was used to calibrate two thermal ionization mass spectrometers in two individual laboratories. The correction factors (i.e., K37/35 = R37/35meas/R37/35calc) were obtained from five cycles of iterative calculations on the basis of calculated and determined R(37Cl/35Cl) values in gravimetric synthetic mixtures. The absolute R(37Cl/35Cl) ratios for NIST SRM 975, NIST 975a, and ISL 354 by the precise calibrated isotopic composition measurements are 0.319876 ± 0.000067, 0.319768 ± 0.000187, and 0.319549 ± 0.000044, respectively. As a result, the atomic weights of chlorine in NIST 975, NIST 975a, and ISL 354 are derived as 35.45284(8), 35.45272(21), and 35.45252(2) individually, which are consistent with the issued values of 35.453(2) by IUPAC in 1999.

C

(electron impact ionization mass spectrometry), CH3Cl+-IRMS (gas source dual inlet isotope ratio mass spectrometry), and Cl−NTIMS (negative thermal ionization mass spectrometry) methods. The inherent disadvantages including complicated pretreatment procedure, strong memory effect, and poor precision in the above techniques could not be resolved until the establishment of positive thermal ionization mass spectrometer (PTIMS)-Cs2Cl+-graphite procedure by Xiao et al., who observed very stable emission of polyatomic Cs2Cl+ ion beam with graphite loading in thermal ionization mass spectrometry (TIMS).15 Among the major techniques men-

hlorine comprises two stable isotopes, 37Cl and 35Cl, and the mean abundance of 35Cl and 37Cl are 75.77% and 24.23%, respectively, in nature as recommended by IUPAC.3,4 With rapid improvement of various measurement techniques, significant fractionations of chlorine isotopes in the geological process have been observed such that δ 37Cl values range from −14‰ to +16‰.5−13 The applications of chlorine isotopic variation have been advanced in geochemistry and environmental chemistry fields, such as tracking original sources and evolution of groundwater,14 investigating chlorine isotopic fractionation between brine and evaporated sediment in salt lakes,15 exploring ore deposit and magmatic and hydrothermal processes,5 tracing organic pollutions, etc.16,17 The isotopic composition of chlorine was determined first by Nier et al. in 1936 and then followed by Aston,18 Owen,19 Horing,20 Shields,3 Taylor,21 Kaufmann,6 and Long22 by employing HCl+-EIIMS © 2012 American Chemical Society

Received: August 29, 2012 Accepted: October 22, 2012 Published: October 22, 2012 10350

dx.doi.org/10.1021/ac302498q | Anal. Chem. 2012, 84, 10350−10358

Analytical Chemistry

Article

tioned above, the procedure of CH3Cl+-IRMS developed by Owen and Schaefler in 1955 and then modified by Kaufmann and Long6,19,22 achieved a total analytical precision as high as 0.09%. The comparison between two leading techniques of chlorine isotopes measurement (i.e., CH3Cl+-IRMS and PTIMS-Cs2Cl+graphite) has been performed by Rosenbaum et al.,23 and it turned out that both the analytic precision and accuracy of chlorine isotopic compositions in five natural samples by the two methods are identical. The PTIMS-Cs2Cl+-graphite method provides a simple pretreatment procedure, high analytical efficiency, and high precision and sensitivity and has been widely used in environmental and geochemistry fields.24−27 It is important to note that the CH3Cl+-IRMS method only provides the chlorine isotopic composition with δ 37Cl values rather than R(37Cl/35Cl) ratios and needs to make a complicated correction for 14C, 2H, and 3H. Therefore, it could not be employed for determination of the atomic weight of chlorine. The first value of chlorine atomic weight was issued in 1882 and then updated several times as shown in Table 1. Clearly, the

expressed as an interval in 2009 and the currently accepted atomic weight of chlorine, Ar(Cl), is [35.446, 35.457].2 The absolute abundance ratio of natural chlorine according to NIST SRM 975 NaCl was given by surface emission mass spectrometry based on the measurement of Cl− ion in 1962.3 The certified reference material of NIST SRM 975 NaCl is used up and unavailable at present, but the substituted standard reference material of chlorine isotopes (NIST SRM 975a) made of commercial NaCl product has not been widely adopted because it does not represent geological significance. Scientists used to adopt chlorine isotopic compositions of seawater as the standard mean value of chlorine in ocean (SMOC) (i.e., δ 37Cl = 0) in different laboratories in the world. However, the values of δ 37 Cl in seawater collected in different areas are not uniform and obvious local differences in R(37Cl/35Cl) values of seawater samples collected in Pacific Ocean, Indian Ocean, and Atlantic Ocean have been observed by Xiao et al.28 They also demonstrated that the variation of δ 37Cl values are apparently related to continental inflow and region. Later on, an isotopic reference material of chlorine as NaCl (i.e., ISL 354 NaCl) was prepared by Xiao et al. using an ion exchange technique from collected seawater in the Pacific Ocean located at 4°18′ N, 161°08′ E,29 and then was recommended as New Reference Materials for Stable Isotope of Chlorine by IAEA in 2004.30 The absolute isotope ratio in ISL 354 NaCl has not been given so far, which greatly restricts its wide application in stable chlorine isotope geochemistry fields in the world. On the basis of investigations discussed above, the CIAAW-IUPAC emphasized the need for new precise calibrated isotopic composition measurements in order to improve the accuracy of the atomic weights of chlorine. In this work, chlorine isotope ratio was accurately determined by a Triton TI thermal ionization mass spectrometer in Nanjing University using the high-precision PTIMS-Cs2Cl+-graphite method. For comparison, the samples were also measured in another laboratory at Institute of Salt Lakes, Chinese Academy of Science (CAS) using a VG 354 thermal ionization mass spectrometer, and the two laboratories gave the same results. The mass discrimination effects were corrected with one series of synthetic isotope mixtures that were gravimetrically prepared with two highly enriched isotope materials of Na35Cl and Na37Cl. After carefully checking the recovery assay and the possible isotopic fractionation of chlorine during the ion-exchange procedure and strictly assessing the measurement uncertainty during isotopic analysis, the absolute isotopic compositions in the accepted reference materials mentioned above were

