Dissociative Recombination of Astrochemically Interesting Ions

Dissociative Recombination of Astrochemically Interesting Ions

Dissociative Recombination of Astrochemically Interesting Ions

Mathias Hamberg

c Mathias Hamberg, Stockholm 2010

ISSN

ISBN 978-91-7447-089-5

Printed in Sweden by US-AB, Stockholm 2010 Distributor: Department of Physics, Stockholm University

To my family...

Abstract

In this thesis the major work described concerns experimental determination of the dissociative recombination (DR) reaction for several molecular ions of astrochemical interest. DR is the process where an electron recombines with a molecular ion to form an excited neutral that disintegrates into two or more neutral fragments to release the gained excess energy. It is very efficient under cold conditions and therefore ubiquitously occurring in interstellar environments such as dark clouds and plays an important role in aeronomical plasmae, lightnings and in man-made plasmas such as in combustion engines and fusion reactors. Although DR reactions are crucial processes in all these environments, product branching fractions of DR reactions have proven to be very unpredictable and present one of the great remaining challenges for theoreticians. The experimental work includes determination of reaction rates and product distribution of DR of complex ions such as protonated alcohols and ethers. The following species have been investigated and are discussed in this thesis:

CH3OH⁺₂ (protonated methanol), CD3OD⁺₂ (deuteronated methanol), CD3OCD2+ (methoxymethyl cation), CD3CDOD⁺ (deuteronated acetaldehyde), CH₃CH₂OH⁺₂ (protonated ethanol) and (CD₃)₂OD⁺ (deuteronated dimethyl ether).

The results of these measurements are used in astrochemical model calculations in which the rates used hitherto greatly have been based on educated guesses. Employing the outcome of the DR investigations of the CH₃OH⁺₂ and CD₃OD⁺₂ ions have shown a great impact on such models.

The DR investigations have been followed up by astronomical observations. Theoretical models and laboratory experiments show that methanol should be formed from CO on cold grains. This scenario was tested by astronomical observations of gas associated with young stellar objects (YSOs). Two independent tests were showing consistency with methanol formation on grain surfaces.

LIST OF INCLUDED PUBLICATIONS

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

I Experimental Studies of dissociative recombination processes for CD3CDOD⁺and CH3CH2OH⁺₂

M. Hamberg, V. Zhaunerchyk, E. Vigren, M. Kaminska, I.

Kashperka, M. Zhang, S. Trippel, F. Österdahl, M. af Ugglas, R.

D. Thomas, A. Källberg, A. Simonsson, A. Paál, J. Semaniak, M. Larsson, W. D. Geppert

Astronomy & Astrophysics, In preparation

II Experimental Studies of dissociative recombination processes for the dimethyl ether ions CD₃OCD⁺₂ and (CD₃)₂OD⁺ M. Hamberg, F. Österdahl, R. D. Thomas, V. Zhaunerchyk, E.

Vigren, M. Kaminska, M. af Ugglas, A. Källberg, A. Simonsson, A. Paál, M. Larsson, W. D. Geppert

Astronomy & Astrophysics, in print.

III Experimental studies on the dissociative recombination of H¹³CO⁺with electrons at energies between 2 - 50 000 meV M. Hamberg, I. Kashperka, F. Österdahl, M. Danielsson, E.

Vigren, R. D. Thomas, V. Zhaunerchyk, M. Kaminska, E.

Roueff, M. af Ugglas, A. Källberg, A. Simonsson, A. Paál, M.

Larsson, W. D. Geppert In preparation

IV Dissociative recombination of protonated methanol

W. D. Geppert, M. Hamberg, R. D. Thomas, F. Österdahl, F.

Hellberg, V. Zhaunerchyk, A. Ehlerding, T. J. Millar, H. Roberts, J. Semaniak, M. af Ugglas, A. Källberg, A. Simonsson, M.

Kaminska, M. Larsson

Faraday Discussions, 133 (Chemical Evolution of the Universe),

177-190, (2006)

V Observational tests of interstellar methanol formation

E. S. Wirström, W. D. Geppert, J. H. Black, Å. Hjalmarson, C.

M. Persson, P. Bergman, T. Millar, M. Hamberg, E. Vigren Astronomy & Astrophysics, in preparation

Reprints were made with permission from the publishers.

PUBLICATIONS NOT INCLUDED

VI Experimental determination of dissociative recombination reaction pathways and absolute reaction cross-sections of CH2OH⁺, CD2OD⁺ and CD2OD⁺₂

M. Hamberg, W. D. Geppert, R. D. Thomas, V. Zhaunerchyk, F. Österdahl, A. Ehlerding, M. Kaminska, J. Semaniak, M. af Ugglas, A. Källberg, A. Paál, A. Simonsson, M. Larsson

Molecular Physics, 105(5-7), 899-906 (2007).

VII Branching ratios and absolute cross sections of dissociative recombination processes of N₂O⁺

M. Hamberg, W. D. Geppert, S. Rosén, F. Hellberg, A.

Ehlerding, V. Zhaunerchyk, M. Kaminska, R. D. Thomas, M. af Ugglas, A. Källberg, A. Simonsson, A. Paál, M. Larsson

Physical Chemistry Chemical Physics, 7(8), 1664-1668, (2005).

VIII Dissociative recombination of protonated formic acid:

implications for molecular cloud and cometary chemistry E. Vigren, M. Hamberg, V. Zhaunerchyk, M. Kaminska, J. Semaniak, M. Larsson, R. D. Thomas, M. af Ugglas, I.

Kashperka, T. J. Millar, C. Walsh, H. Roberts, W. D. Geppert Astrophysical Journal, 709, 1429-1434, (2010)

IX Rotational state effect and fragmentation of small polyatomic molecular ions

V. Zhaunerchyk, W. D. Geppert, M. Hamberg, M. Kaminska, E. Vigren, A. Al-Khalili, S. Rosén, M. Danielsson, F. Österdahl, V. Bednarska, A. Petrignani, E. Bahati, M. E. Bannister, M.

R. Fogle, C. R. Vane, M. Larsson, W. J. van der Zande, R. D.

Thomas

Journal of Physics: Conference Series, 192, (2009)

X The dissociative recombination of protonated acrylonitrile, CH₂CHCNH⁺, with implications for the nitrile chemistry in

dark molecular clouds and the upper atmosphere of Titan E. Vigren, M. Hamberg, V. Zhaunerchyk, M. Kaminska, R. D.

Thomas, M. Larsson, T. J. Millar, C. Walsh, W. D. Geppert Astrophysical Journal, 695(1, Pt. 1), 317-324, (2009)

XI Sequential formation of the CH₃+H+H channel in the dissociative recombination of CH⁺₅

V. Zhaunerchyk, M. Kaminska, E. Vigren, M. Hamberg, W. D.

Geppert, M. Larsson, R. D. Thomas, J. Semaniak

Physical Review A: Atomic, Molecular, and Optical Physics, 79(3, Pt. A), 030701/1-030701/4, (2009)

XII Crossed-Beam and Theoretical Studies of the S(1D) + C₂H₂ Reaction

F. Leonori, R. Petrucci, N. Balucani, K. M. Hickson, M.

Hamberg, W. D. Geppert, P. Casavecchia, Piergiorgio, R.

