An Investigation of the Thermal and Photochemical Reaction Mechanisms of Cycloalkenes and Ferrocenes with Ozone by Matrix Isolation Spectroscopic Analysis and Theoretical Calculations

A dissertation submitted to the

Graduate School

of the University of Cincinnati

in partial fulfillment of the requirements for the degree of

Doctorate of Philosophy

in the Department of Chemistry

of the College of Arts and Sciences

March 20, 2015

by

Laura Pinelo

B.A., Chemistry, Illinois Wesleyan University, 2009

Chair: Dr. Bruce S. Ault

Abstract

The thermal and photochemical reactions of cycloheptene, 1,3- and 1,4-cyclohexadiene with ozone have been studied using a combination of matrix isolation, infrared spectroscopy, and theoretical calculations. The reaction of cycloheptene with ozone resulted in the first observation of early intermediates for a cyclic alkene with conformational isomerism. Spectral evidence supports the presence of the primary ozonide of the chair and boat conformers of cis- cycloheptene, which represents the first time two primary ozonide isomers have been observed for any alkene. In addition, at least one conformer of the Criegee intermediate, as well as stable end products were observed spectroscopically. Experimental and theoretical results for the reaction of ozone with 1,4-cyclohexadiene demonstrate that this reaction does not follow the long-accepted Criegee mechanism. Rather, the reaction of ozone with 1,4-cyclohexadiene leads to the essentially barrierless formation of benzene, C6H6, and H2O3. In addition, it was determined that the reaction of ozone with 1,3-cyclohexadiene follows two pathways, one of which is the Criegee mechanism through a low energy transition state leading to formation of the primary ozonide. With a similar barrier, ozone can also abstracts a single hydrogen from C5 while adding to C1 of 1,3-cyclohexadiene, forming a hydroperoxy intermediate. The study of

1,3- and 1,4-cyclohexadiene presents two of the rare cases in which the Criegee mechanism is not the dominant pathway for the ozonolysis of an alkene as well as the first evidence for dehydrogenation of an alkene by ozone.

The low-energy photochemical reaction of ozone and n-butylferrocene has been studied using a combination of argon-matrix isolation, infrared spectroscopy, and theoretical calculations. The results support the photodissociation of ozone by red light (λ ≥ 600 nm) which produces an atomic oxygen, O(3P), and a molecular oxygen. O(3P) reacts with n-butylferrocene

iii to form products consisting of an iron atom with a coordinated n-butylcyclopentadienyl or cyclopentadienyl ring and either: (1) a pyran, (2) an aldehyde, or (3) a bidentate cyclic aldehyde with a seven-membered ring including the iron atom. The dark deposition reactions of ozone with ferrocene and with n-butylferrocene were studied using matrix isolation, UV-Vis spectroscopy, and theoretical calculations. The co-deposition of ferrocene with ozone and that of n-butylferrocene with ozone led to production of green charge transfer complexes. These charge transfer complexes underwent photochemical reactions upon irradiation with red light (λ ≥ 600 nm). The MO analysis of the long wavelength transitions indicated that the formation of the charge transfer complex with ferrocene or n-butylferrocene affects the how readily the π*-orbital on ozone is populated when red light (λ ≥ 600 nm) is absorbed. All findings given are supported by 18O-labeled ozone infrared experiments and literature spectra. In addition further justification was provided by theoretical calculations at the B3LYP/6-311++G(d,2p) level. Theoretical UV-

Vis spectra were calculated with TD-DFT using a B3LYP functional and the 6-311G++(d,2p) basis set.

iv

v Acknowledgments

To Dr. Bruce Ault, I would like to thank you for all of your patience and unwavering support. Your leadership and guidance has added considerably to my graduate experience. I am very fortunate to have an advisor who is so dedicated to sharing his knowledge and helping his student achieve their full potential. I would like to acknowledge my committee members, Dr.

Thomas Beck and Dr. Anna Gudmundsdottir, for all of their guidance and input on my research.

A special thanks to Dr. Roger Kugel, I am grateful to have had the opportunity to work with you and for all the advice you have provided. I have benefited greatly from your wisdom, instruction, and assistance in editing this thesis.

To Dr. Narendra Jaggi, my undergraduate advisor in physics and life at Illinois Wesleyan

University for your unwavering support. You taught me to how to expand my perception of a problem and consequently to find more solutions.

I would like to acknowledge my undergraduate advisor in chemistry at Illinois Wesleyan

University, Dr. Rebecca Roesner, for her steadfast support and for encouraging me to apply for

Research Experiences for Undergraduates (REU).

To Dr. David Anderson, for giving me the opportunity to work with him during the REU

Program in Energy Science at the University of Wyoming and for introducing me to matrix isolation. In addition for taking me, at the time just a college sophomore, to the Gordon Research

Conference for Physics and Chemistry of Matrix Isolated Species were I would meet Dr. Ault and numerous other contacts.

vi I would like to acknowledge Dr. Sharon Kettwich, my best friend and mentor for all of her support. To Joel Collett without whose support, encouragement and editing assistance, I would not have finished this thesis.

Lastly and most importantly I would like to thank my parents Polly and Armando Pinelo, for their steadfast support, encouragement, and resolute confidence in my potential that they have provided me through my entire life.

Laura Frances Pinelo May 2015

vii

For my parents, Polly and Armando Pinelo

viii Table of Contents

Page

Abstract ii Acknowledgements vi Dedication viii Table of Contents ix List of Figures x List of Tables xiv

Chapter 1 An Introduction to the Ozone, Alkenes, Ferrocenes, and 1 Matrix Isolation

Chapter 2 Experimental Details 15

Chapter 3 Computational Details 24

Chapter 4 Infrared Matrix Isolation and Theoretical Study of the 25 Initial Intermediates in the Reaction of Ozone with Cycloheptene

Chapter 5 Matrix Isolation Study of the Ozonolysis of 1,3- and 1,4- 46 Cyclohexadiene: Identification of Novel Reaction Pathways

Chapter 6 Low-energy Photochemistry of Ozone and n- 72 Butylferrocene: A matrix isolation study

Chapter 7 Charge Transfer Complexes and Photochemistry of Ozone 92 with Ferrocene and n-Butylferrocene: A UV-Vis matrix isolation study

Appendix A Supporting Information for Chapter 4 118

Appendix B Supporting Information for Chapter 5 120

Appendix C Supporting Information for Chapter 6 127

Appendix D Supporting Information for Chapter 7 168

ix List of Figures

Figure Description Page

1.1 Criegee mechanism for the ozonolysis of alkenes. 3

1.2 Calculated optimized structures of the major products of the 7 photochemical reaction of ozone (O3) with ferrocene: pyran (a) and aldehyde (b).

