Transcript Study of tt production at NLO

The top quark at LHC:
status and prospects
Marina Cobal-Grassmann
“Journee ATLAS France”
Londe Les Maure, 3-5 May, 2004
1
P2
Motivations for Top Physics studies


Top quark exists and will be produced abundantly!
In SM: top- and W-mass constrain Higgs mass
Sensitivity through radiative corrections
 Scrutinize SM by precise determination top mass

Summer 2003 result

Beyond SM: New Physics?


direct
Many heavy particles decay in tt
Handle on new physics by detailed
properties of top

Experiment: Top quark useful to
calibrate the detector

Beyond Top: Top quarks will be a
major source of background for
almost every search for physics
beyond the SM
indirect
Marina Cobal - Londe Les Maure 2004
P3
What we know..
LEP+SLD:
VCKM (4)
0,l ,c,b
lept
mZ ,W ,Z ,W , h0 , Rl0,b,c , AFB
, Al ,b,c , sin 2 eff
UA2+Tevatron:
s(1)
GF (1)
SM
mfermions (9)
mH
NuTeV: sin W
2
predictions
(down to 0.1% level)
mbosons (2)
mW ,W , mt
APV:
eeqq l.e.:
QW (Cs)  sin 2 W
 had
No observable directly related to mH. However the dependence can
appear through radiative corrections.  tree level quantities changed
mW2
1  
2
2
mZ cos W

2
mW 
(1  r )
2
2 sin W GF
, r = f [ln(mH/mW), mt2]
The uncertainties on mt, mW are the dominating ones in the electroweak fit
By making precision measurements (already interesting per se):
• one can get information on the missing parameter mH
• one can test the validity of the Standard Model
Marina Cobal - Londe Les Maure 2004
P4
Top mass: Where we are
Marina Cobal - Londe Les Maure 2004
P5
Near future of Mtop
Tevatron only (di-lepton events or lepton+jet ) from W decays
Status of inputs (preliminary):
mt=(178.0  2.7 (stat)  3.3 (syst)) GeV/c2
(latest Tevatron updated combination – RunI data)
mt=(175  17 (stat)  8 (syst)) GeV/c2
(CDF di-leptons – RunII data)
mt=(178+13-9 (stat)  7 (syst)) GeV/c2
(CDF lepton+jets – RunII data)
Matter of statistics (also for the main systematics) and optimized use of the
available information. Each experiment expects 500 b-tagged tt l+jets events/fb
 Mtop ~ 2-3 GeV/c2 for the Tevatron combined (2-4/fb)
mt  2.5 GeV ; mW  30 MeV  mH/mH  35%
In 2009 (if upgrade is respected) from Tevatron: Mtop = 1.5 GeV !!
Marina Cobal - Londe Les Maure 2004
P6
What can we do at LHC?
√s
[TeV]
Luminosity
[cm-2s-1]
∫L
[fb-1/y]
2
<1032
0.3
LHC (low lum)
14
1033
10
LHC (high lum)
14
1034
100
TeVatron
(pb)
Events/s
Events/y
bb
5108
106
1013
Zee
1.5103
~3
107
Wℓ (ℓ=e,μ)
3104
~60
108
WWeX
tt-
6
10-2
105
830
~1.7
107
H(700 GeV/c2)
1
210-3
104
process
-
LHC
TeVatron
Marina Cobal - Londe Les Maure 2004
P7
Top production at LHC
Cross section determined to NLO precision



sˆ  sx1 x2 ; x1 x2 ~ 103
Total NLO(tt) = 834 ± 100 pb
Largest uncertainty from scale variation
Compare to other production processes:

Low lumi
Process N/s
N/year
Total collected
before start LHC
W e
15
108
104 LEP / 107 FNAL
Z ee
1.5
107
107 LEP
tt
1
107
104 Tevatron
bb
106
1012-13
109 Belle/BaBar ?
H (130)
0.02
105
?
LHC is a top factory!
~90% gg
~10% qq

Opposite
@ FNAL
Top production cross
section approximately
100x Tevatron
Marina Cobal - Londe Les Maure 2004
P8
Top decay

In the SM the top decays to W+b
1.
Di-leptons (e/)



2.

All decay channels investigated


τ+X
21%
Single Lepton (e/)

Using ‘fast parameterized’ detector
response
Checks with detailed simulations


3.

