Article pubs.acs.org/ac
Chemical Separation of Mo and W from Terrestrial and Extraterrestrial Samples via Anion Exchange Chromatography Yuichiro Nagai* and Tetsuya Yokoyama Department of Earth and Planetary Sciences, Tokyo Institute of Technology, Ookayama, Tokyo 152-8551, Japan S Supporting Information *
ABSTRACT: A new two-stage chemical separation method was established using an anion exchange resin, Eichrom 1 × 8, to separate Mo and W from four natural rock samples. First, the distribution coefficients of nine elements (Ti, Fe, Zn, Zr, Nb, Mo, Hf, Ta, and W) under various chemical conditions were determined using HCl, HNO3, and HF. On the basis of the obtained distribution coefficients, a new technique for the two-stage chemical separation of Mo and W, along with the group separation of Ti−Zr−Hf, was developed as follows: 0.4 M HCl−0.5 M HF (major elements), 9 M HCl−0.05 M HF (Ti−Zr−Hf), 9 M HCl−1 M HF (W), and 6 M HNO3−3 M HF (Mo). After the chemical procedure, Nb remaining in the W fraction was separated using 9 M HCl−3 M HF. On the other hand, Nb and Zn remaining in the Mo fraction were removed using 2 M HF and 6 M HCl−0.1 M HF. The performance of this technique was evaluated by separating these elements from two terrestrial and two extraterrestrial samples. The recovery yields for Mo, W, Zr, and Hf were nearly 100% for all of the examined samples. The total contents of the Zr, Hf, W, and Mo in the blanks used for the chemical separation procedure were 582, 9, 29, and 396 pg, respectively. Therefore, our new separation technique can be widely used in various fields of geochemistry, cosmochemistry, and environmental sciences and particularly for multi-isotope analysis of these elements from a single sample with significant internal isotope heterogeneities.
R
the surface environment in the Archean and Paleoproterozoic.1,2 The Mo isotope system has also been applied to study isotope heterogeneities in the early solar system. It was found that Mo isotopic compositions in a variety of meteorites were not homogeneous at the ε-level, suggesting a heterogeneous Mo isotope distribution on a planetary scale in the early solar nebula because of the variable contributions of components derived from different nucleosynthetic sources.3,4 A third major application of W isotopes in geochemistry is the evaluation of the short-lived 182Hf−182W system (182Hf half-life = 8.9 Myr) in chondrites and differentiated meteorites to determine the timing of core formation and silicate differentiation in planets and planetesimals.5 In particular, the determination of marginal excesses or deficits of 182W in some Archean rocks provides important clues to understanding the evolution of the Earth’s mantle.6,7 In addition, some iron meteorites have nucleosynthetic W isotope anomalies that generally correlate with their Hf−W ages.8 The extent of natural mass-dependent isotope fractionation is typically at a level of a few per mill, while isotope anomalies in meteorites are sometimes only at marginal levels ( 48. The breakthrough for an element is essentially negligible when 0.1 × Vmax mL of the reagent is eluted. According to Table 2, the distribution coefficients for Ti, Zr, and Hf were relatively high (Kd > 50) for 6 M HCl, 0.4 M HCl−0.5 M HF, and 1 M HF (excluding Kd (Hf) for 6 M HCl), whereas the Kd values for these elements were sufficiently low ( 50) for 6 M HCl, 9 M HCl−0.01 M HF, 9 M HCl−0.05 M HF, 0.4 M HCl−0.5 M HF, and 1 M HF, whereas the Kd value for W was low (103) for Zr and Hf would make it difficult to remove these elements from the resin bed in the next step. Thus, 1 M HF was not chosen as the loading solution. After sample loading, Ti, Zr, and Hf were separated from W and Mo. It was observed that the Kd values for Ti, Zr, and Hf were 50 when 9 M HCl−
Table 3. Procedures of Chemical Separation Technique procedure
reagents
volume [mL]
First column: Separation of Ti-Zr-Hf, Mo, and W (1 mL of Eichrom 1 × 8, 200−400 Mesh) load sample (major elements) 0.4 M HCl + 0.5 M HF 1.0 wash (major elements) 0.4 M HCl + 0.5 M HF 3.0 recover Ti, Zr, and Hf 9 M HCl + 0.05 M HF 5.0 recover W 9 M HCl + 1 M HF 10.0 recover Mo 6 M HNO3 + 3 M HF 5.0 Second column for W: Separation of W and Trace Nb (0.1 mL of Eichrom 1 × 8, 200−400 Mesh) load sample 0.4 M HCl + 0.5 M HF 0.3 wash 0.4 M HCl + 0.5 M HF 0.4 recover W 9 M HCl + 3 M HF 1.2 Second column for Mo: Separation of Zn, Nb, and Mo (0.1 mL of Eichrom 1 × 8, 200−400 Mesh) load sample (remove Zn) 2 M HF 0.3 wash (remove Zn) 2 M HF 0.4 remove Nb 6 M HCl + 0.1 M HF 2.0 recover Mo 6 M HNO3 + 3 M HF 1.0 4859
dx.doi.org/10.1021/ac404223t | Anal. Chem. 2014, 86, 4856−4863
Analytical Chemistry
Article
Figure 1. Elution profiles using the first-column procedure for terrestrial and extraterrestrial samples. (a) JB-3, basalt; (b) JLk-1, lake sediment; (c) Charcas, iron meteorite; (d) Allende, carbonaceous chondrite.
