The Moon has a magmatic and thermal history that is distinct from that of the terrestrial planets¹. Radioisotope dating of lunar samples suggests that most lunar basaltic magmatism ceased by around 2.9–2.8 billion years ago (Ga)^2,3, although younger basalts between 3 Ga and 1 Ga have been suggested by crater-counting chronology, which has large uncertainties owing to the lack of returned samples for calibration^4,5. Here we report a precise lead–lead age of 2,030 ± 4 million years ago for basalt clasts returned by the Chang’e-5 mission, and a ²³⁸U/²⁰⁴Pb ratio (µ value)⁶ of about 680 for a source that evolved through two stages of differentiation. This is the youngest crystallization age reported so far for lunar basalts by radiometric dating, extending the duration of lunar volcanism by approximately 800–900 million years. The µ value of the Chang’e-5 basalt mantle source is within the range of low-titanium and high-titanium basalts from Apollo sites (µ value of about 300–1,000), but notably lower than those of potassium, rare-earth elements and phosphorus (KREEP) and high-aluminium basalts⁷ (µ value of about 2,600–3,700), indicating that the Chang’e-5 basalts were produced by melting of a KREEP-poor source. This age provides a pivotal calibration point for crater-counting chronology in the inner Solar System and provides insight on the volcanic and thermal history of the Moon.

The replacement of rutile by Fe-Ti oxides is a common phenomenon during the retrogression of eclogites. Here, we report an unusual case regarding the replacement of Fe-Ti oxides by rutile during greenschist-facies metamorphic overprinting of veins in amphibolites (retrograded eclogites) from the Dabie ultrahigh-pressure (UHP) terrane, eastern China. The veins mainly consist of plagioclase, Fe-Ti oxides, and quartz, which crystallized from a Ti-rich amphibolite-facies fluid that formed during exhumation of the eclogites. Two types of textures involving the replacement of Fe-Ti oxides by rutile are recognized in the veins: (1) the first type is characterized by the development of rutile coronas (Rt-C) and other silicates (high-Fe epidote, muscovite, and chlorite) around the external boundaries of the Fe-Ti oxide grains, and (2) the second type is characterized by the formation of symplectitic intergrowths of rutile (Rt-S) and magnetite after exsolved hemo-ilmenite (H-Ilm) lamellae in the Fe-Ti oxides. The micro-textures, mineral assemblages, and Zr-in-rutile thermometry indicate that both replacement reactions involved mineral re-equilibration processes in the presence of an infiltrating fluid phase at ~476–515 °C, taking place by an interface-coupled dissolution-precipitation mechanism. Thermodynamic modeling reveals that both replacement reactions occurred during oxidation processes under relatively high-oxygen fugacity (fO₂) conditions, approximately 2.5–4.5 logfO₂ units higher than the fayalite-magnetite-quartz (FMQ) reference buffer. In situ Sr isotopic analyses of epidote (Ep-C) coexisting with the Rt-C suggest that the infiltrating fluid involved in the greenschist-facies replace- ment reactions was externally derived from the surrounding granitic gneisses (the wall rocks of the amphibolites). Compared with the rutile in the UHP eclogites (Rt-E) and amphibolites (Rt-A), the Rt-C is characterized by distinctly lower contents of Nb (<10 ppm) and Ta (<2 ppm) and Nb/Ta ratios (340 ppm) and V (>1580 ppm). These results provide a geochemical fingerprint for distinguishing the low-pressure (LP) rutile from relic high-grade phases in retrograded HP-UHP rocks. Our results reveal that rutile can form during LP retrograde stage in UHP rocks by high-fO₂ fluid-induced replacement reactions. The unusual replacements of Fe-Ti oxides by rutile-bearing assemblages during retrogression provide important constraints on fluid-mineral reactions and fO₂ variations in exhumed UHP slabs.

