Theses Doctoral

The Fidelity of the Mantle Signal in Peridotite Xenoliths: Interactions during Magmatic Ascent

Towbin, William Henry

Peridotite xenoliths, brought to the surface in volcanic eruptions, provide one of the few opportunities to directly sample the lithospheric mantle. They are often used to understand a locality’s lithospheric history and structure, as well as to determine parameters for geophysical and petrological models. Xenoliths have been particularly instrumental in studying mantle metasomatism, the infiltration of low degree mantle melts or fluids, and in determining the H₂O concentration of peridotite in the lithosphere. Rarely however, are the host magmas studied in conjunction with the xenoliths, which can lead to misattribution of the processes responsible for a xenolith’s chemical or textural properties. The studies presented in the following chapters evaluate the potential for the host-magma to overprint some of those properties. It also highlights some of the unique ways in which xenoliths can be used to study magmatic ascent rates. The second and third chapters most directly focus on investigating these relationships, specifically, the origin of xenolith melt veins (Chapter 2), and the potential for magmatic H₂O to rapidly overprint mantle signatures (Chapter 3).

Chapter 1 is a methods paper (Towbin et al., 2023) and broadly applicable to anyone measuring H₂O concentrations in olivine and pyroxenes by SIMS and FTIR. Chapter 1, which is in press in American Mineralogist (Towbin et al., 2023), addresses the challenges in measuring H₂O concentrations in nominally anhydrous minerals (olivine, orthopyroxene and clinopyroxene) by SIMS and NanoSIMS. We resolve a long standing ~40% discrepancy in the primary olivine H2O SIMS and FTIR calibrations. Additionally we derive new, self-consistent concentrations for common orthopyroxene and clinopyroxene standard reference materials. Owing to the limited availability of homogenous olivines with high concentrations of H2O (>100 ppm), we recommend using orthopyroxene standards to calibrate H₂O in olivine by SIMS, on the basis of their similar calibration slopes. Finally, we apply this calibration procedure to previous measurements of H₂O in olivine and glass to derive olivine-melt partition coefficients. By ensuring consistent calibration methods between studies we largely resolve a previously identified discrepancy between experimental and naturally derived partition coefficients resulting in a value of 0.0009 +/- 0.0003 (KdH2Ool/liq) at pressures ~0.2 - 2 GPa.

Chapters 2 and 3 focus on mantle xenoliths in the context of their host magmas erupted from monogenetic cinder cones in the Basin and Range region of the western U.S. Most of the samples studied come from the Sullivan Ranch Locality, a cinder cone located ~20 miles north of the Grand Canyon on the Uinkaret Plateau. Other xenoliths studied were from Vulcan’s Throne (also in the Uinkaret volcanic field) and the Cima Volcanic field in the Mojave Desert. The xenoliths are all spinel-lherzolite or harzburgite bombs, and the host magma samples are rapidly quenched, ~1cm scoria lapilli or melt inclusion bearing phenocrysts sieved from ash.

Chapter 2 investigates melt veins found within xenoliths, a common feature in samples from many localities around the world. The origin of these veins is variably attributed to mantle metasomatism, infiltration of the host magma, or decompression melting. In order to test these origins and constrain the P-T-t path of xenolith ascent, we applied a wide range of analytical techniques to measure the textural and chemical properties of the xenoliths. Image analysis demonstrates that melt veins represent a significant melt fraction of the xenolith ~10% by volume. Melt veins have trace element concentrations that approach those of the host magma within ~1 cm of the xenolith exterior, providing evidence for limited infiltration. The reconstructed bulk melt veins are a basalt to basaltic andesite (50-52% SiO2), and mineral compositions within the melt vein (high Fo olivine, high Cr cpx and high Ti spinel) are distinct from the primary minerals away from melt veins. The reconstructed incongruent melting reactions for two different xenoliths (cpx + opx + sp  melt + ol) are similar to those from equilibrium peridotite melting experiments at 1 GPa, supporting local equilibrium in the melt vein, but disequilibrium at the hand-sample scale (cms). The timescale for this process was derived from Fe-Mg zonation profiles in primary olivines adjacent to melt veins that indicated diffusion over 1-3 weeks. This timescale is similar to that derived for olivines at the xenolith exterior, which were diffusively equilibrating with the host magma for a week, supporting melting during entrainment. Given the P-T-t path of these xenoliths, which should be similar to those erupted from other alkali basalt cinder cones from around the world, decompression melting is inevitable.

