Re-examination of crystal ages in recent Mount St. Helens lavas

Implications for magma reservoir processes

Kari M. Cooper, Mary R. Reid

Research output: Contribution to journalArticle

91 Citations (Scopus)

Abstract

U-series data for recent Mount St. Helens lavas suggest that crystallization preceded eruption by more than 0.5 ka but are complicated by possible evidence of crystal recycling and/or addition of radium to the liquid after crystallization. We report new ion and electron microprobe trace- and major-element data for plagioclase and pyroxene in these recent Mount St. Helens lavas and use these data to reassess 226Ra-230Th crystal ages by taking into account differences in the partitioning behavior of radium and barium and the effects of impurities in mineral separates. Revised 226Ra-230Th model crystallization ages are ∼2-4 ka for plagioclase (with the exception of the 1982 dacite) and ∼0.15-5.7 ka for pyroxene. In contrast to previous interpretations, no late-stage addition of Ra to the liquid after precipitation of the minerals is required. The variability of Ba concentrations measured in plagioclase is too large to be consistent with progressive crystallization from the same liquid or with diffusive re-equilibration of xenocrysts with a new host liquid. Ba heterogeneity limits the residence time of the crystals in a magma at high temperatures and also suggests that in most cases Ra-Th ages have not been significantly modified by Ra diffusion into or out of the crystals. High (226Ra)/Ba in plagioclase in the 1982 dacite relative to the host liquid likely reflects crystallization processes that precluded bulk crystal-liquid chemical equilibrium. One possibility is that of growth entrapment of surface enrichments during rapid crystallization, which could lead to less discrimination between Ra and Ba than predicted by calculated bulk partition coefficients. 226Ra-230Th crystal ages for the Castle Creek andesite and basalt that are younger than 230Th-238U ages of the same crystals could be explained by mixing of crystals into melts with different 230Th/232Th ratios, by combinations of older and younger crystal growth within the same magma, or, for the basalt, by diffusion of Ra into crystals at high magmatic temperatures. Average plagioclase ages in most of the samples of > 2 ka imply that some significant mass fraction of the crystals in each flow was present simultaneously beneath the volcano. This observation could be consistent either with simultaneous storage of physically distinct magma batches or with incorporation of a population of similarly aged crystals into each successive magma batch.

Original languageEnglish (US)
Pages (from-to)149-167
Number of pages19
JournalEarth and Planetary Science Letters
Volume213
Issue number1-2
DOIs
StatePublished - Aug 1 2003
Externally publishedYes

Fingerprint

magma chamber
magma
examination
Crystallization
crystal
Crystals
plagioclase
crystals
crystallization
liquid
Liquids
liquids
Radium
radium
basalt
Minerals
dacite
pyroxene
minerals
Liquid Crystals

Keywords

  • Barium
  • Crystallization
  • Diffusion
  • Plagioclase
  • Radium
  • Uranium-series disequilibria

ASJC Scopus subject areas

  • Geochemistry and Petrology
  • Geophysics

Cite this

Re-examination of crystal ages in recent Mount St. Helens lavas : Implications for magma reservoir processes. / Cooper, Kari M.; Reid, Mary R.

In: Earth and Planetary Science Letters, Vol. 213, No. 1-2, 01.08.2003, p. 149-167.

Research output: Contribution to journalArticle

@article{825802a61a01432c9c8906e3bfa999cc,
title = "Re-examination of crystal ages in recent Mount St. Helens lavas: Implications for magma reservoir processes",
abstract = "U-series data for recent Mount St. Helens lavas suggest that crystallization preceded eruption by more than 0.5 ka but are complicated by possible evidence of crystal recycling and/or addition of radium to the liquid after crystallization. We report new ion and electron microprobe trace- and major-element data for plagioclase and pyroxene in these recent Mount St. Helens lavas and use these data to reassess 226Ra-230Th crystal ages by taking into account differences in the partitioning behavior of radium and barium and the effects of impurities in mineral separates. Revised 226Ra-230Th model crystallization ages are ∼2-4 ka for plagioclase (with the exception of the 1982 dacite) and ∼0.15-5.7 ka for pyroxene. In contrast to previous interpretations, no late-stage addition of Ra to the liquid after precipitation of the minerals is required. The variability of Ba concentrations measured in plagioclase is too large to be consistent with progressive crystallization from the same liquid or with diffusive re-equilibration of xenocrysts with a new host liquid. Ba heterogeneity limits the residence time of the crystals in a magma at high temperatures and also suggests that in most cases Ra-Th ages have not been significantly modified by Ra diffusion into or out of the crystals. High (226Ra)/Ba in plagioclase in the 1982 dacite relative to the host liquid likely reflects crystallization processes that precluded bulk crystal-liquid chemical equilibrium. One possibility is that of growth entrapment of surface enrichments during rapid crystallization, which could lead to less discrimination between Ra and Ba than predicted by calculated bulk partition coefficients. 226Ra-230Th crystal ages for the Castle Creek andesite and basalt that are younger than 230Th-238U ages of the same crystals could be explained by mixing of crystals into melts with different 230Th/232Th ratios, by combinations of older and younger crystal growth within the same magma, or, for the basalt, by diffusion of Ra into crystals at high magmatic temperatures. Average plagioclase ages in most of the samples of > 2 ka imply that some significant mass fraction of the crystals in each flow was present simultaneously beneath the volcano. This observation could be consistent either with simultaneous storage of physically distinct magma batches or with incorporation of a population of similarly aged crystals into each successive magma batch.",
keywords = "Barium, Crystallization, Diffusion, Plagioclase, Radium, Uranium-series disequilibria",
author = "Cooper, {Kari M.} and Reid, {Mary R.}",
year = "2003",
month = "8",
day = "1",
doi = "10.1016/S0012-821X(03)00262-0",
language = "English (US)",
volume = "213",
pages = "149--167",
journal = "Earth and Planetary Sciences Letters",
issn = "0012-821X",
publisher = "Elsevier",
number = "1-2",

}

TY - JOUR

T1 - Re-examination of crystal ages in recent Mount St. Helens lavas

T2 - Implications for magma reservoir processes

AU - Cooper, Kari M.