Marzio

Journal of Physical Chemistry A, 113(16), 4330-4339, (2009)

XIII Investigation into the vibrational yield of OH products in the OH+H+H channel arising from the dissociative recombination of H₃O⁺

V. Zhaunerchyk, W. D. Geppert, Rosén, S., E. Vigren, M.

Hamberg, Kaminska, M., I. Kashperka, M. af Ugglas, J.

Semaniak, M. Larsson, R. D. Thomas

THE JOURNAL OF CHEMICAL PHYSICS, 130(21), (2009)

XIV Dissociative recombination of highly enriched para-H⁺₃ B. A. Tom, V. Zhaunerchyk, M. B. Wiczer, A. A. Mills, K. N.

Crabtree, M. Kaminska, W. D. Geppert, M. Hamberg, M. af Ugglas, E. Vigren, W. J. van der Zande, M. Larsson, R. D.

Thomas, B. J. McCall

Journal of Chemical Physics, 130(3), 031101/1-031101/4, (2009)

XV Results from the optical replica synthesizer at FLASH P. Salén, M. Hamberg, P. van der Meulen, M. Larsson, G.

Angelova-Hamberg, V. Ziemann, H. Schlarb, F. Löhl, E. Saldin, E. Schneidmiller, M. Yurkov, J. Bödewadt, A. Winter, S. Khan, A. Meseck

Proceedings of FEL2009, Liverpool, UK THOB02, 739-745,

2009

XVI Observation of two-dimensional longitudinal-transverse correlations in an electron beam by laser-electron interactions G. Angelova, V. Ziemann, A. Meseck, P. Salén, P. van der Meulen, M. Hamberg, M. Larsson, J. Bödewadt, S. Khan, A.

Winter, H. Schlarb, F. Löhl, E. Saldin, E. Schneidmiller, M.

Yurkov

Physical Review Special Topics–Accelerators and Beams, 11(7), (2008)

XVII Dissociative recombination of BH⁺₂: The dominance of two-body breakup and an understanding of the fragmentation

V. Zhaunerchyk, E. Vigren, W. D. Geppert, M. Hamberg, M.

Danielsson, M. Kaminska, M. Larsson, R. D. Thomas, E. Bahati, C. R. Vane

Physical Review A: Atomic, Molecular, and Optical Physics, 78(2, Pt. B), 024701/1-024701/4, (2008)

XVIII Dissociative recombination of D₃S⁺: product branching fractions and absolute cross sections

M. Kaminska, E. Vigren, V. Zhaunerchyk, W. D. Geppert, H.

Roberts, C. Walsh, T. J. Millar, M. Danielsson, M. Hamberg, R.

D. Thomas, M. Larsson, M. af Ugglas, J. Semaniak Astrophysical Journal, 681(2, Pt. 1), 1717-1724, (2008)

XIX Dissociative recombination of fully deuterated protonated acetonitrile, CD₃CND⁺: product branching fractions, absolute cross section and thermal rate coefficient

E. Vigren, M. Kaminska, M. Hamberg, V. Zhaunerchyk, R.

D. Thomas, M. Danielsson, J. Semaniak, P. U. Andersson, M.

Larsson, W. D. Geppert

Physical Chemistry Chemical Physics, 10(27), 4014-4019, (2008)

XX Formation of biomolecule precursors in space

W. D. Geppert, E. Vigren, M. Hamberg, V. Zhaunerchyk, R. D.

Thomas, M. Kaminska, T. J. Millar, J. Semaniak, H. Roberts, F.

Hellberg, F. Österdahl, A. Ehlerding, M. Larsson Journal of Physics: Conference Series, 88, (2007)

XXI Formation of biomolecule precursors in space?

W. D. Geppert, E. Vigren, M. Hamberg, V. Zhaunerchyk, M.

Kaminska, R. D. Thomas, F. Österdahl, Hellberg, Fredrik, A.

Ehlerding, M. Danielsson D., M. Larsson

In Proceedings IAU symposium No. 251, Organic Matter in Space, S. Kwok and S. Sandford eds, 349, (2008)

XXII Observational constrains on the formation of interstellar methanol

E. S. Wirström, C. Persson M., Å. Hjalmarsson, J. H. Black, P.

Bergman, W. D. Geppert, M. Hamberg, E. Vigren

In Proceedings IAU symposium No. 251, Organic Matter in Space, S. Kwok and S. Sandford eds, 143, (2008)

XXIII The cross-section and branching fractions for dissociative recombination of the diacetylene cation C₄D⁺₂

M. Danielsson, M. Hamberg, V. Zhaunerchyk, A. Ehlerding, M. Kaminska, F. Hellberg, R. D. Thomas, F. Österdahl, M. af Ugglas, A. Källberg, A. Simonsson, A. Paál, M. Larsson, W. D.

Geppert

International Journal of Mass Spectrometry, 273(3), 111-116, (2008)

XXIV Dissociative recombination of D₂H⁺: comparison between recent storage-ring results and theoretical calculations V. Zhaunerchyk, R. D. Thomas, W. D. Geppert, M. Hamberg, M. Kaminska, E. Vigren, M. Larsson

Physical Review A: Atomic, Molecular, and Optical Physics, 77(3, Pt. B), 034701/1-034701/4, (2008)

XXV Dissociative recombination of OPCl⁺ and OPCl⁺₂: Pushing the upper mass limit at CRYRING

V. Zhaunerchyk, R. D. Thomas, W. D. Geppert, M. Hamberg, M. Kaminska, E. Vigren, M. Larsson, A. J. Midey, A. A.

Viggiano

Journal of Chemical Physics, 128(13), 134308/1-134308/7, (2008)

XXVI Experimental investigation of electron impact on Si⁻₂

A. O. Lindahl, P. Andersson, G. F. Collins, D. Hanstorp, D.

J. Pegg, M. Danielsson, W. D. Geppert, M. Hamberg, R. D.

Thomas, V. Zhaunerchyk, C. Diehl, N. D. Gibson, A. Källberg Physical Review A: Atomic, Molecular, and Optical Physics, 77(2, Pt. A), 022710/1-022710/6, (2008)

XXVII Dissociative recombination study of N⁺₃: Cross section and branching fraction measurements

V. Zhaunerchyk, W. D. Geppert, E. Vigren, M. Hamberg, M. Danielsson, M. Larsson, R. D. Thomas, M. Kaminska, F.

Österdahl

Journal of Chemical Physics, 127(1), 014305/1-014305/5, (2007)

XXVIII Rotational State Effects in the Dissociative Recombination of H⁺₂

V. Zhaunerchyk, A. Al-Khalili, R. D. Thomas, W. D. Geppert, V. Bednarska, A. Petrignani, A. Ehlerding, M. Hamberg, M.

Larsson, S. Rosén, W. J. van der Zande

Physical Review Letters, 99(1), 013201/1-013201/4, (2007)

XXIX Dissociative recombination of the deuterated acetaldehyde ion, CD3CDO⁺: product branching fractions, absolute cross sections and thermal rate coefficient

E. Vigren, M. Kaminska, M. Hamberg, V. Zhaunerchyk, R.

D. Thomas, J. Semaniak, M. Danielsson, M. Larsson, W. D.

Geppert

Physical Chemistry Chemical Physics, 9(22), 2856-2861, (2007)

XXX Dissociative recombination branching ratios and their influence on interstellar clouds

W. D. Geppert, R. D. Thomas, A. Ehlerding, F. Hellberg, F.

Österdahl, M. Hamberg, J. Semaniak, V. Zhaunerchyk, M.

