Electroweak Symmetry Breaking Studies at an

1
Electroweak Symmetry Breaking Studies at an
ee- Linear Collider David J.
Miller University College London

• Lecture 1, 20 August 2001
• Why so long a wait?
• Luminosity, and its problems.
• Polarised beams
• Extra machine options
• The detector
• Event simulation
• Profiling a light Higgs boson

• Lecture 2, 21 August 2001
• SUSY
• No Higgs scenarios
• Anomalous W couplings
• Other high-scale physics
• Top mass
• Precision physics with GigaZ
• The Labyrinth
• What next?

• Almost everything in these lectures can be found
  in
• The TDR TESLA Technical Design Report.
  DESY 2001-011(etc.) or http//tesla.desy.de/tdr/
• The RB Resource Book for Snowmass 2001.
  SLAC-R-570(etc.). (Both on
  CD-rom).

2
Why so long coming? Because it needs 10 4 times
SLC Luminosity
Cross sections for constituent in pp
scattering very similar to ee-. In mid 80s -
people said We know how to build a high L pp
machine, but not how to build its detector. For
ee- its the other way round. Hence LHC first,
but now were ready.
LEP gave 1fb-1/expt. in 11 years, with
107 Z0 at LEP1 At ?s 500 GeV we need 500
fb-1 to get 30,000 Zh120
50,000 h120?? 106 WW- At ?s
1,000 GeV we need 1,000 fb-1 to get
6,000 HA (300GeV) 3,000 h500 ??
2,000 WW?? (if no higgs) Smaller samples
not worthwhile. ( Need lots of luminosity to
scan multiple thresholds, vary polarisation, go
to ??, e-?, e-e-)
3
Mental arithmetic A year lasts 107 seconds. 1
barn 10-24 cm2 1015fb. Luminosity L is
measured in cm-2s-1. Size of a run L t
events/cm2 L t 10-39 events/fb ? fb-1. 2
years at 2.5?1034 gives 500 fb-1
A decade of RD, especially at SLAC, also DESY,
KEK etc. has given
.. new problems, including
Wakefields, Emittance growth,
Disruption, Beamstrahlung, Pair production
?100 improvement. More bunches with more
charge.
Accelerator problems
frep bunches/second Ne electrons/bunch
?x rms bunch width ?y rms bunch height

1/100 reduction in ?y to 5 nm. Lower
emittance, demagnify more
Experimenters problems
4
Experimenters Problems Disruption,
Beamstrahlung, Pair production
1010 electrons/bunch, with ? 106
dimensions nanometers

• Disruption
• ee- beams focus each other inwards (L
  enhanced!), then fly apart after collision
• e-e- beams defocus immediately (L reduced).

e-
e-
e-
e-

• Beamstrahlung
• synchrotron radiation in the
• field of the opposing bunch
• gives a smeared spectrum.

• Pair Production
• incoming beam particles scatter from the
  beamstrahlung photons

5
Coping with Beam-Beam Problems

• Disruption and Beamstrahlung
• E and B minimised for a given
• luminosity by making flat beams.
• e.g. ?x 500 nm ?y 5 nm
• Exiting beams must not hit
• quadrupole aperture.
• Forward absorbers and calorimeters
• catch Beamsg photons and pair
• electrons.
• Instrumented mask and W shield stop
• backscatter into main detector

Pairs
B field of the experiment keeps pair electron
trajectories (Rmax plotted above) out of the
vertex detectors
6
Radiation Levels Only serious inside mask,
especially at LCAL.
Electromagnetic energy deposited per crossing ?
Radiation level per year for each plane of a
Silicon Tungsten LCAL.
Also expect 109 neutrons/cm2/year at
microvertex.Probably manageable.
7
Hit rates in detectors are manageable
Pairs into Microvertex, per crossing.
Gammas with Egt10 keV entering TPC. Total per beam
crossing 1325 _at_ 500 GeV, with 3.1 GeV total
energy.
Each layer has 10s of Megapixels!
ECALs have 200 pairs with more than 3 MeV per
crossing.
8
Other new Problems 1. Interaction Energy
Spectrum , can it be measured?
Monitored for any experimental running period
with real events by using acollinearity ?acol of
ee- Bhabha scatters in the region just outside
the mask (my sole contribution, apart from
chairing committees). Precison 1 in 103 easy,
for top or SUSY thresholds, with forward tracking
pixels (see TDR). (1 in 105 for GigaZ
harder, but possible)