2.1 Cell, cold window, and cold head setup diagram. 17

2.2 The sample preparation manifolds. 18

2.3 Twin jet deposition mode. 20

2.4 Merged jet deposition mode. 21

4.1 Criegee mechanism for the ozonolysis of cycloheptene. 27

4.2 Computed structures of key intermediates in the ozonolysis of 29 cycloheptene.

4.3 Infrared spectra from 840 to 1020 cm-1 of matrices formed by the twin jet 32 codeposition of samples of Ar/C7H12 and Ar/O3. Red (middle) is initial deposition and blue (top) is after annealing to 35 K, compared to a blank spectrum of Ar/C7H12 (pink, bottom). Bands marked with an asterisk (*) are product bands.

4.4 Infrared spectra from 1500 to 1900 cm-1 of matrices formed by the twin jet 33 16 18 codeposition of samples of Ar/C7H12 with Ar/ O3, and Ar/ O3. Blue 16 (second from top) is after annealing with O3 while red (top) is after irradiation. Product bands are denoted by an asterisk (*). The lower two 18 traces (pink and green) show the corresponding traces with O3. In the green trace, the spike at 1700 cm-1 is from an electronic interference. The band profile at 1698 cm-1 is nonetheless clear.

4.5 Infrared spectrum of a matrix formed by the merged jet codeposition of a 40 samples of Ar/cycloheptene and Ar/O3 (blue) compared to a blank spectrum of Ar/cycloheptene (red). An electronic noise spike is noticeable at 1700 cm-1.

x 5.1 Infrared spectra from 604 to 1495 cm-1 of matrixes formed by the twin jet 49 codeposition of samples of Ar/1,4-C6H8 and Ar/O3. Red (middle) is initial deposition and blue (top) is after annealing to 35 K, compared to a blank spectrum of Ar/1,4-C6H8 (pink, bottom). Bands marked with an asterisk (*) are product bands

5.2 Calculated structure of the H2O3–benzene complex formed in the reaction 53 of 1,4-cyclohexadiene with ozone. The calculated energy relative to the reactants is given in parentheses.

5.3 Reaction scheme for ozone with 1,4-cyclohexadiene. 55

5.4 Reaction diagram (kcal/mol) for the conversion of the initial weak 56 complex of 1,4-cyclohexadiene and ozone to the secondary ozonide via the Criegee mechanism; as well as the conversion of the initial weak complex of 1,4-cyclohexadiene and ozone to the benzene–H2O3 complex (BHC).

5.5 Infrared spectra from 750 to 1460 cm-1 of matrixes formed by the twin jet 59 codeposition of samples of Ar/1,3-cyclohexadiene and Ar/O3. Red is initial deposition and blue is after annealing to 35 K, compared to a blank spectrum of Ar/1,3-cyclohexadiene (pink). Bands marked with an asterisk are product bands.

5.6 Calculated structures for early intermediates from the monoozonolysis 62 reaction of 1,3-cyclohexadiene. The calculated energies for each intermediate relative to the reactants are given in parentheses.

5.7 Reaction scheme for ozone with 1,3-cyclohexadiene. 64

5.8 Reaction diagram (kcal/mol) for the conversion of the initial weak 67 complex of 1,3-cyclohexadiene and ozone to the secondary ozonide via the Criegee mechanism, as well as the conversion of the initial weak complex of 1,3-cyclohexadiene and ozone to intermediate PI (hydroperoxy intermediate).

6.1 (a) Gaussian optimized structure of n-butylferrocene (nBuFc). (b) The 73 charge density of and the designations for the nBuFc ring carbons.

xi 6.2 Partial infrared absorption spectra (a) (535-1150 cm-1), (b) (1350-1490 cm- 75 1 -1 16 ), and (c) (1570-1700 cm ) of: Ar/ O3/n-butylferrocene after 1 hour red 16 (λ ≥ 600 nm) irradiation (red trace), dark deposited Ar/ O3/n- 16 butylferrocene (green trace), Ar/n-butylferrocene (black trace), Ar/ O3 18 (pink trace), and Ar/ O3/n-butylferrocene after 30 min. red (λ ≥ 600 nm) irradiation (blue trace). Peak assignments: P = pyran, A = aldehyde , R = ring-aldehyde, U = unassigned.

6.3 The regioisomers for the proposed photochemical products: (a) pyrans A- 78 F, (b) (E)- and (Z)-aldehydes A-I, (c) ring-aldehydes A-E, and (d) ketones.

6.4 E/Z conformational isomers of aldehyde-A (ald-A). (a) The E 80 conformation of ald-A with the hydrogen of the aldehyde group and n- butyl group (or the hydrogen, in the case of the other aldehydes) of the adjoining carbon are on opposite sides. (b) The Z conformation of ald-A with the hydrogen of the aldehyde group and n-butyl group (or the hydrogen, in the case of the other aldehydes) of the adjoining carbon are on same sides.

-1 6.5 Partial (1570-1720 cm ) infrared absorption spectra of Ar/O3/nBuFc 84 showing the most intense peaks assigned to the ring-aldehyde, R, at 1583 cm-1 and to the aldehyde, A, at 1680 cm-1 in the blue (middle) trace is after 30 min. of irradiation with λ ≥ 1000 nm and then followed by 30 min. of irradiation with λ ≥ 600 nm shown in the red (top) trace. The black (bottom) trace is the n-butylferrocene blank.

