seminar desy zeuthen 20.01.04alexander oh, cern particle detection with cvd diamond alexander oh,...

Seminar DESY Zeuthen 20.01.04 Alexander Oh, CERN

Particle Detection with CVD Diamond

Alexander Oh, CERN

Recent Developments and Applications


Outline

• Introduction

• Material Studies

• Particle Detector Prototypes

• Applications in HEP

• Summary


Sources• RD42 Collaboration• University Bonn• BaBar, SLAC• CMS, CERN


W. Adam1, E. Berdermann2, P. Bergonzo3, W. deBoer21, F. Bogani4, E. Borchi5, A. Brambilla3,M. Bruzzi5, C. Colledani6, J. Conway7, P. D'Angelo8,W. Dabrowski9, P. Delpierre10, W. Dulinski6,J. Doroshenko7, B. van Eijk12, A. Fallou10, P. Fischer20,F. Fizzotti13, C. Furetta8, K.K. Gan14, N. Ghodbane11,E. Grigoriev21, G. Hallewell10, S. Han14, F. Hartjes12,J. Hrubec1, D. Husson6, H. Kagan14;*, J. Kaplon15,R. Kass14, M. Keil20, K.T. Knoepfle16, T. Koeth7,M. Krammer1, A. Logiudice13, R. Lu13, L. Mac Lynne7,C. Manfredotti13, D. Meier15, D. Menichelli5,S. Meuser20, M. Mishina17, L. Moroni8, J. Noomen12,A. Oh15, M. Pernicka1, L. Perera7, R. Potenza22,J.L. Riester6, S. Roe15, A. Rudge15, S. Sala8,M. Sampietro18, S. Schnetzer7, S. Sciortino5, H. Stelzer2,R. Stone7, C. Sutera22, W. Trischuk19, D. Tromson3,C. Tuve22, B. Vincenzo22, P. Weilhammer15,N. Wermes20, M. Wetstein7, W. Zeuner11, M. Zoeller14

1 HEPHY, Vienna, Austria2 GSI, Darmstadt, Germany3 LETI/DEIN/SPE/CEA Saclay, France4 LENS, Florence, Italy5 University of Florence, Italy6 LEPSI, IN2P3/CNRS-ULP, Strasbourg, France7 Rutgers University, Piscataway, U.S.A.8 INFN, Milano, Italy9 UMM, Cracow, Poland10 CPPM, Marseille, France11 II.Inst. f. Exp. Physik, Hamburg, Germany12 NIKHEF, Amsterdam, Netherlands13 University of Torino, Italy14 Ohio State University, Columbus, OH, U.S.A.15 CERN, Geneva, Switzerland16 MPI f. Kernphysik, Heidelberg, Germany17 FNAL, Batavia, IL, U.S.A.18 Polytechnico Milano, Italy19 University of Toronto, Canada20 Universitaet Bonn, Bonn, Germany21 Universitaet Karlsruhe, Karlsruhe, Germany22 University of Roma, Italy* Spokespersons

The RD42 Collaboration: Institutes from HEP, Heavy Ion Physics, Solid State Physics


Introduction• Historical Overview

– 1913 Röntgen, Joffé, x-ray detection• Resistivity changes of different materials under x-rays exposure

– 1941 Stetten, Counter experiments with natural diamond• Alpha particle detection with natural diamond

– 1945 van Heerden, Crystal Counters• Wide range of crystal counters investigated, also diamond.