Kaminska, A. Källberg, A. Paál, M. Larsson

Journal of Physics: Conference Series, 4 26-31, (2005)

XXXI Status of the Optical Replica Synthesizer at FLASH

E. Saldin, H. Schlarb, B. Schmidt, E. Schneidmiller, M. Yurkov, J. Bodewadt, S. Khan, A. Winter, N. Javahiraly, M. Hamberg, P. van der Meulen, P. Salen, FYSIKUM, G. V. Angelova, V.

Ziemann, and A. Meseck

Conference proceedings, 22nd Particle Accelerator Conference, June 25- 29, USA, (2007)

XXXII Installation of the Optical Replica Synthesizer in FLASH G. Angelova, V. Ziemann, A. Meseck, M. Hamberg, P. Salen, P.

van der Meulen, M. Larsson, S. Khan, J. Bodewadt, A. Winter, E. Saldin, H. Schlarb, B. Schmidt, E. Scheidmiller, M. Yurkov Conference proceedings 29th International Free Electron Laser Conference, Budker INP, Novosibirsk, Russia, 26-31 August, (2007)

In preparation:

XXXIII Experimental Studies of dissociative recombination processes for the Bensen ions C₆D⁺₆ and C₆D⁺₇

M. Hamberg, E. Vigren, V. Zhaunerchyk, M. Kaminska, I.

Kashperka, S. Trippel, Z. Mingwu, M. af Ugglas, R. D. Thomas, A. Källberg, A. Simonsson, A. Paál, M. Larsson, W. D. Geppert Astronomy and Astrophysics, In preparation

XXXIV Dissociative recombination of the acetaldehyde cation, CH3CHO⁺

E. Vigren, M. Hamberg, V. Zhaunerchyk, M. Kaminska, R. D.

Thomas, S. Trippel, Z. Mingwu, I. Kashperka, M. af Ugglas, C.

Walsh, R. Wester, M. Larsson, W. D. Geppert Physical Chemistry Chemical Physics, In preparation

XXXV Crossed-beam and theoretical studies of the N+C₂H₄ reaction

F. Leonori, R. Petrucci, N. Balucani, K. M. Hickson, M.

Hamberg, W. D. Geppert, M. Rosi, P. Casavecchia The Journal of Physical Chemistry A. in preparation

Authors contribution

The five years I have spent on my doctoral studies has resulted in a multitude of publications. The amount and type of work that I have been putting in differ substantially in between the different articles.

For the articles presented in this thesis, paper I-II I have taken the leading role in both the preparation and carrying-out of the experimental work, the data reduction and writing of the papers. This was also the case for paper III with the exception of the imaging part of which I did not do the analysis and write-up.

For paper IV, I was participating in the experiment, and took part in the analysis work and article writing. I spent two weeks with Professor Tom Millar (Queen’s University Belfast) performing astrochemical modeling of dark clouds and comparing the results from molecular calculations including the results our experiments with previous observations.

For paper V I took part in the astronomical observations at the 20 m radio telescope at Onsala observatory. I was personally taking part in the observations for several weeks and during some time solely responsible for the running of the telescope.

For VI-VII paper I have taken an active role in the experimental work, did the writing of the papers and the analysis of the data except for in paper VII where the data reduction for the CD4O⁺ ion was performed by my supervisor, W. D. Geppert.

During my doctoral studies I was also responsible for the ion source of our experiments. Furthermore I have been in charge of the ion implanted detector measurement system, including detector maintenance. Therefore I have also been actively preparing the experiment for many of the papers where I am not first author.

I was also involved in several studies that are not included in this thesis:

I was taking part in investigations of photo-detachment of negative ions in a cold trap. This was done in collaboration with Roland Wester’s group during two weeks in Freiburg, Germany. Other work that is not described in this thesis

encompasses crossed beams experiments carried out together with Professor Piergiorgio Casavecchia during seven weeks in Perugia, Italy. I also spent two weeks at the group of Eric Herbst, Columbus Ohio, USA, doing astrochemical modeling. Furthermore, I worked two months in DESY, Hamburg, Germany participating in the build- up and experiments with the optical replica synthe- sizer (ORS) in Hamburg, Germany. During one week I performed observa- tions of phosphor-containing interstellar molecules using the 12 m Kitt Peak observatory in Arizona, USA.

