Three-body decays of charged D mesons
(Contact person: Natasha Miladinovic,
CP violation has been observed in strange and bottom mesons, but not yet in charm meson decays.
Small but non-zero direct CP asymmetries are expected in Cabibbo suppressed charm decays,
and can be measured as a difference in branching ratios beteen D+ and D-.
We focus on the D+ decays to three charged pions.
The standard model expectation is in the 10-3 range.
The best published experimental limit is -0.017 ± 0.042 from E791
One intermediate result will be the branching ratio for this decay.
The best published measurement to date is from E791:
0.0311 ± 0.0018 +0.0016-0.0026
The data used are 193pb-1 collected from Feb 2002 to Aug 2003 with the
displaced track trigger, including dynamically prescaled triggers with lower thresholds.
Event reconstruction is based on 5.1 data (xbhd0c dataset) and CharmMods to reconstruct secondary vertices.
Standard phantom-layer/B-field/COT error rescaling corrections are used.
This page describes plots that have been PR-blessed at the
B meeting of April 22, 2004.
Loose cuts for D+->K-pi+pi+ reconstruction:
- At least 3 axial hits in different layers of the SVX and 20/16 axial/stereo COT hits
- pT (track) >= 500 MeV/c
- chi2xy <=20
- |d0| (D+ candidate) <= 500 µm
>= 0.18 MeV/c2 to reject background from partially reconstructed D*+
- Trigger confirmation on each candidate
The histogram has been fit with a double Gaussian and a radiative tail for the signal and a second order
polynomial for the background. Reflections from Ds, D*+, Lambdac decays
have not been explicitly modeled in the fit.
D+->K-pi+pi+ Dalitz plot:
In a Dalitz plot, the invariant mass squared of one track combination is plotted versus the
invariant mass square of another track combination.
In decays of a pseudoscalar particle to 3 pseudoscalar daughters,
the Dalitz plot contains all the non-trivial physics information of the decay.
If the decay amplitude is a constant,
the Dalitz plot will be homogeneously populated.
We use the sidebands to subtract the background.
Since the sidebands have a larger reconstructed mass, they occupy a different region in the
Dalitz plot than the signal, which complicates the side-band subtraction.
We use the following trick:
we rescale the invariant mass of each track pair with the ratio of the reconstructed Q-value and the
true Q-value of the decay.
D+->K-pi+pi+ Dalitz plot:
The K*0(890) resonance can be clearly recognized.
Optimized cuts for D+->pi-pi+pi+ reconstruction:
The branching ratio of D+ -> pi-pi+pi+ is
about 30 times smaller than to K-pi+pi+.
With the same loose cuts as for K-pi+pi+,
the background is too high to be acceptable.
We study the sidebands for the properties of the background and a MC sample for the properties of the signal
and tighten the cuts to get the best value for the significance S/sqrt(S+B):
D+ -> K-pi+pi+ mass plot with tight cuts:
- pT (track)>= 700 MeV/c2
- chi2xy <=3
- |d0| (D+ candidate) <= 80 µm
- ct >=250 µm
- pT (D+ candidate) >=5GeV/c2
The yield of D+ -> K-pi+pi+
has been reduced from about 2.8 million to 1.6 million using the tight cuts.
D+ -> pi-pi+pi+ mass plot with tight cuts:
To the right of the D+ -> pi-pi+pi+ is the
Ds -> pi-pi+pi+
peak with a similar yield.
A Dalitz plot for the D+ -> pi-pi+pi+
decay is made in a similar way as for D+ -> K-pi+pi+.
D+ -> pi-pi+pi+ Dalitz plot:
The Dalitz plot clearly shows the rho0(770)->pi+pi- and
f0(980)->pi+pi- intermediate resonances.