A high sensitivity imaging detector for the study of the formation of (anti)hydrogen

(1)

Department of Physics, Chemistry and Biology

Master’s Thesis in Applied Physics

A high sensitivity imaging detector for the study of the

formation of (anti)hydrogen

Karl Berggren

LITH - IFM - A - EX - - 13/2827 - - SE

Research conducted at

CERN, Gen`

eve, Switzerland

Department of Physics, Chemistry and Biology Link¨oping University

(2)

(3)

A high sensitivity imaging detector for the study of the

formation of (anti)hydrogen

Department of Physics, Chemistry and Biology (IFM), Link¨opings Universitet Karl Berggren LITH - IFM - A - EX - - 13/2827 - - SE Thesis project: 30 hp Supervisors: Dr. M. Doser, AEGIS, CERN Dr. C. H¨oglund,

EES AB / Link¨opings Universitet Examiner: Dr. M. Magnusson,

Link¨opings Universitet Link¨oping: September 10, 2013

(4)

(5)

Datum

2013-09-10

Avdelning, institution

Division, Department

Applied Physics

Department of Physics, Chemistry and Biology

Linköping University

URL för elektronisk version

http://urn.kb.se/resolve?

urn=urn:nbn:se:liu:diva-97359

ISBN

ISRN: LITH-IFM-A-EX--13/2827--SE

_________________________________________________________________

Serietitel och serienummer ISSN

Title of series, numbering ______________________________

Språk Language Svenska/Swedish Engelska/English ________________ Rapporttyp Report category Licentiatavhandling Examensarbete C-uppsats D-uppsats Övrig rapport Titel

A high sensitivity imaging detector for the study of the formation of (anti)hydrogen

Författare Karl Berggren

Nyckelord

CERN, AEGIS, MCP, Microchannel plates, cryo, cryogenic, antihydrogen, hydrogen, gain, resistance Sammanfattning

AEGIS (Antimatter Experiment, Gravity, Interferometry and Spectroscopy) is an experiment under development at CERN which will measure earth's gravitational force on antimatter. This will be done by creating a horizontal pulsed beam of low energy antihydrogen, an atom consisting of an antiproton and a positron. The experiment will measure the vertical deflection of the beam through which it is possible to calculate the gravitational constant for antimatter. To characterise the production process in the current state of the experiment it is necessary to develop an imaging detector for single excited hydrogen atoms. This thesis covers the design phase of that detector and includes studies and tests of detector components. Following literature studies, tests and having discarded several potential designs, a baseline design was chosen. The suggested detector will contain a set of ionising rings followed by an electron multiplying microchannel plate, a light emitting phosphor screen, a lens system and finally a CCD camera for readout. The detector will be able to detect single hydrogen atoms, measure their time of flight as well as being able to image electron plasmas and measure the time of flight of the initial particles in such a plasma. Tests were made to determine the behaviour of microchannel plates at the low temperatures used in the experiment. Especially, the resistance and multiplication factor of the microchannel plates have been measured at temperatures down to 14 K.

(6)

(7)

Abstract

AEGIS (Antimatter Experiment, Gravity, Interferometry and Spectroscopy) is an experiment under development at CERN which will measure earth’s gravi-tational force on antimatter. This will be done by creating a horizontal pulsed beam of low energy antihydrogen, an atom consisting of an antiproton and a positron. The experiment will measure the vertical deflection of the beam through which it is possible to calculate the gravitational constant for antimat-ter. To characterise the production process in the current state of the experi-ment it is necessary to develop an imaging detector for single excited hydrogen atoms. This thesis covers the design phase of that detector and includes studies and tests of detector components. Following literature studies, tests and having discarded several potential designs, a baseline design was chosen. The suggested detector will contain a set of ionising rings followed by an electron multiplying microchannel plate, a light emitting phosphor screen, a lens system and finally a CCD camera for readout. The detector will be able to detect single hydro-gen atoms, measure their time of flight as well as being able to image electron plasmas and measure the time of flight of the initial particles in such a plasma. Tests were made to determine the behaviour of microchannel plates at the low temperatures used in the experiment. Especially, the resistance and multipli-cation factor of the microchannel plates have been measured at temperatures down to 14 K.

Keywords: CERN, AEGIS, MCP, Microchannel plates, cryo, cryogenic, anti-hydrogen, anti-hydrogen, gain, resistance

URL for electronic version:

http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-97359

Abstract in Swedish: Sammanfattning

AEGIS (Antimatter Experiment, Gravity, Interferometry and Spectroscopy) är ett experiment under uppbyggnad vid CERN som kommer att mäta jordens gravitationella kraft p˚a antimateria. Detta kommer att göras genom att skapa en pulsad str˚ale av antiväte, en partikel som best˚ar av en antiproton och en positron, och mäta hur denna str˚ale faller. Fr˚an detta är det sedan möjligt att beräkna gravitationskonstanten för antimateria. För att karakterisera processen för antiväteproduktion behövdes en avbildande vätedetektor utvecklas för

(8)

perimentets kommissioneringsfas. Det här arbetet täcker designfasen av den de-tektorn och inkluderar studier och tester av detektorkomponenter. Efter litter-aturstudier, tester och ha eliminerat potentiella designkomponenter har ett de-signförslag tagits fram. Den föreslagna detektorn best˚ar av en grupp joniserande ringar följt av elektronmultiplicerande mikrokanalplattor, en fluoroscerande fos-forskärm, ett linssystem och avslutas med en CCD kamera för utläsning. De-tektorn kommer kunna detektera enskilda väteatomer, mäta deras flygtid och ¨

aven avbilda elektronplasmor och mäta flygtiden för de initiella partiklarna i ett s˚adant plasma. Tester genomfördes för att bestämma mikrokanalplattor-nas egenskaper vid de l˚aga temperatur som används i experimentet. Speciellt s˚a har resistansen och multipliceringfaktorn för mikrokanalplattorna mätts vid temperaturer ner till 14 K.

(9)

Acknowledgements

It might be my thesis but this work couldn’t have been done with all the great people I have around me such as. . .

. . . my mom and dad whom has always supported, encouraged and helped me to do whatever I wanted to do and do it well.

. . . my sister who I can talk to about anything between heaven and earth. . . . my granddad who let me dissect all sorts of electronics when I was young,

without him I would most certainly not be about to become an engineer. . . . my university supervisor, Carina, who has been available to me on Skype

all the time, checked in on me regularly and even visited me at CERN. . . . my examiner Martin, who has helped me with all the details that comes

with a master’s thesis and who also came and visited me.

. . . my supervisor at CERN, Michael. If I ever got lost in my work a talk with him made everything so obvious and clear.

. . . Christian, who has helped me improve in all aspects of experimental physics. . . . Andreas, who had great answers to both the good and the bad questions. . . . all the other amazing people at AEGIS from whom I have learnt so much. . . . my bros Olof and Rikard. Through the fires of Firenze and the frosts of

Gen`eve.

. . . all the great friends I’ve made here in Gen`eve, you make a good life great. And lastly, thank you world for being such an awesome playground.

(10)

(11)

Nomenclature

The most important reoccurring abbreviations and symbols are described here.

