Transmission electron microscopy of coexisting low-tridymite polymorphs

Abstract

Three polymorphs of tridymite (MC, PO-10 and MX) have been examined by transmission electron microscopy (TEM), in crystals which also contain lamellae of cristobalite. MX is the least common, and is highly unstable in the electron beam. It occurs as small regions (< 1 /~m 2) in PO-10 tridymite and at MC/PO-10 boundaries; these regions are interpreted as being strained. A consistent association between PO-10 and the cristobalite lamellae is attributed to the fact that PO-10 can more closely match the low-cristobalite structure than MC can, at boundaries parallel to the plane in which the crystal structures contain layers of SiO4 tetrahedra. An analogous interpretation explains the observation that twin boundaries (associated with 60 ~ or 120 ~ rotation of the layer) are commonly parallel to the layer plane in PO-10 but at high angles to it in MC. Abundant lamellar features, parallel to the same plane in PO-10, are tentatively interpreted as representing twinning by reflection and 180 ~ rotation, which may also account for c* streaking in diffraction patterns.

An electron diffraction study of the polymorphs of siO2-tridymite

“…5 c is indeed now no longer 90 ~ but ~91 ~ and that the satellite reflections approximately perpendicular to the (1120)* direction do not exactly form rows parallel to the e* direction (see the dashed line in Fig. 5 c), as has also been reported by Ashworth (1989).…”

“…the n = 10 or PO-10 variant, is consistent with the observation that in no case were the diffraction spots characteristic of either high or low cristobalite observed for our specimens. Ashworth (1989), following on from an earlier inference of Schneider and F16rke (1986), observed a strong correlation between the occurrence of the PO-n tridymite and the presence of lamellar slabs of cristobalite intercalated into the tridymite. The absence of such lamellar slabs of cristobalite in our specimens would thus appear to be consistent with the absence of the PO-n polymorph/s.…”

“…In no case were the diffraction spots characteristic of either high or low cristobalite observed. Electron beam damage was a significant problem, as has been pointed out by various authors (Carpenter and Wennemer 1985;Ashworth 1989). Nonetheless it was possible to take diffraction patterns and sometimes images before degradation became a serious problem via use of a low beam intensity and by minimizing the irradiation time for any one particular grain.…”

“…On the other hand, the electron diffraction patterns of the PO-10 polymorph shown in Fig. i of Ashworth (1989) are quite distinct.…”

“…That this apparent variability in primary modulation wave-vector and average structure cell angles is, or rather can be, a genuine feature of this polymorph and not an artefact of our particular specimens can be deduced from previous TEM results. Ashworth (1989) for example, has published two <710> SADP's (see his Fig. ld and e) which also appear to require variability in the exact value of the primary modulation wave-vector. Similarly Carpenter and Wennemer (1985) state that in all their specimens the primary modulation wave-vector for Clearly there can be variability in the exact value of the primary modulation wave-vector and in the angle between the (1120)* and e* directions of reciprocal space.…”