ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2021

Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 2008

Materials analysis using MeV-ions:fundamental challenges and in-situapplications

KARIM-ALEXANDROS KANTRE

ISSN 1651-6214ISBN 978-91-513-1127-2urn:nbn:se:uu:diva-433483

Dissertation presented at Uppsala University to be publicly examined in Häggsalen, 10132,Ångströmlaboratoriet, Lägerhyddsvägen 1, Uppsala, Friday, 19 March 2021 at 09:15 forthe degree of Doctor of Philosophy. The examination will be conducted in English. Facultyexaminer: Prof. Dr. Hans-Christian Hofsäss (Georg August University of Göttingen).

AbstractKantre, K.-A. 2021. Materials analysis using MeV-ions: fundamental challenges and in-situ applications. Digital Comprehensive Summaries of Uppsala Dissertations from theFaculty of Science and Technology 2008. 73 pp. Uppsala: Acta Universitatis Upsaliensis.ISBN 978-91-513-1127-2.

The interaction of energetic ions with matter is highly relevant for a wide range of applications.Amongst them, material characterization employing ion beams is widely used due to itscapability of high-resolution composition depth profiling. The non-destructive nature of thesetechniques makes them appealing, although there are still several aspects which can beimproved and thus deserve attention. For example, better understanding of energy depositionof ions in matter, can improve simultaneous depth profiling of light and heavy atomic speciesin a single target. Also, the synthesis of advanced material systems requires complex, multi-step protocols. This situation creates an increased demand for in-situ material characterization,keeping the benefits of ion beam analysis. The present thesis addresses the above mentionedopen aspects which are of both fundamental and applied character.

First, the energy loss of heavy ions in solid matter, at energies relevant for recoil spectrometry,is investigated. The contribution of inelastic and elastic collisions of heavy ions to the totalenergy loss as well as the validity of the single scattering assumption are assessed. Thisanalysis is performed by a combination of experiments using different ions and target materialsand corresponding Monte Carlo simulations. A non-trivial dependence of elastic losses andtrajectory length on probing depth is found. These observations are not accounted for in severalcommon analysis packages and their implications for depth-profiling are discussed.

Second, the potential of tracking material modification processes, such as annealing orreactive thin film deposition, in-situ by MeV ion beams is investigated. An experimental setup,SIGMA - Set-up for In-situ Growth, Materials modification and Analysis, was constructed.SIGMA holds equipment for thin film growth, low energy ion implantation, sputtering,annealing and controlled exposure to reactive gases, while several ion beam analyticaltechniques are available.

Within this thesis, two studies associated with the rapidly growing field of sustainable energywere performed illustrating the capabilities of SIGMA. First, the growth of photochromicyttrium oxyhydride thin films was monitored in-situ. This study established the completesynthesis path of this material class and furthermore showed that the initial oxidation rateaffects the post oxidation rate and the persistence of the photochromic effect. In a secondinvestigation, deuterium implantation in tungsten was combined with in-situ ion beam analysisand thermal desorption spectroscopy to track the deuterium release during annealing. Theemployed combination of techniques permits to correlate depth-resolved information from ex-situ analysis with data accessible at operating future fusion devices.

Keywords: in-situ characterization; ion beam analysis; heavy ions; reactive thin film growth;materials modification; yttrium oxyhydrides; fuel retention

Karim-Alexandros Kantre, Department of Physics and Astronomy, Applied Nuclear Physics,Box 516, Uppsala University, SE-751 20 Uppsala, Sweden.

© Karim-Alexandros Kantre 2021

ISSN 1651-6214ISBN 978-91-513-1127-2urn:nbn:se:uu:diva-433483 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-433483)

To my parents

"Why don't you explain this to me like I'm five?"

Michael Scott

List of Papers

This thesis is based on the following papers, which are referred to in the text by their Roman numerals.

I Investigation of the energy loss of I in Au at energies below

the Bragg peak K. Kantre, V. Paneta and D. Primetzhofer. Nucl. Instruments Methods Phys. Res. Sect. B, 450 (2019) 37–42.

II Assessing electronic energy loss of heavy ions detected in re-flection geometry K. Kantre, M.V. Moro, V. Paneta and D. Primetzhofer. Ma-nuscript submitted for publication.

III SIGMA: A Set-up for In-situ Growth, Material modification and Analysis by ion beams K. Kantre, M.V. Moro, D. Moldarev, D. Johansson, D. Wessman, M. Wolff and D. Primetzhofer. Nucl. Instruments Methods Phys. Res. Sect. B, 463 (2020) 96–100.

IV Synthesis and in-situ characterization of photochromic yt-trium oxyhydride grown by reactive e−-beam evaporation K. Kantre, M.V. Moro, D. Moldarev, M. Wolff and D. Primetz-hofer. Scr. Mater. 186 (2020) 352–356.

V Combined in-situ ion beam analysis and thermal desorption spectroscopy for studying deuterium retention in tungsten K. Kantre, P.S. Szabo, M.V. Moro, C. Cupak, R. Stadlmayr, L. Zendejas Medina, F. Aumayr and D. Primetzhofer. In manuscript form.

Reprints were made with permission from the respective publishers.

My contributions to the included papers:

Paper I I participated in planning the study, performed all experiments and did the data analysis. I interpreted the results together with the other authors. I wrote the initial manuscript. Paper II I participated in planning the study, performed all experiments and did the data analysis. I interpreted the results together with the other authors. I wrote the initial manuscript. Paper III I participated in planning the study. I was involved in the construction of SIGMA as well as in the experiments. I did the data analysis. I interpreted the results together with the other authors. I wrote the initial manuscript. Paper IV I participated in planning the study, performed all experiments and did the data analysis. I interpreted the results together with the other authors. I wrote the initial manuscript. Paper V I participated in planning the study. I also participated in experiments and I was involved in the data analysis. I interpreted the results together with the other authors. I wrote the initial manuscript.

Contents

1  Introduction ......................................................................................... 13 

2  Ion-solid interactions ........................................................................... 17 2.1  Binary collisions ............................................................................. 17 

2.1.1  Elastic scattering kinematics .................................................. 17 2.1.2  Scattering cross section .......................................................... 18 2.1.3  Beyond the Coulomb barrier .................................................. 20 

2.2  Interactions with electrons .............................................................. 21 2.3  Energy loss of energetic ions in solid ............................................. 22 

2.3.1  Ion range distribution ............................................................. 25 

3  Ion beam analysis ................................................................................ 27 3.1  Backscattering spectrometry ........................................................... 27 3.2  Particle induced X-ray emission ..................................................... 30 3.3  Elastic recoil detection analysis ...................................................... 32 

3.3.1  Foil ERDA ............................................................................. 32 3.3.2  HIERDA ................................................................................ 33 

3.4  Nuclear reaction analysis ................................................................ 34 3.4.1  Particle-NRA ......................................................................... 34 3.4.2  γ-NRA .................................................................................... 35 

4  Material analysis by ion beams: Open fundamental aspects ............... 36 4.1  Assessing energy loss of heavy ions in reflection geometry ........... 37 

4.1.1  Extracting information on the energy loss from the SSM ..... 38 4.1.2  Deviation from SSM: The impact of multiple scattering ....... 38 

4.2  Resulting stopping cross sections and implications for HIERDA .. 41 

5  Material analysis by ion beams: In-situ characterization ..................... 44 5.1  Set-up for in-situ IBA: SIGMA ...................................................... 45 

5.1.1  Main chamber: Basic equipment ........................................... 47 5.1.2  Main chamber: Detector systems ........................................... 49 5.1.3  Automatization of data acquisition: SIDAS .......................... 50 

5.2  Experimental capabilities: Thin film growth, oxidation and in-situ characterization .................................................................................. 51 

5.2.1  Characterizing the reactive growth of photochromic YHxOy 51 5.2.2  Summary of results ................................................................ 54 

5.3  Experimental capabilities: Ion irradiation ....................................... 55 

5.3.1  Combination of in-situ ion implantation, annealing, IBA and thermal desorption spectroscopy ................................................... 55 5.3.2  Sputtering ............................................................................... 58 

6  Conclusions ......................................................................................... 59 

7  Sammanfattning på Svenska (Summary in Swedish) .......................... 61 

Acknowledgements ....................................................................................... 64 

Bibliography ................................................................................................. 65 

Abbreviations

e--beam Electron beam EBS Elastic Backscattering Spectrometry ERDA Elastic Recoil Detection Analysis HIERDA Heavy Ions ERDA HPGe High Purity Germanium IBA Ion Beam Analysis MEIS Medium Energy Ion Scattering NRA Nuclear Reaction Analysis PFC Plasma Facing Component PIGE Particle Induced Gamma-ray Emission PIPS Passivated Implanted Planar Silicon PIXE Particle Induced X-ray Emission RBS Rutherford Backscattering Spectrometry RGA Residual Gas Analysis SDD Silicon Drift Detector SEM Scanning Electron Microscopy SIDAS SIGMA Data Acquisition System SIGMA Set-up for In-situ Growth Material modification and AnalysisSSB Silicon Surface Barrier SSM Single Scattering Model TDS Thermal Desorption Spectroscopy ToF-E Time of Flight-Energy

13

1 Introduction

The ever-increasing human energy demand in a world of limited resources has led to an increased focus on the topic of sustainability in research and technol-ogy [1]. This concept has become important both in providing energy, e.g. in the form of electricity by a potential future power plant based on nuclear fu-sion [2] as well as in an effective use of energy.

Materials to be used for the plasma facing components in a fusion reactor have to withstand high temperatures and deterioration due to the continuous exposure to energetic particles escaping the plasma. The selection of the ap-propriate materials to be used as wall components in fusion reactors is a matter of ongoing studies [3,4].

In an effort to reduce costs for climatization or heating in energy efficient buildings, electrochromic and photochromic materials are developed for do-mestic, industrial and personal use [5]. Extensive research is conducted to de-velop materials which feature excellent optical properties, age resistance and reproducible behavior [6].

For both classes of materials, thin film [7] and near-surface material anal-ysis is of crucial importance. For example, the loss of fuel species in the wall material of a fusion device represents one of the major challenges in fusion research and the retention of the species at elevated temperature is investi-gated [8]. For candidate materials for smart windows, light species such as hydrogen or lithium are responsible for the change in the optical parame-ters [9,10].

In a much more general perspective, thin films are nowadays indispensable in a wide range of technological applications (for example in electronic de-vices) as they can monitor and modify the properties of surfaces and inter-faces. Accurate knowledge of the physical properties of these materials as well as their behavior during manufacturing and application is decisive. When in-vestigating their structure and composition, in particular, the depth distribu-tion of film constituents is highly relevant.

Increasing needs of near-surface material characterization lead to the de-velopment of a number of techniques capable of assessing near surface com-position and structure. The origin and basic concepts of these methods can be traced back to different fields of experimental physics. Some examples in-clude the characterization of the crystal structure by diffraction methods (e.g. by using X-ray [11], electron [12] or neutron [13] beams), as well as study of chemistry and composition by spectroscopic methods (for example X-ray

14

photoelectron [14], Auger [15] and energy dispersive X-ray [16] spectrosco-pies). At the same time, imaging techniques (for instance electron [17] and atomic force [18] microscopy) are also developed. While these methods have been extremely successful, they commonly cannot provide straightforward depth-resolved information on sample composition without extensive sample preparation or the sample being consumed in the analysis. Ion Beam Analysis (IBA) [19] provides commonly non-destructive analytical tools, which help to overcome these limitations [20].

The scattering of ions by atoms was first investigated by Ernest Rutherford and co-workers in the famous Geiger-Marsden experiment that revealed the existence of the atomic nucleus [21,22]. In their study, alpha particles emitted by a radioactive source, were found to be scattered at large angles by a thin gold foil and detected in a fluorescence screen. Rapid development and ad-vances in accelerator technology that took place in the mid-20th century, pro-vided laboratories around the world with access to ion beams with a wide va-riety of different ion species and energies. The former gave rise to new tech-niques for material characterization that are based on the interaction of ion beams with matter and the subsequent detection of the resulting particle and photon radiation. The earliest studies of this type date back to the 50s [23].

Nowadays, IBA is widely used as it can provide information on sample composition often in form of high-resolution depth profiles. The detection and identification of the energetic particles employed in these methods has become comparably easy by the use of semiconductor detectors. At the same time, ion-solid interaction in the MeV energy regime, where the majority of the IBA methods is operated, can be described by comparably simple theoretical mod-els. For example, a basic assumption which holds true for many cases is that the probing particles follow straight trajectories before and after a single bi-nary interaction with the atomic nucleus of the target [24].

However, despite this success some challenges of rather fundamental char-acter still deserve to be addressed, to make IBA an even more valuable set of tools. For example, more complex physical models, that take electron screen-ing and multiple/plural scattering into account, need to be considered for low energy light ion beams used for high-resolution near-surface depth profiling. The same applies in the case of heavy ion beams, in the - intermediate - energy range typically accessed by accelerators in medium scale laboratories, that are employed to detect light elements in a heavy matrix. Moreover, for improved accuracy of analysis, the availability and reliability of reference data on the experienced matrix energy loss providing depth perception has to be im-proved. Due to the entanglement of elastic scattering and inelastic excitation of the electronic system into a single observable, i.e. the energy loss along a trajectory, the two mentioned issues are closely related.

In course of its widespread application, IBA has been proven to be com-patible with the increased need of near surface analysis due to the ongoing physical miniaturization of devices. The non-destructive character of the

15

methodology renders it highly relevant, also for analysis of materials modifi-cation processes during synthesis and operation. However, the exposure of samples to ambient conditions during transport from the fabrication site to an IBA facility can be detrimental, as it might lead to unwanted alterations of the sample properties that can, in turn, complicate characterization attempts. For example, surface oxidation affects the subsequent reaction of materials to ther-mal loads or obscure the signature of a specific treatment. At least equally important is the fact, that ex-situ analysis permits only to characterize the final result of a process or alternatively requires a series of samples characterized upon extraction at individual process steps which comes again, with a risk for detrimental effects of the ambient. Furthermore, such ex-situ analysis will not straightforward permit insight in process steps at elevated temperatures as conditions will change until analysis is performed. This situation creates a steadily increasing demand for approaches that combine synthesis and analy-sis in-situ.

The aim of the present work is to address a series of open fundamental and applied questions in order to further extend the potential of IBA for material characterization. First, fundamental aspects of ion-solid interactions for heavy ions with a few to several tens of MeV, as commonly used for depth profiling of light constituents, are investigated. Specifically, Papers I and II study the impact of multiple and plural scattering on energy spectra obtained from IBA employing heavy ion beams. In these works, also electronic interactions of heavy ions are studied, with a focus on the relevance of the reflection geome-try as commonly employed in IBA, which permits to draw conclusions for quantitative depth-profiling of light species in heavy matrices. Addressing ap-plied challenges, Papers III, IV and V demonstrate a set-up for in-situ IBA during thin film growth and modification as well as two studies using this system and demonstrating the potential of in-situ IBA. Specifically, in the first study, we demonstrated monitored growth of yttrium hydrides and their mod-ification by controlled oxidation revealing the exact synthesis pathway of pho-tochromic oxygen-containing yttrium hydride films. In the second study, we investigated hydrogen isotope retention in tungsten in an approach combining IBA with Thermal Desorption Spectroscopy (TDS) [25] for in-situ irradiated samples.

All experiments in association with the present work have been performed using instrumentation at the 5 MV 15-SDH2 tandem accelerator at the Ång-ström Laboratory at Uppsala University [26].

