The SLHC prospects at ATLAS and CMS Introduction Physics motivation LHC machine upgrade Experiment upgrades New inner trackers for SLHC Conclusions Ian Dawson, University of Sheffield EPS 2007, Manchester

Introduction ATLAS ATLAS and CMS will exploit physics in TeV regime. First collisions expected in 2008. Rich physics programme -Higgs, SUSY etc. CMS Extend the physics potential of the LHC with a luminosity upgrade around ~2016 Rich physics programme for ATLAS and CMS.

Time scale of an LHC upgrade Radiation damage limit ~700fb-1 design luminosity ultimate L at end of year time to halve error integrated L Jim Strait, 2003 Hypothetical lumi scenario Life expectancy of LHC inner-triplet (focusing) magnets estimated to be <10 years due to high radiation doses Statistical error halving time exceeds 5 years by 2011-2012. It is reasonable to plan a machine luminosity upgrade based on new inner-triplet focusing magnets around ~2014-2015

Physics motivation The physics potential of a luminosity upgrade will be better known with LHC data. In general: See: Eur.Phys.J.C39(2005)293 Extend the mass reach by 0.5TeV->1Tev (increased statistics of high-x parton interactions) Extra gauge bosons, heavy Higgs-bosons, resonances in extra-dimension models, SuperSymmetry particles (if relatively heavy). For example, extra gauge bosons (W,Z) like appear in various extensions of the SM symmetry group. Consequences of a ten-fold increase in statistics for the experiments can be divided into 3 area: Extending the mass reach is a consequence of hadron colliders - the probability of very high-x parton interactions is increased by a factor of ten. As an example, shown here is a ~1TeV incease in mass reach in the search for additional gauge bosons. Extra gauge bosons appear in many extensions of SM - usually with couplings similar as for SM Z. Example of improved mass reach - taking into account CMS acceptance and e/ reconstruction efficiency and pile-up noise.

Increased precision in Standard Model Physics Higgs couplings Strongly coupled vector boson scattering (if no Higgs) Triple and quartic gauge couplings Rare top decays through FCNC QGC final state statistics with 6000fb-1. Physics beyond SM (if relatively light and discovered at LHC). For example -SUSY Requires 5 years of SLHC Dark Matter? Extend particle spectrum Improve precision and access rare decay channels. 2) Improving precision measurements - for example, tripke and quartic couplings statistically limited at LHC. 3) Basically, if physics BSM then with more data can make more precise and extensive measurements - especially if the physics is on the heavy side. For example, improving SUSY search for heavy Higgs. Measure coupling of neutralino to Higgs. Determine its higgsino component.

LHC machine upgrade Big challenge to increase luminosity by factor of ten. Machine R&D has to start now. Two upgrade scenarios currently being considered - both are based on first replacing the inner triplet magnets to achieve ultimate LHC luminosity of ~2x1034cm-2s-1 parameter 25ns, small * 50ns, long protons per bunch 1.7 4.9 bunch spacing 25 50 beam current 0.86 1.22 longitudinal profile Gauss Flat rms bunch length 7.55 11.8 beta* at IP1&5 0.08 0.25 full crossing angle 381 peak luminosity 15.5 10.7 peak events per crossing 294 403 initial lumi lifetime 2.2 4.5 effective luminosity (Tturnaround=5h) 3.6 3.5 Improve beam focusing Machine magnets inside experiments 25ns or 50ns bunch crossing Increase beam currents More demanding on machine 50ns bunch crossing Reducing the bunch spacing is not possible due to electron cloud heating.

25 ns spacing Peak luminosities different for 25ns versus 50ns scenarios, but average luminosity similar. 50 ns spacing average luminosity The 25ns scenario will require magnets inside the experiments! Peak luminosities may be too difficult for experiments to cope with - possibility of ”luminosity levelling” being explored. For experiments, what’s important is not peak luminosity but average luminosity - which is similar for 50ns and 25ns scenarios. From experiments point of view - not clear which is preferred solution. 25ns greater peak lumi variance and magnets interfering in experiements, but 50ns has more avergae pile-up. The large peak luminosity variations and associated pile-up will be difficult for experiments to cope with - hence the motivation for luminosity levelling. Various methods for doing this being investigated - blah

Experiment upgrades From the LHC to the SLHC 1033 1035 1032 cm-2 s-1 1034 I. Osborne The ATLAs and CMS experiemtents will need to upgrade to work in higher radiation backgrounds. The inner trackers will need to be replaced due to - 1) Radiation damage, 2) Big increases in occupancies. Both ATLAS and CMS in process of determining upgrade requirements - but the inner trackers need replacing and likely to be most of cost.

