Large Hadron electron Collider - LHeC Frank Zimmermann UPHUK-4, Bodrum, 31 August Uluslararası Katılımlı Parçacık Hızlandırıcıları ve Uygulamaları Kongresi (UPHUK4) 30 Ağustos-1 Eylül 2010

outline physics motivation target parameters Linac-Ring & Ring-Ring options e + source options p beam parameters IR design e- beam parameters, e- RLA/ERL layouts LHeC-CLIC high-energy ERL option tentative schedule, conclusions

LHeC – Large Hadron electron Collider motivation: rich physics program: e-q physics at TeV energies  precision QCD & electroweak physics  boosting precision and range of LHC physics results  beyond the Standard Model  high density matter: low x and eA, & also  A collisions Tevatron/LEP/HERA (Fermiscale)  LHC/LC/LHeC (Terascale) 100 fold increase in luminosity, in Q 2 and 1/x w.r.t. HERA status: CERN-ECFA-NuPECC workshops (2008, 2009, 2010: October) Conceptual Design Report in print by spring 2011

distance scales resolved in lepton- hadron scattering experiments since 1950s, and some of the new physics revealed energies and luminosities of existing and proposed future lepton-proton scattering facilities e- energy ~ GeV luminosity ~10 33 cm -2 s -1 LHeC - more physics motivation >5x HERA c.m. energy >>10x HERA luminosity Max Klein & Paul Newman, CERN Courier April 2009 Max Klein & Paul Newman, CERN Courier April 2009

kinematic plane in Bjorken-x and resolving power Q 2, showing the coverage of fixed target experiments, HERA and LHeC >> 10x particle physicists request both e - p &e + p collisions; lepton polarization is also “very much desired” Max Klein & Paul Newman, CERN Courier April 2009

LHeC – several options RR LHeC: new ring in LHC tunnel, with bypasses around experiments RR LHeC e-/e+ injector 10 GeV, 10 min. filling time LR LHeC: recirculating linac with energy recovery, or straight linac

this talk will focus on “linac” options  least interference with LHC infrastructure  compatible with LHC energy upgrade  upgrade potential to higher energy or to LC  lots of synergies with CLIC, ILC and SPL however LR LHeC is not the baseline or not yet

LHeC-RR & LHC interferences „About 350 places to be avoided for e-magnets. Some areas very difficult. Solutions will be looked for, once a solution for the other ~350 obstacles has been found. Note: The experiments 1 and 5 are passed on the outside. That increases the overall length. [beware of Hirata-Keil beam-beam resonances!] We have to cross the transport region to compensate the overlength. This is almost impossible, because it would make any accelerator repair even more difficult.“ K.-H. Mess, February ‘10

→ dense net of resonances in tune space if circumferences are slightly different → spontaneous beam separation Hirata-Keil resonances: beam-beam resonances for circumference ratio K + /K - ≠1 K. Hirata and E. Keil, 1990; compare also overlap knockout resonances at CERN ISR (S. Myers)

Large Hadron electron Collider - Linac-Ring options J. Osborne

TOBB ETU KEK L-R LHeC – collaborating institutes

LHeC LR design contributors S. Bettoni, C. Bracco, O. Brüning, H. Burkhardt, E. Ciapala, B. Goddard, F. Haug, B. Holzer, B. Jeanneret, J. Jowett, J. Osborne, L. Rinolfi, S. Russenschuck, D. Schulte, H. Thiesen, R. Tomas, F. Zimmermann, CERN, Switzerland; C. Adolphsen, M. Sullivan, Y.-P. Sun, SLAC, USA; A.K. Ciftci, R. Ciftci, K. Zengin, Ankara U.,Turkey; H. Aksakal, Nigde U.,T.; E. Eroglu, I. Tapan, Uludag U., Turkey; T. Omori, J. Urakawa, KEK, Japan ; S. Sultansoy, TOBB, Turkey; J. Dainton, M. Klein, Liverpool U., UK; A.Variola, LAL, France; R. Appleby, S. Chattopadhyay, M. Korostelev, Cockcroft Inst., UK; A. Polini, INFN Bologna, Italy; E. Paoloni, INFN Pisa, Italy; P. Kostka, U. Schneekloth, DESY, Germany; R. Calaga, V. Litvinenko, V. Yakimenko, BNL, USA; A. Eide, NTNU, Norway ; A. Bogacz, JLAB, USA more are welcome!