This thesis summarizes the physical and methodological concepts of IBA as well as the key results of the performed studies. First, chapter 2 covers a concise description of the relevant physics governing the interaction of ener-getic ions with matter. Then, state-of-the-art methods in IBA are described in chapter 3. Finally, the scientific results of this thesis work are presented in chapters 4 and 5. Chapter 4 comprises a summary of the earlier mentioned fundamentals studies. In chapter 5, the instrumentation, which was

16

constructed for performing in-situ material modification and IBA, is described together with a series of exemplary investigations. The majority of these re-sults can be also found in the papers or manuscripts attached to this work.

17

2 Ion-solid interactions

Insight into the physical mechanisms of ion solid interactions is of fundamen-tal importance for the understanding of the present work. In this chapter, the physical processes relevant for energetic ions traversing solid matter, are out-lined. I start with the physics of the binary collision between two charged par-ticles, introducing the concept of scattering from target nuclei. Then the inter-actions with target electrons are discussed and, finally, the collective models that describe energy loss of the ion are introduced.

2.1 Binary collisions In the beginning of the present chapter, we will focus on elastic scattering kinematics and its application in a Coulomb potential resulting in the Ruther-ford scattering cross section. Subsequently, we will examine elastic and ine-lastic processes that occur beyond the Coulomb barrier.

2.1.1 Elastic scattering kinematics During the elastic scattering of the incoming ion by an initially stationary par-ticle, kinetic energy is transferred from the projectile to the recoiling particle. Let the mass of the projectile be 𝑀 and its initial energy 𝐸 , while the mass of the initially static particle is 𝑀 . As illustrated in Figure 2.1, after being elastically scattered at an angle 𝜃, with respect to initial direction, the energy 𝐸 of the projectile is decreased by a factor 𝑘 1 according to 𝐸 𝑘 𝐸 . At the same time, the target nucleus is recoiled into an angle 𝜑 90 with energy 𝐸 with 𝐸 𝑘 𝐸 .

18

Figure 2.1. Sketch of an elastic binary collision between an incoming particle with mass M1 and target atom with mass M2. After scattering the initial energy of the pro-jectile is distributed between the scattered particle and the recoil.

Independently of the type of the scattering potential, the kinematics of the elastically scattered and recoiled particles can be calculated considering con-servation of energy and momentum. In the laboratory frame, the kinematic factor of the scattered projectile is given by:

𝑘 ≡𝐸𝐸

𝑀 cos𝜃 𝑀 𝑀 sin 𝜃𝑀 𝑀

. 2.1

Note that, for 𝑀 𝑀 only the + sign is relevant and that 𝐸 is larger for projectiles scattered by heavier particles and smaller 𝜃.

Similarly, the kinematic factor for the recoil is:

𝑘 ≡𝐸𝐸

4𝑀 𝑀𝑀 𝑀

cos 𝜑 . 2.2

2.1.2 Scattering cross section The scattering angle 𝜃 for a given ion energy and ion-nucleus combination is defined uniquely by the impact parameter 𝑏 [27], i.e. the perpendicular dis-tance between the incoming ion path and the scattering center. As illustrated in Figure 2.2, particles with an impact parameter between 𝑏 and 𝑏 𝑑𝑏 are scattered at an angle between 𝜃 and 𝜃 𝑑𝜃. The area defined by the passing primary ions is 𝑑𝜎 2𝜋𝑏𝑑𝑏 and they are scattered into a solid angle 𝑑𝛺2𝜋 sin𝜃 𝑑𝜃. The probability of an ion being scattered into a certain solid angle element 𝑑𝛺 can then be expressed by the differential scattering cross section:

𝑑𝜎𝑑𝛺

𝑏sin𝜃

𝑑𝑏𝑑𝜃

, 2.3

thus, the differential cross section can be calculated for a known relation be-tween 𝑏 and 𝜃. This relation is characteristic of the interaction potential.

19

Figure 2.2. Schematic drawing of the dependence of the scattering angle on the impact parameter. The differential cross section is the probability of an ion to pass through dσ and thus being scattered into dΩ.

One can describe the potential of an atomic nucleus by combining the repul-sive Coulomb and the attractive nuclear forces. As shown in the - sche-matic - Figure 2.3, the electrostatic potential 𝑉 𝑍𝑒 /4𝜋𝜀 𝑟 is dominant at distances 𝑟 outside the nucleus with classical radius 𝑅 (𝑟 𝑅 ) and de-creases with increasing 𝑟, while the nuclear potential can be reasonably ap-proximated by a finite potential well.

Figure 2.3. Simplified illustration of the combined Coulomb and nuclear potential of the nucleus.

20

When the energy of the incoming ion is not large enough to surpass the so-called “Coulomb barrier”, the ion is elastically scattered by the Coulomb potential commonly referred to as Rutherford scattering. The Rutherford dif-ferential scattering cross section can be derived analytically [28] and is, in the laboratory frame, given by:

𝑑𝜎𝑑𝛺

𝛦,𝜃𝑍 𝑍 𝑒8𝜋𝜀 𝛦

1sin 𝜃

1𝛭𝛭 sin𝜃 cos𝜃

1𝛭𝛭 sin𝜃

. 2.4

The Rutherford differential cross section of a target nucleus to be recoiled at an angle 𝜑 is:

𝑑𝜎𝑑𝛺

𝑍 𝑍 𝑒4𝜋𝜀 𝐸

𝑀 𝑀𝑀

1cos 𝜑

. 2.5

2.1.3 Beyond the Coulomb barrier It is possible for a nucleus with sufficiently high energy to approach a nucleus close enough to face the attractive nuclear potential. The interaction with the nuclear potential can lead to the intermediate stage of the formation of a com-pound nucleus, which might be subsequently deexcited by particle and/or pho-ton emission.

In this situation, elastic scattering is still possible, in spite of the fact that the cross section of the interaction is no longer described by the Rutherford formula. Instead of that and depending on the ion-nucleus combination, nu-clear resonances might appear, where the scattering cross section exhibits maxima and minima at specific energies.

As an example, experimental cross section data [29–31] for the scattering of 4He by 16O at 170 degrees in the energy range from 2000 to 4000 keV are plotted in Figure 2.4 (open squares). For comparison, semi-empirical predic-tions from SigmaCalc [32] are also shown (blue solid line). The plotted values are normalized by the Rutherford cross section. It can be seen that above ~2300 keV the cross section is deviating from the Rutherford formula (equa-tion 2.4). The observed resonance at 3037 keV is widely used in IBA to en-hance detection sensitivity of 16O (see section 3.1).

21

Figure 2.4. Experimental scattering cross section data [29–31] and SigmaCalc [32] evaluation for the elastic scattering of 4He projectiles, with energy between 2000 and 4000 keV, scattered by 16O at 170 degrees. The values are normalized to the Ruther-ford cross section.

2.2 Interactions with electrons Coulomb interaction of a projectile ion with the electrons of the target can result in ionization or excitation of the target atoms due to the energy transfer to the electronic system. These processes lead, apart from the loss of energy of the incoming ion, to the emission of characteristic photons and elec-trons [33].

As it is illustrated in Figure 2.5, after the creation of an inner shell vacancy, the atom is deexcited and the vacancy is filled by an outer shell electron. As a result of this transition, a photon with energy equal to the energy difference of the two shells can be emitted (the alternative is transfer of the energy to an Auger electron). The characteristic X-rays are categorized as K, L etc. X-ray lines depending on the shell that is filled and are also divided in subgroups (Kα, Κβ, Lα etc.) according to the origin of the outer shell electron.

22

Figure 2.5. Schematic representation of an X-ray emission process caused by a pri-mary ion traversing matter and exciting the electronic system.

As the electrons are, in contrast to atomic nuclei, delocalized, electrons affect also the interaction between the primary ion and target nuclei. The scattering described in section 2.1 is a simplified picture where the incoming ions are sufficiently fast to not carry any electrons and/or the interaction distances are small enough for the ion to penetrate the electron cloud before being scattered by the nuclei. In the general case, the potential of the nucleus is screened by electrons. Thus, the interaction potential between the ion and the target atom should be written as following:

𝑉 𝑟𝑍 𝑍 𝑒4𝜋𝜀 𝑟

𝜑𝑟𝛼

, 2.6

where 𝜑 𝑟/𝛼 1 is a screening function and 𝛼 is the screening length. Sev-eral theoretical models have been developed to estimate the screening func-tion. The two most commonly used are the model of Thomas-Fermi-Mo-liere [34] and the universal potential by Ziegler-Biersack-Littmark [35]. Both models propose the same screening function 𝜑 𝑟/𝑎 but different screening lengths.

2.3 Energy loss of energetic ions in solid In the previous sections, we presented the physical principles governing bi-nary collisions between charged particles. An ion penetrating a solid is a sys-tem of much higher complexity as the ion undergoes a series of interactions

23

with the electrons and nuclei of the target. In the present section, we focus on the impact of these interactions on the ion energy. When an energetic ion traverses matter, it effectively experiences a resulting retarding force along its trajectory due to continual interactions. This retarding force is commonly re-ferred to as stopping power [36]. It is denoted by 𝑆 and expressed as the mean energy loss of the ion per unit path length:

𝑆𝑑𝐸𝑑𝑥

. 2.7

The stopping power due to the interactions of the projectile with bound or free electrons of the target is called electronic stopping power (𝑆 ), while nuclear stopping power (𝑆 ) is the result of the collisions of the projectile with the nuclei of the target, which, due to the 1/ sin 𝜃 dependence of the Rutherford cross section, typically lead only to small angle scattering. The total stopping power is given by their sum: 𝑆 𝑆 𝑆 .

A more convenient physical quantity that does not depend on the density of the material (𝑁) is the stopping cross section (𝜀), where:

𝜀 𝜀 𝜀1𝛮

𝑆 𝑆 . 2.8

The contribution of the nuclear stopping power to the total energy loss along the trajectory is significant only at small ion velocities and higher atomic num-bers (i.e. where the scattering cross section is increasing) and is neglectable at higher energies.

Even though a first attempt for a semi-classical theoretical model describ-ing the energy loss of fast charged particles in matter was done already by Niels Bohr in the beginning of the 20th century [37,38], there is still no con-sistent theory predicting the electronic stopping power for all energies. Dis-tinct models have been developed to describe the scaling of the stopping power in different velocity regimes.

For velocities significantly larger than the Bohr velocity, the stopping power decreases as the scattering probability of the penetrating ions is rapidly reduced. Bethe introduced a formula in 1930 [39], which was derived from quantum perturbation theory to first order and showed that the electronic stop-ping power is approximately proportional to 𝐸- :

𝑑𝐸𝑑𝑥

4𝜋𝑍 𝑒 𝑛𝑚 𝑣

𝑒4𝜋𝜀

ln2𝑚 𝑣

𝐼, 2.9

where 𝑣 and 𝑍 are the ion velocity and charge respectively, 𝑚 is the electron mass, 𝑛 the electron density and 𝐼 the mean ionization potential.

24

The Bethe formula was later modified to include corrections to the mean ionization potential [40] and higher order perturbation theory [40–43].

At ion velocities on the order of the electron velocities in the solid, the ion captures electrons. As a consequence, the scattering probability with other charged particles is affected. A stopping power model in this energy range thus needs to take charge state effects into account [44,45]. Additionally, due to the lower maximum energy transfer given by scattering kinematics and only discrete atomic orbitals, limitations occur with respect to which electrons can be excited. At low enough energies, the energy loss is dominated by excitation of the - loosely bound - valence and conduction electrons. Thus, in many cases, the electron system of the target can be described by a free electron gas [46] which would yield a stopping power approximately proportional to the ion velocity [47].

In the energy region where these two models meet, the stopping power ex-hibits a maximum referred to as the Bragg peak [48]. As an example, the en-ergy dependence of the electronic and nuclear stopping power of helium ions penetrating yttrium is plotted in Figure 2.6.

For energies above the Bragg peak, the stopping power of an ion in a com-pound is in good approximation predicted by Bragg’s rule, i.e. the weighted sum of the stopping power of the ion in the constituents of the penetrated ma-terial [48].

Accurate knowledge of stopping power is of crucial importance for a wide range of applications, as e.g. radiotherapy employing ion beams for cancer treatment (e.g. proton or carbon therapy) [49,50]. In these applications, the beam energy and species are selected in order to maximize the delivered dose, i.e. at the Bragg peak, at the depth of the tumor.

In the absence of a consistent theoretical model for the entire energy range and/or heavier ions, the scientific community has to rely on accurate experi-mental data. While for certain combinations, in particularly light ions and no-ble metals and common materials like Al, Si or C a substantial amount of tab-ulated experimental stopping powers is available, in many other cases the ex-isting data are scarce if not totally absent [51]. Therefore, data predicted by semi-empirical models like SRIM [52] are often used in applications, which may - in turn - deviate significantly from experimental data [53].

25

Figure 2.6. Electronic and nuclear stopping cross sections of helium ions in yttrium according to SRIM-2013 [52]. The dependence of the stopping power on the ion en-ergy is shown and the two different regions forming the Bragg peak are indicated.

2.3.1 Ion range distribution The motion of energetic ions in a solid continues as long as it takes to deposit their entire excess energy in the material and reach thermal equilibrium. The mean travelled path length of an ion while slowing down is given by:

𝑅1𝑑𝐸𝑑𝑥

𝑑𝐸 , 2.10

where, 𝐸 and 𝐸 are the initial and final energies of the ion, respectively. Given that the ion path deviates from a straight trajectory, the total distance

experienced by the ion, which is calculated by equation 2.10, exceeds the pen-etration depth, i.e. the projected range 𝑅 . This observation is more prominent at lower energies as multiple scattering results in a significant deviation of the ion trajectories from straight lines.

Individual ions of the same initial energy follow different paths in the ma-terial, due to the stochastic nature of the energy loss processes. It is thus useful to refer to the mean projected range 𝑅 and the projected range straggling 𝛥𝑅 , which is given by the width of the resulting distribution of the projected ion ranges.

26

This concept is commonly used for the modification of a material by ion implantation [54], where the initial energy of the ions is selected according to the desired implantation depth. The depth distribution of the implanted ions is therefore given by the gaussian:

𝑁 𝑥𝜑

𝛥𝑅 √2𝜋exp

12

𝑥 𝑅𝛥𝑅

, 2.11

here normalized by the ion fluence 𝜑 . Note, that the deflection mentioned above also results in an average lateral

distribution around the impact site, which however due to the - often macro-scopic - beam size is less relevant in applications.

An example of resulting implantation profiles is shown in Figure 2.7. The irradiation of tungsten by hydrogen isotopes at several energies between 1 and 100 keV was simulated in TRIM and nominal implantation profiles for a total dose of 1017 ions/cm2 are plotted. Note that diffusion in the material during and after implantation is not considered.

Figure 2.7. Implantation profiles for hydrogen and deuterium ions of different ener-gies in tungsten as calculated by the Monte Carlo simulation code TRIM illustrating the stochastic depth distribution of implanted ions resulting in a characteristic average range.

27

3 Ion beam analysis

The detection of charged particle and photon radiation resulting from ion-solid interaction is the basis of material characterization by ion beams. The pro-cesses relevant for IBA have been discussed in the previous chapter and are summarized in Figure 3.1. In the present chapter, the most commonly used IBA techniques as well as their advantages and potential weaknesses are pre-sented. Starting from backscattering spectrometry techniques, we discuss methods that involve atomic excitation, forward recoil spectrometry and the detection of nuclear reaction products.

Figure 3.1. Schematic illustration showing the different processes that can take place in a material when it is irradiated by a beam of energetic ions.