Table 1. Historical Values of Ar(Cl) Recommended by the Commission of CIAAW-IUPAC1,2 date

atomic weight Ar(E)

1882 1909 1925 1961 1969 1985 1999 2009

U[Ar(E)]

35.45 35.46 35.457 35.453 35.453 35.4527 35.453 [35.446, 35.457]

atomic mass criterion 16

O = 16 O = 16 16 O = 16 12 C = 12 12 C = 12 12 C = 12 12 C = 12 12 C = 12

0.1 0.03 0.003 0.001 0.001 0.0009 0.002

16

measurement precision of chlorine isotopes has been improved greatly with continuous advancement of analytical techniques. It is worth noting that the uncertainty in the value of chlorine atomic weight reported in 1999 is worse than that issued in 1985, which indicates that the obvious variation of chlorine isotopes in nature has been confirmed and the uncertainty of Ar(Cl) (atomic weight of chlorine) also has been enlarged correspondingly. The Commission on Isotopic Abundances and Atomic Weights (CIAAW) decided to increase the uncertainty to include most Cl-bearing materials,1 and then, it emphasized that the standard atomic weights of ten elements (i.e., H, Li, B, C, N, O, Si, S, Cl, Tl) are not constants of nature and each atomic weight value is

Table 2. Impurities in Na37Cl and Na35Cl Purchased from Oak Ridge National Laboratorya impurities

Mg

B

Al

Si

Ca

Cr

Li

Be

Rb

Sr

Y

Zr

Nb

Mo

Ru

Pd

37

Na Cl Na35Cl impurities

40.46 47.49 Ag

26.16 95.33 Cd

2.06 0.93 In

6.19 7.13 Sn

0.77 3.54 Sb

0.34 1.09 Te

0.29 0.12 Cs

0.06 0.05 Ba

0.06 0.10 La

0.39 0.68 Ce

0.06 0.05 Pr

0.75 0.65 Nd

0.10 0.09 Sm

0.09 0.23 Eu

0.01 0.01 Gd

b 0.01 Tb

Na37Cl Na35Cl impurities

22.70 3.30 Dy

0.17 b Ho

Na37Cl Na35Cl impurities

0.09 0.07 U

0.09 0.07 S

0.07 0.05 Sc

Na37Cl Na35Cl

0.04 0.04

0.97 1.69

0.001 0.003

a

0.08 0.06 Er

0.45 0.20 Tm

b 0.12 Yb

0.11 0.03 Lu

0.03 0.04 Hf

1.12 1.02 Ta

0.06 0.05 W

0.06 0.05 Re

0.05 0.04 Ir

0.05 0.13 Au

0.08 0.06 Tl

0.04 0.04 Pb

0.08 0.06 Bi

0.05 0.04 Th

0.07 0.06 Ti

0.01 0.01 V

0.02 0.02 Mn

1.25 1.04 Fe

0.34 0.30 Co

0.21 7.31 Ni

0.41 0.01 Cu

0.001 0.01 Zn

0.09 0.07 Ga

0.06 0.03 Ge

0.10 0.13 As

0.08 0.07

0.07 0.06

b 0.23

0.87 0.03

0.08 0.07

0.03 0.14

0.03 0.07

0.09 0.10

0.02 0.01

0.15 0.42

0.02 0.01

0.60 0.08

0.33 0.74

Se 21.58 b

Unit: μg/g. bUnder determination limit. 10351

dx.doi.org/10.1021/ac302498q | Anal. Chem. 2012, 84, 10350−10358

Analytical Chemistry

Article

determined to obtain the accurate values of the atomic weight of chlorine.



EXPERIMENTAL SECTION Materials. Two high-purity chlorine isotope spikes, chlorine37 (sodium chloride, batch: 198501; atomic percent issued by ORNL: x(35Cl) = 1.79 ± 0.01%, x(37Cl) = 98.21 ± 0.01%) and chlorine-35 (sodium chloride, batch: 150301; atomic percent issued by ORNL: x(35Cl) = 99.90 ± 0.05%, x(37Cl) = 0.10 ± 0.05%) were purchased from Oak Ridge National Laboratory, USA. Milli-Q water (resistivity, 18.2 MΩ·cm−1) was used throughout the experiments. Concentrated HCl and HNO3 were purified twice by sub-boiling distillation. Solutions of NIST SRM 975 NaCl, NIST SRM 975a NaCl, and ISL 354 NaCl were prepared using Milli-Q water. The impurities in Na37Cl and Na35Cl were determined using an inductively coupled plasma mass spectrometry (Element II, Thermal Fisher Finnigan, Germany) as shown in Table 2. Weighing Container Designed for Accurate Subtraction Weighing. The mass loss caused by evaporation during weighing of the liquid sample is estimated according to the equation shown below (eq 1.1).31 After substituting the relative parameters into the equation, an evaporation rate (u) of 0.0126 g min−1 (i.e., 12.6 mg min−1) was estimated for a 2.5 mL open beaker with radius of 1.00 cm at room temperature of 20 °C. u = (P / ρ l ) ×

M /(2πRT )

