EXPERIMENTAL SPECTROSCOPIC STUDIES OF METALS WITH ELECTRON, ION, AND OPTICAL TECHNIQUES ARI MÄKINEN

ISBN 978-952-62-0313-3 ISBN 978-952-62-0314-9 (PDF) ISSN 1239-4327

REPORT SERIES IN PHYSICAL SCIENCES Report No. 87 (2013)

EXPERIMENTAL SPECTROSCOPIC STUDIES OF METALS WITH ELECTRON, ION, AND OPTICAL TECHNIQUES

EXPERIMENTAL SPECTROSCOPIC STUDIES OF METALS WITH ELECTRON, ION, AND OPTICAL TECHNIQUES

ARI MÄKINEN

ARI MÄKINEN

Department of Physics University of Oulu Finland

Department of Physics University of Oulu Finland

Academic Dissertation to be presented with the assent of the Faculty of Science, University of Oulu, for public discussion in the Auditorium TA105, on December 10th, 2013, at 12 o’clock noon.

Academic Dissertation to be presented with the assent of the Faculty of Science, University of Oulu, for public discussion in the Auditorium TA105, on December 10th, 2013, at 12 o’clock noon.

Report series in physical sciences Oulu 2013 • University of Oulu

Report No. 87

Report series in physical sciences Oulu 2013 • University of Oulu

Report No. 87

Opponent Prof. Ergo Nõmmiste, University of Tartu, Estonia

Opponent Prof. Ergo Nõmmiste, University of Tartu, Estonia

Reviewers Prof. Mika Valden, Tampere University of Technology, Finland Prof. emer. Juhani Väyrynen, University of Turku, Finland

Reviewers Prof. Mika Valden, Tampere University of Technology, Finland Prof. emer. Juhani Väyrynen, University of Turku, Finland

Custos Prof. emer. Helena Aksela, University of Oulu, Finland

Custos Prof. emer. Helena Aksela, University of Oulu, Finland

ISBN 978-952-62-0313-3 ISBN 978-952-62-0314-9 (PDF) ISSN 1239-4327 Juvenes Print - Suomen Yliopistopaino Oy Oulu 2013

ISBN 978-952-62-0313-3 ISBN 978-952-62-0314-9 (PDF) ISSN 1239-4327 Juvenes Print - Suomen Yliopistopaino Oy Oulu 2013

i

i

Mäkinen, Ari Pekka: Experimental spectroscopic studies of metals with electron, ion, and optical techniques

Mäkinen, Ari Pekka: Experimental spectroscopic studies of metals with electron, ion, and optical techniques

Department of Physics P.O. Box 3000 FIN-90014 University of Oulu FINLAND

Department of Physics P.O. Box 3000 FIN-90014 University of Oulu FINLAND

Abstract

Abstract

In this thesis, different spectroscopic methods are used for studying metals. Electron spectroscopy is applied for the study of binding energy shifts between atomic vapor and solid metals. Photoionization and Auger decay of high temperature aluminum vapors are investigated. Ionization of atomic chromium metal vapor by light absorption is studied with synchrotron radiation and time-of-flight ion mass spectroscopy. Optical spectroscopy is used for studying light emission from electric arc furnace plasma in experimental apparatuses developed during this work. Experimental techniques and sample preparation methods are presented.

In this thesis, different spectroscopic methods are used for studying metals. Electron spectroscopy is applied for the study of binding energy shifts between atomic vapor and solid metals. Photoionization and Auger decay of high temperature aluminum vapors are investigated. Ionization of atomic chromium metal vapor by light absorption is studied with synchrotron radiation and time-of-flight ion mass spectroscopy. Optical spectroscopy is used for studying light emission from electric arc furnace plasma in experimental apparatuses developed during this work. Experimental techniques and sample preparation methods are presented.

Key words: Electron spectroscopy, Auger decay, mass spectroscopy, optical emission spectroscopy, photoionization, electric arc furnace.

Key words: Electron spectroscopy, Auger decay, mass spectroscopy, optical emission spectroscopy, photoionization, electric arc furnace.

ii

ii

Acknowledgments

Acknowledgments

The work presented in this thesis is mainly carried out at the department of physics at University of Oulu. I would like to thank the staff and the head of the department Prof. Matti Weckström for placing the facilities for my disposal. I want to thank my supervisors Prof. emer. Helena Aksela and Prof. Timo Fabritius for guidance and practical advices. This work is a result of teamwork and as such I deeply appreciate the help of all the people involved in this work. I want to thank Prof. Marko Huttula for valuable support and encouragement during the work. I want to thank Prof. emer. Seppo Aksela for introducing and guiding me to electron spectroscopy and the experimental physics. I thank Dr. Minna Patanen and Dr. Johannes Niskanen for their contributions and guidance during the work. Dr. Samuli Urpelainen is also acknowledged for his support and expertise in experiments carried out at MAX-lab. I want to express my gratitude for all the past and present members of the electron spectroscopy group for encouragements and friendly atmosphere. Henri Tikkala is acknowledged for assistance during the laboratory experiments. I especially want to thank and acknowledge Pentti Kovala for supporting device design and development work, as well as the people in the physics department workshop for making mechanical parts needed for experimental research to be possible. I am especially thankful for colleagues Antti Kettunen and Mikko-Heikki Mikkelä for mental support and interesting and intellectual discussions related to work and life in general. Part of the experimental work is carried out at MAX-lab research facility at Lund, Sweden. I thank the board and the staff of the laboratory for providing the experimental time in the facility and for their support. I like to acknowledge the financial support from Tekes via FIMECC-ELEMET program, EffArc project. All the EffArc project participants are acknowledged for their involvement and contributions to the research. I want to thank the members of process metallurgy laboratory at University of Oulu for interesting discussions and information regrading the use of electric arc

The work presented in this thesis is mainly carried out at the department of physics at University of Oulu. I would like to thank the staff and the head of the department Prof. Matti Weckström for placing the facilities for my disposal. I want to thank my supervisors Prof. emer. Helena Aksela and Prof. Timo Fabritius for guidance and practical advices. This work is a result of teamwork and as such I deeply appreciate the help of all the people involved in this work. I want to thank Prof. Marko Huttula for valuable support and encouragement during the work. I want to thank Prof. emer. Seppo Aksela for introducing and guiding me to electron spectroscopy and the experimental physics. I thank Dr. Minna Patanen and Dr. Johannes Niskanen for their contributions and guidance during the work. Dr. Samuli Urpelainen is also acknowledged for his support and expertise in experiments carried out at MAX-lab. I want to express my gratitude for all the past and present members of the electron spectroscopy group for encouragements and friendly atmosphere. Henri Tikkala is acknowledged for assistance during the laboratory experiments. I especially want to thank and acknowledge Pentti Kovala for supporting device design and development work, as well as the people in the physics department workshop for making mechanical parts needed for experimental research to be possible. I am especially thankful for colleagues Antti Kettunen and Mikko-Heikki Mikkelä for mental support and interesting and intellectual discussions related to work and life in general. Part of the experimental work is carried out at MAX-lab research facility at Lund, Sweden. I thank the board and the staff of the laboratory for providing the experimental time in the facility and for their support. I like to acknowledge the financial support from Tekes via FIMECC-ELEMET program, EffArc project. All the EffArc project participants are acknowledged for their involvement and contributions to the research. I want to thank the members of process metallurgy laboratory at University of Oulu for interesting discussions and information regrading the use of electric arc

iii

iii

iv

iv

furnaces for steel production. Especially Olli Mattila and Juha Roininen are acknowledged for their expertise in process metallurgy. Industrial partners of the project including Outokumpu Tornio Works and staff of the plant are acknowledged for support and making part of the research work possible. I want to thank all my friends for encouragements during the course of this work. Finally I want to thank my parents Jorma and Ulla-Maija, and sister Jaana for their support.

furnaces for steel production. Especially Olli Mattila and Juha Roininen are acknowledged for their expertise in process metallurgy. Industrial partners of the project including Outokumpu Tornio Works and staff of the plant are acknowledged for support and making part of the research work possible. I want to thank all my friends for encouragements during the course of this work. Finally I want to thank my parents Jorma and Ulla-Maija, and sister Jaana for their support.

Oulu, December 2013 Ari Mäkinen

Oulu, December 2013 Ari Mäkinen

v

v

LIST OF ORIGINAL PAPERS

LIST OF ORIGINAL PAPERS

This thesis contains an introductory part and the following papers, which will be referred in the text by their Roman numbers.

This thesis contains an introductory part and the following papers, which will be referred in the text by their Roman numbers.

I M. Huttula, K. Jänkälä, A. Mäkinen, H. Aksela, and S. Aksela, Core shell electron spectroscopy on high temperature vapors: 2s photoionization and Auger decay of atomic aluminium, New J. Phys. 10, 013009 (2008).

I M. Huttula, K. Jänkälä, A. Mäkinen, H. Aksela, and S. Aksela, Core shell electron spectroscopy on high temperature vapors: 2s photoionization and Auger decay of atomic aluminium, New J. Phys. 10, 013009 (2008).

II M. Huttula, L. Partanen, A. Mäkinen, T. Kantia, H. Aksela, and S. Aksela, KLL Auger decay in free aluminum atoms, Phys. Rev. A 79, 023412 (2009).

II M. Huttula, L. Partanen, A. Mäkinen, T. Kantia, H. Aksela, and S. Aksela, KLL Auger decay in free aluminum atoms, Phys. Rev. A 79, 023412 (2009).

III S. Aksela, T. Kantia, M. Patanen, A. Mäkinen, S. Urpelainen, and H. Aksela, Accurate free atom–solid binding energy shifts for Au and Ag, J. Electron Spectrosc. Relat. Phenom. 185, 273-277 (2012).

III S. Aksela, T. Kantia, M. Patanen, A. Mäkinen, S. Urpelainen, and H. Aksela, Accurate free atom–solid binding energy shifts for Au and Ag, J. Electron Spectrosc. Relat. Phenom. 185, 273-277 (2012).

IV A. Mäkinen, M. Patanen, S. Aksela, and H. Aksela, Atom-solid 3p level binding energy shift of transition metals Cr, Mn, Fe, Co, and Ni , J. Electron Spectrosc. Relat. Phenom. 185, 573-577 (2012).

IV A. Mäkinen, M. Patanen, S. Aksela, and H. Aksela, Atom-solid 3p level binding energy shift of transition metals Cr, Mn, Fe, Co, and Ni , J. Electron Spectrosc. Relat. Phenom. 185, 573-577 (2012).

V A. Mäkinen, J. Niskanen, and H. Aksela, Relative photoionization cross section of Cr atoms in the valence region, Phys. Rev. A 85, 053411 (2012).

V A. Mäkinen, J. Niskanen, and H. Aksela, Relative photoionization cross section of Cr atoms in the valence region, Phys. Rev. A 85, 053411 (2012).

VI A. Mäkinen, J. Niskanen, H. Tikkala, and H. Aksela, Optical emission from a small scale model electric arc furnace in 250–600 nm region, Rev. Sci. Instrum. 84, 043111 (2013).

VI A. Mäkinen, J. Niskanen, H. Tikkala, and H. Aksela, Optical emission from a small scale model electric arc furnace in 250–600 nm region, Rev. Sci. Instrum. 84, 043111 (2013).

All the articles and the work reported in this thesis are the result of teamwork. The author has been extensively involved in the measurement technique and device development, which include development of experimental techniques utilizing induction and electric arc heating as well as electron and photon emission spectra collection techniques. The author contributed to the experimental work in Papers I and II, and has been responsible for writing and taking the main responsibilities of experimental work in Papers IV, V, and VI. The author made a strong contribution to the data treatment in Paper III and carried out experimental data treatment in Papers IV and V. The author has also been responsible of development of the laboratory scale electric arc furnace presented in Paper VI. As preceding development work, the author has been developing, leading,

All the articles and the work reported in this thesis are the result of teamwork. The author has been extensively involved in the measurement technique and device development, which include development of experimental techniques utilizing induction and electric arc heating as well as electron and photon emission spectra collection techniques. The author contributed to the experimental work in Papers I and II, and has been responsible for writing and taking the main responsibilities of experimental work in Papers IV, V, and VI. The author made a strong contribution to the data treatment in Paper III and carried out experimental data treatment in Papers IV and V. The author has also been responsible of development of the laboratory scale electric arc furnace presented in Paper VI. As preceding development work, the author has been developing, leading,

vi

vi

and participating the experiment [1] involving an electric arc apparatus for optical emission experiment leading to the work presented in Paper VI. In addition to the included papers, as a part of FIMECC-ELEMET program EffArc project, the author has developed an experimental device [2] for collecting light from a large scale electrical industrial furnaces described in Chapter 3.3.3. The author participated in the experimental work at Tornio steel works factory [2] stimulating a separate commercialization spin off project to the EffArc project. The author has also contributed to the following papers, which are not included in this thesis:

and participating the experiment [1] involving an electric arc apparatus for optical emission experiment leading to the work presented in Paper VI. In addition to the included papers, as a part of FIMECC-ELEMET program EffArc project, the author has developed an experimental device [2] for collecting light from a large scale electrical industrial furnaces described in Chapter 3.3.3. The author participated in the experimental work at Tornio steel works factory [2] stimulating a separate commercialization spin off project to the EffArc project. The author has also contributed to the following papers, which are not included in this thesis:

1. T. Rander, J. Schulz, M. Huttula, A. Mäkinen, M. Tchaplyguine, S. Svensson, G. Öhrwall, O. Björneholm, S. Aksela, and H. Aksela, Corelevel electron spectroscopy on the sodium dimer Na 2p level, Phys. Rev. A 75, 032510 (2007).

1. T. Rander, J. Schulz, M. Huttula, A. Mäkinen, M. Tchaplyguine, S. Svensson, G. Öhrwall, O. Björneholm, S. Aksela, and H. Aksela, Corelevel electron spectroscopy on the sodium dimer Na 2p level, Phys. Rev. A 75, 032510 (2007).

2. S. Urpelainen, M. Huttula, P. Kovala, A. Mäkinen, A. Caló, S. Aksela, H. Aksela, High resolution electron energy loss spectrometer for the study of vapor phase samples, J. Electron Spectrosc. Relat. Phenom. 156-158, 145-149 (2007).

2. S. Urpelainen, M. Huttula, P. Kovala, A. Mäkinen, A. Caló, S. Aksela, H. Aksela, High resolution electron energy loss spectrometer for the study of vapor phase samples, J. Electron Spectrosc. Relat. Phenom. 156-158, 145-149 (2007).

3. N. M. Suni, M. Haapala, A. Mäkinen, L. Sainiemi, S. Franssila, E. Färm, E. Puukilainen, M. Ritala, and R. Kostiainen, Selective Surface Patterning with an Electric Discharge in the Fabrication of Microfluidic Structures, Angew. Chem. Int. Ed. 47, 7442-7445 (2008).

3. N. M. Suni, M. Haapala, A. Mäkinen, L. Sainiemi, S. Franssila, E. Färm, E. Puukilainen, M. Ritala, and R. Kostiainen, Selective Surface Patterning with an Electric Discharge in the Fabrication of Microfluidic Structures, Angew. Chem. Int. Ed. 47, 7442-7445 (2008).

4. M. Aula, A. Mäkinen, and T. Fabritius, Analysis of Arc Emission Spectra of Stainless Steel Electric Arc Furnace Slag Affected by Fluctuating Arc Voltage, Appl. Spectrosc. 68 (2014), Accepted for publication.

4. M. Aula, A. Mäkinen, and T. Fabritius, Analysis of Arc Emission Spectra of Stainless Steel Electric Arc Furnace Slag Affected by Fluctuating Arc Voltage, Appl. Spectrosc. 68 (2014), Accepted for publication.

Contents

Contents

Abstract

i

Acknowledgments

iii

Abstract

i

Acknowledgments

iii

List of original papers

v

List of original papers

v

1 Introduction

1

1 Introduction

1

3 3 5 6 7 9

2 Electronic structure and concepts 2.1 Electronic structure of atoms and molecules . . 2.2 Transitions in atoms and molecules . . . . . . . 2.2.1 Photoexcitation and ionization in atoms 2.2.2 Photon and electron emission in atoms . 2.2.3 Transitions in molecules . . . . . . . . .

2 Electronic structure and concepts 2.1 Electronic structure of atoms and molecules . . 2.2 Transitions in atoms and molecules . . . . . . . 2.2.1 Photoexcitation and ionization in atoms 2.2.2 Photon and electron emission in atoms . 2.2.3 Transitions in molecules . . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

3 Vapor production and sample preparation 3.1 Resistive ovens . . . . . . . . . . . . . . . . . . . . . . 3.2 Induction heating . . . . . . . . . . . . . . . . . . . . . 3.2.1 Simultaneous measurement of atomic and solid spectra . . . . . . . . . . . . . . . . . . . . . . . 3.3 Electric arc heating . . . . . . . . . . . . . . . . . . . . 3.3.1 Device for arc emission characterization . . . . . 3.3.2 Laboratory scale miniature EAF for arc plasma acterization . . . . . . . . . . . . . . . . . . . . 3.3.3 Experiment in Tornio Steel Works . . . . . . . .

. . . . . . state . . . . . . . . . char. . . . . .

. 15 . 16 . 17

4 Electron, ion, and photon detection 4.1 Electron spectrometer . . . . . . . . . . 4.2 Time-of-flight spectrometer . . . . . . . 4.3 Optical spectrometer . . . . . . . . . . . 4.3.1 Resolving power and transmission 4.3.2 Detector . . . . . . . . . . . . . .

. . . . .

. . . . .

vii

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

11 . 12 . 12

. 18 . 20 23 23 25 26 26 28

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

3 3 5 6 7 9

3 Vapor production and sample preparation 3.1 Resistive ovens . . . . . . . . . . . . . . . . . . . . . . 3.2 Induction heating . . . . . . . . . . . . . . . . . . . . . 3.2.1 Simultaneous measurement of atomic and solid spectra . . . . . . . . . . . . . . . . . . . . . . . 3.3 Electric arc heating . . . . . . . . . . . . . . . . . . . . 3.3.1 Device for arc emission characterization . . . . . 3.3.2 Laboratory scale miniature EAF for arc plasma acterization . . . . . . . . . . . . . . . . . . . . 3.3.3 Experiment in Tornio Steel Works . . . . . . . .

. . . . . . state . . . . . . . . . char. . . . . .

. 15 . 16 . 17

4 Electron, ion, and photon detection 4.1 Electron spectrometer . . . . . . . . . . 4.2 Time-of-flight spectrometer . . . . . . . 4.3 Optical spectrometer . . . . . . . . . . . 4.3.1 Resolving power and transmission 4.3.2 Detector . . . . . . . . . . . . . .

. . . . .

. . . . .

vii

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

11 . 12 . 12

. 18 . 20 23 23 25 26 26 28

5 Excitation methods 5.1 Electron beam . . . . . . . . . . . . . 5.2 Synchrotron radiation . . . . . . . . 5.2.1 Beamline I411 . . . . . . . . . 5.2.2 Beamline I3 - FINEST branch 5.3 Electric arc . . . . . . . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

29 29 30 32 32 33

5 Excitation methods 5.1 Electron beam . . . . . . . . . . . . . 5.2 Synchrotron radiation . . . . . . . . 5.2.1 Beamline I411 . . . . . . . . . 5.2.2 Beamline I3 - FINEST branch 5.3 Electric arc . . . . . . . . . . . . . .

spectra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

35 35 35 36 36 38 39 40

6 Data handling for electron, ion, and 6.1 Electron spectra . . . . . . . . . . . 6.1.1 Transmission . . . . . . . . 6.1.2 Background substraction . . 6.2 Mass spectra . . . . . . . . . . . . 6.3 Optical spectra . . . . . . . . . . . 6.3.1 Simulated spectra . . . . . . 6.3.2 Background subtraction . .

43 43 44 45 46

7 Summary and discussion of included papers 7.1 High temperature vapors: spectroscopy of aluminum, Papers I - II . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2 Atom-solid binding energy shifts, Papers III - IV . . . . . . . 7.3 Relative photoionization cross section of Cr atoms, Paper V 7.4 Optical emission from electric arc furnace, Paper VI . . . . .

8 Conclusions and future prospects

49

8 Conclusions and future prospects

49

Bibliography

51

Bibliography

51

Original papers

57

Original papers

57

6 Data handling for electron, ion, and 6.1 Electron spectra . . . . . . . . . . . 6.1.1 Transmission . . . . . . . . 6.1.2 Background substraction . . 6.2 Mass spectra . . . . . . . . . . . . 6.3 Optical spectra . . . . . . . . . . . 6.3.1 Simulated spectra . . . . . . 6.3.2 Background subtraction . .

. . . . .

. . . . .

. . . . .

. . . . .

photon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

7 Summary and discussion of included papers 7.1 High temperature vapors: spectroscopy of aluminum, Papers I - II . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2 Atom-solid binding energy shifts, Papers III - IV . . . . . . . 7.3 Relative photoionization cross section of Cr atoms, Paper V 7.4 Optical emission from electric arc furnace, Paper VI . . . . .

viii

. . . .

viii

. . . . .

. . . . .

. . . . .

. . . . .

photon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

29 29 30 32 32 33

spectra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

35 35 35 36 36 38 39 40

. . . . .

. . . . .

. . . . .

. . . . .

43 . . . .

