Quark Compositeness at Hadron Colliders

1
Quark Compositeness atHadron Colliders

• Robert M. Harris
• Fermilab CD/CDF
• Snowmass July 7, 2001
• Updated for Arlington LC Workshop, Jan 10, 2003

1
2
Outline

• Introduction and Early History
• The high ET jet excess at CDF
• Uncertainties in searching for compositeness.
• Searches using jet angular variables at the
  Tevatron.
• Dilepton searches for contact interactions.
• Currently available estimates for Run 2, LHC,
  VLHC.

3
Introduction

• New physics with energy scale L can manifest
  itself at parton collision energies s½ lt L in
  four fermion contact interactions.
• The canonical form is a left handed interaction
  among composite quarks.

L (2p/L2 ) (qLgmqL) (qLgmqL)

• In hadron collisions these interactions produce
  more events with high Jet ET , and more isotropic
  angular distributions, than expected from QCD.

4
Experimental History

• Compositeness was searched for, uneventfully, in
  the ET and angular distributions of jets in pp
  collisions at the SppS at CERN and the Tevatron
  at Fermilab in the 80s and early 90s.
• The ET distribution appeared to be more sensitive
  to a contact interaction, by virtue of the
  increased statistics.

Date Published 1985 1986 1989 1991 1992
Experiment UA2 UA1 CDF UA2 CDF
L Limit (TeV) ET Dist. 0.37 0.4 0.7 0.825 1.4
L Limit (TeV) Angular Dist. - 0.415 0.33 - 1.0
5
Things get interesting

• The first data that could be interpreted as a
  possible signal were published by CDF in 1996
  from run 1A.
• Best fit L 1.6 TeV
• But, large systematics
• Large experimental uncertainties highly
  correlated vs. ET
• Large parton distribution uncertainties.
• Run 1A PRL 77, 5336 (1996)
• Run 1B PRD 64, 032001 (2001)

6
Experimental ET Uncertainties

• Uncertainties in ET are large
• However, CDF D0 data are in good agreement,
  increasing confidence.

7
Parton Distribution Uncertainties

• Parton distributions at the time did not show
  variations that could account for the CDF jet
  excess.
• The CTEQ4HJ parton distribution included the CDF
  jet data, with a heavy weighting, to see if it
  was possible to fit the CDF data while being
  compatible with the worlds parton dist data.

8
Yes gluon, No compositeness, No limit.

• The CDF data compared to QCD using the CTEQ4HJ
  parton distributions showed less excess at high
  ET
• CTEQ4HJ has more gluons in the proton at high x
  than was previously thought possible.
• Compositeness, or other new physics beyond the
  SM was no longer needed.
• L limits were never quoted.

9
Angular Distributions look like QCD

• The dijet angular distributions from CDF in 1996
  were in good agreement with QCD.
• This was compatible with the hypothesis that
  either parton distributions or ET systematic
  uncertainties were responsible for the high ET
  excess.
• PRL 77, 438 (1996)

10
Angular Ratio excludes L1.6 TeV

• The Angular Ratio was used for the search and
  limit on L
• Rc N(c lt 2.5)/N( 2.5 lt c lt 7.5)
• Compositeness gives high Rc at high dijet mass.
• Rc from CDF is flat vs. mass, in good agreement
  with QCD.
• Excludes L 1.6 TeV, the best fit scale from
  the CDF ET data compared to CTEQ4M.
• For the first time the angular distribution limit
  is published with no published limit from ET.

11
Ratio has Small Theory Uncertainties

• The variations in the calculation of Rc are small
  compared to the ET distribution
• Parton distribution uncertainties are very small
• Angular dist is reflection of fundamental QCD
  matrix elements, not parton dists.
• Renormalization scale uncertainties are
  manageable
• Mainly a shift in Rc while a compositeness signal
  produces an emerging excess in Rc at high dijet
  mass.

12
Flavor Symmetric Contact Interactions

• Previous to 1996 all L limits quoted were for a
  model where only u and d quarks were composite.
• Call these limits on Lud
• Lane introduced flavor symmetric contact
  interactions in which all quarks are composite in
  hep-ph/9605257.
• More processes involved in contact interaction.
• Hadron colliders more sensitive to flavor
  symmetric contact.
• From now on I quote numbers for flavor symmetric
  L
• CDF limit of Lud gt 1.6 TeV g L gt 1.8 TeV

13
D0 Angular Distribution Search

• The D0 angular distribution search published in
  1998 was more sensitive than CDF search
• Again an angular ratio is used.
• Less detector systematics (cracks at D0 less
  disruptive than CDF).
• Excludes L gt 2.1 TeV.
• Again the angular distribution limit is published
  without any L limit from the ET distribution.
• Limit from the SET distribution published in
  2000 is less restrictive L gt 2.0 TeV
• PRL 80, 666 (1998)
• PRD 62, 031101 (2000)