Figure 1. Variation of solution mass with time in the designed weighing container at room temperature of 25 °C.

which the evaporation rate of 0.69 μg·min−1 is derived from the slope of the fitting line. Compared with the evaporation rate of 12.6 mg min−1 using an open beaker, the weighing error of 3.45 × 10−8 caused by solution evaporation is not significant when weighing ∼20 g of liquid sample in 1−2 min using the designed weighing container. Therefore, the designed weighing apparatus can be used reliably for accurate solution weighing with two superior advantages of controllable droplet size and minimized evaporation effect. Determination of Chlorine Content in Na35Cl and Na37Cl and Preparation of Primary Solution of Enriched 35 Cl and 37Cl. Two solid Na37Cl and Na35Cl powders were weighed using an UMX 5 balance (Mettler Toledo, ± 0.1 μg) after being sintered in a furnace at 500 °C for 6 h and cooled down in two individual desiccators for 24 h prior to weighing. The Na37Cl and Na35Cl primary solutions containing ∼0.35 mmol·g−1 of NaCl were prepared. Three parallel samples of each primary solution were weighed to measure chlorine content in solid Na37Cl and Na35Cl by the current constant coulometric determination (Tables 3 and 4). The coulometric apparatus was designed by National Institute of Metrology (P. R. China) and consists of a silver anode with 99.999% purity and a platinum cathode. The mass concentration of NaCl (wt %) is calculated by the following equation (eq 1.2):

(1.1)

in which u is the evaporation rate, g·min−1; P is the saturated vapor pressure, 2.34 × 103 Pa; ρl is the solution density, 1000 kg·m−3; R presents the gas constant, 8.31 J mol−1 K−1; M is the molar mass of solution, 18 × 10−3 kg·mol−1; T is the absolute temperature, 293 K. Obviously, solution evaporation in an open beaker makes the weighing error too big to be acceptable. In order to avoid mass loss caused by evaporation, a weighing container was designed in this work as shown in Scheme 1, which comprises two dissociable parts: (a) capillary injector and (b) squeezable container. After recording the mass variation of solution with time in the designed container over 7 days at room temperature, a linear curve of solution mass vs time was obtained as shown in Figure 1, from Scheme 1. Schematic Illustration of the Weighing Container Designed for Eliminating Mass Loss Caused by Evaporation Effect during Weighing Liquid Samples: (a) Capillary Injector with the out Diameter of ∼1.0 mm; (b) Squeezable Container with the Volume of ∼25 mL

w=

QP QT

× 100 =

I ·t ·M E·M·t × 100 = × 100 n·m·F n·R·F ·m (1.2)

where Qp (coulomb, C) is the actual charge used to electrolyze the sample; QT (C) is the theoretical charge; I (A) is the working current through the solution; t (s) is the electrolysis time; m (g) is the mass of solid NaCl; M (g·mol−1) is the molar mass of sodium chlorine; n is the number of electrons transferred; F is the Faraday constant; E (V) represents the potential of the standard electrolysis cell; R (Ω) is the resistance of the standard electrolysis cell. The impurities of the solvent (i.e., Milli-Q water) are checked, and they are under the detection limit of ICPMS and ion chromatography techniques. Considering the current constant coulometric titration gives the total concentration of haloid ions (i.e., Cl−, Br−, I−, etc) and other impurities (i.e., CO32−, F−, SO42−, etc) have no influence on the analysis accuracy of the method, the absolute purity of Cl− in Na37Cl and Na35Cl was 10352

dx.doi.org/10.1021/ac302498q | Anal. Chem. 2012, 84, 10350−10358

Analytical Chemistry

Article

Table 3. Preparation of Highly Enriched Na37Cl and Na35Cl Primary Solution sample

Na37Cl

Na35Cl

enrichment

x( Cl) = 98.21 ± 0.01% x(37Cl) = 98.193 ± 0.003% (2σ, n = 8) x(37Cl) = 98.105 ± 0.005% (2σ, n = 7) 481.9398 22.9234 0.350869 21.0254

x( Cl) = 99.90 ± 0.05% x(35Cl) = 99.857 ± 0.004% (2σ, n = 6) x(35Cl) = 99.834 ± 0.002% (2σ, n = 7) 467.6598 22.7498 0.354696 22.5542

37

Oak Ridge our laboratories

Triton TI VG 354

mass of solid powder (mg) mass of prepared solution (g) NaCl concentration (mmol g−1) NaCl concentration (mg g−1)

35

Table 4. Determination of Chlorine Purity in Two Highly Enriched Na37Cl and Na35Cl by the Current Constant Coulometric Titration mass of primary solution (g) measurement no. 1 2 3

[NaCl] titrated in primary solution (mg g−1)

chlorine purity in solid spike (%)

35

Na Cl

37

Na Cl

35

Na Cl

Na Cl

4.62201 4.58743

4.16701 4.59433 4.80842

19.9177 19.9224

20.8801 20.8812 20.8839

96.9033 96.9261

99.3093 99.3142 99.3273

19.9200

20.8817

96.9147 ± 0.0161

99.3169 ± 0.0093

mean

37

Na35Cl

Na37Cl

Table 5. Absolute Chlorine Purity in Two Highly Enriched Na37Cl and Na35Cl Solutions after Subtracting Br− and I− Impurities from the Total Purity of Cl in NaCl sample

[Cl−] in primary solution titrated (mg g−1)

[Br−] in primary solution (mg g−1)

[I−] in primary solution (mg g−1)

[Cl−] in primary solution corrected (mg g−1)

[NaCl] in primary solution (mmol g−1)

chlorine purity in solid spike (%)

