PHENIX Overview Mickey Chiu Brookhaven National Lab

PHENIX Overview

Mickey Chiu

Brookhaven National Lab

VNI Simulations: Geiger, Longacre, Srivastava, nucl-th/9806102

In the Beginning…

•We want to create a quark gluon plasma and measure it’s properties

•System is about 10-100 fm big (10-14-10-13 m)

•System lasts for ~10-22 s (t ~ L/c where L ~ 10 fm)

•Collisions occur at rates of 10,000 (Au+Au) to 10,000,000 (p+p) times per second•Requires trigger and fast DAQ

•Produces hundreds of particles into your detector•Requires fine segmentation

•Like smashing two swiss watches at incredible speeds and looking at the broken up insides (from California) to figure out how the watch worked

…and a rough idea how things evolve

Pre-equilibrium

Thermalization

QGP phase

Mixed phase

Hadronization (Freeze-out) + Expansion

p K

e

Space (z)

Tim

e

Au Au

etc.. jet

e

“Passage of particles through matter”

http://pdg.lbl.gov

•Bethe-Bloch equations•Governs ionization loss of particles as they traverse through matter•Depends on Z

•Happens on length scales of many Angstroms (10-10 m)

•Strong interaction ~ fermi (10-15 m)

p

e

~10 MeV/cm 1 GeV/m 10-10 J

1.6x10-19 J/eV

Measuring Momentum

•We measure the momentum of a charged particle by determining its trajectory in a known magnetic field.

•Simplest case: constant magnetic field and pB trajectory is a circle with p=0.3Br (GeV/c, T, m)

•We measure the trajectory of the charged particle by measuring its coordinates

•(x, y, z or r, z, , or r, , ) at several points in space.•Simplest case: determine radius of circle with 3 points

•We measure coordinates in space using one or more of the following devices:

•Wire Chamber: (~1-2 mm) •Drift Chamber (or TPC): (~50-250um) •Silicon detector: (~5-50 um)

Cherenkov radiation

• Cherenkov radiation occurs with following conditions– There is a medium with index of n(>1), not

vacuum– The particle traverse the medium with a

speed exceeding the speed of light (c/n)• Cherenkov radiation was discovered by

P. A. Cherenkov in 1934– Interpreted by I. M. Frank and I. E. Tamm

using classical electromagnetic dynamics– All of three won Nobel prize of 1958!

• Electromagnetic radiation always occur!– Radiation is occurred whenever dielectrics

recovers from polarized state produced by the electric field of a moving particle.

– However, the radiation will be erased by a coherence with a radiation from different position and phase.

– If the particle speeds more than c/n , the radiation will not be vanished in a certain direction (cos(c)=(n)-1) like a shockwave, and thus Cherenkov photons can come out

Electron (Positrons) and Photons

•At high energies, electrons bremstrahlung and photons pair produce•We’ll use this to measure high energy particles in a “electromagnetic calorimeter”•Use strong interactions to measure in a “hadronic” calorimeter

So What Would You Build?

By the way, who was right?

dN/dy ~ 220-230 per chargedNK+/dy ~ 40dNp/dy ~ 28Net baryon density at mid-y small, but not 0 B small

PHENIX as ultimately built

The “Global” DetectorsBeam-Beam Counters (BBC) Zero-Degree Calorimeters (ZDC)

Forward Calorimeter (FCAL)Multiplicity and Vertex Detector (MVD)

Where is PHENIX BBC?

NorthSouth

144.35 cm

⊿η = 3.1 ~ 4.0

⊿φ = 2π

Hardware Component

BBC has 64 elements for North and South arm.

Each element is assembled byQuartz Cherenkov radiator(th=.7)

and meshed dynode PMT.

