The lithospheric architecture of Africa consists of several Archean cratons and smaller cratonic fragments, stitched together and flanked by polycyclic fold belts. Here we investigate the structure and chemistry of spinels from lithospheric mantle xenoliths from distinct tectonic settings, i.e. from the the Saharan metacraton in Libya (Waw-En-Namus) which could show archaic chemical features, Cameroon (Barombi Koto and Nyos Lakes) where the Sub Continental Lithospheric Mantle was modified during the Pan-African event and fluxed by asthenospheric melts of the Tertiary Cameroon Volcanic Line and Morocco (Tafraoute, Bou-Ibalrhatene maars) in the Middle Atlas where different metasomatic events have been recorded. From a structural point of view it is to notice that the Libyan spinels can be divided into two groups having different oxygen positional parameter (u > 0.2632 and u < 0.2627, respectively), while those from Cameroon are in between those values as the Moroccan ones already studied by other authors. The intracrystalline closure temperature (Tc) of the here studied spinels is different among the different samples with one Libyan group (LB I) showing Tc in the range 490-640°C and the other 680-950°C (LB II). Cameroon and Morocco spinels show a Tc in the range 630-760°C. About 150 different spinels have been studied for their trace element content and it can be seen that many of them are related to Cr content, while Zn and Co are not and clearly distinguish the occurrences. Differences in the trace element chemistry, in the structural parameters and in the intracrystalline closure temperatures suggest that a different history should be considered for Cameroon, Morocco and LB I and LB II spinels. Even if it was not considered for this purpose, we tentatively used the Fe2+/Fe3+ vs. TiO2 diagram that discriminate between peridotitic and the so-called “magmatic” spinels, i.e. spinel crystallized from melts. LB I and LB II spinels plot in the peridotitic field while Cameroon and Morocco spinels fall in the magmatic one. Consequently, the xenoliths sampled from a probably juvenile SCLM at the edge of the most important lithospheric roots (i.e. Cameroon and Morocco) apparently have spinels possibly fractionated in situ from percolating melts and do not represent a real spinel-peridotite facies. On the contrary mantle xenoliths from Libya exhibit spinels with peridotitic features compatible with a slow ascent of a mantle diapir (plume).
