34
Di-boson Physics at the Tevatron Fourth Workshop on Mass Origin and Supersymmetry Physics Tsukuba, Japan March 6, 2006 Al Goshaw , Duke University (with thanks to CDF and D0 Colleagues)
Di-boson Physics at the Tevatron
Fourth Workshop on Mass Origin and Supersymmetry Physics
Tsukuba, JapanMarch 6, 2006
Al Goshaw , Duke University(with thanks to CDF and D0 Colleagues)
2
Outline
Introduction
Survey of recent measurementsW(l) and Z(l l) WW and W Z studies using leptonic decays
Search for W/Z -> q q signals in di-boson events
Summary and outlook
3
Di-boson physics at the Tevatron
Measurements of di-boson production provide a rich source of electroweak Standard Model tests and are a natural avenue into searches for the Higgs boson.
Vector boson pair production includes (this talk): W W W H
W W Z Z H Z Z Z
The CDF and DØ experiments have completed the first analysis phase based upon ~ 400 pb-1 of p p integrated luminosity. Ultimate sensitivity will be based upon 4-8 fb-1 And of course continuation at the LHC …
4
p p production of W and Z bosons at √s = 1.96 TeVThe general landscape
High statistics W/Z inclusiveand W mass
Lower statistics di-bosons
Limits on H production
ET > 10 GeV R(l) > 0.7
5
Approach to di-boson studies
1. Compare production properties to Standard Model predictions and measure agreement/deviations.
2. Use anomalous coupling parameters as the metric for evaluating the sensitivity to new physics. This assumes the new physics appears as deviations of the W and Z boson from Standard Model point particles.There are of course other sources of new physics that would appear in di-boson production -- perhaps the most likely sources of a discovery.
3. Use the advantage of having both q q and q q ’ collisions to separate out specific triple gauge couplings where possible: q q ’ -> W* -> W WW coupling only q q ’ -> W* -> W Z WWZ coupling only q q -> Z/ -> W W mix of WW and WWZ couplings q q -> Z/ -> Z mix of ZZ and Z couplings q q -> Z/ -> Z Z mix of ZZ and ZZZ couplings
absent in SM
6
Approach to di-boson studies
4. For the triple gauge coupling studies use leptonic decays of the W and Z: W -> l Z -> l+ l- W+ W--> l+ l- W Z -> l’ l+ l- Z Z -> l+ l- l’ + l’ - and l+ l- where l = e or
5. Extend measurements to W/Z hadronic decay channels Specific channels: W/Z(jet-jet) + and W/Z(jet-jet) + W(l) Useful for calibration/improvement of di-jet mass resolution Prototypes for WH and ZH searches
7
W studies using p p -> l + x
QuickTime™ and aTIFF (LZW) decompressor
are needed to see this picture.
CDFp p -> + x
candidate
Z studies using p p -> l+ l- + x
Photon probes of W and Z bosons
8
p p -> l + x Production
The l final states have contributions from quark and lepton bremsstrahlung processes and the direct production W -> l
The first two diagrams involve W boson coupling only to fermions, and are assumed to be described by the Standard Model .
The third diagram depends on the WW coupling
Therefore the production p p -> l + x can be used to measure this coupling.
triple gauge couplinginitial state radiation final state radiation
W
WW
9
p p -> l+ l- + x Production
Within the confines of the SM, the ZZ and Z couplings are zero at tree level.
Destructive interference of the W -> l process with the initial state bremsstrahlung process suppresses the l cross section. For p p collisions at √s = 1.96 TeV the SM expectations are:
[W(l)] / [Z(ll)] ~ 10.7 for ET > 0
[lll] ~ 4.3 (1.5) for ET > 10 (100) GeV with R(l) > 0.7
triple gauge couplinginitial state radiation final state radiation
Z/Z/
Z/l
l
l
Z
Absentin SM
10
Data selection for l and l+ l- events
Events triggered on high ET/PT central electron/muon
Selection of leptons similar to inclusive W/Z measurements
Selection of photons central: ~ 1.0 photon energy: ET > ~ 8 GeV isolated: R(l ) > 0.7
Dominate background from W(l )+jets and Z/(l l)+jets with jet -> fake photon Use jet data samples to measure the jet -> fake rates Correct for real photon content Apply this to jets in W/Z + jet data to get W/Z + fake photons
11
Electron and muon ID efficiencies are evaluated from data using Z -> ee and inclusive decays.