3.1 Backscattering spectrometry Rutherford Backscattering Spectrometry (RBS) [24] is the simplest IBA method and involves detection of ions scattered by the target atoms at a backscattering angle (>90 degrees) with respect to the incident beam direction. An angle larger than 150 degrees is usually selected for better mass separation.

28

Typically, a beam of 4He ions with primary energy on the order of ~2 MeV is used. RBS shows high sensitivity to the detection of heavier elements and is widely used for the determination of near-surface stoichiometry and thin film thickness.

As illustrated in Figure 3.2, the energy of the detected ions scattered by a certain element has a maximum value of 𝐸 𝑘𝐸 that corresponds to backscattered ions originating from the target surface (see scattering kinemat-ics in section 2.1.1). The ions scattered deeper in the sample (in depth 𝑥) are detected with an energy 𝐸 𝐸 , due to the energy loss, on their ways in and out of the target. This applies for every element, heavier than the projectile, present in the target within the probing depth of the ion beam.

A basic assumption in RBS (as well as in most IBA methods) is that of the Single Scattering Model (SSM). As seen in equation 2.4, the underlying Ruth-erford cross section is decreasing for both increasing energy and scattering angle. Large angle scattering is a rare event in the MeV energy range. It is thus often justified to assume that the primary ions are only scattered once by the target nuclei before being detected with their trajectories otherwise unper-turbed. As a result, depth information can be straightforwardly extracted from the energy spectrum. The areal density 𝑁 of a thin film can, given sufficiently high detector resolution 𝛿𝛦, be deduced from the width 𝛥𝛦 of the peak signal as following:

𝛥𝛦 𝛮 𝜀 , 3.1

where 𝜀 is the stopping cross section factor:

𝜀1

cos 𝛼𝑘𝜀 𝛦

1cos 𝛽

𝜀 𝐸 , 3.2

and 𝛼 and 𝛽 are the incident and exit angles respectively and 𝑘 is the scattering kinematic factor.

Note that the stopping cross section factor is a function of the stopping cross sections experienced by ions entering 𝜀 𝛦 and exiting the material 𝜀 𝛦 . In the near surface approximation [24] which is commonly used for suffi-ciently thin films, the energy before scattering is assumed to be equal to the energy of the primary beam and the outcoming energy equal to the ion energy immediately after scattering (𝛦 𝐸 and 𝐸 𝑘𝐸 ). In the case of thicker films, the mean energy approximation is used with 𝛦 𝐸 𝐸 𝑡 /2 and 𝐸 𝑘𝐸 𝑡 𝐸 /2, where 𝑡 is the film thickness.

As seen in Figure 3.2, the RBS spectrum of a thin film with composition AmBn (with M M ) consists of two peaks, each of them corresponding to a different element (A and B), while their width is correlated to the film thick-ness and the stopping power of the ions in the material. A bulk sample with

29

infinite thickness for the beam (the substrate in the case of the illustrated ex-ample) appears as a stepwise signal starting from the energy corresponding to ions scattered from its surface and extending to the low energy channels.

Let 𝑌 and 𝑌 be the total number of detected particles corresponding to elements A and B respectively, the areal density of each element is then given by:

𝑁𝑡 ,𝑌 , cos𝛼

𝑄𝛺𝜎 , 𝐸,𝜃 , 3.3

where 𝑁 is the atomic density of the material, 𝑡 the thickness, 𝛺 the detector solid angle, 𝑄 the number of incident ions, 𝛼 the incident angle and 𝜎 the scattering cross section.

Note that the stoichiometric ratio of the compound does only depend on the ratio of the yield integrals 𝑌 and 𝑌 and the scattering cross sections, as it is:

𝑛𝑚

𝑁𝑁

𝑌 𝜎 𝛦,𝜃𝑌 𝜎 𝛦,𝜃

. 3.4

Figure 3.2. Typical RBS spectrum for a thin film containing two elements of different masses deposited on a light element substrate. The energy of the primary beam ions is E0. The widths of the peaks A and B are determined by the thickness and the stop-ping power of the ions in the film. A scattering by a nucleus A or B from depth x results in a signal contribution to the middle of the respective peak.

30

Even though RBS is a powerful technique with a broad field of applications, it is challenged when it comes to its sensitivity for light elements. The Ruth-erford cross section decreases with decreasing atomic number according to dσ/dΩ proportionality to 𝑍 , while, at the same time, signals from light ele-ments tend to overlap in the low energy channels of the RBS energy spectrum. At the same time, mass resolution is decreasing for heavier elements. To over-come these limitations, a set of complementary methods have been developed.

Amongst them, and most similar to RBS is a variation of the technique, known by the more general term Elastic Backscattering Spectrometry (EBS), focusing on the detection of elastically scattered ions including energies be-yond the Coulomb barrier (see section 2.1.3). It is based on the fact that at these energies scattering cross sections can differ significantly from Ruther-ford’s formula and may exhibit narrow resonances. The latter can be used to enhance the sensitivity for light elements. One of the most widely employed examples involves the narrow nuclear resonance (≈10 keV) of 16O(4He,4He)16O at 3037 keV [29]. In this case, the scattering cross section is ~30 times higher than predicted from the Rutherford formula at a scattering angle of 170 degrees with respect to the beam direction as shown in Figure 2.4. A sufficiently narrow resonance can also enhance depth profiling capa-bilities, as one can probe the amount of the element of interest (in the previous example 16O) in different depths by changing the beam energy.

Moreover, since EBS is identical in basic principles and experimental set-up to RBS, enhancing the sensitivity for specific light isotopes, information typical for RBS on the heavier constituents of the target, is commonly still contained in the spectra. Thus, the two techniques can be combined in one measurement.

3.2 Particle induced X-ray emission In Particle Induced X-ray Emission (PIXE) [55], the focus is on the detection of characteristic X-rays emitted from the sample (the process is described in section 2.2). PIXE is typically performed using beams of light ions at energies of a few MeV down to the range of hundreds of keV (low energy PIXE [56]).

The probing depth of PIXE depends on the range of penetration of the beam ions in the material, as well as the X-ray attenuation coefficient in the sample. However, the technique does not provide direct depth information while indi-rect information is available by adjusting the penetration depth of the primary ions. On the other hand, for bulk samples, it can be used to deduce quantitative information for the sample composition from the intensity of the measured X-ray lines. The yield of a peak corresponding to a characteristic X-ray line of an element present in a sample is:

31

𝑌 𝑁 𝜎 𝛦, 𝜃 𝑄𝛥𝛺𝜀, 3.5

where 𝑁 is the areal density of the element, 𝑄 the number of incident ions, 𝜎 the cross section for X-ray production, 𝛥𝛺 the solid angle and ε the detection efficiency.

An example of a PIXE spectrum is shown in Figure 3.3. A sample from the experimental project described in Paper IV and section 5.2 (YHxOy thin film deposited on CaF2) is measured by employing a 3060 keV 4He beam. The characteristic K and L lines of Y and substrate Ca are detected. Note that, light elements are not detectable by PIXE due to the absorption of low energy X-rays in the detector window.

Figure 3.3. A typical PIXE spectrum obtained by using a 3060 keV 4He beam to irra-diate an YHxOy thin film deposited on CaF2 substrate.

PIXE is often used as a complementary technique to RBS and can help to overcome some of the RBS limitations. More specifically, information ob-tained by PIXE is used to enhance RBS mass separation sensitivity as ele-ments with similar masses can be identified by their characteristic X-rays [57]. In addition, PIXE spectra can reveal the presence of trace elements in amounts that are hard to detect by RBS.

32

3.3 Elastic recoil detection analysis As mentioned, RBS and PIXE feature low sensitivity for light elements. Es-pecially in case of target nuclei equal or lower in mass than the projectile ion (for example hydrogen), these remain undetectable by RBS. Even though EBS can increase sensitivity for light elements in certain cases, it has the disad-vantage that it can be sufficiently applied only for specific isotopes. An alter-native, potentially overcoming these issues, is offered by Elastic Recoil De-tection Analysis (ERDA) [58], an IBA technique employed for the detection of light elements in heavy matrices.

The major difference between ERDA and RBS is that in ERDA the focus is on detection of recoiled species instead of scattered ions. Experiments are performed in forward geometry, with the detector placed at an angle smaller than 90 degrees with respect to the incident beam direction. The assumptions from the SSM, as discussed in section 3.1, are also typically used in the anal-ysis of ERDA data, i.e. a single binary collision and straight trajectories for the primary ion and the recoil.

Accurate knowledge of the stopping power is of utmost importance for ERDA depth profiling and quantification. Indeed, the choice of a stopping power model and its implementation in the analysis of a multi-elemental sam-ple affects ERDA measurements significantly more than corresponding RBS measurements. The stopping power of the recoils in the sample, in addition to that of the primary ions, is also of interest in ERDA. As a result, in the case of bulk samples, the distinct energy loss processes experienced by different par-ticles (scattered ions and recoil species) affect the obtained depth information.

There are two fundamentally different approaches in ERDA, where the set-up is either designed to measure a specific light element (usually hydrogen isotopes), i.e., foil-ERDA, or to allow simultaneous multi-element profiling, i.e., Heavy Ion ERDA (HIERDA).

3.3.1 Foil ERDA In its simplest approach, ERDA is performed by using a single energy detector as for example a Silicon Surface Barrier (SSB) [59,60] or a Passivated Im-planted Planar Silicon (PIPS) [61,62] detector. An absorber foil is put in front of the detector with its thickness selected in order to let the recoils of interest pass but stop other recoil species and, in particular, the large number of for-ward-scattered particles. This type of ERDA is commonly performed by em-ploying a helium beam as a complementary method to RBS, i.e. by using a second energy detector in suitable geometry, mainly for hydrogen detection. On the one hand, this approach requires a simple experimental setting and provides 100% detection efficiency. On the other hand, it is only sensitive to specific recoil species.

33

3.3.2 HIERDA In this approach, a beam of heavy ions in the energy range of tens of MeV is used in combination with coincidence detection systems that allow mass dis-crimination of the detected ion species.

An example of a more sophisticated detection system is Time-of-Flight - Energy coincidence (ToF-E) HIERDA, in which the velocity of both, recoils and scattered ions, is first measured by a ToF detector, to be followed by the detection of the particles by an energy detector (usually solid state or ioniza-tion chamber). The acquired data are plotted in 3D histograms that contain information on both the energy and the velocity of the detected particles. Spe-cies of different masses are separated in the spectra due to the trivial 𝐸𝑚𝑣 /2 dependence.

A typical ToF-E HIERDA coincidence spectrum for a YHxOy thin film de-posited on CaF2 is shown in Figure 3.4. The primary beam is 36 MeV iodine and the particles are detected at 45 degrees with respect to the beam direction. Signal separation due to the different masses of the detected particles results in the characteristic “banana-shaped” curves. Each curve corresponds to a dis-tinct mass. Even though all elements present in the sample are detected in a single measurement, the detection efficiency of the ToF detector for light ele-ments (in particular hydrogen) is low. The ToF-E HIERDA measurements discussed in the present work were performed at the set-up presented in [63].

Indicatively, some typical examples of ion beams that have been used for performing HIERDA in different IBA laboratories around the world include 23 MeV and 36 MeV iodine [57,64] as well as 50 MeV copper [65].

34

Figure 3.4. A typical ToF-E ERDA spectrum of an YHxOy thin film deposited on CaF2 with recoils created by 36 MeV iodine primary ions. Coincident detection of the time of flight and energy of the emitted particles allows mass discrimination. Each curve corresponds to a specific isotope.

3.4 Nuclear reaction analysis Another IBA technique used for quantification and depth profiling of light isotopes is Nuclear Reaction Analysis (NRA). In this case, the employed beam (ion species and energy) is selected in order to induce a nuclear reaction be-tween the primary ions and the nuclei of the element of interest. The nuclear reaction products are subsequently detected. Depending on whether the de-tected species are charged particles or photons, the method is divided into two kinds: particle-NRA and γ-NRA.

NRA is commonly depth sensitive due to the energy loss of either the pri-mary ions and/or the reaction products in the sample. In this context, the nu-clear reactions which exhibit narrow resonances are even more interesting, as high-resolution depth profiling can be achieved by scanning the energy of the ion beam.

3.4.1 Particle-NRA Light ions such as protons or alphas are the most common reaction products from nuclear reactions occurring in the energy range relevant for IBA studies. In particle-NRA, the reaction products are detected by a particle detector.

35

The energy of the detected particles depends on the primary ion energy prior to the reaction. This means that for non-resonant reactions, depth profile information can be extracted from a single spectrum in a data evaluation pro-cess similar to the one applied in the analysis of RBS spectra.

An absorber foil can be placed in front of the detector. As it was discussed in the case of foil-ERDA, the purpose for this is to achieve better separation between the typically strong signal from scattered ions (and - depending on the angle - recoils) and the nuclear reaction products.

A typical example of a particle-NRA reaction is that of D(3He,p)4He, which is widely used for deuterium depth profiling in solid samples [66].

3.4.2 γ-NRA A second type of NRA is the one performed by detecting gamma-rays pro-duced during the deexcitation of the compound nucleus that is formed by the reaction. This approach is also commonly referred to in literature as Particle Induced Gamma-ray Emission (PIGE).

One of the most widespread applications of γ-NRA is hydrogen depth pro-filing by using the very narrow resonance of 1H(15N,αγ)12C at 6.385 MeV [67].

36

4 Material analysis by ion beams: Open fundamental aspects

This chapter is largely based on results presented in Papers I and II. As dis-cussed in section 3.3.2, ERDA employing heavy ions, in particular in the form of ToF-E HIERDA, is a powerful IBA technique due to its high sensitivity for light elements combined with information on the overall composition by sim-ultaneous detection of heavy constituents. However, there are several chal-lenges that can limit the analytical potential of HIERDA. The achievable ac-curacy in composition measurements for bulk samples is limited by the ab-sence of accurate models for the electronic stopping power, a general lack of experimental reference data, as well as often considerable uncertainties for available datasets.

As mentioned in section 3.3, knowledge of stopping power is of crucial im-portance for HIERDA as both primary ion and recoil stopping power enter equations for the scattering yield. Experimental reference data for heavier ions is, however, particularly scarce. For example, the stopping power of two very common primary ions in HIERDA, i.e. copper and iodine, has been measured only for 12 and 10 elements and compounds respectively [51]. At the same time, HIERDA is typically performed at ion energies below the Bragg peak, where theoretical models are additionally challenged by the complex structure of the projectile. Also, Bragg’s rule for estimating the stopping power of com-pounds is not expected to be highly accurate [68,69]. As a result, the most commonly employed option in applications are tabulated data provided by SRIM, which is extrapolating the limited available experimental datasets.

An additional potential source of systematic errors is the expected charge state dependency of electronic stopping [70]. Electronic stopping is expected to scale with effective charge of the ion species [36,71]. In this context, dif-ferences have been observed between the primary charge state distribution of the recoil species immediately after scattering and after traversing some dis-tance in the material, where the equilibrium charge states are obtained [72]. The same applies for primary beam ions entering the material with charge states far from equilibrium. As a result, the detected scattered ions and recoils may experience a significantly altered average stopping power force for scat-tering events occurring in the near-surface regions of the sample.

Additionally, different scattering angles may affect the charge state and in-elastic losses in the collision. Thus, higher mean charge states have to be

37

expected along trajectories which feature shorter interaction distances be-tween ion and target atoms. The latter indicates a potential dependence of the measured stopping power on the experimental geometry. Note, that vice-versa one may expect a non-trivial dependence of the energy loss resulting from scattering in a certain depth of a thin film. At present, the scarce reference data available for heavy ion stopping power are almost exclusively measured in transmission, while ERDA is performed in reflection geometry, with signifi-cant deflection angles of typically around 45 degrees.