Chapter 3 focuses on determining whether H₂O concentrations in xenolith minerals reflect mantle concentrations or equilibration with the host magma during ascent. We measured H₂O concentrations olivine and pyroxenes from the exterior and interior of several xenolith bombs, using the recommended protocols presented in Chapter 1. We also measured volatile concentrations (H₂O, CO₂, S, Cl, F) in melt inclusions from olivine phenocrysts in the host magma, in order to understand the magma’s initial volatile concentrations and H₂O degassing path. The nearly identical core and rim concentrations of xenolith olivines and olivine phenocrysts from the host scoria provide strong evidence for equilibration. The water zonation profiles of olivine phenocrysts can be used to constrain the rapid ascent of the degassing host magma: 6 km in ~12.5 minutes. In the ≥ 6 days of ascent from the Moho (from Chapter 2), xenoliths and host magmas have ample opportunity to equilibrate, given that that complete equilibration of H₂O in the largest clinopyroxenes (the slowest diffusing phase) will occur in only 7 hours at 1200°C. Moreover, within the uncertainties of partitioning models, xenolith clinopyroxenes have water contents that are in equilibrium with their host magmas from all three localities. We conclude that hot (~1200°C) alkali basalts that have ascent rates comparable to those that produce strombolian eruptions and cinder cones (~week ascent time from the mantle) will overprint the water contents of their cargo of mantle xenoliths, and that no vestige of information is retained about the water concentrations in the mantle lithosphere. Previous studies that have inferred mantle water contents based on alkali basalt hosted xenoliths will require substantial re-evaluation.

A common conclusion from both the second and third chapters is that the magmatic ascent history has the potential to overprint mantle signatures in peridotite xenoliths during ascent. In future studies, careful attention should be paid to determine whether the xenoliths from an eruption have been overprinted. As we have shown, this is best accomplished through comprehensive analysis of xenoliths and their host magma. More broadly, when using xenoliths as mantle proxies for H₂O contents, melting or metasomatism, the first assumption should be that the host magma has overprinted the feature of interest unless there is a compelling reason to believe otherwise. The burden of proof should be to demonstrate that the texture or chemistry has not been modified during ascent, rather than assuming ascent is fast enough to avoid overprinting.

While some of the main conclusions of this thesis might be cause for despair in using peridotite xenoliths as mantle proxies, both Chapters 2 and 3 highlight exciting applications for using xenoliths to learn about the dynamics and timescales of magmatic ascent. Using the multiple chronometers from both studies, we determined that degassing of H₂O, starting around 6 km from the surface, drives magma acceleration from <0.1 m/s (determined in Chapter 2 from total entrainment time) to ~8 m/s during the final stages of ascent (using the multiple H+ diffusion chronometers in Chapter 3). This rarely determined magma acceleration is made possible by the fact that the xenolith records both the total entrainment time, based on Fe-Mg diffusion in olivine rims, as well as a final decompression rate, based on diffusion of H+ in pyroxenes from the last stages of ascent.

Geographic Areas


  • thumnail for Towbin_columbia_0054D_17751.pdf Towbin_columbia_0054D_17751.pdf application/pdf 7.85 MB Download File
  • thumnail for Chapter Chapter application/zip 4.21 MB Download File
  • thumnail for Chapter Chapter application/zip 240 KB Download File
  • thumnail for Chapter Chapter application/zip 46.3 MB Download File

More About This Work

Academic Units
Earth and Environmental Sciences
Thesis Advisors
Plank, Terry A.
Ph.D., Columbia University
Published Here
May 10, 2023