Laboratory of Physical ChemistryDepartment of Chemistry

Faculty of ScienceUniversity of Helsinki

Finland

Spectroscopy and photochemistry of hydrogenperoxide and its complexes in solid rare gases

Susanna Pehkonen

ACADEMIC DISSERTATION

To be presented, with the permission of the Faculty of Science of the University of Helsinki forpublic criticism in the Main lecture hall A110 of the Department of Chemistry on June 12th, 2008 at

12 o’clock noon.

Helsinki 2008

ii

SupervisorProfessor Markku Räsänen

Laboratory of Physical ChemistryDepartment of Chemistry

University of HelsinkiFinland

ReviewersProfessor Matti Hotokka

Department of Physical ChemistryÅbo Akademi University

Finland

and

Professor Henrik KunttuDepartment of ChemistryUniversity of Jyväskylä

Finland

OpponentProfessor Bengt Nelander

Division of Chemical PhysicsChemical Center

University of LundSweden

ISBN 978-952-92-3947- 4 (paperback)ISBN 978-952-10-4729-9 (PDF)

http://ethesis.helsinki.fi

Helsinki 2008Yliopistopaino

iii

Abstract

Molecules and atoms can be isolated in low temperature solids, and this offers an opportunity tostudy their properties in monomeric and complexed forms, and in photochemical reactions.Hydrogen peroxide is an important molecule in various oxidation reactions in industrial chemistryand in the atmosphere. In this thesis, the vibrational properties, structure, and photochemistry ofhydrogen peroxide and its complexes have been studied in solid rare gases using infrared andluminescence spectroscopy. Quantum chemical calculations have been carried out to interpret theexperimental results.

Hydrogen peroxide was obtained from urea hydrogen peroxide adduct compound by flushing itwith rare gas. This newly-developed method provides a safe and convenient means of producingpure, gaseous hydrogen peroxide for laboratory studies. The IR spectra of H2O2, HDO2, and D2O2

in solid argon are presented and compared with previous results, and the IR spectra in solid kryptonand xenon are presented for the first time. Complexes of hydrogen peroxide with small atmosphericspecies, such as SO2 and N2 were studied using infrared spectroscopy and quantum chemicalcalculations.

UV photolysis of hydrogen peroxide in solid argon produces a complex between water and aground state oxygen atom, and OH radicals. It is found that the H2O···O(3P) complex can beconverted back to hydrogen peroxide by light. A three-component photokinetic model

2 3

12 2 2H O O H O 2 OH

k k

k

describes the observed features in solid argon. The wavelength dependence of the rate constants ispresented. Additional photoproducts are observed in solid krypton and xenon. It was found that thethree-component model used in solid argon does not describe the experimental results obtained insolid krypton and xenon. Permanent losses of O atoms are proposed, particularly in solid xenon.Permanent losses of O atoms can be connected with the higher mobility of oxygen atoms in theheavier rare gas solids as compared with solid argon. The relatively weak absorption band of theOH radical in solid krypton and its absence in solid xenon upon UV photolysis of H2O2 arediscussed in terms of delayed cage exit. The photochemical reaction H2O···O(3P) H2O2 wasalso observed in solid krypton and xenon.

The H2O···O(3P) complex was also produced from water molecules and thermally mobilizedoxygen atoms in solid krypton, and the photochemical reaction producing H2O2 was also found. Theorigin of this photochemical reaction is suggested to be a photoinduced charge transfer betweenH2O and O with a concomitant recombination to hydrogen peroxide. Theoretical modeling of thereaction pathway supports the proposed charge transfer mechanism. This photochemical reactionmay also have significance in atmospheric chemistry, because it may serve as a new source for theproduction of hydrogen peroxide.

iv

Acknowledgements

This research was carried out in the Laboratory of Physical Chemistry at the University of Helsinki.I would like to express my deepest gratitude to my supervisor Professor Markku Räsänen, whointroduced me to the fascinating field of matrix isolation. His support and advice have been a greathelp during these years. Docent, Dr. Leonid Khriachtchev, is greatly acknowledged for usefuldiscussions, his encouragements and invaluable guidance during the research and at the final stageof the study.

The former and present members our research group and the personnel of the Laboratory ofPhysical Chemistry deserve my thanks for making the laboratory a pleasant working environment. Iwould to express my gratitude to all of the co-authors of the publications I-VII. I am very grateful toDr. Mika Pettersson and Professor Jan Lundell for helpful discussions and guidance, and forcreating a positive atmosphere during this research.

I would like to thank Professor Lauri Halonen for the opportunity to work in the Laboratory ofPhysical Chemistry and for his criticism and helpful comments on the manuscript. I would also liketo thank Professor Matti Hotokka and Professor Henrik Kunttu, the reviewers of this thesis, for theirwork and useful suggestions. Dr. Joseph Guss is acknowledged for revising the English language ofthis thesis. Financial support for this thesis work from the Finnish Cultural Foundation, Universityof Helsinki, Emil Aaltonen Foundation, Acta Chemica Scandinavica, and from The GraduateSchool Laskemo (Ministry of Education) is gratefully acknowledged.

Finally, I want to thank all of my friends, my parents Raili and Matti, my brother Mikko, Liisaand Tapio, and most of all, my husband Mika and our beloved children Laura and Ilmari for theirinvaluable support and enjoyable life beyond the office.

Susanna PehkonenMilan, May 2008

v

List of original publications

This thesis is based on the following publications, which are referred to by the Roman numerals I-VII:

I M. Pettersson, S. Tuominen, M. Räsänen: IR Spectroscopic Study of H2O2, HDO2, andD2O2 Isolated in Ar, Kr, and Xe Matrices. J. Phys. Chem. A 101 (1997) 1166-1171.

II L. Khriachtchev, M. Pettersson, S. Tuominen, M. Räsänen: Photochemistry of HydrogenPeroxide in Solid Argon. J. Chem. Phys. 107 (1997) 7252-7259.

III J. Lundell, S. Pehkonen, M. Pettersson, M. Räsänen: Interaction between HydrogenPeroxide and Molecular Nitrogen. Chem. Phys. Lett. 286 (1998) 382-388.

IV S. Pehkonen, M. Pettersson, J. Lundell, L. Khriachtchev, M. Räsänen: PhotochemicalStudies of Hydrogen Peroxide in Solid Rare gases: The Formation of HOH···O(3P)Complex. J. Phys. Chem. A 102 (1998) 7643-7648.

V L. Khriachtchev, M. Pettersson, S. Jolkkonen, S. Pehkonen, M. Räsänen Photochemistry ofHydrogen Peroxide in Kr and Xe Matrices. J. Chem. Phys. 112 (2000) 2187-2194.

VI S. Pehkonen, J. Lundell, L. Khriachtchev, M. Pettersson, M. Räsänen: Matrix Isolation andQuantum Chemical Studies on the H2O2···SO2 Complex. Phys. Chem. Chem. Phys. 6 (2004)4607-4613.

VII S. Pehkonen, K. Marushkevich, L. Khriachtchev, M. Räsänen, B. L. Grigorenko, A. V.Nemukhin: Photochemical Synthesis of H2O2 from the H2O···O(3P) van der Waals complex:Experimental Observations in Solid Kr and Theoretical Modeling. J. Phys. Chem. A 111(2007) 11444-11449.

The candidate, Susanna Pehkonen, is the corresponding author of publications IV, VI and VII, and aco-author of publications I, II, III, and V. She has performed most of the experimental work inpublications I, III, IV, VI, and VII, and a significant part of the experimental work in publication II.The experimental work consists of preparing samples, measuring infrared spectra, performingphotolysis using an excimer laser, or participating in photolysis by selected wavelengths using anoptical parametric oscillator in cooperation with Dr. Leonid Khriachtchev. The author performedpart of the theoretical calculations in publication IV, whilst Prof. Jan Lundell carried out most of thecalculations in publications III, IV, and VI. The calculations in publication VII were done by Dr.Bella Grigorenko and Prof. Alexander Nemukhin.

vi

Contents

1. Introduction 11.1 Hydrogen peroxide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.2 Matrix isolation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.3 Photodissociation in the solid state . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

1.3.1 Photochemistry of H2O2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

2. Experimental methods 52.1 Sample preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52.2 Gas phase experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52.3 Matrix isolation experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52.4 FTIR measurements and UV photolysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62.5 Computational methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

3. Results and discussion 93.1 Gas phase measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93.2 The infrared spectrum of H2O2 in solid rare gases . . . . . . . . . . . . . . . . . . . . . . . . 103.3 Complexes of H2O2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

3.3.1 H2O2···N2 complex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123.3.2 H2O2···SO2 complex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

3.4 Photochemistry of hydrogen peroxide in solid argon . . . . . . . . . . . . . . . . . . . . . . 173.4.1 Photolysis of H2O2 in solid argon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173.4.2 Kinetics of the H2O2 photolysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213.4.3 Light-induced recovery of H2O2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

3.5 Photochemistry of H2O2 in solid krypton and in solid xenon. . . . . . . . . . . . . . . . . 263.6 Photolysis of the H2O2···SO2 complex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283.7 Synthesis of H2O2 from H2O and N2O in solid krypton . . . . . . . . . . . . . . . . . . . . 29

4 Concluding remarks 34

5 Errata 36

6 References 37

vii

Abbreviations

antis. Antisymmetric (vibration)a. u. Arbitrary unitbend Bending (vibration)BSSE Basis set superposition errorB3LYP Becke3-Lee-Yang-Parr (functional)CASSCF Complete active space self-consistent field (method)CC Coupled cluster (method)CCSD(T) Coupled cluster with single, double, and perturbative triple excitationCI Configuration interaction (method)CIS Configuration interaction including singles (method)DCBS Dimer centered basis setDNA Deoxyribonucleic acidEA Electron affinityEPR Electron paramagnetic resonanceFTIR Fourier transform infrared (spectroscopy)HF Hartree-FockIP Ionization potentialIR Infrared (spectroscopy)LCAO Linear combination of atomic orbitalsLIF Laser induced fluorescenceMCQDPT2 Multi-configurational quasi-degenerate perturbation theory (method)MCSCF Multi-configurational self-consistent field (method)MI Matrix isolationMO Molecular orbitalMP Møller-Plesset perturbation theoryMP2 Second order Møller-Plesset perturbation theory (method)OPO Optical parametric oscillatorPES Potential energy surfacePFA Perfluoroalcoxy (copolymer)SCF Self-consistent field (method)str. Stretching (vibration)sym. Symmetric (vibration)UHP Urea hydrogen peroxideUV Ultraviolet (spectroscopy)Vis Visible (spectroscopy)

1

1 Introduction

Photoinduced chemical reactions of various compounds can be studied in low-temperaturesolids. The use of tunable lasers allows for the selective excitations of electronic andvibrational transitions in order to generate new species. These photochemically producedspecies can be studied using various spectroscopic methods. Quantum chemicalcalculations are used to model the potential energy surfaces, reaction pathways, andgeometries of the species involved in the photochemical reaction to interpret theexperimental results.

This thesis comprises experimental and computational work on photochemical andspectroscopic studies of hydrogen peroxide and its complexes in solid rare gases. Theemphasis is on the experimental work. Quantum chemical calculations are used tounderstand the experimental observations. The development of a safe method to producesmall amounts of pure, gaseous hydrogen peroxide from a urea hydrogen peroxide adductcompound (UHP) made it possible to study hydrogen peroxide. [I] Infrared andluminescence spectroscopy, isotopic substitution, and quantum chemical calculations wereused to characterize hydrogen peroxide, [I] its complexes, [III, VI] and its photolysisproducts. [II, IV, V] Paper VII concludes this thesis by a synthesis, which demonstratesthat hydrogen peroxide can be produced by a photochemical reaction from H2O and N2Oprecursors in solid krypton. This observation may have significance in atmosphericchemistry.