μ+jets
15%
μ+μ e+μ e+e
1% 2% 1%
e+jets
15%
BR=29.6%  2.5x106 ev/y
One top reconstructed
Clean sample
Fully Hadronic

jets
45%
BR≈4.9%  0.4x106 ev/y
No top reconstructed
Clean sample


BR≈45%  3.5x106 ev/y
Two tops reconstructed
Huge QCD background
Large combinatorial bckgnd
Marina Cobal - Londe Les Maure 2004
P9
MTop from lepton+jet
Lepton side

Golden channel

Br(ttbbjjl)=30%
for electron + muon

The reconstruction starts with the
W mass:



Hadron side
Typical selection efficiency: ~5-10%:
•Isolated lepton PT>20 GeV
•ETmiss>20 GeV
Clean trigger from isolated lepton
different ways to pair the right jets
to form the W
jet energies calibrated using mW
Important to tag the b-jets:


enormously reduces background
(physics and combinatorial)
clean up the reconstruction
•4 jets with ET>40 GeV
Background: <2%
•>1 b-jet (b40%, uds10-3, c10-2)
W/Z+jets, WW/ZZ/WZ
Marina Cobal - Londe Les Maure 2004
P 10
Lepton + jet: reconstruct top

Hadronic side



W from jet pair with closest invariant mass
to MW
 Require |MW-Mjj|<20 GeV
Assign a b-jet to the W to reconstruct Mtop
Kinematic fit

Using remaining l+b-jet, the leptonic part is
reconstructed

|mlb -<mjjb>| < 35 GeV
j1

Kinematic fit to the tt hypothesis,
using MW constraints

j2
Selection efficiency 5-10%
b-jet
t
Marina Cobal - Londe Les Maure 2004
P 11
Top mass systematics

Method works:



Biggest uncertainties:




Linear with input Mtop
Largely independent on Top PT
Jet energy calibration
FSR: ‘out of cone’ give
large variations in mass
B-fragmentation
Verified with detailed detector simulation
and realistic calibration
Challenge:
determine the mass of the top
around 1 GeV accuracy in one year of LHC
Source of
uncertainty
Hadronic
Mtop
(GeV)
Fitted
Mtop
(GeV)
Light jet
scale
0.9
0.2
b-jet scale
0.7
0.7
b-quark
fragm
0.1
0.1
ISR
0.1
0.1
FSR
1.9
0.5
Comb bkg
0.4
0.1
Total Marina Cobal
2.3
- Londe Les0.9
Maure 2004
P 12
Alternative mass determination

Select high PT back-to-back top events:



Use the events where both W’s
decay leptonically (Br~5%)



Hemisphere separation
(bckgnd reduction, much less combinatorial)
Higher probability for jet overlapping
Much cleaner environment
Less information available from two ’s
Mtop
Use events where both W’s
decay hadronically (Br~45%)


Difficult ‘jet’ environment
Select PT>200 GeV
Various methods all have different systematics
Marina Cobal - Londe Les Maure 2004
P 13
Jet scale calibration

Calibration demands:

Ultimately jet energy scale calibrated within 1%


Uncertainty on b-jet scale dominates Mtop: light jet scale constrained by mW
At startup jet-energy scale known to lesser precision
±10%
MTop
MTop
Scale light-jet energy
Uncertainty
on light jet scale:
1%
10%
Hadronic
 Mt < 0.7 GeV
 Mt = 3 GeV
Scale b-jet energy
Uncertainty
On b-jet scale:
1%
5%
10%
Hadronic
 Mt = 0.7 GeV
 Mt = 3.5 GeV
 Mt = 7.0 GeV
Marina Cobal - Londe Les Maure 2004
P 14
Alternative methods

Determining Mtop from (tt)?


