Quantitative vesicle analyses and total CO₂ reconstruction in mid-ocean ridge basalts

Highlights

• Stereological methods closely reproduce vesicle size distributions and vesicularities measured by X-ray micro-tomography.

• Revised methods for calculating total CO₂ concentrations in MORB provide more accurate estimates.

• New total CO₂ estimates have implications for estimating mantle carbon concentrations and ridge CO₂ flux.

Abstract

Vesicle textures in submarine lavas have been used to calculate total (pre-eruption) volatile concentrations in mid-ocean ridge basalts (MORB), which provide constraints on upper mantle volatile concentrations and global mid-ocean ridge CO₂ flux. In this study, we evaluate vesicle size distributions (VSDs) and volatile concentrations in a suite of 20 MORB samples that span the range of vesicularities and vesicle number densities observed in MORB globally. We provide recommended best practices for quantifying vesicularity, vesicle number densities, and VSDs based on synthetic vesicle populations and comparisons between traditional 2D estimates and X-ray computed micro-tomography results. For 2D measurements, we recommend analyzing multiple polished fragments with a cumulative area >100 times the area of the largest observed vesicle and including >200 vesicles in stereological VSD reconstructions. For 3D measurements, we recommend analyzing sample volumes >0.01 cm³ at resolutions <2.0 μm/pixel for low vesicularity MORB (i.e., <4 vol.%) and sample volumes >0.1 cm³ with resolutions <5 μm/pixel for higher vesicularity samples. Our validation of vesicularity measurements allows reconstructions of total CO₂ concentrations in MORB using dissolved volatile concentrations, vesicularities, and equations of state. We assess approaches for estimating the exsolved CO₂ concentration in MORB vesicles and find that CO_2(g) density is ~40% lower than previously suggested, likely due to melt contraction during quenching. Based on these results, we recommend using sample eruption pressures, magmatic temperatures, and an equation of state that accounts for non-ideality at high temperatures to calculate exsolved CO₂ concentrations when independent constraints from Raman spectroscopy or laser ablation are unavailable. Our results suggest that some previous studies may have overestimated MORB CO₂ concentrations by as much as 50%, with the greatest differences in samples with the highest vesicularities. These new results imply lower CO₂/Ba of undegassed, enriched-MORB and lower integrated global ridge CO₂ flux than previously inferred.

Introduction

Mid-ocean ridge basalts (MORB) frequently experience incomplete degassing due to the hydrostatic pressure at the seafloor. As a result, MORB volatile concentrations can provide valuable insight into volatile abundances in Earth's upper mantle (e.g., Cartigny et al., 2008; Jones et al., 2019; Michael and Graham, 2015), which influence mantle melting, melt migration, and geophysical properties of the Earth's interior (e.g., Dasgupta and Hirschmann, 2010; Hirth and Kohlstedt, 1996). MORB CO₂ concentrations and vesicle textures have also been used to constrain mid-ocean ridge CO₂ flux (e.g., Chavrit et al., 2014), magma storage conditions within the oceanic crust (e.g., Aubaud et al., 2004; Dixon et al., 1988; le Roux et al., 2006; Sarda and Graham, 1990) and magma ascent and effusion rates during mid-ocean ridge eruptions (Chavrit et al., 2012; Gardner et al., 2016; Jones et al., 2018; Soule et al., 2012). Many of these studies estimate 3D vesicularities and reconstruct total CO₂ concentrations using 2D measurements (e.g., Aubaud et al., 2004; Chavrit et al., 2014; Hekinian et al., 2000; Javoy and Pineau, 1991; Jones et al., 2018; Pineau et al., 2004; Soule et al., 2012). However, the methods for reconstructing 3D vesicle textures and total CO₂ concentrations in MORB have not yet been rigorously tested. Aubry et al. (2013) highlight the sensitivity of CO₂ reconstructions to assumptions regarding the behavior of vesicles during quenching, suggesting that improper assumptions may explain systematic differences between calculated and simulated CO₂ contents in a suite of MORB samples. Several studies compare 2D and 3D vesicularities and vesicle size distributions in subaerial samples (Baker et al., 2011; Giachetti et al., 2011; Gurioli et al., 2008; Hughes et al., 2017); however, MORB have different vesicle characteristics than most subaerial samples, including low vesicularities (i.e., gas volume fractions), low vesicle number densities (i.e., number of vesicles per unit volume), and small vesicle sizes, which motivates a robust evaluation of these methods specific to MORB.

This study examines theoretical and empirical methods for quantifying vesicle populations in a suite of MORB samples and offers new insights into the validity of those methods and best practices for evaluating vesicularity, vesicle size distributions, and CO₂ concentrations in this subgroup of volcanic rocks. We use comparative 2D and 3D measurements of MORB samples along with synthetic data to provide a consistent, comprehensive evaluation of stereological corrections in MORB. We further suggest an improved method for quantifying exsolved CO₂ concentrations based on vesicularity through an evaluation of equations of state, theoretical estimates of vesicle volume change during cooling, and estimates of CO_2(g) density in MORB samples based on Raman spectroscopy.