• Intrinsic momentum spread
• in linac 0.2 ( gtgt LEP)
• can trade off with L.
• Beamstrahlung spread
• Absolute energy of beams.

?acol
Ebeam
Ebeam-Ebrem

• At NLC, with finite beam-crossing angle,
• may use spectrometer on outgoing beam.
• At TESLA, with zero crossing angle,
• will have to develop in-beam spectrometer
• - tried for LEP but needs more RD

9
Rates and Backgrounds
Ron Ruth will discuss the differences for NLC
Much gentler than LHC! Record everything and sort
out offline essentially triggerless. NLC
similar.
Hard virtual photons from beam particles make
high energy ?? collisions. Cross-section
dominated by hadronic photon-photon with vector
meson dominance like minimum bias events at LHC
(but 0.02/BX, c.f. 20/BX_at_LHC). Used HERA
studies to calibrate models for this rate
calculation. Contaminates some signals, e.g. H??,
but good z-resolution often separates primary
from ?? background
10
Polarised beams.
e-
SLC has proved we can polarise e- to 80 hope
for more in 10 years. (Laser onto strained GaAs)
Much harder. Have to make polarised gammas, then
pair produce ee-. Needs e- with gt160 GeV, so
use incoming beam.
e
TESLA e source. If angular acceptances are
restricted, should be possible to get 50 e
polarisation maybe more with reduced L.

• Uses of polarisation
• Turn off SM background processes
• (anything that couples to W?).
• Measure polarisation-dependence of
• signal, e.g. asymmetries like ALR.
• Can be sensitive to CP.

MUST EITHER build very good polarimeters
OR
With random ?P- and ? P can measure both in
Blondel scheme R-B p340
11
Additional Collider Options 1. e-e- is easiest
- add a second e- gun at the e end ( N.B.
permanent magnets!) specially useful if
SUSY sector complicated. Needed for ?? and ?e.
2. ?? and ?e the Compton-Collider use Compton
backscattering of near-visible laser light,
within a few mm of collision point (Novosibirsk
idea Livermore designing laser optics). ?10
on cost 2nd I.R.
Laser
?e collision point (or ?? with 2nd laser)
?
e-
e-
3. GigaZ Include by-pass in linacs to get good
luminosity at MZ and at 2MW (maybe _at_ 2nd I.R
parasitic running). Positron polarisation very
important. By-pass uses 1/4 of linac for main
beams, so can use the rest of e- linac for
positron source. Hope for L 5 1033
12
The Experiment has to do better than ADLOS -
physics demands it, especially for EWSB.
Lessons from LEP and SLD (personal and prejudiced
view, arbitrary order)
OPAL pro - large diameter JET tracker with
best dE/dx, good Pb-glass ECAL,
good Si-W L-cal (0.04_at_Z0), (smallest team at
LEP ? most papers) con - ECAL outside
coil, low B 0.4T, ?vertex cramped, poor z
tracking. ALEPH pro - big TPC tracker with
dE/dx and good dpt/pt2, finegrained ECAL inside
coil, good Si-W L-cal, high B
1.5T, good HCAL uniformity. con -
poor E resolution in gasPb ECAL DELPHI pro -
best LEP Si-strip ?vertex, forward pixels
con - RICH didnt add much - took lots of
space gasPb ECAL aged, TPC
tracker too small. L3 pro - best ECAL E
resolution- BGO inside coil. con -
TEC tracker too small, ? spectrometer
o.t.t. SLD pro - by far best ?vertex (CCDs,
much closer), polarised e- , first Si-W L-cal
con - much lower Luminosity than LEP.
Glossary dE/dx gives particle identification,
as should RICH and CRID (Ring-Imaging Cerenkov
counters) BGO is Bismuth
Germanate, a scintillating crystal. Pb-glass
blocks give Cerenkov light.
JET and TEC are 2-D drift-chambers. TPC is 3-D
Time Projection Chamber. CCD
is Charge Coupled Device, a serial-readout Si
pixel detector, 2-D. Si-strip is (11)-D.
13
1. ?vertext detector for heavy flavour
identification Beauty easiest more energy
release ? bigger kinks c and ? much harder.
Flavour couplings of Higgs are basic, especially
to test SUSY. Many scenarios have different
2/3, -1/3 couplings b versus c best
hope. Improve c-tagging by going much closer
inner pixel layer at 2cm (also reduces c
background in higgs?gluons).
c(b bkgr) is for recognised higgs decays
where we are trying to dig out the c from b.
5 layers of CCDs would give this performance.
Low scattering (only 200 ?m/layer), but slow
readout, accumulate many beam-crossings. Other
technologies in RD smart pixels, CMOS.
14
2. Overall tracking performance Higgs is the
dictator
Whatever its decays, if coupled to Z0 (and light
enough), will see its recoil against Z? ee- or
? ?-. Need momentum resolution!
Alternatives
TESLA TDR (cf US Large) Silicon tracking
inside a big TPC B 3 or 4 T good dE/dx
US Small All silicon B 5 Tesla
Japanese JLC tracker. Big JET chamber 2.3 m
radius good dE/dx, B 2 T
No dE/dx.

