Photoproduction at proton and ion colliders Photoproduction at hadron colliders Photonuclear and  interactions Physics from Photoproduction Unique Possibilities RHIC & Tevatron results Future Prospects: the LHC & RHIC LHC is the highest energy  p collider in the world Conclusions Spencer Klein, LBNL & UC Berkeley  0 photoproduction in STAR

Photons from Hadrons n Relativistic hadrons carry strong electromagnetic fields u Weizsacker-Wiliams: a field of almost-real photons F Virtuality Q 2 < (h/R A ) 2 Q 2  0 significant for e + e - production, and (maybe) with proton beams Photon E max ~  h/R A u 3 GeV with gold at RHIC u 80 GeV with lead at the LHC u ~10%+ of proton energy at both machines n Photon flux ~ Z 2 u Higher intensity with heavy ions F Important for considering multiple interactions between a single ion pair u RHIC & LHC have higher luminosities for light ions --> often better overall rates F “Optimum” ion depends on the problem Au Coupling ~ form factor , P, or meson X

Photonuclear Interactions Photon coupling Z 2  ~ 0.6 for heavy ions u Multi-photon reactions common  -gluon/Pomeron/meson interactions  -Pomeron/meson can be coherent u Exclusive final states u Coupling ~ A 2 (bulk) A 4/3 (surface) u Large cross-sections for vector meson production n Require b > 2R A u no hadronic interactions u ~ fermi at RHIC n Cross sections are huge  ~ 5 barns for coherent  0 production with lead at the LHC u Need plots Au Coupling ~ form factor Au X P/gluons/meson 

 interactions Rate ~  ~ Z 4 u Rates to produce hadrons are much smaller than for photoproduction of vector mesons n Prolific production of e + e - pairs   = 200,000 barns with lead at the LHC u QED process --> proposed as luminosity monitor. u Coulomb corrections are significant n W max ~ 2 k max u 6 GeV with gold at RHIC u 160 GeV with lead at the LHC u Higher for lighter ions Au X    Luminosity

Unique Possibilities at hadron colliders n Highly charged photon emitters u Multi-photon interactions F Impact parameter tagging n Symmetric initial state u Quantum interference n Ion interactions u Pair production with capture

Factorization n Multiple photons are emitted independently (Gupta, 1950) n To lowest order, interactions are independent u (neglecting quantum correlations) True for  + nuclear breakup with heavy ions Seems to work well for e + e - and  0 at RHIC u Cf. Yuri Gorbunov’s talk Au  P Au* 00   G. Baur et al, Nucl. Phys. A729, 787 (2003)

Impact Parameter Tagging n Nuclear excitation ‘tag’s small b n Multiple photon exchange u Mutual Coulomb excitation n Au* decay via neutron emission u Detected in zero degree calorimeters u simple, unbiased trigger n Multiple Interactions probable  P(  0, b=2R) ~ 1% at RHIC u P(2EXC, b=2R) ~ 30% n Non-factorizable diagrams are small for AA Au  P Au*  (1+) 00

Tagged Photon Spectra n Single photon exchange (low energy limit)  N(  ) ~ 1/k u The presence of additional tagging photons biases the collision toward small b n For single tagging (2-photon exchange)  N(  ) ~ independent of k G. Baur et al, Nucl. Phys. A729, 787 (2003) k (GeV) N(k) Photon spectra at RHIC

Strong Coupling/ Multiple Interactions Z 2  ~ 0.6 with gold/lead u ‘Extra’ photons are cheap n Higher order diagrams could be important n Multi-photon reactions are important u They factorize n Mutual Coulomb excitation of colliding nuclei is a useful tag u Simple experimental trigger u Selects events with small impact parameters Au  P Au*  00

Zero Degree Calorimeters & Luminosity measurement n Mutual Coulomb dissociation is used as to monitor luminosity u Calculable in QED F Accuracy limited by uncertainty in hadronic interaction radius n Require 1 neutron in each ZDC (& no activity in central detector) u via Giant Dipole Resonance PHENIX used dA(  ) --> pnA to monitor  *luminosity  -->l + l - proposed at LHC u Uncertainty in hadronic interaction radius limits accuracy

e + e - production w/ STAR n With Coulomb excitation in 200 GeV gold-on-gold collisions u 52 events F M ee > 140 MeV/c 2, |Y ee |< 1 n Equivalent photon method fails for pair p T spectrum u Photon virtuality is important n Cross-section is consistent with all-orders QED calculation u ~ Inconsistent with lowest order calculation F ~ 30% higher than all-orders in STAR acceptance Pair p T Lepton p T Points data Solid line – LO calc Dashed – all orders STAR data, calcs by A. Baltz PRL 100, (2007)

e + e - production in CDF n 1.8 TeV pp collisions n 16 events with M ee > 10 GeV n Fits theoretical calculations u LPAIR n Proposed e + e - pairs as a luminosity monitor at the LHC u QED is well understood u Problem: what is the effective b min, at which there are no hadronic interactions? CDF, PRL 98, (2007)

