goethe university frankfurt gsi helmholtzzentrum für...

09/11/2016 Alberica Toia 1

Nuclear and Particle Physics 4aElectromagnetic Probes

Alberica ToiaGoethe University Frankfurt

GSI Helmholtzzentrum für Schwerionenforschung

Lectures and Exercise Winter Semester 2016-17


Organization● Language: English● Lecture:

● Wednesday 09:00 (c.t.) - 11:00 ● Phys 01.402

● Marks / examination→ only if required / desired● Seminar presentation → schein ● Oral Exam → grade

● Office hours: tbd on demand


Info: Email and Website● E-Mail:

[email protected]

● Website: https://web-docs.gsi.de/~alberica/lectures/KT4a_WS1617.html

mailto:[email protected]


Content● 1) Introduction: Heavy Ion Physics● 2) Detectors● 3) Dielectrons: low energy● 4) Dielectrons: intermediate energy● 5) Dielectrons: high energy● 6) Photons: intermediate energy● 7) Photons: high energy● 8) Dark Matter● 9) ElectroWeak Probes

OCTOBER

NOVEMBER

DECEMBER

JANUARY


Dielectrons: low energy● Motivation

● Dileptons and the QGP ● Dileptons and the Hadron Gas: Chirality, chiral symmetry

breaking and chiral symmetry restoration

● Experimental challenges● combinatorial background

● Dileptons in heavy-ion collisions experiments● Low-energy

– DLS– HADES


The little bang in the lab

• High energy nucleus-nucleus collisions: – fixed target – colliders

• QGP formed in a tiny region

(10-14m) for very short time (10-23s)– Existence of a mixed phase?– Later freeze-out

• Collision dynamics: different observables sensitive to different reaction stages


Probing the QGP

penetrating beam(jets or heavy particles) absorption or scattering pattern

QGP

Rutherford experiment atom discovery of nucleus SLAC electron scatteringe proton discovery of quarks

Penetrating beams created by parton scattering before QGP is formed High transverse momentum particles jets Heavy particles open and hidden charm or bottom

Probe QGP created in Au+Au collisions Calculable in pQCD Calibrated in control experiments: p+p (QCD vacuum), p(d)+A (cold medium)

Produced hadrons lose energy by (gluon) radiation in the traversed medium QCD Energy loss → medium properties

Gluon density Transport coefficient

probe

bulk


Electromagnetic Radiation Thermal black body radiation

Real photons Virtual photons * which appear as dileptonseeor

No strong final state interaction Leave reaction volume undisturbed and reach detector

Emitted at all stages of the space time development Information must be deconvoluted

time

hard parton scattering

AuAu

hadronization

freeze-out

formation and thermalization of quark-gluon

matter?

Space

Time Jet cc e p K


What can we learn from dilepton emission?Emission rate of dilepton per volume

Boltzmann factortemperature

EM correlatorMedium property

ee decay

Hadronic contributionVector Meson Dominance

qq annihilation

Medium modification of mesonChiral restoration

From emission rate of dilepton, one can decode

• medium effect on the EM correlator • temperature of the medium

arXiv:0912.0244

Thermal radiation frompartonic phase (QGP)

q

q

e+

e-

e+

e-

+

-

Photonself-energy

QGP

HadronGas


Theory predictions for dilepton emission rate

Vacuum EM correlatorHadronic Many Body theoryDropping Mass Scenarioq+q → →ee (HTL improved)(q+g → q+→qee not shown)

Theory calculation by Ralf Rapp

dMdydpp

dN

tt

ee at y=0, pT=1.025 GeV/cUsually the dilepton emission is measured and compared as dN/dpTdM

The mass spectrum at low pT is distorted by the virtual photon → ee decay factor 1/M, which causes a steep rise near M=0

qq annihilation contribution is negligible in the low mass region due to the M2 factor of the EM correlator

In the caluculation, partonic photon emission processq+g → q+ → qe+e- is not included

1/M*→ee

qq → * → e+e-

≈(M2e-E/T)×1/M

arXiv:0912.0244


The mass of composite systems

mass given by energy stored in motion of quarks and by energy in colour gluon fields

M mi

binding energyeffect 10-8

atom 10-10 m

M » mi

nucleon 10-15 m

atomic nucleus 10-14 m

M mi

binding energyeffect 10-3

the role of chiral symmetry breaking

• chiral symmetry = fundamental symmetry of QCD for massless quarks

• chiral symmetry broken on hadron level


Chirality

left-handed

right-handed

Chirality (from the Greek word for hand: “”)when an object differs from its mirror image

simplification of chirality: helicity(projection of a particle’s spin on its momentum direction)

massive particles P left and right handed components must exist m>0 particle moves w/ v<c