43 44 45 46

Chapter 1

Chapter 1

Introduction

Introduction

Physics and its applications form the basis for many marvels of modern technology. Tools and methods discovered and introduced during the development of physics have greatly benefited human kind. Atomic physics which studies tiny discrete units of matter called atoms and their interactions has evolved significantly during the course of time. During the last century atomic and molecular physics advanced greatly. The concept of atom itself is very old originating to ancient Greek times. Since then, the knowledge of structure of the atom has greatly improved. The electron was discovered in 1897 by Thomson [3] followed by the discovery of atomic nucleus in 1909 [4] and formulation of the first model of the atom where the electrons orbit the atomic nucleus. This was soon followed by explanation of simple emission spectra of the atom by Niels Bohr [5]. Even before the discovery of the electron, the history of spectroscopy had already begun by Newton in 1666. He used the term spectrum to describe the continuous series of colors produced by his experimental apparatus using prism combined with an aperture, a lens, a screen [6]. Since then the use of spectroscopy as a tool for research and applications has widen to cover energy spectrum of particles. In 1905 Albert Einstein published his explanation of photoelectric effect, in which the electrons are removed from the matter due to photon absorption and the energy of the ejected electrons depends on the wavelength of absorbed photons [7]. One of the fundamental techniques to study the properties of matter at atomic level is electron spectroscopy, which became widely known and used tool for research and applications from the work of Kai Siegbahn et al. in the 1960s [8, 9]. Siegbahn received the Nobel Price for his work in 1981. He had build a high resolution electron spectrometer and established X-ray photoelectron spectroscopy as an important tool for research and applications.

Physics and its applications form the basis for many marvels of modern technology. Tools and methods discovered and introduced during the development of physics have greatly benefited human kind. Atomic physics which studies tiny discrete units of matter called atoms and their interactions has evolved significantly during the course of time. During the last century atomic and molecular physics advanced greatly. The concept of atom itself is very old originating to ancient Greek times. Since then, the knowledge of structure of the atom has greatly improved. The electron was discovered in 1897 by Thomson [3] followed by the discovery of atomic nucleus in 1909 [4] and formulation of the first model of the atom where the electrons orbit the atomic nucleus. This was soon followed by explanation of simple emission spectra of the atom by Niels Bohr [5]. Even before the discovery of the electron, the history of spectroscopy had already begun by Newton in 1666. He used the term spectrum to describe the continuous series of colors produced by his experimental apparatus using prism combined with an aperture, a lens, a screen [6]. Since then the use of spectroscopy as a tool for research and applications has widen to cover energy spectrum of particles. In 1905 Albert Einstein published his explanation of photoelectric effect, in which the electrons are removed from the matter due to photon absorption and the energy of the ejected electrons depends on the wavelength of absorbed photons [7]. One of the fundamental techniques to study the properties of matter at atomic level is electron spectroscopy, which became widely known and used tool for research and applications from the work of Kai Siegbahn et al. in the 1960s [8, 9]. Siegbahn received the Nobel Price for his work in 1981. He had build a high resolution electron spectrometer and established X-ray photoelectron spectroscopy as an important tool for research and applications.

1

1

From the Newton era optical spectroscopy as an experimental method has advanced greatly. Because electrons occupy element specific energy levels in atoms, molecules and solids, the emitted light is characteristic for the studied matter. Nowadays good optical emission line databases (e.g. NIST [10]) are available and they can be used to determine different elements from the observed optical emission spectra. There are several experimental techniques to excite samples to emit light such as lasers [11, 12], synchrotron radiation (SR) [13], induction plasma [14], laser induced plasma [15], flames, and sparks [16]. Electric arc furnaces (EAFs) used in metal manufacturing industry [17–19] can in principle be seen as large scale excitation sources of melted materials as the electric arc produces heat, but also evaporates and excites melted materials. EAF provides high intensity excitation leading to intense optical emission from the arc plasma. In this work different spectroscopic methods are used to investigate properties of metal atoms in vapor and solid phases. The utilized techniques are electron spectroscopy, time-of-flight mass spectroscopy and optical emission spectroscopy (OES). Sample preparation techniques for atomic vapor production and experimental device development for optical emission studies of EAF plasmas are also presented.

From the Newton era optical spectroscopy as an experimental method has advanced greatly. Because electrons occupy element specific energy levels in atoms, molecules and solids, the emitted light is characteristic for the studied matter. Nowadays good optical emission line databases (e.g. NIST [10]) are available and they can be used to determine different elements from the observed optical emission spectra. There are several experimental techniques to excite samples to emit light such as lasers [11, 12], synchrotron radiation (SR) [13], induction plasma [14], laser induced plasma [15], flames, and sparks [16]. Electric arc furnaces (EAFs) used in metal manufacturing industry [17–19] can in principle be seen as large scale excitation sources of melted materials as the electric arc produces heat, but also evaporates and excites melted materials. EAF provides high intensity excitation leading to intense optical emission from the arc plasma. In this work different spectroscopic methods are used to investigate properties of metal atoms in vapor and solid phases. The utilized techniques are electron spectroscopy, time-of-flight mass spectroscopy and optical emission spectroscopy (OES). Sample preparation techniques for atomic vapor production and experimental device development for optical emission studies of EAF plasmas are also presented.

2

2

Chapter 2

Chapter 2

Electronic structure and concepts

Electronic structure and concepts

2.1

2.1

Electronic structure of atoms and molecules

Electronic structure of atoms and molecules

The electronic structure of atoms and molecules largely determines the properties of matter, chemical reactions, and interactions with light and matter. Experimental observations of the behavior of matter can be explained and predicted by the theory of quantum mechanics (QM). QM is the most accurate theory to day to predict the behavior of matter in small scales and it reproduces results of experimental observations with high accuracy. In QM there is a wave function associated with the particles, for example electrons. In 1926 Erwin Schrödinger proposed the use of an equation which related energy state of the system to the wave function of a particle, i.e. the wave equation [20–23]

The electronic structure of atoms and molecules largely determines the properties of matter, chemical reactions, and interactions with light and matter. Experimental observations of the behavior of matter can be explained and predicted by the theory of quantum mechanics (QM). QM is the most accurate theory to day to predict the behavior of matter in small scales and it reproduces results of experimental observations with high accuracy. In QM there is a wave function associated with the particles, for example electrons. In 1926 Erwin Schrödinger proposed the use of an equation which related energy state of the system to the wave function of a particle, i.e. the wave equation [20–23]

ˆ Eψ = Hψ,

ˆ Eψ = Hψ,

(2.1)

(2.1)

ˆ is where E is the total energy of the system, ψ is the wave function, and H ˆ depends on the the Hamiltonian operator. The Hamiltonian of the system H investigated system and contains the interactions of the system. In the case of non-relativistic time independent Schödinger equation of a simple single electron hydrogen like atom, the Hamiltonian is form, where the first part of the sum is the kinetic energy operator and the second one is the potential energy operator.

ˆ is where E is the total energy of the system, ψ is the wave function, and H ˆ depends on the the Hamiltonian operator. The Hamiltonian of the system H investigated system and contains the interactions of the system. In the case of non-relativistic time independent Schödinger equation of a simple single electron hydrogen like atom, the Hamiltonian is form, where the first part of the sum is the kinetic energy operator and the second one is the potential energy operator.

2 ˆ = −~ ∇2 + V (r), H 2µ

2 ˆ = −~ ∇2 + V (r), H 2µ

(2.2)

(2.2)

where ~ is the reduced Planck constant and µ is the reduced mass of the electron. The nucleus of an atom forms an electrical potential well which holds the electron in orbit and is described by term V as a function of distance

where ~ is the reduced Planck constant and µ is the reduced mass of the electron. The nucleus of an atom forms an electrical potential well which holds the electron in orbit and is described by term V as a function of distance

3

3

r from the nucleus. Now, the eigenvalue solutions for equation 2.1 describe allowed discrete energy states of the atom. In general, due to the discrete nature of atomic electron systems, the state of an atom can be characterized using quantum numbers. The state of a bound electron in an atom can be described using four quantum numbers:

r from the nucleus. Now, the eigenvalue solutions for equation 2.1 describe allowed discrete energy states of the atom. In general, due to the discrete nature of atomic electron systems, the state of an atom can be characterized using quantum numbers. The state of a bound electron in an atom can be described using four quantum numbers:

• The principal quantum number n

• The principal quantum number n

• The orbital quantum number (azimuthal quantum number) l

• The orbital quantum number (azimuthal quantum number) l

• The magnetic quantum number ml

• The magnetic quantum number ml

• The spin (projection) quantum number ms

• The spin (projection) quantum number ms

The principal quantum number n defines the electron shell which the electron belongs to and it is described using integers n = 1, 2, 3, . . . . The orbital quantum number l is associated to electron subshell and angular momentum can have values l = 0, 1, 2, . . . , n − 1. Usually subshells are denoted as letters s, p, d, f . . . correspondingly. The magnetic quantum number ml is associated to the orientation of the shape of the subshell and can have the values −l to l. The spin quantum number ms is an inherent property of the electron and can have only two possible states up +1/2 or down −1/2. The distribution of the electrons of an atom is described using an electron configuration. The quantum numbers of the system obey the Pauli exclusion principle [24], which states that the quantum numbers of the electrons in an atom can not be identical. This combined with the fact that the energy of the atom is minimized leads to ground state electron configurations. For example the electron configuration of the iron atom in the ground state is

The principal quantum number n defines the electron shell which the electron belongs to and it is described using integers n = 1, 2, 3, . . . . The orbital quantum number l is associated to electron subshell and angular momentum can have values l = 0, 1, 2, . . . , n − 1. Usually subshells are denoted as letters s, p, d, f . . . correspondingly. The magnetic quantum number ml is associated to the orientation of the shape of the subshell and can have the values −l to l. The spin quantum number ms is an inherent property of the electron and can have only two possible states up +1/2 or down −1/2. The distribution of the electrons of an atom is described using an electron configuration. The quantum numbers of the system obey the Pauli exclusion principle [24], which states that the quantum numbers of the electrons in an atom can not be identical. This combined with the fact that the energy of the atom is minimized leads to ground state electron configurations. For example the electron configuration of the iron atom in the ground state is

1s2 2s2 2p6 3s2 3p6 3d6 4s2 .

1s2 2s2 2p6 3s2 3p6 3d6 4s2 .

Eigenstates of the total angular momentum of the quantum mechanical system are formed by coupling individual angular momenta. In light atoms (Z < 30) a coupling scheme called LS coupling, also known as RussellSaunders coupling [25] can be used. Electron spins interact therefore angular momenta of spins si are combined to form the total spin angular momentum S = Σsi . Also orbital angular momenta li are combined to form the total orbital angular momentum L = Σli . Coupling of L and S forms the total angular momentum J = L + S1 .

Eigenstates of the total angular momentum of the quantum mechanical system are formed by coupling individual angular momenta. In light atoms (Z < 30) a coupling scheme called LS coupling, also known as RussellSaunders coupling [25] can be used. Electron spins interact therefore angular momenta of spins si are combined to form the total spin angular momentum S = Σsi . Also orbital angular momenta li are combined to form the total orbital angular momentum L = Σli . Coupling of L and S forms the total angular momentum J = L + S1 .

1 Note that the total electronic angular momentum is a vector sum and assumes values L + S, L + S − 1, . . . , |L − S|.

1 Note that the total electronic angular momentum is a vector sum and assumes values L + S, L + S − 1, . . . , |L − S|.

4

4

In multielectron atoms angular momentum quantum numbers can be described using Russell-Saunders term symbols which are related to the energy level of the system in some specific electron configuration: 2S+1

LJ ,

In multielectron atoms angular momentum quantum numbers can be described using Russell-Saunders term symbols which are related to the energy level of the system in some specific electron configuration: 2S+1

(2.3)

LJ ,

(2.3)

where 2S + 1 is the spin multiplicity and numerical value of L is replaced with capital letter S, P, D, F, . . . which correspond to values 0, 1, 2, 3, . . . respectively. The ground state of Fe corresponds to term symbol 5 D4 .

where 2S + 1 is the spin multiplicity and numerical value of L is replaced with capital letter S, P, D, F, . . . which correspond to values 0, 1, 2, 3, . . . respectively. The ground state of Fe corresponds to term symbol 5 D4 .

2.2

2.2

Transitions in atoms and molecules

Transitions in atoms and molecules

An atom or a molecule can be excited by photons, electrons, or other particle collisions and interactions. Thermal energy affects the population distribution of energy levels in atoms as well as in solids. The relative population between two energy levels a and b at certain temperature T in an atom can be calculated by using Boltzmann’s Law [26]

An atom or a molecule can be excited by photons, electrons, or other particle collisions and interactions. Thermal energy affects the population distribution of energy levels in atoms as well as in solids. The relative population between two energy levels a and b at certain temperature T in an atom can be calculated by using Boltzmann’s Law [26]

gb Ea −Eb nb = e kT , na ga

gb Ea −Eb nb = e kT , na ga

(2.4)

(2.4)

where k is the Boltzmann constant, na and nb are the populations of lower and higher lever respectively, and ga and gb are the statistical weights of the levels. Statistical weight is the number of sublevels of equal energy and can be calculated from the total angular momentum quantum number J as g = 2J + 1. Since energy level populations of atoms are temperature dependent, it is possible to estimate the temperature via intensities of atomic emission lines in plasmas. In Paper VI the Boltzmann plot method [27] was used to approximate average temperatures of produced arc plasmas. Photon absorption may bring an atom into an excited state where an electron is lifted from a lower energy level to a higher energy level. This excited state can decay via multiple decay paths. With adequately high photon energies the electron can also be ejected completely from the atom and bring the system into a photoionized state which can also be an excited state. The possible decay paths of excited atoms are photon and electron emission. In addition to direct ejection of the electron, the Auger phenomena can occur. In case of molecules, similar processes for excitation are possible, and in some cases chemical reactions can leave the molecule in an excited state.

where k is the Boltzmann constant, na and nb are the populations of lower and higher lever respectively, and ga and gb are the statistical weights of the levels. Statistical weight is the number of sublevels of equal energy and can be calculated from the total angular momentum quantum number J as g = 2J + 1. Since energy level populations of atoms are temperature dependent, it is possible to estimate the temperature via intensities of atomic emission lines in plasmas. In Paper VI the Boltzmann plot method [27] was used to approximate average temperatures of produced arc plasmas. Photon absorption may bring an atom into an excited state where an electron is lifted from a lower energy level to a higher energy level. This excited state can decay via multiple decay paths. With adequately high photon energies the electron can also be ejected completely from the atom and bring the system into a photoionized state which can also be an excited state. The possible decay paths of excited atoms are photon and electron emission. In addition to direct ejection of the electron, the Auger phenomena can occur. In case of molecules, similar processes for excitation are possible, and in some cases chemical reactions can leave the molecule in an excited state.

5

5

2.2.1

Photoexcitation and ionization in atoms

2.2.1

Photoexcitation and ionization in atoms

Light-matter interaction can cause excitation or ionization of matter. If the photon energy matches the energy difference between energy levels a resonance excitation (Figure 2.1) can be possible, depending on selection rules of the photon absorbtion. Another possibility is photoionization, the effect which was reported by Heinrich Hertz 1887 when he noticed charged objects to lose their charge faster under ultraviolet light irradiation [28]. Later in 1905 Albert Einstein formulated a theoretical explanation of this phenomenon, known nowadays as the photoelectric effect, which brought him the Nobel prize [7]. His explanation stated that a quanta of light, known as photon, is fully absorbed as a whole and is capable of removing an electron from the material if the energy hν of the photon is larger than the work function of the material. The work function is the energy which is required to free an electron from a Fermi level of a solid surface to the vacuum zero energy level near the surface [29]. The explanation of the photoelectric effect outlays the basic theoretical basis of photoelectron spectroscopy utilized in this work. Photoelectron spectroscopy is a very sensitive method for investigating chemical environment of atoms because electron binding energies are affected by the surrounding atoms and chemical bonds [30]. This can be clearly demonstrated by comparing photoelectron spectra of bulk and atomic vapor of the same material, as seen in Papers III and IV. In photoionization a photon is absorbed in the atom and the ejection of electron follows due to absorbed energy hν of the photon as depicted in Figure 2.2. The reaction equation for the process is:

Light-matter interaction can cause excitation or ionization of matter. If the photon energy matches the energy difference between energy levels a resonance excitation (Figure 2.1) can be possible, depending on selection rules of the photon absorbtion. Another possibility is photoionization, the effect which was reported by Heinrich Hertz 1887 when he noticed charged objects to lose their charge faster under ultraviolet light irradiation [28]. Later in 1905 Albert Einstein formulated a theoretical explanation of this phenomenon, known nowadays as the photoelectric effect, which brought him the Nobel prize [7]. His explanation stated that a quanta of light, known as photon, is fully absorbed as a whole and is capable of removing an electron from the material if the energy hν of the photon is larger than the work function of the material. The work function is the energy which is required to free an electron from a Fermi level of a solid surface to the vacuum zero energy level near the surface [29]. The explanation of the photoelectric effect outlays the basic theoretical basis of photoelectron spectroscopy utilized in this work. Photoelectron spectroscopy is a very sensitive method for investigating chemical environment of atoms because electron binding energies are affected by the surrounding atoms and chemical bonds [30]. This can be clearly demonstrated by comparing photoelectron spectra of bulk and atomic vapor of the same material, as seen in Papers III and IV. In photoionization a photon is absorbed in the atom and the ejection of electron follows due to absorbed energy hν of the photon as depicted in Figure 2.2. The reaction equation for the process is:

A + hν → A+∗ + e−

A + hν → A+∗ + e−

(2.5)

(2.5)

Depending on which orbital the electron is ejected from, the atom A can be left excited; outermost valence orbital leaves the ion A+ to an electronic ground state and inner orbital ionization leaves the ion in an excited state A+∗ . When an atom or molecule is subject to photon flux, there is a certain likelihood of photon absorption which can be characterized by using the concept of photoionization cross-section. The likelihood of the ionization is photon energy-dependent and the direct photoionization of a free atom needs sufficient photon energy to overcome the binding energy of the specific electron ejection. This can be seen for example in Paper V where ionization channels of Cr atom open up additively after photon energy is higher than the associated binding energy for each level. The kinetic energy of the ejected electron is related to the photon in following manner:

Depending on which orbital the electron is ejected from, the atom A can be left excited; outermost valence orbital leaves the ion A+ to an electronic ground state and inner orbital ionization leaves the ion in an excited state A+∗ . When an atom or molecule is subject to photon flux, there is a certain likelihood of photon absorption which can be characterized by using the concept of photoionization cross-section. The likelihood of the ionization is photon energy-dependent and the direct photoionization of a free atom needs sufficient photon energy to overcome the binding energy of the specific electron ejection. This can be seen for example in Paper V where ionization channels of Cr atom open up additively after photon energy is higher than the associated binding energy for each level. The kinetic energy of the ejected electron is related to the photon in following manner:

6

6

(2.6)

Ekin = Ehν − Ebin ,

(2.6)

Ekin = Ehν − Ebin ,

where Ekin is the kinetic energy of the electron after it is ejected, Ehν is the energy hν of an incoming photon, and Ebin is binding energy of the electron. This relation is the basis of photoelectron spectroscopy. Often, it is convenient to express electron energies in binding energy scale instead of kinetic energy since binging energy relates to energy levels.

where Ekin is the kinetic energy of the electron after it is ejected, Ehν is the energy hν of an incoming photon, and Ebin is binding energy of the electron. This relation is the basis of photoelectron spectroscopy. Often, it is convenient to express electron energies in binding energy scale instead of kinetic energy since binging energy relates to energy levels.

2.2.2

2.2.2

Photon and electron emission in atoms

In general an excited atom A , excited ion A or a molecule will rearrange its electronic configuration after the excitation so that the total energy is minimized. This can lead to emission of photons in fluorescence decay as depicted in Figure 2.1 where electronic transition between two energy levels leads to emission of a photon. Another possible decay path is the emission of electrons in the Auger process [31,32] as shown in Figure 2.2 where a normal Auger transition of a singly ionized excited atom is depicted. +∗

∗

(a)

1

(b)

2

In general an excited atom A∗ , excited ion A+∗ or a molecule will rearrange its electronic configuration after the excitation so that the total energy is minimized. This can lead to emission of photons in fluorescence decay as depicted in Figure 2.1 where electronic transition between two energy levels leads to emission of a photon. Another possible decay path is the emission of electrons in the Auger process [31,32] as shown in Figure 2.2 where a normal Auger transition of a singly ionized excited atom is depicted. (a)

3

4

Photon and electron emission in atoms

1

(b)

2

3

4

Figure 2.1: (a) Resonant photon absorption, 1. photon, and 2. electron transition to higher energy level. (b) Photon emission, 3. electron transition to lower energy level, and 4. photon emission.

Figure 2.1: (a) Resonant photon absorption, 1. photon, and 2. electron transition to higher energy level. (b) Photon emission, 3. electron transition to lower energy level, and 4. photon emission.

The Auger process occurs when an electron hole caused by photoionization is filled by an electron from outer shell and the electromagnetic Coulomb interaction between these electrons, which can be described as virtual photon emission and absorption [33], causes another electron ejection from the atom. The ejected electron is called the Auger electron which kinetic energy equals to the energy difference of initial hole state and final state of the transition.

The Auger process occurs when an electron hole caused by photoionization is filled by an electron from outer shell and the electromagnetic Coulomb interaction between these electrons, which can be described as virtual photon emission and absorption [33], causes another electron ejection from the atom. The ejected electron is called the Auger electron which kinetic energy equals to the energy difference of initial hole state and final state of the transition.