1.1 Hydrogen peroxide

Hydrogen peroxide, H2O2, is a molecule of considerable interest in many fields includingoxidation reactions, [1] combustion chemistry, [2] biological processes, [3] andphotodissociation dynamics. [4-11] H2O2 is also important in industrial chemistry where,for example, it is a frequently used bleaching agent in the production of pulp and paper.H2O2 is considered to be a central oxygen metabolite produced in living cells [12] and ithas been observed to damage DNA. [13]

Hydrogen peroxide also plays an important role in atmospheric chemistry, because it isa source of reactive OH radicals. Gaseous hydrogen peroxide is mainly formed upon thereaction of two hydroperoxyl radicals [14]

HO2 + HO2 H2O2 + O2. (1)

Biomass burning and other combustion processes are also important sources of gas phasehydrogen peroxide. During thunderstorms, high concentrations of H2O2 have beenobserved. [15] The concentration of hydrogen peroxide in the atmosphere depends onphotochemical processes. It was found that the concentration of H2O2 follows the intensityof the light. [14] Chemical reactions in the clouds have a significant effect on thephotochemical processes in the lower troposphere, which also affects the concentration of

2

H2O2. [16] Hydrogen peroxide dissolves easily into aqueous water droplets formingaqueous phase hydrogen peroxide, which is the main sink for gas-phase H2O2. UVphotolysis of H2O2, and the Fenton reaction in the presence of Fe(II), are other sinks ofH2O2. [14]

In space, H2O2 has been observed on the surface of Europa, a satellite of Jupiter, [17]where it has been suggested that it exists in the form of isolated H2O2•2H2O species, [18]and on Enceladus, one of the icy satellites of Saturn. [19] H2O2 has also been detected inthe atmosphere of Mars indicating that water has been photochemically converted to H2O2.[20]

1.2 Matrix isolation

Pimentel and co-workers developed the matrix isolation (MI) technique in 1954 to studyhighly unstable molecules. [21] In matrix isolation experiments, individual molecules aretrapped and isolated from one another in a solid, inert matrix at low temperatures. Underthese conditions, the molecules cannot interact. Therefore, the MI technique is well-suitedto study reactive chemical species such as ions and free radicals. The matrix isolationtechnique also allows for the study of weakly bound systems such as hydrogen-bonded,charge transfer, and van der Waals complexes, which are unstable under room temperatureconditions.

To prepare a frozen solid matrix, the species to be studied is sprayed, with a largeamount of rare gas, onto a cold substrate, which is cooled to temperatures of 4-40 K.Experiments are performed under high vacuum to prevent contamination from unwantedgases freezing onto the cold window. Rare gases are common host materials for MI studiesbecause of their large spectral transparency, which extends from the vacuum ultraviolet tothe far-infrared region. Besides rare gases, molecular oxygen, nitrogen, methane,hydrogen, ammonia, and water have been used as host materials. [22] Reactive species canbe generated either in the gas phase prior to deposition, or after deposition by in-situphotolysis of an appropriate matrix-isolated precursor. Several spectroscopic techniquessuch as infrared (IR), Raman, ultraviolet and visible spectroscopy (UV/Vis), laser inducedfluorescence (LIF), and electron paramagnetic resonance (EPR) can be used in MI studies.[22, 23]

Solid hosts interact weakly with guest molecules. Therefore, the absorption bands ofguest molecules are shifted from the corresponding gas phase values. Aggregation of guestspecies and molecules trapped in several lattice geometries produce additional absorptionbands in the spectra. Even a small amount of an unwanted impurity, such as N2, O2, orH2O, can change the spectrum, since complexation of various forms can take place. Smallmolecules, like H2O, NH3, and hydrogen halides, can rotate in a low-temperature matrix.This complicates the interpretation of the spectra. [24]

Despite the reservations mentioned above, a low-temperature solid environment isoften used to study molecules and molecular complexes because the species can beisolated and selectively studied and because the formation and decomposition of thespecies can be controlled. For example, a low-temperature rare gas environment has been

3

employed to study the structural and vibrational properties of van der Waals complexes ofH2O2. Hydrogen peroxide forms complexes in a low temperature environment, andcomplexes between N2, [II] (CH3)2O, [25] CO, [26] NH3 and N(CH3)3, [27] phosphorusand sulfur bases, [28] O3, [29] hydrogen halides (HF, HCl, and HBr), [30] water, [31]hydrogen peroxide itself, [32] and the peroxy radical have been reported. [33]

1.3 Photodissociation in the solid state

In photodissociation, a molecule absorbs one or more photons, and then decomposes intofragments:

* ) + ) (2)

where is the energy of a photon (angular frequency ), *(AB) describes an excitedstate complex, and the labels and specify the particular internal quantum states of theproducts A and B. [34] The energy required for dissociation depends on the bond strength.Part of the photon energy is consumed in breaking the molecular bond and the excessenergy is distributed among the kinetic, electronic, vibrational, and rotational energy of theproducts. Eventually, the energy is dissipated into the surroundings.

Photodynamics in the solid phase can be distinguished from photodynamics in the gasphase by the importance of many-body interactions. The cage effect is a process where thesolid host prevents or hinders the dissociation of a molecular species. Due to this effect,recombination of fragments, isomerization or reaction with the matrix cage may happen.Solvation by polarizable host species, such as Kr and Xe, can stabilize ionic species andcan therefore reduce significantly the ionization energies. [35]

If the excess energy is large enough, at least one fragment can exit the cage, andpermanent dissociation takes place. The probability of permanent dissociation in solidsurroundings depends on the molecular binding energy and the barrier height of the solidhost, which is typically 1-2 eV in rare gases (1 eV = 1.602 10-19 J). [35] At low photonenergies and for large masses, this often leads to a perfect caging. [36] Sometimes, thecollisions between the guest and the host distort the cage and lower its barrier height, sothat a delayed exit may take place. Delayed cage exit has been observed for HI, [37] HCl,[38] H2O, [39] and H2S, [40] among others. A sudden exit, i.e. the exit of a photofragmentfrom the cage without a significant distortion to the cage, usually occurs with the smallestphotofragments, such as O(1D), F(2P), and S(1D). [36] Permanent photodissociation leadsto stabilization of photofragments at some distance from the parent cage. Furthermore, forsmall atomic fragments, light-induced mobility can take place. This has been observed forthe F atom in the photodissociation of F2 in free-standing Ar crystals [41, 42] and for theO(1D) atom in the photolysis of N2O in a Xe matrix as well as in free-standing Xe crystals.[43 - 45]

4

1.3.1 Photochemistry of H2O2

Hydrogen peroxide absorbs light at wavelengths shorter than 350 nm. The absorptioncross section increases monotonously as the wavelength decreases in the region of 190-350 nm. [46] The gas phase photochemistry of H2O2 has been studied extensively. [4-11]Gaseous hydrogen peroxide dissociates at wavelengths shorter than ~320 nm. [14] Thedominating photolysis products are two ground state hydroxyl radicals with high rotationalquantum numbers. [34] Upon 193 nm photolysis (6.42 eV), the energetically possiblephotoproducts are:

2 2H O + h 2 2H + O < 875 nm (3)

32H O + O( P) < 829 nm (4)

2H OO < 721 nm (5)

2 2OH(X ) + OH(X ) < 555 nm (6)

12H O + O( D) < 358 nm (7)

2H + HO < 325 nm (8)

12H O + O( S) < 213 nm (9)

2 2 +OH(X ) + OH(A ) < 198 nm (10)

The given wavelengths are calculated from the standard enthalpies of formation at 298 Kand at 1 bar (1 bar = 105 Pa). [47] In the gas phase, the quantum yield of the reaction

22 2H O 2OH(X )h was measured to be 1.22 0.13 (193 nm photolysis), [48] which is

lower than 2 (2 is the maximum because the photolysis of H2O2 produces two OHradicals). The quantum yields for other reaction pathways are small. [49] For the H+HO2

products, the quantum yield was determined to be 0.12 [50] and 0.20 0.03. [51]

5

2 Experimental methods

2.1 Sample preparation

Handling of pure hydrogen peroxide requires special techniques and care because of thedanger of explosion, particularly if the concentration of H2O2 exceeds 26 mole %. [52]Commercially available urea hydrogen peroxide adduct compound, CO(NH2)2•H2O2

(UHP), has often been used as a source of hydrogen peroxide in the oxidation reactions oforganic compounds in the liquid phase, because it is safer to use than pure H2O2. [53] Inthis study, urea hydrogen peroxide was used for the first time to obtain gaseous hydrogenperoxide. [I]

2.2 Gas phase experiments

The gases released from UHP were analyzed using an FTIR spectrometer (GASMET,Temet Instruments) designed for quantitative analysis of gas mixtures. [54] Approximately100 mg of UHP was placed in a Pyrex glass tube, which was held in a furnace. One end ofthe tube was connected to the FTIR spectrometer, and carrier gas (Ar, 99.9%) wasprovided from the other end. The flowing gas flowed through a gas cell with an opticalpath length of 1.5 m. The system was purged with the carrier gas to remove water vaporfrom the walls of the gas cell. After stabilization of the system, a background spectrumwas measured. The sample temperature was raised slowly, and spectra were recorded witha resolution of 8 cm 1. The available temperature range was limited because ureadecomposes thermally to NH3 and HNCO. [55] The decomposition temperature of urea,85°C, was obtained in thermogravimetric studies by Ball and Massey. [56]

2.3 Matrix isolation experiments

Hydrogen peroxide was released by flushing UHP with the matrix gas, and the obtainedgas mixture was deposited onto a cold window. The UHP was typically at roomtemperature during the deposition. The deposition line was constructed mainly from PFAplastic (PFA is perfluoroalcoxy copolymer, manufactured by Swagelok) in order to reducecatalytic decomposition of hydrogen peroxide on metal surfaces. [57] Deuteratedperoxides were obtained from deuterated UHP synthesized by mixing D2O with a 30 %water solution of H2O2 and urea, then heating the mixture to 60 °C, and finally pumpingthe water off. [58] In the deuteration process, one must ensure that no unbound H2O2

remains in the vessel before vacuum pumping in order to prevent explosion.In order to prepare matrices, the gas mixture was deposited onto a CsI substrate in a

closed-cycle helium cryostat (APD, DE 202A or Displex CS-202). The typical deposition

6

time was one hour. The temperature of the CsI substrate was controlled with a Lakeshoretemperature controller. The matrix thickness varied between 70 and 150 m. TheH2O2/Rg, H2O2/N2/Ar, H2O2/SO2/Rg, and N2O/H2O/Kr matrices were prepared in thefollowing ways:

H2O2/Rg gaseous samples were obtained by flushing UHP with 99.9999% Ar,99.997% Kr or 99.997% Xe. The total pressure of the 2.0 L bulb containing therare gas was about 360 torr (1 torr = 1 mmHg = 133.322 Pa, 1 bar = 105 Pa). Thesubstrate temperature was 17-18 K. The spectra were measured at 17-18 K. SomeH2O2/Rg samples were studied using a cryostat providing temperatures as low as7.5 K.The H2O2/N2/Ar gaseous samples were prepared by flushing UHP with an N2/Armixture at room temperature. The N2/Ar mixtures were prepared by mixing a smallamount (0-2%) of N2 with Ar. The total pressure of the 2.0 L bulb was about 400torr. The matrix deposition temperature was 7.5 K.H2O2/SO2/Rg gaseous samples were prepared by flushing UHP with SO2/Rg atroom temperature. The total pressure of the 2.0 L bulb was about 400 torr. The gasmixture was deposited onto a CsI substrate kept at 20 K for argon and 25 K forkrypton and xenon. After deposition, the sample was cooled down to 7.5 K and thespectrum was measured.In N2O/H2O/Kr experiments, two bulbs were used. One bulb contained the N2O/Krgas mixture and the other bulb contained the H2O/Kr gas mixture. The bulbs wereconnected to the matrix deposition line through calibrated leak valves (Leybold,Varian) and the gases were mixed in the line before the deposition port. TheN2O/Kr samples were made using standard manometric technique. The totalpressure of the 2.0 L bulb was typically about 400 torr. The N2O/Kr ratio wasvaried from 1/500 to 1/1000, the typical ratio being 1/700 (N2O of technical grade,AGA). The H2O/Kr samples were made by adding a drop of deionized and distilledwater into the 2.0 L bulb and degassing it before the bulb was filled with rare gas.This water-containing bulb was used as a source of constant H2O/rare gascomposition (H2O/rare gas = 1/16) determined by the vapor pressure of water atroom temperature. The ratio between the guest and the host was estimated usingthe known integrated absorptivities and integrated intensities of guest bands. Thegas mixtures were deposited onto a CsI substrate kept at 25 K and then cooleddown to 8.5 K.

2.4 FTIR measurements and UV photolysis

The infrared absorption spectra were measured in the 4000-400 cm 1 spectral region witha Nicolet SX60 FTIR-spectrometer. A liquid nitrogen cooled MCT detector and a Ge/KBrbeam splitter were used to measure the spectra at resolutions of 1 cm 1 or 0.25 cm 1.

