Abstract Eclogites in the high-pressure (HP) and ultrahigh-pressure (UHP) belts record subduction-zone processes; exhumed eclogites of seafloor protoliths record low-temperature (mostly <600°C), high-pressure and ‘wet’ environments: that is, relatively ‘cold’ subduction with highly hydrous minerals such as lawsonite. In contrast, eclogites formed by the continental subduction record relatively ‘hot’ ( T > 650°C) and ‘dry’ ultrahigh-pressure metamorphic (UHPM) conditions with syncollisional magmatism. Here, we investigate some eclogites from two ophiolite sequences that intercalated in the North Qaidam UHPM belt, which is genetically associated with continental subduction/collision. The observations of lawsonite pseudomorphs in garnets, garnet compositional zoning, mineral and fluid inclusions in zircons, and zircons with distinct trace-element patterns and U–Pb ages all suggest that these eclogites represent two exhumation episodes of subduction-zone metamorphic rocks: the early ‘cold’ and ‘wet’ lawsonite eclogite and the late ‘hot’ and ‘dry’ UHP kyanite eclogite. The early lawsonite-bearing eclogite gives metamorphic ages of 470–445 Ma and the later kysnite-bearing eclogite gives metamorphic ages of 438–420 Ma, with a time gap of c. 7–10 myr. This gap may represent the timescale for transition from oceanic subduction and continental subduction to depths greater than 100 km. We conclude that evolution from oceanic subduction to continental collision and subduction was a continuous process. In addition, we find that titanium contents in zircons have a positive correlation with U contents. Ti-in-zircon thermometry is likely to be invalid or limited for low-temperature eclogites.

Forty-six years ago saw the first manned landing on the Moon and the return of the first lunar samples. Since then a vast database has been accumulated with many ideas published on lunar petrogenesis, yet important problems recognized in early days remain under-addressed. In this paper, we first review these problems and emphasize that these problems need resolving before genuine progress can be made. We then discuss that contrary to the prevalent view, the available data do not show the presence of a strong positive Eu anomaly (Eu/Eu* > 1) in the lunar highland crust, but a weak negative one (Eu/Eu* < 1) if any. This observation weakens the plagioclase flotation hypothesis, which is the very foundation of the prevailing lunar magma ocean (LMO) hypothesis. Recent success in the determination of abundant water in lunar glasses and minerals confirms the prediction in the early days of lunar research that the Moon may have been a water-rich planet and may still be so in its interior, which disfavors the dry Moon hypothesis, weakens the LMO hypothesis, and questions many related lunar petrogenesis interpretations. Volatilization (into the vacuum-like lunar “atmosphere”) of lunar magmatism during its early history could have further facilitated plagioclase crystallization and feldspathic crustal formation. The important role and effect of plagioclase crystallization are best manifested by the significant correlation ( R 2 = 0.983 for N = 21) of Eu/Eu* (0.24–1.10) with Sr/ Sr* (0.10–1.12) defined by the lunar samples. Although the anorthositic lunar highlands are expected to have large positive Eu (Eu/Eu* > 1; ~1.99) and Sr (Sr/Sr* > 1; ~2.56) anomalies, their absence inferred from the global remote sensing data is best explained by the widespread but areally and volumetrically insignificant KREEP-like material that is enriched in K, rare earth elements, and P (hence, KREEP) as well as all other incompatible elements with very strong negative Eu (Eu/Eu* << 1; as low as 0.24) and Sr (Sr/Sr* << 1; as low as 0.10) anomalies. The KREEP-like material may have been produced through fractional crystallization enrichment equivalent to processes in advancing, periodically replenished, periodically tapped, continuously fractionated magma chambers. Compared with magmatic rocks on the Earth, lunar rocks are depleted in moderately volatile elements like P, Na, K, Rb, Cs, etc., probably associated with volatilization during the early history of the lunar magmatism. Further work is needed toward an improved understanding of the origin and evolution of the Moon and its magmatism.