Luminosity uncertainty then plays the game (5%?)
huge statistics, totally different systematics
But: Theory uncertainty on the pdfs kills the idea





10% th. uncertainty  mt  4 GeV
Constraining the pdf would be very precious…
(up to a few % might not be a dream !!!)
Luminosity uncertainty
then plays the game (5%?)
Continuous jet algorithm




Reduce dependence on MC
Reduce jet scale uncertainty
Repeat analysis for many cone
sizes R
Sum all determined top mass:
robust estimator top-mass
Marina Cobal - Londe Les Maure 2004
P 15
Top mass from J/

Use exclusive b-decays with high mass products (J/)




Higher correlation with Mtop
Clean reconstruction (background free)
BR(ttqqb+J/)  5 10-5
 ~ 30%  103 ev./100 fb-1
(need high lumi)
MlJ/
Different systematics
(almost no sensitivity
to FSR)
Uncertainty on the bquark fragmentation
function becomes
the dominant error
M(J/+l)
M(J/+l)
Ptto
p
Marina Cobal - Londe Les Maure 2004
P 16
Search for resonances
Many theoretical models include the existence of resonances decaying
to top-topbar




SM Higgs
(but BR smaller with respect to the WW and ZZ decays)
MSSM Higgs (H/A, if mH,mA>2mt, BR(H/A→tt)≈1 for tanβ≈1)
Technicolor Models, strong ElectroWeak Symmetry Breaking, Topcolor, “colorons”
production, […]
Study of a resonance Χ once known σΧ, ΓΧ and BR(Χ→tt)

Reconstruction efficiency for semileptonic channel:


1.6 TeV
resonance
Mtt
20% mtt=400 GeV
15% mtt=2 TeV
σxBR [fb]

830 fb
xBR required for a discovery
30 fb-1
300 fb-1
1 TeV
m [GeV/c2]
Marina Cobal - Londe LesttMaure 2004
P 17
Couplings and decays

Does the top quark behaves as expected in the SM?





According to the SM:


Yukawa coupling to Higgs from ttbarH events
Electric charge
Top spin polarization
CP violation
Br(t Wb)  99.9%, Br(t  Ws)  0.1%, Br(t  Wd)  0.01%
(difficult to measure)
Can probe t W[non-b] by measuring ratio of double b-tag
to single b-tag


Statistics more than sufficient to be sensitive to SM expectation for
Br(t  W + s/d)
need excellent understanding of b-tagging efficiency/purity
Marina Cobal - Londe Les Maure 2004
P 18
Rare decays: FCNC

In the SM the FCNC decays are highly suppressed (Br<10-13-10-10)

Any observation would be sign of new physics

Sensitivity according to ATLAS and CMS studies :

t  Zq (CDF Br<0.137, ALEPH Br<17%, OPAL Br<13.7%)



t  q


Reconstruct t  Zq  (l+l-)j
Sensitivity to Br(t  Zq) = 1 X 10-4
(100 fb-1)
(CDF Br<0.032)
Sensitivity to Br(t  q) = 1 X 10-4
(100 fb-1)
t  gq



Difficult identification because of the huge QCD bakground
One looks for “like-sign” top production (ie. tt)
Sensitivity to Br(t  gq) = 7 X 10-3
(100 fb-1)
Marina Cobal - Londe Les Maure 2004
P 19
Top Charge determination

Can we establish Qtop=2/3?

Currently cannot exclude exotic possibility Qtop=-4/3

Assign the ‘wrong’ W to the b-quark in top decays


tW-b with Qtop=-4/3 instead of tW+b with Qtop=2/3 ?
Technique:

Hard  radiation from top quarks


Radiative top production, pptt cross section proportional to Q2top
Radiative top decay, tWb


On-mass approach for
decaying top:
two processes treated
independently
Matrix elements have
been calculated and fed into
Pythia MC
Radiative top production
Radiative top decay
Marina Cobal - Londe Les Maure 2004
P 20
Top Charge determination

Yield of radiative photons allows
to distinguish top charge
Q=2/3
Q=-4/3
pptt
101 ± 10
295 ± 17
pptt ; tWb
6.2 ± 2.5
2.4 ± 1.5
Total background

Determine charge of b-jet and
combine with lepton


q bjet
38 ± 6
10 fb-1
One year low lumi
Use di-lepton sample
Investigate ‘wrong’ combination
b-jet charge and lepton charge


  κ
q j  pi

i i

  κ
i j  pi
Effective separation b and b-bar
possible in first year LHC
Study systematics in progress
events
pT()
Marina Cobal - Londe Les Maure 2004
P 21
Top spin correlations

In SM with Mtop175 GeV, (t)  1.4 GeV » QCD



Top decays before hadronization, and so can study the decay of ‘bare quark’
Substantial ttbar spin correlations predicted in pair production
Can study polarization effects
through helicity analysis of
daughters