Fluid infiltration into metacarbonates is a key mechanism to induce orogenic decarbonation, which influences the global carbon cycle and long‐term climate evolution. Little is known regarding the fluid pathways during episodic infiltration events and how flow patterns control time‐integrated CO₂ outflux. We investigate the “vein‐like” polycrystalline mineral reaction zones (PMRZs) in dolomite marbles (Mogok metamorphic belt, Myanmar), which are formed by metasomatism via the infiltration of Si–Al–K–Ti–Zr‐bearing fluids. The petrographic textures and mineral U–Pb chronology reveal three episodes of fluid influx in a single PMRZ: (1) the initial episode (Stage‐I) transformed most dolomite into Mg‐rich silicates/oxides and calcite at ∼35–36 Ma indicated by baddeleyite cores; (2) baddeleyite rims gave ages of ∼23–24 Ma, representing a subsequent infiltration episode (Stage‐II) that modified Stage‐I minerals via a dissolution–precipitation mechanism; (3) the final episode (Stage‐III) is recorded by zircon replacing baddeleyite, which yielded ages of ∼17 Ma. Stage‐III fluid has a higher SiO₂ activity and [CO₂/(CO₂ + H₂O)] than Stage‐I/Stage‐II fluids. Thermodynamic and mass‐balance analyses indicate that Stage‐I infiltration causes >62–67% loss of CO₂ by both dolomite‐consuming reactions and calcite dissolution, whereas the latter two infiltration episodes induce <12–18% loss of CO₂ via calcite dissolution. Our results provide compelling evidence that repeated episodes of infiltration (each separated in time by 7–13 Ma) occurred along a single channel in marbles. The initial infiltration episode may create high‐permeability regions, offering favorable channels for later‐stage fluids that transfer obviously less CO₂ than the initial metasomatism. This considerably complicates a quantitative assessment of CO₂ liberation from metacarbonates during orogenesis.

Garnet orthopyroxenites from Maowu (Dabieshan orogen, eastern China) were formed from a refractory harzburgite/dunite protolith. They preserve mineralogical and geochemical evidence of hydration/metasomatism and dehydration at the lower edge of a cold mantle wedge. Abundant polyphase inclusions in the cores of garnet porphyroblasts record the earliest metamorphism and metasomatism in garnet orthopyroxenites. They are mainly composed of pargasitic amphibole, gedrite, chlorite, talc, phlogopite, and Cl-apatite, with minor anhydrous minerals such as orthopyroxene, sapphirine, spinel, and rutile. Most of these phases have high X_Mg, NiO, and Ni/Mg values, implying that they probably inherited the chemistry of pre-existing olivine. Trace element analyses indicate that polyphase inclusions are enriched in large ion lithophile elements (LILE), light rare earth elements (LREE), and high field strength elements (HFSE), with spikes of Ba, Pb, U, and high U/Th. Based on the P–T conditions of formation for the polyphase inclusions (˜1.4 GPa, 720–850°C), we suggest that the protolith likely underwent significant hydration/metasomatism by slab-derived fluid under shallow–wet–cold mantle wedge corner conditions beneath the forearc. When the hydrated rocks were subducted into a deep–cold mantle wedge zone and underwent high-pressure–ultrahigh-pressure (HP–UHP) metamorphism, amphibole, talc, and chlorite dehydrated and garnet, orthopyroxene, Ti-chondrodite, and Ti-clinohumite formed during prograde metamorphism. The majority of LILE (e.g. Ba, U, Pb, Sr, and Th) and LREE were released into the fluid formed by dehydration reactions, whereas HFSE (e.g. Ti, Nb, and Ta) remained in the cold mantle wedge lower margin. Such fluid resembling the trace element characteristics of arc magmas evidently migrates into the overlying, internal, hotter part of the mantle wedge, thus resulting in a high degree of partial melting and the formation of arc magmas.

The mechanism and efficiency of CO2 release from subducting carbonates lie at the heart of quantifying carbon cycling in subduction zones. Most previous studies focused on CO2 release from calcite/aragonite, while decarbonation behavior of another critical carbonate—dolomite—is poorly constrained. Here, we investigated reaction veins and zones in dolomitic marbles from the Dabie ultrahigh-pressure (UHP) terrane, China. Petrology, geochemistry, O-C isotopes, and U-Pb dating reveal that the reaction veins and zones record the metasomatic interaction of dolomitic marbles with episodic Si-Al-Ti–bearing aqueous fluids from UHP eclogites and substantial release of CO2. Mass-balance calculations indicate that 67%–84% of initial CO2 was released from the altered marbles, whereas the losses of Ca (34%–67%) and Mg (almost immobile) are much lower, suggesting that CO2 removal was not caused by simple (bulk) dissolution of dolomite. Instead, it reflects that ~50% of initial CO2 was removed via replacement reactions of dolomite by Mg-silicates + CaCO3, and another ~17%–34% CO2 extraction was by dissolution of reaction-formed CaCO3. Enrichment of solutes (e.g., Si and Al) in slab fluids facilitates the decarbonation by solute load in fluid-dolomite interactions, and pulsed fluid infiltration helps enhance CO2 liberation. Considering the wide presence of dolomite in subducted slabs and its high decarbonation efficiency during fluid infiltration, the findings highlight that slab dolomite plays a significant role in carbon release from subduction zones.