Ultra-high precision ⁴⁰Ar/³⁹Ar ages for Fish Canyon Tuff and Alder Creek Rhyolite sanidine: New dating standards required?

Abstract

The ⁴⁰Ar/³⁹Ar dating technique is a high precision (<0.1%) method with wide application to geological samples. However, the method is predicated on the availability of natural mineral standards of known age. Widely used ⁴⁰Ar/³⁹Ar standards include sanidine from the (ca. 28 Ma) Fish Canyon Tuff (FCT) and the (ca. 1.2 Ma) Alder Creek Rhyolite (ACR). Despite common usage, the ages of FCT and ACR sanidine have proven contentious, with reported values varying by >2%; well outside the ±0.1% aspiration of EARTHTIME (www.earth-time.org). The current study presents ultra-high precision, multi-collector ⁴⁰Ar/³⁹Ar results for FCT and ACR sanidine, using a new generation ARGUSVI mass spectrometer. These analyses demonstrate the significantly higher (by a factor of ∼10) level of precision achievable using the ARGUSVI system, compared to most previous studies. Significantly, ⁴⁰Ar/³⁹Ar step-heating analyses resolve distinct age gradients for both FCT and ACR sanidine. Possible explanations for these gradients include combinations of recoil loss of ³⁹Ar, instrumental bias, isotopic fractionation during step-heating, thermally induced argon loss and/or extraneous argon contamination. Minor recoil loss of ³⁹Ar and partial retention of inherited argon during extended magma residence, are deemed the most plausible causes of the age discordance. Given this assumption, we calculate revised (maximum?) eruption ages of 28.01 ± 0.04 Ma (2σ) for FCT and 1.178 ± 0.002 Ma (2σ) for ACR, relative to the astronomically calibrated age of A1 Tephra sanidine, Crete [Rivera et al. (2011) Earth Planet Sci. Lett. 311, 420–426]. Nonetheless, the ARGUSVI data demonstrate that FCT and ACR sanidine are non-ideal as high precision ⁴⁰Ar/³⁹Ar standards, with different heating protocols likely to induce inter-laboratory bias. The current findings account for at least some of the scatter in inter-calibration ages reported previously for various ⁴⁰Ar/³⁹Ar standards. In broader terms, the results mandate a re-evaluation of the astronomically calibrated ages for these and other sanidine samples, while the new generation of multi-collector mass spectrometers provides the means to further evaluate ⁴⁰Ar/³⁹Ar ages used to define new decay constants and standard ages.

Introduction

The ⁴⁰Ar/³⁹Ar variant of the K–Ar dating technique, developed by Merrihue and Turner (1966), has become one of the most widely used geochronological methods applicable to geological samples (see McDougall and Harrison, 1999 for details). The ⁴⁰Ar/³⁹Ar method involves neutron irradiation of potassium-bearing samples to convert a proportion of the ³⁹K to ³⁹Ar, with the ⁴⁰Ar∗/³⁹Ar_K ratio measured mass spectrometrically (⁴⁰Ar∗ = radiogenic ⁴⁰Ar; Table 1). Natural mineral standards of ‘known’ age are co-irradiated with the samples to determine sample/standard ⁴⁰Ar∗/³⁹Ar ratios (R-values; Renne et al., 1998) and hence the age(s) of the unknown(s). Although the ⁴⁰Ar/³⁹Ar method is capable of high precision (<0.1%) age determinations, the accuracy of the ages is dependent on the veracity of the standard ages and the potassium decay constants. In this study, we focus on the accuracy of two commonly used ⁴⁰Ar/³⁹Ar standards, namely Fish Canyon Tuff (FCT) and Alder Creek Rhyolite (ACR) sanidine. Decay constant issues are discussed in Renne et al., 2010, Renne et al., 2011.

Fish Canyon Tuff (FCT) minerals were first suggested as ⁴⁰Ar/³⁹Ar and fission track standards by Cebula et al. (1986). In recent years, FCT sanidine (ca. 28 Ma) has become the standard of choice in most ⁴⁰Ar/³⁹Ar laboratories, due to its excellent single crystal age reproducibility and exceptionally low atmospheric contamination levels (e.g. Renne et al., 1998, Spell and McDougall, 2003). Although less frequently utilised, Alder Creek Rhyolite (ACR) sanidine (ca. 1.19 Ma; Turrin et al., 1994) is an important Quaternary dating standard and also the type locality for the ∼1.19 Ma Cobb Mountain Normal Polarity Subchron. Despite widespread use, the ages of FCT and ACR sanidine have proven controversial. Suggested K–Ar and ⁴⁰Ar/³⁹Ar ages for FCT sanidine range from 27.54 ± 0.29 Ma to 28.39 ± 0.19 Ma (∼3%) (e.g. Cebula et al., 1986, Renne et al., 1998, Renne et al., 2010, Renne et al., 2011, Lanphere and Baadsgaard, 2001, Spell and McDougall, 2003, Kuiper et al., 2008, Ganerød et al., 2011, Rivera et al., 2011, Hall, 2013), compared to ²⁰⁶Pb/²³⁸U zircon ages that vary from 28.0 to 28.4 Ma (e.g. Schmitz and Bowring, 2001, Bachmann et al., 2007). Reported ages for ACR sanidine are similarly variable, with reported values ranging from 1.15 ± 0.02 to 1.206 ± 0.004 Ma (>4%) (Turrin et al., 1994, Schmitt et al., 2003, Nomade et al., 2005, Renne et al., 2010, Renne et al., 2011).

The variation in age estimates for both FCT and ACR sanidine standards is well outside the ±0.1% accuracy goal of the EARTHTIME consortium (www.earth-time.org). To address this conundrum, we conducted ultra-high precision argon isotopic analyses of FCT and ACR sanidine, using a new generation, multi-collector, ARGUSVI mass spectrometer. Resultant ⁴⁰Ar/³⁹Ar step-heating data reveal significant age discordance for both samples, which complicates the utility of these minerals as high precision ⁴⁰Ar/³⁹Ar dating standards. More generally, our findings account for some of the variation observed in previous inter-calibration studies and raise questions regarding the accuracy of high precision ⁴⁰Ar/³⁹Ar ages referenced to these standards.