Contents

1 Introduction . . . . 1

1.1 The interstellar medium . . . . 1

1.1.1 Diffuse clouds. . . . 1

1.1.2 Dense clouds . . . . 2

1.1.3 Planetary and satellite atmospheres. . . . 3

1.1.4 Model calculations. . . . 4

1.1.5 Surface processes. . . . 4

1.1.6 Gas processes . . . . 5

1.2 Ion electron reactions . . . . 5

1.3 Dissociative recombination and its two mechanisms. . . . 6

1.3.1 Direct process. . . . 7

1.3.2 Indirect process. . . . 7

1.4 Alternative DR measurement techniques . . . . 9

1.4.1 Stationary afterglow. . . . 9

1.4.2 Flowing afterglow. . . . 9

1.4.3 Merged crossed and inclined beams . . . . 10

1.4.4 Electrostatic storage rings. . . . 10

2 Experimental technique . . . . 13

2.1 CRYRING . . . . 13

2.1.1 Ion Production. . . . 14

2.1.2 Electron Cooler. . . . 15

2.1.3 Ion current measurement . . . . 18

2.1.4 0-degree arm and particle detector system . . . . 18

2.1.5 Imaging detection system. . . . 21

2.2 Onsala 20 m telescope . . . . 23

2.3 APEX . . . . 23

3 Data evaluation . . . . 25

3.0.1 Particle energies and velocities . . . . 25

3.0.2 Space charge correction. . . . 26

3.1 Thermal rate constant and cross section. . . . 27

3.1.1 Toroidal correction. . . . 29

3.1.2 Correction for electron energy spread. . . . 29

3.1.3 Thermal rate constant . . . . 30

3.1.4 Error estimation. . . . 31

3.1.5 Summary of the analysis procedure . . . . 31

3.2 Branching fractions . . . . 34

3.2.1 Fragment identification. . . . 36

3.2.2 Matrix formation . . . . 37

3.2.3 Loss compensation . . . . 38

3.2.4 Error evaluation. . . . 40

3.2.5 Toroidal effect correction. . . . 40

4 Results . . . . 41

4.1 Paper I. . . . 41

4.2 Paper II . . . . 47

4.3 Paper III. . . . 49

4.4 Paper IV and V. . . . 55

5 Acknowledgments . . . . 61

A Monte Carlo simulations . . . . 65

A.0.1 Two-body breakup . . . . 65

A.0.2 Three-body breakup. . . . 66

A.0.3 Adapting the results to the lab frame . . . . 67

Bibliography . . . . 69

1. Introduction

This thesis is focusing on dissociative recombination of molecular ions, a reaction known to be among the dominating ion processes found in plasmas cold enough to contain molecular constituents [1]. Plasma denotes a collectively behaving gas state in which a non negligible fraction of the particles are ionized. In our terrestrial environment plasmas can be found in as different environments as combustion processes, industrial etching plasmas, lightning, fluorescent light tubes and fusion reactors which may sound exceptional. However, 99% of the matter in the universe is often said to consist of plasma and even though this estimate may not be accurate a quick glimpse up in the sky a clear night will show us myriads of plasma-containing stars and nebulae indicating that this claim might be well-founded [2].Our experimental investigations aim for a bettter understanding of plasma processes in the interstellar medium as well as star forming regions.

1.1 The interstellar medium

The interstellar medium (ISM) is the gas and dust that pervades the space between stars and their planetary systems within a galaxy. It is made up by 1%

dust and 99% gas (practically plasma) [3]. Molecules can form in the colder parts of the medium consisting of diffuse or dense clouds. Currently more than 150 different molecular species has been detected in the interstellar and circumstellar media and approximately 50 of them contain 6 or more atoms [4].

1.1.1 Diffuse clouds

One type of interstellar object is diffuse clouds [5,6]. They are generally hard to spot for the untrained eye since their visible and UV absorption by dust and gas is weak and hence the ionization rate is higher in these objects than in dark clouds. The UV radiation from surrounding stars is massive. Ionization effi- ciency for species with a lower ionization potential than hydrogen is very high and therefore, e. g., virtually all atomic carbon is ionized. Also, cosmic rays contribute to ionization. Typical temperatures in diffuse clouds range around 50-100 K and the total gas densities are∼10² cm⁻³. In such environments several molecular ions e.g. HCO⁺and HOC⁺have been detected [7].

1.1.2 Dense clouds

Also known as molecular clouds or dark clouds [8,5], they are normally as- sociated with giant molecular clouds (GMC). These huge and often irregular assemblies of gas and dust can have masses ranging up to∼10⁵m_⊙(m_⊙ = 1 solar mass) and are frequently the birthplace of new stars. Due to their irreg- ularity their appearance can vary from translucent to virtually black objects called cores. The clouds consist mainly of H₂molecules and ice-covered dust particles [9, 10]. It should be noted that distinction between the categories dark and translucent clouds has not been uniformly consistent in the literature [8]

TMC-1 is one of the best studied interstellar cloud cores [4]. This type of cloud typically has temperatures of∼10 K and densities ∼10⁴ - 10⁶ cm⁻³. Large molecular species such as HC₇Nand C₅Nhave been found in this en- vironment that show a tendency of producing “exotic” unsaturated hydrogen poor species [11,12,4]. The physical conditions are thought to be very homo- geneous especially since no stars are embedded in this object, although lately observations have showed that TMC-1 contains at least six smaller units / clumps with distinct chemical properties [13].

Similar to the dark cores (in respect to temperature, density and composi- tion) are the densest part of Bok globuli which are small roundish isolated gravitationally bound molecular clouds. They typically have the size of a light year across and have masses of about 10-50 m_⊙. Star formation may take place with typically one to two stars per globule.

As the core evolves with time, star formation plays an increasingly impor- tant role. Pre-stellar cores are formed, in which the temperature still remains low (∼10 K), although a collapse process has started creating a central region with densities of 10⁵⁻⁷ cm⁻³. A typical object in this class is L1544 where ite has been observed that some species are depleted toward the center of the condensation whereas others are not [13].

“Hot cores” are molecular condensations directly associated with young embedded massive stars. Subsequently the term “hot corino” was introduced by Ceccarelli et al. [14] in order to further distinguish the warm inner re- gions of the envelope surrounding the low-mass protostars due to the signif- icant different amount of material involved (and therefore their unlike evolu- tion timescale) [14]. Herein, complex molecules are thought to be produced through two major steps. 1) Hydrogenation of simple molecules such as CO on grain surfaces during the cold pre-collapse cloud phase forms the so called

“parent“ or “first generation molecules” (e.g. CH₃OH). 2) Heating of the newly formed star will cause evaporation of the “parental” molecules that in the gas-phase quickly will undergo further reactions to form “daughter or second generation” molecules. Here the temperatures are 100-300 K and the densities could be around 10⁸cm⁻³depending on the particular core [15,13].

Many complex species are found in these regions such as e.g. dimethyl ether

and ethanol [14,5,16,4]. It is worth noting that in the outer envelope, where the dust temperature is lower and sublimation temperatures are not reached for the dust particles, the chemistry is basically identical to pre-stellar cores [15].

Protoplanetary disks are formed around new-born stars and represent a spe- cial environment. Observations have been made toward the young AA Tauri star that is believed to be similar to hot cores with temperatures over the sub- limation limit of ices. The abundances of H₂Oare about one order of magni- tude higher than what models and observations of hot cores have shown [17].

Analogously, one might therefore believe that this could indicate a molecular synthesis in the AA Tauri where more complex molecules might be produced within the disk. However no complex molecules (including CH3OH) have yet been found in protoplanetary disks. Observations are so far limited to small molecules since the disk sizes and masses are relatively low [4].

Outflows of young stars make an impact on its surrounding molecular cloud with shock waves causing material frozen onto dust grains to sputter off or to be destroyed. The temperatures could go up to a few thousand Kelvin in the shock which could open up endoergic reaction pathways implying a richer chemistry [4].

In the galactic centre molecular clouds exist with characteristics that differ from the ones found in the disk region. Typical temperatures are 50-120 K and densities are in the order of 10⁴cm⁻³ [18]. Also for this cloud type complex molecules such as dimethyl ether (CH₃OCH₃) and ethanol (CH₃CH₂OH) are found. A difference of these objects from hot cores that should be pointed out is that the kinetic temperatures are high (>100 K), whereas the excitation temperatures only range around 10-20 K because of the low H₂densities. Also the dust shows temperatures below 20 K. Generally, the environment close to the centre of the galaxy is quite extreme with high levels of UV-radiation, X- rays, cosmic rays, and the saturation (hydrogenation) of molecules is higher than in hot cores found in the disk [19].

1.1.3 Planetary and satellite atmospheres

Planet and moon atmospheres can have a complex chemistry [20] and their surface pressure, temperatures as well as the intensity of incoming UV radia- tion leading to ionization of their molecules span over a wide range.