ATLAS Muon chambers Forward calorimeters In general, most of calorimetry should be OK FCAL (|η|>3.1) particularly subject to beam radiation Simulations show possible heating of Lar Improve cooling? Or new “warm” FCAL? Background rates very uncertain (factor ~5x) Need LHC experience Effect of slim quads? Use existing chambers as far as possible Beryllium beampipe will reduce by factor of 2.

CMS Calorimeters Level 1 trigger Muon system Electromagnetic calorimeters should be OK at SLHC. Hadronic calorimeter scintillator may suffer radiation damage for >2 R&D required Impact on HF if machine magnet insertions? There may not be enough rejection power using the muon and calorimeter triggers to handle the higher luminosity conditions at SLHC Level 1 Muon Trigger has no discrimination for pT > 20GeV/c, therefore problem to keep Level 1 at ~100kHz. Adding tracking information at Level 1 gives the ability to adjust PT thresholds Muon system High momentum tracks are straighter so pixels line up Search Window γ In general, expected to be quite robust, but some electronics “less” radiation hard and most likely need replacing. As with ATLAS, need experience with running at LHC. 2 layers about 1mm apart could communicate

New trackers for SLHC Both ATLAS and CMS need new inner trackers Started looking at new layouts. Simulations crucial to ascertain: Track reconstruction performance Occupancies Material effects Silicon trackers at LHC use p-in-n sensors - requires full depletion  high depletion voltages already at LHC. n-in-p or n-in-n can operate under-depleted - promising progress being made on n-in-p. For both ATLAS and CMS the inner trackers will need to be replaced. There are two reasons for this: 1) Radiation damage, 2) Big increases in occupancies. Fluences at larger radii dominated by neutron-albedo, greatest near endcaps. Fluences at small radii dominated by particles from interaction point.

Possible sensor technology for the SLHC tracker Long strips (present p+-n or n+-p) Short strips (n+-p) Pixel b-layer (3D? Diamond? thin-Si? Gas?) n+-p pixels The most challenging will be engineering work (cooling, cabling, shielding, other services)

For innermost layers - new technology required 3D silicon sensors ? Uses MEMS (Micro Electro Mechanical Systems) technology to engineer structures. Both electrode types are processed inside the detector bulk instead of being implanted on the wafer's surface. Electric field between electrodes - shorter signal collection distances compared with planar devices: Lower depletion voltage (and therefore power) Faster and more efficient charge collection. Large scale production? Timescale? Cost ? CVD Diamond ? Gas ? Now well established with several applications (eg beam conditions monitoring) Large band gap and strong atomic bonds give excellent radiation hardness Low leakage current and capacitance = low noise Large band gap means ~2 less signal than Si for same X0 For innermost layers, certainly the b-layer - extremely high radiation - several new technologies being looked at. 1) A novel new technology being developed are 3D silicon detectors. The short signal collection distance (d = 50->70m) ensures efficient charge collection. The signal is proportional to Leff(1-exp(-d/Leff)) so if Leff similar between 3D and planar devices (which is case if both Silicon) then 3D is better. Leff is reduced with increasing irradiation because of introduction of traps so that Leff ~ 1/fluence. Alternative 3D geometries/designs are now being investigated, eg single-type-column, partial-columns 2) Diamond is now well established and found several applications (eg BCM, pixel modules fabricated and tested succesfully by ATLAS) Eg GOSSIP (micromegas) low mass but gas issues such as sparking?

Some initial comparison data Plot on left shows signal charge versus irradiated fluence. For inner most layers we expect 1.6x1016 where 3D looks promising. However, it is the signal/noise which is important and the plot on the right gives an indication of noise - diamond has very low noise and seems competitive. Pixel n-in-n or n-in-p may also play a role should the novel new technologies not be ready on time - or too costly - but thinner wafers need to be used to avoid very high voltages for depletion. Planar silicon n-in-p or n-in-n can still play a role for inner most layers should new technologies not be ready Tried and tested solution but high operating voltages unless thin ?

