UMass Geosciences Link
Ultrachron

 

Testing

  • Resolution
  • Precision
  • Accuracy

Analytical resolution relevant to geochronologic analysis can be evaluated by boundary testing of sharply sector zoned monazite (right). These tests were performed with an optimized LaB6 cathode after approximately 3000 hrs. of service life. Tests were done at 10 and 15kV, and 40, 100 and 200 nA beam currents. Beam motion is normal to interface, while ThMa intensity is acquired (below). Scan data can be evaluated by second-derivative analysis to determine resolution (lower-right). Resulting resolution measurements are summarized in figure upper right. See Jercinovic et al. (2008) for details.

 

Sector zoned natural monazite. Sectors are defined by Th zoning (ThMa map), with bright and dark differing by ~2wt.% Th.
Measured focused beam analytical resolution (in nanometers) from boundary measurements. LaB6 (3000hrs.). Calculated PbMa resolution is 99.1% of ThMa.
Savitsky-Golay noise filtered data (40, 100, and 200nA and 10kV). ThMa intensity is measured with distance through interface between Th sector zones. All intensities are scaled to 100nA values for comparison. Second derivative analysis of 10kV curves in figure to left, and calculated resolution values for Th at varying current.

Precision is dramatically improved via the developments realized in the Ultrachron project. The VL spectrometers improve count rates b y about 4X above standard PET (right), without sacrificing spectral resolution. The 160mm Rowland Circle radius is maintained. Measurables relevant to EPMA geochronology are shown in the table below, which compares an analysis of the GSC SHRIMP monazite standard GSC 8153 performed on the Cameca SX50 at UMass with an analysis performed using the SX-Ultrachron at UMass using the same analytical parameters (voltage, current, count time). Precision is more than doubled in the age analysis. For individual element analysis using the VLPETs (see Pb below), precision is improved by a factor of 4. More precise analyses of 3 ppm Pb, 4 ppm U are achieved in 14 pt analyses (15min/pt.) still using 15kV and 200nA.

Note that higher collection efficiency allows not only improved precision, but also permits precise analysis at reduced voltage or current density in beam-sensitive materials (carbonates, phosphates, etc.). See figure lower right.

GSC 8153 monazite analsyis (ca. 505). U (UMb)and Pb (PbMa) were analyzed using PET on the SX50. U was analyzed using LPET, and Pb with VLPET on the Ultrachron.
Comparison of PbMa intensity PET (SX50) vs VLPET (SX-Ultrachron) over a range of beam accelerating potential. PbMa, analysis could be performed at 8kV using VLPET, attaining the same count rates as PET at 15kV. Similarly, current or count time could be lowered if either beam exposure damage or analysis time were important issues.
Accuracy can be assessed by analysis of secondary standards. In the case of EPMA trace element analysis, as is the case with all trace element analysis, reliable secondary standards are exceptionally problematic. Ages can be verified to an extent by comparison to materials analyzed by IDTIMS or ion probe, keeping in mind that the exact domains analyzed are not volumetrically identical. For complex polygenetic materials, the comparison is very difficult. The most reliable attempts are those involving monazite megacrysts from pegmatites. None have been identified that are truly compositionally homogeneous, but the coarse sector-zoning permits comparison of age analyses by various techniques.
Comparisson of ID-TIMS and EPMA geochronology (UMass Ultrachron) for monazite from a range of ages. Samples included are those for which it is believed that TIMS analysis is not compromized by multiple generations of monazite growth in single grains.
Truly accurate analysis involves assessment of all the potential sources of innacuracy, including background curvature and interferences, peak interferences, boundary flourescence (right). Beam damage effects, buildup of internal space charge, and other analytical effects must also be assessed..
  Fluroescence of potassium Ka from adjacent K-feldspar causes interference of K Ka on UMb, and apparent decrease in age near the monazite rim (U to Pb ratio decreases). Orange = adjacent to Ksp, blue=adjacent to quartz (left). Maps inset: Th Ma (left) and U Mb (right). Note K felspar grains lower right and right). Example from sector zoned monazite from the lower gorge of the Grand Canyon, USA. For details see Jercinovic and Williams (2005).
pfE cameca