6.6 Calculated (a) potential energy diagram and (b) structures for n- 85 butylferrocene + O(3P) → pyran-A + ald-A.

6.7 The full reaction scheme proposed when O(3P) reacts with C2 of n- 88 butylferrocene to form triplet van der Waals complex.

7.1 The dark TJ-deposition spectra in green, the reactants subtracted from 95 them (O3 or O2 in blue and ferrocene, Fc, or n-butylferrocene, nBuFc, in red) and resulting difference spectra in black. (a) Ar/Fc/O3, (b) Ar/nBuFc/O3, (c) Ar/Fc/O2, and (c) Ar/nBuFc/O2.

7.2 Calculated structures and the ground state energy relative to the ground 96 state parent species of the charge transfer complexes of O3 with Fc (a and b) and with nBuFc (c and d).

xii 7.3 Calculated structures and the ground state energy relative to the ground 99

state parent species of the van der Waals complexes of O2 with Fc (a) and with nBuFc (b).

7.4 Spectra of the reactions of O3 or O2 with Fc: (a) Ar/Fc/O3 difference 100 spectrum, (b) calculated TD-DFT spectrum of Fc-O3A, (c) calculated TD- DFT spectrum of Fc-O3B, (d) Ar/Fc/O2 difference spectrum, and (e) calculated TD-DFT spectrum of Fc-O2.

7.5 Spectra of the reactions of O3 or O2 and nBuFc: (a) Ar/nBuFc/O3 101 difference spectrum, (b) calculated TD-DFT spectrum of nBuFc-O3A, (c) calculated TD-DFT spectrum of nBuFc-O3B, (d) Ar/nBuFc/O2 difference spectrum, and (e) calculated TD-DFT spectrum of nBuFc-O2.

7.6 Experimental spectra of the reactions of O3 or O2 with ferrocene (Fc) or n- 103 butylferrocene (nBuFc). The dark deposition spectra are in green, spectra after 30 min of red light photolysis are in blue, and spectra after 45 min of red light photolysis are in black. (a) Ar/Fc/O3 and an enlarged region form 550-900 nm (b) Ar/nBuFc/O3 and an enlarged region form 500-900 nm (c) Ar/Fc/O2 (d) Ar/nBuFc/O2.

7.7 The optimized Gaussian 09 structures for the major photochemical 105 products of ferrocene (Fc) and n-butylferrocene (nBuFc): (a) Fc aldehyde (Fc-ald), (b) nBuFc aldehyde (nBuFc-ald), (c) Fc pyran (Fc-pyran), and (d) nBuFc pyran (nBuFc-pyran).

7.8 The difference spectra from 45 min. of red irradiation minus the dark 107 deposition spectra for the (a) Ar/Fc/O3 and (d) Ar/nBuFc/O3 experiments. Calculated TD-DFT spectrum of: (b) Fc-ald, (c) Fc-pyran, (e) nBuFc-ald, and (f) nBuFc-pyran.

7.9 Calculated π*-orbital and n-orbitals of (a) Fc-O3A and (b) Fc-O3B. 109

7.10 Calculated π*-orbital and n-orbitals of (a) nBuFc-O3A and (b) nBuFc- 111 O3B.

7.11 Calculated π*-orbital and n-orbitals of O3. 114

xiii List of Tables

Table Description Page

4.1 Band positions and Assignments for the Initial Intermediates in the Thermal 28 Reaction of Ozone with Cycloheptene.

4.2 Computed Energies of Species Relevant to the Reaction of Cycloheptene 30 with Ozone.

4.3 Band Positions and Assignments for the Initial Intermediates in the 34 Photochemical Reaction of Ozone with Cycloheptene.

4.4 Band Positions and Assignments for the Initial Intermediates in the Merged 41 Jet Reaction of Ozone with Cycloheptene.

5.1 Band Positions and Assignments for the Products in the Thermal Reaction of 48 Ozone with 1,4-Cyclohexadiene.

5.2 Band Positions and Assignments for the Initial Intermediates in the Thermal 58 Reaction of Ozone with 1,3-Cyclohexadiene.

6.1 Band Positions and Assignments for the Products from the Twin Jet 82 Deposition of Ozone with n-Butylferrocene upon Irradiation with Light of λ ≥ 600 nm.

7.1 Calculated Oxygen–Oxygen Bond Lengths of O3. 98

7.2 Assignment of Calculated Adsorption Bands and the Corresponding 110 Excitation Energies of Fc-O3A and Fc-O3B to the Ar/Fc/O3 Experimental Peak at λmax ≈ 765 nm.

7.3 Assignment of Calculated Adsorption Bands and the Corresponding 112 Excitation Energies of nBuFc-O3A and nBuFc-O3B to the Ar/nBuFc/O3 Experimental Peak at λmax ≈ 816 nm.

7.4 Calculated Adsorption Bands and the Corresponding Excitation Energies of 115 O3.

xiv Chapter 1 An Introduction to Ozone, Alkenes, Ferrocenes, and Matrix Isolation

This dissertation will examine two areas of ozone chemistry. The first topic of interest is the thermal and photochemical reactions of ozone with alkenes.1,2 There has been considerable

3 interest in ozone (O3) over the last few decades. O3 is a highly reactive oxidizing agent. It has industrial applications4 and it plays an important role in our atmosphere as an ultraviolet (UV)

1,5 radiation shield. The highest concentration of O3 in the atmosphere is in the stratosphere, which lies between 10 and 50 km above the surface of the earth and contains the ozone layer.6

The O3 layer is essential to life because O3 absorbs harmful UV solar radiation (200 nm ≤ λ ≤

300 nm).7 Ozone is also present in the lowest level of the atmosphere known as the

1 troposphere. The troposphere extents from ground level to ~11 miles above the earth. While O3 in the ozone layer is a beneficial UV radiation shield; ozone present in the troposphere is harmful to living things. In humans, O3 can cause a number of health problems, such as difficulty breathing, reduced lung function, and, with repeated exposure, it can permanently scar lung tissue. In addition, tropospheric ozone can cause reduced agricultural crop yields and lower the

