Institutionen för Fysik, Kemi och Biologi
Examensarbete
Karakterisering av bly-wolfram-oxid kristallers
optiska egenskaper till CERNs CMS ECAL
Nils Nedfors
LITH-IFM-EX--08/2026--SEInstitutionen för Fysik, Kemi och Biologi Linköpings universitet
Examensarbete LITH-IFM-EX--08/2026--SE
Karakterisering av bly-wolfram-oxid kristallers
optiska egenskaper till CERNs CMS ECAL
Nils Nedfors
Handledare: Etiennette Auffray Hillemanns
CERN
Examinator: Hans Arwin
IFM, Linköpings universitet
Avdelning, Institution
Division, Department
Laboratory of Applied Optics
Department of Physics, Chemistry and Biology Linköpings universitet
SE-581 83 Linköping, Sweden
Datum Date 2009-02-05 Språk Language Svenska/Swedish Engelska/English ⊠ Rapporttyp Report category Licentiatavhandling Examensarbete C-uppsats D-uppsats Övrig rapport ⊠
URL för elektronisk version
http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-16595
ISBN
—
ISRN
LITH-IFM-EX--08/2026--SE
Serietitel och serienummer
Title of series, numbering
ISSN
—
Titel
Title
Karakterisering av bly-wolfram-oxid kristallers optiska egenskaper till CERNs CMS ECAL
Characterization of lead tungstate crystals optical properties for CERN CMS ECAL Författare Author Nils Nedfors Sammanfattning Abstract
The Large Hadron Collider (LHC) at CERN have a capacity to produce proton-proton collisions with an energy of 14 TeV. Four particle detectors are included in the LHC with the purpose to detect all the particles that are created in the collisions. In one of these detectors are scintillating lead tungstate crystals used, to detect the energy of photons and electrons created in the collisions. The energy is detected by measuring of the emitted light from the scintillating crystals. As much knowledge as possible about the optical properties of the crystals are desired to be able to analyze the acquired data from the crystals.
This thesis work presents some techniques used for the characterization of the op-tical properties for the crystals. It also presents measurements done on the decay time of lead tungstate crystals and on the temperature influence to the light yield from the crystals. These measurement results are in addition used in an attempt to estimate how big influence the Cherenkov radiation has to the total amount of emitted light from the scintillating crystals.
The influence from the temperature to the light yield is around −2.02 %/◦C for
BTCP and around −1.75 %/◦C for SICa
. No conclusions could been drawn con-cerning the influence from the Cherenkov radiation to the total amount of emitted light from the temperature measurements.
The decay time measurements showed an influence from the Cherenkov radiation to the total amount of emitted light of; 8 % for crystal 1003, 47 % for crystal 1002 and 19 % for crystal 1001.
Nyckelord
Keywords CERN, CMS, Lead tungstate crystals, Optical properties
a
BTCP (Bogoroditsk Technical Chemical Plant) and SIC (Shanghai Institute of Ceramics) are the two different crystal production facilities used for the production of the crystals
Abstract
The Large Hadron Collider (LHC) at CERN have a capacity to produce proton-proton collisions with an energy of 14 TeV. Four particle detectors are included in the LHC with the purpose to detect all the particles that are created in the collisions. In one of these detectors are scintillating lead tungstate crystals used, to detect the energy of photons and electrons created in the collisions. The energy is detected by measuring of the emitted light from the scintillating crystals. As much knowledge as possible about the optical properties of the crystals are desired to be able to analyze the acquired data from the crystals.
This thesis work presents some techniques used for the characterization of the op-tical properties for the crystals. It also presents measurements done on the decay time of lead tungstate crystals and on the temperature influence to the light yield from the crystals. These measurement results are in addition used in an attempt to estimate how big influence the Cherenkov radiation has to the total amount of emitted light from the scintillating crystals.
The influence from the temperature to the light yield is around −2.02 %/◦C for
BTCP and around −1.75 %/◦C for SIC1. No conclusions could been drawn
con-cerning the influence from the Cherenkov radiation to the total amount of emitted light from the temperature measurements.
The decay time measurements showed an influence from the Cherenkov radiation to the total amount of emitted light of; 8 % for crystal 1003, 47 % for crystal 1002 and 19 % for crystal 1001.
1
BTCP (Bogoroditsk Technical Chemical Plant) and SIC (Shanghai Institute of Ceramics) are the two different crystal production facilities used for the production of the crystals
Acknowledgments
I would like to first express my appreciation to my supervisor at CERN, Etien-nette Auffray Hillemanns how have instructed and guided me through all stages of my thesis work and given me useful feedback on my report. I would also like to tell my thankfulness to Alessandro Thea who have helped me with my work and patiently answered many question especially in the early stages of my work. I would also like to thank my other co-workers at CERN, in particular Benjamin Frisch, Matthias Kronberger and Daniel Abler.
I would also like to mention my examiner at Linköpings University, Hans Arwin for the help with the report and for waiting patiently for me to finish my work. Finally I would like to thank all my friends in Geneva and especially my flatmates at Rue du Vuache for making my time in Geneva a fantastic and unforgettable experience.