At the same time and due to higher scattering cross sections, the impact of multiple and plural scattering is much more pronounced in HIERDA - at ion energies accessible by moderate size machines -, in comparison to IBA meth-ods involving light ions, affecting experimental energy spectra. An analytical approximation has been developed for estimation of the multiple scattering contribution [73,74] that is also available as a calculation option in the soft-ware package SIMNRA [75]. However, a deeper investigation and high accu-racy of data analysis can only be achieved by Monte Carlo simulations.

In the present work, energy loss measurements for heavy ions in noble met-als were performed in an effort to address some of these issues. The experi-ments were - to the best of our knowledge - for the first time performed in a geometry relevant for ERDA studies. At the same time, simulations revealed the impact of multiple and plural scattering on the ion trajectories and the shape of the acquired spectra and were used to subtract their contribution to energy loss from the experimental data.

4.1 Assessing energy loss of heavy ions in reflection geometry

In transmission geometry, the energy loss along the ion trajectory is simply given by the difference between the primary and the detected ion energies. This approach, for a given energy, minimizes the plural and multiple scatter-ing contribution as the vast majority of the detected ions follow straight or almost straight trajectories in the target.

The stopping power experienced by the ions can, however, be also assessed in backscattering or forward scattering geometries extracting data from trajec-tories which are far more realistic in IBA using heavier ions. These trajecto-ries, however, include at least one large angle scattering event, and thus in any case significant elastic losses on top of the energy loss along the trajectories. We used the SSM as a starting point that we further improved by Monte Carlo simulations to measure the stopping power of iodine and bromine in the low energy range.

Specifically, in Paper I we introduced the subject and made a first attempt to deduce the stopping power of iodine in gold by two different approaches.

38

This analysis allows to assess how the relation between observed energy loss and depth for scattering and creation of recoils evolves as a function of ion energy. In Paper II, we performed complementary experiments for bromine projectiles in silver and further deepened the analysis. We also analyzed the relevance of our findings on HIERDA analysis as software packages com-monly used for data evaluation, e.g CONTES [76] and Potku [77], are based on the SSM. In the next sections, the employed methodology and the most important results are summarized.

4.1.1 Extracting information on the energy loss from the SSM The principles of the SSM were described in section 3.1. As shown in equation 3.1, the width of energy spectra is proportional to the stopping cross section factor 𝜀 and the thickness of the film. As such, the total electronic energy loss, electronic stopping cross section and film thickness are strongly entan-gled quantities. Thus, 𝜀 can be calculated by extracting the total inelastic energy loss and knowing the areal density of the film (e.g. from RBS). From 𝜀 , one can extract pairs of stopping cross section data (𝜀 and 𝜀 of equa-

tion 3.2) assuming an energy dependence of obtained data. This energy de-pendence can for example be obtained by assuming a similar energy scaling as in SRIM. Also, assuming no energy scaling, the correct scaling can be ob-tained following an iterative procedure, if data is deduced in a wide enough energy range [78]. An assessment of the resulting stopping cross sections in a wide energy range provides, due to the above-mentioned entanglement, a test case for the employed evaluation model, here the SSM.

4.1.2 Deviation from SSM: The impact of multiple scattering To investigate the deviation of typical trajectories, for heavy ions in ToF-E HIERDA, from the assumptions made in SSM, full ion trajectories of iodine and bromine in gold and silver were simulated by the Monte Carlo package TRIM [35]. The selected ion species and energies (with a range from 4 to 36 MeV) as well as the geometry (similar to the set-up described in [63], i.e. 45 degrees detection angle with symmetrical incidence and exit angle of 67.5 degrees) have been chosen due to their relevance for HIERDA.

For each individual ion the path length and the nuclear and electronic con-tributions to the total energy loss were calculated. The calculations were done in a custom MATLAB script [79] by making use of the TRIM collision output files.

In single scattering, the trajectory length (𝑙) of scattered ions is:

39

𝑙𝑑

cos 𝑎𝑑

cos 𝛽, 4.1

where 𝛼 and 𝛽 are the incident and exit angles respectively and 𝑑 is the max-imum depth the ion reached in the material, i.e. where the single large scatter-ing event takes place.

As a first step, in Paper I we show a comparison of Monte Carlo data of reflection and complementary transmission geometries, i.e. where the target thickness in the transmission simulation is equal to the maximum SSM path length. As seen in Paper I-Figure 2, even in the absence of a large scattering event, low energy (4 MeV) iodine exhibits significant elastic losses in gold, which are, however, significantly lower than predicted from tabulated values for nuclear stopping in this system. Accordingly, the followed path lengths are only slightly larger than the straight trajectory. The vast majority of the ions follow a path length less than 3% larger than the film thickness. Together, these results, illustrate the trajectory selectivity of the experimental geometry, which hampers the use of tabulated values for the stopping power in data anal-ysis.

A comparison of calculated path lengths of high and low energy ions to predictions from the SSM for reflection geometry, is plotted in Paper I-Figure 5. The trajectory length of iodine ions in gold is plotted as a function of the maximum depth reached in the material. The black dashed line shows the path length expected from the SSM. For 36 MeV primary energy, the majority of the ions, especially those that were scattered from the near surface region, ex-hibit a path length in good agreement with the SSM, although deviations be-come already pronounced for the largest depth shown here, which is still com-parably low to the probing depth commonly used in HIERDA. Multiple scat-tering drastically broadens the path length distribution for lower ion energies (the case of 4 MeV iodine is indicatively plotted), even for very shallow pen-etration depth. Even though the low energy data are obviously more scattered, an agglomeration of trajectories around a most likely path length is observed and indicated by a blue dashed line to guide the eye.

As a complement, the integral path length distribution of the intermediate case of 12 MeV iodine scattered by 57.8 nm of gold is plotted in panel (a) of Figure 4.1. It can be seen that the distribution extends to values much higher than the maximum path length predicted by the SSM, calculated according to equation 4.1 (see black dashed-line).

40

Figure 4.1. (a) Path length and (b) nuclear energy loss distribution of 12 MeV iodine scattered by Au. The values expected in a single scattering scenario are indicated in the plots.

A striking observation, that can be made from Figure 4.1b showing the corre-sponding nuclear energy loss of the ions in the same simulation, is that even though multiple and plural scattering are drastically increased, the more the experimental conditions deviate from the SSM, the more the nuclear energy loss is found to deviate to values lower than predicted by the SSM. The shad-owed region denotes the expected SSM nuclear energy loss range as defined by ions scattered from the surface and from the backside of the film. This observation is extensively discussed in Paper II with additional data for MeV bromine ions scattered from silver as well as extensive computer simulations. Whereas as an example - TRIM-calculated nuclear losses for 6 MeV bromine scattered by 80.7 nm of silver - is plotted in Paper II-Figure 3.

Only ions scattered from the near-surface region (green histogram) exhibit elastic energy losses close to what is predicted by the SSM. Ions which pene-trate deeper, show a decreased energy loss due to elastic collisions. For larger depths, scattering into the simulated detector direction is almost exclusively taking place via plural and multiple scattering by significantly smaller angles than the total deflection necessary to reach the detector. For the employed heavy ions, this results in drastically decreased nuclear losses, despite the much longer path length. Thus, in analogy to the analysis of trajectories in transmission and reflection geometry, here a clear change in the character of the typical trajectories contributing to the spectra for changing film thickness is observed.

One consequence of this effect is illustrated in Figure 4.2, comparing ex-perimental and simulated spectra of 12 MeV iodine scattered at 45 degrees from gold samples of different thicknesses. In panel (a), the ToF projections of the experimental spectra are plotted together. A shift of the high-energy onset depending on the thickness is observed. Starting from the smallest thick-ness, i.e. for a 11 nm gold film, a high energy tail appears. The right edge of the signal gradually shifts to higher energies, than anticipated by single scat-tering kinematics. This shift continues as the thickness increases and the half-

41

height of the spectrum edge gradually moves to higher energy with thickness. It is only after a certain - critical - film thickness is reached that the maximum intensity is shifting to lower energy and stabilizes at the energy that the edge of a bulk sample is observed. To summarize, different from RBS, the single scattering peak is not found at the same energy as the half-height of the bulk signal.

Similar observations are made for the Monte Carlo simulations. As seen in panel (b) of Figure 4.2, the half-height spectrum shifts to energies higher than 𝑘𝐸 for increasing thickness before it ends up at the anticipated by SSM en-ergy for a film of 180 nm thickness. These observations indicate that, also in the near-surface region, the detection of ions with trajectories that result in energies higher than expected by the SSM is favored.

Figure 4.2. Experimental (panel a) and Monte Carlo (panel b) data showing the de-pendence of the detected ion energy on the film thickness in the case of 12 MeV iodine scattered by gold at 45 degrees.

The discussed effects also affect the extraction of energy loss data from the width of experimental spectra according to the procedure described in sec-tion 4.1.1. As shown in Paper II-Figure 1b, accurate determination of the ex-perimental width becomes challenging at low energies.

4.2 Resulting stopping cross sections and implications for HIERDA

The results from the Monte Carlo simulations have been used to formulate a more advanced model to assess stopping cross sections taking into account the deviations from the SSM. In this context, only the ions that penetrated deeper in the material and experienced the highest energy loss, are considered and associated with the low energy edge of the experimental spectrum.

Stopping cross section values of iodine in gold and bromine in silver as obtained by both approaches are plotted in Paper I-Figure 3 and Paper II-

42

Figure 4 respectively. An updated version of Paper I-Figure 3 is presented in Figure 4.3, where the plotted Monte Carlo results exhibit better statistics and are extended to a larger energy range compared to the two data points that are indicatively plotted in Paper I. Results are compared to SRIM predictions and literature data [71,80–84] that are available in IAEA library [85].

Based on the Monte Carlo results discussed in the previous section, one could expect an evaluation based on SSM being unable to yield reasonable stopping cross sections at low energies. However, the results obtained by SSM and Monte Carlo evaluation methods are within the estimated uncertainties. Moreover, the data are in good agreement with literature data. In turn, these results also imply that despite the above-mentioned observations, which se-verely challenge the SSM, depth profiling based on the SSM is expected to yield reasonable results.

The agreement of the two models can be explained by the deepened analy-sis of the Monte Carlo simulations that shows a strong dependence of the elas-tic losses on the depth reached by the ion in the material as well as an increas-ing deviation of the path length distribution from the SSM. On the one hand, for a given depth, nuclear energy losses at this energy range are significantly lower than the SSM prediction. On the other hand, for scattering from the same depth slab, the integral electronic energy loss is significantly increased due to the deviation of the ion path lengths from the straight trajectory as-sumed in SSM.

The final position of the low energy edge of the spectrum, used for analysis in the present studies, is the result of two processes with counteracting effects on the total energy loss experienced by the ion. Consequently, the obtained stopping cross sections do not differ much irrespective of the complexity of the implemented analysis, given the experimental uncertainties.

43

Figure 4.3. Stopping cross sections of iodine in gold as obtained by the two imple-mented evaluation methods. Data are plotted together with SRIM tabulated values for electronic and nuclear stopping as well as with literature data.

The effects described have important consequences for data analysis using the SSM as commonly employed in ToF-E HIERDA. At first, the effective de-pendence of the exact energy of the high-energy edge of the signal of scattered ions, which is often used for energy or ToF-calibrations, on the film thickness has to be considered. Also, a similar effect on the high-energy onset for recoil species is expected. Calibrating from a mixed series of very thin and bulk tar-gets can thus result in deteriorating effects of the near-surface depth profiles. Moreover, the fact, that the elastic losses experienced by ions scattered from subsurface layers is overestimated by the SSM may affect depth scales, as, ions scattered from deeper layers (assuming single scattering trajectory length) have experienced a lower energy loss from elastic collisions, and could be potentially detected at energies, which will be assigned to a depth closer to the surface.

At the same time, one should expect that the resulting recoil energy will also deviate from the SSM. However, this estimation involves a higher degree of complexity as, for example, smaller recoil angles result in higher recoil en-ergies. Differences have to be expected in the depth profiles of light and heavy elements. For example, for a homogeneous binary compound with light and heavy constituents, e.g. transition metal borides, carbides, nitrides etc., the tra-jectories of the created recoils differ drastically as light ions follow better de-fined trajectories, while the impact of multiple/plural scattering on heavy re-coils is added to the affected trajectory of the primary ions. As a result, the obtained profiles in such cases exhibit artificial gradients on the concentra-tions.

44

5 Material analysis by ion beams: In-situ characterization

Rapid developments in modern material research create an increasing need for accurate assessment of the sample composition on a true nanometer scale. Such measurements are, most commonly, performed ex-situ, with the samples being exposed to ambient conditions between manufacturing and/or modifi-cation and characterization. The issues, caused by limitations of the ex-situ characterization approach, are in many cases addressed in large infrastructures where set-ups dedicated for in-situ and real-time studies of material systems have been developed [86]. This approach allows to fully correlate sample composition with the processing steps and thus straightforward establish com-plete manufacturing protocols. However, other limitations exist in large scale infrastructures, for example in the form of tight time schedules, high opera-tional costs and limited influence of the individual user.

While the development of set-ups capable of advanced in-situ characteri-zation might also be too cost-intensive for individual research groups, medium scale facilities such as typical accelerator laboratories with IBA capabilities are potential hosts for infrastructure for in-situ analysis [87]. IBA offers good analytical capabilities for qualitative and quantitative near-surface infor-mation and can be performed without high demands concerning the required conditions. For example, IBA techniques are not affected by the presence of a reactive gas in the chamber up to comparably high pressures and can be com-bined with sample annealing. Thus, the design of a set-up that combines ma-terial modification processes with the analytical capabilities of IBA methods in-situ is appealing.

The development of a set-up capable for this type of studies represents the applied focus of the present thesis. After construction of the set-up and com-missioning by a series of test experiments (presented in Paper III), the new set-up was used in two applied studies (Papers IV and V), by combining in-situ IBA with thin film growth, exposure to reactive gasses, ion implantation and annealing.

45

5.1 Set-up for in-situ IBA: SIGMA A set-up for in-situ materials modification and characterization by ion beams was constructed and installed at the 5-MV Pelletron Tandem Accelerator at Uppsala University [26]. The name of the new set-up is SIGMA (an acronym standing for Set-up for In-situ Growth, Material modification and Analysis). A summary of the main features of SIGMA, together with some experimental examples showing its potential and capabilities has been published in [88] (Paper III).

A photograph of the set-up, with main components indicated by numbers 1-12, is reproduced in Figure 5.1, while a schematic drawing is shown in Fig-ure 1 in Paper III. The SIGMA beam line consists of three main parts which, presented in the order of the beam direction, are: a pre-chamber, the main chamber and a load-lock system. All equipment used for material modification and characterization is located in the main chamber, while the other two parts, i.e. load-lock and pre-chamber, function as complementary tools to allow sam-ple storage and transportation and improve vacuum.

The portable load-lock chamber is attached to the main chamber as shown in Figure 5.1. The load-lock is mounted on a trolley and is connected to an autonomous AC power supply. It is equipped with a separate and independent turbomolecular pump. Sample transfer between the load-lock and the main chamber is achieved by a mechanical manipulator. The load-lock is mainly used for sample storage, as it holds a carousel sample holder with a capacity of four samples and enables sample installation and exchange without break-ing the vacuum in the main chamber. In addition, stored samples can be trans-ferred in high vacuum conditions to other experimental set-ups, as for example the Medium Energy Ion Scattering (MEIS) [89,90] system of the same labor-atory [91].

