1. Next colliders for HEP
Marco Zanetti, MIT

2. Outline
Disclaimer:
Very partial and “Higgs-centric” overview on HEP collider projects
Even further, well-known projects as the ILC or the muon collider not discussed..
Introduction, the physics scenario after 1st LHC run
Circular e-e+ collider (TLEP)
Very High Energy LHC
Photon collider (Sapphire)

3. Status of HEP after LHC1
“A Higgs Boson” (CERN D.G. dixit) has been found, tremendous success of the Standard Model
No other direct observations of new phenomena at the LHC
Naturalness much less trendy these days
No indirect observations of deviations w.r.t SM predictions (Dark stuff apart)
Higgs couplings => Decoupling Limit
Impressive consistency of flavor sector
Is there a desert in front of us??

4. Status of HEP after LHC1
Is there something over there??

5. What is still worth looking for
The Higgs is the most exotic object found so far
Its properties has to be studied in details
Decoupling limit => ~1% uncertainty required
(HL)LHC will get to ~5% precision, most likely not enough => Higgs factories
Something has to be up there (at least once we get to the Plank scale..)
Step upward in energy is worth anyway => very large proton colliders
Are we sure we haven’t left anything behind?
Dedicated machines, e.g. photon colliders
Beyond (HL)LHC

6. Circular e+e- Colliders
Higgs Factories
Linear Colliders
Circular e+e- Colliders
LEP3
TLEP
Super-Tristan
FNAL Site-filler
IHEP, + …
ILC
CLIC
SLC-type
Adv. Concepts
Not discussed
SAPPHIRE,
CLICHÉ,
SLC like …
- Colliders
Muon Colliders

7. Required features of Higgs factories
Generalities:
Higgs should not be only physics goal (otherwise needs to be cheap..)
Upgradability in energy or diverse physics program
e+ e- colliders
Integrated luminosity is the main parameter
Additional useful (but not crucial) features:
Polarized beams
Low beam background
Energy spread
Muon colliders
Energy spread (s-channel production)
Integrated luminosity
Photon colliders

8. TLEP overview
TLEP: A High-Performance Circular e+e- Collider to Study the Higgs Boson, A. Blondel, M. Zanetti, F. Zimmermann, et al. arXiv:
Get up to √s=350, top-antitop production, L=0.7x1034 cm-2s-1
Higgs factory at Z+H threshold √s=250, L=5x1034 cm-2s-1
GigaZ, L=1036 cm-2s-1, repeat LEP1 program in 5 min
Possibility for several interaction points => multiply L, experimental redundancy
Challenging but well established technology
Cost-wise in the shadow of the proton-proton program (??)

9. 80-km Tunnel Cost Estimate (preliminary!)
Costs
Only the minimum civil requirements (tunnel, shafts and caverns) are included
5.5% for external expert assistance
(underground works only)
Excluded from costing
Other services like cooling/ventilation/
electricity etc
service caverns
beam dumps
radiological protection
Surface structures
Access roads
In-house engineering etc etc
Cost uncertainty = 50%
Next stage should include costing based on technical drawings
CE works
Costs [BCHF]
Underground
Main tunnel (5.6m)
Bypass tunnel & inclined tunnel access
Dewatering tunnel
Small caverns
Detector caverns
Shafts (9m)
Shafts (18m)
Consultancy (5.5%)
TOTAL
~3.1?
(unofficial)
(→raw tunnel cost could be 4.5 BCHF)
John Osborne & Caroline Waaijer (CERN)
21 February 2013

10. Beam lifetime
Bhabha scattering cross section (s~0.215 barn) implies a burn-off lifetime of ~20 min at 1e34 cm-2Hz
Reminder, SuperKEKB: t~6min!
Solution: top-up injection
Fundamental also for Hubner factor => guarantee high integrated L
High lumi => non-negligible beamstrahlung. Can we keep the beams circulating long enough?