The largest of Saturn satellite, Titan, is the only moon known to have a dense atmosphere, which is mostly composed of nitrogen and also contains methane and ethane clouds. Titan´s atmosphere contain nitriles and hydrocar- bon compounds e.g. CO, HCN, HC₃N, CH₃CN and H₂O. In the Martian and Venusian atmosphere, several species such as CO₂, CO, O₂, O₃ and H₂O₂ have been detected. In the latter atmosphere also sulfur and chlorine com- pounds exist. The atmosphere of the giant planets Jupiter and Saturn contain

species such as PH₃, NH₃, H₂O, CO and CO₂ but also large amounts of H₂ that cannot escape due to the massive gravity.

Tenuous atmospheres could be found on e.g. Io (Jupiter moon) and Ence- ladus (moon of Saturn). Also Pluto and Triton (satellite of Neptune) have at- mospheres due to sublimination of volatile ices including N2, CH4 and CO.

The surface pressures in these atmospheres are∼10 µbar. Io (Jupiter moon) have a spatially heterogeneous atmosphere dominated by SO₂and with traces of SO, NaCl and S₂. The pressure∼nanobars is varying with the diurnal cycle and comes from sublimation and volcanic activity. Cryovulcanic plumes have been discovered at the south pole of Saturn’s moon Enceladus containing CO, N₂and CH₄.

1.1.4 Model calculations

Chemical evolution in different environments is modeled with numerical com- puter calculations. Depending on the environment different parameters (tem- perature, initial composition, density etc.) need to be used as input, e.g. in modeling of cold dark clouds [21] and star forming regions [16]. Feasible processes are restricted to barrier-less exoergic reactions in these cold envi- ronments and, thus, radical-and ion induced reactions reactions dominate the chemistry [4]. Mainly, two different computer models are used for simulation of the intricate networks of reactions: the Rate06 / UMIST model (originally developed at University of Manchester Institute of Science and Technology) and the OSU model (developed at Ohio State University) [4]. These models currently involve about 5000 coupled reactions. In a paper from 2007 describ- ing the UMIST model it is stated that 420 species and more than 4500 reac- tion pathways are included[21]. Some of these processes can be investigated by with ab-initio calculations. However, this do not usually apply to DR reac- tions due to their unpredictable behavior and experiments are heavily needed [5,22]. Of the approximately 500 DR-reactions included (at this stage) in the UMIST model, only about 100 have been experimentally determined [21].

Branching ratios of some reactions e.g. the DR of CH3OH⁺₂ and HCS⁺have been shown to have a crucial influence on the predictions of models [22]. The conclusion within the astrochemistry community has been that grain-surface reactions should be considered to get better consistency of model predictions with the observed data. [23,24,25,26].

1.1.5 Surface processes

Formation of molecules can take different pathways. They can be synthesized in the gas phase while others are believed to be generated by surface pro- cesses. Dust grains that consist mainly of silicates and carbonaceous material provide the reactive areas [27]. Their size ranges from∼10 nm to over ∼1µm [28].The general opinion is that grains are mainly negatively charged due to

the rapid movement of electrons in plasmas compared to heavy positive ions [5]. At low temperature these grains are covered by ices whose composition depends on the temperature and the chemical environment. When star forma- tion starts e.g. in a Bok globule, the temperature will slowly increase in the star forming region and hence the mobility of the different molecules will in- crease until they finally evaporate. The enhanced mobility also augments the number of possible reactions. Light species like H₂ can easily move around in the porous ice while larger molecules are more restricted in their move- ment. The porosity itself decreases with rising temperature and experiments on amorphous water surfaces shows a dramatic reduction of porosity at∼45 K [29]. Ion processes can also occur on grain surfaces. Further and more de- tailed studies are needed and, as Watanabe writes, “quantitative information on surface processes such as reaction channels, reaction rates and activation energies has yet to be obtained”[30].

1.1.6 Gas processes

As mentioned 99% of the interstellar medium is in the plasma state in which the degree of ionization may differ significantly. Even though surface pro- cess play an important part of the chemistry it is obvious that it is in the gas phase the bulk of material is found. Clearly, a precise determination of the ongoing gas-phase processes is important. As mentioned, barrier-less ex- oergic processes form the backbone of the chemistry in the cold interstellar medium [4]. Processes occuring there are e.g. ion-electron, ion-neutral and radical-neutral reactions. For molecular cations in the interstellar medium ion- electron reactions are the most important destruction processes, since elec- trons are much more mobile than competing anions. Many interstellar pro- cesses could be characterized using theoretical methods while others request experimental studies. For most ion-electron interactions experimental investi- gations are necessary.

1.2 Ion electron reactions

When a molecular ion ABC⁺reacts with an electron several types of processes can occur. The course of reactions depends on the rotational, vibrational as well as kinetic temperature of the reactants. Other factors are the reaction en- thalpy, the shape of potential surfaces and the possibility of tunneling effects.

The following ion electron processes are often discussed in the literature:

i Scattering processes: ABC⁺ + e⁻(v₀) → ABC⁺ + e⁻(v₁) The neutral intermediate formed by the recombination auto ionizes ("the electron bounces back out"). If the velocity of the emitted electron is the same as the incoming electron, elastic scattering occurs. If the velocity of the

emitted electron is higher or lower the process is called “super elastic scattering” and “inelastic scattering”, respectively.

ii Radiative recombination: ABC⁺+ e⁻→ ABC + hν. The neutral inter- mediate loses the energy gained through the recombination by emitting one (or possibly several) photons. This process is relatively slow (typi- cally 10⁻¹²cm³s⁻¹ [31]) and thus the probability for it to occur at low temperatures is much less than for dissociative recombination. No con- clusive evidence for radiative recombination for molecular ions has so far been found in storage ring experiments. However, it is an important reaction for atomic ions.

iii Dissociative ionization (DI). ABC⁺+ e⁻→ A⁺+ BC⁺ + 2e⁻. Both the resulting fragments are ionic. This process is more common for higher interaction energies since it is highly endoergic.

iv Dissociative excitation (DE). ABC⁺ + e⁻ → A⁺ + BC +e⁻. A neutral and an ion fragment is formed. Such processes are also endoergic.

v Resonant ion pair formation (RIP). ABC⁺+ e⁻→ A⁺+ BC⁻. Two ions are created. This process can compete with DR at lower energies.

vi Dissociative recombination (DR): ABC⁺ + e⁻→ ABC^∗∗→ A + BC + KER (where KER is the kinetic energy released in the reaction). The electron attach to the ion and forms a highly excited neutral which subsequently undergoes fragmentation. This reaction does not normally have any energy barriers.

1.3 Dissociative recombination and its two mechanisms

Since DR normally does not have any energy barrier for activation it will readily occur in any plasma cold enough to contain molecular constituents.

The cross section for the RIP and DE is normally much smaller than for DR [32]. Therefore DR is one of the dominating processes in naturally occurring plasmas such as those encountered in interstellar clouds, comet comae, au- rorae and aeronomical processes as well as man-made plasmas found in e.g.

combustion processes and fusion reactors [1,33,34,35,36,37,38,39,40,41, 42,43]. In the formation processes of many complex molecules in interstellar environments it is often regarded as the terminal step.

One of my main focuses in this thesis have been determining the reaction rate constants of the DR process at different temperatures. Consider a molec- ular ion ABC⁺ where A, B and C denotes different atoms. Depending on relative kinetic energy (between the electron and ion) and the structure of the ion the outcome of the process may be e.g.