46

Figure 5.1. Photo of the SIGMA beam line. The visible components of the set-up are indicated by numbers: (1) Pre-chamber, (2) Main chamber, (3) Load-lock chamber, (4) Turbo pumps, (5) Pressure gauges, (6) Slits, (7) Goniometer, (8) Ion source, (9) e-

-beam evaporator, (10) Gas supply, (11) RGA and (12) Faraday cup.

The pre-chamber (also indicated in Figure 5.1) is connected to a turbomolec-ular pump that allows differential pumping in order to improve the vacuum in the main chamber. Additionally, it separates the main chamber from the beam line when material modification processes that affect the pressure (i.e. con-trolled gas exposure, sample annealing etc.) take place in SIGMA. It is also equipped with a system of slits and beam collimators that adjust the size of the ion beam before it enters the main chamber.

In the following sections, the main equipment and capabilities of SIGMA are summarized. A table demonstrating the potential of the set-up to run pro-cesses in parallel can be found in Paper III.

47

5.1.1 Main chamber: Basic equipment An inside view of the main chamber is shown in Figure 5.2a. This part of the set-up holds equipment to be used for material modification. The samples are mounted on sample holders like the one shown in Figure 5.2c. The sample holder is attached to the plate (Figure 5.2b) of a 5-axis goniometer that allows three translations and two rotations. The ion beam current hitting the sample is measured at the anode indicated in Figure 5.2c. A low voltage of +36 V can be applied on the target to suppress secondary electron emission.

Resistive and electron beam (e--beam) sample annealing is enabled by a tungsten filament located behind the sample (see Figure 5.2b). A voltage of -650 V can be applied to accelerate emitted electrons towards the sample. Samples can be annealed up to ~1400 K and the temperature is measured by a k-type thermocouple in contact with the sample (connections on the sample holder are indicated in Figure 5.2c). At high temperatures the temperature scale can calibrated by an optical pyrometer. A second k-type thermocouple is mounted on the backside of the goniometer plate to keep track of the tem-perature evolution far from the sample holder. Sample cooling to cryogenic temperatures by liquid N2 is also possible.

A Pfeiffer QMG 250F1 PrismaPro [92] mass spectrometer is attached to the main chamber and enables Residual Gas Analysis (RGA) in a mass range from 1 to 100 amu. The mass spectrometer can be used for performing TDS by measuring the release rate of gas species from the sample during annealing as it will be discussed in section 5.3.1. It can also be used to control the expo-sure of the sample to reactive and noble gasses. The gasses are introduced in the chamber via a needle valve connected to gas bottles. So far, the system has been tested for controlled exposure to 1H2, D2, 16O2, 18O2 and Ar.

A turbomolecular pump is connected to the main chamber in addition to the previously mentioned pumps of the pre-chamber and the load-lock. With simultaneous operation of the pumps, the pressure of the main chamber can be kept well below 5x10-9 mbar.

SIGMA equipment for material modification includes also a Focus EFM 3T [93] e--beam evaporator for thin film growth. The evaporator allows sim-ultaneous evaporation from three independent cells. The material of interest is placed either in crucibles or is directly installed in the form of rods. Current is applied to a hot cathode (tungsten filament) in the range of 2-2.5 A and the produced electrons due to thermionic emission are collected by the positive high voltage applied to the target material. The latter is then evaporated and deposited on the surface of the sample. Each of the cells is connected to its own autonomous power supply that regulates the filament current and the high voltage applied to the target material.

The evaporator was tested by growing thin films of gold, copper and silver from different crucibles and cells. The evaporation parameters, such as re-quired heating power, deposition rates and distances, were calibrated.

48

Evaporation in reactive atmosphere was also performed successfully, which enabled the application discussed in section 5.2.

Figure 5.2. (a) Inside view of the main chamber. The goniometer, the evaporator, the ion source and the RBS detector can be seen. (b) basic connections at the goniometer. (c) Sample holder.

A PreVac IS40C1 [94] ion source can be used for sample irradiation by atomic and molecular ion beams of particles energy up to 5 keV at significantly higher current than the probing beam (~μA). Ion beams produced in this en-ergy range enable surface cleaning and thin film removal by ion sputtering as well as studies involving near-surface ion implantation.

A dedicated sample holder design, for which the impinging ion current is measured behind an aperture with 1 mm diameter at a position corresponding to the geometrical center of the sample, was installed in the chamber. This permits to deduce the profile of the ion beam generated by the ion source and to optimize the sample position. Beam profiles of H+, provided by accelerating

49

voltages of 1 and 2 kV, have been mapped and are plotted in Figure 5.3. The full width half maximum was found to be 11 mm at the horizontal axis, while a typical sample size is around 1 cm2.

Figure 5.3. Profiles of hydrogen beam at 2 different energies produced by the ion source and scanned by an adjusted sample holder.

5.1.2 Main chamber: Detector systems The IBA capabilities of the chamber have been systematically upgraded since its construction. In this thesis, the current status (November 2020) of the chamber is presented, which differs in several instances from the description in [88].

The full IBA characterization capabilities of SIGMA are illustrated in Fig-ure 5.4. Three PIPS detectors are mounted on a carrousel at adjustable angles. One detector is typically placed in backscattering geometry to be used for RBS/EBS and particle-NRA. A scattering angle of 170 degrees with respect to the beam direction is usually selected as it combines good mass separation in RBS and enhances the detection sensitivity of EBS for oxygen depth pro-filing by making use of the narrow elastic resonance of 16O at 3037 keV 4He (see section 2.1.3). The detection angle of the backscattering detector can be modified by an external mechanism without breaking the vacuum. The thick-ness of the depletion region of the presently employed detector is 500 μm and its solid angle is 2.2 msr.

The other two detectors are used for performing foil-ERDA and they are placed in forward directions. Their depletion region thickness is 100 μm and they typically aim to hydrogen isotopes detection simultaneously with RBS/EBS. Depending on the beam energy and the detection angle, aluminum foils between 5 and 30 μm have been used as absorbers.

The placement of opaque foils in front of the PIPS detectors allows their operation during sample annealing as their resolution is otherwise distorted due to the light of the filament.

50

A Silicon Drift Detector (SDD) [95] is mounted on the chamber for per-forming PIXE. Finally, a portable High Purity Germanium (HPGe) [96,97] detector available at the laboratory has been used for γ-NRA.

Figure 5.4. Sketch of the IBA capabilities of SIGMA. Three rotatable PIPS detectors are installed in the chamber for RBS/EBS, particle-NRA and ERDA. An SDD is used for PIXE and a portable HPGE for γ-NRA.

5.1.3 Automatization of data acquisition: SIDAS The most recent upgrade in SIGMA was the development of SIGMA Data Acquisition System (SIDAS) [98]. The code, written in LabVIEW [99] rec-ords simultaneously input signals from the vacuum gauges, the thermocouple and beam current reading, the RGA as well as the IBA data acquisition soft-ware. This approach permits an efficient online processing of multiple input data beneficial for in-situ experiments in which the multiple components of SIGMA are operated simultaneously.

51

5.2 Experimental capabilities: Thin film growth, oxidation and in-situ characterization

5.2.1 Characterizing the reactive growth of photochromic YHxOy

Yttrium hydrides (YHx) are known to exhibit photochromism at pressures in the order of GPa [100]. More recently, yttrium oxyhydride (YHxOy) thin films have attracted attention as a reversible photochromic response at ambient con-ditions was reported [101]. Since then, the composition [102] and struc-ture [103] have been related to their optical [104] and electrical [105] proper-ties. A correlation between the photochromic response and the oxygen/hydro-gen content of the material was shown [106]. It was found that YHxOy exhibit photochromic properties when the ratio of oxygen and yttrium is between 0.45 and 1.5. Photochromic properties were also observed for other rare-earth metal oxyhydrides [107,108].

Up to now, YHxOy thin films were grown by reactive magnetron sputtering of yttrium in Ar/H2 atmosphere, followed by uncontrolled oxidation by the exposure of the samples to air and samples were characterized ex-situ. To im-prove the understanding of the photochromic effect, as well as to establish efficient and cost-effective manufacturing pathways, detailed knowledge on the synthesis and aging mechanisms of the photochromic films is required.

To gain this knowledge, YHxOy films have been grown in SIGMA and an-alyzed in-situ during the deposition by reactive e--beam evaporation and con-trolled oxidation. The study performed is presented in [109] (Paper IV).

52

Figure 5.5. Synthesis and in-situ characterization of photochromic YHxOy in SIGMA. The evolution in the color of the sample can be observed in panel (a), while the IBA spectra before and after oxidation are shown in panels (b) and (c).

In the beginning, yttrium dihydride (YH2) was deposited on transparent sub-strates (CaF2 and soda-lime glass) that enable optical characterization. The deposition was performed by e--beam evaporation of yttrium in a hydrogen atmosphere (hydrogen pressure during evaporation: 3x10-6 mbar).

Incorporation of oxygen in samples was achieved either by direct exposure of the samples to atmosphere or by controlled oxidation in SIGMA. The color of the thin films shifts to yellow-transparent upon oxidation (as the initial stage of YH2 is opaque). The resulting change in the optical appearance of a sample that underwent controlled oxidation is shown in the photos of Figure 5.5a.

For the composition characterization of the manufactured samples, we combined in-situ and ex-situ IBA. As one can see in Figure 5.5 panels (b) and (c), the oxygen and hydrogen content of a fully in-situ characterized sample were tracked by EBS/RBS and foil-ERDA respectively by using 3060 keV 4He as probing beam, while samples were also measured ex-situ by ToF-E HIERDA.

For the optical characterization, samples were illuminated by a LED-clus-ter (λ=455 nm) to trigger the photochromic response. The strength of the

53

response was then measured ex-situ by optical spectrophotometry and in-situ by the comparison of photographical images taken under identical conditions. The photodarkening of a sample by in-situ illumination and the results of the image analysis are shown in Paper IV-Figure 3. Structural characterization was performed by grazing incidence X-ray diffraction ex-situ and the crystal structure and lattice parameter found to be similar to sputtered films (see sup-plementary material of Paper IV).

To summarize, one can distinguish the photochromic samples synthesized in SIGMA into three distinct subsets. First (subset 1) are the samples that were oxidized by direct exposure to air briefly after deposition of YH2. Subset 2 consists of samples that were first oxidized in-situ before their exposure to air. Therefore, composition, optical response and structure of the first two subsets were characterized ex-situ. However, a fully in-situ characterization approach was followed for samples belonging in subset 3. All the different pathways are summarized in the flowchart diagram in Figure 5.6.

54

Figure 5.6. Flowchart diagram that illustrates the established protocol for YHxOy thin film synthesis and characterization. The manufactured samples can be divided into three subsets depending on the process flow, e.g., in-situ or ex-situ oxidation and char-acterization. The photo on top right shows the photochromic effect of a sample of subset 1, which was illuminated on the right-hand side. The photo at the bottom shows the in-situ illumination process followed for samples of subset 3.

5.2.2 Summary of results The final results showing the compositions of subsets 1 and 2 as measured ex-situ by ToF-E HIERDA are plotted in the ternary diagram of Paper IV-Fig-ure 2. It was observed, that samples oxidized by direct exposure to air main-tained their photochromic properties in ambient conditions for a longer time compared to those followed the slower path of controlled oxidation. The dif-ference is attributed to different microstructure depending on the specific ox-idation pathway. As seen in the ternary plot, fast oxidation of samples of sub-set 1 (indicated by green numerical notation) appears to hinder further oxygen incorporation by the film, while at the same time samples of subset 2 (red

55

numerical notation) had a higher post-oxidation rate after being exposed to air.

Results of a full in-situ process are also shown in the ternary diagram. The changes in composition of a sample before and after oxidation as well as after being photodarkened in-situ are plotted in blue squares. The oxidation of the sample caused hydrogen release and the evolution of the composition fol-lowed the formula suggested by Moldarev et al. [106]. The sample was finally measured ex-situ by ToF-E HIERDA after 16 days in ambient conditions (also shown in the ternary diagram). Similarly to samples of subset 2, the initial slow oxidation resulted in faster post oxidation rate and loss of the photo-chromic properties.

To conclude, the performed study contributed to the existing knowledge on photochromic YHxOy in the following ways. It was shown that photochromic YHxOy can be synthesized by e--beam evaporation, in which the yttrium atoms are deposited on the substrate by significantly lower energy than in magnetron sputtering. The capability of controlled exposure to pure oxygen allowed the establishment of a synthesis protocol. Most importantly, the novel in-situ char-acterization approach that was followed in this study enabled the assessment of the composition and photochromic properties of the samples at any stage of their life cycle.

5.3 Experimental capabilities: Ion irradiation 5.3.1 Combination of in-situ ion implantation, annealing, IBA

and thermal desorption spectroscopy During commissioning of SIGMA the low energy ion source was employed for test implantations of 5 keV hydrogen in Si(100) combined with subsequent in-situ γ-NRA, as described in Paper III.

Subsequently, a separate study combining capabilities for ion implantation, IBA and TDS in a single system was performed. Specifically, we studied deu-terium retention in tungsten in-situ in collaboration with the Atomic and Plasma physics group at TU Wien. The results from this investigation are de-scribed in detail in Paper V.

This specific topic is of interest for the nuclear fusion community as the extreme conditions of a fusion reactor require extensive investigations on the wall materials which are heavily modified during plasma exposure [110,111]. Particles are escaping the plasma and are implanted in the walls, which in the case of fuel species such as deuterium and in particular tritium leads to fuel retention [112]. Also, local melting might occur due to high tempera-tures [113]. At the same time, a mixed layer is formed on Plasma Facing Com-ponents (PFCs) due to the continuous sputtering and redeposition of the wall

56

materials [114]. IBA methods have been widely used for depth profiling and characterization of the impact of the plasma on reactor walls [26,115].

Tungsten is considered as the main candidate material for PFCs of future fusion reactors as it features several advantages, e.g. high melting point and low sputtering yield [116,117]. Numerous studies have been conducted on the retention of hydrogen isotopes in tungsten [118].

Since the material system of deuterium in tungsten is extensively charac-terized by IBA, it is an ideal test ground for assessing the in-situ capabilities of SIGMA.

The experimental procedure is outlined in Figure 5.7. Deuterium atomic and molecular ions (accelerating voltage: 3 kV) were implanted in polycrys-talline tungsten bulk foils at nominal fluences in the order of 1022 ions/m2. Samples were subsequently annealed up to ~1400 K and their change in pro-tium and deuterium inventories were tracked online by IBA. Annealing was performed in both stepwise and continuous ramp modes. During annealing, deuterium release was also measured by TDS as the partial pressure of D2 in the chamber was tracked by the RGA. Implanted samples were also measured by Scanning Electron Microscopy (SEM) ex-situ.

For the IBA measurements, foil-ERDA was operated with a 3400 keV 4He++ beam to probe deuterium content up to ~350 nm, while a higher energy beam of 6200 keV was used to increase probing depth to approximately 1 μm. PIXE spectra that were simultaneously recorded were used to normalize ERDA data by the ion beam current and get quantitative results.