– 1984 Angus & Hayman• Low pressure, metastable growth of diamond and diamond-like phases

– 1994 Tesarek et al.• Performance of a Diamond Tungsten Calorimeter

– “First thermo-luminescence experiment”:

Robert Boyle, 28th Oct 1663, Royal Society, London:“I also brought it to some kind of glimmering light by taking it into bed with me, and holding it a good while upon a

warm part of my naked body”


Introduction• Motivation

– LHC and SLHC radiation levels at inner tracking layers O(1015 n cm-2)

– Detectors close to IP or at low rapidity• Vertexdetector• Beam monitoring

• Some advantageous properties of Diamond compared to Silicon :


Introduction: Diamond properties

– Low -> low capacitance

– Low Ileak -> low noise

– Room temperature operation– Fast signal collection time

Property______________________Diamond_____________Siliconband gap 5.47 1.12mass density [g/cm3] 3.5 2.33dielectric constant 5.7 11.9resistivity [cm] >1011 2.3e5breakdown [kV/cm] 1e3...20e3 300e mobility [cm2/Vs] 2150 1350h mobility [cm2/Vs] 1700 480therm. conductivity [W / cm K]10..20 1.5radiation length [cm] 12 9.4Energy to create an eh-pair [eV] 13 3.6ionisation density MIP [eh/m] 36 89ion. dens. of a MIP [eh/ 0.1 ‰ X0] 450 840

–MIP signal 1.9 smaller for same X0

–Collection efficiency < 100%


Introduction

• Diamond material– Synthetic diamond– Chemical Vapor Deposition– Polycrystalline films

– New: large homo-epitaxic mono-crystalline films


Introduction: CVD process

• CVD: Chemical Vapor Deposition

• Allows deposition of diamond on large areas in arbitrary geometries.

• CVD Diamond is composed of columnar microcrystals.

• Growth speed m / h


Introduction• Principle of detector operation

e

h

Substrate-Side

Growth-Side

td collected charge

“collection distance”

= Q / Q0collection efficiency


Introduction: Trapping• The charge collection efficiency is for CVDD <100%.

• Free charge carriers are captured at defects in the lattice.

conduction band

valence band


Trapping has two consequences:

• Polarisation

•“Priming”

•In polychristalline material the defect concentration is very heterogenous !

Creation of space charge and partial compensation of the external field

Charge signal decreases

Passivation of traps by occupation with charge carriers. Decrease of active trap concentration and thus increase of charge carrier lifetime.

Charge signal increases

Impact on charge signal !

Introduction: Trapping


Material Studies


1.0E+00

1.0E+01

1.0E+02

1.0E+03

1.0E+04

1.0E+05

1.0E+06

0 200 400 600 800 1000

Distance [µm]

Concentration [ppmat]

1

1

2

3

300 m

Structure• Scanning Electron Microscopy (SEM)

Pictures show a polycrystalline structure with columnar growth.

• A resolved measurement of Bragg reflexes of single crystallites confirms the growth morphology..

• Secondary Ion Mass Spectroscopy allows to analyze the element concentration as a function of position. A increased concentration of foreign atoms at grain boundaries is found.

H

N

O


Characterisation with ionising radiation

• constant ionization density

• measurement of < >Z (x) •Weak polarization

• Bragg ionization density• measurement of (z=S,x) • strong polarization

MIPS, e- (>1,2 MeV), (10 keV)

p (2 MeV), (5,8 MeV)

ElectronsHoles

Z

X• Resolved measurement can be achieved through beam steering or track reconstruction.


• Characterisation with 90Sr-source, E,max = 2.28 MeV

Amplifier

Diamond Si-detector

Digital Scope

Beta particleSr90 source

Electrons are minimum ionising above 1.2 MeV

ionisation density =36 eh/m

Characterisation with ionising radiation


Material Studies• Signal vs applied Field:

•Saturation above 1 V/m.

•Shape governed by (E) dependence.