15
Tracking Ideas
Gaseous Electron Multipliers for 2-D detector at
endplate of TPC
With chevron pads for precision
FTD pixel discs in forward direction track
electrons for Bhabha acollinearity physics
processes. Low density endplate for ?vertex
essential!
16
3. Jets and Energy-Flow Must separate jets and
measure their energies needs tracking
and calorimetry, especially good segmentation to
separate clusters and match extrapolated tracks
correctly to some of them.
E.g. a) Can we separate W?2jets
from Z ?2jets? masses close 80.4 or 91.2
GeV Especially important for V onV scattering
-where EWSB has to appear if nowhere else!
E.g. b) Can we recognise and measure all
the interesting features in
Competing ECAL technologies Si-W gives great
performance but expensive (1/2 cost of this
TESLA TDR detector)
17
Energy Flow Performance Reconstructed dijet
masses in ee-???WW or ??ZZ for assumed jet
energy resolutions ?E/?E a) 30 b) 60
18
Simulation inputs and tools. needed now for
feasibility studies must be even better for
the analyses.

• Theoretical inputs (the LoopVerein)
• Precision calculations of Electroweak, QCD and
  SUSY processes,
• especially higher order loops - at scales up to
  and beyond the E-W scale.
• Vital for signal and background processes -
  c.f. Janots lectures.
• Monte Carlo Generators
• Never good enough! Many do not treat
  polarisation.
• Do not trust feasibility studies done at
  generator level.
• Full detector simulation
• Mostly still FORTRAN based, but OO versions
  under development.
• Needed as basis for proper feasibility
  studies.
• Fast Simulation
• Take parametrised detector performances from
  full simulation to
• run large samples for feasibility studies.
• Analysis and Display
• New OO tools coming on the market.

19
The SM Higgs
Error bars are from feasibility studies, bands
from theory.
bb - WW crossover explains dip in Tevatron
sensitivities.
20
Light Higgs Profile at a Linear Collider
SM (or decoupled SUSY)
(LHC gets mass to 120
MeV, ghWW to 15)
Relative accuracy on branching ratios for 500fb-1
at ?s 500 GeV
use tagging capacity of ?vertex detector and
good tracking to separate decays in
Higgsstrahlung
Total width from ghWW and , to
5 for Mh120 GeV.