Hydrogen peroxide belongs to the C2 point group. H2O2 has 6 fundamental vibrations:the IR active modes are the symmetric and antisymmetric OH stretching vibrations ( 1 and

5, respectively), the antisymmetric OH bending vibration ( 6), and torsion ( 4). The

7

infrared inactive modes are the symmetric OH bending ( 2) and OO stretching ( 3)vibrations. [59] The torsional vibration (254.55 cm-1 in the gas phase [60]) is outside ofthe spectral region investigated in these experiments.

Since hydrogen peroxide was obtained from UHP, its absolute amount could beestimated only from integrated intensities. Calculations based on integrated intensities ofH2O2 show that the H2O2/Ar ratio was ~1/700 - 1/1200 in the prepared matrices, inaccordance with a rough estimation made in paper I. In the H2O2/SO2/Rg experiments, thetypical ratio was (1-2)/1/750, which agrees with the rough estimate presented in paper VI.The values used were the following: 300.7 km mol 1 for the 3 band of N2O, [61] 60 kmmol 1 for the 6 band of H2O2, [62] and 44.6 km mol 1 for the 3 band of H2O. [63]

Irradiation of the matrices was carried out with an excimer laser (MSX-250, MPBTechnologies Inc) operating at 193 nm (ArF) or 248 nm (KrF) with a typical pulse energydensity of ~10 mJ/cm2. Frequency-doubled radiation from an optical parametric oscillator(OPO Sunlite, Continuum) was used at wavelengths from 225 to 380 nm with a typicalpulse energy density of ~16 mJ/cm2.

2.5 Computational methods

Ab initio calculations were performed to support experimental interpretation. Structures,energies, and vibrational frequencies of the H2O···O, H2O2···N2, and H2O2···SO2

complexes, and the corresponding monomers and reaction pathways from (H2O)+O toH2O2 were computed. All of the calculations were performed using GAUSSIAN 94, [64]GAUSSIAN 98, [65] GAMESS, [66] and PC GAMESS [67] programs.

The energy of a molecule may be obtained by solving the Schrödingerequation H E . The analytical solution of the Schrödinger equation for molecularsystems is impossible. Electronic structure methods use various approximations to obtain asolution to the Schrödinger equation. [68] There are two classes of electronic structuremethods: semi-empirical methods that use parameters from experimental data, and abinitio methods. The latter are briefly introduced here as this was the approach adopted inthis research.

In ab initio methods, some approximations are made in order to obtain the solution. Asnuclei are much heavier than electrons, they move much more slowly and may be treatedas stationary while the electrons move quickly around them. The Schrödinger equation cantherefore - to a good approximation - be separated into nuclear and electronic parts. This iscalled the Born-Oppenheimer approximation.

The electronic wave function can be solved exactly for a hydrogen-like system. Formany-electron systems, approximations are needed to solve the Schrödinger equation. Agenerally useful approach is to represent the electronic wave function in a chosen finitebasis set. In this approach, the Schrödinger equation can be transformed into an algebraicequation, which can be solved numerically. In many computational programs, the many-electron wave function is constructed from approximate molecular orbitals (MOs). Theseare represented as linear combinations of atomic orbitals (LCAOs) constructed fromGaussian-type basis functions.

8

The total electronic wave function is antisymmetric with respect to the interchange ofany two electronic coordinates. This can be achieved by building it from Slaterdeterminants. A single Slater determinant is used in the Hartree-Fock (HF) self-consistentfield (SCF) method, which provides the basis for ab initio calculations. In the Hartree-Fock approximation, each electron moves in an averaged potential of the remainingelectrons, and the energy of the electron in the effective field is minimized self-consistently. Because electron repulsion is overestimated in the Hartree-Fock calculations,electron correlation methods have been developed to take the correlation of electrons intoaccount. Examples of electron correlation methods are Møller-Plesset (MP) perturbationtheory, configuration interaction (CI), and coupled cluster (CC) methods.

Electronically excited states require a careful description of electron correlation toobtain reliable energies. A qualitative description of excited states can be achieved by theCI method for a single excited determinant, denoted as CIS. A more sophisticateddescription requires the use of multi-configurational self-consistent field theory (MCSCF)based methods where the orbitals are optimized for each particular state. MCSCF basedmethods are often used to describe the low lying electronic states or bond breaking. [68]

The basis sets used were the split-valence, 6-31G with diffusion and polarizationfunctions to give 6-31+G(d,p), 6-31++G(2d,2p), and 6-311++G(2d,2p) basis sets.Additionally, the properties of the complexes were studied using correlation consistentbasis sets up to triple-zeta quality (cc-pVDZ and cc-pVTZ). Calculations were alsoperformed with the double-zeta basis sets where the diffuse functions were added (aug-cc-pVDZ). An introduction to the basis sets used can be found in Ref. 68. The complexproperties were considered via the supermolecular Møller-Plesset perturbation theory tothe second order (MP2). The harmonic vibrational frequencies were calculated usinganalytical second derivatives. The MP2/6-311++G(2d,2p) level of theory was found to bean acceptable compromise between efficiency and accuracy for weak chemicalinteractions. [69]

The interaction energies were estimated as the difference between the total energy ofthe complex and the monomers at infinite distance, where the monomer wave functionswere derived in the dimer centered basis set (DCBS). This approach corresponds to thecounterpoise correction which was proposed by Boys and Bernardi; it aims to minimizethe basis set superposition error (BSSE). [70]

Complete active space SCF (CASSCF) is a variant of MCSCF, consisting of an SCFcomputation with full CI involving a subset of the orbitals. This subset is known as theactive space. In paper VII, the potential energy surface calculations of (H2O)++O wereperformed using the CASSCF method. The two 1s atomic orbitals of oxygen were keptdoubly occupied in all configurations, while 12 active orbitals were available for partialoccupation of 14 electrons. This set of orbitals was used to qualitatively describe theH2O···O, OH···OH, and H2O2 species participating in the photochemical reaction. Thegeometry optimizations were performed using the CASSCF approximation, and energiesin the point of major interest were also computed using the second order multi-configurational quasi-degenerate perturbation theory (MCQDPT2) as implemented inGAMESS. [66, 67]

9

3 Results and discussion

3.1 Gas phase measurements

The gas phase analysis demonstrated that gaseous H2O2 is released from the UHP adductbelow 92 °C. [I] Above 92 °C, urea decomposes producing NH3 and HNCO. [55] Thistemperature is close to the value of 85°C obtained from thermogravimetric studies by Balland Massey. [56] The decomposition temperature (92 °C) obtained in gas phase studies [I]is inaccurate since a large volume of gas is involved in the experiments. A 1.5 L gas cellwas used, and the whole volume could not be heated uniformly.

The main aim of the gas-phase experiment was to show the release of gaseoushydrogen peroxide below the UHP decomposition temperature. Figure 1 presents a lowresolution FTIR spectrum of H2O2. The fundamentals of H2O2 at 3600 cm 1 ( 1 and 5)and 1260 cm 1 ( 6) are seen in the spectrum. The band at 2345 cm 1 arises from residualcarbon dioxide inside the multipass gas cell. A weak water vapor bending fundamental

2) is seen at 1600 cm 1. The symmetric and antisymmetric stretching bands of watervapor (3657 and 3756 cm-1, respectively [71]) partially overlap with the H2O2 stretchingbands (3600 cm-1), but the intensities of these vibrations are lower than that of the 2

bending vibration of H2O. [63] The very weak H2O bands show that this method can beused to obtain a small amount of pure, gaseous H2O2. This encouraged the use of UHP inmatrix isolation experiments.

3750 3000 2250 1500

0.00

0.02

0.04

H2O2

Abs

orba

nce

Wavenumber (cm-1)

CO2

H2O

H2O

2

Figure 1. FTIR spectrum of gaseous H2O2/Ar mixture measured with a resolution of8 cm 1. The temperature of UHP was 67 °C.

10

3.2 The infrared spectrum of H2O2 in solid rare gases

Matrix isolation experiments in which UHP was the source of H2O2 showed that H2O2 canbe deposited without significant contamination by water. A survey spectrum of H2O2

isolated in solid argon is presented in Figure 2. All the water peaks are due to vibration-rotation bands of monomeric water. [72-77] Due to a concentrated sample, aggregates ofH2O2 are also observed. No indication of urea nor of its decomposition products wasobserved in the matrix. In the following experiments, the amount of H2O2 was kept smalland the amount of multimeric H2O2 was minimized.

Upon deuteration, absorption bands belonging to D2O2 and HDO2 were observed. Theidentification of absorption bands due to H2O2 and its deuterated analogues in solid argonwas based on previous work on H2O2 in solid argon [78, 79], in solid N2, [80] and in gasphase studies. [59, 62, 81 - 85] The IR spectra of H2O2 and its isotopologues in solidkrypton and xenon were identified for the first time in this work. [I] The main infraredabsorption bands of monomeric H2O2, HDO2 and D2O2 are presented in Table 1.

4000 3000 2000 1000

0.0

0.4

0.8

1.2

1.66

2, m

H2O2

+6

d

5,

1

H2O

Abso

rban

ce

Wavenumber (cm-1)

Figure 2: Survey spectrum of H2O2 isolated in an argon matrix. The concentration ofH2O2 is high, and dimeric (d) [32] and multimeric species (m) are present. Impurity waterpeaks are indicated in the figure. The spectral resolution is 1.0 cm 1.

Several gas-phase studies have previously shown that the absorption bands ofhydrogen peroxide are split due to hindered internal rotation. [60, 82, 84] Hougen hasanalyzed the allowed transitions, when both cis and trans tunneling are taken into account.[86] Perrin et al. derived from experimental data the lower trans- and higher cis- energybarriers for the 6 = 1 vibrational state to be 312.32 0.9 cm 1 and 2840.4 120 cm 1,respectively. [85] In the gas phase, the splitting of the ground state is 11.44 cm-1. [83] Dueto tunneling, the gas phase infrared spectrum of hydrogen peroxide consists of doubletstructures.

11

In this study, split absorption bands are observed. [I] This splitting was explained bythe tunneling of a hydrogen atom in analogy with the gas phase. However, splitting due tomatrix site effects is also probable. The assignment of the symmetric and antisymmetricstretching vibrations ( 1 and 5, respectively) is not obvious due to overlapping of thesetwo bands and the interference of impurity water bands in the same spectral region. Theexact assignment of the stretching vibrations of HDO2 is uncertain due to splitting of theHDO2/D2O2 bands and possible interference of the combination 2+ 6 ( 2 and 6 are thesymmetric and antisymmetric bends, respectively) band of H2O2 and water impurities.

Table 1. Main infrared absorption bands (in cm 1) of monomeric H2O2, HDO2 and D2O2

in solid argon, krypton, and xenon.

Ar Kr Xe AssignmentH2O2 3597.0 3583.6 3568.0 1, 5 OH str.H2O2 3587.8 3574.0 3560.0 1, 5 OH str.

H2O2 2653.0 2649.7 2639.6 2+ 6

H2O2 2643.6 2636.3 2628.5 2+ 6

D2O2 2655.7 2645.2 2639 1, 5 OD str.D2O2 2651.3 2633.2 1, 5 OD str.D2O2 2645.3 2628.4 1, 5 OD str.

D2O2 1974.2 - - 2+ 6

D2O2 1965.2 1960.0 1954.3 2+ 6

HDO2 1350.3 1346.4 1341.8 bendingHDO2 1342.6 1339.6 1336.6 bending

H2O2 1277.0 1273.7 1270.3 6 sym. bendH2O2 1270.9 1268.7 1265.7 6 sym. bend

D2O2 951.3 949.9 947.8 6 sym. bend

12

3.3 Complexes of H2O2

3.3.1 H2O2···N2 complex

The interaction between H2O2 and N2 was studied using infrared spectroscopy and abinitio calculations. [III] Computational studies with the H2O2···N2 complex wereperformed at the MP/6-311++G(2d,2p) and MP2/cc-pVTZ levels of theory. The MP2/6-311++G(2d,2p) optimized structure of the complex was used to evaluate the interactionenergy at higher correlated levels (CCSD(T)). A potential energy surface (PES)calculation for the H2O2···N2 complex resulted in only one stationary point, where thestructure of H2O2 is perturbed very little upon complexation compared with the isolatedH2O2. The H2O2···N2 complex is an almost linear hydrogen-bonded complex. The BSSE-corrected interaction energy at the CCSD(T) level, using a geometry derived with MP2/6-311++G(2d,2p), is E(CCSD(T)/6-311++G(2d,2p)//MP2/6-311++G(2d,2p) = 5.5 kJ/mol.When nitrogen is added to an H2O2/Rg sample, new absorption bands appear close to theH2O2 bands, and these features are assigned to the H2O2···N2 complex. Figure 3 presentsthe 6 band of H2O2 as the amount of N2 is increased. The calculated wavenumber shiftsbetween the H2O2 monomer and H2O2···N2. These agree with the experimentally observedshifts, see Table 2.

