A new extraction technique and production rate estimate for in situ cosmognic ¹⁴C in quartz

The potential of in situ cosmogenic ¹⁴C (in situ ¹⁴C) for surficial process studies is widely recognized, yet a realible means of isolating it has proved difficult to develop. We present here a new method for extracting in situ ¹⁴C from quartz that overcomes difficulties encountered with earlier techniques, yielding more reliable production rate estimates. Comparison of ¹⁴C thermal release patterns from surficial and deeply shielded quartz samples (Lifton, 1997) demonstrated that contaminant ¹⁴C is relased at or below 500°C, and that ¹⁴C released between 500 and 1500°C is essentially all in situ-produced. The new technique builds on this key result, using resistance heating of samples in the presence of a lithium metaborate (LiBO₂) flux, and collection of all evolved carbon as CO₂ between 500°C and 1100 to 1200°C. Our improved method has four distinct advantages over other extraction methods: (1) we can identify and quantitatively eliminate atmospheric/organic ¹⁴C contamination; (2) we can identify the in situ ¹⁴C component unambiguously without assumptions of ¹⁴CO/¹⁴CO₂ production proportions within the rock or equilibria on extraction; (3) in situ ¹⁴C is reliably extracted from quartz at lower temperatures and in less time than earlier methods and (4) blank ¹⁴C levels are consistently low ((2.3 ± 0.1) × 10^{5 14}C atoms (1σ)). Our new extraction procedures should thus enable researchers to use in situ ¹⁴C in diverse applications without reservation. We developed our new procedures using samples of wave-cut quartzite benches from the well-dated Bonneville (17.4 ± 0.3 cal ky) and Provo (16.8 ± 0.3 cal ky) shorelines of Pleistocene Lake Bonneville, Utah, and from underlying deeply shielded locations. In situ ¹⁴C was extracted from quartz separated from 2 Bonneville shoreline samples (6 aliquots) and 1 Provo shoreline sample (2 aliquots). Results demonstrate that our new procedures can effectively isolate the in situ ¹⁴C fraction with replicate analytical precision bettern than 2% (1σ, n = 5), while remaining consistent with earlier results. This level of precision and accuracy is comparable to or exceeds those currently obtainable with in situ cosmogenic ¹⁰Be, ²⁶Al, ³He, ²¹Ne, and ³⁶Cl. Resulting weighted mean in situ ¹⁴C site production rates for the Bonneville and Provo shorelines are 52.9 ± 1.7 (¹⁴C atoms/g SiO₂)/y and 48.7 ± 2.8 (¹⁴C atoms/g SiO₂)/y (1σ), respectively-consistent with earlier production rate estimates. Current and previuosly published in situ ¹⁴C site production rate estimates were then scaled to sea level and high geomagnetic latitude using the latitude-altitude scaling models of Lal (1991) and Dunai (2000). Results indicate that both models yield sea level, high-latitude production rates consistent with independent estimates. Our new in situ ¹⁴C data yield integrated late Quaternary production rate estimates at sea level and high latitude of 15.1 ± 0.5 (¹⁴C atoms/g SiO₂)/y using the Lal (1991) model, and 15.8 ± 0.5 (¹⁴C atoms/g SiO₂)/y with that of Dunai (2000). Until significant uncertainties in these models are addressed, however, we prefer the value from the widely-used Lal (1991) model as our best estimate of the integrated late Quaternary production rate for in situ ¹⁴C.