SUSY Higgs Searches in Multi-jet Final States: bh/bbh(bb)


More Information:

Get our conference note! ( PDF , POSTSCRIPT )

Contact persons: Andy Haas (haas@yahoo.com), Avto Kharchilava (avto@fnal.gov), Gordon Watts (gwatts@fnal.gov)

EPS files for Feynman diagrams, LO production at low, high tanB, and masses and widths.


Introduction:

This search is for neutral Higgs bosons in a two-doublet model of Electro-Weak Symmetry Breaking, such as Supersymmetry. We look for a signal in the invariant mass spectrum of the two jets with the highest transverse energy in triple b-tagged multi-jet events. Feynman diagrams for the signal are shown below. The signal cross-section is proportional to the square of tan β, the ratio of the VEV's of the two complex Higgs doublet fields.




Data:

The data used for this analysis was taken between Sept. 2002 July 2003.
131 pb-1 remains after all data quality criteria.

Trigger 1: 75 pb-1
Level 1: Four calorimeter towers with ET >5 GeV.
Level 2: Three jets with ET >8 GeV and | η|<2.4, with total HT >50 GeV.
Level 3: Three jets with ET >15 GeV and | η|<3.

Trigger 2: 56 pb-1
Level 1: Three calorimeter towers with ET >5 GeV.
Level 2: Three jets with ET >8 GeV and | η|<2.4, with total HT >50 GeV.
Level 3: Three jets with ET >15 GeV and | η|<3, with two jets having ET >25 GeV .
(The jet algorithm was improved for trigger 2, and took advantage of the primary vertex Z position to calculate ET more accurately.)

Run II Cone jets, with radius of 0.5 were used.
Initial selection required: 3 jets with ET >15 GeV and | η|<2.5, with one of them having ET >20 GeV .
Afterwards, standard jet quality criteria and the jet energy scale was applied to each jet.
At least 3 taggable jets were then required in each event.
The cuts on the ET
and | η| of jets, and the maximum number of jets, were optimized separately for each candidate Higgs mass, as shown in the table below:




Analysis:

We have derived the composition of the double b-tagged data, from which the triple b-tagged data background was derived.
The double b-tagged data with 4 or more jets is shown in the first figure below.

SVX Extra Loose b-tagging was applied, and events were kept which had 3 or more tags.

The shape of the triple b-tagged background was estimated from the double b-tagged background and extrapolated using a tag-rate-function derived on the full multi-jet data sample. This background shape was then normalized to the triple b-tagged data outside a 1 σ signal window determined for each Higgs mass.

The MCLimit routines as implemented in ROOT were used to set limits on the signal production, by feeding in simulated signal shapes (for each tan β), the normalized background, and the observed data.


Double b-tagged data

The double b-tagged data with 4 or more jets, showing the composition of various sources.
Get the EPS version!


Results:

An example of the excluded signal and normalized background for a Higgs mass of 120 GeV, compared to the data, is shown below.

In the absence of an observed signal, limits have been set in the tan β vs. mA (the mass of the CP-odd neutral Higgs) plane, as shown below.


The triple b-tagged data with 3 or more jets, showing the composition from various sources, and the excluded signal for mA=120 GeV.
Get the EPS version!


Limit Plot
The excluded region in the MSSM Higgs parameter space.
Get the EPS version!



E-mail to haas@fnal.gov