STAR -  0 Photoproduction Exclusive  0 triggered w/ ‘Topology’ trigger  0 -with mutual Coulomb excitation triggered with ZDC coincidence n Select events with 2 primary tracks u + some ‘global’ tracks Au 00  qq STAR, Phys. Rev. C77, (2008)

 0 Photoproduction Mass Spectra n Coherent production events u p T < ~ 150 MeV/c M  fit to  0 + direct  +  - spectrum u w/ interference term u Relativistic Breit-Wigner w/ phase space  0 :direct  +  - ratio same as  HERA Exclusive  0  0 + Mutual Coulomb Excitation  0  0 :  interference Gray -background STAR, Phys. Rev. C77, (2008)

 0 rapidity distribution  0 w/ mutual Coulomb excitation n Photon energy u k = M v /2 exp(y) u 2-fold directional ambiguity leads to symmetric distributions Cross-section depends on  qq-N, which depends on the hadronic model u Saturation models of Goncalves & Machado is ruled out   A->  A (k) can be determined by combining exclusive  and  w/ mutual Coulomb excitation u Errors are large S T A RS T A R STAR, Phys. Rev. C77, (2008)

 0 p T spectra n Coherent + Incoherent form factors u Fit to dual exponential n Incoherent production u b N = 8.8 ±1.0 GeV -2 F nucleon form factor n Coherent production u b Au = ±24.8 GeV -2 F b Au ~ R A 2 F Data sensitive to R A F Measure hadronic radius of gold 3% statistical uncertainty Significant theoretical corrections/uncertainty  (incoh)/  (coh) ~ 0.29 ±0.03 u Extrapolated form factor to t=0 S T A RS T A R p T 2 ~ t STAR, Phys. Rev. C77, (2008)

Polarized Photons n Electric fields parallel to b u photon polarization follows b n Use 1 reaction to determine the polarization, to study a 2nd reaction n Correlated Decay angles Use   -- >  +  - as ‘analyzer   +  - decay plane often parallels photon pol. u reasonable analyzing power  Study VM production angles, single spin  A asymmetries, etc. b,E (transverse view) ++ -- 

Interference n 2 indistinguishable possibilities u Interference!! n Similar to pp bremsstrahlung u no dipole moment, so u no dipole radiation n 2-source interferometer u separation b , , , J/  are J PC = n Amplitudes have opposite signs  ~ |A 1 - A 2 e ip·b | 2 n b is unknown u For p T F destructive interference No Interference Interference         p T (GeV/c) y=0

Entangled Waveforms n VM are short lived u decay before traveling distance b n Decay points are separated in space-time F no interference u OR F the wave functions retain amplitudes for all possible decays, long after the decay occurs n Non-local wave function  non-factorizable:   +  -    +   - n Example of the Einstein-Podolsky-Rosen paradox e+e+ e-e- J/  ++ -- b (transverse view) ++

Einstein-Podolsky-Rosen paradox For each  +  from   decay, can choose to measure x or p u Both x - localize production point (sometimes) u Both p - measure p of VM; see interference; delocalize production The  +  - system is an example of the Einstein- Podolsky-Rosen paradox n (Gedanken) test for superluminal collapse b  -- ++ ++ X/p meas. Klein & Nystrand, 2003 00 00

Interference and Nuclear Excitation n Smaller --> interference at higher u Suppression for p T  0 prod at RHIC: 0.1 < |y| < 0.6 Noint ~ exp(-bt) Int. Noint ~ exp(-bt) Int. t XnXn No breakup criteria t dN/dt

Interference Analysis ‘topology’ (multiplicity+) trigger for exclusive  0, n Neutrons in both ZDCs trigger for mutual Coulomb excitation. Tight cuts -- > clean  0, but lower efficiency u Exactly 2 tracks u 1 vertex with Q=0  0.55 & GeV<M  <0.92 GeV n Fit dN/dt spectrum u Require accurate dN/dt w/o interference u Efficiency is ~ independent of t F Momentum smearing affects two low-t bins Included in Monte Carlo n Interference is maximal at y=0, decreasing as |y| rises u 2 rapidity bins 0/0.05 < |y| < 0.5 & 0.5<|y|<1.0