11. Beamstrahlung
Electrons are lost if they emit a photon with E>hE0 (h momentum acceptance)
Defining:
The number of photons with E>hE0 (i.e. impacting lifetime):
h can be traded off with Nb/sxsz
High lumi and decent lifetime requires either high momentum acceptance or aspect ratio

12. Momentum acceptance ±1.3% ±2.0% ±1.6%
Very preliminary IR designs aimed at high momentum acceptance
2.5-3% feasible?
±1.3%
K. Oide
SLAC/LBNL design
KEK design
±2.0%
±1.6%
FNAL site filler
Y. Cai
T. Sen, E. Gianfelice-Wendt, Y. Alexahin

13. Aspect ratio Initial TLEP proposal matching LEP performances (ke~190)
Need to be more aggressive, currently aim at ke~400
Will benefit a lot from superKEKB and ILC (ATF2)
Still vertical dimension > 100nm

14. BS lifetime
Simulate and track O(109) macroparticles and check the energy spread spectrum (Guinea-Pig)
Lifetime computed from the fraction of particles beyond a given momentum acceptance (h)
Exponential dependence on h
ke=190
ke=370

15. BS Photons TLEP(3) BS photon spectrum is much softer than ILC
Tails up to only a few GeV, compared to tens of GeV for ILC
As a consequence much reduced pairs background
BS g spectrum
pairs spectrum

16. Luminosity profile
Softer BS photon spectrum implies much better luminosity profile
Intrinsic feature of circular high lumi e+e- colliders

17. Charge Compensation(?)
Visionary approach
2 opposite charged bunches per side
Null charge, no beamstrahlung
Could in principle allow much higher luminosities, but
Tried 40 years ago (DCI, Orsay) not a spectacular success (no discernible gain seen):
“small initial bunch displacement errors lead to charge separation” and “a minute deviation from neutrality is amplified as the like-charge beams repel each other”
Beam instrumentation much improved nowadays
The increase in costs for a four-beam solution is substantial
Spurios e+e+ and e-e- collissions

18. Top-up injection SPS-LEP experience:
e± from 3.5 to 20 GeV (later 22 GeV) in 265 ms (~62.26 GeV/s) [K. Cornelis, W. Herr, R. Schmidt]
Injection sequence [P. Collier, G. Roy]:
SPS-> top-up accelerator at 20 GeV
Accelerator from 20 to 120 GeV
Overall acceleration time = 1.6 s
Total cycle time = 10 s looks conservative (→ refilling ~1% of the LEP3 beam, for t~16 min)

19. Top-up cycle beam current in collider (15 min. beam lifetime)
100%
99%
almost constant current
energy of accelerator ring
120 GeV
injection into collider
injection into
accelerator
20 GeV
10 s

20. Top-up performances Super efficient duty cycle achieved at PEPII
Hubner factor not far from 1:
July 3, 2006: H≈0.95
August 2007: H≈0.63
Before top-up
During top-up
J. Seeman,
7 Dec. 2012

21. Synchrotron radiation
2x100 MW supplied to the beams need to be cooled away, heat load non negligible
Previous machines (e.g. PEP-II and SPEAR) coped with much higher heat load per meter
Need to manage higher max photon energy though
N. Kurita, U. Wienands, SLAC

22. Synchrotron radiationpp
A. Fasso
3rd TLEP3 Day
original
LEP design

23. Power consumption
Fixing energy, beam-beam limit and beamstrahlung conditions: => power is linearly proportional to luminosity
For TLEP self imposed limit on power to beams ~200MW, assuming 50% wall to beam efficiency
Complete accounting of power consumers brings the total to beyond 300 MW for TLEP at top energy
To be compared with current max cern site consumption of <200MW
Still margin thanks to possibility of several IPs
Number of IPs affecting BS lifetime

24. Power consumption TLEP-t
Wall-plug RF power
218 (1) [181 w/o RF feedback]
RF cryo power
24 (2)
Magnet system power
6 (3)
Cooling and ventilation
60 (4)
Experiments
25 (5)
General services
15 (5)
SPS & PS as pre-injectors (20 & 3.5 GeV)
5 (6)
e-/e+ source & pre-pre-injector
1(7)
Total
354 [318 w/o RF feedback]
(1): wall power efficiency: power converters: 95%; klystron efficiency: 65%; transmission losses 7%; overall 55% (from the LHeC design report); includes 36 MW for RF feedback margin (which may not be necessary)
(2): 60% of LHeC ; cryo power depends on cavity Q0 (34 kW at 2 K for 1200 cavities with Q0 =2.5e10)
(3): from LHeC ring-ring magnet design; power for 1 magnet (5.4 m, T) = 270 W; assuming 2x80 km of magnets at T (dipole field for 175 GeV beam energy)
(4): TLEP three times more than LHC; maximum capacity for LEP
(5): as for LHC (see appendix)
(6): conservative estimate scaled from higher-energy proton operation
(7): L. Rinolfi, private communication
M.Koratzinos / F. Zimmerman