Photon ID efficiencies are determined from a combination of data (use electrons as proxies for photons) and GEANT-based detector simulations.
Geometric acceptances determined from SM event generators and detector simulations.
Quote cross sections corrected for: full W -> l decay phase space full Z-> l l decay phase space for M(ll) > 30 (40) GeV/c2 D0
(CDF) Photon phase space for R(l ) > 0.7 and ET() > 7 (8) GeV
CDF (D0)
Data corrections for l and l+ l- events
12
(p p -> l + X) (pb) with R(l ) > 0.7
Ndata
(e + )
(l )
exp.
(l )
SM theory
ET
cut
CDF 323 18.1+3.1 19.3+1.4 > 7 GeV
DØ 273 14.8+2.1 16.0+0.4 > 8 GeV
Comparison of l and l+ l- data to Standard Model predictions
(p p -> l l + X) (pb) with R(l ) > 0.7
Ndata
(e + )
(l l )
exp.
(l l )
theory
ET
GeV
M(l l ) GeV/c2
CDF 71 4.6+0.6 4.5+0.3 > 7 > 40
DØ 290 4.2+0.5 3.9+0.2 > 8 > 30
p p at √s = 1.96 TeV
PRL 94, 041803 (2005)
PRL 95, 051802 (2005)
PRL 94, 041803 (2005)
PRD 71, 091108 (2005)
13
Comparison of l signal to Standard Model predictions
W +
final state radiation
DØ
CDF
CDF
14
Comparison of l+ l- signal to Standard Model predictions
CDF
DØ
DØ
final state radiation
Z+
DY +
15
Introduction of anomalous couplingsparameters
Under the assumption of Lorentz and electromagnetic gauge invariance, for massless fermions, the WW coupling can be described in terms of four parameters.
The effective Lagrangian is [ Baur and Berger PRD 41, 1476 (1990) ]
LWW = -ie [(W W A - W A W
)
+ W W F + /MW
2 W W F
+ 2 more CP violating terms
The magnetic dipole and electric quadrapole moments of the W boson are given by:
W = (1 + + )e/2MW and QW = -(- )e/MW2
In the SM at tree level = - 1 = = 0. Estimates of loop
corrections are small: || = 0.008 and || = 0.002 .
Strong constraints from limits on the neutron’s electric
dipole moment
16
The anomalous coupling parameters must be suppressed at high energies to avoid unitarity violations.
For the WW vertex these are assumed to be of the form:
= o/(1 + s/2)2
= o /(1 + s/2)2
where is the scale of the new physics and √s is the W invariant mass.
Similarly, under the assumptions of Lorentz and electromagnetic gauge invariance, the anomalous coupling parameters for the ZV vertex are: hi
V = hioV /(1 + s/2) n where V = s-channel or Z
√s = Z invariant mass i = 1,2 (CP violating), 3,4 (CP conserving) ==> 8 parameters [see e.g. Baur and Berger PRD 47, 4889 (1993)]
Introduction of anomalous couplingsparameters
17
Using l events to put limits on WW couplings (DØ)
Non-zero or lead to enhancement of high ET photons above the SM prediction.
Suppress event with FSR by selecting events with MT(l ) > 90 GeV/c2
Use binned-likelihood fitting ET on o vs ogrid
-0.88 <o < 0.96
-0.20 <o< 0.20
2D 95% C.L.
1D 95% C.L.
= 2 TeV o
o
18
Using l+ l- events to put limits on ZZ and Z couplings
(DØ) No deviations from SM predictions, but very clean data samples can be used to put limits on anomalous couplings.
