April 30, 2005J. Sandweiss RHIC II EM Probes Direct Gammas and HBT Direct photons from the collision are an important, and different, signal carrying information about the collision system. The EM interaction is weak enough so that photons do not appreciably interact after emission. Thus, photons from the initial “QGP” phase can be observed (with the complication that photons from the hadronic phase are also observed). Decoupling is not trivial but there are approaches, especially pt. At pt > 1.5 GeV/c direct subtraction of pi zero photons is feasible. Below 1.5 GeV/c it is not.

April 30, 2005J. Sandweiss RHIC II EM Probes Direct Gamma and HBT The key point is that π 0 gammas will not produce an HBT peak that is detectable. Typical distance between two π 0 s is 25 million Fermis! So the peak in relative momenta of the two gammas is 4 μeV wide. HBT gives space-time information on the collision region. This can be done for different pt, important for disentangling early and late times. The HBT peak is small and good statistics and good systematics are crucial.

April 30, 2005J. Sandweiss RHIC II EM Probes Our Work For several years a group of us have worked on gamma-gamma HBT * Alexei Chikanian Evan Finch Richard Majka Guoji Lin J.S. Interesting results will come from the last run. But a full experiment requires RHIC II upgrades – DAQ 1000, new calorimeter,.1 R.L. converter. This talk focuses on RHIC II.

April 30, 2005J. Sandweiss RHIC II EM Probes Our Approach 1)Produce a model of direct gammas which is “reasonable” estimate. This is not a prediction, but a test case on which we can evaluate experimental requirements. 2)Simulate detection, including effects of particle and gamma background. 3)Use 1 gamma in TPC, 1 gamma in calorimeter.

April 30, 2005J. Sandweiss RHIC II EM Probes Shashlyk Calorimeter 1.Consists of alternating sheets of lead and scintillator with longitudunal wavelength shifting fibers. 2.A resolution of 4%/√E has been achieved NIM A531, 467 (2004) and 3%/√E has recently been reached (M. Zeller, prvt. Comm) 3.Efficiency is essentially 100% for energies down to 100 MeV. 4.Cost for STAR version (estimated!) 5 million$.

April 30, 2005J. Sandweiss RHIC II EM Probes Converter A 0.1 radiation length converter will be a cylinder concentric with the beam line with a radius of 45 cm, which fits inside the inner field cage. We accept only those conversions which occur at this converter. Simulations have been done to calculate the TPC detection efficiency, including reconstruction, as a function of energy and angle. It averages about 7%.

April 30, 2005J. Sandweiss RHIC II EM Probes Calorimeter Cuts The π 0 spectrum is taken from Phenix. Decay photons are then propagated through the converter. TPC efficiency, and calorimeter photons computed. “Seed” photons with energy Above 100 MeV are identified. Photons within 4.5 cm, (.75 Moliere radius) are added to seed. This simulates the effect of undistinguished overlapping photons. Any resulting “photon” is then are kept for HBT if there is no other photon within 9.0 cm (1.5 Moliere rad.)

April 30, 2005J. Sandweiss RHIC II EM Probes π 0 - spectra for central Au+Au at 200 GeV

April 30, 2005J. Sandweiss RHIC II EM Probes  yield predictions

April 30, 2005J. Sandweiss RHIC II EM Probes Single photon conversion event

April 30, 2005J. Sandweiss RHIC II EM Probes Time evolution of Boltzman spectrum T 0 = 500 MeV, τ = 5 fm dN γ /(dE γ dt) ~ E γ · exp(-E γ /T(t)), T(t) = T 0 · exp(-t/τ)

April 30, 2005J. Sandweiss RHIC II EM Probes Direct γ generator algorithm γ - parameters in comoving with the hot matter reference frame x 0 = R σ · g 1, y 0 = R σ · g 2, z 0 = 0 - coordinates, where g i is Gaussian random number with σ = 1, t 0 = -τ · log(√ r 1 ) - time T = T 0 ·√ r 1 - temperature E 0 = -T · log(r 2 ·r 3 ) - energy cos(θ) = 2·r 4 – 1 - polar angle cos φ= 2π·r 5 - azimutal angle P x0 = E 0 ·sin(θ)·cos(φ), P y0 = E 0 ·sin(θ)·sin(φ), P z0 = E 0 ·cos(θ) - momentum components, where r i is flat random number in (0-1) interval

April 30, 2005J. Sandweiss RHIC II EM Probes Direct γ generator algorithm (cont.) Transformation into labframe Y lab = Y max · (2·r 6 -1), where Y max ≈ 1 is the STAR rapidity acceptance. Y 0 = -log(tg(θ/2)), where θ is polar emission angle in the comoving frame ∆Y=Y lab -Y 0 γ = cosh(∆Y), Lorentz gamma factor γ∙β = sinh(∆Y) x = x 0, y = y 0, z = γ∙z 0 + γ∙β∙t 0, t = γ∙t 0 + γ∙β∙z 0 P x = P x0, P y = P y0 P z = γ∙P z0 + γ∙β∙E 0 E = γ∙E 0 + γ∙β∙P z0

April 30, 2005J. Sandweiss RHIC II EM Probes Pair correlation variables Q inv, Q osl, Q xyz Q 2 inv = 2·( |P 1 |·|P 2 | - (P 1 ·P 2 ) ) = 4·E 1 ·E 2 ∙Sin 2 (θ/2 ) Q 2 osl = Q 2 out + Q 2 side + Q 2 long Q out = (P 2 tr1 - P 2 tr2 ) / P tr, where P tr = | P tr1 +P tr2 | - pair transverse momentum Q side = 2· | P tr1 ×P tr2 | / P tr Q long = γ z ·((P z1 -P z2 ) + β z ·(E 1 -E 2 )), γ z, β z – correspond to the reference frame, where P z pair = (P z1 + P z2 ) = 0. Q 2 xyz = Q 2 x + Q 2 y + Q 2 z, Q x =P x1 -P x2, Q y =P y1 -P y2, Q z =P z1 -P z2

April 30, 2005J. Sandweiss RHIC II EM Probes HBT signal from direct photons only 10 6 events, 0 = 1, R σ = 3 fm

April 30, 2005J. Sandweiss RHIC II EM Probes Checking of λ/λ 0 ≈ N 2 dir / N 2 tot relation

April 30, 2005J. Sandweiss RHIC II EM Probes γγ HBT signal

April 30, 2005J. Sandweiss RHIC II EM Probes γγ HBT signal, P t dependence

April 30, 2005J. Sandweiss RHIC II EM Probes Summary A procedure has been developed which permits the measurement of gamma- gamma HBT signals despite the large background of gammas from π 0 mesons This requires the simulation with Monte Carlo of the “no HBT” signal. It appears that this can be done and is not sensitive to the assumed spectra at a significant level.

April 30, 2005J. Sandweiss RHIC II EM Probes summary For the total yield of direct photons, event samples of 200·10 6 events is adequate. For 3D fits, samples about 3 times larger are desirable. P t dependence of the direct gamma yield is still under study. The increased fraction of direct gammas and the possible advantages of triggering would suggest that this can be studied with samples which could be obtained with DAQ Space time information from the HBT would be of great interest.