Modeling Argon Isotopic Behavior in Polygenetic Impact Melt Breccias

Abstract:

Impact melt breccias (IMBs) are complex mixtures of melt products and relict clasts of target rocks. Some show petrographic evidence for multiple phases of melt injection. A thin section of one such specimen from the Apollo 17 sample suite contains three distinctive melt domains. Laser microprobe ⁴⁰Ar/³⁹Ar dating reveals that each has a distinctive age, and the range of apparent melt-injection events spans 100’s of millions of years. Why did the younger melts not completely or partially reset the K-Ar systematics of older melt products? The answer appears to be that the thermal effects of injections of thin veinlets of melt are too transient to produce measurable Ar loss from the host rock. We have coupled a thermal model of the instantaneous injection of a planar melt vein into an infinite rocky medium [1] with a diffusive loss model for spherical grains present in the host material near a vein contact [2]. We used experimentally determined diffusion parameters for Ar in plagioclase [3] in our modeling since it is the least retentive phase in lunar IMBs for which such parameters have been published. At a given distance into the host from the vein contact, the peak temperature and, thus, the peak diffusivity, is controlled by the thickness of the melt vein (i.e., by the thermal mass deposited in the host). Our goal was to constrain the minimum thickness of a melt vein that would cause enough heating for a nearby grain to lose radiogenic ⁴⁰Ar* to shift its age by 1% (a common 2σ analytical precision for ⁴⁰Ar/³⁹Ar dates of lunar samples). For a 3800 Ma old grain, ca. 2.4% of ⁴⁰Ar* must be lost to cause a 38 Ma (1%) reduction in the apparent age. Assuming an ambient temperature of -33 ˚C, a > 24 mm-thick vein of 1500 ˚C melt would be required to cause this level of ⁴⁰Ar* loss from a 20 µm-diameter plagioclase located 100 µm away from the contact. In the sample we studied, typical vein thicknesses are much smaller (< 4 mm). For a grain of the same size and distance from the melt, the dike would need to be at least 2.1 meters thick to completely reset the K-Ar system. The results of our modeling underscore the utility of laser microprobe ⁴⁰Ar/³⁹Ar dating for robust reconstructions of the history of polygenetic IMBs. [1] J. Crank (Oxford University Press, ed. 2, 1979). [2] E. Watson, D. Cherniak (2013) Chem Geol, 335, 93-104. [3] W. Cassata et al. (2009) GCA, 73, 6600-6612.