Na35Cl Na37Cl

12.01896 12.86989

0.00012 0.00012

0.0000261 0.0000290

12.01881 12.86974

0.34357 0.34847

96.9134 99.3156

calculated by subtracting Br− and I− impurities from the total purity of NaCl as shown in Table 5. Preparation of the Gravimetric Synthetic Mixtures. The gravimetric synthetic mixtures were prepared by weighing two highly enriched Na37Cl and Na35Cl primary solutions with a XP 205 balance (Mettler Toledo, ± 0.01 mg). The weight pertained to the difference between the designed weighing container before and after squeezing liquid drops into 2 mL containers. Buoyancy correction was implemented for all weighing data. Portions of primary solutions Na37Cl and Na35Cl were accurately weighed, and the R(37Cl/35Cl) values in gravimetric synthetic mixtures ranged from ca. 0.1 to ca. 9.0. The isotope ratios were calculated based on equation (eq 1.3). The weighing data and the calculated isotopic ratios of eight mixtures are listed in Tables 6 and 7. R37/35 =

WNa37ClC Na37ClfNa37Cl WNa37ClC Na37ClfNa37Cl

37

Cl

35

Cl

+ WNa35ClC Na35ClfNa35Cl + WNa35ClC Na35ClfNa35Cl

Table 7. Calculated Chlorine Isotopic Ratios R(37Cl/35Cl) in Gravimetric Synthetic Solutions R(37Cl/35Cl) calculated

37

Cl Cl

Table 6. Concentration and Isotopic Abundance in Percent of 37 Cl and 35Cl in Primary Solutions isotopic abundance in percent solutions 35

Na Cl

0.34357

Na37Cl

0.34847

MS type Triton TI VG 354 Triton TI VG 354

35

Cl

37

Cl

99.857 ± 0.004

0.143 ± 0.004

99.834 ± 0.002 1.807 ± 0.003

0.166 ± 0.002 98.193 ± 0.003

1.895 ± 0.005

98.105 ± 0.005

weight of solution Na37Cl (g)

weight of solution Na35Cl (g)

Triton TI

VG 354

mixture-1 mixture-2 mixture-3 mixture-4 mixture-5 mixture-6 mixture-7 mixture-8

0.18644 0.23180 0.29855 0.21677 0.28662 0.25075 0.21518 2.01135

1.78662 0.78519 1.01096 0.69135 0.85447 0.70423 0.54336 0.19341

0.10553 0.29422 0.29432 0.31231 0.33388 0.35418 0.39349 8.70798

0.10550 0.29402 0.29417 0.31159 0.33349 0.35356 0.39237 8.62780

where W refers to weight of the primary solutions, g; C is the concentration of the primary solutions, mmol·g−1; f 37Cl and f 35Cl are the isotopic abundance of 37Cl and 35Cl in two primary solutions. Ion-Exchange Procedure to Convert NaCl into CsCl. A two-column ion-exchange procedure was employed to convert NaCl solution into neutral CsCl solution, where the first 0.5 mL of Hydrogen-Form cation resin column connects with the second 0.5 mL Cs-Form cation resin column in series on the basis of the procedure reported by Xiao et al.29 H-Form resin was prepared by regenerating strongly acidic cation resin (Dowex 50W×8, H-Form, USA, 200−400 mesh) through 20 mL of 2 M HNO3 (twice sub-boiling purified) and washing to neutral pH, and Cs-Form resin was prepared by regenerating the Dowex 50W×8 resin through 10 mL of 0.2 M CsCO3 (Sinopharm Chemical Reagents Limited Company, SP) and washing to neutral pH individually.

35

(1.3)

[NaCl] (mmol g−1)

sample no.

10353

dx.doi.org/10.1021/ac302498q | Anal. Chem. 2012, 84, 10350−10358

Analytical Chemistry

Article

Mass Spectrometric Measurement. A Triton TI mass spectrometer (Thermal Fisher Finnigan, Germany) in Nanjing University was used for chlorine isotope analysis by the static double collection method with collecting 133Cs237Cl+ (m/z 303) and 133Cs235Cl+ (m/z 301) ions at Center Cup (m/z 301) and H1 Cup (m/z 303). For comparison, another VG 354 mass spectrometer (Micromass, UK Ltd., Wynthenshawe, Manchester, UK) in Institute of Salt Lakes, CAS was used for the dynamic single jumping collection method with collecting two peaks in sequence at Center Cup. Single tantalum filaments (0.75 cm × 0.076 cm × 0.0025 cm) from H. Cross Company (USA) were degassed under vacuum for 1 h at a current of 3.0 A and then allowed to oxidize in the ambient atmosphere (protected from contamination in closed boxes) for at least 24 h prior to use. Exactly 1.0 μL of graphite slurry was first loaded onto the filament, and then, the sample solution followed after vaporizing the slurry drop at room temperature for 30 s. The high purity graphite (John-Matthew Company, 99.9999% pure) was mixed into a mixture with a volume ratio of 80% ethanol to 20% water to form slurry as a promoter for ionization for positive thermal ionization mass spectrometer (PTIMS) analysis. In the measuring procedure, the filament current was controlled at ∼1100 mA to produce stable emission of Cs2Cl+ ions after being increased to 1000 mA by a step of 100 mA min−1 and then to 1100 mA by a step of 20 mA min−1, and the data acquisition started when around 1.5 V of 133Cs235Cl+ (m/z = 301) ion beam was achieved. Each individual measurement routine lasted for ∼10 min and comprised 10 analytical blocks of 10 cycles (10 blocks × 10 cycles), and the baseline was monitored at the interval of each block. Chlorine isotope ratios could be obtained on the basis of R(37Cl/35Cl) = R303/301, and the internal analytical precision and the external uncertainty of measured R(37Cl/35Cl) ratio of 0.5 μg NIST 975 is ±0.005% and 0.009% computed from individual independent measurements (2σ standard deviation, n = 10). Instrumental mass fractionation was calibrated with the gravimetric synthetic mixtures of two chlorine isotopes. To make two sets of measurements comparable, both three international reference materials (i.e., NIST 975, NIST 975a, ISL 354) and the gravimetric synthetic mixtures were measured in the same analytical session under exactly the same experimental conditions, such as same sample matrix (pure CsCl), same loading amount of chlorine (0.5 μg of Cl for mass Triton TI and 1.0 μg of Cl for mass VG 354), and same ionization temperature, etc.

