Oriented kokchetavite compound rods in clinopyroxene of Kokchetav ultrahigh-pressure rocks

Abstract

Two microdiamond-bearing samples, a dolomite marble and a garnet-clinopyroxene rock, from the Kokchetav ultrahigh-pressure metamorphic terrane were selected in the present study to explore the possible origin of KAlSi3O8 rod inclusions oriented along the c-axis of clinopyroxene host.The KAlSi3O8 rod inclusions at clinopyroxene cores, where K2O content is high in the range of 0.5–1.0wt.%, are mostly fine-grained with a rod width less than 1μm. AEM studies showed that the KAlSi3O8 phase in most rod inclusions is kokchetavite. K-feldspar is present only in a few cases, probably the result of phase transformation/recrystallization from kokchetavite during rock exhumation. Electron diffractions further showed that kokchetavite rods are oriented parallel to clinopyroxene [001] direction and they exhibit the same epitaxial relation with the clinopyroxene host in both samples with the (Al, Si)O4 tetrahedra chains along the hexagonal a-axis of kokchetavite parallel to the single SiO4 chain along the c axis of clinopyroxene; i.e., [12¯10]Ko//[001]Cpx and (0001)Ko//(100)Cpx. It is interesting to note that kokchetavite is always in association with phengite, tremolite, β-cristobalite, Si-rich (Al, K, Ca-bearing) low crystallinity phase, ±Si–Ca (Cl, As) phase, ±calcite, ±apatite, ±lollingite (FeAs2), forming compound rods. Furthermore, all these phases are also present within submicron-scale polyphase inclusion pockets in garnet within garnet-clinopyroxene rock sample. These kokchetavite compound rods are therefore most likely to have resulted from melt/fluid-clinopyroxene interactions leading to epitaxial deposition rather than exsolution sensu stricto from the clinopyroxene host. The suggested melt/fluid would have an “external” and/or an “internal” origin related to rock partial melting involving phengite breakdown.Discrete phlogopite and phengite needle-like inclusions with a needle width less than 1μm, as well as phlogopite–phengite and kokchetavite-mica intergrowth needles, are also not uncommon in clinopyroxene cores. There are specific crystallographic orientation relationships among phases. These micas might have formed earlier than kokchetavite and have exsolved from clinopyroxene host, although the mass balance issue during exsolution and the temporal relation between phlogopite and phengite needles remain to be settled.Clinopyroxene rims generally contain mica needles only. Domains near fractures with coarse-grained kokchetavite compound rods or with coarse-grained phlogopite+quartz needles are not uncommon. These clinopyroxene rims/domains are low in K2O (<0.5wt.%) and their inclusions are most probably a result of precipitation/recrystallization from late-stage(s) infiltrated fluid-clinopyroxene interactions.