Symbols

¯

g Gravitational constant for antimatter 0 Vacuum permittivity

r Relative permittivity

OAR Open Area Ratio

p Proton p Antiproton e− _Electron e+ _Positron H Hydrogen H Antihydrogen Ps Positronium

Abbreviations

CERN European Center for Nuclear Research LHC Large Hadron Collider

AEGIS Antimatter Experiment Gravity, Inferferometry, Spectroscopy AD Antiproton Decelerator

MCP Microchannel plate SFC Segmented Faraday Cup PSU Power Supply Unit

(12)

(13)

Introduction

The following chapter introduces the thesis, presenting the context of where and why the detector is being developed. It starts out with an overview of the layout of the thesis to simplify reading or finding particular information.

1.1 Reading instructions

This thesis was created to both represent the work done in the project as well as being a future reference for the AEGIS experiment. It therefore contains an ex-haustive method chapter as well as an appendix that summarises the important design parameters that should be used in the continued development.

The current chapter introduces the project and the setting in which it has been run. It starts out describing the high level context of the project and then narrows down to a level where the project parameters can be decided. Chapter 2 further describes the project and defines the requirements for the end product. Chapter 3 introduces relevant theory regarding microchannel plates (MCPs) and phosphor screens. Chapter 4 describes the methods used in tests and the data analysis. Chapter 5 summarises and analyses the results from the tests. Chapter 6 presents the different design components that were not chosen for the project and the reasons for not choosing them. Chapter 7 presents the chosen system and its specifications. Chapter 8 discusses the capabilities of the designed system as well as other topics of interest that were encountered during the project.

Appendix A explains different concepts that are important to the AEGIS exper-iment but not directly relevant to the work in this thesis. Appendix B contains a summary of the specifications for the system and is intended to be reference document for future work.

(18)

1.2 CERN

The European Organization for Nuclear Research (CERN) is the worlds largest physics research institute; employing about 2400 people and hosting more than 10,000 guest scientists from 608 different universities. It was founded in 1954 by 12 European nations for research about the nucleus, the centre of atoms. The researchers at CERN were soon able to start probing deeper than just the nucleus and today they study even smaller constituents of matter, also making it known as the European laboratory for particle physics. CERN is made up of and mainly financed by 20 member states but is also funded by financing agencies from other countries and has cooperation agreements and scientific contacts with even more countries. The yearly budget is today over 1 billion CHF (7 billion SEK) and it is mainly used for construction of the equipment needed for the experiments such as accelerators, detectors and computing infrastructure. [1]

1.2.1 The LHC and the accelerator complex

The main workhorse at CERN at the moment is the LHC (Large Hadron Col-lider). The LHC is a circular particle accelerator with a circumference of 27 km that can accelerate either protons or heavy lead ions in two beams in opposite directions. With up to 3.5 TeV per particle in each beam the resulting collisions has a total energy of up to 7 TeV. The LHC cannot accelerate particles from rest and the particles are therefore accelerated in stages in what is known as the accelerator complex (Figure 1.1). The accelerator complex consists of several accelerators where the majority of them act as boosters for the LHC but they are also used independently for other experiments. They also in a way reflect the history of CERN as several of them have at some point been used as the main accelerator.

The LHC supplies the particle beams for the main experiments at CERN but there are also other parts of the complex that serve other experiments. Ex-amples of these are the ISOLDE, the n-TOF and CNGS (CERN Neutrinos to Grand Sasso) targets, the CTF-3 (CLIC Test Facility) and the AD (Antiproton Decelerator).

CTF-3 is a test facility for the next generation of linear colliders where new components and techniques are being evaluated for one of the candidates to replace the LHC after it’s decommissioning.

The AD decelerates antiprotons from close to the speed of light down to only 10% of it. The antiprotons are used as a source for several experiments studying the properties of antimatter.

1.2.2 Experiments

ATLAS (A Toroidal LHC Apparatus) and CMS (Compact Muon Solenoid) are the main experiments at CERN which are searching for new fundamental parti-cles by studying the high energy proton-proton collisions produced by the LHC.

(19)

1.2. CERN 3

Figure 1.1: (colour) An overview of the present accelerator complex at CERN and some of its experiments

(20)

They both have the same mission but are using two different designs for their detectors. This increases the chance of finding new particles as well as giving the ability to double check each others results. The other two large LHC exper-iments are LHCb, which studies differences between matter and antimatter via the bottom quark, and ALICE (A Large Ion Collider Experiment) where quark-gluon plasmas are studied, a state of matter that is thought to have existed just after the big bang. There are also smaller experiments connected to the LHC as for example MOEDAL (Monopole and Exotics Detector at the LHC) which is searching for the hypothetical magnetic monopoles.

There are also a number of non-LHC experiments, some of them make use of other accelerators at CERN while some do not need accelerators at all. There are several experiments studying the standard model such as DIRAC (Dime-son Relativistic Atom Complex) which studies properties of the strong force via pionium decay. Beyond standard model experiments such as CAST (CERN Ax-ion Solar Telescope) and OSQAR (Optical Search for QED Vacuum Bifringence, Axions and Photon Regeneration) which both are searching for the hypothetical axion particle. ISOLDE (Isotope mass Separator On-Line facility) is a facility with dozens of small scale experiments in atomic physics. The multi-purpose detector AMS (Alpha Magnetic Spectrometer) is mounted on the International Space Station. There are also more applied experiments such CLOUD (Cos-mics Leaving OUtdoor Droplets) which is studying the effect of cosmic rays in cloud formation and AWAKE (Proton Driven Plasma Wakefield Acceleration Experiment) which is developing next-generation accelerator technology. The AD complex hosts a range of experiments that are exploring the properties of antimatter by using the low energy antiprotons provided by the antipro-ton decelerator. The AD takes antiproantipro-tons from a metal target where proantipro-ton- proton-antiproton pairs are created in high energy collisions created by the Positron Synchrotron. These antiprotons are then slowed down in the decelerator from several GeV down to MeV levels by using electron and stochastic cooling. The cooled antiprotons are then distributed in bunches to the different experiments in the AD complex. The current experiments working in the AD complex are

• ACE (Antiproton Cell Experiment) is studying how antiprotons affect human cells and its suitability as a future cancer therapy.

• ALPHA captures and studies antihydrogen and compares it to ordinary hydrogen.

• ASACUSA (Atomic Spectroscopy And Collisions Using Slow Antiprotons) is studying antimatter by creating hybrid particles such as ”antiprotonic helium”. These hybrid atoms have longer lifetimes than antihydrogen and can therefore be studied in further extent.

• ATRAP also studies antihydrogen but is using different techniques, such as cold positron cooling, to try to capture the antihydrogen for longer times.

• The AEGIS (Antihydrogen Experiment: Gravity, Interferometry, Spec-troscopy) experiment is studying the gravitational force on antimatter. This thesis covers work that is part of the development of the AEGIS experiment.