Nils Nedfors
Karlskrona, November 2008
Contents
1 Introduction 3
1.1 CERN . . . 3
1.1.1 The Large Hadron Collider - LHC . . . 3
1.1.2 The CMS experiment . . . 4
1.2 Scope of this thesis . . . 5
1.2.1 Problem description and objectives . . . 6
1.2.2 Limitations . . . 6
1.2.3 Outline for this thesis . . . 6
2 Background 9 2.1 The CMS Electromagnetic Calorimeter . . . 9
2.1.1 Photodetectors . . . 10 2.1.2 Preshower detectors . . . 10 2.2 Scintillation . . . 10 2.2.1 Scintillating mechanism . . . 11 2.2.2 Luminescence centers . . . 12 2.2.3 Scintillation properties . . . 15 2.3 Cherenkov radiation . . . 17
2.3.1 Characteristics for the Cherenkov radiation . . . 17
2.4 Lead tungstate crystals . . . 19
2.4.1 Crystal production . . . 19
2.4.2 Optical properties . . . 20
3 Characterization of lead tungstate crystals 25 3.1 Light yield measurements . . . 25
3.1.1 Experimental setup of bench 3 . . . 26
3.1.2 Experimental setup of bench 4 . . . 27
3.1.3 Extraction of light yield parameters . . . 28
3.1.4 The B4 software . . . 33
3.2 Decay time measurements . . . 36
3.2.1 Experimental setup of the decay time bench . . . 36
3.2.2 Extraction of the decay time parameters . . . 37 ix
x Contents
4 Measurements 43
4.1 Description of measurements . . . 43
4.1.1 Decay Time measurements . . . 43
4.1.2 Light yield at different temperatures . . . 44
4.2 Measurement results . . . 44
4.2.1 Results from decay time measurements . . . 44
4.2.2 Results for the measurements of the light yield at different temperatures . . . 47
5 Discussion and conclusions 55 5.1 The influence of the temperature to light yield . . . 55
5.1.1 Discussion . . . 55
5.1.2 Conclusions . . . 55
5.2 The decay time measurements . . . 56
5.2.1 Discussion . . . 56
5.2.2 Conclusions . . . 56
5.3 Future work . . . 57
5.3.1 Determination of the influence from Cherenkov light . . . . 57
5.3.2 The status of Bench 4 . . . 57
List of Figures
1.1 The Large Hadron Collider . . . 4
1.2 The Compact Muon Sloenoid . . . 5
2.1 The electromagnetic calorimeter . . . 11
2.2 Energy bands in a crystal . . . 13
2.3 Potential energy diagram for a luminescent center . . . 14
2.4 A Cherenkov radiation process . . . 18
2.5 An endcap crystal . . . 20
2.6 Distribution of light yield values . . . 23
2.7 Longitudinal transmission and radioluminescence intensity for a PWO crystal . . . 24
3.1 An example of a photomultiplier with its vital parts. . . 26
3.2 A scetch over the electonic setup of bench 3. . . 27
3.3 A scetch over the setup of the electronics used in bench 4. . . 28
3.4 A light yield spectrum from a PWO crystal measured in bench 3. . 29
3.5 The compton scattering process. . . 30
3.6 A plot over the light yield at different points along the crystal . . . 32
3.7 The workflow for operations with the software controlling Bench 4. 34 3.8 A flowchart describing the sequential data acquisition. . . 38
3.9 The classes defined and used by the Bench 4 fitting software . . . . 39
3.10 The flow chart over how the B4 fitting software is running. . . 40
3.11 The flow chart describes how the AutoFit function works. . . 41
3.12 A sketch over the electronics in the decay time bench. . . 42
3.13 An example of a decay time spectrum from a PWO crystal . . . . 42
4.1 Decay time spectra acquired without a filter . . . 46
4.2 Decay time spectra acquired with a UV filter . . . 47
4.3 Crystal 1001 decay time spctra, with and without filter . . . 48
4.4 Crystal 1002 decay time spctra, with and without filter . . . 49
4.5 Crystal 1003 decay time spctra, with and without filter . . . 49
4.6 The summarized decay time spectra . . . 50
4.7 Temperature versus light yield for BTCP crystals . . . 52
4.8 Temperature versus light yield for SIC crystals . . . 52
4.9 Temperature versus light yield for crystal 1001 . . . 53
4.10 Temperature versus light yield for crystal 1003 . . . 53
5.1 Variation of the B4 calibration factor . . . 58
List of Tables
2.1 Lead Tungstate properties . . . 192 Contents
4.1 The different crystals used in the measurements . . . 44 4.2 The fast component decay time parameters . . . 45 4.3 The light yield at different temperatures for the ECAL PWO crystals 50 4.4 The light yield at different temperatures for the crystals 1001 and
1003 . . . 51 4.5 The change of light yield with temperature . . . 51 5.1 Saturation levels for the light intensity over time . . . 57
Chapter 1
Introduction
1.1
CERN
The world’s largest particle physics centre is located outside Geneva close to the Swiss/France border since the 29 September 1954. It is named CERN ‘Conseil Europeen pour la Recherche Nuclear’ and consists of 20 member states, several observer states and organizations.
The main purpose for CERN is to supply scientist from Universities all over the world with the necessary tools to study the building blocks of matter and the forces that holds them together. Extremely high particle energies are required to be able to get a look at the particles holding matter together. To give the particles these high energies they are accelerated in particle accelerators. Therefore CERN supplies scientists all over the world with particle accelerators, particle detectors and computer centers for very powerful data processing.
One of the achievements among many else that have been made during experiments at CERN are the discovery of the W and Z bosons which led to the 1984 Nobel Prize in physics for Carlo Rubbia and Simon van der Meer. World Wide Web was born at CERN originating from a hypertext based network for sharing information between researchers, initiated by Tim Berners-Lee and Robert Cailliu in 1989.
1.1.1
The Large Hadron Collider - LHC
The latest project at CERN is a particle accelerator named The Large Hadron Collider (LHC) which started beam ciculation in September 2008. The collider is built in a tunnel with a circumference of 27 km located 50 to 175 m underground the outskirts of Geneva. The LHC accelerator is capable to provide proton-proton collisions with energy of 7 TeV per beam. To get this collision energy the protons are accelerated in two separated beam lines, in opposite direction to each other, to a velocity of 99.999999 % of the speed of light in vacuum before the collision. In addition to proton-proton collisions, high-energy heavy-ion collision will also occur in the LHC. The two beam lines intersect with each other to form a collision point at four different intersections along the accelerator. An experiment is built at each
4 Introduction
of these collision points with the purpose to detect the particles that are created and the events that occur in the collisions from different point of views. These experiments constitute the four main experiments in LHC and are called: ATLAS, ALICE, CMS and LHCb (See figure 1.1). To be able to circulate the particles in the two beam lines they need to be bended by strong magnetic fields. These fields are created by superconducting magnets that are located along the accelerator. The main purpose for the LHC project is to further increase the understanding of the fundamental structure of the universe, concerning subjects as: dark energy, dark matter and Higgs particle.
Figure 1.1. The Large Hadron Collider with the location of the different experiments along the accelerator. http://atlas.ch/photos/detector-site-surface.html
1.1.2
The CMS experiment
In an underground ‘chamber’ under the village Cessy in France at the opposite side of the LHC tunnel from the Meyrin site is a general particle physics detector built called the Compact Muon Solenoid Detector (CMS). Roughly 2000 scientists from 155 institutes are involved in the CMS project. The detector is constructed as a general purpose detector with the aim to exploit the physics of the proton-proton collisions occurring in LHC over the full range of luminosities expected1 [1]. This means that CMS should be able to measure the energy and momentum of photons, electrons, muons and other charged particles with a high precision.
The completed detector has an overall length of 21.5 m, a radius of 7.5 m and a
1
The luminosity is predicted to 1034
1.2 Scope of this thesis 5
total weight of 12500 ton. The detector is built up by different layers, where each layer has different tasks in the detection of what occurs in the collision. Starting from the central collision point and going outwards are the different layers: an silicon based inner tracking system, a scintillating crystal calorimeter, a sampling hadron calorimeter and a high magnetic-field superconducting solenoid coupled with a multilayer muon system (see figure 1.2).
Figure 1.2. The Compact Muon Solenoid detector [2].
1.2
Scope of this thesis
This report covers a final thesis work for a Master of Science in Applied Physics and Electrical Engineering at Linköping University at the department of Physics, Chemistry and Biology. The work has mainly been done at the PH /CMA division at CERN in Geneva. Supervisor at CERN was Etiennette Auffray Hillemanns. In addition to her supervision, the part of the work concerning the development of the software for bench 4 has been partly supervised by Alessandro Thea. The work started at CERN the 3rd of September 2007 and ended the 31th of October 2008.