7

7

(a)

(b) 1

(a) 4

2 3

(b) 1

5

4

2 3

5

Figure 2.2: (a) Photoelectron emission, 1. photon, and 2. photoelectron. (b) Auger electron emission, 3. electron hole in a lower level, 4. electron transition, and 5. Auger electron ejection.

Figure 2.2: (a) Photoelectron emission, 1. photon, and 2. photoelectron. (b) Auger electron emission, 3. electron hole in a lower level, 4. electron transition, and 5. Auger electron ejection.

An Auger transition is written using nomenclature ABC where symbol A indicates the shell of an initial electron hole, B indicates the initial shell of the second electron, and C indicates the shell where the Auger electron is ejected. For symbols ABC the X-ray notation K, L, M,. . . is used [34]. For example in Figure 2.2 LMM Auger decay is depicted. For the Auger decay to occur the energy of the vacancy state of the atom must be higher than the energy needed to free the Auger electron. The Auger decay is subject to the following selection rule:

An Auger transition is written using nomenclature ABC where symbol A indicates the shell of an initial electron hole, B indicates the initial shell of the second electron, and C indicates the shell where the Auger electron is ejected. For symbols ABC the X-ray notation K, L, M,. . . is used [34]. For example in Figure 2.2 LMM Auger decay is depicted. For the Auger decay to occur the energy of the vacancy state of the atom must be higher than the energy needed to free the Auger electron. The Auger decay is subject to the following selection rule:

∆L = ∆S = ∆J = 0

(2.7)

∆L = ∆S = ∆J = 0

(2.7)

Photon absorption and emission follows dipole transition rule which restricts the amount of possible excitations or decays. In dipole transition final and initial states of the transition obey the following angular momentum rules [35].

Photon absorption and emission follows dipole transition rule which restricts the amount of possible excitations or decays. In dipole transition final and initial states of the transition obey the following angular momentum rules [35].

∆L = 0, ±1 0 = 0 ∆S = 0 ∆J = 0, ±1 0 = 0

∆L = 0, ±1 0 = 0 ∆S = 0 ∆J = 0, ±1 0 = 0

(2.8)

(2.8)

where L is the total orbital angular momentum, S is the total spin momentum, and J is the coupled total momentum of L and S. Orbital angular

where L is the total orbital angular momentum, S is the total spin momentum, and J is the coupled total momentum of L and S. Orbital angular

8

8

momentum l of the electron undergoing optical dipole transition changes and the initial and final states must have opposite parity ∆l = ±1. In Paper VI reported optical emission lines are subject to this transition rule. Example of this can be seen for example in Fe I transitions at 495.72983 nm and 495.75965 nm presented in the paper, where the final and initial electron configurations and spectral term symbols of transitions are respectively 3d6 4s4p, 7 F4 − 3d6 4s5s, 7 D4 and 3d6 4s4p, 7F6 − 3d6 4s5s, 7 D5 .

momentum l of the electron undergoing optical dipole transition changes and the initial and final states must have opposite parity ∆l = ±1. In Paper VI reported optical emission lines are subject to this transition rule. Example of this can be seen for example in Fe I transitions at 495.72983 nm and 495.75965 nm presented in the paper, where the final and initial electron configurations and spectral term symbols of transitions are respectively 3d6 4s4p, 7 F4 − 3d6 4s5s, 7 D4 and 3d6 4s4p, 7F6 − 3d6 4s5s, 7 D5 .

2.2.3

2.2.3

Transitions in molecules

Transitions in molecules

In molecules, the electronic energy level system has more degrees of freedom. This leads to increased number of possible transitions between different energy levels as depicted in Figure 2.3. Because molecules can vibrate and rotate, an additional splitting of electronic energy levels leads to rotational and vibrational fine structure. These additional features are energetically closer to each other than electronic states. Usually optical emission spectra of molecules show clear signs of molecular band structures. The bands of emission lines are caused by many possible transitions from upper levels to several tightly spaced rotational levels. If the excitation energy is high enough it can bring the molecule in the dissociative state which is depicted by the highest potential energy curve in Figure 2.3. In this state it is energetically preferable for the internuclear distance to increase, thus leading to fragmentation of the molecule. These states may be reached if photons with enough energy are absorbed in the molecule or in case of high temperature plasma, thermal energy is sufficiently high to bring the molecule into a dissociative state, as seen with the experimental spectra of electric arc in Paper VI.

In molecules, the electronic energy level system has more degrees of freedom. This leads to increased number of possible transitions between different energy levels as depicted in Figure 2.3. Because molecules can vibrate and rotate, an additional splitting of electronic energy levels leads to rotational and vibrational fine structure. These additional features are energetically closer to each other than electronic states. Usually optical emission spectra of molecules show clear signs of molecular band structures. The bands of emission lines are caused by many possible transitions from upper levels to several tightly spaced rotational levels. If the excitation energy is high enough it can bring the molecule in the dissociative state which is depicted by the highest potential energy curve in Figure 2.3. In this state it is energetically preferable for the internuclear distance to increase, thus leading to fragmentation of the molecule. These states may be reached if photons with enough energy are absorbed in the molecule or in case of high temperature plasma, thermal energy is sufficiently high to bring the molecule into a dissociative state, as seen with the experimental spectra of electric arc in Paper VI.

9

9

(b)

4

3

7 6

Energy

2

(a)

10

1 8

(b)

2

4

3

11 5

7 6

Energy

(a)

10

1

9

8

Inter-nuclear distance

11 5 9

Inter-nuclear distance

Figure 2.3: (a) Diatomic molecule, 1. nuclei, 2. electrons, 3. rotational movement, and 4. vibrational movement. (b) An example of molecular potential energy curves for the molecule, 5. potential energy curve of electronic ground state 6. potential energy curve of electronic excited state, 7. potential energy curve of dissociative state, 8. vibrational energy levels, 9. rotational energy levels, 10. electronic transitions, and 11. photon emission.

Figure 2.3: (a) Diatomic molecule, 1. nuclei, 2. electrons, 3. rotational movement, and 4. vibrational movement. (b) An example of molecular potential energy curves for the molecule, 5. potential energy curve of electronic ground state 6. potential energy curve of electronic excited state, 7. potential energy curve of dissociative state, 8. vibrational energy levels, 9. rotational energy levels, 10. electronic transitions, and 11. photon emission.

10

10

Chapter 3

Chapter 3

Vapor production and sample preparation

Vapor production and sample preparation

Most metals are solid in normal temperature and pressure (NTP) conditions with few exceptions. Electron spectroscopy of free atoms of these sample materials needs a method of producing atomic vapor. Usually atomic vapor is produced by heating solid sample material in a crucible so that its vapor pressure is sufficiently increased for experimentation. The vapor pressure of the sample material depends on the temperature [36]. Depending on the oven configuration and evaporation temperature, the required heating power can vary greatly. When temperature increases, needed heating power to maintain balance between heat loss and heating increases sharply. Since atomic vapor beams are produced in high vacuum environment, most of the heat loss is usually radiative. This however depends on the oven design. Thermal radiation power [37] P depends on the temperature T as follows:

Most metals are solid in normal temperature and pressure (NTP) conditions with few exceptions. Electron spectroscopy of free atoms of these sample materials needs a method of producing atomic vapor. Usually atomic vapor is produced by heating solid sample material in a crucible so that its vapor pressure is sufficiently increased for experimentation. The vapor pressure of the sample material depends on the temperature [36]. Depending on the oven configuration and evaporation temperature, the required heating power can vary greatly. When temperature increases, needed heating power to maintain balance between heat loss and heating increases sharply. Since atomic vapor beams are produced in high vacuum environment, most of the heat loss is usually radiative. This however depends on the oven design. Thermal radiation power [37] P depends on the temperature T as follows:

P = ǫσAT 4 ,

P = ǫσAT 4 ,

(3.1)

(3.1)

where ǫ is the emissivity of the heated material, σ is the Stefan-Boltzmann constant, and A is the surface area of the material. Radiated heat loss can be reduced by using heat shields around a crucible which insulate the heated part of the oven. In addition to heat shielding, water cooling is used in order to prevent excess heating of the outer layer of the oven. This reduces heating of the vacuum chamber walls and other equipment inside the chamber significantly [36]. Atomic vapor beams for experiments can be produced by heating solid sample with different technical means –see [36]. In high vacuum environment for samples which exhibit higher vapor pressure at lower temperatures

where ǫ is the emissivity of the heated material, σ is the Stefan-Boltzmann constant, and A is the surface area of the material. Radiated heat loss can be reduced by using heat shields around a crucible which insulate the heated part of the oven. In addition to heat shielding, water cooling is used in order to prevent excess heating of the outer layer of the oven. This reduces heating of the vacuum chamber walls and other equipment inside the chamber significantly [36]. Atomic vapor beams for experiments can be produced by heating solid sample with different technical means –see [36]. In high vacuum environment for samples which exhibit higher vapor pressure at lower temperatures

11

11

a resistive oven can be used [38]. When atomic vapor is produced in high vacuum where even small amounts of impurities and contaminants from different parts of the oven and the vacuum chamber may raise residual gas pressure and increase pump down time [39]. In order to minimize residual gas pressure during experiments heating of the crucible must be performed so that its support structure or chamber does not heat up and outgas impurities. Metals investigated in the papers of this thesis demanded a high evaporation temperature in order to produce sufficient vapor pressure for the experiments. Many metals, when at high temperature, form a reactive melt which needs to be separated from the crucible with the ceramic lining. This approach was utilized in Papers IV and V by inserting sample inside an Al2 O3 cup within the crucible.

a resistive oven can be used [38]. When atomic vapor is produced in high vacuum where even small amounts of impurities and contaminants from different parts of the oven and the vacuum chamber may raise residual gas pressure and increase pump down time [39]. In order to minimize residual gas pressure during experiments heating of the crucible must be performed so that its support structure or chamber does not heat up and outgas impurities. Metals investigated in the papers of this thesis demanded a high evaporation temperature in order to produce sufficient vapor pressure for the experiments. Many metals, when at high temperature, form a reactive melt which needs to be separated from the crucible with the ceramic lining. This approach was utilized in Papers IV and V by inserting sample inside an Al2 O3 cup within the crucible.

3.1

3.1

Resistive ovens

Resistive ovens

Resistive ovens can be used for vapor production when the needed temperature is not too high for the resistive heating element. For example, oven used for producing Na vapor by Rander et al. [40] consists of a heating wire which heats up a crucible using radiated and convected heat transfer. A schematic view of this type of oven is depicted in Figure 3.1.

Resistive ovens can be used for vapor production when the needed temperature is not too high for the resistive heating element. For example, oven used for producing Na vapor by Rander et al. [40] consists of a heating wire which heats up a crucible using radiated and convected heat transfer. A schematic view of this type of oven is depicted in Figure 3.1.

3.2

3.2

Induction heating

Induction heating

When high heating power is needed to achieve high temperatures, a vapor oven may require different means to heat the crucible. One of the alternative methods is induction heating. It has been found to be very effective method for heating the crucible directly [41] with many advances for producing atomic vapor beams. It is possible to achieve much higher heating power than in resistive ovens due to concentrated non-contact heating and water cooled working coil of the oven which can withstand high electrical currents. Induction heating was used in Papers I-V. Several models of induction heated setups have also been previously designed and used [41]. Induction heating is based on the use of AC current, which generates a time varying magnetic field. This field induces eddy currents to heated electrically conductive materials [42]. Usually when high frequency current is used, the heat is generated inside the top layer of the crucible due to skin effect. The skin effect is caused by a self interaction of the AC electric current via magnetic fields which restricts the current flow inside a conductor [43]. Usually the crucible is made of electrically conductive high melting point

When high heating power is needed to achieve high temperatures, a vapor oven may require different means to heat the crucible. One of the alternative methods is induction heating. It has been found to be very effective method for heating the crucible directly [41] with many advances for producing atomic vapor beams. It is possible to achieve much higher heating power than in resistive ovens due to concentrated non-contact heating and water cooled working coil of the oven which can withstand high electrical currents. Induction heating was used in Papers I-V. Several models of induction heated setups have also been previously designed and used [41]. Induction heating is based on the use of AC current, which generates a time varying magnetic field. This field induces eddy currents to heated electrically conductive materials [42]. Usually when high frequency current is used, the heat is generated inside the top layer of the crucible due to skin effect. The skin effect is caused by a self interaction of the AC electric current via magnetic fields which restricts the current flow inside a conductor [43]. Usually the crucible is made of electrically conductive high melting point

12

12

1

1

3

3

4

4

2

2 5

6

5

6

Figure 3.1: Schematic view of the resistive wire oven. 1 interaction volume of the atomic vapor beam, 2 sample, 3 resistive heating element, 4 heat shields, 5 water cooling, and 6 thermocouple.

Figure 3.1: Schematic view of the resistive wire oven. 1 interaction volume of the atomic vapor beam, 2 sample, 3 resistive heating element, 4 heat shields, 5 water cooling, and 6 thermocouple.

material. For example graphite, molybdenum, and tungsten are suitable crucible materials for high temperature induction heating [36]. In this work a commercial induction heating RF-power supply model TruHeat HF 3005 manufactured by Hüttinger Elektronik was used to provide the AC-current. The power supply was connected to a water cooled working coil wound around the crucible which concentrates the magnetic field to the crucible as depicted in Figure 3.2. Heat is produced only in the crucible and all the other parts of the oven are kept cool by water cooling. The oven is designed so that thermal conduction of the heat is minimised by using a hollow rod holding crucible inside the working coil. A disadvantage of induction heating is that electrons are easily affected by the magnetic and electric fields caused by the induction heating. This will lead to broadened peaks and increased background in electron spectra unless taken care during experiments. The power of an induction heater can

material. For example graphite, molybdenum, and tungsten are suitable crucible materials for high temperature induction heating [36]. In this work a commercial induction heating RF-power supply model TruHeat HF 3005 manufactured by Hüttinger Elektronik was used to provide the AC-current. The power supply was connected to a water cooled working coil wound around the crucible which concentrates the magnetic field to the crucible as depicted in Figure 3.2. Heat is produced only in the crucible and all the other parts of the oven are kept cool by water cooling. The oven is designed so that thermal conduction of the heat is minimised by using a hollow rod holding crucible inside the working coil. A disadvantage of induction heating is that electrons are easily affected by the magnetic and electric fields caused by the induction heating. This will lead to broadened peaks and increased background in electron spectra unless taken care during experiments. The power of an induction heater can

13

13

1

1

3

3 5 2

5 2

4

4

Figure 3.2: Schematic view of the induction oven. 1 target volume, 2 sample, 3 working coil, 4 thermocouple, and 5 water cooling of the casing.

Figure 3.2: Schematic view of the induction oven. 1 target volume, 2 sample, 3 working coil, 4 thermocouple, and 5 water cooling of the casing.

be pulsed on and off so that heating and electron spectra measurement alter rapidly. During the heating pulse the electron signal is vetoed and only the electrons free of the induction field will be collected. Signal vetoing can be realized either by vetoing the detected electron signal or the excitation source. In Paper II electron excitation was vetoed using synchronized electron beam deflection during heating. Another possibility is to use detector signal vetoing which was used when a position sensitive resistive anode detector was used in the experiments with synchrotron excitation. This detector type enabled fast vetoing of the electron signal when vetoing of the excitation source was not available. The pulsing technique is advantageous because heating power can be regulated so that the electron signal collection time is maximized in relation to heating time by using the full power of the power supply while heating is

be pulsed on and off so that heating and electron spectra measurement alter rapidly. During the heating pulse the electron signal is vetoed and only the electrons free of the induction field will be collected. Signal vetoing can be realized either by vetoing the detected electron signal or the excitation source. In Paper II electron excitation was vetoed using synchronized electron beam deflection during heating. Another possibility is to use detector signal vetoing which was used when a position sensitive resistive anode detector was used in the experiments with synchrotron excitation. This detector type enabled fast vetoing of the electron signal when vetoing of the excitation source was not available. The pulsing technique is advantageous because heating power can be regulated so that the electron signal collection time is maximized in relation to heating time by using the full power of the power supply while heating is

14

14

on. The effective average heating power is adjusted by varying heating time in relation to repetition rate. This power adjustment technique is commonly known as pulse width modulation (PWM). A typical heating pulse train during an experiment is shown in Figure 3.3 where the average heating power is ton (3.2) Pavg = Pmax . T Pmax is the power of heating pulses, ton is the length of the power pulses, and T is the period of the pulses. ton T

ton

Pinst Pavg measure

T

Pinst Pavg measure

Power

Pmax

Power

Pmax

on. The effective average heating power is adjusted by varying heating time in relation to repetition rate. This power adjustment technique is commonly known as pulse width modulation (PWM). A typical heating pulse train during an experiment is shown in Figure 3.3 where the average heating power is ton (3.2) Pavg = Pmax . T Pmax is the power of heating pulses, ton is the length of the power pulses, and T is the period of the pulses.

0

0 Time

Time

Figure 3.3: Heating power Pinst as a function of time. Electrons are collected during the off time of the heating indicated by gray rectangles. See text for details.

Figure 3.3: Heating power Pinst as a function of time. Electrons are collected during the off time of the heating indicated by gray rectangles. See text for details.

When PWM power regulation is used the average power Pavg reflects the effective constant heating power. Due to heat capacity of the crucible it acts as a thermal reservoir and averages heating pulses to near constant heating, thus keeping the vapor production stable. PWM heating power regulation was applied in Papers I-IV.

When PWM power regulation is used the average power Pavg reflects the effective constant heating power. Due to heat capacity of the crucible it acts as a thermal reservoir and averages heating pulses to near constant heating, thus keeping the vapor production stable. PWM heating power regulation was applied in Papers I-IV.

3.2.1

3.2.1

Simultaneous measurement of atomic and solid state spectra

Simultaneous measurement of atomic and solid state spectra

Binding energy values of atoms are typically reported in literature in reference to the vacuum level. However, for solids the values are typically given respect to the Fermi-level of the sample and therefore the work-function of the solid sample have to be taken account in order to carry out comparison between atomic and solid state values [44]. This introduces an additional uncertainty for comparisons. However it is possible to measure binding energy of the electrons simultaneously in a single experiment. Direct measurement

Binding energy values of atoms are typically reported in literature in reference to the vacuum level. However, for solids the values are typically given respect to the Fermi-level of the sample and therefore the work-function of the solid sample have to be taken account in order to carry out comparison between atomic and solid state values [44]. This introduces an additional uncertainty for comparisons. However it is possible to measure binding energy of the electrons simultaneously in a single experiment. Direct measurement

15

15

technique of vapor-metal shifts for both Auger and photoelectron spectra has been previously used with X-ray tube and electron gun excitation [44, 45]. The technique used in Papers III and IV involved use of a needle on which atomic vapor from the crucible was deposited on and a clean and continuously renewing surface was formed. Figure 3.4 depicts the experimental setup for this simultaneous measurement technique.

technique of vapor-metal shifts for both Auger and photoelectron spectra has been previously used with X-ray tube and electron gun excitation [44, 45]. The technique used in Papers III and IV involved use of a needle on which atomic vapor from the crucible was deposited on and a clean and continuously renewing surface was formed. Figure 3.4 depicts the experimental setup for this simultaneous measurement technique.

2 3 5

2 3

4

6

5 1

4

6 1

Figure 3.4: Schematic view of the experimental setup for the simultaneous acquisition of the atomic vapor and the solid state photoelectron spectra. 1 oven producing atomic vapor, 2 electron spectrometer entrance, 3 cooled needle, 4 SR-beam, 5 the needle with fresh continuously accumulating surface of the sample, and 6 SR-beam exciting the surface and vapor simultaneously in a small volume.

Figure 3.4: Schematic view of the experimental setup for the simultaneous acquisition of the atomic vapor and the solid state photoelectron spectra. 1 oven producing atomic vapor, 2 electron spectrometer entrance, 3 cooled needle, 4 SR-beam, 5 the needle with fresh continuously accumulating surface of the sample, and 6 SR-beam exciting the surface and vapor simultaneously in a small volume.

3.3

3.3

Electric arc heating

Electric arc heating

In the study of optical emission from the EAF in Paper VI the heat is produced by high temperature plasma. The plasma is heated by high electrical current passing through ionized gas, via particle collisions [46]. If the electric current is sufficiently high it can produce continuous electrical discharge exhibiting high temperature which keeps the discharge channel ionized and electrically conductive. During this work several experimental apparatuses were developed for optical emission studies of electric arcs and the technical description is given in the following sections.

In the study of optical emission from the EAF in Paper VI the heat is produced by high temperature plasma. The plasma is heated by high electrical current passing through ionized gas, via particle collisions [46]. If the electric current is sufficiently high it can produce continuous electrical discharge exhibiting high temperature which keeps the discharge channel ionized and electrically conductive. During this work several experimental apparatuses were developed for optical emission studies of electric arcs and the technical description is given in the following sections.

16

16

3.3.1

Device for arc emission characterization

The device depicted in Figure 3.5 is an earlier simpler development version of the arc furnace presented in Paper VI. This device forms an arc discharge between electrodes one of which is made of sample material and another one of graphite. The power was provided by an adjustable power supply of a welding machine. The chamber was filled with helium gas during the experiments. Once the arc was ignited between the electrodes the optical emission was collected using an optical spectrometer. During the experiments it was soon noticed that the melting of the samples resulted high consumption of the sample materials, because of no crucible to hold the melt. Later addition of this lead to subsequential improvement of the experimental setup and development of a laboratory scale miniature EAF presented in Paper VI.

