Investigation of Natural Fractionation of Stable Mercury
Isotopes by Multi-Collector
Inductively
Coupled Plasma Mass Spectrometry
Bjoern Klaue, Stephen E. Kesler, Joel D. Blum
University of Michigan, Geological Sciences, 2534 CCLittle Building, Ann Arbor, MI
Abstract
Mercury is an intriguing element to study because it: (a) is highly volatile; (b) has seven stable isotopes with a relatively large mass range (196–204 amu); (c) forms bonds that have a high degree of covalent character; and (d) exists in more than one oxidation state. These factors suggest that it may undergo kinetic isotopic fractionation in natural systems (Nier, 1950). The analysis of the Hg isotopic pattern in environmental samples may offer news ways of source and pathway identification for Hg species. However, neither the true isotopic composition nor the fractionation of Hg isotopes has been studied in detail with modern instrumentation. Multi-Collector Inductively Coupled Plasma Mass Spectrometry (MC-ICPMS) allows us for the first time to analyze the isotopic composition of Hg with a precision of better than 0.01%. We will present the methodology as well as preliminary data of Hg isotopic patterns in ore, coal, and environmental samples.
Methodology
Due to mercury’s volatility and high ionization potential, isotopic measurements cannot be performed by thermal ionization mass spectrometry (TIMS). Analysis by gas source mass spectrometry and single-collector inductively coupled mass spectrometry are typically limited to a precision of 0.1%, which is insufficient for the expected levels of natural isotopic fractionation of Hg. Also, the reported abundances by these methods show large inconsistencies (De Bievre 1984, Zadnik 1989).
MC-ICPMS offers an efficient way to achieve high precision isotopic measurements for elements that were previously impracticable or impossible to analyze (e.g., Halliday 1999). Thus, we developed a cold-vapor generation MC-ICPMS method, which allows high precision Hg isotopic analyses down to the 0.005% precision level. Cold-vapor (Hg0) generation with Sn(II) as the reducing reagent allows the fast and selective chemical separation of mercury from the matrix. External mass fractionation correction is achieved by mixing the sample Hg-vapor online with a dry aerosol of the NIST 997 thallium isotopic standard reference material.
The external inter-element fractionation correction is mainly intended to improve the internal precision of MC-ICPMS measurements but typically fails to also accurately correct for absolute isotopic ratios within 0.01% (Rehkämper, 1998, White 2000). The thallium external correction is a well-established method for lead isotopic measurements but even for the chemically and electronically similar elements Tl and Pb the external correction yields absolute ratios that are slightly different from the certified values (Rehkämper 1998). Because there are no certified Hg isotopic standards to test the accuracy of the measurements, we compared the results of the MC-ICPMS ratios (Tl external correction, exponential law fractionation correction) with single-collector ICPMS data (direct measurement). The Finnigan “ELEMENT” single-collector magnetic sector ICPMS used for this study can be tuned to exhibit no or very little absolute mass fractionation using well known standards of Pb (NIST 981) or Tl (NIST 997). Precision levels of 0.05 to 0.1 % can be achieved. However, the VG Elemental “P54” MC-ICPMS typically shows a mass fractionation of ca. 1% per amu in the 200 (amu) mass region which requires external correction procedures. The absolute Hg isotopic composition measured with both instruments agrees within the 0.1% range, which demonstrates that the external Tl correction procedure yields reasonably accurate Hg isotope ratios for MC-ICPMS.
Due to the lack of a Hg isotopic reference material we have chosen a purified native mercury product form the Alamaden mine in Spain (the world largest known Hg deposit) as a relative “zero” standard. Analyses of a number of cinnabar ore samples from the same mine revealed no significant differences between the Hg(0) product and ore samples from the same mine. Table 1 compares the results for the single- and multi-collector ICPMS analyses of the Almaden mine samples with Hg isotopic data from the literature.
Table 1.: Comparison of the results for the single- and multi-collector ICPMS isotope ratio measurements of the Almaden mine (Spain) reference sample with gas source MS data from the literature (De Bievre 1984)
|
|
IUPAC
84 |
Element |
P54 |
Inghram ‘47 |
Hibbs ’49 |
Nier
‘50 |
Dibeler‘55 |
Zadnik
‘89 |
|
196/202 |
0.00506 |
0.00519 |
0.005192 |
0.0052 |
0.0054 |
0.00494 |
0.005263 |
0.005138 |
|
198/202 |
0.3406 |
0.33860 |
0.337567 |
0.3412 |
0.3372 |
0.3386 |
0.3414 |
0.3338 |
|
199/202 |
0.5734 |
0.57095 |
0.569646 |
0.5735 |
0.5693 |
0.5718 |
0.5732 |
0.5650 |
|
200/202 |
0.7791 |
0.77904 |
0.777929 |
0.7825 |
0.7773 |
0.7753 |
0.7783 |
0.7734 |
|
201/202 |
0.4452 |
0.44497 |
0.442776 |
0.4434 |
0.4448 |
0.4448 |
0.4477 |
0.4414 |
|
204/202 |
0.2293 |
0.23036 |
0.229144 |
0.2256 |
0.2302 |
0.2281 |
0.2291 |
0.2299 |
Table 2: Isotopic composition of a Hg concentration standard material (SPEX) and the Almaden Hg isotopic reference sample
|
Isotope ratio R |
Almaden |
%Error (2 RSE) |
SPEX Std. |
%Error (2 RSE) |
D M (M - 202) |
D ‰ |
|
196/202 |
0.005183 |
0.0127 |
0.005195 |
0.0479 |
-6 |
1.463758 |
|
198/202 |
0.337261 |
0.0061 |
0.337461 |
0.0052 |
-4 |
0.542480 |
|
199/202 |
0.569308 |
0.0052 |
0.569642 |
0.0046 |
-3 |
0.515047 |
|
200/202 |
0.777665 |
0.0071 |
0.777924 |
0.0034 |
-2 |
0.203804 |
|
201/202 |
0.442713 |
0.0053 |
0.442737 |
0.0033 |
-1 |
0.283226 |
|
204/202 |
0.229170 |
0.0078 |
0.229202 |
0.0087 |
2 |
-0.240730 |
Table 2 illustrates the mass-dependent fractionation of a commercially available Hg standard (SPEX CertiPrep, FisherScientific) relative to the Alamaden mine isotopic reference sample. The mass fractionation is expressed as D‰ [= 1000*(1-Rs/Rm)] with Rs being the reference value and Rm the measured value of the sample. The mass dependence of the fractionation is plotted in Figure 1. The linear regression of the fractionation as a function of DM yields a slope of ca 0.02% per amu with a correlation coefficient of 0.935.