6 Introduction
1.2.1
Problem description and objectives
The CMS electromagnetic calorimeter will consist of over 70 000 scintillating lead tungstate crystal. The crystals send out light when it absorbs radiation energy and there is a proportional relationship between the energy absorbed by the crystals and the energy emitted. So by knowing this proportional constant it is possible to measure the energy of the incoming radiation by measuring the energy of the emitted light, which is the purpose for the crystals.
To fully understand the acquired data from the CMS ECAL (Electromagnetic Calorimeter) the mechanism behind the emission of light in the crystals must be fully understood. It is therefore desired to know as much as possible about the op-tical properties of the lead tungstate crystals. The aim of this thesis is to describe some of the techniques used for the characterization of the optical properties and to use these measurement methods to get results that can increase the understanding of the crystals. A part of the emitted light comes from Cherenkov radiation which occurs in the crystals when they are exposed to high energy particles. An optical property, of the lead tungstate crystal that will be looked closer at, is how big influence this Cherenkov radiation has on the total emitted light from the crystal. So called light yield measurement benches are used to measure the light yield for separate crystals. Their main principle is to acquire the light emitted by scin-tillating crystal with a photomultiplier tube when the crystals are excited by a radioactive source. A data analyzing software is then used to extract the light yield of the crystal from the acquired light spectrum. A part of this work is to continue the development of the data analyzing software for the light yield in Bench 4.
1.2.2
Limitations
It is desirable, as mentioned in the problem description, to know how big in-fluence the Cherenkov radiation has on the light emitted by the crystal in the CMS ECAL. The measurements done in this work is carried out on measurement benches primarly constructed to measure other crystal properties than the influ-ence from Cherenkov radiation. The results to expect are therefore more to give a hint about the influence from Cherenkov radiation. The experimental setups of the benches used in this work are not optimized for exact these type of mea-surements instead were experimental setups used based on experience from earlier measurements with the benches. To increase the chances to see the Cherenkov Effect the measurements are mainly done with lead tungstate crystals that have their scintillation properties suppressed.
The development of the software for Bench 4 is mainly focusing on the data analyz-ing part, rather than the software that controls the measurements, and therefore is the description of the software focusing on this part.
1.2.3
Outline for this thesis
The lead tungstate scintillating crystals that constitutes most to this thesis work is a part of the CMS detector which is a very big and technical complex system.
1.2 Scope of this thesis 7
It requires therefore a rather detailed description to introduce the reader into a technical context. The reader is introduced to the technical context in chapter 2. This chapter provides also the reader with background information about how scintillation and Cherenkov radiation works and finishes with a description of the lead tungstate crystals.
The techniques used for the characterization of the optical properties used in this thesis work are described in chapter 3. Since a part of the thesis work have been to continue the development of the light yield Bench 4, the description of how the software works for this bench is more detailed than for the other measurement benches.
Chapter 4 includes a description of the different measurements that is done in this work and also a small motivation to why these types of measurements are done. The description is followed by a result section where all the measurement results are presented.
The conclusions drawn from the measurement results are presented in chapter 5 together with discussion about the results. In addition to the discussions and conclusions, a description about future work for the understanding of the influence of Cherenkov radiation is included. There is also a description about the status of Bench 4 and additional improvements that needs to be done for the bench to work properly.
Chapter 2
Background
2.1
The CMS Electromagnetic Calorimeter
The main task for the CMS Electromagnetic Calorimeter (ECAL) is to measure the energy of the electrons and photons that are created in the particle collisions. It shall also contribute to the measurement of hadron showers and missing energy in collaboration with the hadron calorimeter. The ECAL will play an important role in the search for the Higgs particle through the measuring of the two-photon decay mode for mH ≤ 150 GeV , and by measuring the electrons and positrons
from the decay of W’s and Z’s1 originating from the H → ZZ(∗) and H → W W decay chain for 140 ≤ mH≤700 GeV . An excellent energy and angular resolution
are required in order to measure these physical processes. Another requirement for the ECAL comes from the strict LHC operation conditions which mean that the calorimeter must have a fast response time and an optimum resistance to ra-diation [3].
The ECAL is composed of a barrel part which is closed by two endcap parts, one at each end of the central barrel part. The calorimeter consists of 61200 lead tungstate crystals (P bW O4) in the barrel part and by 7324 crystals in each of the two endcaps. The lead tungstate crystal were chosen because it offers the best prospects to meet the requirements for the ECAL; short radiation length, a small Moliere radius (allowing a compact calorimeter), relatively fast respons and easy to produce at two different plants.
The crystals in the barrel part are mounted grouped together in submodules at radius of 1.29 m from the center of the CMS detector. The submodules are then arranged in different modules, 400 - 500 crystals in each module. Four modules are then forming a supermodule containing 1700 crystals. The barrel crystals have a tapered shape with a length of 230 mm and with a square cross-section of 22 × 22 mm2 towards the center and a cross-section of 26 × 26 mm2 at the rear face [2]. The crystals varies slightly in shape depending on were in the barrel they are mounted in order to avoid cracks aligned with particle trajectories.
1
W boson-weak and Z boson-zero, carrier particles for the weak interaction.
10 Background
The endcaps is mounted with its envelope placed 3.15 m from the detectors inter-action point. All the endcap crystals are of the same tapered form with a length of 220 mm, a front face of 28.62 × 28.62 mm2 and a rear face of 30.0 × 30.0 mm2. These crystals are grouped together into units of 25, known as supercrystals. Each endcap is covered with 276 identical supercrystals plus 36 special partial super-crystals. The crystals are arranged in an x-y grid, with the crystals pointing at a focus point 1.30 m beyond the interaction point, giving off-pointing angles ranging from 2 to 8 degrees [2]. See figure 2.1 for an overview of the ECAL detector.
2.1.1
Photodetectors
The photo detectors, which are used to detect the light emitted from the crystals, need to be fast, radiation tolerant and able to work in an electromagnetic field of 4 T. Because of the small light yield for the lead tungstate crystals the detectors also need to amplify their respons and in addition be insensitive to irradiation from particles. These requirements together with the difference in magnetic field configuration and expected level of radiation, between the barrel and endcap parts, lead to two types of photo detectors.
In the barrel part will silicon avalanche photodiodes (APD) be used, developed specially for the ECAL. Two APDs is glued on the back face of each crystal with an active area of 5 × 5 mm2 for each photo detector. A temperature sensor is embedded in every tenth APD pair [2].
Another type of photo detectors are used in the endcaps, called vacuum photo triodes (VPTs). These photo detectors were also specially developed for CMS. They work as a photomultiplier with only one gain stage with an anode of very fine copper mesh which makes them possible to use in the 4 T magnetic field. Each VPT is 25 mm in diameter with an effective area of approximately 280 mm2. One VPT is glued to the back face of each endcap crystal [2].
2.1.2
Preshower detectors
In front of the crystals in the two endcaps are the Preshower detectors located. Their main purpose is to identify neutral pions in the endcaps. They will also help the identification of electrons against minimum ionizing particles and improves the position determination of electrons and photons with high granularity. They consists of layers of lead radiators which initiate electromagnetic showers from incoming electrons and photons. Silicon strip sensors are placed after each radiator element to measure the deposited energies and the transverse shower profiles.