performance targets e- energy ≥ 60 GeV luminosity ~10 33 cm -2 s -1 total electrical power for e-: 100 MW e + p collisions with similar luminosity simultaneous with LHC pp physics e - /e + polarization detector acceptance down to 1 o getting all this at the same time is very challenging

SLCCLIC (3 TeV) ILC (RDR) LHeC Energy1.19 GeV2.86 GeV5 GeV60 GeV e + / bunch at IP40 x x x x10 9 e + / bunch before DR inj.50 x x x 10 9 N/A Bunches / macropulse N/A Macropulse repet. rate120505CW Bunches / second x10 6 e + / second0.06 x x x x Flux of e + X 18 X 65 X 6666 quite a challenge! L. Rinolfi, May 2010

e + source options Compton ring Compton ERL Compton linac undulator based source e - beam on liquid metal-jet target e + recycling

Compton Ring and Compton ERL (ILC) T. Omori, L. Rinolfi, J. Urakawa et al

Compton Linac for e + source 4 GeV N  / N e- = 1 (demonstrated at BNL) N e+ / N  = 0.02 (expected) i.e.  50 gammas to generate 1 e pulses ~5 ns With 5 nC / e - bunch and 10 Compton IP's => 1 nC / e + bunch Data for CLIC: N e+ = 6.4 x 10 9 / bunch ~ 1 nC N e- = 0.32 x / bunch ~ 50 nC For LHeC, one would need to increase the number of targets/capture sections working in parallel in order to reach the requested intensity. => Cost and reliability issues T. Omori, L. Rinolfi, J. Urakawa et al

LH C e - Linac spent e - e + pre- Linac IP 1 e+e+ p+p+ Target LHeC e+ source based on undulator Undulato r  e-e- e + Linac L. Rinolfi et al

- CERN “Merit” (liquid Hq droplet target): possibility to handle MW beam. - Tokyo University S-band gun generating 3nC every 2.8ns for 100 bunches and plan to go up to 1000 bunches LHeC proposal: send ~1 MW average power e- beam on liquid target ~100 Hz, ~1  s trains, 1 A average over macro pulse, 100  A average current; e+/e- yield > 1 with small transverse emittance total wall plug power for e+ source ~3-5MW cost estimate ) ~ 50M$ concern: stability of “droplet” during macropulse high-power e - beam on liquid-metal target V. Yakimenko, June MW e- beam liquid metal target e+

how do we get cm -2 s -1 ? luminosity of LR collider: take highest proton beam brightness permitted by LHeC management (“ultimate” LHC values) assume smallest conceivable proton  * function: - reduced l* (23 m → 10 m) - squeeze only one p beam - new magnet technology Nb 3 Sn maximize geometric overlap factor - head-on collision - small e- emittance (round beams) average e- current !

proton beam brightness N b,p /  p management decision proton bunch spacing (which multiple of 25 ns) is irrelevant for LHeC!

proton round-beam IR optics for ultra-low  * R. Tomas l*=10 m extended e- final-focus optics by Kahraman Zengin, Ankara U.

S. Fartoukh  arc increased by a factor of 2 or 8 in s45/56/81/12 depending on the  * aspect ratio in IP1 and IP5 squeezed optics:  * x/y = 7.5/30 cm (flat beam) alternated in IR1&5 new flat p-p HL-LHC optics w low  * & l*=23 m this flat-beam optics is alternative for LHeC

geometric reduction factor geometric loss factor H hg vs. crossing angle LHeC working point C. Adolphsen H. Braun F. Zimmermann  p *=10 cm (round beams)

interaction region (2008) R. Tomas, F.Z. small e- emittance → relaxed  e * → L e * > L p *, can&must profit from ↓  p * ; single pass & low e-divergence → parasitic collisions of little concern; → head-on e-p collision realized by long dipoles or return loop

IR layout w. head-on collision beam envelopes of 10  (electrons) [solid blue] or 11  (protons) [solid green], the same envelopes with an additional constant margin of 10 mm [dashed], the synchrotron- radiation fan [orange], and the approximate location of the magnet coil between incoming protons and outgoing electron beam [black] detector integrated dipole: field ~0.45 T critical photon energy ~ 1 MeV average SR power = 87 kW 8x10 10  / bunch passage is the SR acceptable for the detector?

prize questions can we shield the SC proton quadrupole from MeV synchrotron radiation? - “this SR is very difficult to shield” K.-H. Mess, LHeC Divonne studies by Husnu Aksakal (2009) and Emre Eroglu (2010) can we reduce back-scattering into the detector to an acceptable level? - studies by Kenan and Rena Ciftci (2010)

