This research began with a simple question: Is molybdenum a chalcophile element – i.e., is it hosted in sulfides? I began to answer this question with three granites at the University of Maryland, then followed Mo through the Archean continental crust, Hawaiian lavas, subducted oceanic crust, the lithospheric mantle, glacial sediments, and ultimately weathering profiles at the University of California Santa Barbara. Each measurement prompted new questions about the geochemical behavior of Mo in different reservoirs. Along the way, we explored other nominally chalcophile elements like Ga, Ge, Mo, Ag, Cd, In, Sn, Sb, W, Tl, and Bi. Partitioning and isotope fractionation studies of several of these elements (e.g., Cd, Sn, Tl, W) are some of the newest frontiers in geochemistry, as their behavior has the potential to inform us of geological processes that shape our crust and mantle.

Nevertheless, molybdenum remained the central theme of these research projects. Using a combination of laser ablation ICP-MS, standard addition single collector ICP-MS and isotope dilution multi collector ICP-MS, I present six studies on the geochemistry of Mo. In the first two chapters we establish that Mo is not hosted in sulfides in the granitic upper crust and in basaltic to intermediate magmas of Kilauea Iki lava lake in Hawaii. We also find that Mo is systematically depleted from each granite analyzed, relative to the LREE, which led to the establishment of three hypotheses: 1) Mo is fractionated during igneous differentiation, 2) Mo is retained in the slab during subduction, or 3) Mo is removed from cooling granite plutons during fluid exsolution to eventually precipitate as molybdenite (MoS2) in epithermal veins. The first hypothesis was the basis for an undergraduate research project that refuted the idea (Appendix G), the second hypothesis formed Chapter III, and the results of these studies ultimately support hypothesis 3, suggesting that, on average, 60% of the Mo delivered to the upper crust in plutons forms MoS2.

Chapters III and VI follow Mo into the lithospheric mantle and lower crust where we find that Mo isotopes are fractionated during subduction, as isotopically light Mo is incorporated into rutile and isotopically heavy Mo is transported into the mantle wedge. This may partially explain why the continental crust is isotopically heavier than peridotites and basalts. Additionally, we find that Mo is added via metasomatic fluids or melts to residual peridotites where it is ubiquitous along grain boundaries. Molybdenum is enriched in metasomatically deposited sulfides, however its abundance in “primary” mantle sulfides is similar to that measured in crustal sulfides. Mo and W-rich fluids derived from subducted slabs may be the metasomatic agent that infiltrated these peridotites.

Finally, Chapters IV and V explore “low-temperature” isotopic fractionation of Mo during continental weathering at the Earth’s surface. We find that Mo isotopes in glacial diamictites that record weathering of the crust over the last three billion years become progressively lighter with time. We hypothesize that this is due to the onset of oxidative weathering at the Great Oxidation Event (2.4 Ga) and test this hypothesis with studies of modern weathering profiles. Data from saprolites and bauxites confirm that when Mo is lost during continental weathering, it is fractionated, and the light Mo isotopes are retained in the regolith, as observed in the diamictites. We conclude that Mo isotopes record the onset of oxidative weathering at ~ 2.4 Ga, and therefore the rise of oxygen in our atmosphere.

The Earth is unique among the terrestrial planets in our solar system, and no other known planet has a similarly bimodal distribution of surface elevations. The thicker, lighter continents are intermediate in composition, as opposed to the denser, mafic oceanic crust. All magmatism is ultimately sourced from the mantle; yet, mantle melting alone produces basalt. Multiple hypotheses have been proposed to explain the composition and evolution of the continental crust, including the incorporation of felsic sediments from the surface into the deep crust. How this happens and the amount of metasedimentary material in the lower continental crust remains unknown. There are four documented processes that can transport surface material to depth: (1) tectonic underplating in a forearc setting (e.g., Jacobson et al., 2011), (2) deep thrust faulting in a retroarc (e.g., Chin et al., 2013), (3) relamination via deep subduction and buoyancy-driven diapiric ascent (e.g., Hacker et al., 2011), (4) deep thrust faulting in a continent-continent collision (e.g., Zhao et al., 2001). Petrochronology of deep-seated metasedimentary rocks can distinguish between these scenarios by revealing the timing and conditions of metamorphism and the circumstances under which sediments were returned to the surface. To that end, three case studies of lower crustal metasedimentary rocks erupted as xenoliths were carried out and are described herein.

Metapelitic xenoliths from New Mexico (USA) and Chihuahua (Mexico) record sediment transport to the lower crust during continent-continent collision associated with the Mesoproterozoic Picuris orogeny. Geothermobarometry and Rio Grande rift-age rutile (Cipar et al., 2020) document equilibration in the local deep crust, which is actively undergoing ultra-high-temperature metamorphism. The original sediments were deposited during the early stages of the Picuris orogeny (1500–1450 Ma), and the transition from detrital zircon accumulation to metamorphic zircon and monazite growth marks the end of subduction and onset of collision. Metamorphic zircon and monazite also grew during Rio Grande rift activity beginning at 32 Ma. Nearly all garnet grains are homogeneous with respect to major and trace elements, and monazite inclusions within garnet indicate that they are Rio Grande rift in age. A single garnet grain is zoned in Zr and Hf, the slowest diffusing elements, with rim concentrations similar to those of other garnets in the sample. The core, however, has lower concentrations, indicating growth at lower temperatures, and is distinct in its abundance of acicular sillimanite inclusions. The presence of sillimanite points to metamorphism under thermal gradients higher than those found in subduction zones, and negative Eu anomalies in all zircon and monazite rule out metamorphism at the ultra-high-pressures that would have preceded relamination. Collectively, these data demonstrate transport of sediments from the surface to the lower continental crust during a collisional orogeny. The sediments resided in the lower deep continental crust for over 1400 Ma, and their presence there defines the location of the suture zone that formed during the Picuris orogeny.