– P looks left handed in the laboratory– P will look right handed in a rest frame

moving faster than P but in the same direction chirality is NOT a conserved quantity

in a massless word m

u = m

d = m

s = 0

– chirality is conserved


QCD and chiral symmetry breaking the QCD Lagrangian:

free gluon field free quarks ofmass mn

explicit chiral symmetry breaking mass term mnnn in the QCD Lagrangian

chiral limit: mu = md = ms = 0 chirality would be conserved

left–handed u,d,s, quarks remain left-handed forever all states have a ‘chiral partner’

(opposite parity and equal mass)

real life: mu and m

d are so small (m

u≈ 4 MeV m

d≈ 7 MeV)

that our world should be very close to chiral limit a1 (JP=1+) is chiral partner of (JP=1-): m≈500 MeV even worse for the nucleon:

N* (½-) and N (½+): m≈600 MeV (small) current quark masses don’t explain this

chiral symmetry is also spontaneously broken spontaneously = dynamically

interaction of quarkswith gluon


Origin of mass constituent quark mass

~95% generated by spontaneous chiral symmetry breaking (QCD mass)

current quark mass generated by spontaneous

symmetry breaking (Higgs mass) contributes ~5% to the visible

(our) mass


Chiral symmetry restoration spontaneous symmetry breaking gives rise to a nonzero ‘order

parameter’ QCD: quark condensate <qq> ≈ -250 MeV3

many models (!): hadron mass and quark condensate are linked

numerical QCD calculations at high temperature and/or high baryon density

→ deconfinement and <qq> → 0 approximate chiral symmetry restoration (CSR)

→ constituent mass approaches current mass

Chiral Symmetry Restoration expect modification of hadron

spectral properties (mass m, width )

explicit relation between (m,) and <qq>?● QCD Lagrangian → parity doublets are degenerate in mass

● how is the degeneracy of chiral partners realized ? ● do the masses drop to zero? ● do the widths increase (melting resonances)?


Hadron massesG. Brown & M. Rho, PRL (1991) 2720

Brown-Rho scaling

Vacuum: Vector and Axial spectral functionswell separated

At Tc: Chiral symmetry restoration


CSR and low mass dileptons what are the best probes for CSR?

requirement: carry hadron spectral properties from (T, B) to detectors relate to hadrons in medium leave medium without final state interaction

dileptons from vector meson decays

best candidate: meson– short lived – decay (and regeneration) in medium– properties of in-medium and of medium itself not well known

meson a special probe for CSR, long lifetime but m(Ф)≈ 2 m(K) simultaneous measurement of φ → ee and φ → KK could be a

powerful tool to evidence in-medium effects

m [MeV] tot [MeV] [fm/c] BR → e+e-

770 150 1.3 4.7 x 10-5

8.6 23 7.2 x 10-5 4.4 44 3.0 x 10-4


Dilepton Signal● What is its temperature?

→ measure thermal photons

● Does it restore chiral symmetry?→ modification of the vector mesons

● How does it affect heavy quarks?→ modification of the intermediate mass region

● All these questions can be answeredby measuring dileptons (e+e− or μ+μ−)

● no strong final state interactions:

● leave collision system unperturbed

● emitted at all stages: need todisentangle contributions


Dilepton Signal Dileptons characterized by 2 variables: M, pT

M: spectral functions and phase space factors pT: pT - dependence of spectral function (dispersion relation)

T - dependence of thermal distribution of “mother” hadron/partonM - dependent radial flow () of “mother” hadron/parton

Note I: M Lorentz-invariant, not changed by flow

Note II: final-state lepton pairs themselves only weakly coupled

dilepton pT spectra superposition of ‘hadron-like’ spectra at fixed T

early emission: high T, low T

late emission: low T, highT

final spectra from space-time folding over T- T history from Ti → Tfo

→ handle on emission region, i.e. nature of emitting source

mT

1/m

T d

N/d

mT

light

heavyT

purely thermalsource

explosivesource

T,

mT

1/m

T d

N/d

mT

light

heavy


Dilepton signal Low Mass Region:

mee < 1.2 GeV/c2

Dalitz decays of pseudo-scalar mesons Direct decays of vector mesons In-medium decay of mesons in the hadronic gas phase

Intermediate Mass Region: 1.2 < mee < 2.9 GeV/c2

correlated semi-leptonic decays of charm quark pairs

Dileptons from the QGP

High Mass Region: mee> 2.9 GeV/c2

Dileptons from hard processes–Drell-Yan process–correlated semi-leptonic decays of heavy quark pairs