(21)

1.3. AEGIS 5

The Higgs boson

One of the best current models for describing our universe is called the Standard Model. It describes three of the four forces in nature, the electromagnetic, the strong and the weak force. The model can be seen as describing the world by the interaction of a set of fundamental particles and these particles make up the universe as we know it, with energy, matter and the forces that connect everything. Until recently, all particles in the Standard Model except one had been observed, the Higgs boson. Named after the man who predicted it, it had been haunting particle physicists for several years and was one of the main reasons for the construction of the LHC. The Higgs boson is a key element of the Standard Model since it is the particle that gives other particles mass. Understanding the origin of mass is an essential factor in a successful theory for understanding our universe. In the summer of 2012 the ATLAS and CMS experiments announced together that they had both discovered a particle with a rest-energy of 126 GeV, which is in the predicted mass range and also with other properties compatible with the ones predicted for the Higgs boson. This particle is today considered to be the Higgs boson but there are still questions regarding its properties. Therefore ATLAS and CMS continue to study the Higgs boson at the same time as the next generation of accelerators are being designed, that will be able to study the Higgs with even higher precision.

Long Shutdown 1

The accelerator complex at CERN is currently shut down for upgrades, the upgrade period being called Long Shutdown 1. After this time the LHC will be able to produce collisons with energies up to 14 TeV compared to the max-imum of 8 TeV reached in 2012. The interval between collisions will also be reduced from 50 ns to 25 ns. At the same time upgrades are being done at the antiproton decelerator, specifically an additional cooling ring is being installed named ELENA. This ring will be able to cool antiprotons even further, reducing the need for degrader foils at which a large portion of the flux is lost otherwise. ELENA will also in contrast to the previous ejector be able to deliver antiproton bunches simultaneously to all experiments, allowing for increased beam time and extended data collection. With the addition of increased beam time, AEGIS is also expected to increase its event rate ten-fold when ELENA starts operating in 2017.[2]

1.3 AEGIS

The AEGIS experiment at CERN seeks to test the weak equivalence principle (WEP1_{) for anti-matter. This is to be done by creating a beam of cold (read:}

slow-moving) anti-hydrogen which is sent through a moir´e deflectometer2_{. This}

1_{The weak equivalence principle is, in its essence, the principle that objects with the same}

starting position and velocity will follow the identical trajectories in a gravitational field regardless of their mass and composition.

2_{A moir´}_{e deflectometer consists of two gratings placed after each other and which creates}

(22)

Figure 1.2: (colour) Schematic images over the basic principle of the AEGIS experiment. Image: AEGIS/CERN

results in a diffraction pattern which can be used to measure the drop of the beam due to gravity, which in turn tells us the earths gravitational force on anti-matter. According to the standard model this should be the same as for ordinary matter but it has never been measured with sufficient precision and is thus an important step in strengthening the support for the WEP. However, a different gravitational force on antimatter would be a major discovery. It could for example help us explain the unbalanced ratio between matter and antimatter in the universe and would force the physics community to reevaluate the standard model. [3][4]

To measure the gravitational fall of antihydrogen the following steps will be done, as also illustrated in Figure 1.2.

• First antiprotons (p) coming from the AD are trapped and cooled to 100 mK in a cylindrical Penning trap3_.

• Shortly after, a beam of positrons (e+_{) will hit a nano-porous material just}

above the trap. In this material the positrons will combine with electrons and form positronium (Ps)

e+_{+ e}− _{→ Ps.}

The positronium ejects from the material in the direction of the antiproton trap.

• To increase their lifetime, the positronium is excited, Ps → Ps∗, using a laser system focused in the region between the positronium converter and the antiproton trap.

• The excited positronium is combined with antiprotons, creating excited antihydrogen atoms and free electrons

p + Ps∗→ H∗+ e−

particles they will fall a different distance between the two grates. This difference gives rise to a displaced diffraction pattern and by measuring the displacement of the pattern one can deduce the gravitational force on the particles.

3_{A Penning trap is set of electric and magnetic fields which together are able to confine}

(23)

1.4. The 1 T Penning trap 7

• The antihydrogen is accelerated into a beam using Stark acceleration, inhomogeneous electric fields that utilise the Stark shift4 _{in the atoms.}

The low temperature of the formed antihydrogen is essential to get a well aligned beam since thermal velocities will cause radial divergence of the atoms.

• The beam goes through a moir´e deflectometer which creates a shifted diffraction pattern depending on the fall of the particles. From this it is possible to calculate the gravitational constant for the antihydrogen. During the current Long Shutdown 1 the AEGIS experiment does not have access to the antiprotons required to create antihydrogen. To keep the devel-opment going the experiment will instead perform measurements on regular hydrogen. This is possible due the symmetry in the process used to create the antihydrogen. Changing the antiprotons to protons in the above reaction, hydrogen and a positron are formed instead

p + Ps∗→ H∗_{+ e}+

A new complication arise from this: hydrogen, being ordinary matter, does not interact as strongly with its surroundings as an antihydrogen atom would. There is therefore a need for a detector which can detect single excited hydrogen atoms and help characterise the processes that will be used to produce antihydrogen. This thesis covers part of the development of that detector. From here on the contents of the thesis will mainly be discussed in regards to the hydrogen production.

1.4 The 1 T Penning trap

The main events of the experiment, the positronium and hydrogen production, will take place in a 1 T Penning trap located in the second half of the experiment chamber (see Figure 1.3 and Figure 1.4). This part of the experiment is sur-rounded by a superconducting solenoid which creates a homogeneous magnetic field of 1 T inside itself. The transition temperature for the superconducting material, 7 K for 5 T magnet and 9 K for the 1 T magnet, is one of the reasons why the experiment is cooled to 4 K by using liquid helium. The other reason is the importance of reducing the thermal velocities of the hydrogen and the 4 K environment is then necessary to be able to cool the hydrogen in the trap down to the required 100 mK.

Inside the trap the protons will be trapped just below the positronium con-verter. Between the target and trap is the focus of the lasers that will excite the positronium and in turn the hydrogen. The trap is constructed as a cylinder consisting of several rings which can be biased at different potentials to control the trapping and eventually also accelerate the formed hydrogen. To reduce any noise from stray particles, the inside of the chamber is also kept at ultra high vacuum with a pressure lower than 10−9 _{mbar and the trap will locally have}

a pressure down to 10−15 _{mbar. The trap ends approximately 4 cm from the}

centre of the hydrogen formation region. From there there is 75 cm available

(24)

A-A ( 1 : 10 )

A

1 1 2 2 3 3 4 4 5 5 6 6 A A B B C C D D A3 1T Magnet2.0.idw 1 Tolerances ISO2768 mK ISO13920 BF surface: N/A N/A Ra Ra

undimensioned edges ISO 13715 surface DIN EN ISO 1302

Quantity: material: scale: creator: C. Loeffler date: 04/09/2012 checked by: date: 1T Magnet2.0 PART NUMBER

State Changes Date Name

1T Magnet2.0.iam middle plane 850 152,08 532,5 15 0 12 5 360,95 936,2

Figure 1.3: Drawing of the 1 T magnet of the experiment and its interior, showing the 1 T trap in the centre and to the left the space in which the hydrogen detector will be placed. The 5 T magnet and the antiproton injection is then placed to the right in the drawing and the moir´e deflectometer will be placed to the left of the drawing. Image: AEGIS/CERN

(25)

1.4. The 1 T Penning trap 9

A-A ( 1 : 20 )

B ( 1 / 5 )

A

B

1 1 2 2 3 3 4 4 5 5 6 6 A A B B C C D D A3 Drawing_space.idw 1 Tolerances _{ISO13920 BF}ISO2768 mK surface: N/A

N/A

Ra Ra

undimensioned edges ISO 13715 surface DIN EN ISO 1302

Quantity: material: scale: creator: C. Loeffler date: 19.01.2013 checked by:

date: Experiment Mounted

PART NUMBER

State Changes Date Name

Experiment Mounted.iam 12 5 15 0 377,35 82,75 294,55 6 6 ,5

Figure 1.4: A close up of the 1 T Penning trap (to the right in the drawing) and the region where the hydrogen detector will be placed. The centre of the image shows the 66.5 mm restriction. Image: AEGIS/CERN

(26)

to the end of magnet with a 66.5 mm restriction in between (see Figure 1.4). It is within this space that the hydrogen detector will need to be placed in the coming commissioning stages of the experiment.