3.3.1

Device for arc emission characterization

The device depicted in Figure 3.5 is an earlier simpler development version of the arc furnace presented in Paper VI. This device forms an arc discharge between electrodes one of which is made of sample material and another one of graphite. The power was provided by an adjustable power supply of a welding machine. The chamber was filled with helium gas during the experiments. Once the arc was ignited between the electrodes the optical emission was collected using an optical spectrometer. During the experiments it was soon noticed that the melting of the samples resulted high consumption of the sample materials, because of no crucible to hold the melt. Later addition of this lead to subsequential improvement of the experimental setup and development of a laboratory scale miniature EAF presented in Paper VI.

8

8

7

2 5

7

6

3

2

4

1

6

5

3

4

1

Figure 3.5: Schematic view of the apparatus for generating electrical discharge for the study of light emission from different sample blocks of metal. 1 electric arc, 2 sample piece, 3 graphite block, 4 water cooling, 5 arc current input feed, 6 gas outlet/pumping 7 gas inlet, and 8 optical fiber.

Figure 3.5: Schematic view of the apparatus for generating electrical discharge for the study of light emission from different sample blocks of metal. 1 electric arc, 2 sample piece, 3 graphite block, 4 water cooling, 5 arc current input feed, 6 gas outlet/pumping 7 gas inlet, and 8 optical fiber.

The described apparatus was utilized in recording of high resolution emission spectra of some elemental compounds of steel in He and air. In the

The described apparatus was utilized in recording of high resolution emission spectra of some elemental compounds of steel in He and air. In the

17

17

Intensity [arb. units]

experiments the arc was formed between a graphite and the sample metal pieces enclosed in the chamber. The arc discharge emitted highly characteristic emission spectra with numerous emission lines of respective elements [1] as can be seen in the part of the iron spectrum presented in Figure 3.6. The highest peak in the spectrum around 495.7 nm was indicated in Paper VI and it originates from an excited neutral iron atom.

Intensity [arb. units]

experiments the arc was formed between a graphite and the sample metal pieces enclosed in the chamber. The arc discharge emitted highly characteristic emission spectra with numerous emission lines of respective elements [1] as can be seen in the part of the iron spectrum presented in Figure 3.6. The highest peak in the spectrum around 495.7 nm was indicated in Paper VI and it originates from an excited neutral iron atom.

470

480

490 500 Wavelength [nm]

510

470

480

490 500 Wavelength [nm]

510

Figure 3.6: A part of Fe emission spectrum measured with the arc device apparatus for emission spectra characterization.

Figure 3.6: A part of Fe emission spectrum measured with the arc device apparatus for emission spectra characterization.

The difference between the spectra measured utilizing the arc emission device and the laboratory scale EAF utilized in Paper VI is that the background consisting of thermal radiation from solid surfaces was less prominent due to lack of hot crucible and the melted pool of metal. Another difference between the experiments was the instrumental resolution, which was selected to be the highest obtainable by the utilized spectrometer introduced in Chapter 4.3: ∼ 0.07nm with 1200 lines/mm grating. However, using higher resolution it was possible to obtain only a part of the visible spectrum simultaneously due to the limited physical width of the CCD detector and fixed number of spectral channels.

The difference between the spectra measured utilizing the arc emission device and the laboratory scale EAF utilized in Paper VI is that the background consisting of thermal radiation from solid surfaces was less prominent due to lack of hot crucible and the melted pool of metal. Another difference between the experiments was the instrumental resolution, which was selected to be the highest obtainable by the utilized spectrometer introduced in Chapter 4.3: ∼ 0.07nm with 1200 lines/mm grating. However, using higher resolution it was possible to obtain only a part of the visible spectrum simultaneously due to the limited physical width of the CCD detector and fixed number of spectral channels.

3.3.2

3.3.2

Laboratory scale miniature EAF for arc plasma characterization

Laboratory scale miniature EAF for arc plasma characterization

In Paper VI a laboratory scale DC-EAF was used for optical emission study of electrical arcs melting substances commonly present in the EAF in stainless steel production. The experimental setup was the next step closer to a real

In Paper VI a laboratory scale DC-EAF was used for optical emission study of electrical arcs melting substances commonly present in the EAF in stainless steel production. The experimental setup was the next step closer to a real

18

18

experimental EAF melting process simulation. This experimental furnace, depicted in Figure 3.7, was developed as a part of this work.

6

experimental EAF melting process simulation. This experimental furnace, depicted in Figure 3.7, was developed as a part of this work.

6 7

3

7

3

1

2

1

2

5

5

4

4

Figure 3.7: Schematic view of the laboratory scale DC electric arc furnace. 1 optical fiber, 2 sample, 3 water cooled graphite electrodes, 4 height manipulator, 5 water cooled heat shield, 6 current input feeds, and 7 position manipulator.

Figure 3.7: Schematic view of the laboratory scale DC electric arc furnace. 1 optical fiber, 2 sample, 3 water cooled graphite electrodes, 4 height manipulator, 5 water cooled heat shield, 6 current input feeds, and 7 position manipulator.

19

19

The furnace was constructed so that it enabled full melting of the samples and the direct view of the arc. The chamber was designed to be vacuum tight so it would be possible to replace the air in the chamber with argon shielding gas. The electrodes were made of graphite which is the material usually used in the electrodes of EAFs [18]. Electrode holder tubes were water cooled as well as heat shielding around the crucible. This enabled continuous use of heating and protected the casing from overheating.

The furnace was constructed so that it enabled full melting of the samples and the direct view of the arc. The chamber was designed to be vacuum tight so it would be possible to replace the air in the chamber with argon shielding gas. The electrodes were made of graphite which is the material usually used in the electrodes of EAFs [18]. Electrode holder tubes were water cooled as well as heat shielding around the crucible. This enabled continuous use of heating and protected the casing from overheating.

3.3.3

3.3.3

Experiment in Tornio Steel Works

Experiment in Tornio Steel Works

Large scale industrial arc furnaces present a challenging environment for optical emission studies due to harsh experimental conditions in these furnaces. During this work an experiment [2] was conducted at industrial scale EAF at Tornio Steel Works factory as a collaboration of the company and electron spectroscopy group at University of Oulu as a part of FIMECC-ELEMET program EffArc project. Optical emission from the EAF depicted in Figure 3.8 was collected using a light collection unit schematically depicted in Figure 3.9 which was developed during this thesis work by the author. The unit utilizes the principle of a pinhole camera [47], which has been used in the history by the first camera and artists as an aid for drawing paintings [48]. Even though the principle itself is old and simple, it has some surprising benefits: the pinhole itself acts as a perfect lens affected only by the diffraction of light and due to lack of physical lens material, it is highly resistant to intense thermal radiation flux appearing in the vicinity of electric arcs in EAFs. Another benefit is that it also limits the intensity of light depending on the physical radius of the pinhole, thus protecting the end of the optical fiber and prevents the possible saturation of the optical spectrometer due to high amount of light. The unit utilizes a highly effective and protective cooling which was designed to perform two main functions; cooling and protection from dust particles. The annular cell structure encloses the whole inner part with coolant and effectively cools the case of the unit. Depending on the thermal load, the outer shell cooling can be water or gas flow cooled. The gas, say air or inert gases such as nitrogen, can be used when water cooling is not needed or desired due to safety reasons in an industrial environment. The inner tube acts as a support for the fiber which is attached to a slide, hold in place by an adjustment rod. The temperature of the fiber is monitored via a thermocouple attached to the slide near the end of the fiber, making possible to get feedback about the cooling of the most heat sensitive part of the unit. Axial position of the end of the fiber, which collects the light, is adjustable enabling control of the light collection angle, as it is affected be the focal

Large scale industrial arc furnaces present a challenging environment for optical emission studies due to harsh experimental conditions in these furnaces. During this work an experiment [2] was conducted at industrial scale EAF at Tornio Steel Works factory as a collaboration of the company and electron spectroscopy group at University of Oulu as a part of FIMECC-ELEMET program EffArc project. Optical emission from the EAF depicted in Figure 3.8 was collected using a light collection unit schematically depicted in Figure 3.9 which was developed during this thesis work by the author. The unit utilizes the principle of a pinhole camera [47], which has been used in the history by the first camera and artists as an aid for drawing paintings [48]. Even though the principle itself is old and simple, it has some surprising benefits: the pinhole itself acts as a perfect lens affected only by the diffraction of light and due to lack of physical lens material, it is highly resistant to intense thermal radiation flux appearing in the vicinity of electric arcs in EAFs. Another benefit is that it also limits the intensity of light depending on the physical radius of the pinhole, thus protecting the end of the optical fiber and prevents the possible saturation of the optical spectrometer due to high amount of light. The unit utilizes a highly effective and protective cooling which was designed to perform two main functions; cooling and protection from dust particles. The annular cell structure encloses the whole inner part with coolant and effectively cools the case of the unit. Depending on the thermal load, the outer shell cooling can be water or gas flow cooled. The gas, say air or inert gases such as nitrogen, can be used when water cooling is not needed or desired due to safety reasons in an industrial environment. The inner tube acts as a support for the fiber which is attached to a slide, hold in place by an adjustment rod. The temperature of the fiber is monitored via a thermocouple attached to the slide near the end of the fiber, making possible to get feedback about the cooling of the most heat sensitive part of the unit. Axial position of the end of the fiber, which collects the light, is adjustable enabling control of the light collection angle, as it is affected be the focal

20

20

length of the system. Apart from the outer casing cooling, an additional cooling feed is attached to the inner tube which feeds the cooling gas inside. The gas performs secondary cooling inside the tube and keeps the internal cavity overpressurized preventing possible dust accumulation from outside. The gas exits through the focused nozzles and the pinhole in the end of the tube, which is directed towards the electric arc. The nozzles form an additional intensified flow zone around the pinhole protecting it from the possible splashes of melted materials. The unit is an essential part of the experiment and development of the method to be applied successfully in large scale industrial EAFs with demanding environment. The initial results [2] showed that it is possible to obtain realtime optical emission spectra from large scale EAF despite the possible experimental challenges in industrial environments. 5

4

length of the system. Apart from the outer casing cooling, an additional cooling feed is attached to the inner tube which feeds the cooling gas inside. The gas performs secondary cooling inside the tube and keeps the internal cavity overpressurized preventing possible dust accumulation from outside. The gas exits through the focused nozzles and the pinhole in the end of the tube, which is directed towards the electric arc. The nozzles form an additional intensified flow zone around the pinhole protecting it from the possible splashes of melted materials. The unit is an essential part of the experiment and development of the method to be applied successfully in large scale industrial EAFs with demanding environment. The initial results [2] showed that it is possible to obtain realtime optical emission spectra from large scale EAF despite the possible experimental challenges in industrial environments. 5

4

2 3

2 3

1

1

Figure 3.8: Schematic view of the light collection setup at Tornio steel works ladle furnace. 1 molten steel, 2 slag layer, 3 electric arc, 4 graphite electrodes, and 5 light collection unit.

Figure 3.8: Schematic view of the light collection setup at Tornio steel works ladle furnace. 1 molten steel, 2 slag layer, 3 electric arc, 4 graphite electrodes, and 5 light collection unit.

21

21

8

7

5

4

8 3

1

2

6

7

5

4

3

1

2

6

Figure 3.9: Schematic view of designed light collection unit. 1 light input, 2 combined pinhole slit and nozzle, 3 additional nozzles, 4 optical fiber input, 5 thermocouple, 6 collection angle adjustment, 7 compressed air/inert gas cooling, and 8 air or water cooling for annular double shell casing.

Figure 3.9: Schematic view of designed light collection unit. 1 light input, 2 combined pinhole slit and nozzle, 3 additional nozzles, 4 optical fiber input, 5 thermocouple, 6 collection angle adjustment, 7 compressed air/inert gas cooling, and 8 air or water cooling for annular double shell casing.

22

22

Chapter 4

Chapter 4

Electron, ion, and photon detection

Electron, ion, and photon detection

Experimental physics and its applications rely on good experimental techniques and devices. In this chapter the main principles of the measuring devices used in this work are introduced. The electron spectroscopy group at University of Oulu has a long history of building and developing instrumentation for experimental setups of electron spectroscopy [49–52].

Experimental physics and its applications rely on good experimental techniques and devices. In this chapter the main principles of the measuring devices used in this work are introduced. The electron spectroscopy group at University of Oulu has a long history of building and developing instrumentation for experimental setups of electron spectroscopy [49–52].

4.1

4.1

Electron spectrometer

Electron spectrometer

The energy separation of the electrons according to their kinetic energies can be done utilizing the Lorentz force:

The energy separation of the electrons according to their kinetic energies can be done utilizing the Lorentz force:

¯ F¯ = q(E¯ + v¯ × B),

¯ F¯ = q(E¯ + v¯ × B),

(4.1)

(4.1)

¯ is the electric field, v¯ is the velocity where q is the charge of the particle, E ¯ of the particle, and B is magnetic field. In electrostatic spectrometers the trajectories of the electrons are determined by the electric fields and kinetic energies of the electrons. This information is used to determine energy spectra of the electrons. There are several different types of electron spectrometers which make use of electric fields [53–57]. In this work commercial Scienta SES-100 and SES-200 hemispherical electron energy analyzers [58] were used. The hemispherical electron spectrometer depicted in Figure 4.1 is custom modified version of the SES-100 spectrometer [59]. The modified SES-100 electron spectrometer (Figure 4.1) works as follows: Free electrons are collected at small opening angle of the electron lens which focuses the electrons to the entrance slit of the spectrometer. The lens

¯ is the electric field, v¯ is the velocity where q is the charge of the particle, E ¯ of the particle, and B is magnetic field. In electrostatic spectrometers the trajectories of the electrons are determined by the electric fields and kinetic energies of the electrons. This information is used to determine energy spectra of the electrons. There are several different types of electron spectrometers which make use of electric fields [53–57]. In this work commercial Scienta SES-100 and SES-200 hemispherical electron energy analyzers [58] were used. The hemispherical electron spectrometer depicted in Figure 4.1 is custom modified version of the SES-100 spectrometer [59]. The modified SES-100 electron spectrometer (Figure 4.1) works as follows: Free electrons are collected at small opening angle of the electron lens which focuses the electrons to the entrance slit of the spectrometer. The lens

23

23

5

5 4

4

2

2

1

1 3

3

Figure 4.1: Schematic view of SES-100 electron spectrometer. 1 sample volume, 2 electron lens, 3 entrance slit, 4 hemisphere, and 5 MCP and RAD detector.

Figure 4.1: Schematic view of SES-100 electron spectrometer. 1 sample volume, 2 electron lens, 3 entrance slit, 4 hemisphere, and 5 MCP and RAD detector.

has two functions; it allows the hemisphere to be located farther from the interaction region and still maintain the collection efficiency of the spectrometer, and it can accelerate or decelerate the electrons. In addition, the lens has electrostatic elements which allows small adjustments of focus position. The hemispherical spectrometer is operated at constant pass mode where the electrons with selected kinetic energy called pass energy Ep can reach the center of the detector and the electrons which have more or less energy will take different paths and end up around the center of the image plane. Constant pass operating mode allows the energy resolution ∆E of the spectrometer to be constant at selected pass energy and it can be calculated approximately using the following equation [60]

has two functions; it allows the hemisphere to be located farther from the interaction region and still maintain the collection efficiency of the spectrometer, and it can accelerate or decelerate the electrons. In addition, the lens has electrostatic elements which allows small adjustments of focus position. The hemispherical spectrometer is operated at constant pass mode where the electrons with selected kinetic energy called pass energy Ep can reach the center of the detector and the electrons which have more or less energy will take different paths and end up around the center of the image plane. Constant pass operating mode allows the energy resolution ∆E of the spectrometer to be constant at selected pass energy and it can be calculated approximately using the following equation [60]

∆E ≈

w Ep , 2R

(4.2)

∆E ≈

w Ep , 2R

(4.2)

where w is the width of the entrance slit and R is the radius of the center hemispherical paths of the electrons with energy Ep . SES-100 electron spectrometer [59] equip with commercial Quantar model 3394A detector package uses multichannel plate (MCP) which amplifies signal of single electrons hits to measurable levels for the resistive anode plate (RAD) which signal is analyzed with Quantar model 2401B unit. The detector allows high response time and position analyzer can be vetoed using an external signal. The veto was used in the experiments in Papers I, III, and IV where SES-100 spectrometer was used with induction heating.

where w is the width of the entrance slit and R is the radius of the center hemispherical paths of the electrons with energy Ep . SES-100 electron spectrometer [59] equip with commercial Quantar model 3394A detector package uses multichannel plate (MCP) which amplifies signal of single electrons hits to measurable levels for the resistive anode plate (RAD) which signal is analyzed with Quantar model 2401B unit. The detector allows high response time and position analyzer can be vetoed using an external signal. The veto was used in the experiments in Papers I, III, and IV where SES-100 spectrometer was used with induction heating.

24

24

4.2

Time-of-flight spectrometer

4.2

In this work photoionization cross-section of atomic chromium vapor was investigated in Paper V using time-of-flight spectroscopy of ions [61]. Timeof-flight mass spectrometer (TOFMS) is based on difference in time-of-flight of ions with different masses from a source to a detector. Wiley-McLaren [62] TOFMS [63] depicted in Figure 4.2 uses a pulsed electrical field or pulsed excitation source over the ion source region to extract ions from the interaction volume at certain interval. The spectrometer has a separate acceleration region where extracted ions are accelerated to final speeds. The separate acceleration region improves the resolution and adjustability of the instrument. The acquired speed of the ions entering the drift tube depends on mass m to charge q ratio m/z of the ions. Final time separation of ions with separate m/z ratio depend on the length of the drift tube and used acceleration voltages. Once the ions arrive to the detector, arrival times relative to the extraction pulse are recorded. Depending on experimental setup the extraction is done repetitively during collection of a mass spectrum allowing all the ions arrive to the detector before the next extraction. A typical repetition rate can be in the order of 15 kHz, depending on the instrumental parameters. 2

3

4

Time-of-flight spectrometer

In this work photoionization cross-section of atomic chromium vapor was investigated in Paper V using time-of-flight spectroscopy of ions [61]. Timeof-flight mass spectrometer (TOFMS) is based on difference in time-of-flight of ions with different masses from a source to a detector. Wiley-McLaren [62] TOFMS [63] depicted in Figure 4.2 uses a pulsed electrical field or pulsed excitation source over the ion source region to extract ions from the interaction volume at certain interval. The spectrometer has a separate acceleration region where extracted ions are accelerated to final speeds. The separate acceleration region improves the resolution and adjustability of the instrument. The acquired speed of the ions entering the drift tube depends on mass m to charge q ratio m/z of the ions. Final time separation of ions with separate m/z ratio depend on the length of the drift tube and used acceleration voltages. Once the ions arrive to the detector, arrival times relative to the extraction pulse are recorded. Depending on experimental setup the extraction is done repetitively during collection of a mass spectrum allowing all the ions arrive to the detector before the next extraction. A typical repetition rate can be in the order of 15 kHz, depending on the instrumental parameters. 2

3

4

5 1

5 1

Figure 4.2: Schematic view of time-of-flight spectrometer. 1 interaction volume, 2 extraction region, 3 acceleration region, 4 drift region, and 5 MCP detector.

Figure 4.2: Schematic view of time-of-flight spectrometer. 1 interaction volume, 2 extraction region, 3 acceleration region, 4 drift region, and 5 MCP detector.

Wiley-McLaren type TOFMS offers space focusing where ions with the same mass and charge state coming from different distance from the detector will arrive almost simultaneously. This is due to extraction field which will, at correct operating voltages, give slightly more energy to the ions travelling longer distance in the direction of the field. This leads to simultaneous arrival times for corresponding ions within resolution limits. The resolution of the time-of-flight instrument is affected by initial position and kinetic energy distributions of the detected ions [62].

Wiley-McLaren type TOFMS offers space focusing where ions with the same mass and charge state coming from different distance from the detector will arrive almost simultaneously. This is due to extraction field which will, at correct operating voltages, give slightly more energy to the ions travelling longer distance in the direction of the field. This leads to simultaneous arrival times for corresponding ions within resolution limits. The resolution of the time-of-flight instrument is affected by initial position and kinetic energy distributions of the detected ions [62].

25

25

4.3

Optical spectrometer

4.3

Optical spectrometer separates the light to its individual spectral components. The dispersive element, separating the wavelengths, can be a prism or grating. Usually, gratings are used in modern spectrometers due to advantages which include adjustable dispersion via line density, possibility to use larger optical surface without increasing amount of material extensively in comparison to prisms, and possibility of cover large wavelength range due to reflection. The commercial Andor SR-500i spectrometer used in Paper VI is based on the Czerny-Turner optical design [64] –see Figure 4.3. The design allows the spectrometer to be relatively compact with good optical properties due optics based on reflection of light which prevents the dispersion of light in the optics other than the grating. The spectrometer has three different gratings with line densities of 150, 600, and 1200 lines/mm enabling either coverage of wide spectral range or high resolution at small range simultaneously.

Optical spectrometer

Optical spectrometer separates the light to its individual spectral components. The dispersive element, separating the wavelengths, can be a prism or grating. Usually, gratings are used in modern spectrometers due to advantages which include adjustable dispersion via line density, possibility to use larger optical surface without increasing amount of material extensively in comparison to prisms, and possibility of cover large wavelength range due to reflection. The commercial Andor SR-500i spectrometer used in Paper VI is based on the Czerny-Turner optical design [64] –see Figure 4.3. The design allows the spectrometer to be relatively compact with good optical properties due optics based on reflection of light which prevents the dispersion of light in the optics other than the grating. The spectrometer has three different gratings with line densities of 150, 600, and 1200 lines/mm enabling either coverage of wide spectral range or high resolution at small range simultaneously.

