When a continent collides with an island arc or other continent, continental crust of the subducted continent may be buried to depths exceeding 100 km, and exposed to pressures that can cause formation of coesite and diamond. This process leads to substantial density increase in SiO₂-rich rocks and, in turn, to a reduction of the buoyancy of the subducted material, which should inhibit exhumation. Nevertheless, coesite- and diamond-bearing continental crustal rocks are known from several occurrences worldwide. We report on the discovery of microdiamond in kyanite-garnet gneiss from allochthonous metasediments of the Seve Nappe Complex in the Scandinavian Caledonides. Our discovery calls for general reconsideration of existing exhumation models of deeply subducted continental crust. We propose that the diamond-bearing rocks were subducted in an arc-continent collision setting, and their exhumation was facilitated by local pressure reduction resulting from extraction of the forearc lithospheric block.

Ultramafic portions of ophiolitic fragments in the Arabian–Nubian Shield (ANS) show pervasive carbonate alteration forming various degrees of carbonated serpentinites and listvenitic rocks. Notwithstanding the extent of the alteration, little is known about the processes that caused it, the source of the CO₂ or the conditions of alteration. This study investigates the mineralogy, stable (O, C) and radiogenic (Sr) isotope composition, and geochemistry of suites of variably carbonate altered ultramafics from the Meatiq area of the Central Eastern Desert (CED) of Egypt. The samples investigated include least-altered lizardite (Lz) serpentinites, antigorite (Atg) serpentinites and listvenitic rocks with associated carbonate and quartz veins. The C, O and Sr isotopes of the vein samples cluster between −8.1‰ and −6.8‰ for δ¹³C, +6.4‰ and +10.5‰ for δ¹⁸O, and ⁸⁷Sr/⁸⁶Sr of 0.7028–0.70344, and plot within the depleted mantle compositional field. The serpentinites isotopic compositions plot on a mixing trend between the depleted-mantle and sedimentary carbonate fields. The carbonate veins contain abundant carbonic (CO₂±CH₄±N₂) and aqueous-carbonic (H₂O-NaCl-CO₂±CH₄±N₂) low salinity fluid, with trapping conditions of 270–300°C and 0.7–1.1 kbar. The serpentinites are enriched in Au, As, S and other fluid-mobile elements relative to primitive and depleted mantle. The extensively carbonated Atg-serpentinites contain significantly lower concentrations of these elements than the Lz-serpentinites suggesting that they were depleted during carbonate alteration. Fluid inclusion and stable isotope compositions of Au deposits in the CED are similar to those from the carbonate veins investigated in the study and we suggest that carbonation of ANS ophiolitic rocks due to influx of mantle-derived CO₂-bearing fluids caused break down of Au-bearing minerals such as pentlandite, releasing Au and S to the hydrothermal fluids that later formed the Au-deposits. This is the first time that gold has been observed to be remobilized from rocks during the lizardite–antigorite transition.

Throughout the high Arctic, from northern Canada (Pearya) to eastern Greenland, Svalbard, Franz Josef Land, Novaya Zemlya, Taimyr and Severnaya Zemlya and, at lower Arctic latitudes, in the Urals and the Scandinavian Caledonides, there is evidence of the Grenville–Sveconorwegian Orogen. The latest orogenic phase (c. 950 Ma) is well exposed in the Arctic, but only minor Mesoproterozoic fragments of this orogen occur on land. However, detrital zircons in Neoproterozoic and Palaeozoic successions provide unambiguous Mesoproterozoic to earliest Neoproterozoic (c. 950 Ma) signatures. This evidence strongly suggests that the Grenville–Sveconorwegian Orogen continues northwards from type areas in southeastern Canada and southwestern Scandinavia, via the North Atlantic margins to the high Arctic continental shelves. The widespread distribution of late Mesoproterozoic detrital zircons far to the north of the Grenville–Sveconorwegian type areas is usually explained in terms of long-distance transport (thousands of kilometres) of either sediments by river systems from source to sink, or of slices of lithosphere (terranes) moved on major transcurrent faults. Both of these interpretations involve much greater complexity than the hypothesis favoured here, the former involving recycling of the zircons from the strata of initial deposition into those of their final residence and the latter requiring a diversity of microcontinents. Neither explains either the fragmentary evidence for the presence of Grenville–Sveconorwegian terranes in the high Arctic, or the composition of the basement of the continental shelves. The presence of the Grenville–Sveconorwegian Orogen in the Arctic, mainly within the hinterland and margins of the Caledonides and Timanides, has profound implications not only for the reconstructions of the Rodinia supercontinent in early Neoproterozoic time, but also the origin of these Neoproterozoic and Palaeozoic mountain belts.