•Metallization typicallyCr/AuTi/AuTi/Pt/AuTi/W (new)


Material Studies• Priming / Pumping

– Increase of signal during radiation– Filling of traps increases free carrier lifetime– Empirical fit function:

–

allows to extract priming fluence 0

– Typical increase factor ~1.5 - 1.8

00

0max exp1)( +⎥⎦

⎤⎢⎣

⎡⎟⎟⎠

⎞⎜⎜⎝

⎛

−−⋅−=


Material Studies• Polarization

– Good contacts are critical– Strong Polarization (probably due to contacts) also observed

with MIPS

-50

0

50

100

150

200

0 10 20 30 40

Time [h]

Collection distance [um]

100 V bias

200 V bias

300 V bias

400 V bias

500 V bias

400 V bias

300 V bias

200 V bias

100 V bias

0 V bias

Luis Fernandez HernandoChristoph IlgnerAlick MacphersonAlexander OhTerry PritchardBob StoneSteve Worm


Material Studies

alpha particle, E=5.7 MeV, S=14 m

• Polarization– Observed with mono-chromatic alpha particles


Material Studies

• Synchrotron-radiation HASYLAB

• 10 keV photons, =1600 m

• Beam diameter 50m .. 100m

• Spatially resolved measurement

Diamond

10 keV photons

Ti-Pt-Au

variable aperture

Digital Scope

beam currentbunchclockAmplifier

HV

xy


Material Studies• Polycrystalline structure has impacts on

charge collection:

Signal Height [au]F

requ

ency


Material Studies• The 2D representation shows strong variations and clusters.

• The dimensions of clusters are similar to typical grain sizes.

• A comparison of back and front side measurement shows similar structures.

1 pixel=50 m x 50 m

1 pixel=50 m x 50 m


Material Studies

Leitungsband

Valenzband

Conduction band

Valence band

• The photon beam deposits roughly 0.5 pC /mm2 / s. • The amount of trapped charge can be measured with the TSC (Thermally Stimulated Current) method.• The analysis of the TSC spectrum gives the energy level and concentration of traps.

Principle• Occupation of traps

through ionisation • Heating of the sample

and measurement of I(T)

• Subtraction of dark current


Material Studies• The photon beam profile defines the volume• After exposure the sample is heated to 300C• The I(T) is recorded• The sample is cooled• The sample is driven to the next scan point

•Irradiation for 30s•Heating time to reach 290C is approx. 30s •Cooling time with LN2 to 0C takes approx. 20s•=> cycle time 1.5 min !


Material Studies• The results of a scan on 1330 x 1050 m with a step size of 70 m.• Also the TSC measurement shows strong non-uniformities !• Signal scales sub-linear with irradiation time.

• The trap concentration and the CCE should be antiproportional• Find correlation of TSC Signal and CCE !

12.18

0.00

2.00

4.00

6.00

8.00

10.00

20

0

2

4

6

8

10

12

14

16

18

150 5 10

Spacial representation

02468

101214

0 200 400 600

Time [s]

Current Integral AU


Material Studies

• First measurement of CCE and the TSC Signal on the same area

• Unfortunately the signal-to-noise is not sufficient to draw a final conclusion….wait for next beam time.


Material Studies• Growth Side of a recent polycrystalline CVD diamond

Courtesy of Element Six200m


Material Studies• Non Uniformities

qualitatively reproduced by modeling


Material Studies• The carrier drift process is simulated for single carriers (electrons and holes) accordingly to the lifetime distribution.• Recombination and trapping is distinguished.• The electric field due to trapped charge is calculated.• The drift path for each carrier is determined by the local electric field and thermal diffusion with 1m resolution.

• Recombination and trapping is distinguished, priming effect can be modelled….

trapped electrons

trapped holespolarisation field


Material Studies• Comparison of track

residuals in the model with and without grain structure

• Model with grain structure can qualitatively reproduce residuals observed with diamond pixel detectors.

Tommaso Lari (INFN)Alexander Oh (CERN)

Norbert Wermes (University Bonn)


Material Studies• In 2000 RD42

launched a research program with Element Six to improve the charge collection properties for pCVD diamond.

• Impressive improvements achieved beyond the goal set by RD42.


Material Studies• Latest Material measured with 90Sr Source:

Research Program was successful !