2.2
Scintillation
What characterizes a scintillating (or radioluminescence) material is the ability to absorb ionizing radiation energy (α, β, γ and X-ray) and reemit the energy in form of visible light. The luminescence can be divided in two different types depending
2.2 Scintillation 11
Figure 2.1. The electromagnetic calorimeter with the barrel part in the middle closed in the ends by the two endcaps [2].
on how fast after the absorption that the light is emitted. If the reemission occurs immediately after the absorption, or precisely within 10−8 s (Being roughly the
time taken for atomic transitions), the process is usually called fluorescence. If it takes longer time than this it is called phosphorescence or afterglow. The time delay between absorption and reemission is due to a metastable excitation state and can last from a few microseconds to hours depending on the material.
2.2.1
Scintillating mechanism
When a scintillator absorbs incident high energy particles the scintilating material falls into a non-equilibrium state. When it relaxes it emits light, this process is called scintillation. The scintillator relaxes towards a new equilibrium state from this non-equilibrium state. The relaxation occurs through a multitude of processes. This scintillation relaxation process can be divided into these five steps [4]:
1. Creation of primary electrons and holes due to absorption of ionizing radia-tion energy.
2. Relaxation of the primary electrons and holes through the production of numerous secondary electrons, holes, photons, plasmon and other electronic excitations.
3. Thermalization of the low-energy secondary electrons (holes) resulting in a number of e-h pairs with energies just above the bandgap energy Eg.
12 Background
4. Energy transfer from the e-h pairs to the luminescence centers and their excitation.
5. Radiative deexcitation of the luminescence centers with emission of light. The first three steps which end in the creation of the e-h pairs are similar in any condensed matter. The second stage in the scintillating process is the most com-plicated process. The primary electron will create an electromagnetic cascade in the solid consisting of an avalanche of electrons, holes, photons and plasmons. The number of secondary particles increases with the depth of the cascade. This mul-tiplication process continues until the electron and photons are unable to create further ionization. At this point the electron - hole pairs relax to an energy near the band gap energy of the solid through thermalization. Some important param-eters are defined from the electromagnetic cascade. One of them is the radiation
length (X0) that can be defined as the mean distance traveled by a high energy electron until its energy has been reduced to a fraction 1/e of the initial energy. Another parameter is the Moliere Radius (RM) that can be understood as the
radius of an infinite cylinder containing 90% of the shower energy.
The fourth and fifth stages in the scintillating process are different for different solids since they involve the luminescence centers in the solid. These centers are the reason for the emission of light and therefore what characterize most of the scintillating properties of a material.
2.2.2
Luminescence centers
The light emitted from scintillating materials originates from radiative transitions between electronic levels. These electronic levels that are responsible for the emis-sion are called luminescent centers. The luminescent centers can either be intrinsic luminescent centers if the electronic levels are from the crystal itself. If the elec-tronic levels corresponds to impurities within the crystal it is classified as extrinsic luminescent centers.
The band theory is a good tool to use when to describe how luminescence in scin-tillators works. This theory says that the electrons of a free standing atom occupy certain discrete energy levels given by Schrodinger’s equation. The energy levels will increase in concentration and form a continuum of energy levels when a large amount of atoms are brought together to form a solid. This continuum can be seen as an energy band of allowed energy levels which can be occupied by the electrons. There are also certain energy values that the electrons cannot have, according to the Schrodinger’s equation. These forbidden energy levels create en-ergy gaps between the allowed enen-ergy levels in an insulator. When an insulator is in a normal state the lower energy bands are fully occupied by electrons, while the higher energy bands are empty. The by electrons filled band with the highest energy is usually called the valence band and the lowest empty band is called the conduction band, separated by an energy gap Eg (See figure 2.2).
This model can only be applied on an insulator with a perfect crystal lattice. In reality for a scintillator there are lattice defects and impurities in the crystal lattices resulting in local electronic levels in the energy gap between the valence
2.2 Scintillation 13
and conduction band (See figure 2.2). These energy levels can be occupied by electrons that are moving freely in the conduction band at the vicinity of these energy levels. These local energy levels can be of three different types:
• Luminescence centers, where the transition to the ground state occurs with a photo emission.
• Quenching centers, from where the transition to the ground state occurs by thermal dissipation.
• Traps, a metastable level from where the electron can either acquire thermal energy from lattice vibration and return to the conduction band or fall to the valence band by dissipate non-radiative energy.
The luminescence and quenching centers comes from impurities, interstitial ions and/or defects. They introduce local discrete energy levels representing the ground and excited state of the center. The traps arise from other lattice defects and introduce additional energy levels for the electrons to occupy below the conduction band.
Figure 2.2. To the left is the energy bands in an insulator with a perfect crystal lattice. To the right is the energy bands together with the local energy levels which occurs in a non perfect crystal.
Mechanism for the luminescence of a center
To describe what is happening when a center emit light, a theoretical model is used where the potential energies of the ground and excited electronic states of the luminescent center is plotted against a configurational coordinate (X) (See figure 2.3). The configurational coordinate is the mean inter-atomic distance between the luminescent centre and neighboring atoms.
14 Background
When the luminescent center is in the ground stable state the center is in position A. If the luminescent center absorbs a photon with enough energy an excitation of the center occurs. This means a transition from point A to point C. The transition is vertical because the excitation and de-excitation of a crystal is fast compared to ionic movements. By thermalization the center will then go to position B to achieve the minimum potential energy of the excited state. The transition back to the ground state, from position B to D, is responsible for the luminescence emission of the center. The amount of time the center spends in position B depends on the probability of optical transition. The center will then go back to the position with the minimum potential energy by non radiative dissipation of excess energy (D to A).
The potential energy curves of the ground and excited states usually intersect or approach each other closely at some point, F1 and F2 in figure 2.3. A luminescent center that reaches point F2 in the excited state can do a non radiative transition back to the ground state point F1. This phenomenon is known as internal thermal quenching and is competing with the emission process. Thermal quenching is more likely the higher the temperature is.
Figure 2.3. Potential energy diagram for a luminescent center. Curve a represent the potential energy for the ground state and curve b the potential energy of the excited state of the luminescent center.
Lead tungsten luminescent centers
Studies have shown that PWO crystals have three luminescence bands [5]:
2.2 Scintillation 15
the crystal. The emission has a wavelength around 420 nm.
• Green band: When anion vacancies are present in W O2−4 centers the irregu-lar W O3anionic molecular complexes can appear which will cause this green luminescence with a wavelength around 490 nm.
• Red band: The presences of Frenkel defects can deform the irregular W O3 center so it emits a red light with a wavelength around 650 nm.
All these centers contribute to the luminescent spectra for the lead tungstate crys-tals. The green and red luminescence centers are usually related to slow compo-nents in the emission spectra therefore have a great effort been made to suppress the influence of theses centers. This has resulted in PWO crystals with scintilla-tion properties where the blue luminescence is dominant with an emission typically peaking at wavelengths around 420 nm.