Study with di-lepton events
Correlation between
helicity angles + and for e+/+ and e-/-
d 2
1  C cos  cos 

 d cos  d cos 
4
1
C  0.34 (degree of spin correlation in helicitybase)

(spin analyserquality : 1 for leptons)
<CosΘ+ · CosΘ->
No helicity correlation
<CosΘ+ · CosΘ->
With helicity correlation
e+/+
+
top
Marina Cobal - Londe Les Maure 2004
P 22
Top spin correlations

Also study spin correlations in
hadronic decays (single lepton
events)

<CosΘ+ · CosΘ->

Least energetic jet from W
decay:  ~ 0.5
30 fb-1
Able to observe spin correlations
in asymmetry C

30 fb-1 of data:



± 0,035 statistical error
± 0,028 systematic error
10 statistical significance for a
non-zero value with 10 fb-1

Ratio between ‘with’ and
‘without’ correlations
Marina Cobal - Londe Les Maure 2004
P 23
Single top production
1) Determination of Vtb
2) Independent mass measurement
Three production mechanisms:
Wg fusion: 245±27 pb
S.Willenbrock et al., Phys.Rev.D56, 5919

+16.6
Wt: 62.2 -3. pb
A.Belyaev, E.Boos, Phys.Rev.D63,
034012
7
Main Background [xBR(W→ℓ), ℓ=e,μ]:



tt
Wbb
Wjj
σ=833 pb
[ 246 pb]
σ=300 pb
[ 66.7 pb]
σ=18·103 pb [4·103 pb]
Wg
Wt
W*
W* 10.2±0.7 pb
M.Smith et al., Phys.Rev.D54, 6696
[54.2 pb]
[17.8 pb]
[2.2 pb]
Marina Cobal - Londe Les Maure 2004
P 24
Single top results

Detector performance critical to
observe signal





Fake lepton rate
b and fake rate id 
Reconstruction and vetoing of
low energy jets
Identification of forward jets


Each of the processes have
different systematic errors for Vtb
and are sensitive to different new
physics


heavy W’  increase in the
s-channel W*
FCNC gu  t  increase in
the W-gluon fusion channel

Signal unambiguous, after 30 fb-1:
Process
Signal
Bckgnd
S/B
Wg fusion
27k
8.5k
3.1
Wt
6.8k
30k
0.22
W*
1.1k
2.4k
0.46
Complementary methods to extract Vtb
Process
Vtb
(stat)
Vtb
(theory)
Wg fusion
0.4%
6%
Wt
1.4%
6%
W*
2.7%
5%
With 30 fb-1 of data, Vtb can be
determined to %-level or better
(experimentally)
Marina Cobal - Londe Les Maure 2004
P 25
Undergoing analyses











CP violation in top events (K. Martens, University of Toronto )
Top spin polarization in di-lepton events (V. Simak et al., Prague)
Top spin polarization in single lepton events (E. Monnier, P. Pralavorio,
F. Hubaut, CPPM)
Single top studies (M. Barisonzi, NIKHEF)
Optimization of kinematic reconstruction in the single lepton channel (V.
Kostioukhine, University of Genova)
Commissioning studies (S. Bentvelsen, NIKHEF)
New MC validation (S. Bentvelsen, E. Monnier, P. Pralavorio)
Full simulation studies of detector effects (A. Etienvre, J. Schwindling,
JP Meyer, Saclay)
Full simulation studies of b-tagging (S. Moed, University of Geneva)
Top mass and calibration studies (D. Pallin, F. Binet, Clermont-Ferrand)
Ttbar resonances (E. Cogneras, Clermont-Ferrand)
Marina Cobal - Londe Les Maure 2004
P 26


What is left before the LHC
starts?
Cover topics still open: cross section, couplings, exotic,
resonances,
Define a strategy for validation of the MC input models (e.g:
UE modeling and subtraction, jet fragmentation properties,
jet energy profiles, b-fragmentation functions..)
see M. Mangano talk at IFAE 2004




Explore the effects of changing detector parameters in
evaluating the top mass.
Perform commissioning studies with top events
Contribute to simulation validation
…
Marina Cobal - Londe Les Maure 2004
P 27
Commissioning the detectors

Determination MTop in initial
phase


Period
Stat Mtop
(GeV)
Stat
/
1 year
0.1
0.2%
1 month
0.2
0.4%
1 week
0.4
2.5%
No background
included
Selection:



Use ‘Golden plated’ lepton+jet
Calibrating detector in comissioning phase
Assume pessimistic scenario:
-) No b-tagging
-) No jet calibration
-) But: Good lepton identification
Isolated lepton with PT>20 GeV
Exactly 4 jets (R=0.4) with
PT>40 GeV
Reconstruction:

Select 3 jets with maximal
resulting PT

Signal can be improved by kinematic
constrained fit

Assuming MW1=MW2 and MT1=MT2
Marina Cobal - Londe Les Maure 2004
P 28
Commissioning the detectors

Most important background for top: W+4 jets

Leptonic decay of W, with 4 extra ‘light’ jets
 Alpgen, Monte Carlo has ‘hard’ matrix element for 4 extra jets
(not available in Pythia/Herwig)

Signal plus background at initial
phase of LHC
L = 150 pb-1
(2/3 days low lumi)
ALPGEN:
W+4 extra light jets
Jet: PT>10, ||<2.5, R>0.4
No lepton cuts
Effective : ~2400 pb
With extreme simple selection
and reconstruction the toppeak should be visible at LHC
measure top mass (to 5-7 GeV)
 give feedback on detector performance
Marina Cobal - Londe Les Maure 2004
P 29
Top in DC2 Tier test

The 10M tier1 events in light of top:

Generation/simulation of 106 top events, inclusive decays, using
MC@NLO



Simulation 500K top events with displaced ID



Same truth generated top events as above
1 cm displacement of ID – check tracking performance
Simulation of 106 W+jet events MC@NLO


Using Herwig for MC + UE
Simulation with full geometry
For W+2 jet background
Simulation of 250K W+4jet events with AlpGen

pT>15 GeV approximately
Marina Cobal - Londe Les Maure 2004
P 30
What is still missing?

Top production is ‘over-weighted’ in 10M sample


One of priorities in ‘post-production’:



Unrealistic to ask for more in this sample
Regenerate half of the top MC@NLO sample
Now using Jimmy UE (much more activity), after tuning
Still on the wish-list:

Top events with spin correlations


Single top events



TopRex available, perhaps AcerMC?
also TopRex
Dedicated samples single- and dilepton top events
Top events with PYTHIA (cross check with DC1)
Marina Cobal - Londe Les Maure 2004
P 31
Conclusions
LHC is top factory
(tt)~830 pb-1
107 events in first year

Precise determination of Mtop is waiting…


Confirmation that top-quark is SM particle


Measure Vtb, charge, CP, spin, decays
Top quarks for commissioning the detectors


Challenge to get Mtop ~ 1 GeV
Top peak should be visible with eyes closed
Today’s signal, tomorrow’s background

Top quarks as main background
for many new physics channels
Marina Cobal - Londe Les Maure 2004
P 32
Rare SM top decays

Direct measurement of Vts, Vtd via decays tsW, tdW

Decay tbWZ is near threshold
(mt~MW+ MZ+mb) 
BRcut(t bWZ)  610-7
(cut on m(ee) is 0.8 MW)

Decay tcWW suppressed by GIM
2
2
factor mb MW
BR(t cWW) ~ 110-13

If Higgs boson is light: tbWH

FCNC decays: tcg, tc, tcZ (BR: 510-11 , 510-13 , 1.310-13 )

Semi-exclusive t-decays tbM
(final state 1 hadron recoiling against a jet:
BR(t b)  410-8, BR(t bDs)  210-7)
Marina Cobal - Londe Les Maure 2004
P 33
Top mass from di-leptons

Use the events where both W’s decay leptonically (Br~5%)



Much cleaner environment
Less information available due to two neutrino’s
Sophisticated procedure for fitting the whole event, i.e. all kinematical
info taken into account (cf D0/CDF)

Compute mean probability as function of top mass hypothesis

Maximal probability corresponds to top mass
Mean probability
80000 events
(tt) = 20 %
S/B = 10
mass
Selection:
2 isolated opposite sign
leptons
Pt>35 and Pt>25
GeV
2 b-tagged jets
Source of
uncertainty
Di-lepton
Mtop (GeV)
statistics
0.3
b-jet scale
0.6
b-quark fragm
0.7
ISR
0.4
FSR
0.6
pdf
1.2
Total
1.7
Marina Cobal - Londe Les Maure 2004
P 34
Top mass from hadronic decay