7

7

6

6 5

5

4

2

4

3

2

1

3 1

Figure 4.3: Schematic view of Andor Shamrock 500i spectrometer. 1 optical fiber, 2 input slit, 3 filter wheel, 4 flip mirror for input port selection, 5 focusing mirrors, 6 triple grating turret, and 7 Peltier element cooled CCD detector.

Figure 4.3: Schematic view of Andor Shamrock 500i spectrometer. 1 optical fiber, 2 input slit, 3 filter wheel, 4 flip mirror for input port selection, 5 focusing mirrors, 6 triple grating turret, and 7 Peltier element cooled CCD detector.

4.3.1

4.3.1

Resolving power and transmission

Resolving power and transmission

The overall optical resolution of the optical spectrometer depends on input slit, grating, wavelength of the light, and possible aberrations of the optics.

The overall optical resolution of the optical spectrometer depends on input slit, grating, wavelength of the light, and possible aberrations of the optics.

26

26

Resolving power R of a grating is defined as R=

λ , ∆λ

Resolving power R of a grating is defined as (4.3)

where the average wavelength of diffracted light is λ and ∆λ is the limit of resolution. ∆λ is defined as a difference of two lines of equal intensity that can be distinguished [65]. The resolving power R for a plane grating can be calculated from equation [65] R=

Nd(sin α + sin β) , λ

(4.4)

where N is the total number of illuminated grooves on the surface of the grating, d is the groove spacing, α is the angle of incoming light, and β is the angle of diffracted light at the center wavelength λ. The optical dispersion of the grating is characterized by linear dispersion relation cos β dλ = , (4.5) dx knL where dλ/dx is derivative of wavelength λ with respect to position x at the image plane, β is angle of diffraction of light at the center wave length λ, k is diffraction order, n is groove density, and L is effective focal length [65]. Gratings of the Andor SR-500i spectrometer are engraved for improved reflection of the first order diffraction. However, small portion of light also falls to higher orders. Therefore the spectrometer is equipped with a filter wheel for cutting off intensity of unwanted wavelengths of light which would otherwise fall to second or higher orders. In Paper VI where the spectrometer was used higher orders were reduced due to absorption of light below 300 nm in the window of the EAF acting as a filter. Transmission is the property of an optical system which describes the relative passage of the amount of light from the source to the detector. Transmission is dependent on all the optical components which alter the intensity of the passing light. If an optical fiber is used with an optical spectrometer, as in all the optical experiments presented in this thesis, the transmission curve of the fiber must be included to the transmission correction of the spectra. Transmission curve Tr (λ) which corrects also for the detection efficiency can be calculated as a function of individual products of separate transmission curves of the optical components. Tr (λ) = Tf b (λ) · Tspectr (λ) · Tdet (λ),

(4.6)

R=

λ , ∆λ

(4.3)

where the average wavelength of diffracted light is λ and ∆λ is the limit of resolution. ∆λ is defined as a difference of two lines of equal intensity that can be distinguished [65]. The resolving power R for a plane grating can be calculated from equation [65] R=

Nd(sin α + sin β) , λ

(4.4)

where N is the total number of illuminated grooves on the surface of the grating, d is the groove spacing, α is the angle of incoming light, and β is the angle of diffracted light at the center wavelength λ. The optical dispersion of the grating is characterized by linear dispersion relation cos β dλ = , (4.5) dx knL where dλ/dx is derivative of wavelength λ with respect to position x at the image plane, β is angle of diffraction of light at the center wave length λ, k is diffraction order, n is groove density, and L is effective focal length [65]. Gratings of the Andor SR-500i spectrometer are engraved for improved reflection of the first order diffraction. However, small portion of light also falls to higher orders. Therefore the spectrometer is equipped with a filter wheel for cutting off intensity of unwanted wavelengths of light which would otherwise fall to second or higher orders. In Paper VI where the spectrometer was used higher orders were reduced due to absorption of light below 300 nm in the window of the EAF acting as a filter. Transmission is the property of an optical system which describes the relative passage of the amount of light from the source to the detector. Transmission is dependent on all the optical components which alter the intensity of the passing light. If an optical fiber is used with an optical spectrometer, as in all the optical experiments presented in this thesis, the transmission curve of the fiber must be included to the transmission correction of the spectra. Transmission curve Tr (λ) which corrects also for the detection efficiency can be calculated as a function of individual products of separate transmission curves of the optical components. Tr (λ) = Tf b (λ) · Tspectr (λ) · Tdet (λ),

(4.6)

where Tf b , Tspectr , and Tdet are the individual transmissions of the fiber, spectrometer, and the detector efficiency.

where Tf b , Tspectr , and Tdet are the individual transmissions of the fiber, spectrometer, and the detector efficiency.

27

27

4.3.2

Detector

4.3.2

Detector

In Andor SR-500i spectrometer photons are detected with a thermoelectrically cooled Andor Newton DU940N-BU2 CCD camera. The back illuminated CCD detector is enhanced for detection of UV light and it can collect photons with wavelength down to 200 nm. The detector is capable of wide range of exposures times and vertical columns can be bind electronically in order to improve signal reading time and reduce electrical read noise. Especially when sequential exposures are taken as in recording of Paper VI spectra binging speed up spectra acquisition. Usually, if a weak signal is detected there is considerable amount of thermal and shot noise [66, 67]. Also cosmic rays can introduce random high intensity false peaks to the spectra unless taken care by comparison of exposures in the control software or data handling. CCD detector usually exhibits random, but permanent variations of background count rates due manufacturing and temperature differences in addition to noise [66]. This uneven additional contribution of the signal can be eliminated via subtraction of a black image, taken with the shutter closed, from the signal. This build-in feature of Andor spectrometer control software was used in Paper VI. The black spectrum is recorded before actual light collection before each acquisition of signal and will be subtracted during the collection.

In Andor SR-500i spectrometer photons are detected with a thermoelectrically cooled Andor Newton DU940N-BU2 CCD camera. The back illuminated CCD detector is enhanced for detection of UV light and it can collect photons with wavelength down to 200 nm. The detector is capable of wide range of exposures times and vertical columns can be bind electronically in order to improve signal reading time and reduce electrical read noise. Especially when sequential exposures are taken as in recording of Paper VI spectra binging speed up spectra acquisition. Usually, if a weak signal is detected there is considerable amount of thermal and shot noise [66, 67]. Also cosmic rays can introduce random high intensity false peaks to the spectra unless taken care by comparison of exposures in the control software or data handling. CCD detector usually exhibits random, but permanent variations of background count rates due manufacturing and temperature differences in addition to noise [66]. This uneven additional contribution of the signal can be eliminated via subtraction of a black image, taken with the shutter closed, from the signal. This build-in feature of Andor spectrometer control software was used in Paper VI. The black spectrum is recorded before actual light collection before each acquisition of signal and will be subtracted during the collection.

28

28

Chapter 5

Chapter 5

Excitation methods

Excitation methods

5.1

5.1

Electron beam

Electron beam

Electron beam excitation differs from the photon excitation because an electron can lose a varying portion of its kinetic energy to the target in a collision. Due to this property the electron beam excitation is well suited for studies of Auger processes. The advantage for using an electron beam source is that most of the devices are compact and provide a good adjustable intensity and energy in a well focused beam. The electron beam with adjustable kinetic energy can be produced by an electron gun depicted in Figure 5.1. Conventional electron guns make use of an electrically heated filament, usually made out materials withstanding high temperature such as tungsten. High temperature is needed to thermally increase the energy of the electrons in the filament material so that the electrons are thermally emitted. Free electrons are then accelerated with an adjustable electric field and focused to a target with a cylindrically symmetric electron lens. The effective source area for the electrons is dependent of the Wehnelt potential relative to the filament. The function of a Wehnelt cap, surrounding the filament, is to allow only the electrons emitted from the tip of the filament to be accelerated toward the output of the gun. A small source area of the electrons allows the beam to be well focusable which is an important feature in many applications of electron beam [68]. In addition to focusing optics, the guns usually have deflection elements for adjusting of the exit angle of the beam. This feature was used in Paper II to provide electron beam veto during the induction heating pulses by deflecting the beam away from the electron collection volume of the electron spectrometer. In the experiment the beam was provided by commercial Specs EQ 22/35 electron gun, and its deflection was synchronized with the heating

Electron beam excitation differs from the photon excitation because an electron can lose a varying portion of its kinetic energy to the target in a collision. Due to this property the electron beam excitation is well suited for studies of Auger processes. The advantage for using an electron beam source is that most of the devices are compact and provide a good adjustable intensity and energy in a well focused beam. The electron beam with adjustable kinetic energy can be produced by an electron gun depicted in Figure 5.1. Conventional electron guns make use of an electrically heated filament, usually made out materials withstanding high temperature such as tungsten. High temperature is needed to thermally increase the energy of the electrons in the filament material so that the electrons are thermally emitted. Free electrons are then accelerated with an adjustable electric field and focused to a target with a cylindrically symmetric electron lens. The effective source area for the electrons is dependent of the Wehnelt potential relative to the filament. The function of a Wehnelt cap, surrounding the filament, is to allow only the electrons emitted from the tip of the filament to be accelerated toward the output of the gun. A small source area of the electrons allows the beam to be well focusable which is an important feature in many applications of electron beam [68]. In addition to focusing optics, the guns usually have deflection elements for adjusting of the exit angle of the beam. This feature was used in Paper II to provide electron beam veto during the induction heating pulses by deflecting the beam away from the electron collection volume of the electron spectrometer. In the experiment the beam was provided by commercial Specs EQ 22/35 electron gun, and its deflection was synchronized with the heating

29

29

2

3

2

3

5

5

1

1 4

4

Figure 5.1: Schematic view of the electron gun. 1 filament, 2 Wehnelt cap, 3 focusing electron lens, 4 x and y deflection, and 5 target.

Figure 5.1: Schematic view of the electron gun. 1 filament, 2 Wehnelt cap, 3 focusing electron lens, 4 x and y deflection, and 5 target.

using a custom computer software wrote by the author. The software used a National Instruments NIDAQ data acquisition board to interface with the electron gun control unit.

using a custom computer software wrote by the author. The software used a National Instruments NIDAQ data acquisition board to interface with the electron gun control unit.

5.2

5.2

Synchrotron radiation

Synchrotron radiation

Advancements in particle acceleration and storage techniques have lead to development of SR sources. SR sources provide radiation for carrying out experiments with many beneficial properties compared to conventional photon radiation sources for example UV-lamps or X-ray tubes. In this thesis two SR sources were used which are located in the MAX-lab synchrotron radiation facility in Lund, Sweden. SR is emitted from electrically charged particles, usually electrons, moving at relativistic speeds and experiencing acceleration due to magnetic fields affecting the path of the particles [69]. A storage ring made for producing SR utilize circular particle accelerator which includes insertion devices in straight sections of the ring as depicted in Figure 5.2. Insertion devices, wigglers or undulators, make use of periodic arrays of magnets to generate the radiation. The electrons circulate the ring in bunches and are steered by bending magnets. Bending magnets generate a broad continuous SR spectrum. Wigglers consist of a few magnetic periods and the radiation characteristics are similar to that of bending magnets with intensified intensity. The radiation causes the electron to lose energy and therefore storage rings include accelerators to maintain the kinetic energy of the electrons. The intensity of SR from insertion devices is several orders of magnitude higher than with conventional sources such as X-ray tubes or ultraviolet discharge lamps. Relativistic Lorentz contraction and Doppler effect affects the dipole radiation from the electrons seen in a laboratory reference frame. The effects

Advancements in particle acceleration and storage techniques have lead to development of SR sources. SR sources provide radiation for carrying out experiments with many beneficial properties compared to conventional photon radiation sources for example UV-lamps or X-ray tubes. In this thesis two SR sources were used which are located in the MAX-lab synchrotron radiation facility in Lund, Sweden. SR is emitted from electrically charged particles, usually electrons, moving at relativistic speeds and experiencing acceleration due to magnetic fields affecting the path of the particles [69]. A storage ring made for producing SR utilize circular particle accelerator which includes insertion devices in straight sections of the ring as depicted in Figure 5.2. Insertion devices, wigglers or undulators, make use of periodic arrays of magnets to generate the radiation. The electrons circulate the ring in bunches and are steered by bending magnets. Bending magnets generate a broad continuous SR spectrum. Wigglers consist of a few magnetic periods and the radiation characteristics are similar to that of bending magnets with intensified intensity. The radiation causes the electron to lose energy and therefore storage rings include accelerators to maintain the kinetic energy of the electrons. The intensity of SR from insertion devices is several orders of magnitude higher than with conventional sources such as X-ray tubes or ultraviolet discharge lamps. Relativistic Lorentz contraction and Doppler effect affects the dipole radiation from the electrons seen in a laboratory reference frame. The effects

30

30

1

1

2

7

2

7

3

6

3

6

4

4

5

5

Figure 5.2: Schematic simplified view of a synchrotron storage ring. 1 undulator, 2 SR output, 3 bending magnet, 4 wiggler, 5 empty section (vacuum tube), 6 rotation direction of the electrons, and 7 acceleration cavity.

Figure 5.2: Schematic simplified view of a synchrotron storage ring. 1 undulator, 2 SR output, 3 bending magnet, 4 wiggler, 5 empty section (vacuum tube), 6 rotation direction of the electrons, and 7 acceleration cavity.

cause the radiation to be emitted in a very narrow forefront angle as depicted in Figure 5.3 and also leads to highly shortened wavelength of the radiation compared to the period of the magnetic field [69].

cause the radiation to be emitted in a very narrow forefront angle as depicted in Figure 5.3 and also leads to highly shortened wavelength of the radiation compared to the period of the magnetic field [69].

1

1

2

3

2

3

Figure 5.3: Schematic view of an undulator magnet array. 1 Halbach array of permanent magnets, 2 relativistic electron beam, and 3 light emitted to a small forefront opening angle.

Figure 5.3: Schematic view of an undulator magnet array. 1 Halbach array of permanent magnets, 2 relativistic electron beam, and 3 light emitted to a small forefront opening angle.

31

31

The radiation spectrum of an undulator is composed of narrow high intensity peaks at specific wavelengths due to specific harmonic oscillations of the electrons in both transverse and the length axis of the undulator [69]. In the undulator radiation interferences with the motion of the electrons and causes narrowing of the peak widths of the emitted radiation. The position of these peaks can be tuned by adjusting the magnetic field strength by changing the gap between the poles of the undulator which makes SR tunable in broad range. Also polarization properties of the radiation can be tuned depending on the magnet array design; usually the radiation is highly linearly polarized in a plane of the storage ring [70]. SR is usually further monochromatized in the beamline optics with a monochromator in order to select one of the emitted wavelengths and further narrow the photon bandwidth of the beam for the experiments.

The radiation spectrum of an undulator is composed of narrow high intensity peaks at specific wavelengths due to specific harmonic oscillations of the electrons in both transverse and the length axis of the undulator [69]. In the undulator radiation interferences with the motion of the electrons and causes narrowing of the peak widths of the emitted radiation. The position of these peaks can be tuned by adjusting the magnetic field strength by changing the gap between the poles of the undulator which makes SR tunable in broad range. Also polarization properties of the radiation can be tuned depending on the magnet array design; usually the radiation is highly linearly polarized in a plane of the storage ring [70]. SR is usually further monochromatized in the beamline optics with a monochromator in order to select one of the emitted wavelengths and further narrow the photon bandwidth of the beam for the experiments.

5.2.1

5.2.1

Beamline I411

Beamline I411

Beamline I411 is part of the 1.5 GeV MAX-II third generation synchrotron storage ring in MAX-lab, Lund in Sweden where the experimental work for Papers I, III, and IV was done. The beamline has 2.65 m long undulator with 88 poles and modified SX700 monochromator [71]. The usable photon energy range of the beamline is 40-1500 eV. Scienta SES-100 electron spectrometer used in the included papers was mounted in the one meter long part of the beamline which provides a place for installing user experiments. The beamline is equipped with differential vacuum pumping employing turbomolecular and ion pumps for gas phase experiments.

Beamline I411 is part of the 1.5 GeV MAX-II third generation synchrotron storage ring in MAX-lab, Lund in Sweden where the experimental work for Papers I, III, and IV was done. The beamline has 2.65 m long undulator with 88 poles and modified SX700 monochromator [71]. The usable photon energy range of the beamline is 40-1500 eV. Scienta SES-100 electron spectrometer used in the included papers was mounted in the one meter long part of the beamline which provides a place for installing user experiments. The beamline is equipped with differential vacuum pumping employing turbomolecular and ion pumps for gas phase experiments.

5.2.2

5.2.2

Beamline I3 - FINEST branch

Beamline I3 - FINEST branch

Beamline I3 [72] is part of the 700 MeV MAX-III synchrotron storage ring [73] in MAX-lab. The beamline uses APPLE-II type elliptically polarizing undulator which can provide variable polarization of the photon beam. The beamline has off-axis eagle type normal incidence monochromator with three gratings which cover the operating photon energy range of 5-50 eV. The beamline has two branches one for solid state photoelectron spectroscopy and another branch includes differential pumping for gas phase experiments. The experimental work in Paper V was carried out at FINEST gas phase branch [74] of I3 beamline.

Beamline I3 [72] is part of the 700 MeV MAX-III synchrotron storage ring [73] in MAX-lab. The beamline uses APPLE-II type elliptically polarizing undulator which can provide variable polarization of the photon beam. The beamline has off-axis eagle type normal incidence monochromator with three gratings which cover the operating photon energy range of 5-50 eV. The beamline has two branches one for solid state photoelectron spectroscopy and another branch includes differential pumping for gas phase experiments. The experimental work in Paper V was carried out at FINEST gas phase branch [74] of I3 beamline.

32

32

5.3

Electric arc

5.3

Excitation of atomic vapor or molecules can be realized with an electric arc discharge. In steel manufacturing EAFs are used to melt raw materials mainly recycled scrap metal. As in many analytical devices such as induction heated plasma torches [14] and laser induced plasma breakdown spectroscopy [15], in EAFs electric arc provides energy for excitation. The difference between analytical devices and EAFs is that analytical plasma devices are designed as excitation sources, where as EAFs are designed to melt materials efficiently. However, in principle the core physics are quite similar and the EAF can be seen as a large plasma excitation source from the spectroscopic point of view.

Electric arc

Excitation of atomic vapor or molecules can be realized with an electric arc discharge. In steel manufacturing EAFs are used to melt raw materials mainly recycled scrap metal. As in many analytical devices such as induction heated plasma torches [14] and laser induced plasma breakdown spectroscopy [15], in EAFs electric arc provides energy for excitation. The difference between analytical devices and EAFs is that analytical plasma devices are designed as excitation sources, where as EAFs are designed to melt materials efficiently. However, in principle the core physics are quite similar and the EAF can be seen as a large plasma excitation source from the spectroscopic point of view.

8

8

7

6

7

6

5 10

9

5 3

2

10

9

3

2

1

1

4

4

Figure 5.4: Schematic view of an industrial scale EAF. 1 molten steel, 2 slag layer, 3 refractory lining, 4 tilting mechanism, 5 water cooled side wall, 6 exhaust gas line, 7 water cooled roof, 8 graphite electrodes, 9 electric arcs, and 10 tapping lip.

Figure 5.4: Schematic view of an industrial scale EAF. 1 molten steel, 2 slag layer, 3 refractory lining, 4 tilting mechanism, 5 water cooled side wall, 6 exhaust gas line, 7 water cooled roof, 8 graphite electrodes, 9 electric arcs, and 10 tapping lip.