Conclusions The implications of a major luminosity upgrade (SLHC) sometime around 2015-2016 are being studied by the LHC machine and experiments. Long lead times require this work starts now. But the main priority is physics exploitation of the LHC. The baseline bunch crossing rate at the SLHC is 20MHz, with 40MHz as a backup. The 80MHz option is no longer considered due to beam heating. ATLAS and CMS need new inner trackers CMS will need to revise Level 1 trigger strategy. ATLAS forward calorimeter unlikely to survive at SLHC A better understanding of upgrade requirements for ATLAS and CMS will be gained from running at LHC.

Backup slides For both ATLAS and CMS the inner trackers will need to be replaced. There are two reasons for this: 1) Radiation damage, 2) Big increases in occupancies.

The European strategy for particle physics “The LHC will be the energy frontier machine for the foreseeable future, maintaining European leadership in the field; the highest priority is to fully exploit the physics potential of the LHC, resources for completion of the initial programme have to be secured such that machine and experiments can operate optimally at their design performance. A subsequent major luminosity upgrade (SLHC), motivated by physics results and operation experience, will be enabled by focussed R&D; to this end, R&D for machine and detectors has to be vigorously pursued now and centrally organized towards a luminosity upgrade by around 2015.” (http://council-strategygroup.web.cern.ch/council-strategygroup/)

Luminosity levelling etc. 25 ns 50 ns average luminosity The relatively fast luminosity decay and high multiplicity call for Luminosity Leveling. …but the issue is how to do it efficiently: dynamic beta*: uses existing hardware; probably complex due to large number of side-effects in IR’s AND arcs. dynamic bunch length: needs new RF; possible side effects in whole machine related to modification of peak current. dynamic crossing angle: using the early separation hardware, no side effects identified. Even better: crabbing.

Simulations for the inner tracker 1 MeV fluences obtained by convolving particle spectra predicted by FLUKA2006 with “displacement-damage” curves (RD48). 1 MeV equivalent neutron fluences assuming an integrated luminosity of 3000fb-1 and 5cm of moderator lining the calorimeters. av18 is current baseline geometry for upgrade studies 5cm neutron moderator lining all the calorimeters no tracker material The beneficial moderating effect of the TRT is lost in an all silicon system. Can be recovered with 5cm of moderator lining barrel (as shown in Genova).

Parameterisation of 1MeV-neq fluences Several requests to parameterise inner tracker backgrounds. Z(cm) a1 a2 a3 a4 1.4x1017 3.7x1015 1.7x1014 -1.0x1012 150 7.0x1016 9.5x1015 9.7x1013 -5.7x1011 300 4.9x1016 1.2x1016 3.0x1014 -2.0x1012 Parameterise also pion and neutron contributions separately. Fluences at small radii dominated by particles from interaction point. Mention no safety factors. Future investigations = moderator, material in front of FCAL Fluences at larger radii dominated by neutron-albedo, greatest near endcaps. Use these types of plots for future investigations (Eg moderator design, impact of extra material etc.)

Uncertainties/Safety-factors? Until we have better benchmarking data, the uncertainties assumed at the LHC still apply. Event generators ~20% Transport codes ~20% Displacement damage cross sections ~50% At the SLHC, the goal seems to be 3000fb-1 for the integrated physics luminosity. What safety factors do we assume? At present, 2 × 3000fb-1 being used Also, until we know what an upgraded detector and machine look like, there may be additional contributions to inner tracker fluences … Concerning safety factors at LHC … if designed to deliver 300fb-1, then by designing tracker to ~700fb-1 and applying a further “simulation” safety factor of 1.5 to give ~1000fb-1, then we have effectively a total safety factor greater than 3.

Impact of additional mass in front of FCAL on Inner Tracker fluences Can ‘alcove’ region in front of FCAL be used for machine magnets? Look at 1MeV-neq fluences in tracker volume. (Saying nothing about FCAL). Fill ‘alcove’ with iron. Should be pessimistic? If so, results below give worse case? av18 5cm poly Z=300cm av18 with Fe mass ~2.7 ~2.0 Factor ~2.7 increase due mainly to increase in neutron albedo.