ReadoutAnalog signal (PMT output) from BBC

Digitized at Front End Module (FEM)• TDC0 (for ADC gate)• TDC1 (for Local Level1)• ADC

Accumulated in AMU

Data Collection Module (DCM)

Event Builder

PHENIX Raw Data Formatother subsystem data

BBC Local Level 1

Global Level 1 decision

Minimum bias at Run4 (Au+Au)( BBCN>=2 & BBCS>=2

& BBCZ<36 [cm] )&

( ZDCN & ZDCS )

BBLL1Selected(ZDC&BBLL1)

ZDC triggered

Slewing Correction

)log()( ADCcADC

baxf

ADC [ch] ADC [ch]

before correction

after correction

Slewing effect was corrected by this empirical function

(Reference time) – (PMT hit time) of typical PMT

a, b, c : constant

ADC : after pedestal subtraction

Intrinsic time resolution : 40±5ps

details in PHENIX Technical Note 393

BBLL1Selected

(ZDC&BBLL1)

ZDC triggered

Z-Vertex and Time zero

TN/S : average hit time, c : light velocity, L : 144.35 cm

2

/22

cLTT

cTT

NS

NS

• Z-Vertex

• Time zero

BBC North

BBC South LL

Vertex position TNTS

Resolution at RUN2 (Au+Au)

222

222

222

PCZDCPCZDC

ZDCBBCZDCBBC

PCBBCPCBBC

Time Zero : 20 [ps]

Z-Vertex : 0.6 [cm]

( at Run2 )

BBCZ - PCZ BBCZ - ZDCZ PCZ - ZDCZ

What does BBC detect?

• 50% of external track was estimated compared to all injected particles using HIJING Au+Au 130GeV events. inner ring = 43%middle ring = 52% outer ring = 57%• Main background source is beam pipeBeryllium (thickness 1.02 [mm]) : < 75 [cm]Stainless Steal (thickness 1.24 [mm]) : < 200 [cm]

(a) : Internal track coming from collision

(b) : External track not coming from collision

beam pipe

Z-direction

R-d

irec

tion

Collision point (a)

(b)

BBC

• inner ring

• middle ring

• outer ring

RING ID

BBC

Dead PMT in South

Reaction Plane by BBC

-- no correction -- ring-by-ring gain correction-- average subtraction (shift correction)-- fluttening

azimuthal angle Φ

Ψ (BBC North)

Ψ (

BB

C S

ou

th)

Corrected Reaction plane correlation

details in PHENIX Analysis Note 151 : S. Esumi et al.

(Univ. of Tsukuba)

Ignore 4 PMTTo keep hexagonal

symmetry

Background SourceThis is all vertex position of each secondary particles that are injected to BBC. Fig.10 is electron or positron at each vertex position. Fig.11 is charged pion. Almost of electron and positron are produced at beam pipe of Stainless steal.

electron, positron at each position (Fig. 10)

π+, π- at each position (Fig. 11)

Beam Pipe

Stainless steal

BBC

inner edge of central magnet inner edge of central magnet

(Fig. 12)

MVD

Beryllium pipe

BBC

Location, Location, Location

•There is a lot of physics at forward rapidities•In a collider, you need to have a DX magnet to steer bunches so they collide

•Spatial Distribution of Charged Particles shown below•Large Separation = Easy Timing = Very Clean Trigger against Beam Gas and Beam Scrape

Hadronic Interaction:Au-Au --> X 6.8 barns-:AuAu --> AuAu + e+e- 33 kbarnsAuAu --> AuAu + 2(e+e-) 680 barnsAuAu --> AuAu + 3(e+e-) 50 barns-N: L(-N )=1029 cm-2s-1 2<E<300GeVAuAu --> Au+Au* 92 barns X+neutronsAuAu --> Au*+Au* 3.670.26 barns X+neutrons Y+neutrons

Collider Processes•You’re probably familiar with the “Hadronic Interactions”•When colliding large nuclei, the Z creates a large photon flux

•ZDC exploits this for a luminosity measurement•Also interesting physics (UCP program, AN063)

Hadronic Interaction

Peripheral Interaction

ZDC Design

ZDC Calorimeter construction:•Tungsten absorber/ fiber (C)sampling•2 Lint/module, 3 modules total•C sampling filters shower secondaries•Uniform response vs. impact point

e,beam

NIM A 470:488-499,2001, nucl-ex/0008005Fiber response vs. angle

(deg) (deg)

Drift Chamber for PHENIX(basic information)

• Main purpose:– Precise measurement of the

charged particle’s momentum– Gives initial information for the

global tracking in PHENIX

• Acceptance:– 2 arms 90º in each – ±90 cm in Z– 0.7 units of

• Location:– Radial :2.02<R<2.48 m– Angular:

• West: -34º < º

• East : 125º < º

Drift Chamber design

• Multiwire jet-type drift chamber (~12800 readout channels)

• 6x80 (r - ) wire nets per arm

• Titanium alloy support frame with 20 C-shell openings (Keystones)

• Independent signal readout from both sides (North, South) DCH Frame

Wires

Keystone

DCH Operation Principles

• To reconstruct charged particle track DC samples a few points in space along the path of the particle. One such point is called a “HIT”

• Registration of one HIT is based on a few physical processes:– When charged particle transverse the gas volume of the DC it creates clusters of

primary ionization on its way– Electrons of primary ionization drift from the point of ionization to anode wires along

electric field lines– Electrons of primary ionization create avalanches in the vicinity of anode wires – Back drift of posistively charged ions generate measurable signal on anode wires which

is amplified, shaped and discriminated

• To register a HIT in the DC:– Carry out drift time measurements: Start - collision time measured by BBC; Stop - time

when signal appears on the anode wire– Drift time (t) can be tranformed into drift distance (x) if calibration curve is known x = x(t)– Working gas is chosen to have an uniform drift velocity in the active region linear xt

relation can be used x = Vdr · t

Wire net configuration

• 6 radial layers of nets (X1,U1,V1,X2,U2,V2)

• X nets – measure coordinate of the track

– 12 anode wires in each X net

• UV (stereo) nets – measure Z coordinate of the track

– 4 anode wires in each UV net

• Cathode nets separate anode nets (see figure)

• Total of 80 anode nets per arm evenly distributed in

Wire net configuration (II)

• Group of 4 anode-cathode

nets makes a keystone

• Stereo nets starts in one keystone (n) and ends in the neighbouring keystone e.g. (n+1) for U, (n-1) for V

• The tilt of UV nets along allows measurement of Z component of the track

Gas mixture choice

• 50% Ar - 50% C2H6 mixture is chosen for operation based on:– uniform drift velocity at E~1

kV/cm– High Gas Gain – Low diffusion coefficient

• In Year2 ~1.5% Ethanol was added to the mixture to improve HV holding of the nets

Drift field configuration

• Specific field configuration around anode wire called drift region is created by “field forming” wires:

– Cathode Wires – Create uniform drift field between anode and cathode

– Field Wires – Create high electric field strength near the anode wire

– Back Wires –Stop drift from one side of the anode wire

– Gate Wires –Also create high field near the anode wire, Localize the drift region width

Cathode

Back

Gate

Anode

Field

Drift Field Configuration (II)

• Here is what happens when the charged particle passes through the wire cell

• Note that only even wires collect charge due to the back wires that block the odd anode wires !

• Back wires solves left-right ambiguity problem

Tracking principles

Main assumptions:

• Track is straight in the detector region

and variables defined on the figure

•Use hough transform – calculate and for all possible combinations of hits and bin those values into hough array – 2D histogram on and

• Look for local maxima in hough array that surpass the threshold

Track Candidates

• The results of the hough transform are track candidates

• Several stages of hit association and track purging follows

• Finally we left with the following tracks

X1X2 and X-only tracking

• First we look for tracks with X1 and X2 hits

• Remaining unassociated hits goes into X1 only and X2 only tracking

• All the track candidates are being liked after this and Z information is being applied to them by PC1-UV-vertex tracking

Final results

Construction and assembling

• Mechanical design and production – PNPI (Russia)

• Front Electronics – SUNYSB• Wire net production, assembling -

PNPI,SUNYSB

PC1 on DCh

•Five planes: East PC1,3 & WEST PC1,2,3=90°, ||=0.35•80m2 MWPC, pixel cathode readout, •172800k readout channels, •1.2% 0 (PC1) with electronics on back

Performance: cosmics

Gain curvesRunning at 1-2*104

Performance: cosmics

efficiency

Position resolution in Z (cosmics)

chamber

Wire dist

(mm)

Z-resol.

(mm)

Perp res

(mm)

Rad.

Thickn.