ABC⁺+ e⁻→ A + BC + KER¹ (1.1)

AB+C + KER2 (1.2)

AC+ B + KER3 (1.3)

A+ B +C + KER₄ (1.4)

Investigation of the branching fractions i.e. the distribution between the dif- ferent products have also been a major focus of the work described here. For astrochemical applications this product distribution is normally of high inter- est since DR may be the dominating production pathway of important inter- stellar molecules. In some case, the investigations yielded the unfeasibility of a reaction sequence including DR to form a particular molecule in the ISM. As we could show the formation of methanol from the protonated form by DR is only occurring for 3% of the reactions [22] and thus too inefficent to produce the observed abundances of this substance. Two different main mechanisms of DR have been invoked:

1.3.1 Direct process

In this process the incoming electron is captured creating a doubly excited state of the formed neutral. The ion potential curve is crossed by the repulsive potential of the neutral state (see fig.1.3.2), along which dissociation occurs.

The DR process has to compete with autoionisation occurring on a femtosec- ond scale and it therefore needs to be very fast in order to occur. The reaction cross section follows an 1/E dependence (where E is the interaction energy).

This reaction mechanism was proposed by Bates in 1950¹[44].

1.3.2 Indirect process

In the indirect process of DR, the electron attaches to the ion generating a rovi- brationally excited Rydberg state (highly singly excited electronic state). The Rydberg state can then be predissociated through coupling to a repulsive dou- bly excited neutral state (see fig.1.3.2). Since a bound rovibrational Rydberg state is formed only at certain discrete energy intervals for the ion electron interaction this process sometimes manifests itself through resonances (struc- tures) in the cross section. This reaction pathway was independently proposed in 1967 by J. N. Bardsley [45] and J. C. Y. Chen and M. H. Mittleman [46].

Also herein autoionisation of the intermediate Rydberg state is possible.

1The science about DR has a rather long and fascinating history and a comprehensive descrip- tion is found in the book by Larsson & Orel [31].

Figure 1.1: Schematic potential curves relevant for the direct mechanism of DR reac- tions.

Figure 1.2: Schematic potential curves relevant for the indirect mechanism of DR reactions.

1.4 Alternative DR measurement techniques

The articles included in this thesis are based on experiments made with the heavy ion-storage ring CRYRING in Stockholm, Sweden and will be de- scribed in detail in the next chapter. The outcome of our experiments has been compared to previous results from alternative measurement methods. Here I would like to give a short introduction into these methods.

1.4.1 Stationary afterglow

The stationary afterglow technique was the first method used to study DR [47].

The principle of this method is introduce gas into a cavity, create a plasma with ions of interest using e.g. a magnetron and subsequently turn off the plasma inducing device. The free electrons in the plasma will then recombine with the ions. Instead of measuring the actual products the method relies on prob- ing the electron density decay by low power microwave field techniques (or a Langmuir probe [48]). By measuring the resonant frequency of the cavity at various times in the afterglow, the variations of the electron density could be determined and therefore also the DR-rates (see e.g. [47,49,31] for a more detailed description). An advantage of this technique is that the ions can eas- ily become thermal obtaining the temperature of the surrounding walls hence T_cavity=T_gas=T_ions=T_e. Therefore, it is possible to successfully probe different temperatures by heating/cooling the surrounding cavity to different tempera- tures prior measuring of the rates [50]. Other advantages of this method is that it has no restrictions in the mass of the ions and the excitation of the products can be found by investigating the radiation of the decaying plasma. Draw- backs of this technique are that there may be excited ions in the interaction region. Another disadvantage is the fact that more than one ion might be pro- duced in the afterglow and the recombination observed is a sum of different recombination processes.

1.4.2 Flowing afterglow

This method represent a further development of the previously described method. A discharge ignites a plasma which is lead out through a tube by means of pressure difference (more detailed descriptions are found elsewhere e.g. [51,52,53,54,55]). Depending on the ion of interest different gases can be injected, in some cases the ion is created by subsequent reactions of the primary ion inside the flow tube. The electron density can be measured by a movable Langmuir prove inside the tube. This method is called Flowing Afterglow Langmuir Probe (FALP). In the end of the flow tube a mass spectrometer is mounted which however may not give a correct representation of the ion composition in the middle of the plasma [31]. Therefore, an alternative configuration with a movable mass spectrometer was constructed [53]. Other techniques that have been combined with the FALP-technique are

UV-absorption, laser induced fluorescence and optical emission spectroscopy techniques [56, 57]. Information about the DR product distribution can be achieved if using these methods, although it has been shown to be difficult and full elucidation about DR branching ratios has rarely been accomplished.

Nevertheless, partial information such the branching fraction of a specific product e.g. OH in the case of DR of H3O⁺, HCO⁺₂ and N2OH⁺ has been obtained [58].

1.4.3 Merged crossed and inclined beams

A total different approach compared to the after glow techniques is to use crossed ion and electron beams. Depending on the angle the technique could be described as crossed, inclined or merged. If the relative ion-electron angle differs from zero it is impossible to reach 0 eV relative kinetic energy. This is why crossed and inclined beam methods predominantly have been used for investigation of high energy reactions. However, in the merged beam tech- nique this angle is zero and changing of the relative energy (and therefore temperature dependence) over a large interval is easy (see e.g. [59] for a de- tailed description). With this method it is also possible to detect and identify the fragments produced in the DR-reaction and the product distribution [60].

A significant drawback of this method is that excited states of probe ions are likely to be present and are difficult to control. It is also difficult to find the product distribution for higher ion masses due to limitations in the detector systems associated with the setup. Another disadvantage is that it measure- ments take relatively long time.

1.4.4 Electrostatic storage rings

Relatively recently electrostatic storage rings has become a tool for study of the DR-reaction. The two main advantages of using storage devices for per- forming experiments is that most molecular ions can be cooled to the lowest vibrational ground state and that merging of an electron beam makes it possi- ble to reach very low collision energies and with a very high resolution [61].

So far the only electrostatic storage ring with an implemented electron target (to enable studies of electron-ion reactions) and a possibility to reach electron- ion interaction energies close to zero eV (and therefore matching interstellar conditions) is the apparatus at KEK in Japan which has been used for study- ing e.g. H⁺₂ [62]. This type of experiment also opened up for measurement relatively complex and heavy biomolecules such as protonated Arginine and Bradykinin (molar mass 175 and 1061 gmol⁻¹ respectively). However such molecules are to heavy for matching electron-ion velocities and was therefore only studied at higher collision energies. This problem could be illustrated by the electron energies at the highest neutral rates are only 5.45, 2.72 and 1.82 eV for 20 keV storage energy of H⁺₂, D⁺₂ and D⁺₃, respectively and is a severe

limitation of electrostatic devices [62]. Another problem has been shown to be the detector resolution, making it virtually impossible to resolve the differ- ent products with standard energy sensitive silicon detector techniques. New detector types such as micro-calorimeter devices has been showing highly re- solved results in the storage energy range of 5-60 MeV which however still are far above the maximum energies in electrostatic storage rings [63].