As shown in Paper V-Figure 3, depth profiles obtained by ERDA reveal the existence of two types of implanted deuterium species. The signal of deu-terium in region A corresponds to species trapped in near-surface defects, while deuterium in region B is diffused deeper in the material and located in interstitial sites, in particular along grain boundaries. SEM on the implanted samples showed the formation of blisters on the samples, an implantation ef-fect well-known in literature [119,120]. As seen in the SEM images in Figure 5.7, the blisters disappear after the total release of deuterium caused by sample annealing. The amount of deuterium detected by IBA is only a small fraction of the nominal implantation fluences, as deuterium is diffused deeper than the probing depth by the employed ion beams.

The measured evolution of the deuterium and protium inventories as a function of temperature in stepwise and continuous approaches is shown in Figures 5 and 6 of Paper V. Hydrogen isotopes show a distinct retention be-havior depending on the region that they are located. The outgassing of -the diffused in the material- region B deuterium as function of temperature is ex-ponential. On the other hand, release of - the near surface - region A species exhibits a threshold-like behavior that is followed by a linear decrease in the signal.

TDS results follow a similar behavior as IBA results indicated for deuter-ium in region A, even though TDS while providing no depth resolution is

57

expected to track the total deuterium release. While this is expected to be par-tially explained by deuterium inventories not accessible such as blisters formed at larger depths, it also demonstrates the necessity of e.g. mandatory degassing of the goniometer or even the whole chamber system to abstract from both deuterium absorbed to the walls or even unintentionally implanted in non-sample materials. For these reasons, the most recent upgrade of SIGMA comprises the above mentioned RGA system, which enables fast re-peated acquisition of complete mass spectra and allows simultaneous obser-vation of several species such as HD, HDO and D2O. The ratio of those as well as its time-dependence is expected to improve the comparability of IBA results and RGA data in future experiments.

Figure 5.7. Illustration of the experimental procedure. Deuterium implantation in SIGMA was followed by in-situ sample annealing with IBA and TDS performed sim-ultaneously. Deuterium and protium inventories of the sample were tracked as a func-tion of temperature. SEM images of a tungsten sample that were obtained ex-situ be-fore and after implantation subsequently to annealing are also shown.

58

5.3.2 Sputtering The ion source has been successfully used for surface cleaning by argon sput-tering. An example showing a thin film removal procedure by sputtering is illustrated in Figure 5.8. Here, a copper thin film was deposited on a Si sub-strate by e--beam evaporation and subsequently irradiated by 5 keV argon un-der an angle of 60 degrees with respect to the surface normal. The effect of argon irradiation on the thickness of the film was tracked in-situ in steps. As shown in panel (a), the thickness was measured by RBS by using a 2 MeV 4He+ beam. For comparison, RBS spectra before and after 30 and 60 minutes of sputtering are plotted together. In panel (b), it can be seen that the thickness of the film decreases linearly with sputtering.

Figure 5.8. Panel (a): RBS spectra of a copper thin film (10.6 nm) after grown by e--beam evaporation (solid black line), and after 2 cycles of 5 keV argon sputtering cleaning (red and blue dashed-lines). Panel (b): Decrease of the copper thin film thick-ness as a function of the sputtering time.

59

6 Conclusions

The aim of the present thesis was to improve ion-based materials analysis us-ing MeV particle beams by dealing with open aspects of both fundamental and more applied character.

Studies on the energy loss of heavy ion beams when traversing solid matter led to interesting observations regarding the role that multiple and plural scat-tering play on the formation of the acquired energy spectra. These effects were, in turn, discussed in the context of assessing electronic energy loss from a spectrum in reflection geometry and the implications in data analysis.

The investigations performed strongly indicated that the beam energies, in which ToF-E HIERDA is commonly performed in many laboratories, are too low to ensure the validity of the single scattering assumption, which is em-ployed in many commonly used evaluation tools. However, the two effects, resulting from the severe deviation from the SSM, namely reduced elastic losses and increased path length may partially or completely compensate each other. Together, they lead to a final energy loss comparable to values antici-pated in the single scattering approximation, despite invalidity of the model itself. Thus, depth scales and quantification in ToF-E HIERDA are commonly not severely affected by the oversimplification in modeling.

However, a stronger residual effect is expected in certain cases such as the depth profiling of samples consisting of light species such as boron or carbon and elements significantly heavier than the probing beam, e.g. tungsten. The additional different kinematics and scattering probability for the respective recoil species will not be compensated by the above-mentioned counteracting effects. In consequence, depth scales for the individual species might be af-fected differently, which in an evaluation using the single scattering approxi-mation lead to the observation of apparent gradients in concentration in the obtained profiles, also for homogeneous bulk samples.

The applied focus of the thesis is mainly on the development in-situ exper-imental processes involving IBA characterization. First, SIGMA was con-structed and equipped with tools for material modification and characteriza-tion. The chamber allows thin film growth by reactive and non-reactive e--beam evaporation, sample annealing, controlled exposure to gases, surface cleaning and ion implantation. In parallel, RBS/EBS, PIXE, ERDA and NRA can be performed at different synthesis steps without exposing the samples to ambient conditions. Α series of more advanced experiments were performed in order to demonstrate SIGMA capabilities. The latter aimed on studying

60

material systems of interest in applications within the rapidly expanding field of sustainable energy.

The in-situ study of the yttrium oxyhydride significantly contributed to the ongoing investigation of optically active rare earth oxyhydrides. First, synthe-sis of photochromic YHxOy was, for the first time, achieved by a deposition technique different than sputtering where the yttrium atoms were deposited on the surface with significantly lower energy. Secondly, the in-situ and ex-situ characterization of the samples at different stages of the processes, led to the establishment of a manufacturing protocol. It was, thus, confirmed that the earlier speculated formation path, i.e. initial formation of yttrium dihydride followed by post-oxidation, sufficiently described the process. In addition, it was observed that the rate of initial oxidation affects the post-oxidation rate and, as a result, the persistence of the photochromic properties. This effect might be attributed to changes in the microstructure of the sample caused by fast oxidation, which, in turn, obstruct incorporation of additional oxygen.

The in-situ study of deuterium retention in tungsten allowed a first test of the correlation of two fundamentally different techniques, i.e. IBA and TDS in a single system. A successful correlation of the data obtained by both meth-ods is of high interest for in-situ diagnostics in actual fusion reactors. As the performance of IBA, and thus the acquisition of depth resolved data, is - in contrast to TDS - not possible during reactor operation. Implanted deuterium was found in two distinct regions in the sample that exhibit different release behaviors. The evolution in the deuterium inventory of both regions was tracked during annealing by ERDA and results were compared to TDS results that comprise the integral deuterium release from the entire sample.

The SIGMA set-up, as the most tangible result from this thesis work has, following the experiments presented in this thesis, been already employed in a series of other investigations ranging from isotopic labelling of ultrathin films to investigation of the thermal stability of hard coatings. From the capa-bilities demonstrated so far and with possible further upgrades it is expected to be used in the near and distant future by researchers at Uppsala University and other users of the tandem laboratory infrastructure for successfully con-ducting in-situ IBA experiments.

61

7 Sammanfattning på Svenska (Summary in Swedish)

Interaktionen mellan energirika joner och material är mycket relevant för en mängd tillämpningar från materialvetenskap och medicin till rymdfysik. Bland dessa används materialkaraktärisering med jonstrålar brett på grund av möjligheten att genomföra högupplöst djupprofilering av den kemiska sam-mansättningen. I och med att metoderna är icke-destruktiva är de tilltalande för bland annat tunnfilmskaraktärisering, som är viktigt vid utvecklingen av många högteknologiska produkter i elektronik, optiska system och belägg-ningsteknologi.

Målet med den här avhandlingen är att ta upp en rad öppna fundamentala och tillämpade frågeställningar för att vidare utveckla jonstråleanalysens möj-ligheter till materialkaraktärisering. Bland de problem som är relevanta för många energirelaterade material finns exempelvis kvantifiering av lätta äm-nen i en tung matris. Sådan kvantifiering kan förbättras drastiskt genom en mer exakt förutsägelse av tunga och lätta energirika joners energideponering i materialen ifråga. Vidare kräver syntes av mer avancerade materialsystem komplexa flerstegsprotokoll. Den pågående miniatyriseringen av materialsy-stem kopplade till strävan efter ett mer resurseffektivt samhälle ökar dock ris-ken för förändringar, till exempel oxidering, mellan tillverkningssteg eller in-nan karaktärisering. Således blir materialkaraktärisering på plats (in-situ) vik-tigare och viktigare, då fördelarna från jonstråleanalys kan bibehållas utan att proverna utsätts för yttre förhållanden. I de första kapitlen av denna avhand-ling summeras fysiken som ligger till grund för interaktionen mellan energi-rika joner och material, och moderna jonstrålemetoder beskrivs. Därefter pre-senteras de genomförda studierna och resultaten av dessa.

Som ett bidrag till en mer fullständig fundamental förståelse av den jon-materialinteraktion som är relevant för jonstråleanalys har energiförlusten för tunga joner i material, vid energier och geometrier relevanta för rekylspektro-metri, undersökts. Denna teknik har, tack vare sin höga känslighet för lätta ämnen, kommit att användas brett för kalibrering av andra foton- eller elektronbaserade verktyg för materialkaraktärisering. Bidraget från inelastisk och elastisk energiöverföring till den totala energiförlusten för tunga joner, såväl som giltigheten av enkelspridningsantagandet granskades. Det är rele-vant, då detta antagande vanligtvis används vid djupupplöst sammansättnings-analys och därmed även i flera populära utvärderingsverktyg. Vid de

62

genomförda experimenten spreds jod- och bromjoner från tunna filmer av guld och silver, och detekterades sedan i reflektionsgeometri. Förståelsen av den direkt mätta totala energiförlusten hos dessa joner fördjupades genom Monte Carlo simuleringar.

Studien kastade nytt ljus över det inflytande multipelspridning har på for-men av upptagna energispektra. Ett icke-trivialt beroende av elastiska förlus-ter och partiklarnas färdsträcka på det uppnådda djupet i materialet upptäcktes vid energier som vanligen används för rekylspektrometri. Det påvisades tyd-ligt att de strålenergier som vanligen nyttjas för elastisk rekylanalys med tunga joner (ToF-HIERDA) vid många laboratorier är för låga för att garantera en-kelspridningsantagandets giltighet. De två huvudsakliga effekterna som upp-står som ett resultat av avvikelser från enkelspridningsmodellen, nämligen re-ducerad elastisk energiförlust och ökad färdsträcka, har dock motsatta effekter på den upplevda energiförlusten. Sammantaget befinns de därmed leda till en total energiförlust som är jämförbar med den som förutsägs via enkelsprid-ningsantagandet, trots modellens begränsade giltighet. Således påverkas van-ligtvis inte djupskalor och kvantifieringar vid ToF-HIERDA avsevärt av över-förenklingen i modellen. En starkare kvarvarande effekt förväntas dock i vissa fall, då djupskalorna för enstaka partikelslag kan påverkas olika starkt, såsom vid djupprofilering av prover uppbyggda av lätta grundämnen säväl som äm-nen mycket tyngre än primärstrålen.

I den tillämpade delen av avhandlingen undersöks möjligheten att följa materialmodifieringsprocesser på plats med MeV-jonstrålar. En experimentell uppställning för syntes, materialmodifiering och analys (Set-up for In-situ Growth, Materials modification and Analysis, SIGMA) har konstruerats. SIGMA innehåller utrusting för reaktiv tunnfilmstillväxt, jonimplantering vid låg energi, sputtring, glödgning och kontrollerad tillförsel av reaktiva gaser, samtidigt som flera jonstråleanalystekniker är tillgängliga. Efter driftsättning och optimering av den installerade utrustningens prestanda genomfördes en serie avancerade experiment vid uppställningen. De två huvudsakliga studi-erna var fokuserade på materialsystem av intresse för tillämpningar i det snabbt växande fältet förnybar energi.

I den första studien undersöktes tillväxt samt utveckling av både samman-sättningen och de fotokroma egenskaperna hos tunna filmer av yttriu-moxyhydrid medan processen pågick. Denna studie förbättrade vår förståelse av fotokroma syrerika hydrider av sällsynta jordartsmetaller, framför allt på följande sätt. För det första klargjordes den fullständiga syntesvägen för materialet och en uppskattning av sammansättningen samt de fotokroma egen-skaperna i alla delar av provernas livscykel möjliggjordes. För det andra oxi-derades den initialt producerade yttriumdihydridfilmen på två olika sätt; ge-nom exponering mot atmosfär och genom kontrollerad exponering för ren syr-gas. Det befanns att den initiala oxideringstakten påverkar egenskaper som uppträder efter oxidering, och som ett resultat av detta påverkas tiden den fo-tokroma effekten dröjer kvar. Vidare visade denna studie att fotokrom YHxOy

63

kan syntetiseras genom olika deponeringsmekanismer, här specifikt reaktiv elektronstråleevaporering. I denna teknik deponeras yttriumatomerna på sub-stratet med signifikant lägre energi än vid magnetronsputtring. Medan de re-sulterande skillnaderna i mikrostruktur var kopplade till den observerade lång-siktiga stabiliteten hos proverna, visade resultaten att den fotokroma mekan-ismen är tydligare relaterad till den kemiska sammansättningen än till detaljer i syntesprocessen.

I den andra studien kombinerades deuteriumimplantering med jonstrålea-nalys och termisk desorptionsspektroskopi i ett och samma system för att mäta den integrerade deuteriumfrigörelsen samt jämföra den med motsvarande djupprofiler. En framgångsrik korrelering av data insamlad med båda meto-derna är högintressant för diagnostik i verkliga fusionsmaskiner där jonstråle-analys, till skillnad från residualgasanalys, inte kan användas under körning. Experimenten visade på närvaron av olika typer av bundet deuterium. Deute-rium hittades i defekter nära provytan såväl som vid lägre koncentration dif-funderat in i provet. Karaktäristiska blåsformationer på grund av stora mäng-der av diffunderade partiklar kunde också påvisas genom kompletterande ana-lys. Experimenten visade tydligt på distinkta beteenden för deuteriumfrigö-relse från olika reservoarer.

64

Acknowledgements

As one of the most exciting periods of my life is coming to an end, I would like to acknowledge the people who played an extremely important role in accomplishing the task with success.

First and foremost, I am grateful to my supervisor Daniel Primetzhofer. He gave me the opportunity to get an insight in advanced scientific research, deepen my knowledge and - most importantly - he was always there for me during the past years. The door of his office was always open to solve any kind of issue, either work- or life-related.

Moreover, I would like to thank my second supervisor Marcos Moro. We spent countless hours together in the lab as well as in planning the projects and discussing data. Everything would have been much harder without his help and support. His company in and out of work context is definitely some-thing to remember! Also, discussing the scientific results with - my third su-pervisor - Max Wolf was a pleasure and his perspective and comments broad-ened our understanding. Finally, Valentina Paneta, my second supervisor dur-ing the first years of my PhD, had an important role in the beginning of my project as well as in helping me adjusting to life in Sweden.

I would also like to acknowledge the technicians at Tandem laboratory. Jonas and Sven thank you for providing me with the proper ion beam every measurement day. Special thanks to Dan Wessman, whose contribution in constructing and running SIGMA was crucial.

A huge thank you to the people of the ion physics group and Tandem lab. Dmitrii for being one of the best officemates ever. Mau and Svenja that we started the Uppsala adventure almost at the same time. Petter who translated the thesis summary in Swedish. Kristina, Radek, Jila, Eduardo, Tuan, Vairavel, Sig, Per, Philipp, Linus, Sotiris, Barbara, Toon, Silma and the list can continue forever. Thank you, guys for the times together both during and after working hours! Also, thanks to all the new friends I made in Uppsala during the last years with a special mention to Nikos, Katerina, Haris, Efstra-tia, Konstantinos and Zisis. You are amazing!

