LFV in the neutrinoless decay tau->3mu at CMS

RWTH Aachen

tau sources at the LHC

For an estimation of the cross section, PYTHIA 6.227 with the parton distribution function CTEQ5L was used. Since the decay tau->3mu is not implemented in PYTHIA, the Standard Model decay tau-> mu nu nu was modified such that both neutrinos have been replaced by muons. With an integrated luminosity of 10fb^-1 per year, even in the low luminosity phase of LHC about 10^12 tau leptons will be produced within the CMS detector

Number of produced tau leptons per year in CMS

production channel W source Z source B_0 B_± B_s D_s
# tau/year (10 fb^-1) 1.7x10^8 3.2x10^7 4.0x10^11 3.8x10^11 7.9x10^10 1.5x10^12

The dominant production sources of the copious lepton production at LHC are the Ds and various B mesons. The W and the Z production sources will provide considerably less tau leptons per year, however more energetic.

Trigger settings of tau->3mu at CMS

The important triggers for this analysis are the dedicated single and di-muon trigger. For the low luminosity phase it is planned to use as thresholds for single muons pT > 14 GeV at L1 and pT > 19 GeV for the HLT. The thresholds for the di-muon trigger will be pT > 3 GeV at L1 and pT > 7 GeV for the HLT.
Studies have shown, that most of the events produced via W and Z will be triggered by these muon triggers. However, due to the low pT of the muons coming from a decay originated from a Ds or B meson decay, the trigger efficiency is very low.
Pt of leading and next-to-leading muons
pT of the leading and next-to-leading muons (generator level) with corresponding trigger thresholds for low luminosity

Studies to improve the efficiency for these events by using a dedicated trigger algorithm are currently ongoing.

tau->3mu mass resolution (Approved in CMS Physics Meeting 30.08.2007)

To simulate the entire detector and for event reconstruction CMSSW_1_3_1 was used.
Wsource 3Mu invariant mass
Reconstructed invariant mass of tau->3mu events
The reconstructed invariant mass of tau->3mu is depicted in the figure above. The mass resolution is about 24 MeV, which ensures a good capability to reduce background events.

Study of background sources

The main sources of muons in CMS that can contribute to the background are decays of heavy mesons from charm and bottom production, gauge bosons (Z,W) including Drell Yan, pile-up effects and cosmic rays. The probability to misidentify an event from pile-up is small and cosmics rays can be rejected by timing. Due to the high momentum of the muons from direct W and Z decays, the contribution to the background is negligible (hep-ex/0210033).
The major background for the decay channel tau->3mu is coming from decays of heavy mesons, due to the considerable high production cross section for charm and even more bottom production at LHC energies in CMS. A previous study (CMS Note 2002/037) suggested that this kind of background events are reducible by background rejection cuts.
One rare decay that can mimic the signal is Ds->phi mu nu followed by an decay phi->mu mu. This background can be reduced by an invariant mass veto around the phi mass. Although with a branching ratio of only 10^-5 , the radiative decay phi->mu mu gamma can not be reduced significantly with the invariant mass cut, since the photon is usually not detected. To decrease the background coming from these radiative decays or other heavy meson decays as well, additional observables like secondary vertex properties and isolation criteria seem to be promising.

Expected limit

Predictions of the achieveable limits are available in an older CMS Note. Their expected upper limit at 95% CL for the W source is B≥ 7.0x10^-8 with 10 fb^-1 and B≥ 3.8x10^-8 with 30 fb^-1 of collected data. For the Z source a limit of B≥ 3.4x10^-7 and for the B meson source a limit of B≥ 2.1x10^-7 was derived. For both sources an integrated luminosity of 30 fb^-1 was considered. The Ds source with the highest tau production rate was not studied in this early paper.