Structure of an industrial scale EAF depicted in Figure 5.4 for scrap metal melting consists of the following main parts: walls, heart, roof, and electrodes. Walls form structural support of the furnace and hold refractory lining made out of bricks with various compositions of oxides and graphite. The refractory lined bottom of the furnace is called heart where the melt will

Structure of an industrial scale EAF depicted in Figure 5.4 for scrap metal melting consists of the following main parts: walls, heart, roof, and electrodes. Walls form structural support of the furnace and hold refractory lining made out of bricks with various compositions of oxides and graphite. The refractory lined bottom of the furnace is called heart where the melt will

33

33

be formed during the operation. Top of the furnace is covered with a roof which has a hole in the center for graphite electrodes. The roof is retractable and is opened for charging of the raw materials. As some of the heat produced during the operation of the furnace is absorbed by the wall and roof, there is water cooling to keep the case of the furnace in safe operation temperature. Usually EAFs are installed above the floor level so that melted steel can be poured to ladles for further processing. Large high power AC furnaces make use of three graphite electrodes which deliver high current order of several tens of kA to electric arcs. Typical transformer power of a large EAF can be order of 100 MW and melting time for one charge around 40 min [17, 18]. One important part of steel making and operation of the EAF is the formation of slag which covers the surface of liquid steel. Slag is formed during the melting and consists of different oxides. Slag layer has several functions: It collects the impurities, acts as a thermal insulation blanket, and can in some situations foam. Foaming of the slag is caused via gas formation of oxidation of carbon in the melt. Since electric arcs radiate heat which can damage the lining of the furnace, foaming can be beneficial because it can cover the arcs and reduce the heat loss improving efficiency of the furnace. During the operation additives such as coke and calcium oxide or magnesium oxide can be added to improve slag formation. Since quality of metal scrap used as raw material can vary, controlling of amounts of needed additives for optimum operation can be challenging [17, 18]. During this thesis emission spectroscopy has been investigated as a one possible tool for getting feedback from the EAF directly. Since electric arcs radiate high intensity light which contain atomic emission lines as in analytical emission spectrometers, spectroscopic method provides composition information about the substances present in the arc. In the EAF a strong electric arcs produce plasmas which emit intense light as the evaporated materials are continuously excited and relaxed via radiative and other processes. In Paper VI excitation of metal and slag component samples were done with DC electric arc between graphite electrodes and sample materials using EAF melting process.

be formed during the operation. Top of the furnace is covered with a roof which has a hole in the center for graphite electrodes. The roof is retractable and is opened for charging of the raw materials. As some of the heat produced during the operation of the furnace is absorbed by the wall and roof, there is water cooling to keep the case of the furnace in safe operation temperature. Usually EAFs are installed above the floor level so that melted steel can be poured to ladles for further processing. Large high power AC furnaces make use of three graphite electrodes which deliver high current order of several tens of kA to electric arcs. Typical transformer power of a large EAF can be order of 100 MW and melting time for one charge around 40 min [17, 18]. One important part of steel making and operation of the EAF is the formation of slag which covers the surface of liquid steel. Slag is formed during the melting and consists of different oxides. Slag layer has several functions: It collects the impurities, acts as a thermal insulation blanket, and can in some situations foam. Foaming of the slag is caused via gas formation of oxidation of carbon in the melt. Since electric arcs radiate heat which can damage the lining of the furnace, foaming can be beneficial because it can cover the arcs and reduce the heat loss improving efficiency of the furnace. During the operation additives such as coke and calcium oxide or magnesium oxide can be added to improve slag formation. Since quality of metal scrap used as raw material can vary, controlling of amounts of needed additives for optimum operation can be challenging [17, 18]. During this thesis emission spectroscopy has been investigated as a one possible tool for getting feedback from the EAF directly. Since electric arcs radiate high intensity light which contain atomic emission lines as in analytical emission spectrometers, spectroscopic method provides composition information about the substances present in the arc. In the EAF a strong electric arcs produce plasmas which emit intense light as the evaporated materials are continuously excited and relaxed via radiative and other processes. In Paper VI excitation of metal and slag component samples were done with DC electric arc between graphite electrodes and sample materials using EAF melting process.

34

34

Chapter 6

Chapter 6

Data handling for electron, ion, and photon spectra

Data handling for electron, ion, and photon spectra

6.1

6.1

Electron spectra

Electron spectra

In Papers I-IV, electron spectra were collected using electron spectrometers. Usually electron spectrometers produce raw spectrum data which must be carefully calibrated and analyzed. In this section some of the electron spectra calibration procedures and principles are introduced.

In Papers I-IV, electron spectra were collected using electron spectrometers. Usually electron spectrometers produce raw spectrum data which must be carefully calibrated and analyzed. In this section some of the electron spectra calibration procedures and principles are introduced.

6.1.1

6.1.1

Transmission

The transmission of an electron spectrometer is defined as a ratio between number of electrons entering the spectrometer and the number of detected electrons. Transmission of hemispherical electron spectrometer varies typically as a function of the kinetic energy of the electrons [75]. This introduces usually unwanted intensity variations to spectra, if not corrected. Well known photoelectron lines of noble gases [76] can be used for calibration. Tunable photon energy of insertion devices allows selection of photoelectron line positions at different kinetic energies. Simultaneously with photoelectrons as subsequent related process, Auger lines may appear. By comparing relative intensity of the Auger lines IAuger and photoelectron lines Iphoto , it is possible to construct the transmission function Ftr (E) of the electron spectrometer [77] P Iphoto (E) . (6.1) Ftr (E) = P IAuger

Transmission

The transmission of an electron spectrometer is defined as a ratio between number of electrons entering the spectrometer and the number of detected electrons. Transmission of hemispherical electron spectrometer varies typically as a function of the kinetic energy of the electrons [75]. This introduces usually unwanted intensity variations to spectra, if not corrected. Well known photoelectron lines of noble gases [76] can be used for calibration. Tunable photon energy of insertion devices allows selection of photoelectron line positions at different kinetic energies. Simultaneously with photoelectrons as subsequent related process, Auger lines may appear. By comparing relative intensity of the Auger lines IAuger and photoelectron lines Iphoto , it is possible to construct the transmission function Ftr (E) of the electron spectrometer [77] P Iphoto (E) . (6.1) Ftr (E) = P IAuger

While Auger lines are located at their respective positions regardless of the exciting photon energy (given that initial hole for Auger process is pro-

While Auger lines are located at their respective positions regardless of the exciting photon energy (given that initial hole for Auger process is pro-

35

35

duced), photoelectron line positions depend on binding energy of the electrons in question and the photon energy E. By moving photoelectron lines at selected steps over the region of transmission function to be determined, one can construct the transmission function. Even if the interaction probability of photons and atoms change when the photon energy is changed, the intensity of the Auger lines changes the same way because the intensity of the Auger lines depends on an electron hole created by the photons [77]. When the transmission function is constructed, the absolute intensity of the lines do not affect the result because transmission function is a ratio between the intensities of the Auger and photoelectron lines, which stays constant regardless of the exciting photon energy. The change in relative intensity between Auger and photoelectron lines follows the electron spectrometer transmission.

duced), photoelectron line positions depend on binding energy of the electrons in question and the photon energy E. By moving photoelectron lines at selected steps over the region of transmission function to be determined, one can construct the transmission function. Even if the interaction probability of photons and atoms change when the photon energy is changed, the intensity of the Auger lines changes the same way because the intensity of the Auger lines depends on an electron hole created by the photons [77]. When the transmission function is constructed, the absolute intensity of the lines do not affect the result because transmission function is a ratio between the intensities of the Auger and photoelectron lines, which stays constant regardless of the exciting photon energy. The change in relative intensity between Auger and photoelectron lines follows the electron spectrometer transmission.

6.1.2

6.1.2

Background substraction

Background substraction

Typically experimental spectra include background which needs to be subtracted in order to improve the presentation and analysis of data. Electron spectra as well as optical spectra may include background. Usually electron spectra have a background due to different factors. These are random counts occurring in the detector, thermal electrons from the heated vapor source, and in case of a solid state spectra contribution from electron collisions with the lattice of sample material when electrons are released inside the sample [78]. The background of solid state spectra can be estimated using a semi-empirical Shirley background shape [79]. This background removal procedure was used in Papers III and IV, where photoelectron spectroscopy was used to study solid samples. The calculation procedure of the Shirley background uses of the signal strength of the electron spectrum as function of energy for accumulative background. The background can be seen as an accumulated signal due to scattered electrons from the solid which have less energy than the nonscattered electrons originating from the photo or Auger electron lines.

Typically experimental spectra include background which needs to be subtracted in order to improve the presentation and analysis of data. Electron spectra as well as optical spectra may include background. Usually electron spectra have a background due to different factors. These are random counts occurring in the detector, thermal electrons from the heated vapor source, and in case of a solid state spectra contribution from electron collisions with the lattice of sample material when electrons are released inside the sample [78]. The background of solid state spectra can be estimated using a semi-empirical Shirley background shape [79]. This background removal procedure was used in Papers III and IV, where photoelectron spectroscopy was used to study solid samples. The calculation procedure of the Shirley background uses of the signal strength of the electron spectrum as function of energy for accumulative background. The background can be seen as an accumulated signal due to scattered electrons from the solid which have less energy than the nonscattered electrons originating from the photo or Auger electron lines.

6.2

6.2

Mass spectra

Mass spectra

The mass spectra acquired for the relative photoionization cross-section presented in Paper V were calibrated using the known time-of-flight relation [80].

The mass spectra acquired for the relative photoionization cross-section presented in Paper V were calibrated using the known time-of-flight relation [80].

r m + b, T =a z

r m + b, T =a z

36

(6.2)

36

(6.2)

H2 O+

where T is the total time-of-flight, a and b are calibration constants, and m/z is the mass to charge ratio of the ion. Calibration of the ions is done by recognising two known peaks from a mass spectrum and then forming an equation pair which will yield calibration constants a and b. After calibration the peaks of time-of-flight spectra can be recognized using known atomic masses of different atoms and their combination for molecules. Intensity [arb. units]

Intensity [arb. units]

where T is the total time-of-flight, a and b are calibration constants, and m/z is the mass to charge ratio of the ion. Calibration of the ions is done by recognising two known peaks from a mass spectrum and then forming an equation pair which will yield calibration constants a and b. After calibration the peaks of time-of-flight spectra can be recognized using known atomic masses of different atoms and their combination for molecules.

Xe+

3000 4000 5000 6000 7000 8000 9000 Time [ns]

H2 O+

Xe+

3000 4000 5000 6000 7000 8000 9000 Time [ns]

Figure 6.1: Example of a time-of-flight spectrum collected at beam line I3 FINEST at MAX-lab.

Figure 6.1: Example of a time-of-flight spectrum collected at beam line I3 FINEST at MAX-lab.

A time-of-flight instrument produces ion count rate which is proportional to the interaction cross-section of exciting photons and target atoms, as well as sample vapor pressure. By applying a calibration procedure which involves photon flux and a noble gas calibration signal collected at certain constant photon energy, a relative photo-ionization cross-section can be obtained –see Paper V. The error analysis of obtained data can be performed with the total differential method [81] which is a standard for estimating error in experimental data when the final calculated results of an experiment depend on multiple experimental values. This method takes relative ratios of error in values used in calculation into account and propagates them trough data handling formulas into final error estimate. In general, for final result f depending on three measured variables a, b, c and their respective errors being ∆a, ∆b, ∆c the error ∆f of value f can be estimated as follows

A time-of-flight instrument produces ion count rate which is proportional to the interaction cross-section of exciting photons and target atoms, as well as sample vapor pressure. By applying a calibration procedure which involves photon flux and a noble gas calibration signal collected at certain constant photon energy, a relative photo-ionization cross-section can be obtained –see Paper V. The error analysis of obtained data can be performed with the total differential method [81] which is a standard for estimating error in experimental data when the final calculated results of an experiment depend on multiple experimental values. This method takes relative ratios of error in values used in calculation into account and propagates them trough data handling formulas into final error estimate. In general, for final result f depending on three measured variables a, b, c and their respective errors being ∆a, ∆b, ∆c the error ∆f of value f can be estimated as follows

∆f ≤ |

∂f ∂f ∂f |∆a + | |∆b + | |∆c. ∂a ∂b ∂c

(6.3)

∆f ≤ |

∂f ∂f ∂f |∆a + | |∆b + | |∆c. ∂a ∂b ∂c

(6.3)

Error analysis in Paper V was done with this method using a Matlab data handling code wrote by the author.

Error analysis in Paper V was done with this method using a Matlab data handling code wrote by the author.

37

37

Optical spectra

6.3

Intensity [arb. units]

The laboratory scale EAF presented in Paper VI was used for optical emission study of electric arcs for different substances used in stainless steel production. Emission from arcs were collected using a commercial Andor SR-500i spectrometer which allows collecting individual spectra as a time series with constant acquisition period. Time series collection permits greater flexibility in data handling in case of changing emission profile and allows continuous monitoring of the emission during the recoding as shown in Figure 6.2.

0 20

300

30

400

40 Time [spectrum number]

500

50 60

600

Wavelength [nm]

Optical spectra

The laboratory scale EAF presented in Paper VI was used for optical emission study of electric arcs for different substances used in stainless steel production. Emission from arcs were collected using a commercial Andor SR-500i spectrometer which allows collecting individual spectra as a time series with constant acquisition period. Time series collection permits greater flexibility in data handling in case of changing emission profile and allows continuous monitoring of the emission during the recoding as shown in Figure 6.2.

Intensity [arb. units]

6.3

0 20

300

30

400

40 Time [spectrum number]

500

50 60

600

Wavelength [nm]

Figure 6.2: An example of time series spectra of a large scale EAF light emission. The EAF turns on during the recording of the time series as can be seen by appearance of characteristic emission spectra. Recoding interval for spectra is 0.55 s and total recoding time is 22 s from spectrum number 20 to 60. Note, that some of the spectra are omitted for clearer visual presentation of the series.

Figure 6.2: An example of time series spectra of a large scale EAF light emission. The EAF turns on during the recording of the time series as can be seen by appearance of characteristic emission spectra. Recoding interval for spectra is 0.55 s and total recoding time is 22 s from spectrum number 20 to 60. Note, that some of the spectra are omitted for clearer visual presentation of the series.

In addition to the home laboratory measurements, the experiment was performed at Tornio Steel Works factory [2] as introduced in Chapter 3.3.3. The collected spectra were recoded as a time series (Figure 6.2) to see apparent changes of optical emission during the melting process. The experiment validated the application of the OES in the industrial scale. The acquired spectra contained large number of characteristic emission peaks and contained apparent changes of emission profile during the melting process [2,82].

In addition to the home laboratory measurements, the experiment was performed at Tornio Steel Works factory [2] as introduced in Chapter 3.3.3. The collected spectra were recoded as a time series (Figure 6.2) to see apparent changes of optical emission during the melting process. The experiment validated the application of the OES in the industrial scale. The acquired spectra contained large number of characteristic emission peaks and contained apparent changes of emission profile during the melting process [2,82].

38

38

Industrial EAF Intensity [arb. units]

Intensity [arb. units]

Industrial EAF

Fe+Cr

Fe Lab. EAF

Fe+Cr

Fe Lab. EAF

Cr Lab. EAF

420

440

Cr Lab. EAF

460

480

500 520 540 Wavelength [nm]

560

580

420

440

460

480

500 520 540 Wavelength [nm]

560

580

Figure 6.3: Example comparison of an optical emission spectrum of an industrial scale ladle EAF heating steel and laboratory scale EAF spectra. Spectra of Fe and Cr are presented in Paper VI. See text for details.

Figure 6.3: Example comparison of an optical emission spectrum of an industrial scale ladle EAF heating steel and laboratory scale EAF spectra. Spectra of Fe and Cr are presented in Paper VI. See text for details.

Figure 6.3 shows an example comparison of an optical emission spectrum from an industrial scale EAF, collected with setup presented in Chapter 3.3.3, to spectra of Fe and Cr presented in Paper VI. The spectra are collected using the same optical spectrometer and the identical experimental resolution. One can see clear resemblance of the industrial scale EAF spectrum and the laboratory scale EAF spectra with the sum spectrum of Fe+Cr. The sum spectrum which is calculated from experimental spectra of Fe and Cr demonstrates the linear combination approximation presented in the paper. The comparison points out the importance of experimental reference data which can be collected using laboratory scale instruments with known samples in addition to available emission line databases.

Figure 6.3 shows an example comparison of an optical emission spectrum from an industrial scale EAF, collected with setup presented in Chapter 3.3.3, to spectra of Fe and Cr presented in Paper VI. The spectra are collected using the same optical spectrometer and the identical experimental resolution. One can see clear resemblance of the industrial scale EAF spectrum and the laboratory scale EAF spectra with the sum spectrum of Fe+Cr. The sum spectrum which is calculated from experimental spectra of Fe and Cr demonstrates the linear combination approximation presented in the paper. The comparison points out the importance of experimental reference data which can be collected using laboratory scale instruments with known samples in addition to available emission line databases.

6.3.1

6.3.1

Simulated spectra

Simulated spectra

Spectral feature recognition from experimental spectra can be assisted with simulated spectra which are constructed from results of theoretical calculations or database values. Since experimental spectra always include line broadening, the spectral intensities and features may be overlapped and appear different compared to pure spectral line intensity and position informa-

Spectral feature recognition from experimental spectra can be assisted with simulated spectra which are constructed from results of theoretical calculations or database values. Since experimental spectra always include line broadening, the spectral intensities and features may be overlapped and appear different compared to pure spectral line intensity and position informa-

39

39

tion.

tion.

Simulated spectrum can be constructed by calculating convolution between line shape model data resembling the experimental peak shape and a delta peak line spectrum of theoretical or database information. The result is simulated spectra with delta peaks broadened in a similar way compared to experimental one. In Paper VI the spectra of Fe and Cr were compared to simulated spectra constructed from the fitted experimental peak shape and line spectra information obtained from the NIST database [10]. Experimental line shapes are usually approximated with Voight shape [39, 83] which is a combination of Gaussian and Lorentzian peak shape.

Simulated spectrum can be constructed by calculating convolution between line shape model data resembling the experimental peak shape and a delta peak line spectrum of theoretical or database information. The result is simulated spectra with delta peaks broadened in a similar way compared to experimental one. In Paper VI the spectra of Fe and Cr were compared to simulated spectra constructed from the fitted experimental peak shape and line spectra information obtained from the NIST database [10]. Experimental line shapes are usually approximated with Voight shape [39, 83] which is a combination of Gaussian and Lorentzian peak shape.

6.3.2

6.3.2

Background subtraction

Raw optical emission spectra obtained in Paper VI include prominent black body radiation background due to high temperatures of melted material and electrodes. In order to make the spectra and characteristic emission lines more comparable, the background was subtracted by implementing a fitting procedure. In the procedure Planck’s law [84] in the form presented1 below as a photon radiance was utilized. 2cλ−4

Background subtraction

Raw optical emission spectra obtained in Paper VI include prominent black body radiation background due to high temperatures of melted material and electrodes. In order to make the spectra and characteristic emission lines more comparable, the background was subtracted by implementing a fitting procedure. In the procedure Planck’s law [84] in the form presented1 below as a photon radiance was utilized. 2cλ−4

(6.4)

−1 where Lp (λ, T ) is photon radiance as a function of wavelength λ and absolute temperature T , c is the speed of light, h is the Planck constant, k is the Boltzmann’s constant, and ǫ is an additional intensity scaling parameter known as emissivity which represents the deviation from the ideal black body. In Paper VI, the scaling of the background assumed a constant emissivity and, thus emissivity value does not affect the final shape of the background. In Figure 6.4 plotted curves show example of photon radiance curves calculated using Equation 6.4 by assuming constant unit emissivity (ǫ = 1) for radiating heated mass, therefore representing ideal blackbody radiation.

, (6.4) −1 where Lp (λ, T ) is photon radiance as a function of wavelength λ and absolute temperature T , c is the speed of light, h is the Planck constant, k is the Boltzmann’s constant, and ǫ is an additional intensity scaling parameter known as emissivity which represents the deviation from the ideal black body. In Paper VI, the scaling of the background assumed a constant emissivity and, thus emissivity value does not affect the final shape of the background. In Figure 6.4 plotted curves show example of photon radiance curves calculated using Equation 6.4 by assuming constant unit emissivity (ǫ = 1) for radiating heated mass, therefore representing ideal blackbody radiation.

1 Note that there are several versions of Planck’s Law in different units. The presented equation is used because the signal in the CCD detector of the optical spectrometer is proportional to the number of collected photons.

1 Note that there are several versions of Planck’s Law in different units. The presented equation is used because the signal in the CCD detector of the optical spectrometer is proportional to the number of collected photons.

40

40

Lp (λ, T ) = ǫ

e(ch)/(λkT )

,

Lp (λ, T ) = ǫ

e(ch)/(λkT )

visible

Intensity x1030 [photons s−1 sr−1 m−2 ]

Intensity x1030 [photons s−1 sr−1 m−2 ]

50

1000 K 2000 K 3000 K 4000 K 5000 K

40 30 20 10 0 400

800

1200

1600

2000

Wavelength [nm]

50

visible

1000 K 2000 K 3000 K 4000 K 5000 K

40 30 20 10 0 400

800

1200

1600

2000

Wavelength [nm]

Figure 6.4: Calculated blackbody photon radiance in different temperatures as a function of wavelength.

Figure 6.4: Calculated blackbody photon radiance in different temperatures as a function of wavelength.

41

41

42

42

Chapter 7

Chapter 7

Summary and discussion of included papers

Summary and discussion of included papers

7.1

7.1

High temperature vapors: spectroscopy of aluminum, Papers I - II

High temperature vapors: spectroscopy of aluminum, Papers I - II

Aluminum is a very commonly used metal in the industry as raw material due to its good physical properties: it is light, almost inert, and a good conductor of heat and electricity. In Paper I, 2s photoelectron and resulting Auger spectrum of atomic aluminum vapor were recorded using modified SES-100 electron spectrometer [59] at the undulator beam line I411 at MAX-II storage ring in Lund, Sweden. Induction heating method was used for atomic vapor production, utilizing electron signal veto in the detector during heating pulses. Aluminum formed a reactive melt and therefore tungsten crucible was used to prevent reaction of the sample with the crucible material. Recorded spectra were interpreted using atomic calculations as an aid. The result showed that modern third generation synchrotron radiation sources are capable of providing information from the high-temperature low ionization cross-section atomic vapors when used with the high-temperature vapor production experimental setup. In Paper II, KLL Auger spectrum of atomic aluminum was recorded with electron impact for the creation of 1s core hole states. The experimental results were compared to predictions obtained with ab initio calculations for spectral structures. The fluctuating electric and magnetic fields during heating pulses needed to be vetoed. Since Scienta SES-200 electron spectrometer was equipped with a 15 frames per second capable camera detector which could not be vetoed, the electron beam deflection veto method had to be used. This provided an effective mean to avoid interference from the

Aluminum is a very commonly used metal in the industry as raw material due to its good physical properties: it is light, almost inert, and a good conductor of heat and electricity. In Paper I, 2s photoelectron and resulting Auger spectrum of atomic aluminum vapor were recorded using modified SES-100 electron spectrometer [59] at the undulator beam line I411 at MAX-II storage ring in Lund, Sweden. Induction heating method was used for atomic vapor production, utilizing electron signal veto in the detector during heating pulses. Aluminum formed a reactive melt and therefore tungsten crucible was used to prevent reaction of the sample with the crucible material. Recorded spectra were interpreted using atomic calculations as an aid. The result showed that modern third generation synchrotron radiation sources are capable of providing information from the high-temperature low ionization cross-section atomic vapors when used with the high-temperature vapor production experimental setup. In Paper II, KLL Auger spectrum of atomic aluminum was recorded with electron impact for the creation of 1s core hole states. The experimental results were compared to predictions obtained with ab initio calculations for spectral structures. The fluctuating electric and magnetic fields during heating pulses needed to be vetoed. Since Scienta SES-200 electron spectrometer was equipped with a 15 frames per second capable camera detector which could not be vetoed, the electron beam deflection veto method had to be used. This provided an effective mean to avoid interference from the

43

43

heating. The experiment resulted a good quality KLL Auger spectrum and intensity rations of different Auger groups were obtained.

heating. The experiment resulted a good quality KLL Auger spectrum and intensity rations of different Auger groups were obtained.