Material Studies

Thinning Experiment

• successive thinning of sample and remeasuring collection distance

•Gain knowledge on collection properties during growth process


Material Studies• Large Wafer Production (5”) possible


Material Studies• Radiation Hardness

– studied with Protons and Pions on pCVD Strip Detectors

– Fluences of 2-3 1015 particles/cm2

– Generally decrease of leakage current with dose observed.

– Resolution of Strip detectors increase with Fluence.

– Pions damage more than protons.– 50% loss of S/N at 2.9 x 1015 pions/cm2.– No loss seen for EM radiation up to 10MGy.

(Behnke et al., Nucl.Instrum.Meth. A489 (2002) 230-240.)


Material Studies• Proton Irradiation

15% loss of S/N 35% improvement in resolution


Material Studies• Pion Irradiation

52% loss of S/N 23% improvement in resolution


Material Studies• Weaknesses of polycrystalline CVD diamond:

– Many grain boundaries -> defects– Non-uniformity of collection properties

• Mono-crystalline CVD diamond is a solution:– No grain boundaries -> less defects – Uniform collection properties – First samples available!


Material Studies• Mono-crystalline CVD

•Perfectly separated from 0e•Narrow Landau distribution•Average 15,000 e


Material Studies• Mono-crystalline CVD

•Saturation already at 0.2 V/m•Collection Distance equals Thickness• ~100% efficient

Diamond Sample CD135 Diamond Sample CD71415b diamonds have been synthesized by Element Six at Ascot


Material Studies

• Mono-crystalline CVD– No significant priming

observed– Trap concentration low

Diamond Sample CD135 diamonds have been synthesized by Element Six at Ascot

QuickTime™ and aTIFF (LZW) decompressor

are needed to see this picture.


Particle Detector Prototypes


Particle Detector Prototypes• Dot detectors

– Characterization

• Strip detectors– Tracking– Slow VA2 and fast LHC electronics

• Pixel detectors– Tracking – CMS and Atlas patterns / electronics


Strip Detectors


• CERN test-beam Setup for Diamond Telescope


• Two planes of the Diamond Telescope


Uniformity in Charge Collection of CVD Diamonds

•Measured with MIPS

•Similar patterns observed as with photon beam measurement


• Next Step: • Biased intermediate

strips to benefit from charge sharing.

• Should improve resolution.


Pixel Detectors


• Diamond Pixel Detectors


• Efficiency = 80%• Resolution = digital

• Results from Atlas Diamond Pixel Detectors

=115m=14m


• Large track residuals• Non-uniformity of response qualitatively reproduces by modeling

• Results from Atlas Diamond Pixel Detectors

Tommaso Lari (INFN)Alexander Oh (CERN)

Norbert Wermes (University Bonn)


• Efficiency = 89% • Resolution = digital

• Results from CMS Diamond Pixel Detectors


Applications in HEP


Applications in HEP• Vertex detectors with CVD Diamond are not

considered yet as an option for LHC.

• For Beam monitoring CVD Diamond is an option for CMS and ATLAS at the LHC.

• BaBar and Belle test already CVD Diamond in their beam monitoring system.


Beam monitoring• For Silicon Vertex systems careful monitoring

is crucial.• Inherently, beam monitors have to be

radiation hard.• Abort Beam when monitors signal dangerous

beam conditions.– False signals must be avoided.– Monitor must be reliable.

• Requirements on the monitoring system depend on the accelerator and vertex system.


One of the first diamond beam monitors

Bergonzo, P., A. Brambilla, et al. (1999). "Semitransparent CVD diamond detectors for in situ synchrotron radiation beam monitoring." Diamond and Related Materials 8(2-5): 920-6.

• Synchrotron beam– E() = 1keV - 30keV– Position determination

without disturbing primary beam

– Fast response (> GHz)


• CMS and Atlas investigate diamond as a sensor for a beam loss and conditions measurements close to the beam pipe.