SR shielding - FLUKA simulation beam energy [GeV] dipole field [T] offset at LHC triplet [cm] distance IP& p-triplet [m]10 lead shielding in front of triplet vacuum dipole example FLUKA results [GeV/cm 3 ] SR code (linked to FLUKA) calculates #SR photos per m, per energy bin Husnu Aksakal, Nigde U., cm shielding is enough

LEAD (Pb) Energy deposition in lead target. Max. value is GeV/cm 3 /  For 25 cm of lead no SR photon penetrates through the target. Emre Eroglu, Uludag U., 2010 FLUKA simulation again 30 cm shielding is enough

back scattering of e - and e + from lead target Max. energy is about Gev/cm 3 /  for electron (up), and about GeV/cm 3 /  for positron (down). Emre Eroglu, Uludag U., 2010 FLUKA simulations e-e- e+e+

Back scattering of  ’s from lead target. Max. energy is about GeV/cm 3 /  for photon LEAD (Pb) Emre Eroglu, Uludag U., 2010 

reflection of  ’s by mirrors Surface of the magnet-coil protection shield should be sloped to reduce power density. Acceptable: ~10-20 W/mm 2. Proposal: introduce mirrors with a shallow grazing angle to reflect part of the photons into electron beam channel or special SR extraction holes Kenan & Rena Ciftci, Ankara U., 2010

“supercon” type magnet p e- d~8.7 cm S. Russenschuck gradient ~250 T/m Nb-Ti ~310 T/m Nb 3 Sn

“mirror” half quadrupole magnet p e- d~6.3 cm S. Russenschuck gradient ~145 T/m Nb-Ti ~175 T/m Nb 3 Sn

detector integrated dipole SC solenoid coil SC coil split and tilted B B Stephan Russenschuck, Simona Bettoni, Eugenio Paoloni IP

how about 2nd LHC proton beam? 2nd beam must be transported across LHeC IR two possibilities: (1) common IR vacuum chamber; second beam unsqueezed (2) detector with a bypass hole (c.f. Tevatron D0)

we did what we could on p side how about e-? e- emittances and  * not critical (protons are big, ~7  m!) most important parameter: average beam current in addition: bunch structure and polarization

target luminosity we need about 6 mA CLIC design current ~ 0.01 mA (factor 600 missing) lowering voltage, raise bunch charge & rep rate → 0.06 mA (NIMA 2007) CLIC drive beam ILC design current ~ 0.05 mA (factor ~100 missing)

SC linacs can provide more average current at lower energy (& lower gradient) e.g. by increasing the duty factor times, or even running cw example design average currents: CERN HP-SPL: ~2.5 mA (50 Hz) Cornell ERL ~100 mA (cw) eRHIC ERL ~ 50 mA at 20 GeV (cw) LHeC ERL needs 6 mA at 60 GeV

LHeC ERL site power main contributions: RF power dissipation + RF power to compensate ERL inefficiency Cryo power for SC linac (RF gradient) SR energy losses in return arcs (radius) g: gradient E: final energy D: duty factor A, B and C: coefficients  : bending radius

one more ingredient! choice of SC linac RF frequency: 1.3 GHz (ILC)? ~700 MHz! requires less cryo-power (~2 times less from BCS theory); true difference ↔ residual resistance, [J. Tückmantel, E. Ciapala] synergy with SPL, eRHIC and ESS

1.67 km 0.34 km 30-GeV linac LHC p 1.0 km 2.0 km 10-GeV linac p-60erl LHC p 70-GeV linac 3.9 km 2.0 km p-140 injector dump injectordump injector IP 140-GeV linac p-140’ injector dump IP 7.8 km LHeC –Linac-Ring configurations high luminosity high energy “least expensive"

LHeC – general parameters e- beamRRLR ERLLR “p-140” e- energy at IP[GeV] luminosity [10 32 cm -2 s -1 ] polarization [%] bunch population [10 9 ] e- bunch length [  m] bunch interval [ns]2550 transv. emit.  x,y [mm] 0.58, rms IP beam size  x,y [  m] 30, 1677 e- IP beta funct.  * x,y [m] 0.18, full crossing angle [mrad] geometric reduction H hg repetition rate [Hz]N/A 10 beam pulse length [ms]N/A 5 ER efficiencyN/A94%N/A average current [mA] tot. wall plug power[MW]100 p- beamRRLR bunch pop. [10 11 ]1.7 tr.emit.  x,y [  m] 3.75 spot size  x,y [  m] 30, 167  * x,y [m] 1.8, $ bunch spacing [ns]25 $ smaller LR p-  * value than for nominal LHC (0.55 m): - reduced l* (23 → 10 m) - only one p beam squeezed - new IR quads as for HL-LHC B. Holzer, M. Klein, F. Zimmermann