Form factors imposed to preserve unitarity with n= 3 for i = 1,3 and n = 4 for i = 2,4 (see form on page 17).
Use binned-likelihood fitting to ET on 2D grid of (h10V vs h20
V) and (h30
V vs h40V). Set 95% C/L. for = 1 TeV. (D0 PRL and
hep-ex/0502036).
19
W+ W- studies using p p -> l+ l’- +
x
CDFp p -> e + x
candidate
e
‘s
WZ studies using p p -> l’ l+ l- +
xDØ
p p -> + xcandidate
WWZ probes of W and Z bosons
20
p p -> W+ W- + x Production
W+ W- pair production is sensitive to both the WW and WWZ) couplings.
The SM expectation for p p - > W+ W- +X at √s = 1.96 TeV is ~ 12.4 pbTo date CDF and DØ have used channels.: l l‘ with l , l‘ = e or BR ~ 4.6% small BR, good S/B q q’ l with l = e or BR ~ 29% good BR, poor S/B
WW Fraction %
qq’ q q’ 46.2%
q q’ l 43.5%
l l‘ 10.3%
WWZ and WW
destructive interferencecauses WW
cross section suppression
21
Comparison of W+W- to Standard Model predictions using l+ l’- decays
Measurement of p p -> W+ W- +X at √s = 1.96 TeV with BR corrections
SM theory predictions at NLO with MCFMExperimental details can be found at: DO hep-ex/ ( Published in PRL) CDF hep-ex/ (Published in PRL)
pb .)(9.0
.)(.)(8.13)( 2.19.0
3.48.3
lum
sysstatWW
±= +
−+−
pb .)(9.0
.)(.)(6.14)( 8.10.3
8.51.5
lum
sysstatWW
±= +
−+−
CDF Run II: PRL 94, 211801 (2005)DØ Run II: PRL 94, 151801 (2005)
SM (WW) = 12.4 + 0.8 pb
22
This analysis is used to simultaneously extract event channels leading to opposite sign di-lepton events: t t -> l+ l- + MET + high ET jets W+ W- -> l+ l- + MET + soft jets Z -> -> l+ l- + low MET + soft jets X+ X- -> l+ l- + ? (search for new physics)
Select events with opposite sign di-leptons using standard cuts: e+ e-
e+ e-
Fit these events for signals in a 2 dimensional phase space of MET vs number of jets with ET > 15 GeV and | | < 2.5
Minimal selection cuts => maximum sensitivity to signals
New measurement from CDF (Feb. 2006)An Inclusive Di-lepton Analysis (AIDA)
ET (PT) > 20 GeV (GeV/c)
23
Backgrounds Drell-Yan Z/ - > ee/ use PYTHIA MC cross checked with data W , WZ and ZZ use SM predictions Fake leptons in W + jets obtain jet -> lepton fake rate from data
New measurement from CDF (Feb. 2006)An Inclusive Di-lepton Analysis (AIDA)
24
New measurement from CDF (Feb. 2006)An Inclusive Di-lepton Analysis (AIDA)
25
New measurement from CDF (Feb. 2006)An Inclusive Di-lepton Analysis (AIDA)
New CDF (WW) = 16.7 pb+5.1
-4.3
26
Comparison of WZ (and ZZ) to Standard Model predictions using all leptonic
decays
[p p -> WZ (ZZ) + x ] at √s = 1.96 TeV corrected for W/Z branching ratios
SM theory predictions at NLO with MCFMExperimental details can be found at: DØ hep-ex/0504019 (Submitted to PRL) CDF PRD 71, 091105 (2005)
Bottom line: consistent with the SM
Integrated Luminosity
~ 300 pb-1 (DØ) ~ 200 pb-1 (CDF)
Ndata Nbackgrond SM theory
cross section (pb)
Experimental
cross section (pb)
95%
C.L.
