46126-AC8
New Analytical Advancements In K-Ca Geochronology and Applications to Sedimentary Dating
Our research group has made considerable progress toward the goals of measuring radiogenic 40Ca using multi-collector inductively-coupled plasma mass spectrometry (MC-ICP-MS), and dating geologic materials using the K-Ca radioisotope system. The ultimate objective of this research is to apply K-Ca geochronology to authigenic sedimentary minerals in an effort to constrain the ages of deposits that are difficult to date using conventional methods. Recent advancements made by our lab are detailed below and include improvement in the precision and reproducibility of calcium isotope measurements, reduction of critical interferences using ICP-MS, creation and calibration of a new calcium spike, and measurement of radiogenic calcium ratios in a geologic materials. 1. Precision and reproducibility of standard calcium isotope measurements Upon conception of this work, we performed mass spectrometric tests of the calcium isotope ratios of NIST SRM 915b. In keeping with the work done by DePaolo and his group, we chose 40Ca/42Ca as our ratio of interest and measured it repeatedly using standard methods to assess the precision and reproducibility achievable using ICP-MS. Errors on individual analyses were small - about 0.02% at the 2( level. Repeated analysis produced an average 40Ca/42Ca value of 150.70 ( 0.15, which corresponds to a reproducibility of ~ 0.1%, or 10.2 units of (Ca. Maximum reproducibility using thermal ionization mass spectrometry (TIMS) is reported to be 1.5 units of (Ca. We have been working towards improving the reproducibility of measurements using ICP-MS by designing instrumental experiments that gauge the effect of various parameters on ratio stability. We found that reproducibility could be marginally improved through optimization of both the hexapole RF amplitude and the position of the plasma torch. Most importantly, however, we found that use of a desolvating nebulizer greatly improved the repeatability of our measurements, without causing a significant decrease in sensitivity. Together, these changes resulted in an improvement in reproducibility from ( 10.2 to ( 2.9 units of (Ca. Although this is not as good as TIMS work, it is sufficient for the measurement of geologic materials having a suitable K/Ca elemental ratio. The internal precision of each analysis could not be improved. 2. Reduction of critical interferences Both K and Ca isotopes are difficult to measure using ICP-MS because of interferences at masses 39 and 40 from the argon carrier gas. We addressed this problem by attenuating argon molecular ions through gas-phase collisions in the hexapole cell. H2 gas, bled into the hexapole at a rate of 2mL/min, collides with 40Ar+ to form ions and neutral molecules that do not interfere with the masses of interest. At best, we can get the background level at mass 40 to less than 10mV. Because we could not improve upon that background level, we focused instead on increasing solution concentrations and measurement sensitivity such that 10mV becomes an insignificant fraction of the total measured mass 40 signal. For a typical calcium standard analysis, the signal at mass 40 is orders of magnitude higher than background. Through experimentation, we found that introducing a small flow of He gas, in addition to H2, is most effective at reducing argon interferences when measuring K. Although not fully understood, H2 reaction with Ar alone does not decrease backgrounds at mass 41 to the same extent that it does at mass 40, perhaps due to the creation of 38ArH+. Introduction of small amounts of He gas dampens that effect and reduces the interference at mass 39. 4. Ca isotopic measurement of geologic materials and age dating Having overcome most of the major analytical hurdles, we have been working on a suite of samples ranging from Permian evaporites to Proterozoic granites. Although we are still unraveling some problems with data reduction, which has proven difficult given the high degree of instrumental mass fractionation using ICP-MS, the measured 40Ca/42Ca ratios of our samples are what we expect given their known ages and elemental concentrations. In the case of sylvite, which is particularly well-suited for this technique because it does not incorporate "common" calcium, we have produced a K-Ca age (247 Ma) that is in agreement with an Ar-Ar age of 245.5 ( 1.1 Ma from the same deposit (Renne et al., 2001). We anticipate successfully measuring sedimentary minerals, like sylvite, that have sufficiently high K/Ca elemental ratios. Our K-Ca ages of previously dated authigenic glauconites are older than expected, but we think this is likely due to problems with sample preparation and purification of glauconite grains. We hope to not only obtain K-Ca of these sediments, but also to compare them to anomalously young Rb-Sr ages in an effort to determine the relative behaviors of Ca and Sr and document the robustness of the K-Ca method.