Figure 2. Elution and recovery curve of chlorine with two connected columns consisting of H-Form cation resin (Dowex 50W×8) and CsForm cation resin (Dowex 50W×8).

loading procedure, the instrument-induced mass fractionation during the measurement cannot be corrected internally using the conventional exponential law.33−36 Instrumental mass fractionation was calibrated with the gravimetric synthetic mixtures of two chlorine isotopes. In order to find the correction factor, K37/35, for the mass discrimination effect, the R(37Cl/35Cl) ratios of the synthetic isotope mixtures were determined at the same analytical session by PTIMS-Cs2Cl+-graphite method using a Triton TI and a VG 354 mass spectrometer in two individual laboratories as shown in Table 8. With testing a linear law and an exponential law on the two sets of data, it turned out the linear law provided the best fitting (R2 = 0.999999) as shown in Figure 3. The correction factors (K) of mass discrimination were calculated with the linear equation (eq 1.4): K37/35 = 37

r(37Cl/ 35Cl)meas r(37Cl/ 35Cl)calc

(1.4)

35

where r( Cl/ Cl)calc is the calculated chlorine isotope-amount ratio with eq 1.3 as listed in Table 7 and r(37Cl/35Cl)meas is the measured chlorine isotope-amount ratio determined from the gravimetric synthetic solutions. By assessing the correction factors obtained from two mass spectrometers, it turns out that the uncertainty of the average K value from sample Nos. 2−7 is obviously superior to that taken from all samples. It indicates that the effect of loading blank on the determined R(37Cl/35Cl) values in sample No. 1, No. 8, and two enriched chlorine isotopes (Na37Cl and Na35Cl) is remarkable, especially for the data from mass spectrometer of VG 354. In order to eliminate possible mass fractionation due to different conditions, the same experimental conditions for each sample were controlled critically in two individual laboratories. The effect of loading blank on sample Nos. 2−7 with natural R(37Cl/35Cl) ratios is negligible as evidenced by two identical Kmean values, i.e., 0.9982 ± 0.0006 from mass Triton TI and 0.9986 ± 0.0008 from mass VG 354. Therefore, among eight gravimetric synthetic mixtures, sample No. 1 with R(37Cl/35Cl) ratio of ∼0.1 (i.e., 37Cl (%): 10%) and No. 8 with the ratio of ∼9.0 (i.e., 37Cl (%): 90%) were designed for correcting the chlorine isotope R(37Cl/35Cl) ratios in Na37Cl and Na35Cl spikes, while sample Nos. 2−7 with R(37Cl/35Cl) ratio of ∼0.3 (i.e., close to the isotopic abundance of chlorine in nature) were assigned to correct R(37Cl/35Cl) ratios in natural standard reference materials.



RESULTS AND DISCUSSION Recovery Assay and Isotopic Fractionation Examination of Ion-Exchange Procedure. In order to check the recovery assay of the ion-exchange procedure, the elution curve of chlorine ions (i.e., Cl−) during the elution step has been examined to achieve full conversion of NaCl into CsCl and avoid possible isotopic fractionation. The eluted solution from each step was collected in individual vials, and chlorine content was determined using ion chromatography with Metrosep A Supp 4250 type column.32 As shown in Figure 2, the concentration of Cl− ions increases initially and decreases later with an increase of elution volume and gradually reaches a blank level, and an average recovery of ∼98.96% was obtained with the procedure, which indicates that the general elution volume of ∼2 mL ensures the full recovery of ∼2.0 mg of Cl− ions. Calculation of the Correction Factor of Mass Discrimination. As chlorine has only two isotopes and other single ions (or polyatomic ions) have not been detected with the sample 10354

dx.doi.org/10.1021/ac302498q | Anal. Chem. 2012, 84, 10350−10358

Analytical Chemistry

Article

Table 8. Determined Isotopic Ratios of Chlorine in the Gravimetric Synthetic Mixtures by Two Mass Spectrometers R(37Cl/35Cl)measda

R(37Cl/35Cl)calcd sample

a

Triton TI

VG354

Triton TI

1 2 3 4 5 6 7 8

0.10553 0.29422 0.29432 0.31231 0.33388 0.35418 0.39349 8.70798

0.10572 0.29423 0.29433 0.31230 0.33384 0.35412 0.39337 8.63754

0.10532 (0.00013) 0.29375 (0.00011) 0.29382 (0.00020) 0.31199 (0.00054) 0.33339 (0.00010) 0.35364 (0.00017) 0.39241 (0.00021) 8.66227 (0.00465)

mean (SD)

nos. 1, 8 nos. 2, 3, 4, 5, 6, 7

correction factor (K) VG354

Triton TI

0.10550 (0.00011) 0.29402 (0.00028) 0.29417 (0.00126) 0.31159 (0.00016) 0.33349 (0.00040) 0.35356 (0.00005) 0.39237 (0.00013) 8.62780 (0.01340)

0.9980 0.9984 0.9983 0.9989 0.9985 0.9985 0.9973 0.9948

0.9963 ± 0.0020 0.9982 ± 0.0006

VG354 0.9979 0.9993 0.9995 0.9977 0.9990 0.9987 0.9974 0.9983 0.9981 ± 0.0003 0.9986 ± 0.0008

The standard deviations of the determined 37Cl/35Cl ratios (2σ, n ≥ 6).