2.2.3
Scintillation properties
Light yield
One property for a scintillator that is very interesting to know, especially when it is used in a calorimeter, is the relationship between the amounts of light the scintillator emits in proportion to the incident energy. This relationship is given by the light yield LY for the crystal and is expressed as the number of photons emitted from the crystal per MeV of incident energy:
LY = nph Eγ
(2.1) The number of photons emitted by the scintillator is nph in (2.1) and Eγ is the
energy of the incident gamma-particle. The number of photons emitted by a scintillator can be written as:
nph= ne−hSQ where 0 < S, Q < 1 (2.2)
The number of electron hole pairs created is ne−h in the expression (2.2). S is the
efficiencies to transfer the energy from the electron-hole pairs to the luminescent center and Q the efficiencies of light emission from the luminescence centers, step 4 and 5 in section (2.2.1). The energy needed to create an electron hole pair Ee−h
is given by Ee−h = BEg, where Eg is the band gap energy of the scintillator and
B is a material constant. The number of electron hole pairs that can be created
with the incident energy Eγ can therefore be expressed as:
ne−h= BEEγ g
(2.3) It is possible to write an expression for nph with the use of (2.2) and (2.3):
nph= ne−hSQ =
Eγ
16 Background
An expression for the maximum light yield can be written with the use of (2.1) and (2.4): LY =nph Eγ = 1 BEg SQ (2.5)
Lead tungstate crystals have an energy gap, Eg = 4.5 eV and the value for B
is around 7. The ideal case, S = Q = 1, gives LYmax = 32000 ph/M eV for
PWO crystals. This rough estimation is more than 300 times higher than the real light yield measured experimentally. This huge difference is mainly because of the thermal quenching characteristic for lead tungstate resulting in a S << 1. Even if this theoretical value is far from reality it gives a hint about the potential for a scintillating crystal when it comes to light yield.
Decay time
The decay time of a scintillator is a parameter that tells the amount of light that is emitted by the scintillator after a certain time. It is usually defined as the time after which the intensity of the emitted light decays to 1/e of its initial value. The light intensity dependence with the time t can be expressed by a sum of exponential decay functions (2.6). I(t) = n X i Aie−t/τi (2.6)
Ai is the initial intensity of component i in (2.6) and τi is the decay time
pa-rameter for component i. A god conformity with the model (2.6) and the actual intensity dependence with time are reached when each luminescent center of the scintillator is represented by one component in the model. From this model two more parameters can be defined; the contribution from each component Ci (2.7)
and the mean decay time τm(2.8). The contribution from each component is often
multiplied by 100 to give the result as a percentage.
Ci= Aiτi Pn j=1Ajτj (2.7) τm= Pn i=1Aiτi2 Pn j=1Ajτj (2.8) The total number of exponential components used to describe the light intensity is n in (2.7) and (2.8).
The decay time of a scintillator is related to the mean lifetime of the excited states and is therefore inversely proportional to the probability of de-excitation. The de-excitation can be either radiative or non-radiative giving:
τ ∝ 1 pr+ pnr
(2.9) where prand pnr are the probability for radiative and non-radiative de-excitation.
2.3 Cherenkov radiation 17
2.3
Cherenkov radiation
When a charged particle moves through a solid material, with a speed faster than the speed of light in that material, the material starts to emit light. This radiation is called Cherenkov radiation after the physicist P. A. Cherenkov who was the first to predict this phenomenon in 1934 [6]. For Cherenkov radiation to be possible the solid material must be non-conducting and transparent for an observable ra-diation.
The movement of a charged particle through a solid disrupts the local electro-magnetic field in the solid causing the electrons in the solid to be displaced and polarized. When the electrons go back to their equilibrium state, after the charged particle has passed, they emits photons. Under normal circumstances, these pho-tons are destructively interfering with each other and therefore result in a non observable radiation. In the case where a charged particle moves through the material with a speed faster than the speed of light, constructively interference between the emitted photons is instead occurring which result in an intensified radiation. The Cherenkov radiation is analogous to the emission of a shock wave by a projectile, since in both cases the velocity of the object moving through the medium is faster than the resulting wave disturbance in the medium [6]. The ve-locity that must be exceeded by the charged particle is the phase veve-locity rather than the group velocity. The phase velocity of the light can be altered dramati-cally by employing a periodic medium and therefore decrease the required speed for the charged particle to create Cherenkov radiation.
The Cherenkov radiation is emitted at a constant angle theta to the direction of the moving charged particle. An expression for this angle theta can be determined with the aid of figure 2.4. If the charged particle is moving with a velocity of v , it is during the time t moving the distance:
xcp= vt (2.10)
During the same time t the Cherenkov radiated photons are moving the distance:
xph=
c
nt (2.11)
In (2.11), c is the speed of light in vacuum and n the refractive index of the medium. Then angle θ can then be described as:
cos(θ) = c
nv (2.12)
2.3.1
Characteristics for the Cherenkov radiation
The Cherenkov radiation shows a proportional behavior between the overall in-tensity of the radiation and the velocity of the charged particles and the amount of charged particles. The emitted light shows a spectra that varies with the wave-length as λ−2. This means that shorter wavelengths (higher frequencies) results
18 Background
Figure 2.4. The thick arrow represents the charged particle moving through a non-conducting material causing Cherenkov radiation which is represented in the figure with the small dashed arrows.
in the ultraviolet spectrum. The intensity of the radiation does not increase with ever shorter wavelengths since the refractive index is a function that changes with wavelength. The refractive index becomes lower than one at X-ray wavelengths and therefore no Cherenkov radiation is observable at these wavelengths. This means that there is a cut off wavelength where equation (2.12) no longer holds for wavelengths shorter than the cut off wavelength.
Some distinctive differences between the characteristics of Cherenkov light and scintillating light can be drawn [7]:
1. Directionality. The Cherenkov light is emitted at a fixed angle from the velocity vector of the particle that generates it. The scintillating light on the other hand is emitted isotropically.
2. Time structure. Cherenkov light is prompt while scintillating light has one or several characteristic decay times.
3. The spectrum. The Cherenkov light is emitted with a characteristic λ−2
spectrum. The scintillating light has instead a spectrum that varies between scintillating processes.
4. Polarization. Cherenkov light is polarized in contrary to scintillating light. 5. Temperature. The scintillating light is very temperature sensitive compared
2.4 Lead tungstate crystals 19
2.4
Lead tungstate crystals
After four years of study with comparison between different inorganic scintillating crystals the lead tungstate crystals (P bW O4 usually written PWO) was chosen for the ECAL [1]. The main advantages for the lead tungstate crystal are its high density which implies a short radiation length and short Moliere radius which al-low the construction of a compact calorimeter [2]. Other advantages are a fast response time, low production cost and that the production facilities to grow full size crystals were already available at the time when the decision was made. The main drawback for the lead tungstate crystal is the low light yield. This problem is overcome by the high performance of the photo detectors used in the calorimeter. Another problem is that the scintillation in the crystals are very temperature de-pendent. This requires a temperature stabilization system to maintain a constant light yield. Figure 2.5 shows an example of a lead tungstate crystal.