CDF(WZ+ZZ) 3 1.02+0.24 5.0 + 0.4 4.3 +5.0 - 2.6 < 15.2 pb
DØ (WZ) 3 0.71+0.08 3.65 + 0.26 4.5 +5.1 - 3.3 < 13.3 pb
27
Search for W/Z boson hadronic decays
Searches for W/Z-> q qusing W/Z + W(l ) events
28
Searches for p p -> W/Z(q q) + XMotivation
New physics searches using dijets depend critically on a good understanding of the jet-jet invariant mass resolution.
One calibration source is W/Z -> q q -> jet jet
This requires a trigger that does not bias the W/Z mass peak, and allows low mass side bands for background subtraction.
Using diboson events of the type W/Z(q q) and W/Z(q q) W(l ) allows a trigger selection based upon the a high Et photon or lepton, and provides an unbiased look at the jet-jet spectrum for extraction of W/Z -> q q .
Also, the W/Z(q q) W(l ) channels are very similar to those used for Higgs searches in H(b b) W(l ) and provide a SM calibration line.
29
Search for p p -> W/Z(q q) + W(l )
The channel p p -> W(l ) + W/Z (q q ) has been studied by CDF Advantages: larger branching ratio Disadvantages: much higher backgrounds BUT anomalous signals appear at
high ET of the W where backgrounds lower
W + jet jet QCD background is constrained by fitting to dijet mass spectrum around MW plus MZ peak.
Fit to data:Ndata = 109 + 110 + 54 events
data(WW + WZ) < 36 pb
With the SM expectationNSM ~ 160 events
SM(WW + WZ) = 16.5 pb
New CDF March 2006
30
Fits to anomalous couplings require assumptions here since 5 parameters contribute to WW plus WZ production: g1
z , z , z, and
Assume g01z= 0 and let o = oz = oand o = oz= o
The PT of the W(l ) is found to be the most sensitive distribution since anomalous VV pairs are produced at high PT .
Limits set:
- 0.51 < o < + 0.45
- 0.29 < o < + 0.29
Search for p p -> W/Z(q q) + W(l )
New CDF March 2006
31
Summary and Outlook
32
All rates and kinematic distributions are consistent with SM predictions
(table uses nominal SM predictions with no theory uncertainties)
SUMMARY: Comparison of diboson production to SM predictions
Channel (l =e (data -SM )/SM
W [l ] -0.06 + 0.16 CDF
-0.06 + 0.16 DØ
Z [l l ] +0.02 + 0.13 CDF
+0.08 + 0.13 DØ
WW [l l ] +0.35 + 0.42 CDF
+0.10 + 0.32 DØ
cross section limits data (95% C.L.)/ SM
WZ [l l l ] 3.3 DØ
WZ + WW [l qq] 2.2 CDF
ZW + ZZ [l l (l or )] 3.0 CDF
Tevatron Run IIp p at √s = 1.96 TeV
200-400 pb-1
33
Coupling limits at 95% CL Energy scale
WW - 0.88 < o < 0.96
2 TeV
- 0.20 < o< 0.20
WWZ - 0.49 <g01z <
0.66 1.5 TeV
- 0.48 <oz< 0.48
ZZ |h10,30| < 0.23 1 TeV
|h20,40| < 0.019
ZZZ |hz10,30| < 0.23 1 TeV
|hz10,30| < 0.020
WWZ, WW - 0.51 < o < 0.45
1.5 TeV
- 0.29 < o< 0.29
SUMMARY: Limits on anomalous couplings
Analyses just starting on individual channels
Need to combine channels and CDF+D0 measurements
34
New physics sources (anomalous couplings, new fermions or gauge bosons) contribute to the high PT tails of W/Z/ production.
At high PT most sources of background (jets faking photons and leptons) fall rapidly.
Therefore the sensitivity to new physics is almost entirely statistics driven.
The Tevatron is ramping up according to its design plan, and the CDF and DØ detectors are operating with good efficiency.
The data sets presented here represent 5-10% of the potential of the Tevatron.
Di-boson channels will be some of the first measurements at the LHC where there is good hope that the SM signals will be badly polluted with new physics!
SUMMARY: Physics beyond the SM