14
D0 Mass Ratio Search

• 1999 D0 Ratio of dijet mass distributions in two
  h bins
• ds/dM( hlt0.5 ) / ds/dM(0.5lthlt1)
• Like the angular ratio variable it is insensitive
  to parton dists has small expt. systematics
• Uses more of the D0 data since there are no cuts
  in h or hboost and the h variable gives larger
  triggering efficiency.
• More statistical sensitivity.
• Limit L gt 2.7 TeV
• PRL 82, 2457 (1999)

15
Lessons from the past for jets

• Uncertainties in jet ET dist made it a poor probe
  of quark compositeness
• Significant experimental and parton distribution
  uncertainties.
• The jet ET distribution was best used to measure
  the gluon at high x.
• Recent work to provide parton distribution with
  uncertainties will help.
• Giele, Keller Kosower, hep-ph/0104052.
• CTEQ Pumplin et al., JHEP 0207012,2002,
  hep-ph/0201195.

2.0
CTEQ5HJ
Systematic Uncertainty Band
Gluon / CTEQ6 Gluon
1.0
MRST2001
CTEQ5M1
0.5
.01
.1
1
X

• Angular variables are currently the best probe of
  contact interactions.
• Small experimental and theoretical uncertainties.
• Possible to have similar statistical sensitivity
  as ET distribution.

16
Common Quark-Lepton Compositeness

• If quarks and leptons share common constituents
  then contact interactions arise in di-lepton
  production at hadron colliders.
• Dielectrons are relatively background free.
  Better 95 CL limits.
• Combination of Tevatron, HERA, LEP, nN, and APV
  (Cheung, PL517,167,2001)
• L(eedd,uu) gt 11.1, 23.3TeV and L-(eedd,uu) gt
  26.4,12.5 TeV
• CDF has estimated the 95 CL reach at Tevatron in
  run 2 (hep-ph/9610382)
• Run 2A (2fb-1) L(eeqq) gt 6.5 TeV and L-(eeqq) gt
  10 TeV
• TeV33 (30fb-1) L(eeqq) gt 14 TeV and L-(eeqq) gt
  20 TeV
• LHC sensitive to L30 TeV at 5s with 100 fb-1
  (Eur. Phys. J. direct C4 N1, 2002.)
• 0.5 TeV LC with polarization 500 fb-1 sensitive
  up to L68 TeV, 95 CL
• If LHC sees a signal then LC will study it
  (Linear Collider Physics, SLAC-R-570,June 2001)

Experiment CDF D0 Aleph Opal Delphi L3
L(eeqq) L-(eeqq) 2.5 3.7 3.3 4.2 5.4 6.2 5.5 3.1 2.4 2.8 4.2 2.8
17
LHC Jet Sensitivity for 300 fb-1

• ATLAS studies
• Eur. Phys. J. direct C4 N1, 2002.
• ET Distribution
• Considers stat and systematic err.
• Parton distribution sys of 10 (low).
• Jet ET sys estimated to be in the range 1.5 - 5
  at ET 3 TeV.
• Produces 15 - 50 deviation.
• Always less than L20 TeV signal
• Sensitive to L20 TeV at 95 CL.
• Angular Distribution
• Systematic errors found negligible.
• Sensitive to L40 TeV at 95 CL.
• Discovery of L14 TeV after 1 month of operation
  at L1033 cm-2s-1.

18
Looking towards the future

• Possible that LHC finds a contact interaction
  around L 25 TeV
• A VLHC could determine that the contact
  interaction is due to quark compositeness.
• VLHC could see the excited states of composite
  quarks (q g qg) around M 25 TeV.
• Direct observation of resonances.
• R. Harris, Snowmass 96, hep-ph/9609319.

19
Conclusions

• Jet ET and angular distributions have been used
  for two decades to search for contact
  interactions at hadron colliders.
• At UA1, UA2, CDF and D0 best limit currently L
  gt2.7 TeV at 95 CL.
• Contact effects in jet ET are hard to find due to
  large systematics.
• A program of quantifying the parton distribution
  uncertainties is underway and should help give
  more reliable searches using ET in the future.
• Angular distributions are more sensitive due to
  lower systematics.
• Dileptons are a good search channel due to low
  backgrounds.
• If LHC sees a signal here, LC can study the
  inverse process in detail.
• Possibilities for contact interactions at future
  hadron colliders.
• An LHC could see a L 25 TeV contact interaction
  in dijets.
• A VLHC could then directly see the physics behind
  the contact.