CCompensating ompensating

HHadronadron CCalorimetryalorimetryWhy do we need it?

How it may work?Does it really work?

If not, can we survive?Aldo Penzo, INFN-Trieste

FERMILAB - Research Techniques Seminar – Aug. 9, 2007

What kind of talk?

•Not a general “review”…• Rather an account of

– “episodes” in Calorimetry R&D– examples of “compensation”

methods• Therefore…

– only partial coverage of the topic– (necessarily) subjective– (hopefully) not too biased

• Purpose: outline the changing scope of what is called

“compensation”

Joule's apparatus for measuring the mechanical equivalent of heat in which the "work" of a falling weight is converted into the "heat" of agitation in the water.

Calorimeter used by J. Joule for his (final) determination of the mechanical equivalent of heat in which he obtained a value of 4.1538 Joules/cal. (Museum of the Department of Natural Philosophy at Glasgow University, Scotland  (1876)

No need of high-tech for a good

result...

...just go for it! Take the best data you

can...

1991 - Reconfigurable-stack calorimeter

“Hanging File”: Scintillator-Absorber plates

Shower development and energy

measurement

What’s going on in a developing shower?

Everything that can be found in standard textbooks on

Radiation Interactions in Matter (f.i. Landau-Lifschitz, Bruno Rossi, etc.)

• Photons/electrons undergo EM interactions and produce

• EM showers

Hadrons interact strongly and produce hadronic cascades

• copiously produced o generate pairs, originating EM showers inside hadronic cascades

• 2 scales for hadronic cascade development:– abs for strongly interacting part– Xo for the EM part

hadron

2 scales of shower development

0.1

1

10

100

0 10 20 30 40 50 60 70 80 90 100

X0

I

X0,

λ I [c

m]

Z

Development of hadron cascade

Large fluctuations due to o production

Dominance of low energy particles

Lateral cascade distribution

Lead-scintillating fibrecalorimeter

The EM Componentmore concentrated on the central part of the shower: EM core

Hadronic vs EM response

For instance in lead (Pb):Nuclear break-up (invisible) energy: 42%Ionization energy: 43%Slow neutrons (E ~ 1 MeV): 12%Low energy ’s (Eγ ~ 1 MeV): 3%

Not all hadronic energy is “visible”: Lost nuclear binding energy neutrino energy Slow neutrons, …

Hadronic shower reduced vs EM (and much broader)

e/h = 1/

Interim outline High precision hadron calorimeters should have equal

response to electromagnetic and strongly interacting particles (compensation condition e/h =1) in showers generated by incoming hadrons, in order to achieve:

• linear response in energy to hadrons, • gaussian energy distribution for mono-energetic

hadrons, • electron-to-pion ratio close to unity, constant with

energy,• relative energy resolution (dE/E), improving as

sqrt(1/E). This is of prime relevance for the measurement of jets,

involving various particles of different energies, with a substantial fraction of neutral pions..

In practice…

• Significant episodes of the straggle for compensation in hadron calorimetry will be summarized and different methods will be compared.

• Recent progress aiming to differentiate and separately measure the shower products (interacting electromagnetically or strongly) will be discussed and technological solutions outlined

Compensation methods(mainly sampling calorimeters)

• Intrinsic compensation: • Recover part of the “invisible energy”• Decrease the electromagnetic contribution (often using composite passive materials)• Off-line compensation: • Weighting methods• Multiple shower measurements (with 2 or more active media, selective to EM,etc)

High Z material as absorber• Full compensation can be achieved with high Z

material;• large part of EM components will be deposited in

absorber decreasing the EM energy deposition in active medium

• Tuning the thickness of the absorber and active layer

• High - absorber that can partially recover the invisible hadronic energy via nuclear and collisions processes.

238U as passive, scintillator as active media.

• 238U: Absorber with high Z decreases EM response;

• Slow neutrons induce fission in the 238U, that compensates losses due to “invisible” energy

• Slow neutrons can be captured in the nucleus of 238U which emits a low energy γ’s and can further recover the “invisible” energy

• Scintillator: Slow neutrons loose their kinetic energy via elastic collisions with hydrogen in the scintillator

(LAr - Fabjan, Willis, 1977)

Zeus U-scintillatorCalorimeter

EE

E %35)(

EE

E %18)(

Had

EM

Response for e/h = 1

If e/h = 1 than: Hadron response linear Energy distribution “Poisson”

E

cb

E

a

E

E

)(

Statistical fluctuations

Constant term (calibration, non-linearity, etc

Noise, etc

Tuning Pb thickness for e/ =1

So much about scintillator calorimeters:

I will revisit later on…

Now let’s turn to Si detectors in calorimetry:it so happened that such systems gave preciseand interesting data on very detailed compensation effects…