25. Power Consumption TLEP-H
Wall-plug RF power
44 (1)
RF cryo power
6 (2)
Magnet system power
6 (3)
Cooling and ventilation
30 (4)
Experiments
25 (5)
General services
15 (5)
SPS & PS as pre-injectors (20 & 3.5 GeV)
5 (6)
e-/e+ source & pre-pre-injector
1(7)
Total
132
*: low power version with 1e34 cm^-2/s-1 in each of 4 IPs
(1): wall power efficiency: power converters: 95%; klystron efficiency: 65%; transmission losses 7%; overall 55% (from the LHeC design report); including RF feedback margin
(2): 60% of LHeC ; cryo power depends on cavity Q0 (34 kW at 2 K for 1200 cavities with Q0 =2.5e10)
(3): from LHeC ring-ring magnet design; power for 1 magnet (5.5 m, T) = 400 W; assuming 2x80 km of magnets at T (dipole field for 175 GeV beam energy)
(4): TLEP three times more than LHC; maximum capacity for LEP ; reduced by a factor of 2 for low power
(5): as for LHC (see appendix)
(6): conservative estimate scaled from higher-energy proton operation
(7): L. Rinolfi, power of LPI, private communication
M.Koratzinos / F. Zimmerman

26. Polarization
LHeC equilibrium polarisation vs ring energy, full 3-D spin tracking results [D. Barber, U. Wienands, in LHeC CDR]
Up to 80% at Z pole
For TLEP conceivable to extend sizable polarization up to ~80 GeV

27. VHE-LHC
VHE-LHC
TLEP

28. VHE-LHC
Main reason for building an 80 km tunnel, extremely challenging machine
Aim at √s>80 TeV pp collisions, need substantial SC magnet R&D (being pursued already for HE-LHC)
Leveled lumi ~5e34 cm-2Hz => 200 pileup events
SR heat load
33 W/m, need photon stops
Arc quads (if same length as LHC)
223 T/m x (50/7) = T/m at 50 TeV
Collimation and Machine Protection
New injection chain
Beam Dump System
4.5 GJ beam energy, ~3km dump system

29. parameters – 1 smaller?! (x1/4?) available now at LHC! >3.0 ?
O. Dominguez, L. Rossi, F.Zimmermann

30. parameters – 2
33?
*100?
O. Dominguez, L. Rossi, F.Zimmermann

31. H->gg
Outlier in the Higgs measurement at the LHC (now only in ATLAS though)
Triggered a lot of attention on the Hgg coupling
Sensitive to BSM physics in the decay loop
Might well be a fluctuation, should not get too excited..
The inverse process remains anyhow very interesting!

32. gg colliders in a nutshell
e-e- colliders equipped with high power laser beams
Compton backscattering of the lasers photon off the electron beams
Energy-angle correlation of the scatter photons => collimated gg collisions at ~0.8√see
Several possible layouts, in particular coupled to Linear Colliders (PLC, CLICHÉ)
In the following focus on a LHeC spinoff

33. gg collider basics Compton scattering energy:
Max energy for back scattering g energy:
Avoid e+e- pair production => limit √s of backscattered photon and laser => limit on x=> limit on laser photon w:
Compton cross section depend on relative g and e- polarization (l and P):
For the Higgs production need polarized beams

34. gg collider Luminosity
gg Luminosity depends on:
The number of beam electrons scattering at least once on laser beam
“thickness” k of the laser beam (related to the laser power and pulse duration)
Normalized distance r of the Compton Scattering point to the interaction point
Quasi monochromaticity of gg interaction energy:
Thanks to energy-angle correlation, q~1/g

35. gg collider Luminosity
Typically electron scatter more than once, low energy tails get populated
Level of low energy component can be regulated by tuning the CP-IP distance
Order of a few mm
Spent beams could provide additional luminosity (e-e-, e-g)