first and second eluents, a minuscule breakthrough of W for the Allende sample (98.3%, Table 4. Recovery Yields of Zr, Mo, Hf, and W Zr [%] JB3 JLk-1
Figure 2. Elution profiles using the W-second chemical separation procedure for the W−Nb mixed solution.
Charcas Allende
that for the first column (0.4 M HCl−0.5 M HF) in order to completely remove some unwanted elements (e.g., Fe) remaining in the W fraction. Subsequently, W was collected while adsorbing Nb onto the resin. It appeared that Nb was retained on the resin until >1.4 mL of 9 M HCl−3 M HF was eluted. In this procedure, elements that cause isobaric interference for W isotope analysis (e.g., Os) may not be removed completely. However, ionization of Os during N-TIMS does not significantly affect the W isotope analysis, because Os (OsO3−) can be evaporated completely above 1000 °C,23 while ionization of W (WO3−) occurs above 1300 °C.12 In the case of multicollector ICPMS (MC-ICPMS), an additional treatment is required to remove Os before isotope analysis (e.g., evaporation of volatile OsO4 using an oxidizing acid such as HClO4). Molybdenum was separated from Zn and Nb in the Mosecond stage (Figure 3). During the loading and washing step (2 M HF), Zn was not adsorbed onto the resin because Zn behaves as a cation (Zn2+) in HF solution.22 Although Nb could be removed in the next step (6 M HCl−0.1 M HF), Mo breakthrough (>1%) occurred when more than 2.5 mL of this solution was eluted. Tungsten that existed in the Mo fraction from the first column was separated from the Mo in this step. The total recovery of W could possibly increase by merging this fraction with the W fraction from the first column before the W-second stage. With regard to the potential for isobaric interference during the Mo isotope analysis, nearly 100% of the Zr was removed in the first column and the Mo-second stage
Mo [%]
100.4 ± 0.3 99.6 ± 1.3 99.4 ± 1.6 98.4 ± 0.5
Hf [%]
Terrestrial Rocks 101.7 ± 0.6 100.0 ± 0.7 103.0 ± 2.3 101.1 ± 1.7 Meteorites 101.6 ± 1.2 n.d. 99.8 ± 1.3 98.3 ± 1.4
W [%] 102.0 ± 1.0 98.7 ± 0.9 96.2 ± 2.2 97.3 ± 3.6
and >96.2% for Zr, Mo, Hf, and W, respectively. The Hf yield for Charcas was not determined because the abundance of lithophile elements in iron meteorites is generally negligible. Although the recovery of Ti was not determined in this study, it behaves like Zr and Hf, and thus, the recovery is expected to be similar to those for these elements. The recovery of Mo was nearly 100% for all of the samples when analytical uncertainties were taken into account. In contrast, the other elements, particularly W, could not be recovered completely in some cases (∼96% for Charcas and ∼97% for Allende). As is described above, the lower recovery was due to the breakthrough of W into the Mo fraction in the first column procedure, which was then recovered by collecting the Nb fraction (6 M HCl−0.1 M HF) in the Mo-second stage. Notably, the recovery yields for Zr, Hf, and W for the Allende sample were less than 100%. The reason for these low recoveries is unclear at this point. However, they may be due to the carbonaceous characteristics of this meteorite, which contains a significant amount of organic matter compared to terrestrial rocks. To evaluate the capacity of the column separation procedure developed in this study, the elution profiles and recovery yields for Zr, Mo, Hf, and W were determined using 250−1000 mg of the JB-3 rock. For this test, the same separation conditions were employed (i.e., resin volumes, amount of reagents, etc.), other than the volume of the loading solution for the first column (1 4861
dx.doi.org/10.1021/ac404223t | Anal. Chem. 2014, 86, 4856−4863
Analytical Chemistry
Article
two-stage chemical separation of Mo and W, along with the group separation of Ti−Zr−Hf, was developed. The performance of this new technique was evaluated by separating these elements from several terrestrial and extraterrestrial samples. The recovery yields for Mo, W, Zr, and Hf were nearly 100% for these samples. Therefore, our new separation technique can be widely used in various fields of geochemistry, cosmochemistry, and environmental sciences, which require multi-isotope analysis of these elements using a single sample.