7.2

7.2

Atom-solid binding energy shifts, Papers III - IV

Atom-solid binding energy shifts, Papers III - IV

In Paper III binding energy shifts between atomic and solid states were observed with improved accuracy for Au 4f and Ag 3d core levels using a simultaneous measurement method. Experimental setup used inductively heated vapor source and a needle target to allow simultaneous recording of vapor and solid phases. Modified SES-100 electron spectrometer [59] was used for collecting the spectra and photon excitation was provided by I411 beamline in MAX-II storage ring at MAX-lab, Lund. The 3d photoelectron spectrum and binding energies of free Ag atoms were observed for the first time. New values were obtained and free atom to bulk solid shifts deviated considerably from the previous estimates for Au but agreed reasonably well for Ag. The valence bands were also found to shift rather accurately by the same amount as the core levels. In Paper IV binding energy shift of 3p sub-shell electrons were experimentally determined using a simultaneous measurement of atomic vapor and solid phase with similar experimental setup as in Paper III. In vapor production reactivity of high temperature melts of studied metals with molybdenum crucible material were prevented using Al2 O3 lining. Binding energy shifts between atomic and solid phases for transition metals Cr, Mn, Fe, Co, and Ni were obtained for 3p subshells from simultaneously recorded atomic vapor and solid state spectra. This experimental technique provided higher accuracy in comparison to separate measurements and allowed direct determination of the shifts. The observed shift values were compared to values obtained using the semiempirical Born-Haber cycle method and the peak structures of the solid state photoelectron spectra were compared to atomic 3p spectra. The binding energy shift of Cr was found to be much smaller than that of the other studied elements. Comparison between experimental peak-to-peak shift values and the semiempirical Born-Haber cycle model showed a similar trend. As the BornHaber cycle fully omits the multiplet structure, calculations do not reproduce the experimental values entirely. The small shift value of Cr compared to other studied elements could be explained with the electron configuration change between solid and atom in with the others while in Cr the configuration remained the same. The relative positions of broad structural features in

In Paper III binding energy shifts between atomic and solid states were observed with improved accuracy for Au 4f and Ag 3d core levels using a simultaneous measurement method. Experimental setup used inductively heated vapor source and a needle target to allow simultaneous recording of vapor and solid phases. Modified SES-100 electron spectrometer [59] was used for collecting the spectra and photon excitation was provided by I411 beamline in MAX-II storage ring at MAX-lab, Lund. The 3d photoelectron spectrum and binding energies of free Ag atoms were observed for the first time. New values were obtained and free atom to bulk solid shifts deviated considerably from the previous estimates for Au but agreed reasonably well for Ag. The valence bands were also found to shift rather accurately by the same amount as the core levels. In Paper IV binding energy shift of 3p sub-shell electrons were experimentally determined using a simultaneous measurement of atomic vapor and solid phase with similar experimental setup as in Paper III. In vapor production reactivity of high temperature melts of studied metals with molybdenum crucible material were prevented using Al2 O3 lining. Binding energy shifts between atomic and solid phases for transition metals Cr, Mn, Fe, Co, and Ni were obtained for 3p subshells from simultaneously recorded atomic vapor and solid state spectra. This experimental technique provided higher accuracy in comparison to separate measurements and allowed direct determination of the shifts. The observed shift values were compared to values obtained using the semiempirical Born-Haber cycle method and the peak structures of the solid state photoelectron spectra were compared to atomic 3p spectra. The binding energy shift of Cr was found to be much smaller than that of the other studied elements. Comparison between experimental peak-to-peak shift values and the semiempirical Born-Haber cycle model showed a similar trend. As the BornHaber cycle fully omits the multiplet structure, calculations do not reproduce the experimental values entirely. The small shift value of Cr compared to other studied elements could be explained with the electron configuration change between solid and atom in with the others while in Cr the configuration remained the same. The relative positions of broad structural features in

44

44

the solid spectra showed similarities to the atomic spectra of next elements in the period. In conclusion, the electron configuration change seemed to explain the shift trend. Lack of clear multiplet structure at higher binding energies in solids was explained to be most probably due to opening of SCK channels for components still appearing in atomic spectra. In this paper we also reported absolute binding energies of solid state peaks with respect to vacuum level.

the solid spectra showed similarities to the atomic spectra of next elements in the period. In conclusion, the electron configuration change seemed to explain the shift trend. Lack of clear multiplet structure at higher binding energies in solids was explained to be most probably due to opening of SCK channels for components still appearing in atomic spectra. In this paper we also reported absolute binding energies of solid state peaks with respect to vacuum level.

7.3

7.3

Relative photoionization cross section of Cr atoms, Paper V

Relative photoionization cross section of Cr atoms, Paper V

In Paper V total single photoionization of chromium was investigated by means of time-of-flight spectroscopy. The experiment was carried out at MAX-lab in gas phase branch of beamline I3 [72] with a VUV energy range of 10 to 35 eV. Atomic chromium vapor was produced with an inductively heated oven at continuous heating mode making stable vapor beam. A WileyMcLaren type of TOFMS was used to acquire mass spectra at selected photon energies as well as reference spectra of Ar+ and beam off spectra for calibration.

In Paper V total single photoionization of chromium was investigated by means of time-of-flight spectroscopy. The experiment was carried out at MAX-lab in gas phase branch of beamline I3 [72] with a VUV energy range of 10 to 35 eV. Atomic chromium vapor was produced with an inductively heated oven at continuous heating mode making stable vapor beam. A WileyMcLaren type of TOFMS was used to acquire mass spectra at selected photon energies as well as reference spectra of Ar+ and beam off spectra for calibration.

Collected data was handled using a Matlab code written by the author during this work in order to evaluate large number of spectra systematically and evaluate the error limits utilizing a total differential method. The experimental result was compared to a semiempirical model of ionization cross-section which used combined cross-section calculations from the literature [85] and final states from the NIST database [10].

Collected data was handled using a Matlab code written by the author during this work in order to evaluate large number of spectra systematically and evaluate the error limits utilizing a total differential method. The experimental result was compared to a semiempirical model of ionization cross-section which used combined cross-section calculations from the literature [85] and final states from the NIST database [10].

While the wide energy splitting of the singly ionized valence states explained the behavior of the ionization cross section in the lower-energy end, reasons for the mismatch between the calculations and the experiment in the higher-energy end were discussed. Possible reason for difference was considered to be the open shell nature of chromium which may affect the results of the calculation as possible interchannel coupling was omitted in used reference data [85]. Also the magnitude of 3d and 4s ionization cross sections was found to have different values in reference [86] when it was compared to reference [85]. The conclusion was that experimental results and theoretical models differ in the literature and more experimental and theoretical work would be needed in order to clarify the differences.

While the wide energy splitting of the singly ionized valence states explained the behavior of the ionization cross section in the lower-energy end, reasons for the mismatch between the calculations and the experiment in the higher-energy end were discussed. Possible reason for difference was considered to be the open shell nature of chromium which may affect the results of the calculation as possible interchannel coupling was omitted in used reference data [85]. Also the magnitude of 3d and 4s ionization cross sections was found to have different values in reference [86] when it was compared to reference [85]. The conclusion was that experimental results and theoretical models differ in the literature and more experimental and theoretical work would be needed in order to clarify the differences.

45

45

7.4

Optical emission from electric arc furnace, Paper VI

7.4

Optical emission from electric arc furnace, Paper VI

In Paper VI a design for a laboratory scale DC electric arc furnace developed during this work for optical emission studies related to industrial EAF melting process was presented. The furnace was a continuation of other experimental apparatuses developed and used during this work. For demonstration several spectra of substances present in stainless steel melting process were recorded with a commercial Andor Technology SR500i spectrometer. The resolution of the optical spectrometer was selected so that it was possible to record simultaneously optical emission from the furnace in wide optical range covering the visible spectrum including the ultraviolet regime down to 250 nm. The furnace was designed to emulate a larger scale EAF melting process and enabled control for a composition of the gas atmosphere inside the chamber. This allows wide variety of possible experiments to be performed with different combinations of melted materials in controlled gas atmospheres. Several aspects of the recorded emission spectra were studied, including plasma temperature determination, signs of molecular species in the plasma, and comparison of the spectra to simulated atomic emission spectra constructed utilizing NIST [10] reference data. Plasma temperatures were determined from the collected spectra by Boltzmann plot method [27] and several emission lines for each collected spectra were used. The results show temperatures in range of 6000K to 16000K, with relatively large error limits possibly due to fluctuating experimental conditions and self-absorption. However, temperature estimation provided rough estimate of average plasma temperatures. It was found that the spectra of the iron group elements contain a large number of emission lines which cover wide range of optical regime. In the experiment several compounds present in typical EAF slag were recoded separately in addition to main metals of stainless steel. Optical emission spectra from the slag oxides display several high intensity emission peaks while some of the high intensity peaks were found to be missing due to selfabsorption. Spectra of iron and silica show clear molecular Swan bands of diatomic carbon molecule and the spectrum of aluminum oxide contained molecular AlO bands. A linear combination sum spectrum of collected individual spectra was constructed by scaling intensity of individual component spectra to resemble the composition of a typical slag. In addition to collecting separate spectra of Fe and Cr, emission spectrum of equal mixture Fe and Cr was collected and compared to separately collected spectra to test the linear combination ap-

In Paper VI a design for a laboratory scale DC electric arc furnace developed during this work for optical emission studies related to industrial EAF melting process was presented. The furnace was a continuation of other experimental apparatuses developed and used during this work. For demonstration several spectra of substances present in stainless steel melting process were recorded with a commercial Andor Technology SR500i spectrometer. The resolution of the optical spectrometer was selected so that it was possible to record simultaneously optical emission from the furnace in wide optical range covering the visible spectrum including the ultraviolet regime down to 250 nm. The furnace was designed to emulate a larger scale EAF melting process and enabled control for a composition of the gas atmosphere inside the chamber. This allows wide variety of possible experiments to be performed with different combinations of melted materials in controlled gas atmospheres. Several aspects of the recorded emission spectra were studied, including plasma temperature determination, signs of molecular species in the plasma, and comparison of the spectra to simulated atomic emission spectra constructed utilizing NIST [10] reference data. Plasma temperatures were determined from the collected spectra by Boltzmann plot method [27] and several emission lines for each collected spectra were used. The results show temperatures in range of 6000K to 16000K, with relatively large error limits possibly due to fluctuating experimental conditions and self-absorption. However, temperature estimation provided rough estimate of average plasma temperatures. It was found that the spectra of the iron group elements contain a large number of emission lines which cover wide range of optical regime. In the experiment several compounds present in typical EAF slag were recoded separately in addition to main metals of stainless steel. Optical emission spectra from the slag oxides display several high intensity emission peaks while some of the high intensity peaks were found to be missing due to selfabsorption. Spectra of iron and silica show clear molecular Swan bands of diatomic carbon molecule and the spectrum of aluminum oxide contained molecular AlO bands. A linear combination sum spectrum of collected individual spectra was constructed by scaling intensity of individual component spectra to resemble the composition of a typical slag. In addition to collecting separate spectra of Fe and Cr, emission spectrum of equal mixture Fe and Cr was collected and compared to separately collected spectra to test the linear combination ap-

46

46

proximation. The combination spectra demonstrated that despite relatively moderate resolution it is possible to find suitable indicator spectrum lines for different compounds investigated in the paper. The furnace was found to be suitable for further optical emission studies of EAF melting process involving variety of different compounds and it allowed the materials to form a pool of melt as in a real EAF process in addition to providing a direct view of the arc for study. It was confirmed that OES can in principle be used for real-time monitoring and it provides a broad range of information directly from the EAF to be used in the metal industry for improving process control and efficiency.

proximation. The combination spectra demonstrated that despite relatively moderate resolution it is possible to find suitable indicator spectrum lines for different compounds investigated in the paper. The furnace was found to be suitable for further optical emission studies of EAF melting process involving variety of different compounds and it allowed the materials to form a pool of melt as in a real EAF process in addition to providing a direct view of the arc for study. It was confirmed that OES can in principle be used for real-time monitoring and it provides a broad range of information directly from the EAF to be used in the metal industry for improving process control and efficiency.

47

47

48

48

Chapter 8

Chapter 8

Conclusions and future prospects

Conclusions and future prospects

The investigations in this thesis provide an experimental view of different techniques for studying metal vapors. This work is interdisciplinary and covers atomic physics as well as applied physics in the field of process metallurgy. Papers I and II dealt with high temperature atomic vapors and photo as well as Auger electron spectroscopy. The induction heating technique used in the papers was found to be well suited for producing atomic metal vapors and provides good experimental technique for possible future investigations of atomic vapors as well as metal cluster studies [87]. The measurement of binding energy shifts of metals studied in Papers III and IV gave valuable information about the change in electron binding energies between atomic and solid phases. Obtained results can be used in free cluster studies as a reference for two ends of the size scale which are infinite solid (bulk) and single atom. The photoionization study of atomic Cr vapor presented in Paper V utilized a technique which can be used for determination of relative total ionization cross-sections for metal vapors. The miniature EAF presented in Paper VI was found to provide an excellent experimental environment for studies of optical emission directly from various metals and oxides which are present in stainless steel production. Since the light for the EAF contains highly characteristic emission lines specific for different substances and elements, the technique is highly selective and indicative if an appropriate instrumental resolution is used. The validity and suitability of using the linear combination approximation of overlapping lower resolution spectra, as in Paper VI, for chemical analysis in EAF environments could be studied in the future to see is it possible to ease the resolution requirement and still provide valid results. This could enable the use of less expensive spectrographic devices for EAF state monitoring, and in general increase the use of the technique. The OES method and developed apparatuses in this thesis can be used

The investigations in this thesis provide an experimental view of different techniques for studying metal vapors. This work is interdisciplinary and covers atomic physics as well as applied physics in the field of process metallurgy. Papers I and II dealt with high temperature atomic vapors and photo as well as Auger electron spectroscopy. The induction heating technique used in the papers was found to be well suited for producing atomic metal vapors and provides good experimental technique for possible future investigations of atomic vapors as well as metal cluster studies [87]. The measurement of binding energy shifts of metals studied in Papers III and IV gave valuable information about the change in electron binding energies between atomic and solid phases. Obtained results can be used in free cluster studies as a reference for two ends of the size scale which are infinite solid (bulk) and single atom. The photoionization study of atomic Cr vapor presented in Paper V utilized a technique which can be used for determination of relative total ionization cross-sections for metal vapors. The miniature EAF presented in Paper VI was found to provide an excellent experimental environment for studies of optical emission directly from various metals and oxides which are present in stainless steel production. Since the light for the EAF contains highly characteristic emission lines specific for different substances and elements, the technique is highly selective and indicative if an appropriate instrumental resolution is used. The validity and suitability of using the linear combination approximation of overlapping lower resolution spectra, as in Paper VI, for chemical analysis in EAF environments could be studied in the future to see is it possible to ease the resolution requirement and still provide valid results. This could enable the use of less expensive spectrographic devices for EAF state monitoring, and in general increase the use of the technique. The OES method and developed apparatuses in this thesis can be used

49

49

as a probe to study the evaporated and excited species present in industrial scale EAFs. Obtained information could in principle be combined with other industrial process online data such as EAF off-gas analyser data and post melting composition analysis of slag and steel. This could improve process control and ease the difficulties related to raw material quality variations [88] in the industrial scale stainless steel production. Utilizing the laboratory scale vacuum tight miniature EAF presented in Paper VI one could in principle study the effects of EAF gas atmosphere to optical emission as well as make reference data for emission profiles of wide variety of different compositions of slags and metals. In conclusion, the laboratory scale EAF apparatus and OES technique enables a range of application and fundamental science studies, beneficial for both industry as well as science.

as a probe to study the evaporated and excited species present in industrial scale EAFs. Obtained information could in principle be combined with other industrial process online data such as EAF off-gas analyser data and post melting composition analysis of slag and steel. This could improve process control and ease the difficulties related to raw material quality variations [88] in the industrial scale stainless steel production. Utilizing the laboratory scale vacuum tight miniature EAF presented in Paper VI one could in principle study the effects of EAF gas atmosphere to optical emission as well as make reference data for emission profiles of wide variety of different compositions of slags and metals. In conclusion, the laboratory scale EAF apparatus and OES technique enables a range of application and fundamental science studies, beneficial for both industry as well as science.

50

50

Bibliography

Bibliography

[1] M. Hauru, Internal report: Grafiitin, raudan, nikkelin ja teräksen fluoresenssispektrejä valokaaresta, University of Oulu, 2010.

[1] M. Hauru, Internal report: Grafiitin, raudan, nikkelin ja teräksen fluoresenssispektrejä valokaaresta, University of Oulu, 2010.

[2] A. Mäkinen and J. Niskanen, Internal report: Tornion terästehtaalla senkka-asemalla tehdyt fluoresenssimittaukset, University of Oulu, 2011.

[2] A. Mäkinen and J. Niskanen, Internal report: Tornion terästehtaalla senkka-asemalla tehdyt fluoresenssimittaukset, University of Oulu, 2011.

[3] J. Thomson, Philos. Mag. 44, 293 (1897).

[3] J. Thomson, Philos. Mag. 44, 293 (1897).

[4] H. Geiger and E. Marsden, Proc. R. Soc. A 82, 495 (1909).

[4] H. Geiger and E. Marsden, Proc. R. Soc. A 82, 495 (1909).

[5] N. Bohr, Philos. Mag. 26, 1 (1913).

[5] N. Bohr, Philos. Mag. 26, 1 (1913).

[6] M. J. Osler, Notes & Records of The Royal Society 60, 291 (2006).

[6] M. J. Osler, Notes & Records of The Royal Society 60, 291 (2006).

[7] A. Einstein, Ann. Physics 17, 132 (1905).

[7] A. Einstein, Ann. Physics 17, 132 (1905).

[8] K. Siegbahn, C. Nordling, A. Fahlman, R. Nordberg, K. Hamrin, J. Hedman, G. Johansson, T. Bergmark, S.-E. Karlsson, I. Lindgren, and B. Lindberg, ESCA - atomic, Molecular and Solid State Structure Studied by Means of Electron Spectroscopy (Almqvist and Wiksells, Uppsala, Sweden, 1967).

[8] K. Siegbahn, C. Nordling, A. Fahlman, R. Nordberg, K. Hamrin, J. Hedman, G. Johansson, T. Bergmark, S.-E. Karlsson, I. Lindgren, and B. Lindberg, ESCA - atomic, Molecular and Solid State Structure Studied by Means of Electron Spectroscopy (Almqvist and Wiksells, Uppsala, Sweden, 1967).

[9] K. Siegbahn, C. Nordling, G. Johansson, J. Hedman, P. F. Hedén, K. Hamrin, U. Gelius, T. Bergmark, L. Werme, R. Manne, and Y. Baer, ESCA applied to free molecules (North-Holland publishing company, Amsterdam, Netherlands, 1969).

[9] K. Siegbahn, C. Nordling, G. Johansson, J. Hedman, P. F. Hedén, K. Hamrin, U. Gelius, T. Bergmark, L. Werme, R. Manne, and Y. Baer, ESCA applied to free molecules (North-Holland publishing company, Amsterdam, Netherlands, 1969).

[10] Y. Ralchenko, A. Kramida, J. Reader, and N. A. Team, NIST atomic spectra database, National Institute of Standards and Technology.

[10] Y. Ralchenko, A. Kramida, J. Reader, and N. A. Team, NIST atomic spectra database, National Institute of Standards and Technology.

[11] T. H. Mainman, Nature 187, 493 (1960).

[11] T. H. Mainman, Nature 187, 493 (1960).

[12] A. Schultz, H. W. Cruse, and R. N. Zare, J. Chem. Phys. 57, 1354 (1972).

[12] A. Schultz, H. W. Cruse, and R. N. Zare, J. Chem. Phys. 57, 1354 (1972).

51

51

[13] J. Álvarez Ruiz, E. Melero-García, A. Kivimäki, M. Corenoa, P. Erman, E. Rachlew, and R. Richter, J. Phys. B-At. Mol. Opt. 38, 387 (2004).

[13] J. Álvarez Ruiz, E. Melero-García, A. Kivimäki, M. Corenoa, P. Erman, E. Rachlew, and R. Richter, J. Phys. B-At. Mol. Opt. 38, 387 (2004).