1290 1285 1280 1275 1270 1265 12601290 1285 1280 1275 1270 1265 12601290 1285 1280 1275 1270 1265 12601290 1280 1270 1260

*

*

2.0 % N2

1.0 % N2

0.5 % N2

Wavenumber (cm-1)

0 % N2

*

Abs

orba

nce

(a. u

.)

Figure 3. Spectra of H2O2 with N2 in solid argon as a function of the N2 concentration.The H2O2···N2 complex band is marked with an asterisk.

13

Table 2. Calculated and observed wavenumbers of H2O2 and H2O2···N2 (in cm 1).Underlined values are used in the calculations of the observed shift between the H2O2

monomer and the H2O2···N2 complex.

Assignment H2O2

(in Ar)MP2/cc-pVTZ

H2O2···N2

(in Ar)MP2/cc-pVTZ

Calc.shift

Observedshift

5 OH str. 3588.2 3820.3 3587.2 3818.2 2.1 1.01 OH str 3586.4 3819.3 3582.1 3810.0 9.3 4.32 6 2653.0

2643.6 2656.52 sym.bend 1437.4 1450.8 +13.46 antis.bend 1277.0 1280.0 1337.8 +9.8 +9.1

1270.9 1328.0

3.3.2 H2O2···SO2 complex

The interaction between H2O2 and SO2 was studied both experimentally andcomputationally. [VI] Computationally, two structures of the H2O2···SO2 complex wereobtained.

Figure 4. Calculated geometry of the lower energy form of the H2O2···SO2 complex.

H1

O2

O3

H4O5

S6

O7

2.046 Å

14

The lower energy structure forms a five membered planar ring with an interaction energyof -19.5 kJ/mol at the CCSD(T)/6-311++G(2d,2p)//MP2/6-311++G(2d,2p) level oftheory. The calculated interaction bond length (H4-O5) at the MP2//6-311++G(2d,2p)level of theory is 2.046 Å (1 Å = 10-10 m), see Figure 4. The higher energy structureforms a twisted 5-membered ring, and it could be considered to be a higher energyconformer of the lower energy structure with a rotation of the dangling H-atom aroundthe O-O bond of H2O2. The calculated CCSD(T) interaction energy of this structure is-11.2 kJ/mol. The computationally obtained structure in Figure 4 is similar to onereported earlier by Vincent et al. [87] The structure is also similar to the experimentallyobtained structure of HOO···SO2 by Svensson and Nelander. [88]

In H2O2/SO2/Rg matrices, a complex of H2O2 and SO2 is formed, as demonstrated inFigure 5 for solid argon (spectrum c). Several additional bands near the absorption bandsthat belong to isolated H2O2 and SO2 provide evidence for this complex. Comparing theIR spectrum of the complex with the spectra of the H2O2 and SO2 monomers, [I, 89] theseadditional absorption bands are at 3572.8 cm 1, a triplet at 3518.7, 3511.2, and 3504.7cm 1, and at 1340.3, 1280.2, and 1149.9 cm 1. These new bands are seen at relatively lowguest/host ratios indicating that they belong to the 1:1 complexes. When dimeric H2O2 ispresent, an additional band appears at 1290.8 cm-1. This band is close to the reportedH2O2 dimer band (1293.5 cm-1 [32]) and is probably due to the (H2O2)2···SO2 complex.Previously, Tso and Lee tentatively assigned two OH stretching vibrations at 3568 and3548 cm 1 to the 1:1 H2O2···SO2 complex in their photolysis experiments of H2S in solidO2. [90]

Table 3. Observed infrared absorption bands of the H2O2···SO2 complex, H2O2, SO2, andthe vibrational shifts ( ) in solid argon, krypton, and xenon. All values are in cm-1.

Monomers H2O2···SO2

Ar Kr Xe Ar (Ar) Kr (Kr) Xe (Xe)H2O2 str. 3587.7 3574.0 3560.0 3572.8 -15.0 3570.3 -3.7 3551.4 -8.6H2O2 str. 3518.7 -69.1 3533.1 -40.9 3531.3 -28.7

3511.2 -76.63504.7 -83.1

3 SO2 1355.2 1350.1 1344.9 1340.3 -14.9 1337.2 -12.9 1334.0 -10.96 H2O2 1270.9 1268.7 1265.7 1280.2 9.3 1278.2 9.5 1273.4 7.71 SO2 1152.0 1149.8 1148.9 1149.9 -2.1 1145.2 -4.6 1142.0 -6.92 SO2 519.5 519.0 517.0 526.2 -7.2 523.2 -6.2

Figure 6 presents infrared spectra of the H2O2···SO2 complex in the OH stretchingregion in solid argon, krypton, and xenon. The main difference between the spectra is theshift of the OH stretching mode. In solid krypton and xenon, the free OH stretching modeis shifted a few reciprocal centimeters from the corresponding monomeric stretchingmode in the same matrices, and the bonded OH stretching mode is shifted

15

-40.9 and -28.7 cm 1 for solid krypton and xenon, respectively. In both matrices, the non-complexed OH bond is sensitive to the change of the environment, while the complexedOH bond is not. Similar behaviour was also observed in the case of the H2O2···COcomplex in solid argon, krypton, and xenon. [26] However, in solid argon, the situationfor the H2O2···SO2 complex is different: the bonded OH stretching of the H2O2···SO2

complex is split and the shift from monomeric OH stretching mode is abnormal. Onepossible reason for this may be the size of the matrix lattice site compared to theembedded H2O2···SO2 complex. The argon matrix cage may be too small for thiscomplex; this changes the complex structure and causes spectral splitting of theabsorption. The triplet structure in solid argon is most probably due to the matrix siteeffect. Observed infrared absorption bands of the H2O2···SO2 complex, H2O2, SO2 and thevibrational shifts in solid argon, krypton, and xenon are presented in Table 3.

-0.2

0.0

0.2

0.4

0.6

1350 1320 1290 1260

3600 3560 3520 3480

0.00

0.04

0.08

*(H2O2)2

SO2

H2O2

°*

SO2

Abso

rban

ce

a

b

c

c

b

a

H2O2

*** *

Wavenumber (cm-1)

Figure 5. FTIR spectra of SO2/Ar (spectrum a), H2O2/Ar (spectrum b) and H2O2/SO2/Ar(spectrum c) matrices. The H2O2···SO2 bands are marked with an asterisk. The bandmarked with a circle (°) is tentatively assigned to the (H2O2)2···SO2 complex.

16

3600 3575 3550 3525 3500

Ar

Wavenumber (cm-1)

Kr

Abso

rban

ce (a

. u.)

*

*

** * *

*

*

Xe

Figure 6. FTIR spectra of the H2O2···SO2 complex in the OH stretching region in solidargon, krypton, and xenon. The bands marked with asterisks belong to the H2O2···SO2

complex.

In Ar matrices, the free OH stretching mode of the H2O2···SO2 complex shifts by -15cm 1 from the H2O2 fundamental (3578.8 cm 1), while the complexed stretching mode hasa larger red shift of 69-83 cm 1, depending on the matrix site. Computationally, at theMP2/6-311++G(2d,2p) level of theory, a more stable complex (Figure 4) was predicted toinduce shifts of -21 cm 1 and -91 cm 1 for the free and hydrogen-bonded OH-groups ofH2O2, respectively. The predicted vibrational wavenumbers for the H2O2 subunit agreewith experimental results. For the SO2 subunit, the agreement is not so good, since onlythe symmetric stretching mode ( 1) of SO2 was adequately predicted by the calculations.This disagreement could be due to insufficient electron correlation and incompleteness ofthe basis set used to describe the SO2 monomer in the complex. The addition of ananharmonic correction to the wavenumber calculations might solve this problem. In fact,B3LYP anharmonic frequency calculations of the SO2···HO2 complex by Wang and Hou[91] are in agreement with the experimental work on this complex in solid argon. [88]

17

3.4 Photochemistry of hydrogen peroxide in solid argon

Photolysis of hydrogen peroxide in solid argon, krypton, and xenon is described in thefollowing sections, and differences between the solid hosts are discussed. A novel resultobtained from the photodissociation of H2O2 is the formation of the H2O···O complex inlow-temperature solids. In this complex, water is bonded to a ground state oxygen atomforming a hydrogen-bonded complex (HOH···O). Interestingly, the H2O···O complex canbe converted back to H2O2 by UV light. The mechanism of the light-inducedphotochemical recovery reaction is discussed.

3.4.1 Photolysis of H2O2 in solid argon

In the gas phase, H2O2 decomposes at wavelengths shorter than 320 nm forming two OHradicals. [92] In solid argon, UV photolysis of H2O2 produces (i) OH radicals and (ii) theH2O···O complex originating from the OH + OH reaction in the cage. A differencespectrum showing the results of 193 nm photolysis of H2O2 in solid argon is presented inFigure 7. The negative bands correspond to the decomposition of H2O2, and new IRabsorption bands appear at 3730.1, 3633.0, and 3554.1 cm 1. [II, IV]

3800 3700 3600 3500

-0.1

0.0

0.1

0.2

H2O2

Abso

rban

ce

Wavenumber (cm-1)

OH

H2O...O

Figure 7. Difference spectrum of the 193 nm photolysis of H2O2 in solid argon at 17 K.

The assignment of the OH radical spectrum is based on infrared and luminescencespectroscopy. Under irradiation of the photolyzed sample at 285 nm, a strongluminescence band at 340 nm and a weaker sideband at 395 nm were observed. [II]According to the results by Goodman and Brus, this luminescence originates from matrix-isolated OH radicals. [93] The observed infrared absorption of the OH radical at 3554cm 1 is close to the previous assignment by Cheng et al. [94] Observation of the OH

18

radicals shows that some escape from the parent cage upon UV photolysis of H2O2, whileother OH radical pairs remain in the parent cage, and their concomitant reaction producesH2O and O: [IV]

2OH + OH H O + O . (11)In the matrix cage, the (H2O + O) fragments form a van der Waals complex, which wasobserved in the infrared spectrum. Because the amount of monomeric water does notincrease as a result of photolysis, cage exit of an oxygen atom after the reaction plays aminor role. The sharp bands at 3730.1 and 3633.0 cm 1 were assigned to the H2O···Ocomplex (see Figure 7). Similar spectra of the H2O···O complex were observed uponphotolysis of hydrogen peroxide in krypton and xenon matrices at 18 K. The H2O2

photolysis products in various matrices are listed in Table 4.

2800 2750 2700 2650-0.08

0.00

0.08

0.16

3750 3700 3650 3600

0.00

0.05

0.10

Wavenumber (cm-1)

Abso

rban

ce

HDO2 + D2O2

DOD···OHOD···O

DOD···O

H2O2

HOH···OHOD···O

DOH···O HOH···O

Figure 8. Photolysis of H2O2 /HDO2 /D2O2 in solid argon at 17 K (difference spectra).The negative bands are from decomposed H2O2, HDO2 or D2O2. Interacting atoms areindicated in bold.

19

The assignment of the H2O···O complex spectrum was based on results obtained fromIR spectroscopy, isotopic substitution, and quantum chemical calculations. Upon UVphotolysis of deuterated peroxide, sharp bands appear at 2766.5, 2653, 2691.8, 3699.8,and 3664.9 cm 1 (see Figure 8). These absorption bands were integrated and compared atdifferent stages of the photolysis and were found to grow (and decrease) synchronously.With a low deuteration degree, where H2O2 and HDO2 were dominant before photolysis,only one OD stretching absorption band at 2691.8 cm 1 was observed. [IV]

The observed frequencies of 3730.1 and 3633.0 cm 1 are close to monomeric waterband centers in solid argon (3734.3 and 3638.0 cm 1). [72] Also, the isotope H/Dfrequency ratios of the new bands, 1.348 and 1.369, are close to the corresponding valuesfor monomeric water [74] and complexed water [95] in solid argon. The new bands weresuggested to have arisen from complexed water molecules. Because the precursor wasH2O2, a complex of H2O and an oxygen atom was suggested.