The lead tungstate crystal have a sheelite structure belonging to the space group I4 1/a or monoclinic raspite with a tetragonal unit cell [1], [3]. The dimension of the unit cell is a = b = 0.5466 nm and c = 1.2020 nm. Table 2.1 summarize the properties for the lead tungstate crystals.
Parameter Value Unit
Density 8.28 [g/cm3]
Radiation length 0.89 [cm]
Interaction length 22.4 [cm]
Moliere Radius 2.19 [cm]
Light decay time 5(39 %) [ns]
15(60 %) 100(1 %) Refractive Index at emission peak 2.30
Maximum of emission 420 - 440 [nm] Temperature coefficient -2 [%/◦C]
Light Yield ≈ 10 [pe/M ev]
≈ 100 [ph]
Melting point 1123 [◦C]
Hardness 4 [M oh]
Table 2.1. Lead Tungstate properties
2.4.1
Crystal production
The lead tungstate crystals used for the CMS ECAL are produced by two different manufactures; Bogoroditsk Techno-Chemical Plant (BTCP) in Tula Russia and The Shanghai Institute of Ceramics (SIC) in China. The producers together with the CMS team had to put a great effort in to the understanding of how the crystal work to be able to mass produce full size high quality PWO crystals. The P bW O4 crystals are grown from a mixture of 50 % lead oxide (P bO) and 50 % tungsten
20 Background
Figure 2.5. An endcap crystal with an vaccum phototriode detector attached to the back face of the crystal [2].
oxide (W O3). Two different methods are used for the growth; Choralski used by BTCP and Bridgman-Stockbarger used by SIC.
2.4.2
Optical properties
To be able to use and understand the data collected from the ECAL in the best way as much knowledge as possible about the optical properties for the lead tungstate crystals is desirable. Considering the enormous amount of crystals that is involved in the detector some restriction has to be made to the optical characterizing of the crystals. Therefore only simple and easy measurable properties are studied. These properties are: light yield, decay time and light transmission. Knowledge about these optical properties also gives a confirmation that the requirements on the crystals are fulfilled and therefore results in a scintillating crystal calorimeter that can ensure the performance of the detector. This section summarize the re-quirements set for the optical properties of the crystals in the CMS ECAL together with some results from measurements of the crystals in the CMS ECAL with the automatic measurement systems ACCOCE 1 and ACCOCE 2, see section 3 for a short description.
Light yield
The low light yield for the PWO crystals is one of the major drawbacks for using this type of crystals in the electromagnetic calorimeter. A huge effort has been
2.4 Lead tungstate crystals 21
done in attempts to increase the light yield of the crystals. This has resulted in crystals with a mean value for the light yield of 10.2 pe/M eV for BTCP crystals and 12.3 pe/M eV for the SIC crystals [8]. This light yield is obtained when the crystals are measured with a radioactive source placed 7.5 cm from the small end of the crystal and a Philips XP2262B photomultiplier tube. The rear surface of the crystal is in contact with the photomultiplier trough a thin layer of optical glue. Figure 2.6 shows a distribution of light yield for 61267 BTCP crystals and 1825 SIC crystals [8]. One of the major reasons for the low light yield is due to the strong thermal quenching characteristic of PWO. This strong thermal quenching also results in the very temperature dependent light yield for the crystals (2%/◦C).
The high refractive index of the crystal material is also a reason for the small light yield [8].
A light yield of 8 pe/M eV is required for the crystal to be used in the calorimeter when measured under the conditions mentioned above. The required light yield measured under these circumstances corresponds to a light yield around 4 pe/M eV when detected by two APD’s, glued to the rear face of the crystals in test beam measurements.
Light transmission
The light transmission is a measure of the difference in the amount of radiant flux that enters one side of the crystal and exits the opposite side. A longitudinal transmission spectrum is shown in figure 2.7 together with a radio luminescence spectrum. Photon energies above the band gap energy of the crystal will be ab-sorbed by the material and there is therefore also a threshold wavelength where photons with energies above this value will be absorbed by the crystal. In addition to this absorption caused by band structure there exist internal absorption caused by impurities in the crystal lattice. This internal absorption will shift the thresh-old wavelength towards longer wavelengths (lower energies) and will influence the spectrum more the longer the crystal is.
According to the radio luminescence spectra in figure 2.7, the PWO crystals emit blue-green scintillation light with a broad maximum at 420 - 430 nm. To achieve as high light yield as possible it is desirable to have as good transmission as possible around this light emission peak. The requirements set on the crystals for the lon-gitudinal light transmission at 420 nm is therefore set to a threshold transmission of 55% [1]. The longitudinal light transmission spectra can tell more things about the quality of the crystal. The transmission at especially 360 nm has been proven to have connection to the presence of defects which are correlated to the radiation tolerance of the crystal. It has also been shown that a low transmission at higher wavelengths indicates the presence of a core defect that may affect light collection. Therefore there are two more requirements for minimum threshold transmission; 25% at 360 nm and 65% at 620 nm [1].
Decay time
The short decay time that characterizes the lead tungsten crystals is one of the ma-jor reasons why this scintillating crystal have been chosen for the calorimeter. The
22 Background
decay time for the PWO crystals consist of three components since the scintillation light for the crystals originate from three different types of luminescent centers. The components are: a fast ∼ 5 ns, a medium ∼ 15 ns, and a slow ∼ 100 ns. These three components contribute to the total emitted light as; ∼ 39 % fast,
2.4 Lead tungstate crystals 23
Figure 2.6. Distribution of light yield values for the BTCP (top) and SIC (bottom) crystals [8].
24 Background
Figure 2.7. Longitudinal transmission (1, left scale) and radioluminescence intensity (2, right scale) for a P bW O4 crystal [2].
Chapter 3
Characterization of lead
tungstate crystals
The characterization of the PWO crystals include light yield measurements with two different benches working in a similar way. In addition to this decay time measurements have been done with the use of a decay time bench. The setup of these benches will be described in this chapter together with an explanation of how the measurements are done and the resulting data is extracted.
An automatic measurement system has been used to measure, in a reasonable amount of time, the optical properties of the nearly 80000 PWO crystals in the ECAL detector. The characterization has been carried out on two automatic mea-surement systems and they are called ACCoCE1 and ACCoCE2. They work in a similar way and perform in an automatic sequence: a crystal dimension measure-ment, a longitudinal transmission measuremeasure-ment, a decay time measurement and finally a transversal transmission measurement. The light yield is then extracted from the results from the decay time measurement. Light yield data from the classical light yield benches 3 and 4, which are described further in this chapter, are used for the calibration of the ACCoCE benches. The ACCoCE benches are not described further since they are not used in this thesis work.
3.1
Light yield measurements
The basic principle for the light yield measurements is that the light emitted from a PWO crystal is detected by a photomultiplier tube when the crystal is excited by a radioactive source in. The sources used for the measurements in this work was 60CO. The rear face of the crystal is placed upon the photomultiplier tube
window with a layer of optical glue (n ∼ 1.5) between the two surfaces. The crystal is placed in a tyvek1 envelope so that all the crystal surfaces except the rear face are covered. A cooling unit is also integrated in the benches so that the crystal can maintain a constant temperature during measurement, usually 18◦C.