Lower crustal xenoliths from the Great Falls tectonic zone (GFTZ) in Montana (USA) record a multi-billion-year geologic history prior to their entrainment and eruption during the Cenozoic Laramide orogeny. Metapelitic samples equilibrated in the lower crust, and mafic xenoliths with an arc cumulate origin document conditions in excess of the current crustal thickness. Rutile U-Pb ages date the eruption of the host rocks to ~50 Ma and demonstrate residence of the samples in the deep crust until that time. Metamorphic zircon and monazite in all metapelitic samples record sediment transport to depth during a major metamorphic event beginning ~1800 Ma, while one sample also documents an event at 2200–2000 Ma related to burial and metamorphism during rifting. Younger metapelite xenoliths show a clear transition from detrital zircon formation to metamorphic zircon and monazite growth at 1800 Ma, corresponding the end of subduction and beginning of collision during the Paleoproterozoic Great Falls orogeny. Their occurrence in the deep crust of the Great Falls tectonic zone (GFTZ) provides evidence that the GFTZ is a suture zone between the Medicine Hat block and Wyoming craton. U-Pb and trace-element depth profiles of monazite and zircon also record metasomatism in the deep crust at ~60 Ma during the Laramide orogeny, which is consistent with recent suggestions for Farallon slab fluid infiltration at this time. These rocks remained in the lower continental crust for over 1700 Ma.

Metapelitic xenoliths from the Bournac volcanic pipe in the Massif Central (France) record the incorporation of sediments into the lower continental crust during the collisional phase of the Paleozoic Variscan orogeny. Geothermobarometry yields pressure-temperature conditions consistent with a geothermal gradient in excess of those associated with a continental geotherm, and rutile in these samples remained above the Pb closure temperature until eruption of the xenoliths at 11.6 Ma. Detrital zircon date back to the Archean, and the sediments were deposited after 393 Ma based on the youngest detrital zircon and monazite ages. Metamorphic zircon and monazite document the transport of the sediments to depth beginning ~330 Ma, which is consistent with the known timing of continent-continent collision during the Variscan orogeny just prior to an episode of UHT metamorphism in the lower crust that peaked at 313 Ma (Laurent et al., 2023). Post-orogenic collapse began by 300 Ma, and zircon and monazite crystallization continued until ~265 Ma as the crust thinned and cooled. Later rifting events are recorded by even younger metamorphic zircon and monazite dates (as young as 207 and 74 Ma, respectively). These data demonstrate that the original sediments were transported to depth during continent-continent collision and resided in the stable lower continental crust for over 300 Ma.

Temperatures of the lower crust and Moho are critical boundary conditions for models of heat transfer and production in the lithosphere, as well as models of the bulk composition of the continental crust, yet they remain uncertain. This dissertation focuses on xenoliths—fragments of in situ deep crust transported rapidly to the surface by volcanism—and accessory phase U-Pb and trace-element petrochronology in order to provide direct insights into the long-term thermal history of lower continental crust. Chapter I constitutes a study of xenoliths and basement outcrops from northern Tanzania aimed at unraveling crustal vs. mantle heat contributions to variable surface heat flow across the tectonically-stable craton and adjacent rifting crust. Lower-crustal xenoliths erupted on the craton edge contain rutile with U-Pb dates as old as 1000 Ma, whereas xenoliths from the rift have rutile and apatite with near-zero Ma U-Pb dates but significantly older titanite dates (up to 560 Ma). These and other data suggest that Moho temperatures beneath the craton margin are cooler than below the rifting crust, such that regional differences in surface heat flow reflect variable mantle heat flow. Chapter II focuses on kimberlite-borne xenoliths from the central Siberian craton. Rutile and apatite in garnet-granulite xenoliths preserve U-Pb dates between 1.8 Ga and 360 Ma (the timing of kimberlite eruption). Such spreads in U-Pb dates have been interpreted previously to reflect partial Pb loss during slow cooling through the Pb closure temperatures of rutile/apatite, but could also result from brief heating pulses in the lower crust prior to eruption. Laser ablation depth profiling reveals that U-Pb age and elemental gradients in rutile are coupled, contrary to diffusive decoupling expected for thermally-mediated volume diffusion over billions of years. These data are instead best explained by multiple transient heating episodes during distinct thermotectonic events, including immediately prior to kimberlite eruption. Finally, Chapter III presents TIMS and LASS-ICP-MS data for potential reference apatites. Using microbeam methods to obtain precise and accurate U-Pb, Sm-Nd, Sr isotopes and elemental abundances from apatite—a petrologically diverse thermochronometer—requires well-characterized and matrix-matched reference materials, yet suitable reference apatites are scarce. The development of the reference apatite suite provides new benchmarks for in situ apatite isotopic analyses and inter-laboratory calibrations.