The Heidelberg Cryogenic Storage Ring (CSR) is an electrostatic appara- tus currently under construction. It is aiming to store ion beams with kinetic energies between 20 to 300 keV. The machine is expected to operate at as low temperatures as∼2 K and should therefore be able provide fast beams of molecular ions in a single quantum state [64]. This will open up new aspects of the DR science.

2. Experimental technique

The experiments were made at the heavy-ion storage ring CRYRING in Stock- holm, Sweden.

2.1 CRYRING

CRYRING (CRYsis¹ connected to a storage RING) is a heavy-ion storage ring situated at the Manne Siegbahn laboratory in Stockholm, Sweden (see fig.

2.1). It is built up as a regular dodecagon² and therefore has twelve equally long sections of Ø100 mm steel tubes. The perimeter of the ringlike structure is 51.63 m. CRYRING is a synchrotron which means that the acceleration of the ions is synchronized with the magnetic field of the bending magnets sit- ting in between each straight section. Six of these sections have focusing mag- netic elements such as quadropoles and sextupoles whereas the other six are used for injection, acceleration, diagnostics/measurements and electron cool- ing. The ions are produced in a hollow-cathode ion source (see description below) and are then accelerated by∼40 kV into a bending magnet used for mass selection. To ensure efficient pressure reduction between the ion source and the storage ring the ions pass a comparatively long distance (more than 10 m) before they are injected into the ring. The injection is performed through a ten times multi turn procedure utilizing four pairs of electrostatic kicker plates that create a local closed-orbit bump on the injection section. This bump is re- duced for each turn which in conjunction with betatron oscillations of the al- ready injected ions ensures an efficient filling [65]. The ions are subsequently accelerated in a 2.7 m long driven drift tube RF system [66] and the maximum energy gained after acceleration is 96(q/A)² MeV u⁻¹ [65] which is accom- plished after∼1 s. This energy is limited by maximum magnetic field in the dipole magnets (B=1.2 T) and the bending radius (ρ=1.2 m) which together determine the magnetic rigidity of Bρ=1.44 Tm. The pressure in the ring it- self is in the order of∼10⁻¹¹Torr which minimizes the rest gas collisions and ensures ion lifetimes of several seconds. Therefore there is plenty of time for excited states in infrared-active vibrational modes to cool off radiatively by spontaneous photon emission. This normally occurs within a timescale of 0.1- 1 s [67]. When sufficient time for cooling has passed, the neutral fragments

1CRYSIS stands for CRYogenic Stockholm Ion Source. This type of ion source was not used during the experiments described in this thesis.

2A dodecagon is a twelve segmented polygon

Figure 2.1: The heavy ion storage ring CRYRING. Seeing the ring as a clock we can identify the injection at 12 o’clock; the RF-system for acceleration at ten o’clock; the electron cooler six o’clock. The zero degree arm is leading tangentially out of the electron cooler and it is also here the IID is mounted. The ion source is located in the right room just before the last bending magnet.

emerging from the DR-reactions in the electron cooler are detected by an ion implanted detector (IID).

2.1.1 Ion Production

During the experiments the most commonly used ion source was of the hollow cathode type (see fig. 2.2). Gases are injected (either directly from gas tubes or evaporated from fluids) in a pulse of∼30-100 ms. Thereafter a discharge is created in the ion source with a voltage of∼300-2000 V over a similar timescale creating the plasma forming the ions. The time between two injections (i.e. the ring cycle) is normally 5-10 s. Having the ion source working in a pulsed mode serves to significantly reduce the following problems: high sample consumption, ion source heating, wear and soot formation. Therefore no water cooling of the ion source is necessary and observed operation times of these sources usually are in the order of a week.

Normal working pressures measured∼1.5 m after the source is ∼10⁻⁵ Torr before gas injection (base pressure) and ∼5-50×10⁻⁵ Torr at the peaks of the sample gas flow. This relatively high peak pressure enables collisional quenching, which is particularly important for ions excited in IR-inactive modes that cannot radiatively cool. The base pressure could be adjusted by having a second gas continuously flowing into the source. However, this may also affect ionization processes which could be electron impact, protonation,

Table 2.1: Ion and precursor gases.

Ion Gas1 Gas2

H¹³CO^{+ 13}CO H₂

H¹³CO⁺ H¹³COOH H₂ CD₃CDOD⁺ CD₃CD₂OD D₂ CH3CH2OH⁺₂ CH3CH2OH H2

CD3OCD⁺₂ CD3OCD3 D2

(CD₃)₂OD⁺ CD₃OCD₃ D₂ CH₃OH⁺₂ CH₃OH H₂ CD₃OD⁺₂ CD₃OD D₂

penning ionization and ion impact ionization. The ions presented in this thesis were produced from the gas mixtures in table2.1.

2.1.2 Electron Cooler

Electron cooling denotes a technique used to reduce the momentum spread of stored particle beams. Ions moving in the storage ring do not have iden- tical velocities. This spread of momentum can be seen as a thermal motion.

The electron cooler (see fig2.3) is mounted in one of the straight sections in CRYRING and essentially generates a cold electron beam that is merged with the ion beam. In total it serves three purposes [67,68,69,70,71,72,73]:

• Increase of the phase space density of the stored ions, i.e. narrowing the energy distribution of the ions.

• Cooling of long lived excited internal degrees of freedom including infrared inactive modes with super-elastic scattering process [67].

• Providing an interaction region between ions and electrons during DR data collection.

In the electron cooler electrons are emitted from a small (Ø4 mm) cathode by an electron gun. The temperature of the cathode is in the range of 900^◦C which leaves the electrons with an energy of∼0.1 eV. The velocity distribu- tion can therefore be described by an isotropic Maxwellian distribution:

Figure 2.2: JIMIS hollow cathode type ion source

f(v_x, v_y, v_z) =

 m_e 2πkBT

3/2

e⁻

me(v2x+v2 y +v2

z )

2kBT (2.1)

where vx−zis the velocity in respective direction, meis the electron mass and k_BBoltzmann’s constant. However, due to the acceleration of the electrons the velocity spread in the longitudinal direction parallel with the electron beam is drastically reduced to (T_k)∼0.1 meV [74].

The cathode is surrounded by an axial magnetic field of 3 T created by a su- perconducting solenoid. This field is gradually diminishes by a factor of 100 to 0.03 T toward the rest of the electron cooler [72]. In the process the electron beam adiabatically expands from Ø4 mm to Ø40 mm and since the ratio of the transversal kinetic energy toward the magnetic field strength (E_⊥/B) re- mains constant it follows that E_⊥ is reduced by a factor of 100 as well [73].

The expected transverse energy spread is therefore (T_⊥) ∼2 meV when the electron and ion beams are well aligned. The electron velocity distribution is anisotropic and can be described as:

f(v_⊥, v_k) = me

2πk_BT_⊥

 me

2πkBT_k

1/2

e⁻

mev2⊥

2kBT⊥−_2kBTk^mev2k

(2.2) where v_⊥²= v_x²+v_y²and v_k= v_z.