The ab initio calculations at the MP2/6-311++G(2d,2p) level of theory on the van derWaals complex H2O···O and its deuterated analogues support the assignment given above.The calculated shifts between the monomeric water band and the correspondingcomplexed band agree with the experimental results. For H2O···O in solid argon, theexperimental shifts of the OH stretching modes are 4 and 5 cm 1 from the waterfundamentals and the corresponding computational shifts are 16 and 18 cm 1. ForD2O···O, the experimental shifts of the OD stretching are 4 and 5 cm 1 and thecorresponding computational shifts 8 and 8 cm 1. The H2O···I complex in solid argonstudied by Engdahl and Nelander also exhibits similar absorption bands to the H2O···Ocomplex assigned here. [95]

Figure 9. Optimized structure of the H2O···O complex. The bond lengths and bond anglesare calculated at the MP2//6-311++G(2d,2p) level of theory. [IV] The values inparentheses are the corresponding results of the CASSCF calculations. [VII]

105°(103°)

168°(167°)

2.23 Å(2.54 Å)

20

Table 4. The main products of 193 nm photolysis of H2O2, HDO2, and D2O2 in solidargon, krypton, and xenon, and the calculated harmonic wavenumbers of the H2O···Ocomplex at the MP2/6-311++G(2d,2p) level of theory. The H(D) atom, to which theoxygen atom is bound, is indicated in bold to distinguish the HDO···O complexes. Thewavenumbers are in cm 1, and the calculated harmonic wavenumbers are scaled by afactor of 0.95. The calculated intensities (in parentheses) are given in km mol 1.

Ar Kr Xe Calculated Assignment3747.5 3790 (74) H2O

3730.1 3718.6 3704.3 3773 (129) HOH...O 3

3699.8 3739 (46) HOD...O3664.9 3708 (153) DOH...O

3734 (36) HOD3695.8 3676 (10) H2O

3633.2 3622.4 3607.3 3658 (52) HOH...O 1

3554.1 3547.1 3530 OH3550.2 3541.8 OH

3538.6 OH2777 (45) D2O

2766.5 2758.8 2749.1 2769 (79) DOD…O 3

2714 (18) DOH...O2710 (21) HOD

2691.8 2689.5 2691 (74) HOD...O2649 (7) D2O

2653 2674.8 2641 (27) DOD...O 1

2620.1 2615.9 OD2617.3 2613.5 OD

1630 broad HOH ?1578 (66) H2O1572 (57) HOH...O

1592.9 1589.9 1586.7 HOH ?1383.9 HOO

1385 (53) DOH...O1383 (57) HOD1374 (50) HOD...O

1174 11721155 (35) D2O1150 (31) DOD...O

21

The computationally predicted stable structure of the H2O···O complex is presented inFigure 9, and the computed harmonic wavenumbers are presented in Table 4. Thecomputed wavenumbers are similar to the experimental values. Two separate OHstretching bands are predicted for every isotopologue of the water···oxygen atom complex(HOH···O, DOD···O, HOD···O, and DOH···O). In analogy to the H2O2 photolysisexperiments, the two stronger HDO···O bands at 3699.8 and 2691.8 cm 1 were assigned tothe HOD···O complexes, and the weaker band at 3664.9 cm 1 to the DOH···O complex, assuggested by computations. This is in agreement with the usually preferred complexationroute through the D side rather than the H side. [96] The OD stretching absorption of theDOH···O complex was not experimentally observed. According to calculations, it islocated at ~2714 cm 1, and its intensity is much weaker than that of the correspondingband of HOD···O. Neither the bending vibration of the H2O···O complex nor the bendingvibration of its isotopologues was found in these experiments. Engdahl and Nelander alsofailed to observe the bending vibration of water in the H2O···I complex. According to themthe bending vibration of the H2O···I complex could be hidden under water absorption. [95]

3.4.2 Kinetics of the H2O2 photolysis

The H2O2 photolysis kinetics was followed by integrating the infrared bands of H2O2,H2O···O, and OH. The concentrations were assumed to be proportional to the area underthe IR absorption bands. At the beginning of the photolysis, the H2O···O concentrationincreases, but after some time it begins to decrease. At 193 nm, the only stablephotolysis product of H2O2 is the OH radical. Summing up the normalized concentrationsof H2O2, H2O···O, and OH, it is observed that - within error limits - the total concentrationremains constant, as shown in Figure 10. The kinetic model of the photolysis of H2O2 insolid argon can be described as follows:

2 3

12 2 2H O O H O 2 OH

k k

k (12)

where ki are the rate coefficients of the corresponding processes. Minor photolysischannels and the reaction producing H2O2 from two OH radicals (OH + OH H2O2) areomitted in equation (12). The sum (k1+k3) gives the H2O2 photodissociation rate. Afterdissociation, the initially created pair of OH radicals can become separated from eachother in the solid host or, if they remain in the same cage, they can react with each other toproduce H2O···O (k1). In turn, the H2O···O complex can be converted into H2O2 by light.

The quantum yield of OH radical production in 193 - 230 nm photolysis in solid argon,k3 /(k1+k3), is (0.17±0.02), indicating that 83 % of the OH radicals remain in the cageforming the H2O···O complex. However, cage exit may be even less likely because thepresent analysis excludes the recovery of hydrogen peroxide from two OH radicals.

The absolute rate of H2O2 decomposition increases with photon energy, and thespectral trend for (k1+k3) follows the known absorption spectrum of gaseous H2O2. [14]

22

Using irradiation from 193 nm to 320 nm, the yield of H2O···O increases with increasingphoton energy as presented in Figure 10 (lower panel).

0 5 10 150.0

0.2

0.4

0.6

0.8

1.0

H2O2

H2O... O

2OH

SUM

Con

cent

ratio

n

Number of pulses (x104)

200 250 3000.0

0.2

0.4

0.6

H2O

...O

max

imum

Wavelength (nm)

Figure 10. Kinetics of the 230 nm photolysis of H2O2 in solid argon at 18 K. The averagepulse energy density was ~1 mJ/cm2. Upper panel: Relative concentrations of H2O2,H2O···O and OH as a function of irradiation time. The strongest absorption bands of H2O2

(1270.0 cm 1), H2O···O (3730.1 cm 1), and OH (3554 cm 1) were used to determine theconcentrations. The lines are fits to the kinetic model (equation 12). Lower panel: Themaximum concentration of the H2O···O complex as a function of photolysis wavelength.

3.4.3 Light-induced recovery of H2O2

Light-induced recovery of hydrogen peroxide from the water-oxygen complex wasobserved in solid argon. Irradiation at 300 nm destroys the H2O···O complex and recovershydrogen peroxide (see Figure 11). The recovery kinetics upon 300 nm irradiation ispresented in Figure 12. The concentration of OH radicals remains constant during the 300nm irradiation.

23

3750 3700 3650 3600 3550 3500

H2O

2

(c)

(b)

(a)

H2O2

Wavenumber (cm-1)

Abs

orba

nce

H2O···O

H2O···O

OH

Figure 11. Photolysis and recovery of H2O2 in solid argon: (a) after deposition, (b)difference spectrum showing the result of the 193 nm irradiation, (c) difference spectrumshowing the result of the 300 nm irradiation. Note that the amount of OH radicals isunchanged by the 300 nm irradiation.

0.0 0.5 1.0 1.50.0

0.2

0.4

0.6

0.8

H2O...O

2OH

H2O2

Con

cent

ratio

n

Number of pulses (x104)

Figure 12. Kinetics of the H2O···O H2O2 reaction in solid argon at 18 K under300 nm irradiation. The H2O2/Ar matrix was preliminarily photolyzed at 193 nm.

24

200 240 280 3200.0

0.5

1.0

KrAr

k 2 (a.

u.)

Wavelength (nm)

Figure 13. Excitation profile of the H2O···O H2O2 photoreaction in solid argon andkrypton.

The rate coefficient k2 for the H2O···O H2O2 photoreaction was determined as afunction of the photolysis wavelength in solid argon (see Figure 13). The maximum of theH2O2 recovery (k2) was obtained at ~260 nm (4.8 eV). Due to the broadness of theabsorption profile, photoinduced charge transfer was suggested as an initial step of thephotoreaction: [II]

H2O···O + h (H2O)+ + O (13)

Charge transfer could occur between water and the oxygen atom. The ionization potentialof argon (15.759 eV) [47] is too high to be considered at the excitation energies used. ForCoulombically bound exciplexes, the transition energy required for charge transfer can beestimated by the following expression: [97, 98]

tr sol2 e= IP(H O) - EA(O) - ( ) -E U R E , (14)

where IP and EA are the ionization potential and the electron affinity, U(Re) is theelectrostatic potential of the (H2O)+···O– complex, solE is the difference between theground and the excited state solvation energies. IP(H2O) = 12.6 eV, EA(O) = 1.46 eV, [47]and the electrostatic potential energy of 5.35 eV obtained from ab initio calculations areused. The solvation energy, solE , was estimated by employing a simple classical cavity cellmodel by Fajardo and Apkarian: [97, 98]

2sol

30

1 8( -1) ( )=(2 +1)

Ed

, (15)

25

where is the dielectric constant of the medium, d is the cavity diameter, and 2( ) is thedifference between the squares of the dipole moments in the ground and excited states.The estimate of solvation energy yielded 0.51 eV. As a result, trE was calculated to beabout 5.28 eV (235 nm), which is in reasonable agreement with experiment.

The (H2O)+···O– charge transfer complex was suggested to relax to H2O2 via oxywater,H2OO. [II] Oxywater is theoretically predicted to be stable, although it has not beenobserved experimentally. [99] The oxywater can form hydrogen peroxide by a simple1,2-hydrogen shift with a low energy barrier. [100, 101]

However, in the recent quantum mechanical calculations shown below, the (H2O)+O–

charge transfer state is found to relax to H2O2 via a complex between two OH radicals.[VII] The relaxation of the charge transfer species, (H2O)+ + O , to H2O2 was modeledcomputationally using ab initio CASSCF calculations. [VII] Using MCQDPT2calculations at the CASSCF geometry, the vertical excitation energies were calculated toestimate the accuracy of the calculations. The calculated excitation energies of the twolowest excited singlet states of H2O2, 5.66 eV and 6.62 eV with an estimated maximumerror of 0.5 eV, agree with previous calculations by Liu et al. [102]

Figure 14. Energy diagram for the formation of H2O2 from the charge transfer (H2O)++O–

state. Singlet states are marked with blue bars, and triplet states with red bars.

The diagram in the Figure 14 summarizes the possible transformations in the H2O2

system. Initially, the H2O···O(3P) complex is in the triplet state (lower left corner in Fig.

+-

01234567

E(eV)

S=1

S=1

S=1

S=1, S=0

S=0

26

14). The fourth triplet state (E ~7 eV) corresponds to the excited state complex,(H2O)++O– , i.e. the charge transfer state. The computed vertical excitation energy, 5.8 eV,was obtained by averaging over four states in the CASSCF procedure followed by theMCQDPT2 calculations. The obtained gas-phase value agrees with the semiempiricalestimate obtained in Ref. II (5.28 eV) described above. In the solid environment, if matrixsolvation effects and possible errors in computation (0.5 eV) are taken into account, theenergetic scale is fairly consistent with the experimental results (4.8 eV). It shows that anexcitation of about 5 eV of the H2O···O complex in the matrix would allow the complex toreach the charge transfer (H2O)++O state.

A possible path from (H2O)++O– to H2O2 was investigated by performingunconstrained geometry optimizations starting from the (H2O)++O– state. Upon descentalong the PES, a structure consisting of HO···H···O was found, as seen in Figure 14. Thepotential energy surface leads from the HO···H···O structure to a complex between two OHradicals. The rearrangement of a complex of two OH radicals to H2O2 should be easilyaccomplished in the solid state. In conclusion, the CASSCF calculations suggest that (i)the (H2O)++O– charge transfer state can be excited by the photon energies used, and (ii)the formation of H2O2 is probable from this excited state.

3.5 Photochemistry of H2O2 in solid krypton and in solid xenon

In this section, the results obtained in different matrices are compared. Severalphotochemical studies on H2O2 were performed in solid krypton and xenon. The studiesdiscussed above were carried out in solid argon at 18 K. [II, IV] When the temperature islower (7.5 K), all of the photolysis products in solid argon agree with the measurements athigher temperatures. On the other hand, some new features in the photolysis of H2O2 areobserved in solid krypton and in solid xenon. [V] There are three main differences in thephotolysis of H2O2 in solid argon as compared to solid krypton and xenon. [V] First,additional absorption bands are observed in solid krypton and xenon. Secondly,differences are found in the amount of the H2O···O complex formed upon photolysis in thedifferent matrices. The third difference concerns the OH radical as the photolysis product.