1.5 Purpose

This thesis is a part of the development of a hydrogen detector for the AEGIS experiment that will be part of the commissioning process of the experiment. The final detector will be used to evaluate and characterise the hydrogen pro-duction of the experiment as a step towards the antihydrogen propro-duction that will start when the AD starts operating again in the second half of 2014. With antihydrogen the AEGIS experiment will measure the gravitational force on an-timatter to help us increase our understanding about anan-timatter and possibly explain mysteries as for example the baryon asymmetry.

The master’s thesis is also a practical and theoretical learning experience where previous education is applied to solve new problems and new knowledge has to be taken on to complement the existing.

(27)

Chapter 2

Objective and layout of

project

The following chapter presents the state of the development at the start of the project and general and use-case specific requirements for the detector. It is concluded with an overview of how the project was structured and performed.

2.1 Initial state of development

Upon the start of this thesis there had already been some development done regarding the hydrogen detector, mainly with respect to the initial ionising stage and tests of a microchannel plate (MCP) delay line detector.

The principle of detection had at this point been decided to work by ionising the Rydberg hydrogen1 _{coming out of the 1 T Penning trap. This is done by}

placing a set of rings with a high potential difference close to each other. At high enough electric field gradients this can ionise excited hydrogen atoms, allowing their constituents to be more easily controlled and accelerated. The lifetime of the excited states is not exactly known in the conditions of the experiment but is judged to at least be in the order of 100µs and possibly even longer. Nevertheless, due to the atoms low speed (coming from a sub-K region), it is still important for the ionising rings to be as close as possible to the trap but without disturbing the sensitive processes taking place inside it. With a ring design one gets a radial dependence of the ionised states and if imaged precisely enough it is considered possible to measure the distribution of the excited states. The ionising rings were also researched and developed further in parallel with the development of the rest of the detector. [5]

1_{A Rydberg atom (also explained in Appendix A.3) is an atom with a high quantum}

number and with energy levels that are well described by the Bohr model

(28)

2.2 Objective

The objective of the project was to design and develop a detector that could analyse the hydrogen production in the AEGIS experiment. The process is the same as the one that will be used when the experiment starts producing antihydrogen after Long Shutdown 1 and understanding its characteristics is an important part of developing the experiment as a whole.

The requirements can be divided into general requirements about the detector, due to the environment it is supposed to operate in, and in four main use cases. The general requirements on the detector were as follows

• It must work in a 4 K environment and output little or no heat to its surroundings. This is so that the device does not interfere with the sub-K regions in the 1 T trap and that it has a minimal load on the cryogenic system of the experiment.

• It should also be able to handle temperature cycling up to 600 K to allow for vacuum baking. Vacuum baking is a process of heating components to make them release molecules attached to their surfaces (mainly H2O).

This is crucial to reach the vacuum levels required in the AEGIS experi-ment.

• It must work in a 1 T magnetic field. The 1 T magnetic field is an essential part of the Penning trap and the detector must be placed within the same magnet.

• It must work in a ultra high vacuum, between 10−15_{mbar and 10}−9_mbar.

The vacuum is necessary to keep unwanted elements out of the process and several different components also requires a high vacuum to operate. • It must fit in the AEGIS experiment chamber and has to be easy to mount. The space inside the 1 T magnet is limited and this must be taken into consideration.

• The development and material costs should be kept within the budget of the experiment.

The properties that are to be studied with the detector is the hydrogen pro-duction rate and the characteristics of that hydrogen. The main characteristic of interest is the temperature of the hydrogen since it will have a strong ef-fect on the collimation of the atom beam. Another interesting characteristic is the principle quantum number of the hydrogen which affects both the dipole momentum of the hydrogen (which is important for the Stark acceleration, see Appendix A.4) as well as the lifetime of the excited state. The use cases pre-sented below represent different measurements that will be done at different stages in the production. They are here presented in their chronological order in the experiment.

(29)

2.2. Objective 13

2.2.1 Direct detection of Rydberg atoms

The first use case is to directly detect Rydberg atoms without the use of the ionising rings. The ionising rings will not be able to ionise all the excited states and is expected to reach an ionisation rate of 25% for the n = 18 state[5]. It is therefore of interest to try to detect any hydrogen that has passed through the first stage of the detector. The main requirements for this are that

• The stage after the ionising rings should also be able to ionise more deeply bound Rydberg hydrogen to some degree.

• The detector should be able to detect single particles at a rate in the order of 10 atoms in the time span of 10 µs to 1 ms after the hydrogen production pulse. This is the approximate rate at which the experiment expects to produce hydrogen in the beginning minus the approximate number of atoms ionised.

• The events should preferably be timed but not necessarily matched with the imaged particles. The time of arrival gives the time of flight which in turn can be used to calculate the velocity of the particles and thus also be used to estimate their temperature.

2.2.2 Detection of ionised Rydberg atoms

In the current design of the ionising rings it will be possible to trap the electrons from the hydrogen ionisation. This gives the possibility to delay the arrival of these electrons and do a separate measurement from that of the un-ionised hydrogen. This requires the expected signal in the order of 20 electrons to be detected simultaneously after being released from the trap. Since they have been delayed, the arrival times of single particles is not as interesting as in the first case.

2.2.3 Time of flight analysis of trapped electrons

Once the hydrogen production is done there will be surplus electrons/positrons from one of the antiproton/proton cooling mechanisms inside the trap, these can be used to further analyse the production process and the environment it takes place in. The temperature of these electrons can be measured by slowly letting them out of the trap and then measuring the distribution in the time of flight of these electrons. To do this the detector needs to be able to precisely measure the arrival time of the first arriving particles with an exponentially increasing number of arrivals (with an expected total of 108_{electrons). No imaging of the}

particles is required for the time of flight measurement.

2.2.4 Imaging of electron plasma distribution

Another property to analyse is the spatial distribution of the electron plasma. This requires a spatial resolution that can distinguish 0.1 mm details and that

(30)

the equipment can handle the rate of 108 _{electrons within a time down to 100}

ns.

