Comment on "Limits on Dark Matter Using Ancient Mica"

The search for WIMP-recoil tracks in mica using Atomic Force Microscopy [1] promises important improvements on current exclusion limits. For this to be fulfilled, background sources that could mimic the WIMP-recoil signature must be rejected. In the case of naturally occurring α-recoil tracks, the authors rely on the longer expected depth of their etch-pits, since they result from the sum of 8 (6) successive and spatially-connected recoils of nuclei in the 238 U Th (232) chains. An isolated α-recoil would be essentially indistinguishable from a typical WIMP-recoil: for instance, the ~ 70 keV 234 Th α-recoil has a range P ~ 290 ± 55 ± 45 Å in mica (the errors are longitudinal and lateral straggling respectively), while the recoil induced by a 100 (500) GeV/ c 2 WIMP scattering from a potassium nucleus has an average kinetic energy of ~ 30 (42) keV, with P ~ 270 ± 90 ± 70 (365 ± 120 ± 90) Å. The larger total stopping power S t for the 234 Th α-recoil (S t ~ 13.6 vs ~3.8 GeV cm 2 /g) guarantees a comparatively higher pit-revealing efficiency in the etching process, and a larger resistance to thermal annealing. The probability that all the consecutive α-recoil damage tracks retrace each other's trajectories imitating the WIMP signature is tiny. This has prompted the authors to claim that their method is at present not background-limited. I would like to suggest a possible weakness in this reasoning. The 238 U chain proceeds as 238 234 234 234 230 U Th Pa U Th α β β α  →   →   →   →  − − ... with α-emission half-lives T y T y 1 2 238 9 1 2 234 5 4 4 10 2 4 10 / /. ,. = ⋅ = ⋅ (the very fast β − decays produce no observable recoil tracks). The long T 1 2 234 / results in some 234 Th recoils not being accompanied by a second one, thereby mimicking a WIMP recoil. These single α-recoils are not negligible in this search: for a mica age A ~ < 10 9 y, one can approximate that the 238 U decay is taking place at a constant rate R 238 ~ N λ 238 , where λ 238 1 2 238 = (ln / 2)/T and N is the number of 238 U …