Search for New Physics in Dileptons+X

A Signature Based Search for New Physics in Dileptons+X

I describe a search for new, heavy objects decaying into high \pt~ dileptons using 929 pb^-1 of data. Starting with the inclusive electron and muon samples, I maximize acceptance by including all three generations of leptons, loosening fiducial requirements and loosening the p_T requirement of the 2nd lepton from E_t> 20 Gev to E_t > 12 GeV. To distinguish new physics involving the 3rd generation, I include b-tagging from the start. Initially using the e-mu and same sign channels, and to probe the reach of the search, I set a cross section limit on a 300 GeV right handed down type quark predicted by Bjorken, Pakvasa and Tuan. The limit is 1.4*sigma_Q where sigma_Q=0.289 pb is the expected cross section.

Contacts: Collin Wolfe, Sasha Paramonov, Steve Levy, Carla Pilcher, Henry Frisch
Data: Run II, 929 pb^-1
Documentation: Public Note PS

This analysis is a search for anomalous dilepton+X events where X can be:
- Large HT
- High ET Jets
- b tags
- Third leptons
- Large Missing Et
Because these dilepton+X signatures are rare in the SM, we've increased acceptance by improving lepton ID acceptance.
b-tagging is included from the start to separate 3rd Generation.

To probe the reach of our search, we set a limit on a 300 GeV right handed down type quark predicted by Bjorken, Pakvasa, and Tuan. Details about the model:

Reference: Bjorken, Pakvasa, Tuan: Yet Another Extension of the Standard Model: Oases in the Desert?, hep-ph/0206116
Idea is to take an ansatz to the question: ``Why does the CKM matrix have the structure it does?'' Then take the simplest physics which explains the ansatz and examine the consequences of the physics.
Assume CKM matrix is due to a diagonal up quark mass matrix and a mostly diagonal down quark mass matrix with a small perturbation (the so-called Stech texture) due to mixing of three down type iso-singlet right-handed quarks.
We take the BPT model for concreteness, but this is a signature based search broad enough to catch many models such as a $b'$ suggested by Rosner, $t'$, Extra Dimensions, SUSY...
We've taken the mass to be 300 GeV to probe our reach.
The new quarks correspond naturally to the three families, with decays:
- $D_i \rightarrow u_i W$
- $D_i \rightarrow d_i Z$
- $D_i \rightarrow d_i H$
- where $i=d,s,b$
- These have relative branching ratios for $W:Z:H$ of approximately $2:1:1$
Note: If $D_d$ is the lightest of the heavy quarks, we would find a heavy quark decaying to a d quark, not a b quark.
Down type quarks produced via qcd exactly like top, implies if the Higgs mass is less than the $D_d$ mass, we are producing the Higgs with strong cross section!
This model has been implemented in MadGraph (see CDF7516)

My control regions are e-e with Ht < 200 GeV, mu-mu with Ht < 200 GeV, e-mu with Ht < 200 GeV and e-mu+btag with no cut on Ht. By understanding the variables which go into Ht in these regions, I have confidence that I understand the Ht in the signal region. The optimized cuts (on S/sqrt(B))for this analysis are Ht>400 GeV and two jets with Et > 50 GeV. With these cuts, in 929 pb^-1, we expect 2.9 +/- 1.5 events from Standard Model backgrounds and 1.9 +/- 0.2 events from the BPT model. We see 2 events and set a Bayesian limit at 90% CL of 2.1*sigma_Q where sigma_Q=0.289 pb is the expected cross section. In addition to the e-mu channel, the model will contribute to same sign (SS) leptons at large Ht and high Et jets. At large Ht, most of the SM contribution comes from W+jets where a jet fakes a lepton. From the e-mu channel, we know this is well understood. Using the same signal region as in the e-mu channel, we expect 1.5 +/- 0.75 events from SM backgrounds and 0.9 +/- 0.09 events from the BPT model in 929 pb^-1. This gives a limit of 2.1*sigma_Q. Combining the SS and e-mu channels together, we find a cross section limit at 90% CL of 1.4*sigma_Q where sigma_Q=0.289 pb is the expected cross section. The errors are systematic and due almost entirely to the Jet Energy Scale.

Plots and Tables:

Control Region Plots
	Comparison of SM MC and Data in Ht in ee events. Linear Scale GIF. EPS.
	Comparison of SM MC and Data in Ht in ee events. Log Scale GIF. EPS.
	Comparison of SM MC and Data in Ht in mu-mu events. Linear Scale GIF. EPS.
	Comparison of SM MC and Data in Ht in mu-mu events. Log Scale GIF. EPS.
	Comparison of SM MC and Data in M_e-mu in e-mu events. Linear Scale GIF. EPS.
	Comparison of SM MC and Data in M_e-mu in e-mu events. Log Scale GIF. EPS.
	Comparison of SM MC and Data in Ht in e-mu events. Linear Scale GIF. EPS.
	Comparison of SM MC and Data in Ht in e-mu events. Log Scale GIF. EPS.
	Comparison of SM MC and Data in Ht in e-mu events out to 1000 GeV (no overflow). Log Scale GIF. EPS.
	Event display of 864 GeV Ht e-mu event. GIF. EPS.
	Event display of 864 GeV Ht e-mu event. GIF. EPS.
	Event display of 864 GeV Ht e-mu event. GIF. EPS.
	Event display of 864 GeV Ht e-mu event. GIF. EPS.

*High Pt objects in 864 GeV Ht e-mu event*
What	ET	eta	phi	px	py	pz
mu^-	159	0.23	2.28	-103.3	120.6	36.6
e^+	28.4	0.38	3.68	-24.4	-14.5	11.0
b	216	-0.02	5.13	88.7	-197.9	-4.1
j	199	0.13	2.89	-193	48.7	27.0
j	156	1.15	0.09	155.5	14.7	223.0
j	15	0.81	5.67	12.1	-8.5	13.4
met	89.5	--	0.70	85.7	25.7	--

*Expected Events in 305/pb from BPT Signal and total SM background for Ht > 400 GeV and two jets with Et > 50 GeV.*
Channel	Expected Events
QQ -> ddHH	0.0246 +/- 0.0017
QQ -> ddHZ	0.0535 +/- 0.0053
QQ -> ddZZ	0.0107 +/- 0.0024
QQ -> udWH	0.157 +/- 0.009
QQ -> udWZ	0.0834 +/- 0.009
QQ -> uuWW	0.197 +/- 0.020
Total (with Systematics)	0.536 +/- 0.058
SM (with Systematics)	0.802 +/- 0.440