36. SAPPHiRE
LHC
SAPPHiRE: a Small Gamma-Gamma Higgs Factory, S.A. Bogacz, J. Ellis, D. Schulte, T. Takahashi, M. Velasco, M. Zanetti, F. Zimmermann, arXiv:

37. SAPPHiRE parameters 4 arcs for each beam on each side
Overall length 8 km
1/6 of the geometrical e+e- lumi
10k Higgs per year

38. Laser system The most technologically challenging component
Need high power lasers for efficient conversion (k~1)
Pulse energy of a few J, 5 ps long pulse, 1MW average power
Fiber laser
Stacked passive optical cavity pumped by a laser via a semi-transparent mirror
Amplification of

39. MightyLaser experiment at KEK-ATF
non-planar high finesse four mirror Fabry-Perot cavity;
first Compton collisions observed in October 2010
I. Chaikovska, N. Delerue, A. Variola, F. Zomer et al
Vacuum vessel for Fabry-Perot cavity installed at ATF
Optical system used for laser power amplification and to inject laser into FPC
Plan:
improve laser
and FPC mirrors
& gain several
orders
Comparison of measured and simulated gamma-ray energy spectra from Compton scattering
Gamma ray spectrum for different FPC stored laser power

40. Alternative approach (FEL)
Possibility of coupling the setup to a free-electron laser is very interesting
Get synchronization for free
Reduction in cost and complexity

41. Energy Loss Energy loss per arc:
Sapphire (LHeC) case, r=764m => electrons lose 4 GeV
Can be compensated by increase the linacs to 10.5 GeV
beam energy [ GeV]
DEarc [GeV]
DsE [MeV] 10
0.0006
0.038 20
0.009
0.43 30
0.05
1.7 40
0.15
5.0 50
0.36 60
0.75 70
1.39 35 80
1.19 27
total
3.89
57 (0.071%)

42. e- beam emittance growth
Horizontal emittance growth of the electron beam may be a severe limitation:
LHeC optics (lbend=10m, r=764m, <H>=1.2e-3m) lead to a too high emittance growth (DeN=13mm)
80 GeV instead of 60, high price due to 6th power of energy
<H> scale as l3bend/r2 => shorten lbend by factor 4=> down to 1mm at 80 GeV

43. Flat polarized e- guns FNAL A0 line injector test facility:
starting with ge~4-5 mm at 0.5 nC, achieved emittances of 40 mm horizontally and 0.4 mm vertically (ex/ey~100)
SAPPHiRE needs only ex/ey=10 but:
Larger bunch charge 1.6 nC and smaller initial ge (~1.5 mm)
Altogether, within state of the art
Main question is whether we can get polarized beams with those parameters
Ongoing efforts:
low-emittance DC guns
MIT-Bates, Cornell, JAEA, KEK [E. Tsentalovich, I. Bazarov, et al]
polarized SRF guns
FZD, BNL, etc. [J. Teichert, J. Kewisch, et al]

44. Spent beam Crossing angle in a crab-waist schema to get rid of e- beam
Minimize spurious collisions
Disruption angle (12mrad), outer radius of the last quadrupole (5-10 cm) and its distance from the IP (4m) determine the mininal crossing angle: ~25mrad
Beamstrahlung to be watched out
Special decay paths for hard photons

45. Spent beams
Integration of the laser, the incoming e- beams and the spent beams with the detector (need good vertexing capabilities) is quite challenging
Beam background need to be dealt with:
Low energy remnants
pileup
Incoherent e+e- pairs
neutron background

46. Potential concerns (Telnov)
Conservation of polarization in rings is a problem (due to energy spread).
The bunch length (σz = 30 μm) is very close to condition of coherent radiation in arcs.
With 80GeV e- and short bunches very large disruption angle (q~Esz-0.5) leading to unacceptable background levels.
Probably not the best layout for a gg collider (e.g. dedicated linac would be shorter)
It can only address a few aspects of Higgs program

47. (very) Tentative (CERN-centric) timeline

48. Summary
Mandate from Physics: study the Higgs, explore higher energy scale, don’t forget what could have left behind
VHE-LHC as next step towards the Plank scale (M. Mangano dixit)
A bit like going to the top of the building if the goal is reaching the moon
Anyhow a very challenging machine
ILC well established for addressing the Higgs and study phenomena up to TeV ranges
Watch out, it may be the desert out there
TLEP could be best option for precision Higgs physics and SM precision testing. Several challenges:
100 MW synchrotron radiation, activation by high SR(1.5MeV)
Momentum acceptance, beamstrahlung, aspect ration
Interface to detectors (synchrotron radiation, background, 4 beams xing)
gg colliders (Sapphire) technologically challenging
Offering many physics and application opportunities
Low energy precursor (e.g. IRIDE) will be instrumental!