A new absorber was observed upon UV photolysis of hydrogen peroxide at 7.5 K. Thefeature appears at 1588 cm 1 in solid krypton, and at 1587 cm 1 in solid xenon. Theintensities of these bands decrease to zero upon annealing at about 15 K, while theH2O···O complex bands grow. During annealing, there were no observed changes in theintensities of H2O2 and OH. The band observed in solid krypton and xenon is close to theabsorption band position of non-rotating water (1589.2 cm 1 in solid krypton), [72] whichsupports the idea that the band originates from complexed water. The band is probably dueto a complex between water and the oxygen atom with different geometry compared to thevan der Waals complex H2O···O described earlier. This new band was assigned to a loosecomplex between H2O and O, denoted as H2O//O. [V] The irreversible conversion of theloose complex H2O//O to the H2O···O complex occurs at about 15 K and does not involveglobal mobility of the O atom. The formation of the loose H2O//O complex represents anexample of the short-range stabilization discussed by Khriachtchev et al. [103] The

27

photolysis of H2O2 produces H2O and an energetic O atom in a cage. The energetic Oatom can be stabilized at different distances from H2O thus forming H2O···O and H2O//O,or it can escape from the vicinity of the H2O molecule.

The second difference from argon matrices concerns the long-range escape of Oatoms. It was found that the three-component model used in solid argon does not describethe experimental results obtained in solid krypton and xenon. Permanent losses of O atomscan be proposed. In solid xenon in particular, permanent losses dominate in 193 nmphotolysis. The increased losses in solid krypton and xenon can be connected with thehigher mobility of oxygen atoms therein as compared with solid argon. The highermobility of oxygen atoms favors the cage exit of an energetic O atom, and with theexistence of additional photolysis products. Light-induced mobility of O atoms can alsocontribute to the permanent losses.

The third difference concerns the concentration of OH radicals. In solid argon, OHradicals are the only photostable photolysis product. Upon UV photolysis of H2O2 in solidkrypton, the absorption band of the OH radical is relatively weak and is absent in solidxenon. The absence of OH radicals in solid xenon was verified by measuring the LIFspectrum, which showed no sign of OH radicals in xenon after the 225-320 nm photolysis.[V] There are reasons for the absence of OH radicals. First, the delayed cage exit of afragment is strongly influenced by energy transfer between the guest and the host. [36]Activation of the surroundings needed for cage exit takes several attempts, and the cageexit can be interrupted by the reaction between the OH radicals (OH + OH). [II] Thenumber of activation attempts should be much larger for a Xe matrix compared with Kr orAr matrices due to the larger mass of Xe atoms. The light fragment loses only a small partof its energy in one collision with the lattice atoms. Therefore, the cage exit probability isappreciable for a large number of attempts. [104, 105] In this approach, the quantum yieldof the (OH + OH) separation decreases for heavier rare gases, in a qualitative agreementwith the experiment. Second, the Xe surroundings can catalyze additional photolysischannels. The photolysis of H2O2 in solid xenon produced absorption that can be assignedto the HOO species (1383.9 cm-1), in analogy with HOO absorption bands reported insolid argon. [106-108] Furthermore, the potential energy surfaces of HXeOOH andHOXeOH might participate in the photolysis dynamics.

The H2O···O H2O2 photoreaction was also observed in solid krypton and xenon.First, the tight H2O···O complex was prepared by 266 nm irradiation and subsequent low-temperature annealing (15-20 K), which promotes the formation of the tight complex fromthe loose complex. The tight H2O···O complex was then converted into H2O2 by irradiatingin the 230-320 nm spectral region. The rate coefficient k2 for the photorecovery reactionwas determined as a function of wavelength in solid argon and krypton (see Figure 13,chapter 3.4.3). The maximum of the spectral profile in solid krypton is at about 280 nm,shifted to the red ~20 nanometers from the value obtained in solid argon. This similarity ofthe recovery efficiency and the spectral position support the conclusion that the chargetransfer step is intrinsic to the H2O···O complex. The role of the surrounding may becrucial for the H2O+···O– H2O2 reaction. The small red shift of the profile in solidkrypton from that in solid argon can be partially explained by the charge transfer model,when the change of solvation energy due to the different dielectric constants

28

( = 1.00052 in Ar, and = 1.0078 in Kr) is taken into account. [47] In solid xenon, asimilar recovery was found in the same spectral region as for solid argon and krypton, butthe profile was not measured due to large permanent O atom losses. No recovery wasobserved from the loose complex H2O//O. [V] Furthermore, if it is assumed that the host iswater ice, the charge transfer model described above predicts that the maximum of theH2O···O absorption profile would shift even further to longer wavelengths. This suggeststhat the H2O···O H2O2 photoreaction could be significant in the production ofatmospheric hydrogen peroxide.

3.6 Photolysis of the H2O2…SO2 complex

1440 1410 1380 490 480 470

-0.01

0.00

0.01

0.02

3750 3700 3650 3600 3550 3500

-0.01

0.00

0.01

0.02

H2O···SO3

Abs

orba

nce

Wavenumber (cm-1)

iOH

H2O···SO3H2O···O

i

H2O2 ****

Figure 15. FTIR spectra after 300 nm photolysis of a H2O2/SO2/Ar sample. The negativebands indicate decomposition of the H2O2···SO2 complex (marked with asterisks). TheH2O···SO3 complex is formed in the photolysis of the H2O2···SO2 complex. The H2O···Ocomplex is formed from the photolysis of monomeric H2O2. At the beginning of thephotolysis, two lines appear at 3599.1 and 3514.7 cm 1. They belong to an intermediatespecies (marked with i), which decomposes upon further irradiation.

29

UV photolysis of the H2O2···SO2 complex in rare gas matrices at 7.5 K mainlyproduces a complex between water and sulfur trioxide, H2O···SO3. The H2O···SO3 complexhas previously been identified by Bondybey and English in Ne, Ar, and Xe matrices. [109]Irradiation at 300 nm decomposes H2O2···SO2, but SO2 is not observed to dissociatesignificantly. The IR spectrum obtained by the 300 nm photolysis of H2O2···SO2 in solidargon is presented in Figure 15.

In solid argon, a two-step process in a matrix cage for the formation of H2O···SO3 wasproposed. [VI] First, H2O2 is photolyzed to yield H2O···O. The so-formed ternary complex(H2O···O···SO2) is isolated in the matrix cage, which is expected given the small cage exitprobability for an oxygen atom in solid argon. The OH stretching vibrations of thetentatively assigned H2O···O···SO2 complex are shifted a few reciprocal centimeters fromthe OH stretching absorption of the H2O···O complex, see Fig. 15. In this ternary complex,the oxygen atom further reacts with SO2 to produce SO3 in the vicinity of a watermolecule, forming the H2O···SO3 complex. At the CCSD(T)//MP2 level of theory, theH2O···SO3 complex was found to be 174 kJ/mol lower in energy than H2O2···SO2.

In solid krypton and xenon, the main photolysis product of the H2O2···SO2 complex isalso the H2O···SO3 complex, but the total amount is smaller compared to solid argon.Permanent losses of O atoms, as discussed in the H2O2 photolysis case, probably explainthis observation.

3.7 Synthesis of H2O2 from H2O and N2O in solid krypton

The photochemical synthesis of H2O2 from the H2O···O complex was studied in solidkrypton. H2O2 was synthesized from H2O and N2O precursors in three steps. [VII] First,the 193 nm photolysis of N2O yields oxygen atoms in solid krypton. Next, upon annealingat ~ 25 K, mobile oxygen atoms react with water forming the H2O···O complex. Finally,the H2O···O complex is converted to H2O2 by irradiation at 300 nm. The synthesis can bedescribed as follows:

N2O 193 nm N2 + O(1D) (16)

O(1D) t O(3P) (17)

O(3P) + H2O annealing H2O···O (18)

H2O···O 300 nm (H2O)+···O H2O2 (19)

It is supposed that the light-induced conversion of H2O···O to H2O2 proceeds via a chargetransfer state as described in chapter 3.4.3.

30

In these experiments, the H2O and N2O precursors in solid krypton were mainlymonomeric species (see Figure 16). [VII] The positions of the infrared absorption bands ofH2O [72] and N2O [110-111] agree with previous matrix isolation studies. Photolysis at193 nm efficiently decomposes N2O in solid krypton whereas the concentration of H2Oremains practically unchanged. This was verified by integrating and summing the waterbands. In this spectrum, the negative peaks at 3765.5 and 3700.7 cm–1, and the positivepeak at 3747.7 cm–1 belong to different rotational transitions of monomeric water. Theobserved changes in the monomeric water absorption intensities are due to ortho-paranuclear spin conversion. [112] After photolysis, no IR active products, for example NO,N2O2, or O3, are detected in solid krypton. The only photolysis channel of N2O producesmolecular nitrogen and atomic oxygen in analogy with the gas phase and solid state resultsfrom N2O. [113, 114] The O(1D) atoms thus formed relax to the ground state. The lifetimeof O(1D) was found to be 780 ms in solid SF6, [115] while the lifetime is 110-115 s in thegas phase. [116] In some experiments, a relaxation time of up to 3 hours was employedbefore the annealing step. This ensures that the thermally mobilized oxygen atoms are inthe triplet state.

3800 3750 3700 3650 1350 1300 1250-0.4

-0.2

0.0

0.2

0.4

Abso

rban

ce

Wavenumber (cm- 1 )

N2O/H2O/Kr

(b) 193 nm

(c) annealing

(a) deposition

H2O

N2O

H2O···O

* *

Figure 16. Photolysis and annealing of the H2O/N2O/Kr (1/1/700) matrix. (a) IR spectrumof the N2O/H2O/Kr matrix. A small amount of the N2O···H2O complex is present, [117]marked with an asterisk. (b) Difference spectrum with respect to spectrum (a) showing theresults of 193 nm photolysis. No change of the H2O concentration occurs at this stage. (c)Difference spectrum with respect to spectrum (b) showing the result of annealing at 23 K.The H2O concentration decreases at this stage. The sample temperature was 8.5 K duringthe spectral measurements.

31

The O atoms trapped in solid krypton were mobilized by annealing. Upon annealing ofa photolyzed sample at 14 K, new absorption bands were observed at 3718 and 3623cm–1 indicating the formation of the H2O···O complex, see Figure 16c. The formation ofthe H2O···O complex takes place in two stages (see Figure 17a). The first stage occurs at14 - 18 K, and it presumably originates from H2O and O fragments that are in closevicinity to each other (short-range mobility). The second stage occurs at temperaturesabove 22 K, where oxygen atoms move globally. The amount of monomeric waterdecreases at this step of the synthesis indicating a reaction between H2O and mobileoxygen atoms. Ozone (absorption at 1033 cm–1 in solid Kr) [118] was formed in smallquantities and especially after annealing above 25 K. Danilychev and Apkarian studied therecombination of two oxygen atoms in free standing Kr crystals. They found that oxygenatom recombination is activated at 20 K and 30 K. [114] Because the guest/rare gas ratio issmaller in free standing crystals, the obtained temperatures (14 - 18 and 22 K) agree withprevious results. In a xenon matrix, oxygen atom reaction was also observed to occur attwo stages: first at 18 - 24 K from close O···O2 pairs, and at 27 K due to the globalmobility of oxygen atoms. [119]

The kinetics of the global oxygen atom mobility was studied as follows. First, thephotolyzed sample was annealed at 21 K for several minutes to allow all of the atomswithin a short range to react. Secondly, the sample was annealed at 22 - 25 K, and theH2O···O concentration was measured as a function of annealing time. Finally, the samplewas annealed to 28 K to let all the available O atoms react. The H2O···O concentrationobtained at 28 K was used to normalize the concentrations (Figure 17). The kinetic curvesat 22, 23, and 24 K are presented in Figure 17b. The formation of the H2O···O complexwas fast above 25 K, and the formation kinetics could not be measured. The naturallogarithm of the characteristic time (in minutes) of the formation of the H2O···O complexas a function of reciprocal annealing temperature is presented in Figure 17c. A linear fit tothese data yielded an Arrhenius activation energy of (70 6) meV (550 cm–1) for themobility of the O atoms.

The mobility of the oxygen atom in the solid state has been studied in manyexperimental and theoretical papers. [36, 43-45, 114, 120-127] According to the modelsuggested by Danilychev and Apkarian, the mobile oxygen atoms are in the triplet groundstate. [114, 124] The activation energy obtained in the present work, (70 6) meV, agreeswith the experimental site dependent activation energies (0.05 and 0.13 eV) reported byDanilychev and Apkarian. [124]

32

0 100 200 300

0.4

0.6

0.8

1.0

16 20 24

0.2

0.4

0.6

0.8

1.0

0.040 0.042 0.044 0.046

2

4

6

Time of annealing (min)

23 K

22 K

Con

cent

ratio

n 24 K

(b)

Temperature (K)

H2O

...O

con

c.