8 growth and survival rate of tree seedlings. Most of the tropospheric O3 is the result of human

9 activates and is classified as a secondary pollutant. Tropospheric O3 is known as a secondary pollutant because this O3 is produced by the chemical reaction of nitrogen oxides (NOx) and volatile organic compounds (VOCs) initiated by sunlight. Emissions from power plants, vehicles and industrial facilities are major sources of NOx and anthropogenic VOCs. However, the majority of VOCs are biogenic and come from vegetation. VOCs included alkanes, alkenes, and

10 other hydrocarbons. While O3 does not react significantly with alkanes; it does readily react with alkenes in the troposphere from both biogenic and anthropogenic sources. It should be

1 noted that solar radiation either directly or indirectly is a significant factor in the chemistry of the atmosphere. O3 can undergo photodecomposition resulting in the formation of a highly reactive atomic oxygen and molecular oxygen.11-13 As a result, both the thermal and the photochemical reactions of ozone were investigated in this work. Studies of anthropogenic and biogenic alkene ozonolysis in the atmosphere have concluded that these reactions have a complex impact on air pollution and the overall composition of the atmosphere.14-19 Many of these studies cite the need for a more complete understanding of the mechanism of the ozone-alkene reaction.14-17

The reaction mechanism for ozone with alkenes first proposed by Criegee in 194920,21 is widely accepted19,22,23 based on considerable indirect experimental evidence and theoretical calculations (Figure 1.1).24-34 It should be noted that other mechanisms have been proposed and some studies have observed non-Criegee products, such as in the solution phase ozonolysis of

1,3-cyclohexadiene where a small amount of phenol was produced.22,35 The first step of the

Criegee mechanism involves the formation of the primary ozonide (POZ) by a 1,3 polar addition across the double bond of the alkene yielding a 1,2,3-trioxolane. For a reaction where the starting alkene is acyclic, the POZ decomposes in the second step to the Criegee intermediate

(CI), a carbonyl oxide, and an aldehyde or ketone.22,30 When the starting alkene is a cycloalkene, the POZ decomposes into a ring-opened CI where the aldehyde/or ketone portion remains at one end of the molecule with the carbonyl oxide at the other end. In the third step, a secondary ozonide (SOZ) or 1,2,4-trioxolane can form by the recombination of the two ends in the case of a cyclic alkene. For a CI formed from a straight-chain alkene, a SOZ can form when the separate aldehyde/or ketone and carbonyl oxide molecules recombine.22 Theoretical and experimental studies show that the initial step of the reaction is very exothermic; the resulting POZ possesses excess energy of about 50 kcal/mol.

2

3 The activation energy required for the POZ to form the CI was calculated to be approximately 19 kcal/mol. Thus, there is sufficient energy for the reaction to proceed to a range of final products.22,25,28-30,32-34 In order to characterize the POZ, CI, and/or SOZ and provide verification of the Criegee mechanism, these species must be isolated and the excess energy must be dissipated.30,35,36 Matrix isolation has been successfully used to isolate and characterize a wide variety of reactive intermediates; such as radicals, ions, and molecular complexes.36,37 This technique has successfully led to the isolation and characterization of the primary ozonide, the secondary ozonide, and/or Criegee intermediate for several simple alkenes.29,30,38

While cycloheptene, 1,3-cyclohexadiene and 1,4-cyclohexadiene, the alkenes investigated here, are not prevalent in the atmosphere; a detailed reaction mechanism for the ozonolysis of these cycloalkenes could provide critical insights into the mechanistic features of other ozone-cycloalkene reactions and contribute to the knowledge of tropospheric chemistry.39

There have been a few mechanistic studies of the ozonolysis of cycloalkenes; primarily these studies relied on indirect experimental evidence and theoretical calculations.18,24,30,35,39-41 In addition, cycloheptene has multiple conformers42 which have the potential to add extra complexity not seen in the cycloalkene-ozone reactions previously studied by this group.30,41

The goal of this study was to isolate and characterize the early intermediates and late stable products formed by the thermal and photochemical reactions O3 with cycloheptene, 1,3- cyclohexadiene and 1,4-cyclohexadiene. Matrix isolation coupled with Fourier transform infrared (FT-IR) spectroscopy and theoretical calculations were used to isolate and characterize the reaction intermediates and final products. Earlier work of this group studied the reactions of

30,41 cyclopentene, cyclohexene, and cyclopentadiene with O3. A further goal of this project was

4 to compare and contrast the results of the studies of O3 with cycloheptene, 1,3-cyclohexadiene and 1,4-cyclohexadiene to those of the previously studied cycloalkenes.

The second area of ozone chemistry investigated in this work focused on the thermal and photochemical reactions of ozone with ferrocenes. The use of organometallics, such as ferrocene, as precursors for the formation of metal oxide thin films has become an area of intense study over the last decade.43-47 In general, metal oxide thin films play an important part in our lives; they are in flat panel displays, energy saving window coatings, solar cells, and a variety of other devices.48-50 Thin films are primarily generated by techniques that are either a form of chemical vapor deposition (CVD) or physical vapor deposition (PVD).51-53 PVD techniques, such as evaporation or sputtering, involve the adsorption of the vaporized material onto the substrate to form the thin film.51 CVD is a process that involves the formation of a thin film of solid reaction product on a substrate by the reaction of gas phase precursors.52,53 In CVD metal oxide thin films may be prepared by reacting a volatile organometallic compound with an oxygen source.53 Increasingly, ozone is being used as the oxygen source because it has a high oxidizing potential, it lacks hydrogens, it is volatile, and it is highly reactive leading to rapid and effective metal oxide thin film formation.3,54-56