Figure 2.3: Electron cooler: The solid arrow depicts the ion beam while the dashed line show the parallel interaction region. Note also the two bending magnets bending in and out the electron beam giving rise to the toroidal sections where the interaction energy is higher.

Since the resulting electron beam is not only cold but constantly renewed the ion beam can lose part of its heat (temporal spread) due to Coulombic drag forces between the electrons and the ions. This process is more efficient the lower the mass-to-charge ratio is. Therefore the effect of the phase space cooling is limited for the ions investigated in this thesis. Even so the velocity spread of the ions is approximately vi×10⁻³and its contribution to the energy resolution is negligible [32,75].

Two dipole bending magnets are located in the cooler that guide the electron beam in and out from the ∼ 85 cm long interaction region of before it is collected. The magnetic fields in these regions are to weak to affect the ion pathway but will have an influence on the ion-electron impact energy since the trajectory angles there are not zero. The electron current Ie, is measured with a digital multimeter when the gun voltage is constantly set to the cooling value.

Ieis assumed to remain constant if small changes of a reasonably high cathode voltage are implemented, which has been shown³to be a good approximation.

3Hamberg and Danielsson, unpublished.

2.1.3 Ion current measurement

The ion current measured in CRYRING for the different experiments can dif- fer significantly depending on the nature of the ion and its production pathway.

Ion beam intensities can range between 1nA and 10µA. During the accelera- tion of the ion beam by the RF-system ion bunches are formed. The absolute ion current is typically measured within 0.1 s after the end of the RF acceler- ation when the ions are still bunched.

The ion current measurement system is built up of two instruments: A Bergoz integrating current transformer (ICT) that delivers an absolute value of the ion current with a sensitivity of 1 nA_rmsand a capacitive pick-up (PU) that gives a relative current value, although with a sensitivity down to 100 pA_rms. Therefore the PU is calibrated with the ICT signal at relatively high ion cur- rents. The current is subsequently lowered down to a value suitable for the DR-measurements which will proceed with ion currents as low as 500 pA_rms [76].

Using this method one obtains a reliable absolute ion current value deter- mined just after acceleration. Within a 0.1 s after this measurement is made the bunching of the ion beam will disappear and the ion beam will slowly (de- pending on the ion lifetime) degrade due to rest-gas collisions and DR events and therefore the methods of ion current measurement described above are not longer suitable. In order to obtain an ion current throughout the total mea- surement sequence we used a micro channel plate (MCP) situated at the end of a straight section after one of the bending magnets, which collects signals from neutral particles from background events deriving from ion interaction with residual gas in the ring (mostly H₂). The neutrals are unaffected by the bending magnets and will therefore leave the storage ring tangentially imping- ing on the MCP. A multi channel scaler (MCS) with well defined small time intervals (typically 2 ms, called bins) is then used to record the signal. The ion current throughout the measurement sequence is finally obtained by scal- ing this MCS data with the previously achieved absolute ion current under the assumption that this scaling factor stays constant.

2.1.4 0-degree arm and particle detector system

As mentioned above the DR processes results in neutral fragments that will leave the ring in a tangential direction. In conjunction with the bending magnet directly after the electron cooler a straight additional fing section is mounted tangentially, the so called 0-degree arm, where the DR neutral fragment de- tectors are set up.

During most experimental DR investigations at CRYRING energy sensi- tive ion implanted detectors (IID, see fig.2.4) have been used. The detectors are in principle built up like a PN diode connected in reverse mode to a bias voltage. When a neutral fragment impinges on this surface it will penetrate into the underlying PN-layer and create electron-hole pairs on its way. Due to

Figure 2.4: Ion implanted detector, together with manipulator arm and signal cable.

the bias voltage this charges are soon annihilated by a small current passing through the detector. In the ion kinetic energy interval present in our experi- ments the number of electron-hole pairs is proportional to the kinetic energy of the hitting ion. This effect is linear and since all the neutral fragments are traveling with essentially the same velocity the detector will give signals that are linearly depending on the mass of the incoming particles.

Two detectors with different diameter Ø34 mm and Ø62 mm, respectively, are used (see their respective characteristics in table2.2). Due to the energy released in the DR-processes lighter fragments can get enough transversal ki- netic energy to surpass the active detector area. This effect will be more promi- nent for heavier ions traveling slowly in the storage ring and especially when the neutral fragments created is light such H and D. Therefore the detectors are mounted as close as possible to the electron cooler. The larger of the detec- tors shows a slightly poorer resolution but has the advantage that less particles with high transversal kinetic energy miss detection. Consequently the latter device is more often used for lighter ions. The small detector is more sen- sitive to losses of light fragments but has a superior resolution. For the ions investigated in this thesis only the small detector has been used since a good resolution is crucial to obtain reliable branching fractions with the compara- tively heavy ions studied.

Another element implemented in the 0-degree arm is a metal flag sitting in front of the detectors. The flag only opens during the phases of the measuring cycle when the flux of neutral particles is sufficiently low to not be harmful.

Table 2.2: Ortec ULTRA^{T M} ion implanted detectors and their characteristics mea- sured at 21±1^◦C.

Detector BU-025-900-300^a BU-055-3000-300

Diameter Ø34 mm Ø62 mm

Alpha Resolution^b 25 keV 55 keV

Noise level 24.0 keV 11.9 keV

Used Preamplifier 142B 142C

a Ortec BU-025-900-300 is read: B = Rear microdot connector, U = ULTRA^{T M}, 025 = guaranteed maximum resolution in keV, 900 = sur- face area in mm², 300 is ion depletion layer inµm.

b Alpha resolution is specified as the maximum peak width for a standard alpha source measured at one-half the peak height (FWHM) expressed in keV. The total system alpha resolution is measured and warranted for 5.486-MeV alphas from²⁴¹Am with an ORTEC preamplifier chosen to be consistent with the detector capacitance and an ORTEC amplifier using equal differential and integral time constants of 1.0µs [77]

A typical setup scenario can be seen in fig.2.5. The outgoing signal from these detectors is preamplified with ORTEC 142B respective 142C model preamplifier. The signal is then further amplified and shaped before it is sent forward depending on the envisaged measurement:

• Branching fraction determination: In this type of experiment we normally run the electrons at constant centre-of-mass kinetic energy relative (∼0 eV) to the ions. Unfortunately the detector system is not able to resolve different fragments emerging from the same DR event due to the short time interval between the impacts of these particles and the long response time of the detector(i.e. shaping time of the main amplifier), which typically is in the range of µs. A metal grid is therefore inserted in front of the detector, it is made of metal thick enough (∼50µm) to stop the neutral fragments.

Throughout its surface area it has holes that are big enough (Ø80 µm) to easily let fragments through without affecting them. The transmission probability of the grid is 0.297±0.015⁴[78]. The signal from the detector corresponds linearly to the sum of the masses of the fragments impinging on the detector after a certain DR event. These signals are augmented by the amplifier and are then recorded by a multi channel analyzer (MCA, see

4The threeσconfidence interval