Hi John,
I collected 88 analyses of RTP on a grid measuring approximately 5.5 x 3.9 mm. I analyzed for Ca, Na, K, and Cs using a 15 kV potential, 100 nA beam current, and 10-micron beam diameter; respective standards were anorthite, albite, adularia, and pollucite. I counted for 120 s peak and 120 s background for each element and interpolated linearly between background offsets. I used LiF to measure Cs La in order to eliminate interference from Ti Ka. As I noted above, I applied pulse amplitude discrimination to suppress interference from P Ka(2) and P Kb(2) at the Na Ka background offset positions. Average concentrations and standard deviations were: CaO, 108 and 18 ppm; Na2O, 44 and 19 ppm; K2O, 202 and 13 ppm. For Cs2O, I ended up with average k-ratio = -0.000053. For each of the oxides CaO, Na2O, K2O, and Cs2O, I calculate respective detection limit (3 std deviations above background) of 37, 67, 33, and 600 ppm for a given analysis. The only trace elements I could "see" in wavelength scans were K and Ca.