mL of 0.4 M HCl−0.5 M HF per 250 mg of sample). The elution profiles for larger sample sizes (not shown) were generally consistent with those obtained when 80−120 mg of terrestrial and extraterrestrial samples were used (Figures 1−3). In fact, nearly 100% Zr, Hf, and W were recovered in all cases, although the recovery of Mo decreased slightly (Table 5). The Table 5. Recovery Yields of Zr, Mo, Hf, and W for Larger Amounts of JB-3 amount of JB3, mg 250 500 750 1000
Zr [%] 98.9 100.3 100.2 99.6
± ± ± ±
1.2 1.7 1.7 2.8
Mo [%] 97.2 93.0 96.3 94.4
± ± ± ±
1.8 1.4 1.9 2.1
Hf [%] 99.7 100.4 100.6 99.9
± ± ± ±
1.6 1.7 1.6 2.0
■
W [%] 98.5 97.2 97.9 98.2
± ± ± ±
ASSOCIATED CONTENT
S Supporting Information *
2.5 1.9 2.3 3.6
Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
decrease in the Mo recovery may be due to the formation of an unknown Mo complex that acts differently than the majority (∼95%) of the Mo during the column chemistry procedure. It is important to note that the extent of Mo isotope fractionation due to the loss of 5% Mo during separation should be examined when investigating natural mass-dependent Mo isotope fractionation, although such an investigation was beyond the scope of this study. The concentrations of Zr, Hf, W, and Mo in the procedural blanks used for the chemical separation procedure were 582, 9, 29, and 396 pg, respectively. The procedural blank for Mo was comparable to the blanks described in Burkhardt et al. (0.1−1 ng of “total” procedural blanks),3 while the blank for W was slightly higher than that reported in Kleine et al. (sub 10 pg, using 250−700 mg of powdered rock samples).13 Application of the Chemical Separation Technique for Isotope Analysis. The developed chemical separation technique can be applied for the multiple-isotope analysis of various types of rock samples, including igneous rocks, sediments, and extraterrestrial materials. Because of the efficient removal of unwanted elements and the lower concentrations of elements in the chemical blanks compared to those used in previous studies, this method is suitable for analyzing meteorites (and their components) that have extremely low Mo and W concentrations (e.g., achondrites and chondrules). Moreover, this method is specifically useful for determining multielemental isotope compositions in a single sample with internal isotope heterogeneities. For example, primitive chondrites are known to carry isotopically anomalous presolar grains with sizes of a few μm or less. In some cases, duplicate analyses of a single chondrite sample give inconsistent isotope data because different batches of the sample powder include different amounts of presolar grains, particularly when the amount of sample is extremely small (i.e., the nugget effect). Therefore, in order to determine the isotopic correlations between different elements, simultaneous analysis of the isotope compositions of multiple elements from a single batch of a sample is strongly recommended. Thus, our new technique is concluded to be useful for the isotope analysis of Mo and W, as well as Ti, Zr, and Hf.
*E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS We would like to thank R. J. Walker for his assistance and encouragement during the early stage of this study. This research was supported by a grant for the Global COE Program, “From the Earth to “Earths”,” from the Ministry of Education, Culture, Sports, Science, and Technology of Japan, and KAKENHI Grant Number 23340171 from the Japan Society for the Promotion of Science (JSPS).
