Jc. Stormer et al., VARIATION OF F-X-RAY AND CL-X-RAY INTENSITY DUE TO ANISOTROPIC DIFFUSION IN APATITE DURING ELECTRON-MICROPROBE ANALYSIS, The American mineralogist, 78(5-6), 1993, pp. 641-648
Under conditions routinely used for electron microprobe analyses (15 k
V, sample current of 0.015 muA, beam diameter of 5-50 mum) FKalpha X-r
ay intensities of Durango fluorapatite sections with the surface perpe
ndicular to the c axis increase up to 100% during the first 60 s of ex
posure to the beam. After longer periods of exposure to an electron be
am the intensity falls to values below the initial intensity. This eff
ect is strongly anisotropic. Sections parallel to the c axis show a si
milar behavior but on a time scale approximately 20 times longer. Ther
e appears to be no relaxation or decay of the effect when the same spo
t is reanalyzed after periods of up to several weeks. Under similar mi
croprobe operating conditions, topaz shows no change in FKalpha intens
ity, and fluorite shows a decline in intensity with increasing duratio
n of beam exposure. The intensity variation and its anisotropy can be
explained by diffusion of F to the surface driven by the electrical fi
eld produced by primary beam electrons implanted at a depth below the
analyzed region. The increase in intensity is believed to be due to th
e diffusion of F ions to sites near the surface where their X-rays are
subject to much less absorption than in the apatite matrix. The aniso
tropy of the effect is explained by the known, structurally controlled
anisotropy of diffusion in apatite. Cl appears to behave in a similar
manner except that the initial enhancement of X-ray intensity is less
pronounced, since Cl X-rays are not so strongly absorbed in the apati
te matrix. Accurate analysis of apatite will require a series of analy
ses on the same spot extrapolated to a value at the initiation of beam
exposure. Because of permanent long-term changes in X-ray intensity w
ith cumulative beam exposure, apatite should not be used as a primary
standard. Samples exposed to an electron beam for cathodoluminescence
studies should not be analyzed without removing the altered surface la
yer (approximately 5 mum).