1983: Si-W EM prototype

24 Xo’s tungsten; 12 Si detectors (5x5cm)

EE

E %18)(

EM

SICAPO: Silicon Calorimeter

Another hanging file reconfigurable calorimeter prototype

Large scale SICAPO test calorimeter

The calorimeter vessel could accommodate:• the Si mosaic detector planes and• a number of absorbers between the detector planes• The passive plates were W, Pb (High Z)• as well as Fe, …,G10 (Low Z) (arranged in any desired configuration)

The readout of this calorimeter consisted of 30 mosaic planes (of about 1600 cm2 active area) each made of 400 silicon detectors. A detector had an area of about 4 cm2, is 400 m thick, and operated at full depletion.

Support plate for Si detectors

Assembling the Si detectors

Tuning of absorbers for SICAPO

The passive plates’ configuration could be arranged in order to achieve:

• local hardening effect (introducing thin low Z sheets near the detector’s planes)• filtering effect (by using combinations of High and Low Z absorbers)

Hardening

• G10 plates inserted close to the detector planes….

Filtering

PbFe–Si–PbFe configuration as a function of the thickness of Pb in the absorber (the overall absorber thickness, including the Fe plates is 23 mm).

Resolution for U-scint. Cal.

EE

E %35)(

Evidence for Compensation in a Si Hadron CalorimeterE. Borchi, M. Bosetti, C. Leroy, S. Pensotti, A. Penzo, P.G. Rancoita, M. Rattagi, G. TerziIEEE TRANS. ON NUCLEAR SCIENCE, 40 (1993)

Resolution vs sampling step

Linearity for Scintillator Cal.

EnergyResolution

(for Scint. Cal.)

Offspring: PAMELA

EE

E %18)(

Another example: in this case highly non-compensating

• In a quest for rad-hard materials as active parts for calorimetry, quartz fibers

• Only Cerenkov light, only (low energy) e+/-

The HF calorimeter

• Steel absorber with embedded fused-silica-core optical fibers where Cherenkov radiation forms the basis of signal generation.

• Thus, the detector is essentially sensitive only to the electromagnetic shower core and is highly non-compensating (e/h 5).

• Above Cherenkov threshold (E = 190 keV for electrons) they generate Cherenkov light, thereby rendering the calorimeter mostly sensitive to the electromagnetic component of showers.

CMS – HF: A full-scale technological benchmark

Quartz Fiber Calorimeter

~ 1000 km quartz fibers 1 HF weights ~ 250 tons

PMT readout (magn. field ~ 0 )

Co60 source calibration ≤ 5%

/e values for SPACAL and HF

Resolution of HF• The electromagnetic energy resolution is

dominated by photoelectron statistics and can be expressed in the customary form. The stochastic term a = 198% and the constant term b = 9%.

• The hadronic energy resolution is largely determined by the fluctuations in the neutral pion production in showers, and when it is expressed

as in the electromagnetic case, a = 280% and

b = 11%.

- Beam test results from a fine-sampling quartz fiber calorimeter for electron, photon and hadron detection - N. Akchurin et al.- Nucl.Instrum.Meth.A399:202-226,1997 - Test beam results of CMS quartz fibre calorimeter prototype and simulation of response to high-energy hadron jets - N. Akchurin et al. - Nucl.Instrum.Meth.A409:593-597,1998 -On the differences between high-energy proton and pion showers and their signals in a non-compensating calorimeter - N. Akchurin et al. - Nucl.Instrum.Meth.A408:380-396,1998

CMS NOTE 2006/044 - Design, Performance, and Calibration of CMS Forward Calorimeter Wedges, G. Baiatian et al.

The challenge….Mockett 1983 SLAC Summer Institute A technique is needed that is sensitive to the relative fraction of electromagnetic energy and hadronic energy deposited by the shower. This could be done hypothetically if the energy were sampled by two media: one which was sensitive to the beta equals one electrons and another which was sensitive to both the electrons and other charged particles. For example one sampler could be lucite which is sensitive only to the fast particles, while the other sampler could be scintillator.

See also Erik Ramberg et al. , Dave Winn et al. , …

….and the DREAM…

Main theme: multiple measurements of every shower to suppress fluctuations

•Spatial changes in

density of local

energy deposit

•Fluctuations in EM

fraction of total

shower energy

•Binding energy losses

from nuclear break-up

•fine spatial sampling

with SciFi every 2mm

•clear fibers measuring

only EM component of

shower via Cherenkov

light from electrons

(Eth = 0.25 MeV)

•measure MeV neutron

component of shower.

