diffusion
The discussion above shows that, for the local-PU free data sample only, a significant correction has to be applied to the lifetime measurement assuming our baseline deuterium concentration of 1 ppm (as mentioned we also have initiated R&D to reduce it to the 0.1 ppm level) . A credible correction cannot only rely on simulation, but must come from direct observations. Fortunately we can self calibrate the diffusion correction, if we are able to monitor the deuterium concentration in the experiment (it could even change during the course of a run due to outgasing from the walls).
Our main correction method relies on direct observation of diffusion. During
our test run we have demonstrated that a signal/background ratio of 100:1 was
achieved for diffusion events, characterized by a displaced vertex between muon
and electron. This data corresponds to natural hydrogen with 
140 ppm.
Accordingly we expect a signal/background ratio of 1:1 at 1 ppm, possibly
4:1 accounting for the larger solid angle coverage of the electron detector.
For a selected good solid angle of 30% for TPC tracking we expect a statistics
of 
events. This should readily allow a calibration of diffusion
at the 5% level.
In addition we are exploring two other methods. One is the detection of 
fusion events, where more statistics is needed to decide whether sufficient
background suppression can be achieved. The expected statistics of
of 
He
MeV) appears sufficent for a few % accuracy.
Another method is suggested by the radial dependence of the correction estimated
in table 9. In first order the correction scales with C
,
suggesting that C can be extracted from the difference of the
observed decay rates for different radial cuts.