Figure 1.: Plot of the mass fractionation of the data from Table 1 as a function of the mass difference
Results
The MC-ICPMS method was applied to a number of terrestrial, extraterrestrial, ore, and environmental samples to establish the range of isotopic variation for Hg. We analyzed cinnabar and native mercury samples from 8 mines throughout the world and found fractionation ranging from 0.064 to -0.12 % for the Hg 198/204 ratio as listed in Table 3.
While the cinnabar samples from most mines displayed only small differences in the 0.01 – 0.03% range, we observed significant fractionation for native Hg and for some of the cinnabar samples. This is consistent with the possibility that partial condensation or evaporation during ore deposition and decomposition is a likely mechanism for Hg fractionation.
Table 3: Comparison of observed isotopic fractionation in D‰ of the Hg198/204 isotope ratio relative to the Almaden mine reference sample.
|
Sample |
D‰
198/204 |
|
error 2s |
|
Almaden
reference |
0.00 |
|
|
|
HgSnAg alloy, Bolivia |
0.00 |
± |
0.07 |
|
Cinnabar, New Almaden, CA |
0.64 |
± |
0.07 |
|
Cinnabar, Salana, Hungary |
0.07 |
± |
0.07 |
|
Cinnabar, Monte Armiata, Italy |
-0.31 |
± |
0.07 |
|
Cinnabar, Kern County, CA |
0.20 |
± |
0.07 |
|
Native Hg, Napa, CA |
0.55 |
± |
0.07 |
|
Cinnabar, Coso Hot Spring, CA |
-0.98 |
± |
0.07 |
|
Cinnabar, Idria, Yugoslavia |
0.04 |
± |
0.07 |
|
Native Hg, Idria, Yugoslavia |
-1.17 |
± |
0.07 |
|
|
|
|
|
|
various fish, Aberjona, MA |
3.10 |
± |
0.2 |
|
DORM2, NRC |
-0.09 |
± |
0.1 |
|
DOLT2, NRC |
0.98 |
± |
0.1 |
Analyses of the certified standard reference materials dogfish muscle and dogfish liver (DORM-2 and DOLT-2, NRC-Canada) did not yield a conclusive result concerning biological fractionation of Hg isotopes. Both standards were produced from the same fish and the major difference between the two materials is the difference in inorganic and methyl mercury concentrations. Dogfish muscle contains basically pure methyl mercury whereas the dogfish liver holds 75% of the total mercury in the inorganic form. The isotopic composition of neither sample displayed a clear correlation of mass fractionation with mass difference. The same problem was observed for the fresh water fish samples from the Aberjona watershed in MA. Odd isotopic patterns for Hg due to nuclear reaction are very unlikely (Heymann 1986). Therefore, we will need to perform additional tests to rule out analytical problems due to the much lower concentration of Hg in the fish samples. We did not observe this problem for any of the highly concentrated ore samples.
Discussion
The first results of Hg isotopic analyses by MC-ICPMS clearly indicate that there is significant fractionation of Hg isotopes in nature. Nevertheless, the degree of fractionation is very small compared to other radiogenic and (light) stable isotope systems. MC-ICPMS is currently the only analytical method capable of resolving those small differences in order to study systematic effects in the environment. We are now in the process of studying numerous sample materials in order to establish the range of terrestrial Hg isotopic values. This includes ores, important technical products, coal, emission and deposition samples, and sediments. Based on these data we hope to exploit the isotopic variations for source and process identification of Hg in the environment
References
Rehkämper
M., Halliday A.N. (1998) Int. J. Mass
Spectrom. 181, 123-133.
Nier,
A.O. Phys. Rev, 79, (1950) 450
De
Bievre P., Gallet M., Holden N.E., Barnes I. L. (1984) J. Phys. Chem. Ref. Data, Vol. 13, No. 3, 877.
Zadnik
M.G., Specht S., Begemann F. (1989) Int.
J. of Mass Spectrom. and Ion Proc. 89, 103-110.
Halliday,
A.N., Christensen, J.N., Lee, D-C., Rehkämper, M., Hall, C.M., and Luo, X.
(1999) In: Inorganic mass spectrometry, fundamentals and applications, Marcel
Dekker Inc.,New York, Chapter 13.
White
M.W., Albarede F., Telouk P (2000) Chemical Geology 167 (2000) 257-70.
Heymann D., Liffman K. (1986) Meteoritics 21 (1) 95-108.