(a)

ln

1/T (1/K)

(c)

Figure 17. (a) Formation of the H2O···O complex in solid krypton as a function ofannealing temperature. The temperature was increased at a rate of 1 K/min. The spectrawere measured at elevated temperatures. Short-range (14 - 18 K) and long-range ( 22 K)formation of the H2O···O complex can be distinguished. Before annealing, theN2O/H2O/Kr = 1/1/700 matrix was photolyzed at 193 nm. The H2O···O concentration wasobtained by integration of the 3718 cm–1 band and normalized by the value after additionalannealing to 28 K for ~10 min. (b) Kinetic curves of the H2O···O formation at variousannealing temperatures. (c) Characteristic time (in minutes) for the formation of H2O···Oas a function of the annealing temperature. The averaged slope gives an activation energyof ~ 70 meV.

33

Upon 300 nm irradiation, the H2O···O complex absorption decreases, and the H2O2

absorption increases (see Figure 18). The amounts of photolyzed N2O and newly-formedH2O2 were compared by integrating the corresponding IR absorption bands and usingknown band intensities. [61 - 63] The estimate shows that 50-60 % of oxygen atoms fromdecomposed N2O are transferred to H2O2, which is a high yield. In addition, a smallportion of the oxygen atoms may react with each other or with O2 impurities to produceO3, which is observed as a minor product at 1033.6 cm–1. The formation of O2 in similarexperiments has been observed previously. [26]

3800 3700 3600 3500

-0.05

0.00

0.05

Abs

orba

nce

Wavenumber (cm-1)

H2O2

H2O···O

(a)H2O

Figure 18. Formation of H2O2 from H2O···O in solid krypton upon irradiation at 300 nm.The figure shows a difference spectrum, where the spectrum after 28 K annealing is usedas a background. The changes of the monomeric water band intensities are due to spinconversion, [112] and the concentration of H2O does not change.

34

4 Concluding remarks

The photolysis and spectroscopy of H2O2 and its complexes in solid rare gases (Ar, Kr,and Xe) were studied. Hydrogen peroxide was obtained by flushing rare gas over UHP,which provided a new and safe technique to obtain pure, gaseous hydrogen peroxide.Infrared and luminescence spectroscopy, isotopic substitution, and quantum chemicalcalculations were used to interpret the spectra of H2O2, its complexes, and its photolysisproducts.

The UV photolysis of H2O2 and its deuterated analogues was studied in solid raregases. In solid argon, the 193 nm photolysis of H2O2 produces a van der Waals complexbetween water and a ground state oxygen atom, H2O···O(3P), and free OH radicals. TheH2O···O(3P) complex was characterized in solid argon, krypton, and xenon for the firsttime using infrared spectroscopy, isotopic substitution, and quantum chemicalcalculations. The photolysis kinetics of H2O2 was studied. In solid argon, the photolysis ofH2O2 follows a three-component model:

2 3

12 2 2H O O H O 2 OHk k

k (20)

The rate constants are wavelength dependent. During photolysis in solid argon, the sum ofthe normalized concentrations of H2O···O, H2O2 and the OH radical remains unity.

There are three main differences in the photochemistry of H2O2 in solid argoncompared to solid krypton and xenon. (i) A loose water oxygen complex, H2O//O wasobserved in the photolysis of H2O2 in solid krypton and xenon. In the H2O//O complex,the oxygen atom is stabilized at a different distance from the H2O molecule whencompared with the H2O···O complex. The loose H2O//O complex decomposes below 20 Kforming the H2O···O complex. (ii) The second difference is the long-range escape of Oatoms. In solid krypton and xenon, the experimental results of H2O2 photolysis cannot bedescribed with the same model as that used for solid argon: the sum of the normalizedconcentrations of H2O···O, H2O//O, H2O2 and OH radical decreases during the photolysisindicating losses from the three-component model. The total amount of the H2O···Ocomplex is smaller in solid krypton than in solid argon, and it is formed only in smallamounts in solid xenon. (iii) The amount of OH radicals formed differs in the rare gassolids studied. In solid argon, the only photostable product is the OH radical. In solidkrypton, the amount of OH radicals is much smaller, and it is absent in solid xenon. Thedifferences in H2O2 photolysis compared to solid argon are partially due to the higheroxygen atom mobility in solid krypton and xenon. The cage-exit probability of the O atomafter the OH + OH reaction in the cage is larger in solid krypton and xenon. The HOOspecies formed in the photolysis in solid xenon indicates the presence of additionalphotolysis channels.

Hydrogen peroxide was also produced from H2O and N2O precursors using UVphotolysis and low temperature annealing in solid krypton. The 193 nm photolysis of N2O

35

produces oxygen atoms. Upon annealing of the sample at about 25 K, oxygen atoms aremobilized and form the H2O···O complex. Subsequent UV irradiation converts the H2O···Ocomplex into H2O2. The activation energy of the H2O···O formation, i.e. oxygen atommobility, was estimated to be about 70 meV. This agrees with previous experimentalstudies.

A light-induced recovery of H2O2 from the H2O···O(3P) complex was experimentallydemonstrated in solid argon, krypton, and xenon. The origin of the recovery reaction stemsfrom a photoinduced charge transfer between H2O and O with concomitant recombinationinto H2O2. The concentration of OH radicals remains constant, indicating that they do nottake part in the recovery reaction.

Relaxation of the (H2O)++O charge transfer state to H2O2 was studied by the ab initioCASSCF calculations. It was estimated that an excitation energy of about 5 eV in the solidenvironment is needed to excite the H2O···O complex to the charge transfer state, inagreement with the experimentally estimated energy. Upon relaxation of the (H2O)++Osystem, the possible reaction path passes through the HO···H···O and OH···OHconfigurations on its way to H2O2. Computational modeling supports the experimentalresults concerning the excitation energy and the role of the charge transfer state.

A simple cavity model can be applied to estimate the energy needed for the chargetransfer. According to the model, an increase in the polarizability of the host shifts thephotorecovery reaction profile to longer wavelengths. This is in agreement withexperiments in solid argon and in solid krypton. If the polarizability of water is included inthe model, the maximum of the absorbance profile is shifted to longer wavelengths. Thisindicates that the studied photoreaction can produce hydrogen peroxide in condensedwater. This finding may be significant in the formation of hydrogen peroxide in theEarth’s atmosphere or in planetary ices.

36

5. Errata

Publication II:

As a result, we obtain trE = 4.84 eV, which corresponds to 256 nm for the transitionwavelenght, in agreement with the maximum of the observed profile 2( )k . This sentenceon page 7258 should read: As a result, we obtain trE = 5.28 eV, which corresponds to 235nm for the transition wavelenght, in agreement with the maximum of the observed profile

2 ( )k .

37

6. References

[1] Cooper, M. S.; Heaney, H.; Newbold, A. J.; Sanderson, W. R. Synlett. 1990, 533.[2] Tsang, W.; Hampson, R. F. J. Phys. Chem. Ref. Data 1986, 15, 1087.[3] Beckman, K. B.; Ames, B. N. Physiol. Rev. 1998, 78, 547.[4] Ondrey, G.; van Veen, N.; Bersohn, R. J. Chem. Phys. 1983, 78, 3732.[5] Docker, M. P.; Hodgson, A.; Simons, J. P. Chem. Phys. Lett. 1986, 128, 264.[6] Gericke, K-H.; Klee, S.; Comes, F. J.; Dixon, R. N. J. Chem. Phys. 1986, 85, 4463.[7] Klee, S.; Gericke, K.-H.; Comes, F. J. J. Chem. Phys. 1986, 85, 40.[8] Klee, S.; Gericke, K.-H.; Comes, F. J. Ber. Bunsenges Phys. Chem. 1988, 92, 429.[9] Zhang, D. H.; Zhang, J. Z. J. Chem. Phys. 1993, 98, 6276.[10] Riffault, V.; Gierczak, T.; Burkholder, J. B.; Ravishankara, A. R. Phys. Chem. Chem.

Phys. 2006, 8, 1079.[11] Thiebault, J.; Aluculesei, A.; Fittschen, C. J. Chem. Phys. 2007, 126, 186101-1.[12] Giorgio, M.; Trinei, M.; Migliaccio, E.; Pelicci, P. G. Nat. Rev. Mol. Cell Biol. 2007,

8, 722.[13] Meneghini, R.; Hoffmann, M. E. Biochem. Biophys. Acta 1980, 608, 167.[14] Finlayson-Pitts, B. J.; Pitts, J. N. Chemistry of the Upper and Lower Atmosphere,

Academic Press, San Diego, 2000.[15] Zuo, Y.; Deng, Y. Geochim. et Cosmochim. Acta 1999, 63, 3451.[16] Lelieveld, J.; Crutzen, P. J. J. Atmos. Chem. 1991, 12, 229.[17] Carlson, R. W.; Anderson, M. S.; Johnson, R. E.; Smythe, W. D.; Hendrix, A. R.;

Barth, C. A.; Söderblom, L. A.; Hansen, G. B.; McCord, T. B.; Dalton, J. B.; Clark, R.N.; Shirley, J. H.; Ocampo, A. C.; Matson, D. L. Science 283 1999, 283, 2062.

[18] Loeffer, M. J.; Baragiola, R. A. Geophys. Res. Lett. 2005, 32, L17202.[20] Clancy, R. T.; Sandor, B. J.; Moriarty-Schieven, G.-H. Icarus 2004, 168, 116.[19] Newman, S. F.; Buratti, B. J.; Jaumann, R.; Bauer, J. M.; Momary, T. W. Astrophys.

J. 2007, 670, L143.[21] Whittle, E.; Dows, D. A.; Pimentel, G. C. J. Chem. Phys. 1954, 22, 1943.[22] Hallam, H. E. Vibrational Spectroscopy of Trapped species, John Wiley & Sons, New

York, 1973.[23] Andrews, L.; Moskovits, M. Chemistry and Physics of Matrix-Isolated Species,

North-Holland, Amsterdam, 1989.[24] Almond, M. J.; Wiltshire, K. S..Annual Reports on the Progress of Chemistry, Section

C: Physical Chemistry 2001, 97, 3.[25] Goebel, J.; Ault, B. S.; Del Bene, J. E. J. Phys. Chem. A 2000 104, 2033.[26] Lundell, J.; Jolkkonen, S.; Khriachtchev, L.; Pettersson M.; Räsänen, M. Chem. Eur.

J. 2001, 7, 1670.[27] Goebel, J. R.; Ault, B. S.; Del Bene, J. E. J. Phys. Chem. A 2001, 105, 6430.[28] Goebel, J. R.; Ault, B. S.; Del Bene, J. E. J. Phys. Chem. A 2001, 105, 11365.[29] Engdahl A.; Nelander, B. Chem. Phys. 2003, 293, 203.[30] Goebel, J. R.; Antle, K. A.; Ault, B. S.; Del Bene, J. E. J. Phys. Chem. A 2002, 106,

6406.[31] Engdahl, A.;. Nelander, B. Phys. Chem. Chem. Phys. 2000, 2, 3967.

38

[32] Engdahl, A.;. Nelander, B.; Karlström, G. J. Phys. Chem. A 2001, 105, 8393.[33] Engdahl, A.; Nelander, B. Phys. Chem. Chem. Phys. 2004, 6, 730.[34] Schinke, R. Photodissociation Dynamics, Spectroscopy and Fragmentation of Small

Polyatomic Molecules, Cambridge monographs on atomic, molecular and chemicalphysics, Cambridge University Press, 1995.

[35] Bondybey, V. E.; Räsänen, M.; Lammers, A. Ann. Rep. Prog. Chem., Sect C; PhysChem. 1999, 95, 331.

[36] Apkarian, V. A.; Schwentner, N. Chem. Rev. 1999, 99, 1481.[37] Alimi, R.; Gerber, R. B.; Apkarian, V. A. J. Chem. Phys. 1988, 89, 174.[38] Gersonde, I. H.; Gabriel, H. J. Chem. Phys. 1994, 101, 9558.[39] Schriever, R.; Chergui, M.; Kunz, H.; Stepanenko, K. V.; Schwentner, N. J. Chem.

