EFFECT OF SULFUR IN MONAZITE ON ELECTRON MICROPROBE AGE ESTIMATES

Accounting for peak interferences in wavelength dispersive spectra is a critical issue for obtaining accurate Th-U-Pb age estimates by electron microprobe, especially for REE-bearing minerals such as monazite. Under-correction for interfering peaks on low-concentration elements will produce younger age estimates for Pb peaks and older estimates for U peaks. The overlapping of peak spectra with background positions will also produce spurious results. Until recently, routine EMP analysis for age determination at the Nagoya University Center for Chronological Research involved measurements on Th, U, Pb and Y peaks, with two-point background measurements for each peak. Peak interferences were corrected and age estimates were obtained using the Chemical Total-Pb-Th-U isochron method (CHIME). More recently, other elements have been included in routine analysis, including Ca and Si (to confirm stoichiometric coupled substitution of Th+Ca and Th+Si in the monazite structure) and K (to eliminate analysis with activation volumes that overlap minerals adjacent to monazite grains).

In addition, the effect of sulfur on age estimates has been considered. It has been observed that monazite data with detectable amounts of sulfur do not fit on an isochron. Measured sulfur can come from galena or sphalerite inclusions; however, substitution of sulfate for phosphate complexes in the monazite structure is also possible. Sulfate-bearing monazites have been reported in carbonatites, for which the double substitution REE3+(PO4)3- = Ca2+(SO4)2- has been proposed.

During our research on detrital monazites from the Upper Silesia Coal Basin in Poland, age populations tended to be younger than the stratigraphic ages of the host sedimentary units. Monazite from the samples were re-analyzed with a new procedure that includes the measurement of Ca, K and S. Analyses with detectable sulfur (> 50 ppm) were found to yield apparent ages younger than S-free analyses. We conclude that S measurements should be included in routine EMP analysis for Th-U-Pb chronology, in order to filter out spurious age data.

This research has been carried out with the support of JSPS grant no P05729.