Thin films containing iron oxide have been of particular interest and the subject of numerous research papers. They have been used in photochemical cells, batteries, and gas sensors.57-62 Metallocenes are one class of iron precursors used in the CVD of iron-oxide- containing thin films.44 This class includes ferrocene (Fc) and n-butylferrocene (nBuFc), which have been investigated for use as an iron precursor in CVD.43,44,63 n-Butylferrocene, like ferrocene, has several advantages over other iron precursors since it is relatively inexpensive, thermally stable, neither air nor moisture sensitive, and vaporizes cleanly. In addition, nBuFc has

5 the added advantage of being liquid at room temperature.44 n-Butylferrocene has been used in

CVD as a co-precursor with nickelocene and oxygen to form a thin film of nickel ferrite

(NiFe2O4). The nickel ferrite thin film is of particular interest due to its potential for application in microelectronic devices.63

Despite the importance of metal oxide thin films, little is known about the chemical reactions of ozone with the organometallic compounds used to make these films.64,65 Matrix isolation has been successfully used in conjunction with theoretical calculations to determine the key initial steps in the reaction mechanisms of a number of chemical systems.66-68 Recently a

69 matrix isolation study of ferrocene (Fc) and O3 was conducted by Kugel, et al. This study used matrix-isolation combined with IR spectroscopy and theoretical calculations to show that a photochemical reaction of O3 with Fc occurs upon irradiation with red light (λ ≥ 600 nm). This low-energy photochemical reaction leads to the production of atomic oxygen O(3P) which subsequently reacts with Fc. The major products consist of an iron cyclopentadienyl ring moiety and either a pyran or an aldehyde (Figure 1.2). A dark green matrix and slightly red-shifted O3 infrared absorptions observed prior to irradiation were theorized to be indicative of the formation a charge transfer complex between O3 and Fc (Fc-O3). The green color of the matrix suggested strong red and blue absorptions in the visible spectrum.69 The slightly red-shifted infrared

37 absorptions of ozone are characteristic of perturbed O3. There have been a few studies that have investigated the generation of atomic oxygen by the photodissociation upon red-visible irradiation of small charge transfer complexes involving O3. For example, the irradiation of the

70,71 charge transfer complex O3:Br2 led to the identification of new halogen oxide compounds.

6

7 In addition, the photochemistry of O3 in the visible range, specifically in the Chappuis band (420 nm ≤ λ ≤ 700 nm), has been of particular interest due to application in atmospheric O3 monitoring.13 However, the reaction mechanisms of metallocenes with ozone are largely uninvestigated and initial products formed in these reactions remain unknown.

In general, the goal of this second area of study was to determine the initial products formed when O3 reacts with n-butylferrocene. Earlier work in this lab focused on the

69 photochemical products formed by O3 and ferrocene upon red irradiation, as discussed above.

For the study of the reaction of n-butylferrocene and O3 the goals were fourfold: first, to use matrix isolation coupled with FT-IR spectroscopy to isolate and characterize the photochemical products formed upon irradiation of the matrix; second, to use matrix isolation coupled with UV-

Vis spectroscopy to isolate, characterize, and identify the initial products formed during dark deposition of ferrocene and n-butylferrocene with O3 prior to irradiation; third, to probe the electronic transitions of the initial products formed during the dark deposition of ferrocene and n- butylferrocene with O3 involved in the formation of the photochemical products upon irradiation through the use UV-Vis spectroscopy and computational methods; and fourth, to compare and contrast the results for the reaction of O3 with n-butylferrocene to those of the reaction of O3 with ferrocene and evaluate the effect of the n-butyl group on the reaction. n-Butylferrocene was specifically chosen as the substituted ferrocene to investigate since it can be easily vaporized.

The technique of matrix isolation has been successfully used to isolate and characterize a wide variety of reactive intermediates; such as radicals, ions, and molecular complexes. Matrix isolation achieves this by trapping these reactive species in a cage or matrix made of a chemically inert substance (such as argon) at cryogenic temperatures (for an argon matrix ~12-15

K). Both the inert matrix and the cryogenic temperatures contribute to dissipate the excess

8 energy of the reactive species and stabilize them. Typically, reactants are diluted with a large excess of the inert substance (often argon with dilution rations between 100/1 and 1000/1). This gives the resulting matrix enough rigidity to prevent the diffusion of trapped species and bimolecular reactions from occurring.36,37

The species trapped in the matrix can be characterized using spectroscopic methods, such as infrared and UV-Vis spectroscopy. The primary method of analysis used during the course of this research was FT-IR spectroscopy. This method of spectroscopic analysis is advantageous since it benefits from of the band sharpening (bandwidth are typically ~1 cm-1) that results from the low temperatures and inert environment of the matrix.

A brief summary of this dissertation is as follows: Chapter 2 provides an in-depth description of the matrix isolation technique, sample preparation and deposition methods, photolysis procedures, and spectroscopic methods. In Chapter 3 computational methods used to augment the experimental work will be examined. Chapter 4 will explore the thermal and photochemical reaction of O3 and cycloheptene. Chapter 5 covers thermal and photochemical reactions of O3 with 1,3-cyclohexadiene and 1,4-cyclohexadiene. In Chapter 6 the photochemical reaction of O3 with n-butylferrocene will be discussed. Finally, Chapter 7 will cover the results of a UV-Vis study on the dark deposition products formed during the reaction of

O3 with ferrocene and n-butylferrocene.

References

(1) Horváth, M.; Bilitzky, L.; Hüttner, J. Ozone; Elsevier: Amsterdam; New York, 1985.

(2) Society, A. C.; Leedy, H. A. Ozone Chemistry and Technology; American Chemical Society, 1959.

9 (3) Sundstrom, H.; Gottschalk, C. The Evolution of Ozone Subsystems. Semiconductor International 2006, 2.

(4) Cook, G. A. Industrial Uses of Ozone. J Chem Educ 1982, 59, 392.

(5) Finlayson-Pitts, B. J.; Pitts, J. N. Chemistry of the Upper and Lower Atmosphere: Theory, Experiments, and Applications; Academic Press, 2000.

(6) Taylor, F. W. The Stratosphere. Philosophical Transactions: Mathematical, Physical and Engineering Sciences 2003, 361, 11.