PC1 8.4 1.7 2.5 1.2%

PC2 13.6 3.1 3.9 2.4%

PC3 16.0 3.6 4.6 2.4%

measured

Performance Au+Au central

What is RICH? Cerenkov photons from e+ or e- are detected by array of PMTs

mirror

Most hadrons do not emit Cerenkov light

Electrons emit Cerenkov photonsin RICH.

Central Magnet

ＲＩＣＨ

PMT arrayPMT array

•Primary electron ID device of PHENIX

•Hadron rejection at 104 level for single track

•Full acceptance for central arms•|y| < 0.35 ; = 90 degrees x 2

•Threshold gas Cherenkov•Using CO2( th ~ 35)•eID pt range : 0.2 ~ 4.9 GeV/c

•PMT array readout•pixel size ~ 1 degree x 1 degree

Why cherenkov radiation is Ring in RICH?

• Pure Optics of the photon!• Suppose that the charged particle

emerge to the radial direction• Mirror is spherical, so..

– Spherical mirror focuses all the photons emitted to a certain angle (c) with refer to the radial direction

– The focus point is at a distance of the half of the diameter of spherical mirror and at a angle of c with refer to the radial direction

• If a particle is not moving on the straight line from vertex point(0,0), the focused ring will be distorted!

112

222

2

22

_ 370,sin

eVcmcmr

zdE

cmr

zLN

eedc

eeelectronphoto

L : path length of particles in radiator

εc : collecting Cherenkov light efficiency

εd : quantum efficiency of photo electron conversion

Gas Vessel

The vessels are designed and fabricated at Florida State University.

• Two RICH detectors– One for each arm– Weight: 7250 kg /

arm– Gas volume: 40 m3 / arm– Radiator length: 0.9 - 1.5 m

• Mirror system – Radius : 403 cm– Surface area: 20 m2 / arm

• Photon detector– 2560 PMTs/arm

• Radiation length– CO2: 0.41%– Windows: 0.2%– Mirror panels: 0.53%– Mirror support: 1.0%– Total: 2.14%

RICH Mirror• Segmented spherical

mirror• Reflection surface

– Aluminum made• Mirror mounts are

adjusted so that all optical targets are within 0.25 mm of the designed spherical surface.– graphite fiber epoxy

• Mirror support structure– graphite fiber, Delrin

Rohacell foam core (1.25 cm thick)

Gel-coat (0.05 mm thick)

4 ply graphite-epoxy (0.7 mm thick)

Structure of the mirrorCompleted mirror array of the first RICH

Design of 3 points mirror mounts

Mirror panels are mounted by adjustable 3 point mounts on the frame bars

Mirror, mirror, mirror…

RICH (mirror alignment)

• After mirrors are installed, the RICH vessel is rotated up in the same orientation as on PHENIX carriage

• Positions of optical targets placed on mirror surface were surveyed with a computerized theodolite system (MANCAT).

BNL survey crew were measuring the optical targets on the mirror during the mirror alignment.

Alignment calibration (Once in RUN)•Accumulate all the hit PMTs around tracks for totally 56 mirrors

•14(mirrors)*2(side)*2(arm)•Adjust mirror positions so that ring centers match projected points of tracks

RICH in Operation! from Year-1 (I)

High PT electron candidate is seen!

Candidate selected with RICH, DC, and EMCal.PHENIX RUN 12280 SEQ 0014 EVENT 850

View from North Side

South Side

East Arm West Arm

RICH EMCal

RICH ring(6 PMT hit)

EMCal hit(2.5GeV)6 PMT RICH ring

2.55 GeV/c track2.5 GeV EMCal hitelectron candidate

EMCal

RICHPC1

DC

EMCal

RICHPC1

DC

TOF

TECPC3

How to identify electrons using RICH

• Starting from track• Calculate projected track point on

PMT according to the mirror alignment– Mirror alignment calibration is very

important!• Define Minimum and Maximum

radius of possible ring by the track– Minimum and Maximum are predicted

based on the simulation study• Check several quantities in the region

of (min)<r<(max)– How many number of PMT hit?– How many cherenkov photons hit?– How is the ring shape?– How is the timing of signal?