49. BACKUP

50. luminosity formulae & constraints
SR radiation
power limit
beam-beam limit
>30 min beamstrahlung lifetime (Telnov) → Nb,bx
→minimize ke=ey/ex, by~bx(ey/ex) and respect by≥sz

51. LEP3/TLEP parameters -1 soon at SuperKEKB: bx*=0.03 m, bY*=0.03 cm
LEP2
LHeC
LEP3
TLEP-Z
TLEP-H
TLEP-t
beam energy Eb [GeV]
circumference [km]
beam current [mA]
#bunches/beam
#e−/beam [1012]
horizontal emittance [nm]
vertical emittance [nm]
bending radius [km]
partition number Jε
momentum comp. αc [10−5]
SR power/beam [MW]
β∗x [m]
β∗y [cm]
σ∗x [μm]
σ∗y [μm]
hourglass Fhg
ΔESRloss/turn [GeV]
104.5
26.7 4
2.3 48
0.25
3.1
1.1
18.5 11
1.5 5
270
3.5
0.98
3.41 60
100
2808 56
2.5
2.6
8.1 44
0.18 10 30 16
0.99
0.44
120
7.2
4.0 25
0.10 50
0.2
0.1 71
0.32
0.59
6.99
45.5 80
1180
2625
2000
30.8
0.15
9.0
1.0 78
0.39
0.71
0.04
24.3
40.5
9.4
0.05 43
0.22
0.75
2.1
175
5.4 12 20 63
0.65
9.3
SuperKEKB:ey/ex=0.25%

52. LEP3/TLEP parameters -2 LEP2 was not beam-beam limited LEP2 LHeC LEP3
LEP2
LHeC
LEP3
TLEP-Z
TLEP-H
TLEP-t
VRF,tot [GV]
dmax,RF [%]
ξx/IP
ξy/IP
fs [kHz]
Eacc [MV/m]
eff. RF length [m]
fRF [MHz]
δSRrms [%]
σSRz,rms [cm]
L/IP[1032cm−2s−1]
number of IPs
Rad.Bhabha b.lifetime [min]
ϒBS [10−4]
nγ/collision
DdBS/collision [MeV]
DdBSrms/collision [MeV]
critical SR energy [MeV]
3.64
0.77
0.025
0.065
1.6
7.5
485
352
0.22
1.61
1.25 4
360
0.2
0.08
0.1
0.3
0.81
0.5
0.66
N/A
0.65
11.9 42
721
0.12
0.69 1
0.05
0.16
0.02
0.07
0.18
12.0
5.7
0.09
2.19 20
600
700
0.23
0.31 94 2 18 9
0.60 31 44
1.47
2.0
4.0
1.29
100
0.06
0.19
10335 74
0.41
3.6
6.2
6.0
9.4
0.10
0.44
300
0.15
0.17
490 32 15
0.50 65
0.43
4.9
0.25 54
0.51 61 95
1.32
LEP data for GeV consistently suggest a beam-beam limit of ~0.115 (R.Assmann, K. C.)

53. Power efficiency

54. Physics performances: Higgs
Sub percent precision on the Higgs couplings
Total width accessible via both ZZ decay and VBF production

55. Physics performances: Higgs

56. Physics performances: low √s
Unprecedented precision on EW observables:
s(mW)~0.2 MeV, predict top mass at 100 MeV
Probe the loop structure, ultimate closure test of SM
Beam energy assessed by means of resonant depolarization
Dedicate one bunch during physics operation, no extrapolation needed

57. Upgradeability to higher √s
People contest the non-upgradeability in √s of a circular e-e+ collider.
Can a liner collider be upgraded to O(100) pp collider??
No doubts about the superiority of VLHC+TLEP in terms of physics program.