Na35Cl, respectively, and the mean correction factor was obtained from mixture solution Nos. 2−7 with near natural R(37Cl/35Cl) ratios of ∼0.3. After substituting R′A‑37 and R′B‑35 into eq 1.3, the iterative calculation turns out new calculated isotopic ratios R(37Cl/35Cl) in the gravimetric synthetic mixtures and so on. The iterative calculation continues until the relative difference between two consecutive K37/35 values is less than 10−7. As a result, the true isotopic compositions in two enriched isotope materials Na37Cl and Na35Cl were obtained after five iterations as shown in Table 9, and the mean correction factors of 0.99806 and 0.99852 by Mass Triton TI and Mass VG 354 would be adopted to correct chlorine isotope ratios in standard reference materials that were determined under exactly the same measurement conditions as the synthetic mixtures in the same analytical session. Absolute Isotope Ratio and Atomic Weight of Chlorine. Three isotope reference samples (i.e., NIST 975, NIST 975a, and ISL 354 NaCl) were selected for isotopic composition determination in this work. The data of corrected R(37Cl/35Cl) ratios of three standard materials are given in Tables 10, 11, and 12. The absolute R(37Cl/35Cl) ratios for NIST SRM 975 and NIST 975a by the high-precision measurement are 0.319876 ± 0.000067 and 0.319768 ± 0.000187, and that of 0.319549 ± 0.000044 for ISL 354 from this work is in a good agreement but with a higher precision when compared with the value of 0.319644 ± 0.00917 given by Xiao et al. in 2002.29 On the basis of the absolute measurement, the atomic weights of chlorine in NIST 975, NIST 975a, and ISL 354 are calculated as 35.45284(8), 35.45272(21), and 35.45252(2) individually, which are consistent with the issued values of 35.4527(9) and 35.453(2) by IUPAC in 1985 and 1999.1 Uncertainty Evaluation. According to the recommendations for evaluation of isotope-ratio-measurement uncertainty in “Atomic weights of the elements 2007” (IUPAC Technical Report),38 all certainties reported in this paper were evaluated including random error and potential systematic errors. For the uncertainty evaluation of chlorine isotopic ratios, the internal precision, external reproducibility of measurement, and the correction factors were taken into account. In this study, the uncertainties related to the purity of Na37Cl and Na35Cl salts have been reduced effectively with high accuracy determination of sodium chlorine content by the standard method (i.e., the current constant coulometric titration). The uncertainties occurring during the weighing process also were scaled down by designing a weighing container and weighing with high precision balances. The systematic errors in the measurement

Figure 3. The n(37Cl)/n(35Cl)Determined vs n(37Cl)/n(35Cl)Calculated in the gravimetric synthetic solutions (Nos. 2−7) by MS Triton TI and MS VG 354.

As shown in Scheme 2, initially, the calculated isotopic ratios R(37Cl/35Cl) in the gravimetric synthetic mixtures are derived from the determined R(37Cl/35Cl) values (or atomic abundance) in two spikes, Na37Cl (i.e., RA‑37) and Na35Cl ((i.e., RB‑35), as shown in Table 9. Then, the correction factors K37/35 obtained from mixture solution No. 8 with R(37Cl/35Cl) ratio of ∼9.0 and No. 1 with R(37Cl/35Cl) ratio of ∼0.1 are used to correct the true isotopic composition in two highly enriched spikes, Na37Cl and 10355

dx.doi.org/10.1021/ac302498q | Anal. Chem. 2012, 84, 10350−10358

Analytical Chemistry

Article

Scheme 2. Schematic Diagram for Deriving the Mean Correction Factor from the Iterative Calculation on the Basis of Calculated and Determined R(37Cl/35Cl) Values in Gravimetric Synthetic Mixturesa

a

The iterative calculation continues until the relative difference between two consecutive K37/35 values is less than 10−7.

Table 9. Corrected Isotopic Composition of Highly Enriched Na37Cl and Na35Cl from 5 Cycles of Iterative Calculation for Triton TI and VG 354 Na37Cl Triton

Na35Cl VG 354

Triton

KMean

VG 354

ordinal number of iteration

x(37Cl) %

x(35Cl) %

x(37Cl) %

x(35Cl) %

x(35Cl) %

x(37Cl) %

x(35Cl) %

x(37Cl) %

Triton

VG 354

0 1 2 3 4 5

98.18819 98.19227 98.19887 98.19915 98.19918 98.19920

1.81181 1.80273 1.80113 1.80095 1.80082 1.80080

98.10509 98.10827 98.10885 98.10896 98.10898 98.10898

1.89491 1.89173 1.89115 1.89104 1.89102 1.89102

99.85671 99.85637 99.85636 99.85636 99.85636 99.85636

0.14329 0.14362 0.14364 0.14364 0.14364 0.14364

99.83388 99.83353 99.83352 99.83352 99.83352 99.83352

0.16612 0.16647 0.16648 0.16648 0.16648 0.16648

0.9982221 0.9980865 0.9980638 0.9980603 0.9980598 0.9980598

0.9985912 0.9985344 0.9985261 0.9985245 0.9985245 0.9985245

Table 10. Corrected Isotopic Ratios of Chlorine in Standard Reference Materials Mass 1 (Triton TI)a