Phys. 1989, 91, 4128.[40] Khriachtchev, L.; Pettersson, M.; Isoniemi, E.; Räsänen, M. J. Chem. Phys. 1998,

108, 5747.[41] Kunttu, H.; Lawrence, W. G.; Apkarian, V. A. J. Chem. Phys. 1991, 94, 1692.[42] Zoval, J.; Apkarian, V. A. J. Chem. Phys. 1994, 98, 7945.[43] Ryan, E. T.; Weitz, E. J. Chem. Phys. 1993, 99, 8628.[44] Ryan, E. T.; Weitz, E. J. Chem. Phys. 1993, 99, 1004.[45] Lawrence, W. G.; Apkarian, V. A. J. Chem. Phys. 1992, 97, 6199.[46] Sander, S. P.; Friedl, R. R.; Golden, D. M.; Kurylo, M. J.; Huie, R. E.; Orkin, V. L.;

Moortgat, G. K.; Ravishankara, A. R.; Kolb, C. E.; Molina, M. J.; Finlayson-Pitts, B.J. Chemical Kinetics and Photochemical Data for Use in Atmospheric Studies,Evaluation No. 14, JPL Publications 02-25, 2003.

[47] CRC-Handbook of Chemistry and Physics, CRC Press, 88th edition, 2007-2008.[48] Schiffman, S.; Nelson, D. D.; Nesbitt, D. J. J. Chem. Phys. 1993, 98, 6935.[49] Atkinson, R.; Baulch, D. L.; Cox, R. A.; Crowley, J. N.; Hampson, R. F., Jr.; Hynes,

R. G.; Jenkin, M. E.; Rossi, M. J. Troe, J. Atmos. Chem Phys. 2004, 4, 1461.[50] Gerlach-Meyer, V.; Linnebach, E.; Kleinermanns, K.; Wolfrum J. Chem. Phys. Lett.

1987, 133, 113.[51] Nakayama, T.; Takahashi, K.; Matsumi, Y. Int. J. Chem. Kin. 2005, 37, 751.[52] Satterfield, C. N.; Kavanagh, G. M.; Resnick, H. Ind. Eng. Chem. 1951, 43, 2507.[53] Heaney, H. Aldrichim. Acta 1993, 26, 35.[54] Saarinen, P.; Kauppinen, J. Appl. Spectrosc. 1991, 45, 953.[55] Greenwood, N. N., Earnshaw, A. Chemistry of the elements, 1st ed., Pergamon Press,

Oxford, 1994, p 343.[56] Ball, M. C.; Massey, S. Thermochim. Acta 1995, 261, 95.[57] Elvers, B.; Hawkins, S.; Ravenscroft, M.; Schultz, G. Ullmans Encyclopedia of

Industrial Chemistry, 5th edn., VCH Verlagsgesellschaft mbH: Veinheim, 1989, VolA13, p. 462.

[58] Lu, C.-S.; Hughes, E. W.; Giguère, P. A. J. Am. Chem. Soc. 1941, 63, 1507.[59] Bain, O.; Giguère, P. A. Can. J. Chem. 1955, 33, 527.[60] Hunt, R. H.; Leacock, R. A.; Peters, C. W.; Hecht, K. T. J. Chem. Phys. 1965, 42,

1931.[61] Toth, R. A. Appl. Opt. 1991, 32, 7326.

39

[62] Perrin, A V.; Valentin, A.; Flaud, J.-M.; Camy-Peyret, C.; Schriver, L; Schriver, A.;Arcas, P. H. J. Mol. Spectrosc. 1995, 171, 358.

[63] Galapov, B. S.; T. Dudev, T. Vibrational Intensities, Elsevier, Amsterdam, 1996,p.185.

[64] Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.; Johnson, B. G.; Robb,M. A.; Cheeseman, J. R.; Keith, T.; Petersson, G. A.; Montgomery, J. A.;Raghavachari, K.; Al-Laham, M. A.; Zakrzewski, V. G.; Ortiz, J. V.; Foresman, J. B.;Closlowski, J.; Stefanov, B. B.; Nanayakkara, A.; Challacombe, M.; Peng, C. Y.;Ayalas, P. Y.; Chen, W.; Wong, M. W.; Andres, J.L.; Replogle, E. S.; Gomperts, R.;Martin, R. L.; Fox, D. J.; Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart, J. P.; Heads-Gordon, M.; Gonzalez, C.; Pople, J. A. Gaussian 94, Revision B.1, Gaussian, Inc.,Pittsburgh, PA, 1995.

[65] Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery J. A. Jr.; Stratmann, R. E.; Burant,J. C.; Dapprich, S.; Millam, J. M.; Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas,O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo,C.; Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma,K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.;Ortiz, J. V.; Baboul, A. G.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.;Komaromi, I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.;Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.;Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon, M.; Replogle,E. S.; Pople, J. A. Gaussian 98 (Revision A.9), Gaussian, Inc., Pittsburgh PA, 1998.

[66] Schmidt, M. W.; Baldridge, K. K.; Boatz J. A.; Elbert, S. T.; Gordon, M. S.; Jensen,J.H.; Koseki, S.; Matsunaga, N.; Nguyen, K. A.; Su, S. J.; Windus, T. L.; Dupuis, M.;Montgomery, J. A. J. Comp. Chem. 1993, 14, 1347.

[67] Nemukhin, A. V.; Grigorenko, B. L.; Granovsky, A. A. Moscow State Univ. Res.Bull. Khimia 2004, 45, 75.

[68] Jensen, F. Introduction to computational chemistry, John Wiley & Sons, 2nd edn.,2007.

[69] Lundell, J. J. Phys. Chem. 1995, 99, 14290.[70] Boys, S. F.; Bernardi, F. Mol. Phys. 1970, 19, 553.[71] Schimanouchi, T. Tables of Vibrational Frequencies Consolidated, National Bureau

of Standards 1972, 1, 1-160.[72] Engdahl, A.; Nelander, B., J. Mol. Spectrosc. 1989, 193, 101.[73] Ayers, G. P.; Pullin, A. D. E. Chem. Phys. Lett. 1974, 29, 609.[74] Ayers, G. P.; Pullin, A. D. E. Spectrochimica Acta 1976, 32A, 1641.[75] Ayers, G. P.; Pullin, A. D. E. Spectrochimica Acta 1976, 32A, 1689, 1695.[76] Ayers, G. P.; Pullin, A. D. E. Spectrochimica Acta 1976, 32A, 1629, 1695.[77] Bentwood, R. M.; Barnes, A.J.; Orville-Thomas, W. J. J. Mol. Spectrosc. 1980, 84,

391.[78] Lannon, J. A.; Verderame, F. D.; Anderson, R. W. Jr. J. Chem. Phys. 1971, 54, 2212.[79] Giguère, P. A.; Srinivasan, T. K. K. Chem. Phys. Lett. 1975, 33, 479.[80] Catalano, E.; Sanborn, R. H. J. Chem. Phys. 1963, 38, 2273.

40

[81] Giguère, P. J. Chem. Phys. 1949, 18, 88.[82] Redington, R. L.; Olson, W. B.; Cross, P. C. J. Chem. Phys. 1962, 36, 1311.[83] Hillman, J. J.; Jennings, D. E.; Olson, W. B.; Goldman, A. J. Mol. Spectrosc. 1986,

117, 46.[84] Olson, W. B.; Hunt, R. H.; Young, B. W.; Maki, A. G.; Brault J. W. J. Mol.

Spectrosc. 1988, 127, 12.[85] Perrin, A.; Flaud, J.-M.; Camy-Peyret, C.; Goldman, A.; Murcray, F. J.; Blatherwick,

R. D. J. Mol. Spectros. 1990, 142, 129.[86] Hougen, J. T. Can. J. Phys. 1984, 62, 1392.[87] Vincent, M. A.; Palmer, I. J.; Hillier, I. H.; Akhmatskaya, E. J. Am. Chem. Soc. 1998,

120, 3431.[88] Svensson T.; Nelander, B. Chem. Phys. 2003, 286, 347.[89] Maillard, D. Allavena, M.; Perchard, J. P. Spectrochimica Acta 1975, 31A, 1523.[90] Tso, T.-L.; Lee, E. K. C. J. Phys. Chem. 1984, 88, 2776.[91] Wang, B.; Hou, H. Chem. Phys. Lett. 2005, 410, 235.[92] Okabe, H. Photochemistry of small molecules, Wiley, New York, 1st edn., 1978.[93] Goodman, J.; Brus, L. E. J. Chem. Phys. 1977, 67, 4858.[94] Cheng M. B.-M.; Lee, Y.-P.; Ogilvie, J. F. Chem. Phys. Lett. 1988, 151, 109.[95] Engdahl, A.; Nelander, B. Chem. Phys. 1985, 100, 273.[96] Engdahl, A.; Nelander, B. J. Chem. Phys. 1987, 86, 1819.[97] Fajardo, M. E.; Apkarian, V. A. J. Chem. Phys. 1986, 85, 5660.[98] Fajardo, M. E.; Apkarian, V. A. J. Chem. Phys. 1988, 89, 4102.[99] Sennikov, P. G.; Ignatov, S. K.; Schrems, O. Chem. Phys. Chem. 2005, 6, 392.[100] Huang, H. H.; Xie, Y.; Schaefer III, H. F. J. Phys. Chem. 1996, 100, 6076.[101] Xie, Y.; Allen, W. D.; Yamaguchi, Y.; Shaefer III, H. F. III J. Chem. Phys. 1996,

104, 7615.[102] Liu, Y.-J.; Persson, P.; Lunell, S. Mol. Phys. 2004, 102, 2575.[103] Khriahtchev, L.; Pettersson, M.; Pehkonen, S.; Isoniemi, E.; Räsänen, M. J. Chem.

Phys. 1999, 111, 1650.[104] Alimi, R.; Brokman, A.; Gerber, R. B. J. Chem. Phys. 1989, 91, 1611.[105] Tarasova, E. I.; Ratner, A. M.; Stepanenko, V. M.; Fugol, I. Ya.; Chergui, M.;

Schriever, R.; Schwentner, N. J. Chem. Phys. 1993, 98, 7786.[106] Milligan, D. E.; Jacox, M. E. J. Chem. Phys. 1963, 38, 2627.[107] Jacox, M. E.; Milligan, D. E. J. Mol. Structr. 1972, 42, 495.[108] Smith, D. W.; Andrews, L. J. Chem. Phys. 1972, 56, 4433.[109] Bondybey, V. E.; English, J. H. J. Mol. Spectrosc. 1985, 109, 221.[110] Lapinski, A.; Spanget-Larsen, J.; Waluk, J.; Radziszewski, J. J. Chem. Phys. 2001,

115, 1757.[111] Lawrence, W. G. and Apkarian, V. A. J. Chem. Phys. 1992, 97, 2224.[112] Redington, R. L.; Milligan, D., E. J. Chem. Phys. 1963, 39, 1276.[113] Adams, S. F.; DeJoseph, C. A., Jr.; Carter, C. C.; Miller, T. A.; Williamson, J. M. J.

Phys. Chem. A 2001, 105, 5977.[114] Danilychev, A. V.; Apkarian, V. A. J. Chem. Phys. 1993, 99, 8617.[115] Mohammed, H. H. J. Chem. Phys. 1990, 93, 412.

41

[116] Bhardwaj, A.; Haider, S. A. Adv. Space Res. 2002, 29, 745.[117] Gimmler, G.; Havenith, M. J. Mol. Spectrosc. 2002, 216, 315.[118] Spoliti, M.; Nunziante Cesaro, S.; Mariti, B. J. Chem. Phys. 1973, 59, 985.[119] Kiviniemi, T.; Pettersson, M.; Khriachtchev, L.; Räsänen, M.; Runeberg, N. J.

Chem. Phys. 2004, 121, 1839.[120] Kruger, H.; Weitz, E. J. Chem. Phys. 1992, 96, 2846.[121] Lawrence, W. G.; Apkarian, V. A. J. Chem. Phys. 1992, 97, 2229.[122] Ford, M. B.; Foxworthy; A. D.; Mains, G. J.; Raff, L. M. J. Phys. Chem. 1993, 97,

12134.[123] Danilychev, A. V.; Apkarian, V. A., J. Chem. Phys. 1994, 100, 5556.[124] Benderskii, A. V.; Wight, C. A. J. Chem. Phys. 1996, 104, 85.[125] Ning, X.-J.; Zhang, L.-N.; Chen, M.-H.; Qin, Q.-Z. J. Chem. Phys. 2000, 112, 386.[126] Danilychev, A. V.; Apkarian, V. A. Low. Temp. Phys. 2000, 26, 669.[127] Ning, X.-J.; Qin, Q.-Z. J. Chem. Phys. 2001, 114, 9969.