2.3 Project layout

From these requirements the possible options needed to be investigated and researched. The project started out by evaluating several different technolo-gies where the main part of them was discarded for different reasons. These technologies are discussed in Chapter 6. The evaluation then resulted in a can-didate system which needed to be investigated further, the main concern being the performance of MCPs at low temperature and high rates. The main topic covered in this thesis is the tests and analysis needed to understand the MCP performance in the setting of the experiment. In addition to this the other components that were decided upon after the evaluation phase, the phosphor screen and optical readout, are also discussed.

(31)

Chapter 3

Theory

This chapter presents relevant theory for the project, it mainly covers the prin-ciples of different components and also discusses different models. The chapter mainly covers microchannel plates and phosphor screens.

Microchannel plates and phosphor screens are commonly used components in particle detection and imaging. Both require additional hardware to create a complete detector and they can also be used together as part of a detector. The microchannel plate is a highly grained electron multiplier which behind it requires some sort of imaging device, either purely electrical by using for example a delay line detector or segmented Faraday cup, or by electron-photon conversion using a phosphor screen and then a regular camera readout. The phosphor screen acts as mentioned as an electron to photon converter and requires some sort of imaging device after it such as a CCD or photographic plate. With a correct setup the microchannel plate is able to multiply single particles while the phosphor screen requires several energetic electrons to produce enough light for imaging. However, together they can be used for single particle detection and imaging with high accuracy.

3.1 Microchannel plates

The microchannel plate (MCP) is a fine grained electron multiplying device which can be used for particle detectors with high spatial resolution. The prin-ciple of operation is schematically shown in Figure 3.1.

The multiplication process start with an impinging ionising particle, e.g. an electron, ion or photon, having enough enough kinetic energy to liberate one or more electrons from the wall of one of the MCPs channels. This impact releases electrons from the surface which are then accelerated by an electric field which is induced by applying a high voltage difference over the MCP. The accelerated electrons knock out more electrons from the channel walls; creating an avalanche of electrons which is then emitted from the back of the MCP. The gain factor (number of electrons emitted for each incoming particle) is usually in the range of 103_{to 10}5_{for a single MCP. This factor can be increased by stacking multiple}

(32)

Figure 3.1: A schematic picture of the structure of a MCP and its principle of operation. The channels are usually slightly angled from the perpendicular of the MCP to prevent particles from passing straight through. Image: Hama-matsu Photonics K.K.

(33)

3.1. Microchannel plates 17

MCPs on each other. The gain increase is, however, not linear to the number of MCPs but the method is known to be effective for stacks of up to three MCPs. There are several manufacturing methods for making MCPs, one method as specified in [6] is based on creating hexagonal arrays of microfibers which in turn can be fused together to larger sizes. These arrays of fibers are then cut at a slight angle, 8◦ _{to 15}◦_{, to give the channel bias angle. Afterwards the}

plate goes through different process to lower the work function of the channel surfaces. Lastly, the front and back of the plate are coated in a metal alloy to make the surfaces conducting and to improve detection efficiency. The main differences between different brands of MCPs are the materials used in the bulk of the plate, the surface coatings and the final processing method of the channel walls resulting in different work functions. It is also possible to improve the performance in certain applications by making thicker MCPs (for gamma detection) or applying different types of surface coatings (for example for X-ray detection).

Each channel is in the order of 2 − 25 µm in diameter and has a length to diameter ratio, L/d, in the order of 40 to 80 for standard MCPs. L/d and the secondary emission factor of the material inside the channel are the main factors that determine the gain of the MCP [6]. Since it is the L/d factor that affects the gain one can produce MCPs with smaller channel diameter while maintaining the same gain. This is relevant for extended dynamic range MCPs, which have smaller channel diameters to be able to handle larger input signals. The small channel diameter of MCPs gives a high position resolution making them a good choice in several high sensitivity imaging applications. Examples of these are camera sensors with high demands on low-light sensitivity such as high-speed or night-vision cameras.

3.1.1 Electrical model

One simple, yet effective, method of modelling an MCP considers each channel as an RC-circuit as was done by Fraser et. al. [7]. Characterising each channel with Rchand Cchthe time constant of the system will be τch= kRchCchwhere

k is constant. Analysing this structure one also finds that the time constant must be same as the time constant of the entire MCP, τM CP = τch.

Using some assumptions regarding the gain and operation of the MCP, Fraser arrives at the equation

Ip

Is

= G₁c(0)M RM CP/V0

N + kRM CPCM CP

where Ipis the output current, Isis the strip current, Gc(0) is the charge gain for

low count rates, M is the number of channels, V0is the applied voltage and N is

the count rate. This equation shows the saturation behaviour of the MCP, the ratio between the output pulse and the strip current of the MCP is limited for high rates. This means that a MCP with a limited current supply can at most output pulses corresponding to a certain fraction of the total current through the MCP. Hamamatsu [8] specifies this maximum ratio to approximately 7%.

(34)

3.1.2 Capacitance

One of the important properties of the MCP is the capacitance. It influences the RC time constant of the MCP but it can also be used to estimate the total charge available in a channel. Measuring the capacitance is usually complicated due to other capacitances affecting the measuring circuit and calculating it is usually precise enough and more consistent when comparing different MCPs. The capacitance of a MCP can be approximately calculated using a parallel plate capacitor model [7]

CM CP = 0(r(1 − aopen) + aopen)

AM CP

L

where 0is the vacuum permittivity, r is the relative permittivity of the glass

used, aopen is the open area ratio, AM CP is the area of the MCP and L is the

thickness. This formula assumes that the whole MCP is active, i.e. covered with channels. However, most MCPs have an additional rim not containing any channels but with the conductive coating. To calculate the capacitance of a MCP with a rim, the above expression has to be modified slightly by adding the capacitance of the rim (since it can be considered to be a parallel coupled capacitance). CM CP = 0(r(1 − aopen) + aopen) Aactive L + 0r Arim L

Since the MCPs used in this report are circular it is more convenient to use the total MCP diameter, dM CP, and active diameter, dactive, which are well

specified properties. CM CP = 0(r(1 − aopen) + aopen) πd2 active 4L + 0r π(d2 M CP − d2active) 4L

Using the values for a Hamamatsu F1217-01 and estimating the relative permit-tivity of the glass to be that of lead glass we arrive at at a total capacitance of approximately 300 pF, where the contribution from the active area is approxi-mately 150 pF, i.e. half of the total capacitance.

3.1.3 Resistance

There are three main materials that contribute to the resistivity of a MCP; the insulating bulk glass, the conductive surface coating and the semiconducting surface of the channels. The two surfaces can be considered as two resistances coupled in series around the parallel coupled bulk and channel coating. Due to the well insulating properties of the lead-glass in the bulk and the relatively high conductivity of the surfaces the MCP will mainly get its resistance be-haviour from the channel walls and thus behave as a semiconductor in terms of conductivity. The nature of semiconductors also determines the temperature dependence of the resistivity. Since the conductive properties of a semiconduc-tor mainly come from thermally excited electrons that move to the valence band the resistance should scale with the Boltzmann factor, e/kT_{. This gives}

semi-conductors an exponentially increasing resistance as the temperature decreases and at 0 K it is essentially a perfect insulator.