[14] X. Hou and B. Jones, Encyclopedia of Analytical Chemistry: Inductively Coupled Plasma/Optical Emission Spectrometry (John-Wiley & Sons Ltd, Chichester, England, 2000), pp. 9468–9485.

[14] X. Hou and B. Jones, Encyclopedia of Analytical Chemistry: Inductively Coupled Plasma/Optical Emission Spectrometry (John-Wiley & Sons Ltd, Chichester, England, 2000), pp. 9468–9485.

[15] D. Cremers and L. Radziemski, Handbook of Laser-Induced Breakdown Spectroscopy (John-Wiley & Sons Ltd, Chichester, England, 2006).

[15] D. Cremers and L. Radziemski, Handbook of Laser-Induced Breakdown Spectroscopy (John-Wiley & Sons Ltd, Chichester, England, 2006).

[16] D. A. Skoog, F. J. Holler, and S. R. Crouch, Principles of Instrumental Analysis Sixth Edition (Thomson Brooks/Cole, Belmont, USA, 1998).

[16] D. A. Skoog, F. J. Holler, and S. R. Crouch, Principles of Instrumental Analysis Sixth Edition (Thomson Brooks/Cole, Belmont, USA, 1998).

[17] Y. N. Toulouevski and I. Y. Zinurov, Innovation in Electric Arc Furnaces (Springer Heidelberg Dordrecht, London, England, 2010).

[17] Y. N. Toulouevski and I. Y. Zinurov, Innovation in Electric Arc Furnaces (Springer Heidelberg Dordrecht, London, England, 2010).

[18] E. Plöckinger and O. Etterich, Electric Furnace Steel Production (John Wiley & Sons Ltd, New York, USA, 1985).

[18] E. Plöckinger and O. Etterich, Electric Furnace Steel Production (John Wiley & Sons Ltd, New York, USA, 1985).

[19] B. Bowman and K. Krüger, Arc Furnace Physics (Stahleisen, Düsseldorf, Germany, 2009).

[19] B. Bowman and K. Krüger, Arc Furnace Physics (Stahleisen, Düsseldorf, Germany, 2009).

[20] E. Schrödinger, Ann. Phys. 79, 361 (1926).

[20] E. Schrödinger, Ann. Phys. 79, 361 (1926).

[21] E. Schrödinger, Ann. Phys. 79, 489 (1926).

[21] E. Schrödinger, Ann. Phys. 79, 489 (1926).

[22] E. Schrödinger, Ann. Phys. 80, 437 (1926).

[22] E. Schrödinger, Ann. Phys. 80, 437 (1926).

[23] E. Schrödinger, Ann. Phys. 81, 109 (1926).

[23] E. Schrödinger, Ann. Phys. 81, 109 (1926).

[24] W. Pauli, Z. Physik 31, 765 (1925).

[24] W. Pauli, Z. Physik 31, 765 (1925).

[25] R. Cowan, The Theory of Atomic Structure and Spectra (University of California Press, Berkeley, USA, 1981).

[25] R. Cowan, The Theory of Atomic Structure and Spectra (University of California Press, Berkeley, USA, 1981).

[26] W. T. Silfvast, Laser Fundamentals Second Edition (Cambridge University Press, Cambridge, England, 2004), pp. 201–202.

[26] W. T. Silfvast, Laser Fundamentals Second Edition (Cambridge University Press, Cambridge, England, 2004), pp. 201–202.

[27] R. H. Tourin, Spectroscopic Gas Temperature Measurement Pyrometry of Hot Gases and Plasmas (Elsevier, Amsterdam, Netherlands, 1966), pp. 47–48.

[27] R. H. Tourin, Spectroscopic Gas Temperature Measurement Pyrometry of Hot Gases and Plasmas (Elsevier, Amsterdam, Netherlands, 1966), pp. 47–48.

[28] H. Hertz, Ann. Phys. 267, 983 (1887).

[28] H. Hertz, Ann. Phys. 267, 983 (1887).

[29] C. Kittel, Introduction to Solid State Physics (John Wiley & Sons Inc., New York, USA, 1995).

[29] C. Kittel, Introduction to Solid State Physics (John Wiley & Sons Inc., New York, USA, 1995).

52

52

[30] S. Hagström, C. Nordling, and K. Siegbahn, Z. Phys. 178, 439 (1964).

[30] S. Hagström, C. Nordling, and K. Siegbahn, Z. Phys. 178, 439 (1964).

[31] L. Meitner, Z. Phys. A 9, 131 (1922).

[31] L. Meitner, Z. Phys. A 9, 131 (1922).

[32] P. Auger, J. Phys. Rad. 6, 205 (1925).

[32] P. Auger, J. Phys. Rad. 6, 205 (1925).

[33] D. Chattarji, The Theory of Auger Transitions (Academic Press Inc., New York, USA, 1976).

[33] D. Chattarji, The Theory of Auger Transitions (Academic Press Inc., New York, USA, 1976).

[34] E. Burhop, The Auger Effect and Other Radiationless Transitions (Cambridge University Press, Cambridge, England, 1952).

[34] E. Burhop, The Auger Effect and Other Radiationless Transitions (Cambridge University Press, Cambridge, England, 1952).

[35] F. H.G. Kuhn, Atomic Spectra (Longmans, Green and Co Ltd, London, England, 1962).

[35] F. H.G. Kuhn, Atomic Spectra (Longmans, Green and Co Ltd, London, England, 1962).

[36] K. Ross and B. Sonntag, Rev. Sci. Instrum. 66, 4409 (1995).

[36] K. Ross and B. Sonntag, Rev. Sci. Instrum. 66, 4409 (1995).

[37] R. Siegel and J. Howell, Thermal Radiation Heat Transfer Third Edition (Hemisphere Publishing Corporation, Washington, USA, 1992).

[37] R. Siegel and J. Howell, Thermal Radiation Heat Transfer Third Edition (Hemisphere Publishing Corporation, Washington, USA, 1992).

[38] T. Rander, J. Schulz, M. Huttula, A. Mäkinen, M. Tchaplyquine, S. Svensson, G. Öhrwall, O. Björneholm, S. Aksela, and H. Aksela, Phys. Rev. A 75, 032510 (2007).

[38] T. Rander, J. Schulz, M. Huttula, A. Mäkinen, M. Tchaplyquine, S. Svensson, G. Öhrwall, O. Björneholm, S. Aksela, and H. Aksela, Phys. Rev. A 75, 032510 (2007).

[39] J. Moore, C. Davis, M. Coplan, and S. Green, Building Scientific Apparatus (Cambridge University Press, Cambridge, England, 2009).

[39] J. Moore, C. Davis, M. Coplan, and S. Green, Building Scientific Apparatus (Cambridge University Press, Cambridge, England, 2009).

[40] T. Rander, J. Schulz, M. Huttula, A. Mäkinen, M. Tchaplyguine, S. Svensson, G. Öhrwall, O. Björneholm, S. Aksela, and H. Aksela, Phys. Rev. A , 032510 (2007).

[40] T. Rander, J. Schulz, M. Huttula, A. Mäkinen, M. Tchaplyguine, S. Svensson, G. Öhrwall, O. Björneholm, S. Aksela, and H. Aksela, Phys. Rev. A , 032510 (2007).

[41] M. Harkoma, Phd thesis: Instrumentation for electron spectroscopy of high temperature vapours, Oulu University Press, 2002.

[41] M. Harkoma, Phd thesis: Instrumentation for electron spectroscopy of high temperature vapours, Oulu University Press, 2002.

[42] G. Wouch and A. E. L. Jr, Am. J. Phys. 46, 464 (1978).

[42] G. Wouch and A. E. L. Jr, Am. J. Phys. 46, 464 (1978).

[43] S. Zinn and S. L. Semiatin, Elements of Induction Heating: Design, Control, and Applications (ASM International, Materials Park, USA, 1988).

[43] S. Zinn and S. L. Semiatin, Elements of Induction Heating: Design, Control, and Applications (ASM International, Materials Park, USA, 1988).

[44] R. Kumpula, J. Väyrynen, T. Rantala, and S. Aksela, J. Phys. C 12, L809 (1979).

[44] R. Kumpula, J. Väyrynen, T. Rantala, and S. Aksela, J. Phys. C 12, L809 (1979).

[45] J. Väyrynen, S. Aksela, and H. Aksela, Phys. Scripta 16, 452 (1977).

[45] J. Väyrynen, S. Aksela, and H. Aksela, Phys. Scripta 16, 452 (1977).

53

53

[46] A. Fridman, Plasma Chemistry (Chambridge University Press, Cambridge, England, 2008).

[46] A. Fridman, Plasma Chemistry (Chambridge University Press, Cambridge, England, 2008).

[47] E. Hecht, Optics Fourth edition (Addison-Wesley, Boston, USA, 1998).

[47] E. Hecht, Optics Fourth edition (Addison-Wesley, Boston, USA, 1998).

[48] A. Davenport, The History of Photography (University of New Mexico Press, Boston, USA, 1999).

[48] A. Davenport, The History of Photography (University of New Mexico Press, Boston, USA, 1999).

[49] M. Harkoma and S. Aksela, J. Electron Spectrosc. 122, 209 (2002).

[49] M. Harkoma and S. Aksela, J. Electron Spectrosc. 122, 209 (2002).

[50] M. Huttula, M. Harkoma, E. N˜ommiste, and S. Aksela, Nucl. Instrum. Meth. A 467-468, 1514 (2001).

[50] M. Huttula, M. Harkoma, E. N˜ommiste, and S. Aksela, Nucl. Instrum. Meth. A 467-468, 1514 (2001).

[51] S. Aksela, M. Karras, M. Pessa, and E. Suoninen, Rev. Sci. Instrum. 41, 351 (1970).

[51] S. Aksela, M. Karras, M. Pessa, and E. Suoninen, Rev. Sci. Instrum. 41, 351 (1970).

[52] S. Aksela, Rev. Sci. Instrum. 43, 1350 (1972).

[52] S. Aksela, Rev. Sci. Instrum. 43, 1350 (1972).

[53] P. Palmberg, G. Bohm, and J. Tracy, Appl. Phys. Lett. 15, 254 (1969).

[53] P. Palmberg, G. Bohm, and J. Tracy, Appl. Phys. Lett. 15, 254 (1969).

[54] J. Preston, M. Hender, and J. McConkey, J. Phys. E Sci. Instrum. 6, 661 (1973).

[54] J. Preston, M. Hender, and J. McConkey, J. Phys. E Sci. Instrum. 6, 661 (1973).

[55] F. Pauty, G. Matula, and P. Vernier, Rev. Sci. Instrum. 45, 1203 (1974).

[55] F. Pauty, G. Matula, and P. Vernier, Rev. Sci. Instrum. 45, 1203 (1974).

[56] D. Roy and D. Tremblay, Rep. Prog. Phys. 53, 1621 (1990).

[56] D. Roy and D. Tremblay, Rep. Prog. Phys. 53, 1621 (1990).

[57] S. Hüfner, Photoelectron Spectroscopy Third Edition (Springer, Berlin, Germany, 2003), pp. 20–23.

[57] S. Hüfner, Photoelectron Spectroscopy Third Edition (Springer, Berlin, Germany, 2003), pp. 20–23.

[58] P. J. Bassett, T. E. Gallon, and M. Brutton, J. Phys. E Sci. Instrum. 5, 1008 (1972).

[58] P. J. Bassett, T. E. Gallon, and M. Brutton, J. Phys. E Sci. Instrum. 5, 1008 (1972).

[59] M. Huttula, S. Heinäsmäki, H. Aksela, E. Kukk, and S. Aksela, J. Electron Spectrosc. 156-158, 270 (2007).

[59] M. Huttula, S. Heinäsmäki, H. Aksela, E. Kukk, and S. Aksela, J. Electron Spectrosc. 156-158, 270 (2007).

[60] S. Urpelainen, Phd thesis: Instrumentation for spectroscopy and experimental studies of some atoms, molecules and custers, Oulu University Press, 2009.

[60] S. Urpelainen, Phd thesis: Instrumentation for spectroscopy and experimental studies of some atoms, molecules and custers, Oulu University Press, 2009.

[61] W. Stephens, Phys. Rev. 69, 691 (1946).

[61] W. Stephens, Phys. Rev. 69, 691 (1946).

[62] W. Wiley and I. McLaren, Rev. Sci. Instrum. 26, 1150 (1950).

[62] W. Wiley and I. McLaren, Rev. Sci. Instrum. 26, 1150 (1950).

[63] M. Huttula, M. Harkoma, E. N˜ommiste, and S. Aksela, Nucl. Instrum. Methods 467-468, 1514 (2001).

[63] M. Huttula, M. Harkoma, E. N˜ommiste, and S. Aksela, Nucl. Instrum. Methods 467-468, 1514 (2001).

54

54

[64] J. James and R. S. Sternberg, The design of optical spectrometers (Chapman and Hall Ltd, London, England, 1969), p. 73.

[64] J. James and R. S. Sternberg, The design of optical spectrometers (Chapman and Hall Ltd, London, England, 1969), p. 73.

[65] C. Palmer, Diffraction Grating Handbook fifth edition (Thermo RGL, New York, USA, 2002).

[65] C. Palmer, Diffraction Grating Handbook fifth edition (Thermo RGL, New York, USA, 2002).

[66] J. R. Janesick, Fundamentals of Scientific Charge-coupled Devices (SPIE - The International Society for Optical Enginering, Bellingham, USA, 2001), pp. 608–716.

[66] J. R. Janesick, Fundamentals of Scientific Charge-coupled Devices (SPIE - The International Society for Optical Enginering, Bellingham, USA, 2001), pp. 608–716.

[67] R. B. Bilhorn, J. V. Sweedler, P. M. Epperson, and M. B. Denton, Appl. Spectrosc. 41, 1114 (1987).

[67] R. B. Bilhorn, J. V. Sweedler, P. M. Epperson, and M. B. Denton, Appl. Spectrosc. 41, 1114 (1987).

[68] J. B. L. Poole, Nucl. Instrum. Methods 187, 241 (1981).

[68] J. B. L. Poole, Nucl. Instrum. Methods 187, 241 (1981).

[69] A. Hofmann, The Physics of Synchron Radiation (Cambridge University Press, Cambridge, England, 2004).

[69] A. Hofmann, The Physics of Synchron Radiation (Cambridge University Press, Cambridge, England, 2004).

[70] D. Attwood, Soft X-rays and extreme ultraviolet radiation Principles and Applications (Cambridge University Press, Cambridge, England, 1999).

[70] D. Attwood, Soft X-rays and extreme ultraviolet radiation Principles and Applications (Cambridge University Press, Cambridge, England, 1999).

[71] M. Bässler, A. Ausmees, M. Jurvansuu, R. Feifel, J.-O. Forsell, P. de Tarso Fonseca, A. Kivimäki, S. Sundin, S. Sorensen, R. Nyholm, O. Björneholm, S. Aksela, and S. Svensson, Nucl. Instrum. Meth. A 469, 382 (2001).

[71] M. Bässler, A. Ausmees, M. Jurvansuu, R. Feifel, J.-O. Forsell, P. de Tarso Fonseca, A. Kivimäki, S. Sundin, S. Sorensen, R. Nyholm, O. Björneholm, S. Aksela, and S. Svensson, Nucl. Instrum. Meth. A 469, 382 (2001).

[72] T. Balasubramanian, B. Jensen, S. Urpelainen, B. Sommarin, U. Johansson, M. Huttula, R. Sankari, E. N˜ommiste, S. Aksela, H. Aksela, and R. Nyholm, AIP Conf. Proc. 1234, 661 (2010).

[72] T. Balasubramanian, B. Jensen, S. Urpelainen, B. Sommarin, U. Johansson, M. Huttula, R. Sankari, E. N˜ommiste, S. Aksela, H. Aksela, and R. Nyholm, AIP Conf. Proc. 1234, 661 (2010).

[73] M. Sjöström, E. Wallén, M. Eriksson, and L.-J. Lindgren, Nucl. Instrum. Meth. A 601, 229 (2009).

[73] M. Sjöström, E. Wallén, M. Eriksson, and L.-J. Lindgren, Nucl. Instrum. Meth. A 601, 229 (2009).

[74] S. Urpelainen, M. Huttula, T. Balasubramanian, R. Sankari, P. Kovala, E. Kukk, E. N˜ommiste, S. Aksela, R. Nyholm, and H. Aksela, AIP Conf. Proc. 1234, 411 (2010).

[74] S. Urpelainen, M. Huttula, T. Balasubramanian, R. Sankari, P. Kovala, E. Kukk, E. N˜ommiste, S. Aksela, R. Nyholm, and H. Aksela, AIP Conf. Proc. 1234, 411 (2010).

[75] P. J. Bassett, T. E. Gallon, and M. Prutton, J. Phys. E Sci. Instrum. 5, 1008 (1972).

[75] P. J. Bassett, T. E. Gallon, and M. Prutton, J. Phys. E Sci. Instrum. 5, 1008 (1972).

[76] M. Jurvansuu, A. Kivimäki, and S. Aksela, Phys. Rev. A 64, 012502 (2001).

[76] M. Jurvansuu, A. Kivimäki, and S. Aksela, Phys. Rev. A 64, 012502 (2001).

[77] J. Jauhiainen, A. Ausmees, A. Kivimäki, S. Osborne, A. N. de Briton, S. Aksela, S. Svensson, and H. Aksela, J. Electron Spectrosc. 69, 181 (1994).

[77] J. Jauhiainen, A. Ausmees, A. Kivimäki, S. Osborne, A. N. de Briton, S. Aksela, S. Svensson, and H. Aksela, J. Electron Spectrosc. 69, 181 (1994).

55

55

[78] S. Hüfner, Photoelectron Spectroscopy Third Edition (Springer, Germany, 2003), p. 6.

[78] S. Hüfner, Photoelectron Spectroscopy Third Edition (Springer, Germany, 2003), p. 6.

[79] J. Végh, J. Electron Spectrosc. 151, 159 (2006).

[79] J. Végh, J. Electron Spectrosc. 151, 159 (2006).

[80] E. de Hoffmann and V. Stroobant, Mass Spectrometry Principles and Applications (John Wiley & Sons Ltd, New York, USA, 2007), p. 128.

[80] E. de Hoffmann and V. Stroobant, Mass Spectrometry Principles and Applications (John Wiley & Sons Ltd, New York, USA, 2007), p. 128.

[81] J. R. Taylor, An Introduction to Error Analysis (University Science Books, Sausalito, USA, 1997), p. 79.

[81] J. R. Taylor, An Introduction to Error Analysis (University Science Books, Sausalito, USA, 1997), p. 79.

[82] H. Tikkala, Master thesis: Valokaarianalyysiä optisen emission avulla, University of Oulu, 2012.

[82] H. Tikkala, Master thesis: Valokaarianalyysiä optisen emission avulla, University of Oulu, 2012.

[83] G. Wertheim, M. Butler, K. West, and D. Buchanan, Rev. Sci. Instrum. 45, 1369 (1974).

[83] G. Wertheim, M. Butler, K. West, and D. Buchanan, Rev. Sci. Instrum. 45, 1369 (1974).

[84] M. J. Klein, Arch. Hist. of Exact Sci. 1, 459 (1961).

[84] M. J. Klein, Arch. Hist. of Exact Sci. 1, 459 (1961).

[85] J. J. Yeh and I. Lindau, At. Data Nucl. Data Tables 32, 1 (1985).

[85] J. J. Yeh and I. Lindau, At. Data Nucl. Data Tables 32, 1 (1985).

[86] S. B. Whitfield, J. Phys. B 37, 3435 (2004).

[86] S. B. Whitfield, J. Phys. B 37, 3435 (2004).

[87] M. Huttula, M.-H. Mikkelä, M. Tchaplyquine, and O. Björneholm, J. Electron Spectrosc. 181, 145 (2010).

[87] M. Huttula, M.-H. Mikkelä, M. Tchaplyquine, and O. Björneholm, J. Electron Spectrosc. 181, 145 (2010).

[88] D. Janke, L. Savov, H.-J. Weddige, and E. Schulz, Mater. Technol. 34, 387 (2000).

[88] D. Janke, L. Savov, H.-J. Weddige, and E. Schulz, Mater. Technol. 34, 387 (2000).

56

56

Original papers

Original papers

The original papers have been reprinted with the permission of

The original papers have been reprinted with the permission of

Elsevier Science Publishing (www.elsevier.com)

Elsevier Science Publishing (www.elsevier.com)

Journal of Electron spectroscopy and Related Phenomena

Journal of Electron spectroscopy and Related Phenomena

IOP Science (iopscience.iop.org)

IOP Science (iopscience.iop.org)

New Journal of Physics

New Journal of Physics

The American Institute of Physics (www.aip.org)

The American Institute of Physics (www.aip.org)

Review of Scientific Instruments

Review of Scientific Instruments

The American Physical Society (publish.aps.org)

The American Physical Society (publish.aps.org)

Physical Review A

Physical Review A

Original publications are not included in the electronic version of the dissertation.

57

57

EXPERIMENTAL SPECTROSCOPIC STUDIES OF METALS WITH ELECTRON, ION, AND OPTICAL TECHNIQUES ARI MÄKINEN

ISBN 978-952-62-0313-3 ISBN 978-952-62-0314-9 (PDF) ISSN 1239-4327

REPORT SERIES IN PHYSICAL SCIENCES Report No. 87 (2013)