AU - Reid, Mary R.

PY - 2003/8/1

Y1 - 2003/8/1

N2 - U-series data for recent Mount St. Helens lavas suggest that crystallization preceded eruption by more than 0.5 ka but are complicated by possible evidence of crystal recycling and/or addition of radium to the liquid after crystallization. We report new ion and electron microprobe trace- and major-element data for plagioclase and pyroxene in these recent Mount St. Helens lavas and use these data to reassess 226Ra-230Th crystal ages by taking into account differences in the partitioning behavior of radium and barium and the effects of impurities in mineral separates. Revised 226Ra-230Th model crystallization ages are ∼2-4 ka for plagioclase (with the exception of the 1982 dacite) and ∼0.15-5.7 ka for pyroxene. In contrast to previous interpretations, no late-stage addition of Ra to the liquid after precipitation of the minerals is required. The variability of Ba concentrations measured in plagioclase is too large to be consistent with progressive crystallization from the same liquid or with diffusive re-equilibration of xenocrysts with a new host liquid. Ba heterogeneity limits the residence time of the crystals in a magma at high temperatures and also suggests that in most cases Ra-Th ages have not been significantly modified by Ra diffusion into or out of the crystals. High (226Ra)/Ba in plagioclase in the 1982 dacite relative to the host liquid likely reflects crystallization processes that precluded bulk crystal-liquid chemical equilibrium. One possibility is that of growth entrapment of surface enrichments during rapid crystallization, which could lead to less discrimination between Ra and Ba than predicted by calculated bulk partition coefficients. 226Ra-230Th crystal ages for the Castle Creek andesite and basalt that are younger than 230Th-238U ages of the same crystals could be explained by mixing of crystals into melts with different 230Th/232Th ratios, by combinations of older and younger crystal growth within the same magma, or, for the basalt, by diffusion of Ra into crystals at high magmatic temperatures. Average plagioclase ages in most of the samples of > 2 ka imply that some significant mass fraction of the crystals in each flow was present simultaneously beneath the volcano. This observation could be consistent either with simultaneous storage of physically distinct magma batches or with incorporation of a population of similarly aged crystals into each successive magma batch.

AB - U-series data for recent Mount St. Helens lavas suggest that crystallization preceded eruption by more than 0.5 ka but are complicated by possible evidence of crystal recycling and/or addition of radium to the liquid after crystallization. We report new ion and electron microprobe trace- and major-element data for plagioclase and pyroxene in these recent Mount St. Helens lavas and use these data to reassess 226Ra-230Th crystal ages by taking into account differences in the partitioning behavior of radium and barium and the effects of impurities in mineral separates. Revised 226Ra-230Th model crystallization ages are ∼2-4 ka for plagioclase (with the exception of the 1982 dacite) and ∼0.15-5.7 ka for pyroxene. In contrast to previous interpretations, no late-stage addition of Ra to the liquid after precipitation of the minerals is required. The variability of Ba concentrations measured in plagioclase is too large to be consistent with progressive crystallization from the same liquid or with diffusive re-equilibration of xenocrysts with a new host liquid. Ba heterogeneity limits the residence time of the crystals in a magma at high temperatures and also suggests that in most cases Ra-Th ages have not been significantly modified by Ra diffusion into or out of the crystals. High (226Ra)/Ba in plagioclase in the 1982 dacite relative to the host liquid likely reflects crystallization processes that precluded bulk crystal-liquid chemical equilibrium. One possibility is that of growth entrapment of surface enrichments during rapid crystallization, which could lead to less discrimination between Ra and Ba than predicted by calculated bulk partition coefficients. 226Ra-230Th crystal ages for the Castle Creek andesite and basalt that are younger than 230Th-238U ages of the same crystals could be explained by mixing of crystals into melts with different 230Th/232Th ratios, by combinations of older and younger crystal growth within the same magma, or, for the basalt, by diffusion of Ra into crystals at high magmatic temperatures. Average plagioclase ages in most of the samples of > 2 ka imply that some significant mass fraction of the crystals in each flow was present simultaneously beneath the volcano. This observation could be consistent either with simultaneous storage of physically distinct magma batches or with incorporation of a population of similarly aged crystals into each successive magma batch.

KW - Barium

KW - Crystallization

KW - Diffusion

KW - Plagioclase

KW - Radium

KW - Uranium-series disequilibria

UR - http://www.scopus.com/inward/record.url?scp=0043200957&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0043200957&partnerID=8YFLogxK

U2 - 10.1016/S0012-821X(03)00262-0

DO - 10.1016/S0012-821X(03)00262-0

M3 - Article

VL - 213

SP - 149

EP - 167

JO - Earth and Planetary Sciences Letters

JF - Earth and Planetary Sciences Letters

SN - 0012-821X

IS - 1-2

ER -