Finally, I would like to thank my family for the support and especially my sister Natalia who created the cover picture of the present thesis.

Uppsala, January 31, 2021

Karim

65

Bibliography

[1] S. Chu, A. Majumdar, Opportunities and challenges for a sustainable energy future, Nature. 488 (2012) 294–303. https://doi.org/10.1038/nature11475.

[2] J. Ongena, G. Van Oost, Energy for future centuries: Prospects for fusion power as a future energy source, in: Fusion Sci. Technol., American Nuclear Society, 2012: pp. 3–16. https://doi.org/10.13182/FST12-A13488.

[3] H. Bolt, V. Barabash, G. Federici, J. Linke, A. Loarte, J. Roth, K. Sato, Plasma facing and high heat flux materials - Needs for ITER and beyond, J. Nucl. Mater. 307–311 (2002) 43–52. https://doi.org/10.1016/S0022-3115(02)01175-3.

[4] H. Bolt, V. Barabash, W. Krauss, J. Linke, R. Neu, S. Suzuki, N. Yoshida, Materials for the plasma-facing components of fusion reactors, in: J. Nucl. Mater., 2004: pp. 66–73. https://doi.org/10.1016/j.jnucmat.2004.04.005.

[5] G.A. Niklasson, C.G. Granqvist, Electrochromics for smart windows: Thin films of tungsten oxide and nickel oxide, and devices based on these, J. Mater. Chem. 17 (2007) 127–156. https://doi.org/10.1039/b612174h.

[6] C.G. Granqvist, S. Green, G.A. Niklasson, N.R. Mlyuka, S. von Kræmer, P. Georén, Advances in chromogenic materials and devices, Thin Solid Films. 518 (2010) 3046–3053. https://doi.org/10.1016/j.tsf.2009.08.058.

[7] H. Frey, H.R. Khan, Handbook of Thin-Film Technology, Springer Berlin Heidelberg, 2015. https://doi.org/10.1007/978-3-642-05430-3.

[8] R.A. Causey, Hydrogen isotope retention and recycling in fusion reactor plasma-facing components, J. Nucl. Mater. 300 (2002) 91–117. https://doi.org/10.1016/S0022-3115(01)00732-2.

[9] C.G. Granqvist, E. Avendaño, A. Azens, Electrochromic coatings and devices: Survey of some recent advances, in: Thin Solid Films, Elsevier, 2003: pp. 201–211. https://doi.org/10.1016/S0040-6090(03)00983-0.

[10] J.N. Huiberts, R. Griessen, J.H. Rector, R.J. Wijngaarden, J.P. Dekker, D.G. De Groot, N.J. Koeman, Yttrium and lanthanum hydride films with switchable optical properties, Nature. 380 (1996) 231–234. https://doi.org/10.1038/380231a0.

[11] Y. Waseda, E. Matsubara, K. Shinoda, X-Ray Diffraction Crystallography, Springer Berlin Heidelberg, 2011. https://doi.org/10.1007/978-3-642-16635-8.

[12] J.B. Pendry, Low Energy Electron Diffraction: The Theory and Its Application to Determination of Surface Structure, Academic Press, New York, 1974.

[13] A.T. Boothroyd, Principles of Neutron Scattering from Condensed Matter, 1st ed., Oxford University Press, Oxford, 2020.

66

[14] P. van der Heide, X-Ray Photoelectron Spectroscopy: An Introduction to Principles and Practices, John Wiley & Sons, Inc., Hoboken, NJ, USA, 2011. https://doi.org/10.1002/9781118162897.

[15] C.L. Briant, R.P. Messmer, eds., Auger electron spectroscopy, Academic Press, 1988.

[16] J.C. Russ, Fundamentals of Energy Dispersive X-ray Analysis, 1st ed., Butterworth-Heinemann, 1984. https://doi.org/https://doi.org/10.1016/C2013-0-04162-X.

[17] S. Amelinckx, D. Van Dyck, J. Van Landuyt, G. Van Tendeloo, Electron Microscopy: Principles and Fundamentals, Wiley, 1997. https://doi.org/10.1002/9783527614561.

[18] G. Haugstad, Atomic Force Microscopy: Understanding Basic Modes and Advanced Applications, John Wiley & Sons, Inc., Hoboken, NJ, USA, 2012. https://doi.org/10.1002/9781118360668.

[19] E. Nastasi, Y. Wang, Handbook of Modern Ion Beam Analysis, 2nd editio, Materials Research Society, 2009.

[20] G. Demortier, Ion Beam Techniques for the Non-destructive Analysis of Archaelogical Materials, in: X-Rays Archaeol., Springer-Verlag, 2005: pp. 67–100. https://doi.org/10.1007/1-4020-3581-0_4.

[21] H. Gegier, E. Marsden, On a diffuse reflection of the α-particles, Proc. R. Soc. London. Ser. A, Contain. Pap. a Math. Phys. Character. 82 (1909) 495–500. https://doi.org/10.1098/rspa.1909.0054.

[22] E. Rutherford, LXXIX. The scattering of α and β particles by matter and the structure of the atom, London, Edinburgh, Dublin Philos. Mag. J. Sci. 21 (1911) 669–688. https://doi.org/10.1080/14786440508637080.

[23] S. Rubin, T.O. Passell, L.E. Bailey, Chemical Analysis of Surfaces by Nuclear Methods, Anal. Chem. 29 (1957) 736–743. https://doi.org/10.1021/ac60125a001.

[24] W.-K. Chu, J.W. Mayer, M.-A. Nicolet, Backscattering Spectrometry, Academic Press, 1978. https://doi.org/https://doi.org/10.1016/B978-0-121-73850-1.X5001-7.

[25] S. Ogura, K. Fukutani, Thermal Desorption Spectroscopy, in: Surf. Sci. Soc. Japan Compend. Surf. Interface Anal., Springer, Singapore, 2018. https://doi.org/https://doi.org/10.1007/978-981-10-6156-1_116.

[26] M. Mayer, S. Möeller, M. Rubel, A. Widdowson, S. Charisopoulos, T. Ahlgren, E. Alves, G. Apostolopoulos, N.P. Barradas, S.E. Donnelly, S. Fazinic, K. Heinola, O. Kakuee, H. Khodja, A. Kimura, A. Lagogiannis, M. Li, S. Markelj, M. Mudrinic, P. Petersson, I. Portnykh, D. Primetzhofer, P. Reichart, D. Ridikas, T. Silva, S.M. Gonzalez de Vicente, Y. Wang, Ion beam analysis of fusion plasma-facing materials and components: Facilities and research challenges, Nucl. Fusion. 60 (2019). https://doi.org/10.1088/1741-4326/ab5817.

[27] J.R. Taylor, Classical Mechanics, University Science Books, 2005. [28] K.S. Krane, Introductory Nuclear Physics, 3rd ed., John Wiley & Sons Inc,

New York, 1987.

67

[29] J.A. Leavitt, L.C. McIntyre, M.D. Ashbaugh, J.G. Oder, Z. Lin, B. Dezfouly-Arjomandy, Cross sections for 170.5° backscattering of 4He from oxygen for 4He energies between 1.8 and 5.0 MeV, Nucl. Inst. Methods Phys. Res. B. 44 (1990) 260–265. https://doi.org/10.1016/0168-583X(90)90637-A.

[30] C. Huan-sheng, S. Hao, T. Jiayong, Y. Fujia, Cross sections for 170° backscattering of 4He from oxygen in the energy range of 2.0-9.0 MeV, Nucl. Inst. Methods Phys. Res. B. 83 (1993) 449–453. https://doi.org/10.1016/0168-583X(93)95870-B.

[31] J. Demarche, G. Terwagne, Precise measurement of the differential cross section from the 16O (α,α) 16O elastic reaction at 165° and 170° between 2.4 and 6.0 MeV, J. Appl. Phys. 100 (2006). https://doi.org/10.1063/1.2402868.

[32] A.F. Gurbich, SigmaCalc recent development and present status of the evaluated cross-sections for IBA, Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms. 371 (2016) 27–32. https://doi.org/10.1016/j.nimb.2015.09.035.

[33] J.D. Garcia, R.J. Fortner, T.M. Kavanagh, Inner-shell vacancy production in ion-atom collisions, Rev. Mod. Phys. 45 (1973) 111–177. https://doi.org/10.1103/RevModPhys.45.111.

[34] G. Moliere, Theorie der Streuung schneller geladener Teilchen I. Einzelstreuung am abgeschirmten Coulomb-Feld1, Zeitschrift Fur Naturforsch. - Sect. A J. Phys. Sci. 2 (1947) 133–145. https://doi.org/10.1515/zna-1947-0302.

[35] J.F. Ziegler, J.P. Biersack, U. Littmark, The Stopping and Range of Ions in Solids, Pergamon Press, New York, 1985.

[36] P. Sigmund, Particle Penetration and Radiation Effects - General Aspects and Stopping of Swift Point Charges, 1st ed., Springer-Verlag Berlin Heidelberg, 2006. https://doi.org/10.1007/3-540-31718-X.

[37] N. Bohr, On the Theory of the Decrease of Velocity of Moving Electrified Particles on Passing Through Matter, Phil. Mag. 25 (1913).

[38] N. Bohr, On the decrease of velocity of swiftly moving electrified particles in passing through matter, Phil. Mag. 30 (1915) 581–612.

[39] H. Bethe, Zur Theorie des Durchgangs schneller Korpuskularstrahlen durch Materie, Ann. Phys. 397 (1930) 325–400. https://doi.org/10.1002/andp.19303970303.

[40] F. Bloch, Zur Bremsung rasch bewegter Teilchen beim Durchgang durch Materie, Ann. Phys. 408 (1933) 285–320. https://doi.org/10.1002/andp.19334080303.

[41] W.H. Barkas, J.N. Dyer, H.H. Heckman, Resolution of the Σ- Mass anomaly, Phys. Rev. Lett. 11 (1963) 26–28. https://doi.org/10.1103/PhysRevLett.11.26.

[42] H.H. Andersen, H. Simonsen, H. Sørensen, An experimental investigation of charge-dependent deviations from the Bethe stopping power formula, Nucl. Physics, Sect. A. 125 (1969) 171–175. https://doi.org/10.1016/0375-9474(69)90836-7.

68

[43] H.H. Andersen, J.F. Bak, H. Knudsen, P. Møller Petersen, B.R. Nielsen, Experimental investigation of higher-order Z1 corrections to the Bethe stopping-power formula, Nucl. Instruments Methods. 140 (1977) 537–540. https://doi.org/10.1016/0029-554X(77)90373-1.

[44] W. Brandt, M. Kitagawa, Effective stopping-power charges of swift ions in condensed matter, Phys. Rev. B. 25 (1982) 5631–5637. https://doi.org/10.1103/PhysRevB.25.5631.

[45] P.M. Echenique, I. Nagy, A. Arnau, Interaction of slow ions with matter, Int. J. Quantum Chem. 36 (1989) 521–543. https://doi.org/10.1002/qua.560360854.

[46] E. Fermi, E. Teller, The capture of negative mesotrons in matter, Phys. Rev. 72 (1947) 399–408. https://doi.org/10.1103/PhysRev.72.399.

[47] J. Lindhard, M. Scharff, Energy dissipation by ions in the kev region, Phys. Rev. 124 (1961) 128–130. https://doi.org/10.1103/PhysRev.124.128.

[48] W.H. Bragg, R. Kleeman, XXXIX. On the α particles of radium, and their loss of range in passing through various atoms and molecules, London, Edinburgh, Dublin Philos. Mag. J. Sci. 10 (1905) 312–340.

[49] G. Kraft, Tumor therapy with heavy charged particles, Prog. Part. Nucl. Phys. 45 (2000) S473–S544. https://doi.org/10.1016/s0146-6410(00)00112-5.

[50] M. Durante, J.S. Loeffler, Charged particles in radiation oncology, Nat. Rev. Clin. Oncol. 7 (2010) 37–43. https://doi.org/10.1038/nrclinonc.2009.183.

[51] C. Montanari, Electronic Stopping Power of Matter for Ions, Int. At. Energy Agency / Nucl. Data Sect. (n.d.). https://www-nds.iaea.org/stopping/ (accessed October 6, 2020).

[52] J.F. Ziegler, M.D. Ziegler, J.P. Biersack, SRIM - The stopping and range of ions in matter (2010), Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms. 268 (2010) 1818–1823. https://doi.org/10.1016/j.nimb.2010.02.091.

[53] K. Wittmaack, Misconceptions impairing the validity of the stopping power tables in the SRIM library and suggestions for doing better in the future, Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms. 380 (2016) 57–70. https://doi.org/10.1016/j.nimb.2016.04.057.

[54] M. Nastasi, J.W. Mayer, Ion implantation and synthesis of materials, 2006. https://doi.org/10.1007/978-3-540-45298-0.

[55] S.A.E. Johansson, T.B. Johansson, Analytical application of particle induced X-ray emission, Nucl. Instruments Methods. 137 (1976) 473–516. https://doi.org/10.1016/0029-554X(76)90470-5.

[56] J. Miranda, Low energy PIXE: Advantages, drawbacks, and applications, Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms. 118 (1996) 346–351. https://doi.org/10.1016/0168-583X(95)01176-5.

[57] M. V. Moro, R. Holeňák, L. Zendejas Medina, U. Jansson, D. Primetzhofer, Accurate high-resolution depth profiling of magnetron sputtered transition metal alloy films containing light species: A multi-method approach, Thin Solid Films. 686 (2019). https://doi.org/10.1016/j.tsf.2019.137416.

69

[58] J. Tirira, Y. Serruys, P. Trocellier, Forward Recoil Spectrometry, Springer US, 1996. https://doi.org/10.1007/978-1-4613-0353-4.

[59] J.W. Mayer, Performance of Germanium and Silicon Surface Barrier Diodes as Alpha-Particle Spectrometers, J. Appl. Phys. 30 (1959) 1937–1944. https://doi.org/10.1063/1.1735093.

[60] S. Gotoh, Z. Takagi, Silicon Surface Barrier Detector, J. Nucl. Sci. Technol. 1 (1964) 311–315. https://doi.org/10.1080/18811248.1964.9732130.

[61] J. Kemmer, Fabrication of low noise silicon radiation detectors by the planar process, Nucl. Instruments Methods. 169 (1980) 499–502. https://doi.org/10.1016/0029-554X(80)90948-9.

[62] J. Kemmer, Improvement of detector fabrication by the planar process, Nucl. Inst. Methods Phys. Res. A. 226 (1984) 89–93. https://doi.org/10.1016/0168-9002(84)90173-6.

[63] P. Ström, P. Petersson, M. Rubel, G. Possnert, A combined segmented anode gas ionization chamber and time-of-flight detector for heavy ion elastic recoil detection analysis, Rev. Sci. Instrum. 87 (2016) 103303. https://doi.org/10.1063/1.4963709.

[64] Z. Siketić, I. Bogdanović Radović, I. Sudić, M. Jakšić, Surface analysis and depth profiling using time-of-flight elastic recoil detection analysis with argon sputtering, Sci. Rep. 8 (2018) 10392. https://doi.org/10.1038/s41598-018-28726-x.

[65] M. Chicoine, F. Schiettekatte, M.I. Laitinen, T. Sajavaara, Oxy-nitrides characterization with a new ERD-TOF system, Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms. 406 (2017) 112–114. https://doi.org/10.1016/j.nimb.2017.02.041.

