A TEM study of the oriented orthopyroxene and forsterite inclusions in garnet from Otrøy garnet peridotite,WGR, Norway: new insights on crystallographic characteristics and growth energetics of exsolved pyroxene in relict majoritic garnet

Regularly oriented orthopyroxene (opx) and forsterite (fo) inclusions occur as opx + rutile (rt) or fo + rt inclusion domains in garnet (grt) from Otrøy peridotite. Electron diffraction characterization shows that forsterite inclusions do not have any specific crystallographic orientation relationships (COR) with the garnet host. In contrast, orthopyroxene inclusions have two sets of COR, that is, COR‐I: <111>_grt//<001>_opx and {110}_grt～//～{100}_opx (～13° off) and COR‐II: <111>_grt//<011>_opx and {110}_grt～//～{100}_opx (～14° off), in four garnet grains analysed. Both variants of orthopyroxene have a blade‐like habit with one pair of broad crystal faces parallel/sub‐parallel to {110}_grt plane and the long axis of the crystal, <001>_opx for COR‐I and <011>_opx for COR‐II, along <111>_grt direction. Whereas the lack of specific COR between forsterite and garnet, along with the presence of abundant infiltrating trails/veinlets decorated by fo + rt at garnet edges, provide compelling evidence for the formation of forsterite inclusions in garnet through the sequential cleaving–infiltrating–precipitating–healing process at low temperatures, the origin of the epitaxial orthopyroxene inclusions in garnet is not so obvious. In this connection, the reported COR, the crystal habit and the crystal growth energetics of the exsolved orthopyroxene in relict majoritic garnet were reviewed/clarified. The exsolved orthopyroxene in a relict majoritic garnet follows COR‐III: {112}_grt//{100}_opx and <111>_grt//<001>_opx. Based on the detailed trace analysis on published SEM images, these exsolved orthopyroxene inclusions are shown to have the crystal habit with one pair of broad crystal faces parallel to {112}_grt//{100}_opx and the long crystal axis along <111>_grt//<001>_opx. Such a crystal habit can be rationalized by the differences in oxygen sub‐lattices of both structures and represents the energetically favoured crystal shape of orthopyroxene inclusions in garnet formed by solid‐state exsolution mechanism. Considering the very different COR, crystal habit, as well as crystal growth direction, the orthopyroxene inclusions in garnet of the present sample most likely had been formed by mechanism(s) other than solid‐state exsolution, regardless of their regularly oriented appearance in garnet and the COR specification between orthopyroxene and garnet. In fact, the crystallographic characteristics of orthopyroxene and the similar chemical compositions of garnet at opx + rt inclusion domains, fo + rt inclusion domains/trails and garnet rim suggest that the orthopyroxene inclusions in the garnet are most likely formed by similar cleaving‐infiltration process as forsterite inclusions, though probably at an earlier stage of metamorphism. This work demonstrates that the oriented inclusions in host minerals, with or without specific COR, can arise from mechanism(s) other than solid‐state exsolution. Caution is thus needed in the interpretation of such COR, so that an erroneous identification of exhumation from UHP depths would not be made.