example (old) RLA optics for 4-pass ERL option w. 500 MeV injection energy (=dump energy) 0.5 → → 60 → → 0.5 GeV LHeC question: feeding groups of linac quadrupoles by one power converter? or no quads?! Vladimir Litivinenko Hugues Thiesen new optics under development by Alex Bogacz (JLAB) Anders Eide

W. Meng et al PAC’09 common vacuum chamber for eRHIC return loop with several layers of small-gap magnets, profiting from small e - emittance LHeC questions: same power converter feeding magnets for several loops with different number of coils or separate power converter for each loop? strength of dipoles and quad’s tapered to local beam energy (SR)? details of return loop design Hugues Thiesen

Relative emittance for three final beam energies (60 GeV, 100 GeV and 140 GeV); here test initial emittance is 2 μm. For E = 140 GeV, the absolute normalized emittance growth is roughly 50 μm, which was confirmed by other simulation studies with higher start emittance & by analytical estimates SR & chromatic emittance growth Y.-P. Sun  =100  m at 140 GeV and 50  m at 60 GeV are conservative values SR emittance growth: few % at 60 GeV, & 100% at 140 GeV SR emittance growth OK!

ERL collective effects multi-beam wakes & beam break up modified PLACET code (D. Schulte) ion accumulation and ion-driven instabilities needs gaps and/or clearing electrodes electron-cloud instability for e+ beam simulations planned (G. Rumolo)

beam dump without energy recovery we will have to dump a ~50 MW e- beam = 2-3 x maximum ILC beam power ! CERN dump experts: feasible ! (Brennan Goddard, Chiara Bracco)

future work updated/new optics for ERL study of ERL collective effects detector-integrated dipole design flat-beam IR optics conclusion on IR synchrotron radiation e ± -A/  -p/  -A options, esp. IR layouts & laser e+ source baseline solution …

Higher performance w CLIC? luminosity of LR collider: raising p bunch population (round beams) allows reducing average e- current

higher luminosities with CLIC? extreme possibility: dramatically increase N b,p by putting all protons in single bunch of length ≤ CLIC train length: “proton superbunch” (D. Schulte, F. Zimmermann, CLIC Notes 589 & 608) collision region then extends over ≤ 23 m this would bring luminosity factor ~3000! † unfortunately ATLAS and CMS rejected superbunches in a strongly worded joint official statement at CARE HHH-2004

highest-energy LHeC ERL option High luminosity LHeC with nearly 100% energy efficient ERL. The main high-energy e- beam propagates from left to right. In the 1 st linac it gains ~150 GeV (N=15), collides with the hadron beam and is then decelerated in the second linac. Such ERL could push LHeC luminosity to cm -2 s -1 level. V. Litvinenko, 2 nd LHeC workshop Divonne 2009 this looks a lot like CLIC 2-beam technology high energy e- beam is not bent; could be converted into LC?

LHeC-LR & CLIC synergies original hope was to later extend linac & convert into LC (staged LC construction, with “cheap” starting point) however, this may not be possible as the CLIC or ILC tunnels are positioned & oriented differently (J. Osborne) many LHeC LR subsystems and technologies are similar to those of ILC or CLIC linac, RF power source, e- injector, e+ source, final focus, highest-energy option (!), but most requirements are more relaxed (emittance, spot size); gain experience! equipment could be reused for CLIC or ILC many of the same people work on LC and LR LHeC

LHeC – possible schedule : installation of (ring or linac) LHeC, during HL-LHC upgrade shutdown : ~10 years of operation with LHC [p/A] colliding with E e ≈ 60 GeV [e - /e + ]: ~100 fb -1 after 2030: possible extension to high E e LHeC, during HE-LHC upgrade shutdown and long term operation with 16.5 TeV p colliding with e.g. E e = 140 GeV [e - /e + ]

conclusions ERL with 2x10 GeV SC linac proposed for high- luminosity 60-GeV Linac-Ring LHeC possibility of extension to >100 GeV e- energies good progress with linac & IR designs many contributions from Turkish universities & TAC synergies with SPL, eRHIC, EIC, CLIC, ESS, … LHeC CDR due by coming October

thank you!