Gopher: Draft random thoughts

• From e-mail exchange with Bruce Cleveland:
  We have recently measured the 210Po content of the bricks we have by surrounding the detector endcap with lead and looking at the intensity of the 803-keV peak. From the rate in this peak and a GEANT4 simulation to get the efficiency, we infer the 210Po specific activity of our lead to be 49 ± 11 Bq/kg. Further, from the magnitude of the 210Bi spectrum in this measurement and calculations of the effectiveness of copper shielding in NIMB 117 189 (1996), I infer that the internal bremsstrahlung from 210Bi that gets through our 50-mm copper inner liner is about 30%-40% of the background we see in this detector at low energy. Obviously this detector would benefit if the inner lead had lower 210Po content, but it would be sufficient to use Pb with about 5 Bq/kg, and 0.3 Bq/kg would be overkill.
  • 803.10 keV peak from Po-210 has a ratio of 0.00121%. That's why I don't usually bother with it.
  • NIMB 117 189 (1996) and its abstract:
    A method for fast simulations (based on GEANT) of background of low-level Ge γ-spectrometers induced by 210Pb/210Bi in shielding lead is described. Particles emitted from the inner surface of contaminated lead are simulated instead of decays of 210Bi nuclei. This approach speeds up the computation more than thousand times. The used generator of particles emitted from the lead surface is described in detail. The surface flux approach gives results equivalent to a full simulation for rectangular shields. For cylindrical shields, deviations are smaller than 2%. The method is demonstrated by investigating the reduction of the background with layers of radio-pure lead and copper of varying thickness. A layer 18 mm thick reduces the maximum in the background spectrum 100 times and good results are achieved even with a layer 5 mm thick. A general attenuation coefficient of 0.714 ± 0.001 cm−1 was found for reduction of the background by copper layers used for inner lining.

Comparing Gopher's spectrum with the one from Diode-M, it's noticeable that Pb-210 peak at 46.5 keV is not that much higher in our spectrum, whereas the continuum is much more (unproportionally) elevated. Does this suggest that Pb-210 is not our main issue?