• If above checked quantities fulfilled conditions, tag it as electron!– ex. Number of PMT hits >=3– ex. Number of photo-electrons>9

Electrons in Ratio of Energy and Momentum

• Ratio of energy (E) and momentum (p) of associated track

• p and E are measured with DC and EMCal, respectively

• Condition required– PMT hits of more than

two in the ring of 3.4cm<r<8.4cm

– Good ring shape

• Peak is seen at E/p=1, which corresponds to electrons

• There is NO electron peak until Cherenkov hit is requiredGreen: Raw spectra Black: Cherenkov hit required

Blue: Estimated background Red: Background subtracted

0.3GeV<p<0.4GeV 0.6GeV<p<0.7GeV

0.8GeV<p<0.9GeV 1.1GeV<p<1.2GeV

Integration of RICH

• Gas vessel • Mirror• PMT and arrays• Electronics

Purpose of this sectionBe familiar with RICH components

Particle (n)

How do we use gas?

• When a particle travels in gas volume, – If particle speed exceeding the speed of light in gas (c/n)– Photons are radiated – Cherenkov lights!

GAS (radiator) If momentum is known,

we can IDENTIFY particle using Cherenkov light measurements.

Index:n

c

Threshold of light emissionn

Emission anglecosc =(n)-1

Light YieldProportional to L and sin2c

L

PMT and arrays

• Photon detection device– Hamamatsu H3171S

• Cathode Diameter: 25 mm• Tube Diameter: 29 mm• Cathode: Bialkali• Gain: > 107

• Operation Voltage: -1400 ~ -1800V• Rise Time: < 2.5ns• Transit Time Spread: < 750ps

• A Winstone cone – shaped conical mirror is attached to each

PMT – Entrance: 50 mm, Cut off: 30

• Supermodule (2x16 PMTs grouped)– 40 super-modules per one side– 4 sides * 40 * 2* 16 = 5120 PMTs

• 8 PMTs share the same HV channel pixel size 1 degree x 1 degree

The arrays are fabricated at SUNY

Summary of hardware

• Electron go thorough the RICH gas volume,– Cherenkov photons emitted– Photons are reflected by

mirror and focused on the PMT surface

– PMT detect photons

Cerenkovphotons from e+ or e- are detected by array of PMTs

mirror


Electrons emit Cerenkovphotonsin RICH.

Central Magnet

ＲＩＣＨ

PMT arrayPMT array

Cerenkovphotons from e+ or e- are detected by array of PMTs

mirror


Electrons emit Cerenkovphotonsin RICH.

Central Magnet

ＲＩＣＨ

PMT arrayPMT array

Identify electron from 0.02 to 4.9 GeV/c.

What can we do with RICH?

• Measure several kinds of proposed QGP signals. – Deconfinement

• J/ toe+e-– Chiral Symmetry Restoration

• Mass, width, branching ratio of to e+e-, K+K- with M < 5 MeV

– Thermal Radiation of Hot Gas• Prompt * to e+e-

– Strangeness and Charm Production• Production of , J/, D mesons• Single electron

Electron ID is essential for above measurements.

Low mass e+e- pair

• Chiral Symmetry Restoration– Mass, width of to e+e-

• Thermal Radiation of Hot Gas– Prompt * to e+e-

We did not yet get significant results in AuAu.NEED more statistics

]2 invriant mass [GeV/c-e+e0.6 0.7 0.8 0.9 1 1.1 1.20

20

40

60

80

100

120

140

160

180

200

220

Unlike-sign Pair mass HmassERT2Entries 9612Mean 0.6634RMS 0.5068

Could be

Recently, we have an indication of peaks in dAu data.

Y. Tsuchimoto

How to identify electrons?

• Main parameters in RICH– Distance between ring center and track

projection (disp)– Number of hit PMT in a region (n0)– Number of photo-electron in a region

(npe0)– Ring shape (chi2/npe0)

1. According to the track information, projection point is calculated.

2. Find PMT hits near projection point in the region (3.4 cm < r < 8.4 cm)Numbers come from position resolution of PMT

hits.

3. Above parameters are calculated using projection point and PMT hit information

How good it is.

• Energy / momentum– For electron, it should be 1.

• RICH cut suppress back ground by about factor 100.

Energy / Momentum

All Charged particle

Apply RICH cut

BGRICH works well.We can continue making progress in QGP physics using electron measurements

Extension of Charged Hadron PID Capability

Aerogel together with TOF can extend the PID capability < 10 GeV/c• Without TOF, no K-proton separation at pT < 5 GeV/c.