■
REFERENCES
(1) Duan, Y.; Anbar, A. D.; Arnold, G. L.; Lyons, T. W.; Gordon, G. W.; Kendall, B. Geochim. Cosmochim. Acta 2010, 74, 6655−6668. (2) Kendall, B.; Gordon, G. W.; Poulton, S. W.; Anbar, A. D. Earth Planet. Sci. Lett. 2011, 307, 450−460. (3) Burkhardt, C.; Kleine, T.; Oberli, F.; Pack, A.; Bourdon, B.; Wieler, R. Earth Planet. Sci. Lett. 2011, 312, 390−400. (4) Burkhardt, C.; Kleine, T.; Dauphas, N.; Wieler, R. Earth Planet. Sci. Lett. 2012, 357−358, 298−307. (5) Kleine, T.; Mezger, K.; Münker, C.; Palme, H.; Bischoff, A. Geochim. Cosmochim. Acta 2004, 68, 2935−2946. (6) Willbold, M.; Elliott, T.; Moorbath, S. Nature 2011, 477, 195− 199. (7) Touboul, M.; Puchtel, I. S.; Walker, R. J. Science 2012, 335, 1065−1069. (8) Schulz, T.; Münker, C.; Peters, S. T. M. Earth Planet. Sci. Lett. 2010, 362, 246−257. (9) Kraus, K. A.; Nelson, F.; Moore, G. E. J. Am. Chem. Soc. 1955, 77, 3972−3977. (10) Kronenberg, A.; Mohapatra, P. K.; Kratz, J. V.; Pfrepper, G.; Pfrepper, R. Radiochim. Acta 2004, 92, 395−403. (11) Dauphas, N.; Reisberg, L.; Marty, B. Aanal. Chem. 2001, 73, 2613−2616. (12) Touboul, M.; Walker, R. J. Int. J. Mass Spectrom. 2011, 309, 109−117. (13) Kleine, T.; Hans, U.; Irving, A. J.; Bourdon, B. Geochim. Cosmochim. Acta 2012, 84, 186−203. (14) Kruijer, T.; Sprung, P.; Kleine, T.; Leya, I.; Burkhardt, C.; Wieler, R. Geochim. Cosmochim. Acta 2012, 99, 287−304. (15) Qin, L.; Dauphas, N.; Janney, P. E.; Wadhwa, M. Anal. Chem. 2007, 79, 3148−3154. (16) Pietruszka, A. J.; Reznik, A. D. Int. J. Mass Spectrom. 2008, 270, 23−30. (17) Makishima, A.; Nakamura, E.; Nakano, T. Anal. Chem. 1997, 69, 3754−3759.
■
CONCLUDING REMARKS The distribution coefficients of nine elements were determined for an anion exchange resin, Eichrom 1 × 8, under various chemical conditions using HCl, HNO3, and HF. On the basis of the obtained distribution coefficients, a new technique for the 4862
dx.doi.org/10.1021/ac404223t | Anal. Chem. 2014, 86, 4856−4863
Analytical Chemistry
Article
(18) Münker, C.; Weyer, S.; Scherer, E.; Mezger, K. Geochem. Geophys. Geosyst., 2001, 2, DOI: 10.1029/2001GC000183. (19) Schönbächler, M.; Lee, D.-C.; Rehkämper, M.; Halliday, A. N.; Fehr, M. A.; Hattendorf, B.; Günther, D. Earth Planet. Sci. Lett. 2003, 216, 467−481. (20) Sprung, P.; Scherer, E. E.; Upadhyay, D.; Leya, I.; Mezger, K. Earth Planet. Sci. Lett. 2010, 295, 1−11. (21) Leya, I.; Schönbächler, M.; Wiechert, U.; Krähenbühl, U.; Halliday, A. N. Earth Planet. Sci. Lett. 2008, 266, 233−244. (22) Makishima, A.; Tanaka, R.; Nakamura, E. Anal. Sci. 2009, 25, 1181−1187. (23) Völkening, J.; Walczyk, T.; Heumann, K. G. Int. J. Mass Spectrom. Ion Phys. 1991, 105, 147−159. (24) Nagai, Y.; Yokoyama, T. Molybdenum Isotope Anomalies in Allende and Murchison Meteorites. 44th Lunar and Planetary Science Conference, The Woodlands, Texas, March 18−22, 2013, #2373. (25) Imai, N.; Terashima, S.; Itoh, S.; Ando, S. Geostand. Newsl. 1995, 19 (2), 135−213. (26) Jarosewich, E.; Clarke, R., Jr.; Barrows, J. In Smithsonian Contributions to the Earth Sciences; Smithsonian Institution Press: Washington, DC, 1987; Vol. 27. (27) Moore, C. B.; Lewis, C. F.; Nava, D. In Meteorite Research; Millman, P. M., Ed.; D. Reidel: Dordrecht, 1969; pp 738−748. (28) Scott, E. R. D.; Wasson, J. T. Geochim. Cosmochim. Acta 1973, 37, 1957−1983.
4863
dx.doi.org/10.1021/ac404223t | Anal. Chem. 2014, 86, 4856−4863