• Same type of device considered by CMS & Atlas inside their tracker– CMS: ~ z=±1.8m and r=4cm– Atlas: ~ z=±3.5m close to beam pipe

• Act as part of a radiation monitoring system for equipment safety and radiation level/beam monitoring

• A beam conditions monitor can in particular address the following issues:– Allow to protect equipment during instabilities / accidents– Providing feedback to the machine thereby helping them to routinely provide

optimum conditions– Monitor the instantaneous dose during operation

• Advantages of diamond for this application– Radiation hard, low leakage currents at room temperature, fast signal

response• The goal is to detect signs of beam losses and monitor beam conditions

Beam loss and conditions monitor for CMS and Atlas


CMS beam monitor• Diamond activity started.• Possible location of Beam Condition Monitors

(BCM) in CMS.

Beam condition monitors

Looking for increase over normal rate

Monitors to be within CMS volume and feed into

machine interlock

Mika Huhtinen, CERN

Expected dose per accident ~1 Gy <=> 1012 protons


BCM: First test beam program

Test of response to beam loss: T7 PS testbeam

Read out through 16m long RG58 cable connected directly to diamond (no electronics close to beam).

BCM tests done byVladimir CindroLuis Fernandez HernandoChristoph IlgnerAlick MacphersonAlexander OhHeinz PerneggerTerry PritchardBob StoneSteve Worm

Beam: Train of 40ns-wide bunches from CERN PS with 262ns gapUsed to emulate an asynchron beam abort at CMS

Beam intensity: ~8x1011protons per bunch (24GeV)~4x1010 protons/cm2/bunch at the centre of the beam spot ~1x108 protons/cm2/bunch in the halo

Samples:Polycrystalline diamond samples for response to conditions in CMS/ATLAS


BCM Preliminary Results Diamond signal response to high intensity bunch

Almost identical diamond response to PS beam monitor response (pulse length 40ns)

Single pulses from diamond Bias on Diamond = +1 V/um Readout of signal:

16m of cable, no electronics Attenuation on signal cables at scope input Ch 1:20dB (factor 10) Ch 2: 24dB (factor

16)


BCM Preliminary Results Diamond signal response to high intensity bunch

•Bunch train of 7 bunches resolved•Signal stable over bunch train if electric field can be maintained


BaBar beam monitor• For production Si PIN diodes are used.

• Ubias = 50V, Ileak increases with 1nA/krad

• After 100fb-1, noise 50A, signal 10nA

• Since 4 month CVD diamond beam monitor prototype installed

• Package must fulfill space constraints

• Robustness


CVD Diamond installed and working in BaBar


BaBar beam monitor• Promising results!

– Stable operation– Follows closely diode signal


BaBar beam monitor


BELLE Diamond Beam Monitor

CVD diamond is installed and working


Summary• Charge Collection

– Poly-crystalline CVD diamond:• Most probable signal of ~8000e reached (pCVD)• 99% of charge distribution > 3000e• FWHM / MP ~ 0.95

– Mono-crystalline CVD diamond : • MP signal 13,000e• 99% of charge distribution > 10,000e• FWHM / MP ~ 0.3


Summary• Proto-type Detectors

– Dots / Strips / Pixel– Good resolution and S/N 8:1 obtained with rad-

hard electronics– Intermediate Strips are tested this July

• Radiation Hardness– 50% loss of S/N after 2.9 x 1015 pions/cm2

– No loss seen for EM radiation up to 10MGy.– Will be repeated with newest samples


Summary

• Application in HEP– Beam monitoring in BaBar, Belle– Option for CMS and ATLAS Beam monitoring

• Future– Mono-crystalline CVD diamond– Continue research on poly-crystalline diamond to

reach 300m collection distance.

seminar desy zeuthen 20.01.04alexander oh, cern particle detection with cvd diamond alexander oh,...

Documents

cern slide

seminar desy zeuthen

cern particle detection

cvd diamond alexander

university of florence

university of torino

university of toronto

university of roma