(7) Daumont, D.; Brion, J.; Charbonnier, J.; Malicet, J. Ozone UV spectroscopy I: Absorption cross-sections at room temperature. J Atmos Chem 1992, 15, 145.

(8) EPA, U. S. Ozone: Good Up High, Bad Nearby; Government Printing Office: Washington, 2003.

(9) Marston, G. Tropospheric ozone. Physical chemistry chemical physics : PCCP 2011, 13, 10847.

(10) Goldstein, A. H.; Galbally, I. E. Known and unknown organic constituents in the Earth' s atmosphere. Environ Sci Technol 2007, 41, 1514.

(11) Leighton, P. Photochemistry of Air Pollution; Academic Press: New York, 1961.

(12) Bahou, M.; SchriverMazzuoli, L.; CamyPeyret, C.; Schriver, A. Photolysis of ozone at 693 nm in solid oxygen. Isotopic effects in ozone reformation. Chemical Physics Letters 1997, 273, 31.

(13) Anderson, S. M.; Mauersberger, K. Laser Measurements of Ozone Absorption Cross-Sections in the Chappuis Band. Geophysical Research Letters 1992, 19, 933.

(14) Atkinson, R.; Tuazon, E. C.; Aschmann, S. M. Products of the gas-phase reactions of O3 with alkenes. Environ Sci Technol 1995, 29, 1860.

(15) Ariya, P. A.; Sander, R.; Crutzen, P. J. Significance of HOx and peroxides production due to alkene ozonolysis during fall and winter: A modeling study. Journal of Geophysical Research 2000, 105, 17721.

(16) Paulson, S. E.; Orlando, J. J. The reactions of ozone with alkenes: An important source of HOx in the boundary layer. Geophysical Research Letters 1996, 23, 3727.

(17) Ravishankara, A. R. Introduction: Atmospheric chemistry--long-term issues. Chem Rev 2003, 103, 4505.

(18) Hatakeyama, S.; Ohno, M.; Weng, J. H.; Takagi, H.; Akimoto, H. Mechanism for the Formation of Gaseous and Particulate Products from Ozone-Cycloalkene Reactions in Air. Environmental Science & Technology 1987, 21, 52.

10 (19) Zaikov, G. E.; Rakovsky, S. K. Ozonation of Organic and Polymer Compounds; Smithers Rapra Technology, 2009.

(20) Criegee, R. The Course of Ozonization of Unsaturated Compounds. Record Chem.Progr.(Kresge-Hooker Sci.Lib.) 1957, 18, 111.

(21) Criegee, R. Mechanismus der Ozonolyse. Angewandte Chemie 1975, 87, 765.

(22) Bailey, P. S. Ozonation in Organic Chemistry; Elsevier Science, 2012; Vol. 1.

(23) Scott, G. Atmospheric Oxidation and Antioxidants; Elsevier Science, 2012.

(24) Chuong, B.; Zhang, J.; Donahue, N. M. Cycloalkene ozonolysis: collisionally mediated mechanistic branching. J Am Chem Soc 2004, 126, 12363.

(25) Olzmann, M.; Kraka, E.; Cremer, D.; Gutbrod, R.; Andersson, S. Energetics, kinetics, and product distributions of the reactions of ozone with ethene and 2,3-dimethyl-2- butene. J Phys Chem A 1997, 101, 9421.

(26) Johnson, D.; Marston, G. The gas-phase ozonolysis of unsaturated volatile organic compounds in the troposphere. Chem Soc Rev 2008, 37, 699.

(27) Horie, O.; Moortgat, G. K. Decomposition Pathways of the Excited Criegee Intermediates in the Ozonolysis of Simple Alkenes. Atmos Environ a-Gen 1991, 25, 1881.

(28) Anglada, J. M.; Crehuet, R.; Bofill, J. M. The ozonolysis of ethylene: A theoretical study of the gas-phase reaction mechanism. Chemistry-A European Journal 1999, 5, 1809.

(29) Clay, M.; Ault, B. S. Infrared matrix isolation and theoretical study of the initial intermediates in the reaction of ozone with cis-2-butene. J. Phys. Chem. A 2010, 114, 2799.

(30) Hoops, M. D.; Ault, B. S. Matrix isolation study of the early intermediates in the ozonolysis of cyclopentene and cyclopentadiene: observation of two Criegee intermediates. J Am Chem Soc 2009, 131, 2853.

(31) Chan, W. T.; Hamilton, I. P. Mechanisms for the ozonolysis of ethene and propene: Reliability of quantum chemical predictions. Journal of Chemical Physics 2003, 118, 1688.

(32) Hendrickx, M. F. A.; Vinckier, C. 1,3-Cycloaddition of Ozone to Ethylene, Benzene, and Phenol: A Comparative ab Initio Study. The Journal of Physical Chemistry A 2003, 107, 7574.

(33) Ljubic, I.; Sabljic, A. Theoretical study of the mechanism and kinetics of gas- phase ozone additions to ethene, fluoroethene, and chloroethene: A multireference approach. J Phys Chem A 2002, 106, 4745.

11 (34) Rathman, W. C. D.; Claxton, T. A.; Rickard, A. R.; Marston, G. A theoretical investigation of OH formation in the gas-phase ozonolysis of E-but-2-ene and Z-but-2-ene. Physical Chemistry Chemical Physics 1999, 1, 3981.

(35) Griesbaum, K.; Jung, I. C.; Mertens, H. Difunctional and Heterocyclic Products from the Ozonolysis of Conjugated C5-C8 Cyclodienes. J Org Chem 1990, 55, 6024.

(36) Cradock, S.; Hinchcliffe, A. J. Matrix isolation : A technique for the study of reactive inorganic species; Cambridge University Press: New York, 1975.

(37) Andrews, L.; Moskovits, M. Chemistry and physics of matrix isolated species; North-Holland, 1989.

(38) Coleman, B. E.; Ault, B. S. Matrix isolation investigation of the ozonolysis of propene. Journal of Molecular Structure 2010, 976, 249.