•Like SPACAL

•Like HF

• Triple Readout

DREAM = SPACAL + HF

1) Muon detection with a dual-readout calorimeter. N. Akchurin, K. Carrell, J. Hauptman, H. Kim, H.P. Paar, A. Penzo, R. Thomas, R. Wigmans - Nucl.Instrum.Meth.A533:305-321,2004 2) Electron detection with a dual-readout calorimeter. N. Akchurin, K. Carrell, H. Kim, R. Thomas, R. Wigmans, J. Hauptman, H.P. Paar, A. Penzo - Nucl.Instrum.Meth.A536:29-51,2005 3) Hadron and jet detection with a dual-readout calorimeter. N. Akchurin, K. Carrell, J. Hauptman, H. Kim, H.P. Paar, A. Penzo, R. Thomas, R. Wigmans - Nucl.Instrum.Meth.A537:537-561,2005 4) Separation of scintillation and Cherenkov light in an optical calorimeter. N. Akchurin, O. Atramentov, K. Carrell, K.Z. Gumus, J. Hauptman, H. Kim, H.P. Paar, A. Penzo, R. Wigmans - Nucl.Instrum.Meth.A550:185-200,20055) Comparison of High-Energy Electromagnetic Shower Profiles Measured with Scintillation and Cherenkov Light. N. Akchurin, K. Carrell, J. Hauptman, H. Kim, A. Penzo, R. Thomas and R. Wigmans - Nucl.Instrum.Meth.A548:336-354,2005

DREAM Published Results (www.phys.ttu.edu/dream)

DREAM [Dual REAdout Module] prototype is 1.5 ton heavy

Cell

[basic element of detector]

2m long extruded copper rod, [4 mm x 4 mm]; 2.5mm hole contains 7 fibers:3 scintillator & 4 quartz(or acrylic plastic).

In total, 5580 copper rods (1130Kg) and 90 km optical fibers.

Composition (volume) Cu: S : Q : air = 69.3 : 9.4 :12.6 : 8.7 (%)

Effective Rad. length (X0)=20.1mm;Moliere radius(rM)=20.35mm Nuclear Inter. length ( int )=200mm;10 int depth Cu.

Filling fraction = 31.7%; Sampling fraction = 2.1%

(S, Q fibers 0.8 mm )

Fig. a : fiber bundles for read-out PMT; 38 bundles of fibersFig b : front face of detector with rear end illuminated: shows 3 rings of honey-comb hexagonal structure..

Tower : readout unit Hexagonal shape with 270 cells (Fig. b); Readout 2 types of fibers to PMTs (PMT: Hamamatsu R580) (Fig. a)

Detector : 3 groups of towers (Fig. b) center(1), inner(6) & outer(12) rings;Signals of 19 towers routed to 38 PMT

Effective radius 162 mm (0.8 int, 8 rM )

Test Beam Results

c)

d)

a), b) energy distributions from scintillating

and Cerenkov fibers for 100GeV single -

asymmetric, broad, smaller signal than e- typical features of non-comp. calorimeter.c) energy resolution (%) vs beam energyd) Scintillation signal response vs energy

After (Q+S)/E correction, the signal distributions are described well by gaussian distribution and energy resolution was dramatically improved.( 12.3% resolution became 2.6% for 100GeV beam). ( Fig. a & b)

Energy resolution as a function of beam energy(Fig. c) are well described by

E

a

bE1/2

( b is related to sampling non-uniformity depending on impact point of the beam. )

c)

DREAM calibrated with 40 GeV e- into center of each tower,

recover linear hadronic response up to 300 GeV for - and “jets”

b. Method using the directionality of Cerenkov light

•Measure the light from both ends of fibers (F and B).•B/F = ratio of light emitted in the backward and forward.•Clear difference was found in B/F ratio between Cerenkov and scintillation light.•Scintillation light is generated isotropically and only 20% difference in backward and forward signals.•Cerenkov light in forward direction is ~6 times larger than in backward direction.

B/F ratio for scintillator (a) and Cerenkov (b) signals generated by 80GeV electrons.

New issues and options

• (Light-Emitting) Active Media

Study Xstals, Cerenkov radiators,

neutron sensitive scintillators• (Photon-Sensing) Detectors

Develop SiPM (popular objective…

… need good technology partners) • (Time-Domain) Signal Processing

Fast Pulse Shape digitizer

Compensating Calorimetry might be

the art of compromise and balance…

…and elegance instead of brute force, like…

Appareil à deux globes de verre inventorié dans les archives de l'Ecole polytechnique comme " appareil de Gay-Lussac " dans la catégorie des " appareils ayant trait à la chaleur "......it was never used for a real measurement....

Maybe go back to really old stuff?

Provando e riprovando….