Mass 2 (VG 354)b

materials

R(37Cl/35Cl)measuredc

R(37Cl/35Cl)corrected

R(37Cl/35Cl)measuredc

R(37Cl/35Cl)corrected

NIST 975 NIST 975 a ISL 354 (stored) ISL 354-31 ISL 354-141 ISL 354-237 ISL 354−327 ISL 354-424 ISL 354 mean

0.319208 (0.000165) 0.319280 (0.000204) 0.318988 (0.000131) 0.318816 (0.000306) 0.318940 (0.000218) 0.318945 (0.000064) 0.318937 (0.000043) 0.318764 (0.000146) 0.318898 (0.000087)

0.319828 0.319900 0.319608 0.319436 0.319560 0.319565 0.319557 0.319383 0.319518 (0.000087)

0.319450 (0.000067) 0.319163 (0.000054) 0.319045 (0.000036) 0.319120 (0.000053) 0.319098 (0.000108) 0.319191 (0.000053) 0.319061 (0.000074) 0.319126 (0.000077) 0.319107 (0.000052)

0.319923 0.319636 0.319518 0.319593 0.319571 0.319664 0.319534 0.319599 0.319580 (0.000052)

a

The correction factor K37/35 for Mass 1 (Triton TI) is 0.99806. bThe correction factor K37/35 for Mass 2 (VG 354) is 0.99852. cThe overall uncertainties are derived based on 95% confidence limits for the mean value.



caused by the procedure blank and the stability of the instruments were reflected in the internal precision and external reproducibility of the measurement of chlorine isotopic compositions. The isotopic abundance variations in natural standard materials were taken into account for the uncertainty evaluation of chlorine atomic weight.

CONCLUDING REMARKS

The absolute isotopic composition of chlorine was determined by high-precision PTIMS-Cs2Cl+-graphite method with two thermal ionization mass spectrometers in two individual laboratories. The absolute mass spectrometry method described in this work ensures high precision measurement of chlorine 10356

dx.doi.org/10.1021/ac302498q | Anal. Chem. 2012, 84, 10350−10358

Analytical Chemistry

Article

Table 11. Results of Absolute Isotopic Abundance and Atomic Weight of Chlorine in Three International Standard Materials Determined by Two Mass Spectrometersa Mass 1 (Triton TI)

a

Mass 2 (VG 354)

materials

x(37Cl) %

x(35Cl) %

atomic weight

x(37Cl) %

x(35Cl) %

atomic weight

NIST 975 NIST 975 a ISL 354

0.242325515 0.242366846 0.242199199

0.757674485 0.757633154 0.757800801

35.45278881 35.45287135 35.45253655

24.238004793 24.221527755 24.218311887

75.761995207 75.778472245 75.781688113

35.45289771 35.45256866 35.45250443

The atomic mass of 37Cl and 35Cl are 36.96590259u and 34.96885268u, respectively.37

020614330005), and the grant (2009-II-8) from the State Key Laboratory for Mineral Deposits Research (Nanjing University).

Table 12. Mean Values of Absolute Isotopic Composition and Atomic Weight of Chlorine in Three International Standard Materials materials

absolute R(37Cl/35Cl) (2σ)

atomic weight (2σ)

NIST 975 NIST 975a ISL 354

0.319876 ± 0.000067 0.319768 ± 0.000187 0.319549 ± 0.000044

35.452843 ± 0.000077 35.452720 ± 0.000214 35.452520 ± 0.000023



(1) Coplen, T. B. Pure Appl. Chem. 2001, 73 (4), 667−683. (2) Wieser, M. E.; Coplen, T. B. Pure Appl. Chem. 2011, 83 (2), 359− 396. (3) Shields, W. S.; Murphy, E. T.; Garner, E. L.; Dibeler, V. H. J. Am. Chem. Soc. 1962, 84, 1519−1522. (4) Rosman, K. J. R.; Taylor, P. D. P. Pure Appl. Chem. 1998, 70, 217− 235. (5) Banks, D. A.; Green, R.; Cliff, R. A.; Yardley, B. W. D. Geochim. Cosmochim. Acta 2000, 64, 1785−1789. (6) Kaufmann, R.; Long, A.; Bentley, H.; Davis, S. Nature 1984, 309, 338−340. (7) Gaudette, H. Geol. Soc. Am. Abstr. Prog. 1990, 22, A173. (8) Eggenkamp, H. G. M.; Kraulen, R.; Koster van Groos, A. F. Geochim. Cosmochim. Acta 1995, 59, 5169−5175. (9) Magenheim, A. J.; Spivack, A. J.; Volpe, C.; Ransom, B. Geochim. Cosmochim. Acta 1994, 58, 3117−3121. (10) Volpe, C.; Spivack, A. J. Geophys. Res. Lett. 1994, 21, 1161−1164. (11) Godon, A.; Jendrzejewsk, N. Chem. Geol. 2004, 207, 1−12. (12) Stewart, M. A.; Spivack, A. J. Rev. Miner. Geochem. 2004, 22, 231− 254. (13) Willmore, C. C.; Boudreau, A. E.; Spivack, A. J.; Kruger, F. J. Chem. Geol. 2002, 182, 503−511. (14) Ransom, B.; Spivack, A. J.; Kastner, M. Geology 1995, 23, 715− 718. (15) Xiao, Y. K.; Zhang, C. G. Int. J. Mass Spectrom. Ion Proc. 1992, 116, 183−192. (16) Henry, H.; Andersson, P. Anal. Chem. 2004, 76 (8), 2336−2342. (17) Numata, M.; Nakamura, N.; Koshikawa, H.; Terashima, Y. Anal. Chim. Acta 2002, 455, 1−9. (18) Aston, F. W. Mass Spectra and Isotopes; Longmans Green and Co.: New York, 1941. (19) Owen, H. T.; Schacffer, O. A. J. Am. Chem. Soc. 1955, 77, 898− 903. (20) Hoering, T. C.; Parker, P. L. Geochim. Cosmochim. Acta 1961, 23, 186−199. (21) Taylor, J. W.; Grimsrud, E. P. Anal. Chem. 1969, 41, 805−810. (22) Long, A.; Estone, C. J.; Kaufmenn, R. S.; Martin, J. G.; Wir, L.; Fenley, J. B. Geochim. Cosmochim. Acta 1993, 57, 2907−2912. (23) Rosenbaum, J. M.; Cliff, R. A.; Coleman, M. L. Anal. Chem. 2000, 72, 2261−2264. (24) Holmstrand, H.; Andersson, P.; Gustafsson, O. Anal. Chem. 2004, 76, 2336−2342. (25) Holmstrand, H.; Gadomski, D. Environ. Sci. Technol. 2006, 40, 3730−3735. (26) Shirodkar, P. V.; Xiao, Y. K.; Sarkar, A.; Dalal, S. G.; Chivas, A. R. Environ. Int. 2006, 32 (2), 235−239. (27) Tan, H. B.; Xiao, Y. K. Sci. China, Ser. D 2005, 35 (3), 235−240. (28) Xiao, Y. K.; Zhou, Y. M.; Liu, W. G.; Hong, A. S.; Wang, Q. Z.; Wang, Y. H.; Wei, H. Z.; Shirodkar, P. V. Geol. Rev. 2002, 48 (supp), 264−270. (29) Xiao, Y. K.; Zhou, Y. M.; Liu, W. G.; Wang, Y. H.; Wei, H. Z. Chem. Geol. 2002, 182, 655−661.