(39) Fenske, J. D.; Kuwata, K. T.; Houk, K. N.; Paulson, S. E. OH radical yields from the ozone reaction with cycloalkenes. J Phys Chem A 2000, 104, 7246.

(40) Wang, Z. Y.; Zvlichovsky, G. Selectivity in Ozonolyses of Cyclic 1,3-Dienes. Tetrahedron Letters 1990, 31, 5579.

(41) Hoops, M. D.; Ault, B. S. Matrix isolation study of the photochemical reaction of cyclohexane, cyclohexene, and cyclopropane with ozone. Journal of Molecular Structure 2009, 929, 22.

(42) Leong, M. K.; Mastryukov, V. S.; Boggs, J. E. Structure and conformation of cyclopentene, cycloheptene and trans-cyclooctene. Journal of Molecular Structure 1998, 445, 149.

(43) Pflitsch, C.; Viefhaus, D.; Bergmann, U.; Kravets, V.; Nienhaus, H.; Atakan, B. Growth of thin iron oxide films on Si(100) by MOCVD. Journal of the Electrochemical Society 2006, 153, C546.

(44) Singh, M. K.; Yang, Y.; Takoudis, C. G. Low-Pressure Metallorganic Chemical Vapor Deposition of Fe2O3 Thin Films on Si(100) Using n-Butylferrocene and Oxygen. Journal of the Electrochemical Society 2008, 155, D618.

(45) George, S. M.; Park, B. K.; Kim, C. G.; Chung, T. M. Heteroleptic Group 2 Metal Precursors for Metal Oxide Thin Films. European Journal of Inorganic Chemistry 2014, 2014, 2002.

(46) Conley, J. F.; Ono, Y.; Tweet, D. J.; Solanki, R. Pulsed deposition of metal–oxide thin films using dual metal precursors. Applied Physics Letters 2004, 84, 398.

(47) Roura, P.; Farjas, J.; Eloussifi, H.; Carreras, L.; Ricart, S.; Puig, T.; Obradors, X. Thermal analysis of metal organic precursors for functional oxide preparation: Thin films versus powders. Thermochimica Acta 2015, 601, 1.

12 (48) Collman, J. P. Principles and applications of organotransition metal chemistry; University Science Books: Mill Valley, Calif., 1987.

(49) Moss, S. J.; Ledwith, A. Chemistry of the Semiconductor Industry; Springer, 1989.

(50) Tan, B.; Wu, Y. Dye-sensitized solar cells based on anatase TiO2 nanoparticle/nanowire composites. The journal of physical chemistry. B 2006, 110, 15932.

(51) Mattox, D. M.Chapter 1 - Introduction. In Handbook of Physical Vapor Deposition (PVD) Processing (Second Edition); Mattox, D. M., Ed.; William Andrew Publishing: Boston, 2010, pp 1.

(52) Kääriäinen, T.; Cameron, D.; Kääriäinen, M.-L.; Sherman, A.Fundamentals of Atomic Layer Deposition. In Atomic Layer Deposition; John Wiley & Sons, Inc.: 2013, pp 1.

(53) Jones, A. H. M.Overview of Chemical Vapour Deposition. In Chemical vapour deposition: precursors, processes and applications; Jones, A. H. M., Ed.; Royal Society of Chemistry: Cambridge, 2009, pp 1.

(54) Kim, S. K.; Hwang, C. S.; Park, S. H. K.; Yun, S. J. Comparison between ZnO films grown by atomic layer deposition using H2O or O-3 as oxidant. Thin Solid Films 2005, 478, 103.

(55) Ha, S.-C.; Choi, E.; Kim, S.-H.; Roh, J. S. Influence of oxidant source on the property of atomic layer deposited Al2O3 on hydrogen-terminated Si substrate. Thin Solid Films 2005, 476, 252.

(56) Park, H. B.; Cho, M. J.; Park, J.; Lee, S. W.; Hwang, C. S.; Kim, J. P.; Lee, J. H.; Lee, N. I.; Kang, H. K.; Lee, J. C.; Oh, S. J. Comparison of HfO2 films grown by atomic layer deposition using HfCl4 and H2O or O-3 as the oxidant. Journal of Applied Physics 2003, 94, 3641.

(57) Široký, K.; Jirešová, J.; Hudec, L. Iron oxide thin film gas sensor. Thin Solid Films 1994, 245, 211.

(58) Lin, Y. M.; Abel, P. R.; Heller, A.; Mullins, C. B. α-Fe2O3 nanorods as anode material for lithium ion batteries. The Journal of Physical Chemistry Letters 2011, 2, 2885.

(59) Jiang, J.; Li, Y.; Liu, J.; Huang, X.; Yuan, C.; Lou, X. W. Recent advances in metal oxide-based electrode architecture design for electrochemical energy storage. Adv Mater 2012, 24, 5166.

(60) Sivula, K.; Le Formal, F.; Grätzel, M. Solar water splitting: progress using hematite (α‐Fe2O3) photoelectrodes. ChemSusChem 2011, 4, 432.

13 (61) Beermann, N.; Vayssieres, L.; Lindquist, S. E.; Hagfeldt, A. Photoelectrochemical studies of oriented nanorod thin films of hematite. Journal of the Electrochemical Society 2000, 147, 2456.

(62) Murphy, A. B.; Barnes, P. R. F.; Randeniya, L. K.; Plumb, I. C.; Grey, I. E.; Horne, M. D.; Glasscock, J. A. Efficiency of solar water splitting using semiconductor electrodes. International Journal of Hydrogen Energy 2006, 31, 1999.

(63) Yang, Y.; Tao, Q.; Srinivasan, G.; Takoudis, C. G. Cyclic Chemical Vapor Deposition of Nickel Ferrite Thin Films Using Organometallic Precursor Combination. Ecs Journal of Solid State Science and Technology 2014, 3, P345.

(64) Hartley, F. R.; Patai, S. The Chemistry of the Metal-Carbon Bond; John Wiley & Sons, Ltd.: Toronto, 1985; Vol. 2.