58. Beamstrahlung Beamstrahlung dependencies:
Flat beams, vertical size affects only luminosity
For a given bunch length, horizontal size and particles per bunch drive the BS effects
Same dependencies for the BS photon energy
Circular collider parameters designed to lead to smaller BS
N (1010)
sz (cm)
sx (mm)
sy (mm)
eNx (10-6 mrad)
eNY (10-6 mrad)
bx (m)
by (cm)
ILC 2
0.03
0.75
0.008 10
0.035
0.013
0.04
LEP3
100
0.23 71
0.32
6000 28
0.2
0.1
TLEP-H 50 43
0.22
2200 12

59. Integration with the Experiments
Need to arrange the top up accelerator nearby the experiment
Hole in the detector not acceptable
Long bypass around the experiments would impact sizably on the overall cost
O(10)x4 additional km
Accelerator and collider intersecting each other at the IP sharing a common beam pipe
Definitely not straightforward..

60. Dealing with BS Scan relevant BS parameters:
B*x to scale horizontal beam dimension
Number of particle per bunch
BS lifetime for nominal parameters (assuming h=0.04):
LEP3: >~ 30 min
TLEP-H: ~day
>4h for h=0.03, ~4 min for h=0.02
LEP3, h=0.02
LEP3, h=0.04

61. Power
The spectrum is softer and ng is smaller than ILC, but (T)LEP(3) have up to ~x100 more particles per bunch.
Comparable power dissipation for ILC and circular colliders, O(10) kW
Most of the power dissipated at very small angle
LEP3
Power (W/0.2 mrad)

62. circular HFs – arc lattice
KEK design
IHEP design
Q. Qin
K. Oide
FNAL site filler
SLAC/LBNL design
Y. Cai
T. Sen, E. Gianfelice-Wendt, Y. Alexahin

63. circular HFs – final-focus design
KEK design
IHEP design
K. Oide
Q. Qin
βx*=20cm,
βy*=0.5cm
FNAL site filler
SLAC/LBNL design
T. Sen, E. Gianfelice-Wendt, Y. Alexahin
Y. Cai

64. Summary of issues SR handling and radiation shielding
optics effect energy sawtooth [separate arcs?! (K. Oide)]
beam-beam interaction for large Qs and significant hourglass effect
IR design with even larger momentum acceptance
integration in LHC tunnel (LEP3)
Pretzel scheme for TERA-Z operation?
impedance effects for high-current running at Z pole

65. Extrapolation LEP3 TLEP-H LEP2→TLEP-H SLC→ILC 250 peak luminosity x400
energy
x1.15
x2.5
vertical geom. emittance
x1/5
x1/400
vert. IP beam size
x1/15
x1/150
e+ production rate
x1/2 !
x65
commissioning time
<1 year → ?
>10 years →?

67. M. Peskin statement on TLEP

68. collimation challenges higher energy density
→ need for more robust materials
cross section for single diffractive scattering
increases with energy → degraded cleaning efficiency
smaller beam sizes & smaller gaps → higher precision in collimator control
(warm? or shielded SC) magnets in the collimator insertions
VHE-LHC: 99 W/m
dedicated photon stops
R. Assmann, HE-LHC’10

69. State of the art (Garching MPI) : ~70kW, 2ps (F~5600) stored in a cavity (O.L.35(2010)2052) ~20kW, 200fs
From a feedback point of view:
Locking a ‘150m’ cavity to finesse~ nm is the same as Locking a 4m 800nm to ~25000 finesse
R&D done at Orsay
2ps Tis:apph 76MHz oscillator (~0.2nm spectrum)
cavity finesse ~28000

70. Summary Fabry-Perot cavity Optical issues Advantages
Very high gain (eventually)
‘easy’ laser-electron synchronization
Stable transverse & longitudinal modes
Though painful, laser/cavity feedback techniques are well know
Disadvantages for Sapphire
Very long cavity
technical noise (?)
Tight feedback as difficult as a highest finesse table top experiment…
(BW may be limited by the laser frep)
Very small laser waists & circ. Polar. (?)  careful optical design of the geometry and mirror shape
Optical issues
High peak power
coating damage threshold large mirrors
Large average power: thermal load effects
Thermal lens in the coupling mirror (cf VIRGO upgrade with >600kW)