Annealing radiation damage and the recovery of cathodoluminescence

• Published in: Chemical geology 2002-11, Vol.191 (1), p.121-140
• Main Authors: Nasdala, Lutz, Lengauer, Christian L, Hanchar, John M, Kronz, Andreas, Wirth, Richard, Blanc, Philippe, Kennedy, Allen K, Seydoux-Guillaume, Anne-Magali
• Format: Article
• Language: English
• Subjects: Cathodoluminescence; Earth Sciences; Mineralogy; Radiation damage; Raman spectroscopy; Sciences of the Universe; SHRIMP ion probe analysis; Thermal annealing; Transmission electron microscopy; X-ray powder diffraction
• ISSN: 0009-2541; 1872-6836
• DOI: 10.1016/S0009-2541(02)00152-3

The structural recovery upon heat treatment of a highly metamict, actinide-rich zircon (U≈6000 ppm) has been studied in detail using a range of techniques including X-ray powder diffraction, Raman spectroscopy, SHRIMP ion probe, electron microprobe, transmission electron microscopy and cathodoluminescence analysis. The structural regeneration of the amorphous starting material depends on random nucleation. It starts between 800 and 950 °C when amorphous ZrSiO 4 decomposes to form crystalline ZrO 2 and amorphous SiO 2. At around 1100 °C, well-crystallised ZrSiO 4 grows at the expense of the oxides. U has been retained in the newly grown zircon whereas Pb was evaporated during the heat treatment. This process is in marked opposition to the reconstitution of moderately metamict minerals, which experience a gradual recovery controlled by the epitaxial growth at the crystalline–amorphous boundaries. Both of these recovery processes are not the direct inverse of metamictisation. The structural regeneration was found to be connected with a significant increase in the emission of CL. In all cases (annealing heavily damaged zircon and moderately damaged zircon and monazite), we observe that the final, well-crystallised annealing products emit more intense CL than their radiation-damaged starting minerals, although having almost identical elemental composition. Our observations are taken as evidence that the CL is not only determined by the chemical composition of the sample but is also strongly controlled by structural parameters such as crystallinity or the presence of defect centres.