(35)

Earlier studies of MCPs by Roth and Fraser [9] suggest that there are regimes with different dominating conduction mechanisms which can be modelled using different model parameters. Still assuming that the resistance is dependent on the number of occupied states one can refine the previous suggested model by introducing different activation energies, , for different intervals of T . The data by Roth and Fraser suggests two different regimes giving a piecewise resistance function

R(T ) =

R1e1/kT , T ≤ T0

R2e2/kT , T > T0

with the restriction that the function is continuous.

3.1.4 Gain

The gain is the number of particles that are emitted for one electron being released from the wall at the beginning of the channel and it is one of the most important properties of a MCP. For single particle detection it is important that this number is high enough to generate sufficiently strong signals to be read out by standard electronics. It is therefore also essential to understand how the gain changes with temperature. The gain mainly depends on the channel width to length ratio, bias voltage, work function of the surface and the charge available in the channel.

Schecker et al. [10] has developed a rate and temperature dependent gain model starting from a previous model

G = (κEc)n

where κ is a constant, Ec the collisional energy of electrons to the channel wall

and n is the number of stages (cascades) down a channel, giving the average gain κEc per stage. Without consideration of rate and temperature this model

can be written as

G = (κV0 n )

n

where the parameters Ec = V2

0

4α2 and n =

4α2

V0 have been derived in earlier papers. α is the length to diameter ratio and is the electron energy when released from the wall.

To model the rate dependence the authors assume that the main charge deple-tion occurs in the last multiplicadeple-tion stage in the channel and that the charge in this stage has a capacitive recharge. The multiplication factor of the last stage is replaced as κEc → κEc(1 − e−dt/τ) where dt is the average time between

events and τ is the recharge time constant. This yields the new gain function G = (κV0

n )

n_{(1 − e}−dt/τ₎

where a rate dependence has been introduced via dt and a temperature via τ since we have τ = RchCch and Rch is a strongly temperature dependent

function. An important property from this model is, as long as we assume that there is no or only very weak temperature dependence in κ and n, that the low rate gain, dt → ∞, is temperature independent.

(36)

3.1.5 Time constant of channels and full MCP

To account for differences between the time constant of a perfect RC-circuit, τ = RC and the actual time constant Fraser [7] introduces an additional factor such that τ = kτch= kRchCch. k is an MCP dependent factor and τch can be

considered the natural RC time constant of the channel. Fraser assumes that the MCP can be considered an array of parallel RC circuits representing each channel, and under this assumption the natural time constant of an individual channel is equal to the time constant of the whole MCP. However, this does not take into account the non-active area of the MCP which mainly gives rise to additional capacitance for the entire MCP rather than affecting individual channels. Adding another capacitance, Crim, for the rim of the MCP yields

τch = RM CP(CM CP − Crim), i.e. it is the capacitance of the active area that

determines the channel time constant. This should be taken into consideration when determining or predicting the time constant of a MCP. One cannot directly use the time constant of the MCP in an electrical circuit as the time constant of individual channels. The time constant of the entire MCP is for example relevant when considering fast bias changes in a MCP system and the channel constant strongly affects the high rate behaviour of the MCP.

3.1.6 Dark counts

Undesired signals, known as dark counts, can stem from from residual ions in the vacuum (external) or, more commonly, from thermally emitted electrons (internal) from the MCP. The MCPs used in this report are manufactured by Hamamatsu Photonics K.K. and the internal dark counts are specified to be lower than 3 Hz/cm2 _{giving that the total dark counts should be lower than}

42 Hz with the currently used MCP (F1217-01) [8]. The specified dark count rate comes from several factors which can cause the MCP channels to fire by them self. An internal source of dark counts not accounted for in this number is the ion back-scattering, which can arise at high bias voltages. The effect comes from positive ions being knocked out, accelerated backwards in the channel and hitting the channel wall at a point close to entry, causing a new cascade in the channel. Electrically these events can be discerned from their pulse profile, giving a characteristic double peak. Using only a time-integrated image, as the one from a phosphor screen, this is much harder.

External dark counts can originate from radiation and gas as well as mechanical disturbances. MCPs are sensitive to UV-light, X-rays and gamma radiation, even though the detection efficiencies for these are considerably lower than for charged or heavy particles. They are still a noteworthy source of noise, especially in the AEGIS experiment since the hydrogen production utilises UV-lasers and there is 511 keV gamma radiation from the positronium decay. Alpha and beta radiation are clearly detectable by the MCP but in its operating environment there are no such sources and this should not cause a problem. Cosmic rays should in theory be detectable as well but are negligible due to their low rates. Sporadic gas molecules which ionise in the MCP channel can also be a source of noise, however, the ultra high vacuum and low temperature used in the experiment chamber reduces this source of noise considerably.

(37)

3.1. Microchannel plates 21 100 50 20 10 5 2 1 DETECTION EFFICIENCY (%)

ELECTRON ENERGY (keV)

0.05 0.1 0.2 0.5 1 2 5 10

0.02 20 50

RELATIVE DETECTION EFFICIENCY (a.u.)

PHOTON ENERGY (keV)

0.1 1.0 10 100

0 0.5 1.0

Figure 3.2: These two graphs show the non-trivial energy dependence on the detection efficiency. Electrons to the left and photons in the UV and X-ray range to the right. Image: Hamamatsu Photonics K.K.

3.1.7 Detection efficiencies

For non-charged massive particles the detection efficiency is clearly proportional to the open area ratio but also depends on factors such as kinetic energy, po-tential to ionise in the MCP, etc. For charged particles the detection efficiency depends on the electric field in and around the MCP, since the field lines can creep into the channels when there is a bias difference over the MCP.

In the case of detecting electrons and photons there are several different ef-fects that influence the detection efficiency. For electrons, for example, higher energies will make the electrons penetrate deeper into the bulk before transfer-ring their energy. Du to this, the secondary electrons are more likely to stop in the bulk rather than escaping the surface. For photons there are different phenomena through which energy can be transferred to electrons, for example the photoelectric effect, Thomson scattering and Compton scattering. That the detection efficiencies are non-linear with respect to the particles energies is also illustrated in Figure 3.2. The figure shows the detection efficiencies energy dependency for electrons and photons in selected ranges.

In the experiment it is also of interest to be able to detect Rydberg hydrogen atoms (i.e. excited hydrogen). Hydrogen normally requires very high electric fields to be ionised but the required field is significantly lower for excited states. Figure 3.3 shows how likely different states are to ionise in a 2 MV/m electric field which is the strength of the field inside a channel at full gain settings. However, since the particles are neutral they are not guided by the electric or magnetic field and the chance of a hydrogen atom entering a channel is approximately equal to the open area ratio of the MCP, 60%.