PHENIX-MRPC: Detail

• 6 gaps (230 micron).• Gas mixture: R134A (95%), Isobutane (5%) at 60 cc/min.• HV: 7.5 kV

Gas gap = 0.23 mm

Readout strip thickness = 0.5 mm

Total active area width = 11.2 cmHoneycomb width = 12 cm

Glass

Electrode

MylarPC boardReadout pad

HoneycombStandoff

Inner glass width = 11.2 cm

Outer glass width = 11.5 cm

PCB width = 13 cm

Outer glass = 1.1 mm

Inner glass = 0.55 mm

carbon tape = 0.9 mm Mylar thickness = 0.25 mm

PCB thickness = 1.5 mm

Honeycomb thickness = 9.5 mm

Strip width = 2.81 cmStrip interval = 0.3 cm

TOFW (Final Configuration & Position)

• Coverage Area of 4 m2

• 128 Chambers• 1024 Readout

Channels• Two Sectors

PID Upgrade Completed Track Correlation Methods

How does it (TOF) look like ?

- 960 plastic scintillators with 1920 PMT’s - locates at 5 meter from the vertex- Rapidity (-0.35~0.35), 45 degree in phi, ~1/3Sr

Phenix (East-Arm)

“panel” “slat”

“E1 sector”

“E0 sector”

Basics of measurements

222

222

12

22

1

12

22

10

tv

ttvx

tttt

lightlight

light

light

vtt

x

vltt

t

2

221

0

210 100ps

then 1.6cm

Precise TOF and Hit position

Double hit - Lose timing information must occupancy level low - Consistency between ratio of ADC and TDC diff. light

aa

vtt

x

lx

2

log2

21

2

1

What does it give now ? (First of all, ) Hadron

Identification

- Pion, Kaon, proton, and deuteron are clearly identified. - Overall, ~120ps (overall, in m2) time resolution is achieved

(anti-)deuteron

Kaon up to 2 GeV/c, Proton/pion(0) up to 4 GeV/c.Proton/pion(0) changes from “< 1.” to “>1” at ~ 2GeV/c.

What does it give now ? (e.g.#2) Particle Ratio

Before the summary, an idea of Phenix PID upgrade.

Continuous measurement from ~0.5 GeV/c up to ~ several GeV/cfor pi/K/p identification !

Principle of Timing measurement

v

xT 0 v

xLT

0

L

L – xx

PMT1 PMT2

(T0,x0)T1 T2

2

/)( 21 vLTT

vTT

221

TOF

Y position

Cartoon

T1, T2 : Timing measured by PMT1,2L : slat lengthv : light velocity in scintillator

Hadron identification

• Pion, Kaon, Proton and deuteron are clearly identified !– Overall ~ 120 ps (overall, in mass2) time resolution is achieved.

Concept

PID in high pT region• Cherenkov Radiation

Cherenkov Radiator

• Low refractive index• Best index with RICH(CO2) is n ~ 1.01.

Requirements- Refractive index : n~1.01 – Momentum threshold- Light yield : >10 p.e. – Resolving power- Uniformity of the light yield : Needed. – Easy handling - Occupancy in Au+Au collisions : <10% – S/N

Installation Purpose

To enhance the PID capability of PHENIX !!

TOF

Aerogel (+ TOF or RICH)

RICH

Momentum[GeV/c]

K

p

1 2 3 4 5 6 7

0.5 2.5

~10

4.2

Aerogel : (n=1.011.)TOF : 100 ps time resolutionRICH : CO2, (n = 1.00041)

5.53.7

What is Aerogel Counter ?? ( I ) Outline

Cherenkov Counter (non-ring-imaging type)• Cherenkov radiator is Silica Aerogel. (MATSUSHITA, SP-12M)

• Photon is detected by 2 PMTs. (HAMAMATSU, R6233)

• All inner surface is covered with DRP Reflector. (Goretex)

• Integration cube for uniformity of light yield. (Air)

PMT (3inch)

PMT (3inch)

Aerogel (index~1.011

)

Integration Cube (Air)

Reflector (Goretex)

