Search

Cosmic ray muons penetrate deeply into rock where they interact with atoms to produce cosmogenic nuclides. Incorporation of the muon contribution to the production rates of cosmogenic nuclides increases the accuracy of exposure dates, burial ages, and erosion rates inferred from measured nuclide concentrations. In the absence of empirical evidence, it is generally assumed that muons do not produce ³He, a cosmogenic nuclide commonly used for exposure dating. Here we assess whether muons produce ³He by measuring He isotope concentrations in pyroxene and ilmenite from a ∼300 m deep drill core and other subsurface samples of the mid-Miocene Columbia River Basalt in Washington, USA. ³He concentrations in our samples exhibit an exponential decline with depth with an e-folding length of 32.4 m, which corresponds to an attenuation length for ³He production of 8780 g cm⁻². The deeply penetrating exponential is diagnostic of ³He production by cosmic ray muons. Assuming no erosion, we constrain the minimum surface muonogenic production rate to be 0.23 atom g⁻¹ pyroxene yr⁻¹, whereas when incorporating erosion the production rate is 0.45 atom g⁻¹ pyroxene yr⁻¹. ³He concentrations in samples deeper than ∼100 meters are consistent with model-based estimates of depth-independent nucleogenic production from the capture by ⁶Li of neutrons produced by alpha particle reactions on light elements. Measurements in other subsurface samples indicate that muon-produced ³He is prevalent across the Columbia Plateau. The penetration depth of muonogenic ³He production is substantially deeper, and the ratio of muon- to spallation-produced ³He is substantially lower, than found for other cosmogenic nuclides. Our results provide the first definitive empirical evidence for ³He production by muons, which has several implications for quantifying the timing and rates of Earth surface change and interpreting He isotope ratios. Importantly, despite the low production rates, landforms in the Channeled Scablands, which were formed by incision of the Columbia River Basalt by the late-Pleistocene Missoula floods, have high concentrations of ³He inherited from post-Miocene muon exposure. Hence ³He production by muons must be considered, particularly when dating rapid erosional events in old bedrock. Our findings indicate samples with less than several tens of meters of shielding by overlying rock will contain cosmogenic ³He that elevates ³He/⁴He ratios. Hence caution should be used when using ³He/⁴He ratios from samples at shallower depths to infer mantle sources of basalt.