Volatiles in basaltic glasses from the East Pacific Rise at 21°N: implications for MORB sources and submarine lava flow morphology

Abstract

Pillow-rim glasses from a suite of moderately evolved lavas erupted along the axis of the East Pacific Rise (EPR) at 21°N were analyzed by high-temperature mass spectrometry for volatile content. Concentrations of H 2O, F and Cl in the 21°N glasses are low and correlate inversely with Mg# yielding well-defined trends. These and other geochemical data indicate derivation of the 21°N glasses from similar parental magmas produced from a highly depleted, nearly homogeneous source [1]. Total volatile, H 2O, F and S contents are lower and CO 2 content is somewhat higher in these samples than in previously analyzed Mid-Atlantic Ridge (MAR) glasses. 21°N glasses are similar in volatile content to glasses from the Galapagos Spreading Center (GSC) at 95°W except in Cl (lower) and CO 2 (higher). Unlike most MAR and GSC glasses, CO 2 is the dominant volatile in all of the 21°N glasses with Mg# > 62. All of the EPR and GSC glasses contain reduced carbon species (CO and CH 4), unlike most previously analyzed MAR samples. These data indicate that sources for mid-ocean ridge basalts are extremely volatile-poor (< 0.10 wt.%) and are probably dominantly reduced (below quartz-fayalite-magnetite buffer). Sheet flow and pillow basalts contain identical volatile contents. Thus, volatile abundance is not a factor controlling flow morphology. Extrusion rate and/or surface topography are probably the most important factors influencing flow morphology in submarine basalts. H 2O-release patterns are however related to flow morphology. Mass pyrograms for sheet flow and pillow basalt glasses both show bimodal H 2O-release behavior at the same temperatures (800°C and1000°C ± 50°C). However, the dominant H 2O-release peak for sheet flow glasses is at the lower temperature; for pillow-rim glasses it is at the higher temperature. Infrared spectroscopic studies indicate that H 2O in these glasses is present only as hydroxyl groups. Thus, the cause of differences in the bimodal H 2O-release and how it is related to flow morphology is unclear.