isotopic ratios in Na37Cl and Na35Cl spikes and synthetic isotope mixtures as well as in natural standard reference materials. After eliminating the weighing error caused from evaporation by designing a special weighing container and accurately determining the chlorine contents in two highly enriched Na37Cl and Na35Cl solutions by the current constant coulometric titration, one series of gravimetric synthetic mixtures with R(37Cl/35Cl) values ranging from ca. 0.1 to ca. 9.0 has been prepared. The experiment on recovery assay confirms the full conversion of NaCl into CsCl with the ion-exchange procedure. The isotopic compositions of chlorine in a whole bunch of samples have been determined by two individual mass spectrometers. The corrected isotopic ratios were used to calculate the absolute isotopic abundance and atomic weight of chlorine in natural standard materials with the typical correction factors that were calculated by an iteration cycle method. The absolute R(37Cl/35Cl) ratios for NIST SRM 975, NIST 975a, and ISL 354 by the precise calibrated isotopic composition measurements are 0.319876 ± 0.000067, 0.319768 ± 0.000187, and 0.319549 ± 0.000044. As a result, the atomic weights of chlorine in NIST 975, NIST 975a, and ISL 354 are derived as 35.45284(8), 35.45272(21), and 35.45252(2) individually, which are consistent with the issued value of 35.453(2) by IUPAC in 1999.



REFERENCES

AUTHOR INFORMATION

Corresponding Author

*Address: Department of Earth Science, Nanjing University, 22 Hankou Road, Nanjing 210093, Jiangsu, PR China. Phone: +86 (25) 83596832. Fax: +86 (25) 83592393. E-mail: shyjiang@nju. edu.cn. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We are grateful for the suggestions from two anonymous reviewers and Prof. T. G. M. van de Ven in Department of Chemistry, McGill University, for exhaustive grammatical editing. This study is supported by the National 973 project (2012CB416706), the National Natural Science Foundation of China (No. 41073002 and No. 40973002), the Fundamental Research Funds for the Central Universities (No. 10357

dx.doi.org/10.1021/ac302498q | Anal. Chem. 2012, 84, 10350−10358

Analytical Chemistry

Article

(30) IAEA Reference products for environment and trade, ISL-354, Sodium Chloride, Material with known 37Cl/35Cl isotopic composition. Website: http://nucleus.iaea.org/rpst/ReferenceProducts/ ReferenceMaterials/Stable_Isotopes/37Cl35l/ISL-354.htm. Accessed 11/2012. (31) Eyring, H. Physical Chemistry, Academic Press, 1975. (32) Mou, S. F.; Liu, K. N.; Ding, X. J. Method and application of ion chromatography; 2nd, Chemistry Industry Press: Beijing, 2005. (33) Hart, S. R.; Zindler, A. Int. J. Mass Spectrom. Ion Proc. 1989, 89, 287−301. (34) Kasemann, S. A.; Jeffcoate, A. B.; Elliott, T. Anal. Chem. 2005, 77, 5251−5257. (35) Caro, G.; Papanastassiou, D. A.; Wasserburg, G. J. Earth Planet. Sci. Lett. 2010, 296, 124−132. (36) Caro, G.; Bourdon, B.; Birck, J.; Moorbath, S. Geochim. Cosmochim. Acta 2006, 70, 164−191. (37) Audi, G.; Wapstra, A. H.; Thibault, C. Nucl. Phys. 2003, A729, 337−676. (38) Michael, E. W.; Berglund, M. Pure Appl. Chem. 2009, 11, 2131− 2156.

10358

dx.doi.org/10.1021/ac302498q | Anal. Chem. 2012, 84, 10350−10358