Use events where both W’s decay hadronically (Br~45%)
Difficult ‘jet’ environment



Perform kinematic fit on whole event


(QCD, Pt>100) ~ 1.73 mb
(signal) ~ 370 pb
b-jet to W assignment for combination
that minimize top mass difference
Selection
6 jets (R=0.4), Pt>40 GeV
2 b-tagged jets
Note: Event shape
variables like HT, A, S, C,
etc not effective at LHC
(contrast to Tevatron)
Increase S/B:

Require pT(tops)>200 GeV
3300 events selected:
(tt)
= 0.63 %
(QCD)= 2·10-5 %
S/B
= 18
Source of
uncertainty
Hadronic Mtop
(GeV)
Statistics
0.2
Light jet scale
0.8
b-jet scale
0.7
b-quark fragm
0.3
ISR
0.4
FSR
2.8
Total Marina Cobal - Londe3.0
Les Maure 2004
P 35
High Pt sample

The high pT selected sample deserves independent analysis:



Hemisphere separation (bckgnd reduction, much less combinatorial)
Higher probability for jet overlapping
Use all clusters in a large cone R=[0.8-1.2] around the
reconstructed top- direction


Less prone to QCD, FSR,
calibration
UE can be subtracted
j2
j1
b-jet
t
Mtop
Mtop
Statistics seems OK and syst. under control
R
Marina Cobal - Londe Les Maure 2004
P 36
Jet scale calibration

Calibration demands:

Ultimately jet energy scale calibrated within 1%


Uncertainty on b-jet scale dominates Mtop: light jet scale constrained by mW
At startup jet-energy scale known to lesser precision
±10%
MTop
MTop
Scale light-jet energy
Uncertainty
on light jet scale:
1%
10%
Hadronic
 Mt < 0.7 GeV
 Mt = 3 GeV
Scale b-jet energy
Uncertainty
On b-jet scale:
1%
5%
10%
Hadronic
 Mt = 0.7 GeV
 Mt = 3.5 GeV
 Mt = 7.0 GeV
P 37
Rare decays: topWbZ


Since Mtop~MW+Mb+MZ
With present error mt  5 GeV,
BR varies over a factor  3
B-jet too soft to be efficiently identified
 “semi-inclusive” study for a WZ near
threshold, with Z  l+l- and W ->jj


G. Mahlon hep-ph/9810485
(tWbZ)/(tWb)
Interesting: branching ratio depends
strongly on Mtop
Requiring 3 leptons reduces
the Z+jets background
M(top) (GeV)

Sensitivity to Br(t  WbZ)  10-3 for 1 year at low lumi.

Even at high L can’t reach Sm predictions ( 10-7 - 10-6)
Marina Cobal - Londe Les Maure 2004
P 38
topHq
Signal
ttH

Various approaches studied




tt
Previously: ttbarHq Wb(b-bbar)j(lb)
for m(H) = 115 GeV
Sensitivity to Br(t  Hq) = 4.5 X 10-3
(100 fb-1)
New results for:
t tbarHq WbWW*q Wb(l lj) (lb)
≥ 3 isolated lepton with pT(lep) > 30 GeV
 pTmiss > 45 GeV
 ≥ 2 jets with pT(j) > 30 GeV,
incl. ≥ 1 jet con b-tag
 Kinematical cuts making use of
angular correlations


Signal
ttH
tt
Sensitive to Br(t  Hq) = 2.4 X 10-3
for m(H) = 160 GeV (100 fb-1)
Marina Cobal - Londe Les Maure 2004
P 39
Non-SM Decays of Top


4thfermion family
Constraints on Vtqrelaxed:
Supersymmetry (MSSM)

BR(t  W  b( W  c )) ~ 10 3 at mb  100GeV
Observed bosons and fermions would have superpartners 
2-body decays into squarks and gauginos (t  H+ b )
t  t1 g, t  b1  1 , t  t1  10
(
)
Big impact on 1 loop FCNC
two Higgs doublets






H LEP limit 77.4 GeV (LEP WG 2000)
Decay t  H+ b can compete with t  W+ b
5 states (h0,H0,A0,H+,H-) survive after giving W & Z masses
H couples to heaviest fermions  detection through breakdown of e / m / t
universality in tt production
Marina Cobal - Londe Les Maure 2004