–Charmonia –Upsilons

→ HMR probe the initial stage Little contribution from thermal radiation

• LMR: mee < 1.2 GeV/c2 o LMR I (pT >> mee)

quasi-real virtual photon region. Low mass pairs produced by higher order QED correction to the real photon emission

o LMR II (pT<1GeV)Enhancement of dilepton discovered at SPS (CERES, NA60)


HI low mass dileptons at a glance

(KEK E235) CERES

DLS

NA60

HADES

CBM

90 95 1000 0585

PHENIX

● Time scale of experiments

● Energy scale of experiments

(KEK E235)

CERES

DLS

NA60

HADES

CBM PHENIX

10 158 [A GeV]

17 [GeV]√sNN200

// // //

// // //

ALICE

[A TeV]

1


Dilepton Analysis Challenges● Experimental Challenge

● Need to detect a very weak source of pairs ~ 10-6 /π0 ● in the presence of hundreds of charged particles in central AA

collision● and several pairs per event from trivial origin

π0 Dalitz decays ~ 10-2/π0 + γ conversions (assume 1% radiation length) 2x10-2 /π0 → huge combinatorial background (dN∝

ch/dy)2

● Analysis Challenge

● Electron pairs are emitted through the whole history of the collision (from the QGP phase, mixed phase, HG phase and after freeze-out)

– need to disentangle the different sources.

– need excellent reference pp and dA data.

– need independent information about the known sources in nuclear collisions


Dilepton Analysis Steps● Tracking + Momentum reconstruction → Resolution

(position, momentum)● Particle Identification → Purity● Rejection close pairs → Significance,

Signal/Background

● Pairing: mee

= [2 p1p

2 (1-cosθ)]1/2

● Subtraction of Background(mixed events, like-sign)

● Efficiency Correction● Mass Spectrum


Remarks on S/B● how is the signal obtained?

● unlike-sign pairs: F● combinatorial background: B (like-sign pairs or event mixing)

→ S = F – B

● statistical error of S● depends on magnitude of B, not S

S ≈ √2B (for S<<B)

● “background free equivalent” signal Seq

● signal with same relative error in a situation with zero backgroundSeq = S * S/2Bexample: S = 104 pairs with S/B = 1/250 → Seq = 20

● systematic uncertainty of S● dominated by systematic uncertainty of B

S/S = B/B * B/Sexample: B/B = 0.25% precision, S/B = 1/250 → S/S = ~60% systematic uncertainty of S

S2 = F2 + B2 = S2 + B2 + B2

≈ 2B2 = 2B2

B = √B

Seq

/ Seq

= √2B / S

Seq

= √2Seq

/ √S

eq = √2B / S

/ S

eq = 2B / S2


Separating Signal from Background


SIS/BEVALAC energy regime: 1-2 AGeV

• Final state (Freeze-out) in heavy ion

collisions

– approximately 10-15 % pions, baryon

resonances ((1232)) dominated

– up to 200 charged particles (Au+Au)

• Enhancement of baryon density in "fireball"

– Comparable to \ life times :

V=1.3\23 fm/c

15 fm/c

Dense matterFreeze-outFirst chance

collisionsAu + Au @ 1 AGeV

UrQMD:J. Phys. G: Nucl. Part. Phys. 25 (1999) 1859–

T60=80MeV


Meson Production at SIS/BEVALAC energies

● Production of vector mesons close to the threshold : SNN< STHR=2MN + m

(Ekinthresh = 1.92 GeV)

● co-operative process :NN N, N NN/

or N, N N*() N/

● production confined to high density phase

– Low production rates: One vector meson decaying into lepton pair per 10 Million reactions !

● Investigation of NN and N collisions is prerequisite for HI !

mT=( m2 + p2 T)1/2

W. Cassing, E.L. Bratkovskaya / Phys.Rep. 308 (1999) 65Ð233Yield

(arb. units)


DLS puzzle

Data: R.J. Porter et al.: PRL 79(97)1229

Model: E.L. Bratkovskaya et al.: NP A634(98)168, BUU, vacuum spectral function

DLS puzzle:

• Not explained even by in-medium mass shifts and broadening

• Shape of enhancement consistent with e+e- but cross section to low (TAPS)

Strong dilepton enhancement over hadronic cocktails comparable to top SPS energies (CERES)!