 0 p T spectrum in UPCs n Scattering (Pomeron) p T u Modified nucleon form factor F Glauber calculation for absorption F position-dependent photon flux n Photon p T u Weizsacker-Williams + form factor n Add components in quadrature n w/o interference dN/dt ~ exp(-bt) u b = x * R A 2 F STAR simulations F Woods-Saxon + Glauber calculation (H. Alvensleben et al., 1970) u Modifications change x n Interference u Only cause for dN/dt -->0 at small t Woods-Saxon Form Factor

n 2 Monte Carlo samples: u Interference u No interference u w/ detector simulation F Detector Effects Small n Data matches Int n Inconsistent with Noint n Interference clearly observed n 973 events n Fit to dN/dt = A exp(-kt)* [1+c(R(t)-1)] u Exponential for nuclear form factor u R = Int(t)/NoInt(t) u Separates nuclear form factor (exponential) & interference Analysis Technique dN/dt (a.u.) Data (w/ fit) Noint Int LS Background STAR Preliminary t R = Int(t)/Noint(t) t (GeV 2 ) = p T 2 dN/dt for XnXn, 0.1<|y|<0.6

Results n Efficiency corrected spectra n Two rapidity bins, 2 triggers u Cut |y|<0.05 for topology to remove cosmic rays

Results  2 /DOF > 1 for 2 samples u Much study, no answers…  Scale errors up by   2 /DOF (PDG prescription) n Systematic errors due to trigger (10% for topology), other detector effects (4%), background (1%), fitting & nuclear radius(4%), theoretical issues (5%) Final, combined result: c=0.87 ± 0.05 (stat.) ± 8 (syst.) % u Limit on decoherence, due to decay or other factors < 23% at a 90% confidence level

k values n photon flux ~ 1/b 2 P(b) ~ N  (b)  (b) n For b ~ few R A, production is asymmetric u Smaller apparent size  Small change in  (tot) n ~ 3 R A for minimum bias data n F(t) is Fourier transform of production density u Parameter b ~ (diameter) 2 n Affects the interference pattern u Impact parameter b eff < b geom u Neglected here N  (b) ~ 1/(b-R A ) 2 N  (b) ~ 1/b 2 N  (b) ~ 1/(b+R A ) 2 Au Woods-Saxon (b) WS weighted by 1/b 2 b (fm) P(b) =  (b)N  (b) A linear (toy) model

n Photon is almost always emitted by gold  No two-fold ambiguity--> easy  extraction  d --> d  , pn  0 both seen n Deuteron dissociation tagged by neutron in Zero Degree calorimeter u Allows cleaner trigger n Coherent reaction has larger t-slope n Incoherent process reduced at small t u Not enough momentum transfer for dissociation  0 photoproduction in deuteron-gold interactions Neutron-tagged data t, GeV 2 S T A RS T A R

Photoproduction of  Expected to be largely through a radially excited    and/or  n Peak at low p T from coherent enhancement n Studies of resonant substructure are in progress STAR preliminary M  (GeV) p T  (GeV/c) S T A RS T A R B. Grube, Wkshp. on HE Photon Collisions at the LHC S T A RS T A R

J/  photoproduction at RHIC n e + e - pair + 1 nucleus breakup u Nuclear breakup needed for trigger  J/  + continuum  --> e + e - u ~ 12 events in peak Cross sections for both J/  and continuum e + e - ~ as expected D’Enterria, PHENIX, nucl-ex/

J/  photoproduction at the Tevatron CDF selects exclusive J/  photoproduction is sensitive to gluon structure of nuclei   ~ |g(x,M V 2 /4)| exclusive  +  - signal events. u “Background” from double Pomeron production   c -->  J/   Some  ’ n Cross section determination in progress M(  ) (GeV/c 2 ) J. Pinfold, Wkshp. on HE Photon Collisions at the LHC

UPCs at the LHC n CMS, ALICE and ATLAS plan programs n “Yellow Book” gives physics case u K. Hencken et al., Phys. Rept. 458, 1 (2008). n Gluon structure Functions at low-x u Including saturation tests n The ‘black disc’ regime of QCD n Search for exotica/new physics   --> Higgs, Magnetic monoples, etc. n Some techniques, and some final states overlap with pp diffraction (double-Pomeron) studies u Roman pots can be used for full kinematic reconstruction in pp  t_perp can help separate ,  P and PP interactions

Structure Functions at the LHC n Many photoproduction reactions probe structure functions   --> qq; the quarks interact with target gluons u Q 2 ~ (M final state /2) 2 u x ~ at midrapidity u x ~ in forward regions J/ ,  ’,  states n Open charm/bottom/top? n Dijets The twofold ambiguity can be avoided using pA collisions or by comparing d  /dy with and w/o mutual Coulomb excitation

J/  photoproduction  ~ g(x,Q 2 ) 2  x ~ few for J/ the LHC  x ~ few for J/ RHIC u Q 2 ~ M v 2 /4 n Coherent production u p T < h/R A n High rates u 3.2 Hz production with Pb n Detection is easy n Shadowing is ~ factor of several  Theoretical uncertainties cancel in  (pp)/  (AA) or  (pA) M. Strikman, F. Strikman and M. Zhalov, PL B540, 220 (2002) y y