(38)

1 4 7 10 13 16 19 22 25 28 31 34 37 40 0 10 20 30 40 50 60 70 80 90 100

Percentage of n-state ionization in F =2.0e+06 V/m

Quantum number n P er ce n ta g e io n iz ed

Figure 3.3: (colour) The percentage of atoms that are ionised by an electric field equivalent to that in a MCP channel. Image: Olof Ahl´en/CERN

Type Range Efficiency [%]

Electrons 0.2 keV to 2 keV 50 - 85 2 keV to 50 keV 10 - 60 UV 300 ˚A - 1100 ˚A 5 - 15 1100 ˚A - 1500 ˚A 1 - 5 γ 511 keV 0.03 H* n ≤ 12 0 n = 13 11 n = 14 50 n ≥ 15 60

Table 3.1: Various detection efficiencies. Hydrogen detection assumes a 2 kV bias voltage across a chevron MCP stack

(39)

Parameter Value Unit

OAR 0.6 -dact 42 mm L 11 cm (α/ρ)shield 8.445 · 10−2 [11] cm2/g ρshield 2.699 [11] g/cm3 hshield 2 mm (α/ρ)M CP 1.429 · 10−1 [11] cm2/g ρM CP 6.220 [11] g/cm3 hM CP 1 mm N0 108 events

Table 3.2: MCP absorption parameters, values represent the Hamamatsu F1217-01 MCP used in the tests

3.1.8 MCP gamma events

In the hydrogen production process using positronium the main portion of the positronium will never interact with protons and simply annihilate, creating two 511 keV gamma photons. The positronium is very short lived and only produced for a short instant so that the annihilations result in a flash of gamma radiation. To estimate the number of events in the MCP that the gamma flash is expected to create, there are several factors to take into consideration. Factors such as the amount of absorbed gammas in the MCP, the type of absorption and how many detectable events that these create.

The number of gammas absorbed, Nabs, from an initial number N0 depends

mainly on three factors: the MCPs solid angle from the flash centre, attenuation due to shielding materials and absorption in the MCP. This gives

Nabs= (1 − OAR) d2 act 16L2e −αshieldhshield_{(1 − e}−αM CPhM CP_)N 0

where OAR is the open area ratio, dact is the active diameter of the MCP,

L is the distance between the MCP and the centre of the flash, αshield is the

attenuation coefficient for the shielding, hshield is the shield thickness, αM CP

is the attenuation coefficient for the MCP active area and hM CP is the MCP

thickness. It is assumed that all gammas travel perpendicular to the MCP surface. The attenuation coefficients are usually given as the attenuation per density unit, α/ρ so that α = α

ρρ. From the values in Table 3.1.8 we get

Nabs≈ 30000 which is about 0.4% of the 7350000 channels in total.

Secondly we will analyse the behaviour of the photons in the MCP bulk. The main interaction mechanism at these energies is Compton scattering, producing a spectra of electrons in the bulk of the MCP. These electrons will have a mean free path and a certain probability to knock out electrons from the surface of a channel and thus creating a cascade. The interaction is approximately equally likely at any point in the channel, creating an exponentially distributed pulse height spectrum. Only signals above a certain threshold give a significant output. A principle sketch of the distributions of dark counts and a standard signal charge can be seen in Figure 3.4.

(40)

Number of events

Pulse charge [C] Detection treshold

Figure 3.4: (colour) The charge spectrum from dark counts will be exponentially distributed (blue) in comparison to the gaussian distribution of a proper signal (red).

The maximum electron energy, the Compton edge, is given by ECompton=

2E2

mec2+ 2E

Since the photon energy comes from an electron/positron annihilation the en-ergy is the same as the rest enen-ergy for an electron, giving ECompton = 2/3 ·

mec2= 341 keV.

Using the NIST ESTAR data resource, the continuous slowing down approxima-tion range of 350 keV electrons is calculated to be 290 µm, which is significantly longer than the channel pitch and one can therefore assume that most of the absorbed photons will create a channel event.

Nevertheless, the interactions are equally distributed throughout the channel and therefore only a fraction create events that give signals that look like non-photon events. An estimation of the region that creates indistinguishable signals is the visible depth (looking straight at a channel) which for a 12 µm channel with a 8◦_{angle is 85 µm. Compared to the full length of channels in a chevron}

stack, 1 mm, this is 8.5%. A rough estimate is then that approximately 10%, 3000, of the absorbed gammas could be mistaken as hydrogen events.

3.1.9 Resolution

The resolution (size of discernible details) of the MCP is mainly determined by its channel pitch, which is 15 µm for the relevant MCPs.

(41)

3.2. Phosphor screens 25

The spread of light in the phosphor screen can be approximated by the contin-uous slowing down approximation range in phosphor which for 1 keV electrons is in the order of 1 µm, much lower than the size of the electron cloud from the MCP (which is at least 12 µm).

With the camera placed 170 mm from the last lens and a field of view of 70◦_,

each pixel covers approximately 230 µm. The camera resolution is the clear bottleneck of the system and actually gives a slightly worse resolution than specified in the requirements. Nevertheless it is within the right order and the performance loss is acceptable.

3.1.10 Gain change in magnetic fields

It is clear that strong magnetic fields have an impact on the behaviour of MCPs. This has been previously studied by Schecker et al. [10] and is also known by MCP manufacturers [8] which conclude that it is possible to operate MCPs in magnetic fields but with some restrictions. The main restrictions are the strength and orientation of the magnetic field compared to the MCP. Magnetic fields with non-perpendicular components compared to the MCPs surface will severely lower the gain of the MCP even at low field strengths. However, axially oriented fields, as is the case in the AEGIS experiment, can actually increase the gain up to a certain point. The properties of the MCPs used in the study by Schecker et al are similar to the ones used in this study and they found that at a 1 T axial field the gain is approximately the same as without a magnetic field. This suggests that any gain change due to magnetic fields in the AEGIS experiment should be minimal as long as the MCPs are placed with their surface perpendicular to the magnetic field lines.

3.2 Phosphor screens

Phosphor screens are commonly used for imaging charged particles. They con-sist of a phosphor covered glass plate with a metal coating on top, usually aluminium or indium-tin-oxide. By accelerating the electrons enough (to about 1 keV) they can pass through the 250 − 500 ˚A thick aluminium layer and in-teract with the phosphor. The electrons give rise to photo-luminescence in the phosphor; exciting electrons to levels above the conduction band. These non-radiatively relax to the conduction band of the phosphor and then non-radiatively relax into their ground state, emitting photons with a wavelength corresponding to the band gap. The kinetic excitation from the impinging electrons indicates that the output light is approximately linear (above a threshold) to both the number of incoming electrons and their energy. Aluminium is a common choice due to its reflective properties, allowing photons emitted back towards the anode to be reflected and thereby increasing the total light emitted.

(42)

(43)

Chapter 4

Methods

This section covers the methods used in the different parts of the project. In par-ticular, this section covers the different test setups used and how measurements were done.

The chapter is divided into five parts, the first four covering the main tests and the fifth covering the data analysis. The main test setups are presented in chronological order: first cryogenic measurements, measurements at room temperature using the same setup, second cryogenic measurements using a new setup, and room temperature tests with an optical readout added to the new setup.

4.1 First MCP low temperature test

To test the behaviour of MCPs at low temperatures a basic test setup was constructed and mounted to a cryocooler. In principle, as illustrated in Figure 4.1, the setup consists of an electron source that illuminates an area with known size on the MCP. The electrons are then multiplied in the MCP and emitted on to a Faraday cup from which the signal can be read out. The final setup used in the cryotests can be seen in Figure 4.2.