TAPSTAPS – Two Arms Photon Spectrometer

Electromagnetic calorimeter : m : m =(2E =(2E11EE22(1-cos((1-cos(1122))))1/21/2

C+C @ 1 AGeV

• small acceptance (mid rapidity)

Acc~10-3


Pi0 and eta from TAPSCross sections measured and extrapolated

to full solid assuming isotropic thermal source at rest in NN CM frame

Converted into pair yields (Cocktail):compared to data


The DiLepton Spectrometer at LBL

• 1988-1993 at Bevalac• 2 Arm-Spectrometer

– Minimum opening angle: 40°– Each arm: 40° in Φ, 7.5° in Θ– Trigger on electron-pairs – Mass resolution– 30-40% systematical error

• pp/pd, Ca+Ca• 1993 C+C 1.04 AGeV,

– Mid-Rapidity: 0.69– Acceptance– Statistics

e+e-

Pair YieldsComb Net pairs Syst. Error

4760 1919 2841 ± 82 ±30%


HADESSide View

START

FWFW

Acceptance: Full azimuth, polar angles 18o - 85o

Pair acceptance 0.35

Particle identification: RICH: CsI solid photo cathode, C4F10 radiator, TOF: 384 scintillator rodsTOFino: 24 scintillator paddles MUL limitation, high granularity RPC from 2008Pre-Shower: 18 pad chambers & lead converters)2' Level single leg electron trigger (Me>=1)e+e->=92%, evt. reduction : 20(pp) -3(ArKCl)

Momentum measurementMagnet: B = 0.36 Tm + MDC: 24 midi drift chambers, single-cell resolution 140 mm, 2005-6: set-up completed (MDC IV)


Phase coverage: HADES vs DLS

mid-rapidity mid-rapidity

• Thermal π0 and η events: HADES and DLS acceptance• HADES acceptance larger but for low masses (M<0.14 GeV/c2) part of phase space covered by DLS not covered by HADES !

Red dashed lines: constant pair momenta in steps of 100 MeV/c


Hadron ID• momentum vs velocity (momentum vs velocity (ββ) measurement: ) measurement:

181800 < <<45<4500 TOFINO TOFINO 454500 < <<85<8500 TOF TOF~450 ps~450 ps ~120 ps~120 ps


Electron ID

hadron blind had t < lep

C4F10 : t = 18.3

pGeV/c

e+

e-0

~ 15.20

Dalitz decay

20% electrons: pi0 Dalitz

e+

e- ~ 2.20

Conversion

70% : electrons conv.

One ring:Two rings (if Ị needs high res. MDC!


Electron ID• Spatial correlations

– RICH rings ↔ MDC tracks– MDC tracks ↔ TOF and PreShower

hits• PID : e+, e-

– β vs momentum correlation– PreShower condition

DATA

C + C @ 2AGeV

• hadron admixture < 3% at 1000 MeV/c

e-

e+

velocity vs. momentum


Combinatorial BackgroundCombinatorial background:

M < 150 MeV/c2 - sLS

sLs = 2

(checked with MC for HADES )

M > 150 MeV/c2 - mixedOpposite Sign (mOS)

MMeeee > 150MeV/c > 150MeV/c22

MMeeee > 150MeV/c > 150MeV/c22

eeee NN

CC 2AGeV

• Normalization done between 150-550 MeV/c2 Mee

• sLS and mOS background show same behavior for Mee > 150 MeV/c2

• For Mee < 150 MeV/c2

deviations due to correlated background eeX

Pair cut: opening angle > 9º


C+C @ 2 AGeV

~ 2300023000 signal pairs for full Mee range

~ 30003000 signal pairs for full Mee > 150 MeV/c2

(M(Meeee) ~9 % @ M) ~9 % @ Meeee~0.8 GeV/c~0.8 GeV/c22

2002 set-up with 2 inner MDC only

•Spectra before efficiency correctionnormalizated to ½(+ + -) yield


C+C pion production

mmt t spectraspectra

C +C @ 2AGeVC +C @ 2AGeVanisotropiesanisotropies )(cos1

)(cos2

2 cmCM

Ad

dN

2 AGeV 1 AGeV

HADESN< -+>/Apart = 0.137 ± 0.015

Apart =8.4 <b>=3.2 fm

N< -+>/Apart = 0.069 ± 0.008

Apart =8.6 <b>=3.0 fm

Yields extrapolated (LVL1) to 4Yields extrapolated (LVL1) to 4::