（ 11x22x20 cm3

）

What is Aerogel Counter ?? ( III ) Where

RED: AerogelYELLOW: Integration sphereGREEN: PMT

vertex

particle track

z (beam

) dire

ction

azimuthal angle

160 segments

- 4.5m from vertex

- 4.0m along z direction

- 15 deg. in phi

Half of them in Run4

Mechanical Design ( I ) Silica Aerogel

Many many Aerogel tiles!! Y.Miake

Characteristic• Refractive index ~ 1.0114 +/- 0.0008

• Density ~ 40 mg/cm3

• Hydrophobic

• Long term stability ( KEK-Belle )

• Transparent for 10mm thickness - 64% @ 400nm, 88% @ 550nm

• Very fragile

- Silica aerogel with lowest refractive index commercially available !!

He is godfather of aerogel !!

Where is the EMCal?

PbSc

PbGl

East

West(arm=0)

East(arm=1)

0

1

2

3

0

1

2

3

Principles of Detection:

• Electrons and Photons interact electromagnetically (bremsstrahlung and pair production)

electromagnetic shower• Strongly interacting particles: hadronic shower, MIP

• Calorimeter measures energy, position, and TOF

• PbSc – sampling calorimeter, layers of lead and scintillator

• PbGl – homogeneous calorimeter, lead-glass Cherenkov radiator

• Light read by PMT

• Charged shower particles generate Cherenkov photons in the PbGl

• The Ch. Photons propagate with a wavelength dependent

attenuation to the PMT • Shower depth:

• Number of generated Cherenkov photons:

Principles: PbGl

tE

Eln

X

X

c

max 0

0ENCherenkov

Non-Linearity Effects

Absorption

Leakage

photons

electrons

Non-linearity effects have to be corrected

Principles: PbSc

Pb + Scintillator

generateshower

generatelight

•Absorber : Pb•Scintillator: 1.5 % PTP / 0.01 % POPOP

Light collection

Non-Linearity in the PbSc

finite light attenuation length in WS fiber

energy leakage

0

0

E

EEmeas

Two parts make one detector!

• PbSc: – Excels in timing– Better linearity in response– In principle, response to hadrons better understood

• PbGl: – Excels in energy measurement– Better granularity– Proven system (WA98)

• Two detectors = different systematics increase confidence level of physics results

The Leadglass Detector

PbGl-Sector

• 2 Sectors PbGl

• 1 PbGl Sector• 16x12 supermodules (SM)

• 1 PbGl SM• 6x4 towers • Separate reference system

• 1 FEM • Reads out 2x3 supermodules or 12x12 towers

• TF1 PbGlass • 51% Pb-Oxide • Wrapped with aluminized mylar foil•New developed HV-bases

PbGl Structure - Module

1 PbGl tower = 1 PbGl module

PbGl Structure II

PbGl Structure III

The Lead Scintillator

PbSc Structure• 1 Sector = 6x3 Supermodules (SM)• 1 PbSc SM = 12x12 towers • PbSc towers: 5.52 x 5.52 x 33 cm3 (18 X0)• 15552 blocks total

1 PbSc tower: • 66 sampling cells• 1.5 mm Pb, 4 mm Sc• Ganged together by penetrating wavelength shifting fibers for light collection• Readout: FEU115M phototubes

1 FEM reads out 1 Supermodule

PbSc Supermodule

Hadron ID with TOF

• Measure space points

• Deduce o Vertex locationo Good momentum resolutiono Decay lengthso Distance of Closest Approach (DCA)

What is measured?What is measured?

Introduction to Semiconductor Introduction to Semiconductor DetectorsDetectors

L

Primary vertex

Secondary vertex

Example:

L = (p/m) c

• By measuring the decay length L, and the momentum, p, the lifetime of the particle can be determined

• Need accuracy on both production and decay point

Also, by measuring the decay length, L, and knowing the lifetime of the particle, the momentum can be determined

• Decay lengths


0sKJ/ΨB

Introduction to Semiconductor Introduction to Semiconductor DetectorDetector

D± = 312 m, D0 = 123 m B± = 501 m, B0 = 460 m

• Distance of Closest Approach (DCA)

b = distance of closest approach of a reconstructed track to the true interaction point


beam

Beam b


DCA distribution for single simulated pions in 3<pT<4 GeV/c. Simulation is done with 200 micron pixel layers and 650 micron strip layer. The passive material is 1.0% per pixel layer and 2.75% per strip layer.

