[Halld-physics] Eta Proposal

[Gan, Liping]

Hi, Matt,

As pointed out in Dave's note, he wanted to study the upper limit only. Therefore, he did not apply any rare decay specific trigger and cuts at all. In the eta rare experiment, we take beam above 9 GeV. There should be a hardware cut in the trigger which requires the total energy deposit in the FCAL greater than 8 GeV, which will remove most of background in Dave's figure 2.  In the offline analysis, the coincident window will not be 100 ns, it should be at level of 4 ns! We used about 4 ns coincident window in PrimEx-I analysis even without individual TDC. With Flash ADC in Hall D, we can only do better. 

As for eta -> pipi, I don't not aware any  particular models for non-SM CP violation prediction, neither  one would see an effect in eta -> 2 pi0 but not in eta -> pi+ pi-. New experiment results may encourage new theoretical calculations.   I do not agree with you that BESIII result can make final statement on those channels. I have discussed with the group who are working on this subject last year . They agreed that their background will be orders of magnitude worse than Jlab.  With same number of eta decays, our actual detection efficiency will be much higher than any facilities, including BESIII, WASA, CB-MIMI, and  future KLOE-2.

I agree with you that  to distinguish eta->pi0+2gamma and eta-2pi0 is not a easy job. However, we are not the first one to do this business either. With Hall D tagged high energy photon beam and high resolution PWO calorimeter, we expect that we can do better.

Liping




________________________________________
From: Matthew Shepherd [mashephe at indiana.edu]
Sent: Tuesday, May 01, 2012 7:56 AM
To: Gan, Liping
Cc: Alex Somov; GlueX Physics
Subject: Re: Eta Proposal

Hi Liping,

Thanks -- here are a few responses:

On Apr 30, 2012, at 10:25 PM, Gan, Liping wrote:

> The beam related background has be done by Dave Lawrence (see attached file), we will include it in the proposal in the next a couple of days.

Dave's plot suggests the potential for some background.  It is difficult to know how this will extrapolate out to the signal region.  The rate is in Hz; and your run will have about 10^7 seconds of live time.  So out where there are only 1 Hz in David's plot there will be ten million events (in 2 MeV bins) in the final data run. If I look at the slope it seems to be decreasing logarithmically at a rate of about 1 decade per 100 MeV.  If that rate continues then you expect 3-4 orders of magnitude drop out to the eta mass.  That still leaves 1000-10000 events per 2 MeV bin in the region of the eta mass -- that's a lot.  I understand simulating such durations of background is impossible.  I wonder if more background could be simulated so you don't have to try to extrapolate over 7 orders of magnitude.  If for example you had another 2-3 orders of magnitude more BG MC then you could see if the logarithmic trend continued over the next hundred MeV or so.  Bottom line is that David's plot doesn't assure me that EM backgrounds are insignificant.

> Our goal is to improve the current experimental  limit on  eta->2pi0. Can we really see new source of CP violation appeared in eta->2pi0? I will not able to answer. Only God knows.

I should have asked my question a separate way.  Are there particular models for non-SM CP violation where one would see an effect in eta -> 2 pi0 but not in eta -> pi+ pi-?  I understand there is probably a model for any effect, but it would be good to know what model space you are targeting.  Any model that predicts eta -> pi0 pi0 at the same level as eta -> pi+ pi- has already been ruled out in the majority of the region you hope to test (and may be completely ruled out in the entire region with more statistics at BESIII by the time the eta data are collected).  Do you have some references to go with the physics argument around Eq. 3 (bottom of page 7)?

> To distinguish eta->pi0+2gamma and eta-2pi0, one can check the invariant mass of two gammas in all combinations of pair.

Sure, but this is much easier said than done since there are combinatoric backgrounds.  It would be nice to demonstrate this with a sample of eta -> pi0 2 gamma and eta -> 2 pi0 in the paper (assuming you can isolate just the eta -> 4 gamma sample).  It may enhance your argument for a high resolution calorimeter.

Matt