4.1.1 Setup details

The setup consisted of the following components • Chevron stack of Hamamatsu F1217-01 MCPs.

• Faraday cup made from a cleaned copper coated circuit board. • Electron gun, Kimball Physics ES-015 BaO-coated disc cathode. • 22Ω heating resistor.

• Cernox temperature sensor.

(44)

Al-screen with Ø2 mm hole Electron gun

Double stack MCP Faraday Cup

Figure 4.1: A schematic drawing of the setup used to test the MCPs • Internal holder and spacing rings in MACOR and aluminium oxide. • External holder in aluminium.

• Aluminium screen with a 2mm hole in the center.

• Heat bridge made of copper and electrically isolating polymer film. • Coaxial cables for Faraday cup, MCP and electron gun housing. • Cables for heater and electron gun filament.

The peripheral equipment used was

• Cryocooler coldhead, Sumitomo RDK-408D2. • High voltage PSU, CAEN N470.

• Linear Amplifier, TENNELEC TC205A. • Scaler and counter timer, CAEN N1145 • Picoammeter and voltage source, Keithley 487 • Sourcemeter, Keithley 2410 1100V

• Oscilloscope, LeCroy Wavepro 7100 • Programmable DC PSU, TTi TSX3510P • Waveform generator, Agilent 33250A

(45)

4.1. First MCP low temperature test 29

Figure 4.2: (colour) A picture of the setup used in the cryotests. From the left one can see the holder, the black MCP, a collimator screen and the electron gun. The Faraday cup is placed inside the holder behind the MCP. The heater, temperature sensor and ground connections are not in the photo.

(46)

• Programmable DC PSU for heater • Pfeiffer ion gauge pressure sensor • Lakeshore temperature readout

• Computer with LabView and GPIB interface

In addition to this a number of filters were applied to different in and outputs to reduce noise from various sources, mainly the cryocooler and the high voltage power supply. The wiring diagram is shown in Figure 4.3 with filters included.

4.1.2 Measurement complications

Initially there was a lot of noise in the setup coming both from electrical and mechanical sources. The electrical noise coming both from the high voltage power supply and the cryocooler were significantly reduced by introducing ap-propriately designed filters at the output of the sources. The mechanical noise, originating from the pump in the cryocooler, was harder to get rid off. However, due the low frequency of the mechanical noise, compared to the signals, it was still possible to measure the electrical with good precision by just adjusting the time resolution of the oscilloscope.

The objective of the measurement was to reach an as low temperature as possible (approximately 4 K using the cooler) but the setup only reached 67 K at its lowest and 70 K when operated with full gain and the electron gun. The problem in getting down in temperature was mainly due to a to strong heat bridge between the cold head and the setup and that the cables to the setup were not thermalised. This was noticed and improved for the second cryomeasurement.

4.1.3 Verifying bias difference for Faraday Cup

To determine the needed bias difference between the back of the MCP stack and the Faraday cup a test had to be made where the Faraday cup output was measured for different biases, keeping the rest of the setup constant. In our setup the Faraday cup was connected in such a way that the output was the count rate in number of events per second, which should become constant above a certain bias threshold. It was difficult to get a constant output from the electron source since it takes a long time for it to stabilise. During the measurements the electron output was slowly increasing but had a low flux. The values were registered manually and can be seen in Figure 4.4. From the figure it was concluded that a voltage difference of at least 100 V should be used to make sure that all signals go to the Faraday cup.

4.1.4 Measuring resistance over MCP versus temperature

To measure the resistance over the MCP stack it was necessary to measure the voltage or current while controlling the other. The resistance is then given by Ohms law. The most convenient method in our setup was to set a fixed

(47)

4.1. First MCP low temperature test 31 Electron NgunNfilament TTi KeithleyN48 7 100Nk Ω 4M7NnF KeithleyN48 7 KeithleyN24 10 50Nk Ω 50NnF 5NnF 100Nk Ω CAENNN8 70 PANCSA NtauN=N250Nns Signal Bias Output Test _CAENNN8 70 AgilentNwavef ormNgenerato r square Oscilloscope AgilentNLV Heate r 22N Ω 100NnF 100NnF 100NnF 100NnF 1NpF 80N Ω Electron NgunNhousing MCP FaradayNCu p Figure 4.3: The electrical diagram for the main tests together with filter sp ecifications

(48)

Figure 4.4: (colour) The increase in registered counts on the Faraday cup as its bias voltage compared to the back of the MCP increases. There is a slight increase in the MCP output with increase biased voltage due to the electron gun heating up while measuring. The error bars are defined as a percentage of the measured rate.

voltage over the stack and then measure the current output from the Keithley sourcemeter. The voltage and current were measured continuously while the apparatus was cooling down. There was, however, a risk that the resistance had a slight voltage dependence. Therefore the supply voltage was cycled up and down through the values 50, 100, 150, 200, 250 and 300 V, staying at each voltage for 60 s to allow any capacitors in the system to stabilise. The voltage, current, time, pressure and temperature were registered and stored over GPIB using a LabView program.

4.1.5 Measuring MCP gain

To measure the gain, the Faraday cup was connected to an amplifier which in turn was connected to an oscilloscope. With a low rate (order 10-50 kHz) it was possible to directly measure signals from single electrons via the pulse output from the MCP onto the Faraday Cup. Averaging these pulses for a certain supply voltage one can read out the pulse height from the oscilloscope. The Faraday cup signal was replaced with an artificial signal pulse from a waveform generator and the amplitude of the artificial pulse was tuned to the same level as the one previously read out from the Faraday cup. The pulse height from

A high sensitivity imaging detector for the study of the formation of (anti)hydrogen

Department of Physics, Chemistry and Biology

Master’s Thesis in Applied Physics

A high sensitivity imaging detector for the study of the

formation of (anti)hydrogen

Karl Berggren

CERN, Gen`

eve, Switzerland

A high sensitivity imaging detector for the study of the

formation of (anti)hydrogen

Datum

2013-09-10

Applied Physics

Department of Physics, Chemistry and Biology

Linköping University

URL för elektronisk version

http://urn.kb.se/resolve?

urn=urn:nbn:se:liu:diva-97359

ISBN

ISRN: LITH-IFM-A-EX--13/2827--SE

Abstract

Abstract in Swedish: Sammanfattning

Acknowledgements

Nomenclature

Symbols

Abbreviations

Contents

Chapter 1

Introduction

1.1

Reading instructions

1.2

CERN

1.2.1

The LHC and the accelerator complex

1.2.2

Experiments

1.3

AEGIS

1.4

The 1 T Penning trap

A-A ( 1 : 10 )

A

A

A-A ( 1 : 20 )

B ( 1 / 5 )

A

A

B

1.5

Purpose

Chapter 2

Objective and layout of

project

2.1

Initial state of development

2.2

Objective

2.2.1

Direct detection of Rydberg atoms

2.2.2

Detection of ionised Rydberg atoms

2.2.3

Time of flight analysis of trapped electrons

2.2.4

Imaging of electron plasma distribution

2.3

Project layout

Chapter 3

Theory

3.1

Microchannel plates

3.1.1

Electrical model

3.1.2

Capacitance

3.1.3

Resistance

3.1.4

Gain