Expected DCA Expected DCA resolution of VTXresolution of VTX

Au+Auat 200 GeV

~ 40 m

Primary vertex Secondary

vertex

• Semiconductor with moderate bandgap (1.12 eV)

• Thermal energy = 1/40 evo Little cooling required

• Energy to create e/h pair (signal quanta) = 3.6 eV3.6 eV c.f Argon gas = 15 eV15 eVo High carrier yield o Better energy resolution and high signal

no gain stage required

Why silicon?Why silicon?


• High density and atomic number o Higher specific energy losso Thinner detectors o Reduced range of secondary particles

Better spatial resolution

• High carrier mobility Fast!o Less than 30 ns to collect entire signal

• Industrial fabrication technique available• Advanced simulation packages

o Processing developmentso Optimization of geometry o Limiting high voltage breakdown o Understanding radiation damage

Why silicon?Why silicon?


• Cost of Area coveredo Detector material could be cheap – Standard

Sio Most cost in readout channels

• Material budget o Radiation length can be significant

Tracking due to multiple scattering

• Radiation damageo Replace often or design very well

Disadvantages?Disadvantages?


P-N JunctionP-N Junction

One of the crucial keys to solid state electronics is the nature of the P-N junction. When p-type and n-type materials are placed in contact with each other, the junction behaves very differently than either type of material alone. Specifically, current will flow readily in one direction (forward bias), creating the basic diode.

Near the junction, electrons diffuse across to combine with holes, creating a "depletion region".


http://hyperphysics.phy-astr.gsu.edu/hbase/solids/dope.html#c4

http://hyperphysics.phy-astr.gsu.edu/hbase/solids/dope.html#c3

http://hyperphysics.phy-astr.gsu.edu/hbase/solids/diod.html#c3

http://hyperphysics.phy-astr.gsu.edu/hbase/solids/pnjun.html#c2

• Charge particles– Bethe-Bloch

• Not covered o Neutronso Gamma Rays

Compton scattering, pair production, etc…

Energy depositionEnergy deposition


Status of the VTX Project in PHENIXStatus of the VTX Project in PHENIX

Central Silicon Vertex Trackers

“VTX”

Pixel

Strippixel

VTX Layer R1 R2 R3 R4

Geometrical dimensions

R (cm) 2.5 5 10 14

z (cm) 21.8 21.8 31.8 38.2

Area (cm2) 280 560 1960 3400

Channel count Sensor sizeR z (cm2)

1.28 1.36(256 × 32 pixels)

3.43 × 6.36(384 × 2 strips)

Channel size 50 425 m2 80 m 3 cm(effective 80 1000 m2)

Sensors/ladder 4 4 5 6

Ladders 10 20 18 26

Sensors 160 320 90 156

Readout chips 160 320 1080 1872

Readout channels 1,310,720 2,621,440 138,240 239,616

Radiation length(X/X0)

Sensor 0.22% 0.67 %

Readout 0.16% 0.64 %

Bus 0.28%

Ladder & cooling 0.78% 0.78 %

Total 1.44% 2.1 %

Pixel detectorPixel detector Strip detectorStrip detectorBarrel VTX ParametersBarrel VTX Parameters

BEAM

Strip

Pixel

Layer radius Detector Occupancy in Central Au+Au collision

Layer 1 2.5 cm Pixel 0.53 %

Layer 2 5.0 cm Pixel 0.16%

Layer 3 10.0 cm

Strip 4.5 % (x-strip) 4.7 % (u-strip)

Layer 4 14.0 cm

Strip 2.5 % (x-strip) 2.7 % (u-strip)

Precursors to PHENIX / 12 - the verdict in the 2nd round

(Not too much time spent on mincing the words…)

Reject all three because of major deficiencies

Emphasis should be on QGP photons and electrons (implicit: STAR already does hadrons)

Get together (lead by Sam Aronson), and design one single detector

Maximum cost $30M

Documents

PHENIX Overview Mickey Chiu Brookhaven National Lab