Ultra-precise titanium stable isotope measurements by double-spike high resolution MC-ICP-MS

In this contribution, we present a new technique for the ultra-precise determination of titanium stable isotope composition (expressed as delta Ti-49 or deviation of the Ti-49/Ti-47 ratio to the reference standard) of geological samples by multi-collection plasma source mass spectrometry (MC-ICPMS) using the double spike method to correct for instrumental mass bias. Tails of polyatomic spectral interferences on Ti-46 are accounted for by using sample-standard bracketing in high-resolution mode. Choice of ideal double and triple spike composition is investigated and results show that analytical error for a single measurement is optimised for a Ti-47-Ti-49 double spike composed of ca. 50% of each spike and mixed with ca. 52% of sample. Measurements of pure Ti solution show that internal error on single measurements of ca. 0.010 parts per thousand (95% c.i.) is attainable on delta Ti-49, in agreement with the error model. Due to the lack of a widely available reference isotopic standard for titanium, all results are expressed as deviations relative to newly created reference material (CL-Ti standing for Origins Laboratory - titanium) prepared from an ultra-pure titanium metal rod. A range of analytical tests demonstrates the robustness of our method. An external reproducibility of ca. 0.020 parts per thousand (2sd) is routinely achievable for Ti stable isotopes. Data for a range of basaltic rock standards as well as a subduction zone basalt reference suite is presented and show that the Ti stable isotope compositions of terrestrial basalt show resolvable variations but are overall very close to the OL-Ti reference standard. The average Ti isotopic composition of the basalts studied here is the present best estimate of the upper mantle composition; delta Ti-49 = +0.004 +/- 0.062 parts per thousand (2sd).