1
A high density Polyethylen fibres material.
26 Characterization of lead tungstate crystals
A photon that is detected by a photomultiplier tube is absorbed by the photocath-ode. The photocathode will then transfer the photon energy to an electron flux through the photoelectric effect. The efficiency this transformation occurs with is called the quantum efficiency (QE) and is defined as:
QE = N umber of photoelectrons emitted
N umber of incident photons (3.1)
The electron flux is then focused and accelerated towards the first dynode by an electro-optical input system. The first dynode multiplies the electron flux which is then multiplied further through a series of dynodes before it is collected by the anode. The anode converts the electron flux into an output signal. Figure 3.1 shows an example of a photomultiplier.
Figure 3.1. An example of a photomultiplier with its vital parts.
3.1.1
Experimental setup of bench 3
The purpose for bench 3 is to measure the light yield for a scintillating crystal when excited by a radioactive source. The source used for the measurements in this work is a60CO. The crystal holder in the bench has dimensions to suit the ECAL barrel
crystals. The light emitted from the crystal is detected by a XP2262B Photonis photomultiplier tube supplied with a high voltage of 2170 V. In figure 3.2 is a sketch over the bench and the electronics processing the outgoing signal from the PMT. The source is attached to a stepping motor so that the source can be moved
3.1 Light yield measurements 27
Figure 3.2. A scetch over the electonic setup of bench 3.
along the crystal. A cooling unit is also integrated in the bench so that the crystal can maintain a constant temperature during measurement, usually 18◦C.
When the PMT generates a signal, which is discriminated from background noise, it starts the time unit which gives a gate signal for 150 ns. During this time the indiscriminate signal from the PMT is integrated by the QVT2 unit working in Q mode. The magnitude of the resulting analog output from the QVT unit represents the energy from the PMT and is converted to a 1024 bits digital value by a analog to digital converter and then transferred to a PC. All the signals from the PMT are collected in a spectrum in this way. The pedestal in the spectra is created by a pulse generator which creates random pulses that triggers the gate.
3.1.2
Experimental setup of bench 4
The main principle of bench 4 is the same as for bench 3, to measure the light yield for scintillating crystals when excited by a radioactive source. The source used in bench 4 is also a60CO. The crystal holder in B4 is a bit bigger and therefore able
to measure the ECAL endcap crystals which have larger dimensions compared to the barrel crystals. The light yield is measured with a XP2262B Photonis photomultiplier tube supplied with a high voltage of 2200 V. Figure 3.3 shows a sketch over B4 and how the signal from the PMT goes through the electronics and finally is recorded by a computer. The position of the source can be moved along the crystal with a stepping motor. The bench has a cooling unit so that the crystals can maintain a constant temperature throughout the measurements. The data acquisition can either work in pedestal gate mode or self gate mode, which mode to use is controlled through an I/O register. The gate signal is created by a time unit independent of the signal from the PMT in pedestal gate mode. This will result in a spectrum with one pedestal representing zero energy. If the bench is working in self gate mode, the gate signal is generated when the PMT
2
A multichannel analyzer that can analyze the gate signal in three modes: Charge (area, Q), Voltage (peak, V) and time to digital conversion (start/stop, T)
28 Characterization of lead tungstate crystals
detects light. The signal from the PMT are then integrated during the time the gate signal is high and the resulting value corresponds to the photoelectron energy from the PMT. This energy value is then converted to a digital value and stored in a spectrum by a connected PC. In this way two different spectrums are created one for each mode; pedestal mode consisting of only a pedestal corresponding to zero signal energy and self gate mode with a spectrum consisting of a single electron peak and the spectrum over the events created by the scintillating crystal.
Figure 3.3. A scetch over the setup of the electronics used in bench 4.
3.1.3
Extraction of light yield parameters
The resulting light spectrum from the data extraction (fig 3.4) is composed of three peaks: pedestal, single photoelectron and photoelectric peak. The 60CO
source used in the light yield measurements emits gamma rays of two different energies; 1.17 M eV and 1.33 M eV . This ought to result in a spectrum with two distinguished photoelectric peaks but the light yield of the crystal is not high enough to allow the separation of the two peaks. The spectrum shows instead a single photoelectric peak that is composed of the two peaks. The pedestal represents gate signals when there is no signal coming from the photomultiplier and therefore it represent the level of zero energy. The width of the pedestal depends on how big the electronic noise is. The single electron peak come from the normal conduction electrons within the photocathode, which can occasionally have enough energy to escape from the photocathode if they are close enough to the surface. The difference between the position of the pedestal and the single electron peak gives the number of channels representing the energy from one photoelectron and can therefore be used to calibrate the channels in terms of photoelectrons. To be able to do this calibration and finally determine the light yield, the position of the different peaks must be estimated. The position of the pedestal is easy to estimate thanks to its distinctive shape. The single electron peak has also a rather distinctive shape so its position can be determined with the fit of a Gaussian
3.1 Light yield measurements 29
Figure 3.4. A light yield spectrum from a PWO crystal measured in bench 3.
function. The position of the pedestal is subtracted from the value for the position of the single electron peak and the result is the calibration factor CADC or the
number of ADC channels that corresponds to one photoelectron.
CADC= Single electron peak position − P edestal position (3.2)
Some of the photons that enter the crystal can undergo Compton scattering. The incident photon is interacting with an electron and deflected an angle theta with respect to its original direction when Compton scattering occurs. A part of the photon energy is transferred to the electron during this interaction. A sketch of the Compton scattering process can be seen in figure 3.5. The following equation is derived from the conservation of energy and momentum:
E′ γ = Eγ 1 + Eγ mec2(1 − cosθ) (3.3) In (3.3) are; E′
γ the energy of the compton scattered photon, Eγ the energy of the
incident photon, θ the angle between the direction of the incident photon and the scattered one, methe free electron mass and c is the velocity of light. There is a
chance that the scattered photon leaves the scintillator and in this case, only the energy of the Compton electron is transferred to the scintillation.
The energy of the scattered electron (3.3) varies with the angle θ with a max-imum when the scattered photon has its minmax-imum value (θ = π). The maxmax-imum Compton energy Ecfor the Compton scattered electron can be derived from (3.3):
30 Characterization of lead tungstate crystals
Figure 3.5. The compton scattering process.
Ec = Eγ 1 1 + mec2 2Eγ ! (3.4) The energy spectrum of these events is almost flat and is present in the part of the spectrum to the left of the photoelectric peak, i.e. at lower energies than the photoelectric peak. A compton edge is located for energies around Ec. No
compton scattering will occur above this energy so the number of events caused by compton scattering in the spectrum will suddenly drop to zero and form a compton edge in the spectrum. This edge is smeared out due to resolution effect and by secondary photon interactions.