[66] V.K. Alimov, M. Mayer, J. Roth, Differential cross-section of the D(3He, p)4He nuclear reaction and depth profiling of deuterium up to large depths, Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms. 234 (2005) 169–175. https://doi.org/10.1016/j.nimb.2005.01.009.

[67] M. Wilde, K. Fukutani, Hydrogen detection near surfaces and shallow interfaces with resonant nuclear reaction analysis, Surf. Sci. Rep. 69 (2014) 196–295. https://doi.org/10.1016/j.surfrep.2014.08.002.

[68] J.T. Park, E.. Zimmerman, Stopping Cross Sections of Some Hydrocarbon Gases for 40-250 keV Protons and Helium Ions, Phys. Rev. 131 (1963) 1611. https://doi.org/10.1103/PhysRev.138.A1317.

[69] D.I. Thwaites, Bragg’s rule of stopping power additivity: A compilation and summary of results, Radiat. Res. 95 (1983) 495–518. https://doi.org/10.2307/3576096.

[70] A. Arnau, M. Peñalba, P.M. Echenique, F. Flores, A charge state approach to the stopping power of ions in solids, Nucl. Inst. Methods Phys. Res. B. 69 (1992) 102–107. https://doi.org/10.1016/0168-583X(92)95744-C.

[71] J.M. Anthony, W.A. Lanford, Stopping power and effective charge of heavy ions in solids, Phys. Rev. A. 25 (1982) 1868–1879. https://doi.org/10.1103/PhysRevA.25.1868.

70

[72] G. Dollinger, M. Boulouednine, A. Bergmaier, T. Faestermann, Nonequilibrium charge states of recoil ions in high resolution elastic recoil detection analysis, Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms. 136–138 (1998) 574–578. https://doi.org/10.1016/S0168-583X(97)00662-9.

[73] E. Szilágyi, F. Pászti, G. Amsel, Theoretical approximations for depth resolution calculations in IBA methods, Nucl. Inst. Methods Phys. Res. B. 100 (1995) 103–121. https://doi.org/10.1016/0168-583X(95)00186-7.

[74] G. Amsel, G. Battistig, A. L’Hoir, Small angle multiple scattering of fast ions, physics, stochastic theory and numerical calculations, Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms. 201 (2003) 325–388. https://doi.org/10.1016/S0168-583X(02)01740-8.

[75] M. Mayer, SIMNRA, a simulation program for the analysis of NRA, RBS and ERDA, AIP Conf. Proc. 475 (1999) 541–544. https://doi.org/10.1063/1.59188.

[76] M.S. Janson, CONTES Instruction Manual , Internal Report, Uppsala, 2004. [77] K. Arstila, J. Julin, M.I. Laitinen, J. Aalto, T. Konu, S. Kärkkäinen, S.

Rahkonen, M. Raunio, J. Itkonen, J.P. Santanen, T. Tuovinen, T. Sajavaara, Potku - New analysis software for heavy ion elastic recoil detection analysis, Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms. 331 (2014) 34–41. https://doi.org/10.1016/j.nimb.2014.02.016.

[78] M. V. Moro, B. Bruckner, P.L. Grande, M.H. Tabacniks, P. Bauer, D. Primetzhofer, Stopping cross section of vanadium for H+ and He+ ions in a large energy interval deduced from backscattering spectra, Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms. 424 (2018) 43–51. https://doi.org/10.1016/j.nimb.2018.03.032.

[79] MATLAB R2019b, (2019). [80] M. Abdesselam, J.P. Stoquert, G. Guillaume, M. Hage-Ali, J.J. Grob, P.

Siffert, Cu, I and Au stopping powers in solid targets, Nucl. Inst. Methods Phys. Res. B. 72 (1992) 7–15. https://doi.org/10.1016/0168-583X(92)95273-T.

[81] S. Ouichaoui, L. Rosier, E. Hourany, R. Bimbot, N. Redjdal, H. Beaumevieille, Stopping powers of Al, Cu, Ag and Au media for 1.47 MeV/u 127I, 2 MeV/u 32S and 79Br ions, Nucl. Inst. Methods Phys. Res. B. 95 (1995) 463–469. https://doi.org/10.1016/0168-583X(94)00854-X.

[82] G.N. Knyazheva, S. V. Khlebnikov, E.M. Kozulin, T.E. Kuzmina, V.G. Lyapin, M. Mutterer, J. Perkowski, W.H. Trzaska, Energy losses of 252Cf fission fragments in thin foils, Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms. 248 (2006) 7–15. https://doi.org/10.1016/j.nimb.2006.04.071.

[83] C.D. Moak, M.D. Brown, Some heavy-ion stopping powers, Phys. Rev. 149 (1966) 244–245. https://doi.org/10.1103/PhysRev.149.244.

[84] L.B. Bridwell, L.C. Northcliffe, S. Datz, C.D. Moak, H.O. Lutz, Stopping powers for iodine ions at energies up to 200 MeV, Phys. Rev. 159 (1967) 276–277. https://doi.org/10.1103/PhysRev.159.276.

[85] IAEA Nuclear Data Services, (n.d.). https://www-nds.iaea.org/ (accessed June 24, 2020).

71

[86] W. Kreuzpaintner, B. Wiedemann, J. Stahn, J.F. Moulin, S. Mayr, T. Mairoser, A. Schmehl, A. Herrnberger, P. Korelis, M. Haese, J. Ye, M. Pomm, P. Böni, J. Mannhart, In situ Polarized Neutron Reflectometry: Epitaxial Thin-Film Growth of Fe on Cu(001) by dc Magnetron Sputtering, Phys. Rev. Appl. 7 (2017) 054004. https://doi.org/10.1103/PhysRevApplied.7.054004.

[87] C. Linsmeier, P. Goldstraß, K.U. Klages, ARTOSS - A new surface science experiment to study the hydrogen inventory in multi-component materials, in: Phys. Scr. T, IOP Publishing, 2001: pp. 28–33. https://doi.org/10.1238/Physica.Topical.094a00028.

[88] K. Kantre, M.V. Moro, D. Moldarev, D. Johansson, D. Wessman, M. Wolff, D. Primetzhofer, SIGMA: A Set-up for In-situ Growth, Material modification and Analysis by ion beams, Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms. 463 (2020) 96–100. https://doi.org/10.1016/j.nimb.2019.11.007.

[89] R.G. Smeenk, R.M. Tromp, H.H. Kersten, A.J.H. Boerboom, F.W. Saris, Angle resolved detection of charged particles with a novel type toroidal electrostatic analyser, Nucl. Instruments Methods. 195 (1982) 581–586. https://doi.org/10.1016/0029-554X(82)90022-2.

[90] J.F. van der Veen, Ion beam crystallography of surfaces and interfaces, Surf. Sci. Rep. 5 (1985) 199–287. https://doi.org/10.1016/0167-5729(85)90001-9.

[91] M.K. Linnarsson, A. Hallén, J. Åström, D. Primetzhofer, S. Legendre, G. Possnert, New beam line for time-of-flight medium energy ion scattering with large area position sensitive detector, Rev. Sci. Instrum. 83 (2012). https://doi.org/10.1063/1.4750195.

[92] Pfeiffer QMG 250F1 PrismaPro, (n.d.). https://www.pfeiffer-vacuum.com/en/products/measurement-analysis/analysis-equipment/residual-gas-analysis/residual-gas-analysis-in-high-vacuum/prismapro-qmg-250-f1-1-100-u-open-ion-source/32123/prismapro-qmg-250-f1-1-100-u-open-ion-source-yttriated-iridiu.

[93] EFM 3T - FOCUS GmbH, (n.d.). https://www.focus-gmbh.com/efm-evaporators/efm-3t/ (accessed December 3, 2020).

[94] Ion Source IS 40C1, (n.d.). https://www.prevac.eu/en/2,offer/37,instruments/122,ion-source-40c1.html (accessed December 3, 2020).

[95] E. Gatti, P. Rehak, Semiconductor drift chamber - An application of a novel charge transport scheme, Nucl. Instruments Methods Phys. Res. 225 (1984) 608–614. https://doi.org/10.1016/0167-5087(84)90113-3.

[96] R.D. Baertsch, R.N. Hall, Gamma Ray Detectors Made From High Purity Germanium, IEEE Trans. Nucl. Sci. 17 (1970) 235–240. https://doi.org/10.1109/TNS.1970.4325695.

[97] R.H. Pehl, R.C. Cordi, F.S. Goulding, High-purity germanium: Detector fabrication and performance, IEEE Trans. Nucl. Sci. 19 (1972) 265–269. https://doi.org/10.1109/TNS.1972.4326520.

[98] E. Petersson, Study of integrated digital interface solutions for the SIGMA-setup at the Tandem laboratory, Uppsala, 2020.

72

[99] J. Travis, L. Wells, LabVIEW for Everyone: Graphical Programming Made Easy and Fun, Prentice Hall, 2002.

[100] A. Ohmura, A. MacHida, T. Watanuki, K. Aoki, S. Nakano, K. Takemura, Photochromism in yttrium hydride, Appl. Phys. Lett. 91 (2007) 151904. https://doi.org/10.1063/1.2794755.

[101] T. Mongstad, C. Platzer-Björkman, J.P. Maehlen, L.P.A. Mooij, Y. Pivak, B. Dam, E.S. Marstein, B.C. Hauback, S.Z. Karazhanov, A new thin film photochromic material: Oxygen-containing yttrium hydride, Sol. Energy Mater. Sol. Cells. 95 (2011) 3596–3599. https://doi.org/10.1016/j.solmat.2011.08.018.

[102] D. Moldarev, D. Primetzhofer, C.C. You, S.Z. Karazhanov, J. Montero, F. Martinsen, T. Mongstad, E.S. Marstein, M. Wolff, Composition of photochromic oxygen-containing yttrium hydride films, Sol. Energy Mater. Sol. Cells. 177 (2018) 66–69. https://doi.org/10.1016/j.solmat.2017.05.052.

[103] S. Cornelius, G. Colombi, F. Nafezarefi, H. Schreuders, R. Heller, F. Munnik, B. Dam, Oxyhydride Nature of Rare-Earth-Based Photochromic Thin Films, J. Phys. Chem. Lett. 10 (2019) 1342–1348. https://doi.org/10.1021/acs.jpclett.9b00088.

[104] C.C. You, T. Mongstad, J.P. Maehlen, S. Karazhanov, Engineering of the band gap and optical properties of thin films of yttrium hydride, Appl. Phys. Lett. 105 (2014). https://doi.org/10.1063/1.4891175.

[105] C.C. You, T. Mongstad, E.S. Marstein, S.Z. Karazhanov, The dependence of structural, electrical and optical properties on the composition of photochromic yttrium oxyhydride thin films, Materialia. 6 (2019). https://doi.org/10.1016/j.mtla.2019.100307.

[106] D. Moldarev, M. V Moro, C.C. You, E.M. Baba, S.Z. Karazhanov, M. Wolff, D. Primetzhofer, Yttrium oxyhydrides for photochromic applications: Correlating composition and optical response, Phys. Rev. Mater. 2 (2018) 115203. https://doi.org/10.1103/PhysRevMaterials.2.115203.

[107] F. Nafezarefi, H. Schreuders, B. Dam, S. Cornelius, Photochromism of rare-earth metal-oxy-hydrides, Appl. Phys. Lett. 111 (2017). https://doi.org/10.1063/1.4995081.

[108] S.M. Aðalsteinsson, M. V. Moro, D. Moldarev, S. Droulias, M. Wolff, D. Primetzhofer, Correlating chemical composition and optical properties of photochromic rare-earth oxyhydrides using ion beam analysis, Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms. 485 (2020) 36–40. https://doi.org/10.1016/j.nimb.2020.09.016.

[109] K. Kantre, M.V. Moro, D. Moldarev, M. Wolff, D. Primetzhofer, Synthesis and in-situ characterization of photochromic yttrium oxyhydride grown by reactive e−-beam evaporation, Scr. Mater. 186 (2020) 352–356. https://doi.org/10.1016/j.scriptamat.2020.05.009.

[110] D. Naujoks, Plasma-Material Interaction in Controlled Fusion, Springer Berlin Heidelberg, 2006. https://doi.org/10.1007/3-540-32149-7.

73

[111] G. Federici, C.H. Skinner, J.N. Brooks, J.P. Coad, C. Grisolia, A.A. Haasz, A. Hassanein, V. Philipps, C.S. Pitcher, J. Roth, W.R. Wampler, D.G. Whyte, Plasma-material interactions in current tokamaks and their implications for next step fusion reactors, Nucl. Fusion. 41 (2001) 1967–2137. https://doi.org/10.1088/0029-5515/41/12/218.

[112] T. Loarer, Fuel retention in tokamaks, J. Nucl. Mater. 390–391 (2009) 20–28. https://doi.org/10.1016/j.jnucmat.2009.01.039.

[113] G.F. Matthews, B. Bazylev, A. Baron-Wiechec, J. Coenen, K. Heinola, V. Kiptily, H. Maier, C. Reux, V. Riccardo, F. Rimini, G. Sergienko, V. Thompson, A. Widdowson, Melt damage to the JET ITER-like Wall and divertor, in: Phys. Scr., Institute of Physics Publishing, 2016: p. 014070. https://doi.org/10.1088/0031-8949/T167/1/014070.

[114] C. Linsmeier, M. Reinelt, K. Schmid, Surface chemistry of first wall materials - From fundamental data to modeling, in: J. Nucl. Mater., Elsevier B.V., 2011. https://doi.org/10.1016/j.jnucmat.2010.08.056.

[115] P. Petersson, M. Rubel, G. Possnert, S. Brezinsek, B. Pegourie, Nuclear reaction and heavy ion ERD analysis of wall materials from controlled fusion devices: Deuterium and nitrogen-15 studies, in: Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms, North-Holland, 2012: pp. 113–117. https://doi.org/10.1016/j.nimb.2011.07.053.

[116] W. Eckstein, C. Garcia-Rosales, J. Roth, W. Ottenberger, Sputtering data, Max-Planck-Institut für Plasmaphysik, 1993. http://hdl.handle.net/11858/00-001M-0000-0027-6324-6 (accessed March 11, 2020).

[117] K. Sugiyama, K. Schmid, W. Jacob, Sputtering of iron, chromium and tungsten by energetic deuterium ion bombardment, Nucl. Mater. Energy. 8 (2016) 1–7. https://doi.org/10.1016/j.nme.2016.05.016.

[118] T. Tanabe, Review of hydrogen retention in tungsten, Phys. Scr. (2014) 14044. https://doi.org/10.1088/0031-8949/2014/T159/014044.

[119] W. Wang, J. Roth, S. Lindig, C.H. Wu, Blister formation of tungsten due to ion bombardment, J. Nucl. Mater. 299 (2001) 124–131. https://doi.org/10.1016/S0022-3115(01)00679-1.

[120] M. Balden, S. Lindig, A. Manhard, J.H. You, D2 gas-filled blisters on deuterium-bombarded tungsten, J. Nucl. Mater. 414 (2011) 69–72. https://doi.org/10.1016/j.jnucmat.2011.04.031.

Acta Universitatis UpsaliensisDigital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 2008

Editor: The Dean of the Faculty of Science and Technology

A doctoral dissertation from the Faculty of Science andTechnology, Uppsala University, is usually a summary of anumber of papers. A few copies of the complete dissertationare kept at major Swedish research libraries, while thesummary alone is distributed internationally throughthe series Digital Comprehensive Summaries of UppsalaDissertations from the Faculty of Science and Technology.(Prior to January, 2005, the series was published under thetitle “Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology”.)

Distribution: publications.uu.seurn:nbn:se:uu:diva-433483

ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2021