Dijets n With calorimetery to |y|<3, probe down to x~10 -4 in 1 month n Use standard jet triggers L=4*10 26 /cm 2 /s M. Strikman, R. Vogt and S. White, PRL86, (2006)

“Black Disk” Regime n At high enough energies/low enough x, nuclei look like black (absorptive) disks u Studies of HERA data suggest that the Black Disk Regime (BDR) should be visible at the LHC for Q 2 < 4 GeV 2 n Photons fluctuate to qq dipoles, with separation d   dipole-target ~  d 2 F Smaller and smaller dipoles interact  Photons contain many small dipoles --> increasing number of interactions -->  tot (  p) rises rapidly u Increase in high p T interactions u Fraction of diffractive events increases M. Strikman

e + e - pairs at the LHC  ~ 200,00 barns with Pb beams u Significant background for small-radius detectors u 20 pairs/beam crossing (at /cm 2 /s & 100 nsec crossing) n e +, e - have small p T (typically m e ) u Hard to trigger n ALICE strategy – random events u Expect ~ MHz rates in inner detectors u Measure cross-section, M ee, p T spectra, etc. n Multiple pairs likely from 1 collision

LHC – Plans & Issues J/ ,  ’,  --> leptons is relatively easy u CMS, ALICE, ATLAS are pursuing n Di-jets u ATLAS is pursuing n Triggering is problematic u Beam gas, neutrons, grazing hadronic collisions... even cosmic ray muons u ZDC coincidences might help with backgrounds F Not always available at Level 0 n FP420, TOTEM & other forward detectors  Use tagging to select  --> Higgs, sleptons…

Bound Free Pair Production n A + A --> A + Ae - + e + u 1e - atom has lower Z/A F Less bending in dipoles u Momentum ~ unchanged --> beam  ~ 280 barns w/ lead at the LHC L u 280,000 ions/s at L = /cm 2 /s F 28 watts! u Hits beampipe ~ 136 m from the IP F Enough energy to quench superconducting magnets? n Limits LHC luminosity w/ heavy ions u Limit is ~ planned luminosity F Quench calculations are touchy IP Pb 82+ Pb 81+ SK, NIM A459, 51 (2001); R. Bruce et al. PRL 99, (2007)

Observation of BFPP at RHIC n Copper beams at RHIC   theory ~ 200 mb   magnetic rigidity) ~ 1/29 F 1e Au defocuses before striking beampipe n 1e Cu beam hits beampipe ~ 144 m from interaction point n Look for showers using pre-existing PIN diodes u Small enough that acceptance simulations are problematic n Observe correlations between luminosity (measured at IP) and PIN diode rates R. Bruce et al. PRL 99, (2007)

BFPP results Colors are PIN photodiode signals at different locations Blue is optimal location BFPP has been observed, with cross section ~ expected RHIC (ZDC) luminosity monitor

Future Directions at RHIC n STAR “DAQ1000” increases data rate by a factor of 10 & TOF system improves particle identification u Substructure in 4-pion events   0  0 pairs n Roman pots for STAR u Mostly for pp diffraction, but might be able to measure scattered deuteron in dA collisions, improving UPC reconstruction F Phase I optimized for elastic scattering Requires special low-  beam optics F Phase II has larger acceptance

 0  0 production 4 diagrams interfere Depends on p T sum and difference between two  n Away from y=0, the top and bottom diagrams dominate u Stimulated emission at low p T u Possibility to look for stimulated decays Needs more luminosity, and background rejection from  ’ decays, but looks possible

Angular Correlations in  0  0 & mutual nuclear excitation to Giant Dipole Resonances n ~ any states decaying via a vector decay In linearly polarized  0 decays, the angle between the  0 polarization and the  + /  - p T (wrt the direction of motion) follows cos(    + /  - direction acts as an analyzer The  0 polarization follows the  polarization The angle between the  + /  - p T in  0  0 decays is the convolution of the two cos(  ) distributions  C(  ) = 1 + 1/2cos(2  ) n Possible way to study linear polarization Angle between  + /  - p T in double VM decays G. Baur et al.,Nucl. Phys. A729, 787 (2003)

Conclusions n Many photoproduction reactions can be studied at hadron colliders. u STAR & PHENIX at RHIC, and CDF at the Tevatron have studied a variety of topics, including vector meson production and e + e - pair production. n Many unique reactions can be studied with heavy ion collisions. u Multiple interactions among a single ion pair. n Bound-free e + e - production will limit the LHC luminosity with lead. n The LHC and RHIC have promising future programs.  Until the ILC is complete, the LHC will be the highest energy  and  p collider in the world.