The part of the spectrum that is caused by the Compton scattering must be included in the fitting to estimate the position of the photoelectric peak. This is accomplished with the use of a Fermi-dirac distribution with a Fermi level of
Ec. Since the photoelectric peak is distributed around a central value with a
limited number of photoelectrons it is fitted with a Poisson distribution. These two distributions are combined to form an analytical expression for the fitting function [9]: NN E CADC pe ·e−Npe E CADC ! + LC eE−EcTe + 1 (3.5)
3.1 Light yield measurements 31
The first term represent the Poisson function, while the second represent the Fermi-Dirac distribution. There are five free parameters:
1. N is a normalization factor that takes into account the integral of the pho-topeak
2. Nperepresent the average number of photoelectrons produced
3. CADC represent the number of ADC channels that correspond to one
pho-toelectron, i.e. the calibration factor
4. LC is the mean level of the Compton background
5. Te is a parameter that controls the slope of the Fermi-Dirac distribution
around the Compton edge
The function (3.5) is fitted towards a defined region of the spectrum around the photopeak. After the fit is done, the parameter Npe gives the number of
pho-toelectrons that correspond to the fitted photopeak. This value is the number of photoelectrons that is produced in the photocathode when the photons emitted by the crystal are detected by the photomultiplier. To know the real number of pho-tons (Nph) that reaches the PMT the quantum efficiency (3.1) of the PMT must be
known. The quantum efficiency is strongly dependent with the wavelength of the incident photons. Studies in the laboratory have shown a quantum efficiency that peaks at wavelengths around 400 nm with a peak quantum efficiency of ∼ 13% for the XP 2262B photomultiplier in B3. It is hard to get a precise value for the quantum efficiency and therefore to have more precise values for the light yield it is expressed in photo electrons instead of photons.
As mentioned in section 2.2.3 the light yield is expressed as the number of pho-toelectrons per MeV (2.1). An average value of 1.25 M eV , for the gamma energy emitted by the 60CO source, is used since no distinction can be made between
the 1.17 M eV and 1.33 M eV gamma rays. This gives the following expression to calculate LY:
LY [ Npe M eV] =
Npe
1.25 (3.6)
The Light Yield is measured with the source placed at different positions along the crystal usually at 20 different points with a distance of 10 mm between each point. The light yield values at the different points are then plotted towards where along the crystals they are measured, see figure 3.6. Two linear functions are fitted towards the points in the plot to get a measure of the light yield uniformity along the crystal. The first one is fitted against the nine points measured at closest distance to the PMT (3.5 − 11.5 cm from the PMT), i.e the rear face of the crystal. The second function is fitted against the nine points measured at distances of 11.5 − 19.5 cm from the PMT,i.e. the front face of the crystal. The two fitted functions are written in the form:
yF(x) = aFx + bF , x ∈ [11.5, 19.5]
32 Characterization of lead tungstate crystals 9.4 9.6 9.8 10 10.2 10.4 10.6 10.8 11 11.2 0 2.5 5 7.5 10 12.5 15 17.5 20 22.5 25 BOUND values UNBOUND values Dist. from PMT (cm) Npe/MeV
Figure 3.6. A plot over the light yield at different points along the crystal [1].
A reference light yield is defined from (3.7) as the average value for the two lines at x = 11.5 cm:
yref =
yF(11.5) + yR(11.5)
2 (3.8)
The slope of these two linear functions gives a value for the light yield uniformity. They are called the Rear Non Uniformity (Rnuf) for the first function and Front Non Uniformity (Fnuf) for the second function. They are expressed in units as %/X0 instead of MeVNpe /cm (3.9). F nuf [%/X0] = aF 100 yref X0 Rnuf [%/X0] = aR 100 yref X0 (3.9)
3.1 Light yield measurements 33
The electromagnetic shower described in section 2.2.1 reaches its maximum at a deep of 8 X0 from the front face of the PWO crystal [1]. This deep corresponds to a distance of 15.5 cm from the PMT or the rear face of the crystal. The light yield for the crystals at this shower maximum is of interest and is calculated from the fitted function at the front face of the crystal, LY at 8X0 = yF(15.5) and is
often refered as the light yield of the crystal LY.
These three parameters; Fnuf, Rnuf and LY (at 8X0) are the results that are of the biggest interest from the measurements with the light yield benches.
3.1.4
The B4 software
The software in B4 controls the measurements and extracts the desired results from the acquired data. All the required software is installed on one PC that is connected to a VME3 crate on B4 and the motor that controls the movement of the source. The software can be divided in to two distinctive parts. One part controls the measurements and acquires the data. This part is done with the National Instruments LabView software. The other part is extracting the measurement results from the acquired data. This part is done with a C++ based object oriented data analysis framework called ROOT developed for data analysis at CERN.
The measurement control software
Measurements in B4 are controlled by the LabView program NewFastDAQPanel.vi and its sub vi’s4. The LabView software is communicating with the motor in B4 through a Parker Digiplan PDX-series motor driver which is connected to the com-port of the computer. It is also communicating, through a National Instruments MXI-2 bus, with an 16 channel I/O register and an LeCroy 1182 ADC 5 unit placed in a VME crate. Fig 3.7 shows the workflow for the software controlling B4. Four basic operations can be reached from the main frame of NewFastDAQ-Panel.vi. The motor control is a window from where the motor that moves the source can be controlled. All the settings for the measurements are done in the configuration window. It consists of two parts; one part where all the hardware settings are being done and one part where the settings for the data acquisition is being done. The hardware settings are both for the motor control (PDX settings) and for the LeCroy 1182 ADC device which transfers the data from the bench to the computer. In the configuration panel is; information about where the data should be stored, spectrum settings, number of source positions and settings for the fitting of the acquired spectrum done.
The data acquisition in B4 can occur in two different modes either with a pedestal run or a self gate mode run. See section 3.1.2 for a description over how the different modes operate. A measurement run can be built up by a sequence of different acquisition modes. This sequence is defined together with the size of
3
VME is a computer bus standard 4
Programs that runs under LabView has the file extension .vi and a sub vi is a LabView program integrated in to another LabView program
5
34 Characterization of lead tungstate crystals
Figure 3.7. The workflow for operations with the software controlling Bench 4.
each acquisition directly in the main frame controlling B4. The LabView software is then selecting between the different modes through the I/O register. A high value on channel 9 enables Pedestal Run while a high value on channel 8 enables Self Gate run. The sequential data acquisition occurs according to the flowchart in figure 3.8. When a sequence has started a header file is created, consisting of information about the measurement in the folder specified in the configuration window. The source is then moved to the position specified in the configuration window and the data acquisition begins in the mode defined by the first sequen-tial continuing until the desired measurement size is fulfilled. The next acquisition mode specified in the sequence is then started. After the completion of a sequence, the source is moved to the next position specified in the configuration window and the measurement sequence starts again. The complete measurement is finished when all the source positions has been measured.
The software for extraction of the light yield parameters
The measurement results that are extracted from the acquired data are the light yield at each measured point along the crystal. The LY at 8 X0, the Fnuf and