SYS error =11%


Comparison to Physics GeneratorsExperimentExperiment

RawRaw DataData

HADESHADES

AnalysisAnalysis

PhysicsPhysics

PairPair SpectraSpectra

Event GeneratorEvent Generator

EfficiencyEfficiency CorrectionCorrection

dN/dMdN/dM

AcceptanceAcceptance FilterFilter

time consuming but done to cross-time consuming but done to cross-check eff.corrections and acc. filterscheck eff.corrections and acc. filters

final comparison: only final comparison: only in HADES acceptancein HADES acceptance

comparison incl. comparison incl. efficiency factors efficiency factors

AccAcc±± (p, (p, , , ΦΦ))


Acceptance and Efficiency matrices

acc

rec

N

NpEff ),,(

• pair production and decay is described by 6 dof (3 production and 3 decay)p (0 – 2 GeV/c), Φ (0o – 60o), Θ (0o – 90o)

and similarlyand similarly

all

acc

N

NpAcc ),,(

acc

rec

N

NpEff ),,(

CC @ 2 AGeVCC @ 2 AGeV


Coctail A (long lived mesonic components)

0 thermal source, anisotropic angular distribution according to measured +-

isotropic mT scaling

18 %18 % 21 %21 %

systematic errors:systematic errors:15 % - efficiency correction 10 % - combinatorial background11 % - 0 normalization

• Cocktail A: 0 + η + ω

• Cocktail B: Cocktail A + Δ(Ne+e-) + ρ

short lived component

A. Agakichiev Phys.Rev. Lett 98(2007) 052302

C+C @ 2 AGeV


C+C @ 1 AGeV

Cocktail A: 0 + η + ω

= “long-lived” components only

Large excess yield

•Good agreement in π0 region•Underestimates the data for Mee > 0.15 GeV/c2

Cocktail B: Cocktail A + Δ + ρ

Contribution from short lived resonances (ρ, Δ, N*)

•Reduced discrepancy for Mee>0.15 GeV/c2

•Cocktail B underestimates dataPR

ELIMINARY


HADES vs DLS direct comparison

Fit:

Mee < 0.15 GeV/c² 0.15 GeV/c² ≤ Mee ≤ 0.55 GeV/c²

PRELIMINARY

mapping of the measured HADES pair yield onto the DLS acceptance, defined inthe 3-d space spanned by the pair variables Mee, Pee and the rapidity Yee⊥HADES coverage is larger and almost fully contains the DLS acceptance

1) projecting out 2d slices (p vs.y) of the efficiency-and ⊥acceptance-corrected pair yield for different mass bins2) fitting a reasonable 2d function 3) using the resulting fits to extrapolate in 3d phase space, mass slice by mass slice, to small p and large y ⊥


HADES vs DLS direct comparison

DLS: Porter et al., PRL 79 (1997) 1229HADES: Agakishiev et al., PLB 663 (2008) 43

HADES fully confirmed highly controversialDLS findings in C+C !


Measuring the reference

●shape of the mass spectra changes dramaticallywhen going from p+p to n+p interactions

● 0.15 <m<0.35 GeV/c2: n+p yield ~10x p+p yield one would expect 2x if were the only relevant source

● the tail at high invariant mass extends much further pn / pp ~ 100 at 0.5 GeV/c2.

●OBE effective models reproduce p+p but not (yet) p+n

HADES, PLB 690 (2010)


Adding higher order diagrams


Measuring the reference

+

Mass distribution (-subtracted) similar at 1 and 2 AGeV (except only high mass)

½ (pp + pn) = C+C

DLS puzzle: has its origin in a insufficient treatment of radiation from elementary N+N collisions.No compelling evidence for in-medium effects in C+C


Putting the puzzle togetherRecent transport calculations:enhanced NN bremsstrahlung , in line with recent OBE calculationsHSD: Bratkovskaya et al. NPA 807214 (2008)

The DLS puzzle seems to be reduced to an understanting of the elementarycontributions to NN reactions.


Heavier Systems

HADES, Nuclear Physics A 830 (2009)

Both for Ar+KCl and Au+Au:● A+A > ½ (pp+pn) = C+C● Strong excess of dielectron pairs● No mere superposition of elementary pp and np collisions● Almost exponential spectrum up tovector meson region


Comparison with transport modelCoarse-graining UrQMD1) Average over many UrQMD transport events2) Determine local temperature & density in a grid of space-time cells3) Use Hadronic Many Body Theory (HMBT) & spectral functions to compute EM emission rates4) Sum up all cells → thermal dilepton radiation5) Add freeze-out contributions → non-thermal part(Endres, van Hees, Weil & Bleicher, PRC 92 (2015) 014911)


The meson in-medium spectral function


/ from p+Nb vs p+p @ 3.5 GeV


HMBT vs. chiral symmetry restoration

goethe university frankfurt gsi helmholtzzentrum für...

Documents