Using delayed nucleation in melt inclusions to constrain final growth durations of magmatic crystals: An introduction and example using the May 18, 1980 White Pumice of Mt. St. Helens

Abstract Knowledge of the timing and duration of magmatic events provides important constraints on the evolution of igneous rocks and on volcanic eruption mechanisms. A new method is proposed to constrain the time duration between the saturation of a suite of melt inclusions trapped in a crystal with daughter minerals and the quenching of the melt inclusions to conditions where crystals cannot form. Classical nucleation theory predicts that crystal nucleation delay and subsequent growth is a function of the volume of the initial melt. This suggests that the crystallinity (defined in this work as the presence or absence of crystals) in a suite of melt inclusions of differing volumes can be used to constrain the duration of time between saturation of the melt inclusions and their quenching. As an example to demonstrate the applicability of this technique, a set of 402 melt inclusions found in 5 plagioclase crystals erupted in the Mt. St. Helens White Pumice of May 18th, 1980 were studied using a petrographic microscope and a scanning electron microscope. An ellipse was fit to each melt inclusion, its axes measured, and the presence, or absence, of crystals within the inclusions noted. The volume of each melt inclusion was estimated by approximating the size of the inclusion as an ellipsoid of revolution using the measured ellipse axes with rotation about the major axis. These measurements were used to calculate the probability of melt inclusion crystallization as a function of melt inclusion size and fit to the classical nucleation theory model for delayed nucleation in small volumes. This fit yielded the product of the time-averaged nucleation rate, <J>, and the time duration, Δ, or Δ. This information is combined with previous experimental studies on plagioclase nucleation rates, <J>, and crystal habits to calculate that melt inclusions most probably became supersaturated (which may have occurred soon after trapping) only days to months before the May 18th eruption. This time estimate is consistent with the observed seismic and gas emission records prior to the eruption and with diffusion modeling of Fe-Mg zoning in orthopyroxene crystals from Mt. St. Helen’s eruptions in the 1980’s.