(65) Bulgakov, R. G.; Sharapova, L. I.; Sharipov, G. L.; Bikbaeva, G. G. Spectral studies of the mechanism of oxidation of Cp2Fe by ozone. Russian Chemical Bulletin 1999, 48, 790.

(66) Muthukrishnan, S.; Sankaranarayanan, J.; Klima, R. F.; Pace, T. C.; Bohne, C.; Gudmundsdottir, A. D. Intramolecular H-atom abstraction in gamma-azido-butyrophenones: formation of 1,5 ketyl iminyl radicals. Org Lett 2009, 11, 2345.

(67) Klima, R. F.; Jadhav, A. V.; Singh, P. N. D.; Chang, M.; Vanos, C.; Sankaranarayanan, J.; Vu, M.; Ibrahim, N.; Ross, E.; McCloskey, S. Photoinduced CN bond cleavage in 2-Azido-1,3-diphenyl-propan-1-one derivatives: Photorelease of hydrazoic acid. The Journal of organic chemistry 2007, 72, 6372.

(68) Zhang, X.; Sarkar, S. K.; Weragoda, G. K.; Rajam, S.; Ault, B. S.; Gudmundsdottir, A. D. Comparison of the Photochemistry of 3-Methyl-2-phenyl-2H-azirine and 2-Methyl-3-phenyl-2H-azirine. The Journal of organic chemistry 2014, 79, 653.

(69) Kugel, R. W.; Pinelo, L. F.; Ault, B. S. Infrared Matrix-Isolation and Theoretical Studies of the Reactions of Ferrocene with Ozone. J Phys Chem A 2014.

(70) Schriver-Mazzuoli, L. Ozone photochemistry in the condensed phase. Phys Chem Earth Pt C 2001, 26, 495.

(71) Bahou, M.; SchriverMazzuoli, L.; Schriver, A.; Chaquin, P. Structure and selective visible photodissociation of the O-3:Br2 and O-3:BrCl complexes: An infrared matrix isolation and ab initio study. Chemical Physics 1997, 216, 105.

14 Chapter 2

Experimental Details

Matrix Isolation System:

All experiments were conducted using a standard matrix isolation system that has been previously described in the literature.1 The experimental setup consists of a closed cycle helium refrigeration system with cold head apparatus, a sample cell, sample preparation manifolds mechanical and diffusion pumps, a photolysis lamp, and an infrared or UV-Vis spectrometer.

The cryogenic temperatures were achieved using a closed cycle helium Cryodyne® refrigeration system with a Model 22 Cold Head and Model 8200 Compressor. The cold head and compressor were connected by two braided stainless steel lines (the helium supply and return lines). The matrices were deposited on a CsI cold window. The cold window was sandwiched between two indium gaskets and mounted in a copper frame. The copper cold window holder was bolted to the second stage cold station2 of the cold head with an indium gasket in between them. The temperature of the cold window was measured using a silicon diode sensor that was bolted to the cold window holder with an indium gasket in the middle. The temperature was monitored with a Lake Shore Cryotronics, Inc. Model DRC 80C Temperature Controller.

Operating temperatures during the course of an experiment were between 12 and 15 K. The indium gaskets were used to maximize thermal contact that could be achieved between the parts.

An annealing temperature of ~36 K was achieved with a 20 W button heater also mounted on the cold window holder. Voltage to the 20 W button heater was regulated using a Variac®. An aluminum radiation shield was bolted to the first stage cold station2 and extends to the end of cold window holder. The radiation shield has two holes cut on opposite sides so spectra of the

15 matrices can be collected. The modified cold head is seated into a custom built cold cell. The top flange of the cold head forms a seal with the cell using a rubber O-ring and vacuum grease. The cold cell has two KBr windows lined up with the holes in the radiation shield and cold window; which allows the beam of the spectrometer to pass through the matrix. Black wax (Apiezon W) is used to form a seal between the KBr windows and the cell (Figure 2.1). For a minimum of 24 hours prior to and during the course of an experiment, the cell was pumped on by a Varian M-2 diffusion pump backed by a Model 1402 Duo-Seal Welch vacuum pump. A vacuum of about 10-

7 torr was typically achieved and measured by Varian cold cathode gauge mounted on the trap of diffusion pump.

There are two sample preparation manifolds (Figure 2.2). The two manifolds each have a

Nupro valve that connects by a copper tee to a Model 1402 Duo-Seal Welch vacuum pump. A vacuum of ~1x10-4 torr could be achieved and measured using a Varian thermocouple gauge.

Each manifold has a deposition line connected by a Nupro needle valve and an Ashcroft

Duragauge. As well as cold finger, a 2 L stainless steel sample can, and a line to a shared argon cylinder each connected to the manifold with by Nupro valve. All components of the sample preparation manifolds, except for the deposition lines, were connected with ¼ in. o.d. (outer diameter) stainless steel tubing and ¼ in. Swagelok® tube fittings. The ¼ in. o.d. Teflon® FEP sample deposition lines were connected to the preparation manifolds with ¼ in. Ultra-Torr® fittings. The needle valve at end of each sample preparation manifold just before the deposition lines was used to regulate the deposition rate. A deposition rate of 2 mmol/hr from each sample manifold was used.

16

17

18 The Ashcroft Duragauge measured, in inches of mercury, the pressure in the sample manifold and was used to set the deposition rate. The 2.0 L sample cans held the argon or argon sample mixture during deposition. In addition, the sample preparation manifold used for O3 was

16 18 connected to via a Nupro valve to a O2 or O2 cylinder.

The cell had two ¼ in. Ultra-Torr® deposition jet ports by which the different deposition jet configurations were inserted into the cell. There were two deposition jet configurations used for sample deposition during the course of this research. The first deposition configuration is twin jet, in which the two ¼ in. o.d. deposition lines enter cell separately and almost perpendicular to same side of the cold window Figure 2.3. Twin jet allows for only a very brief mixing/reaction time prior to deposition on to the cold window. The second deposition configuration is known as merged jet in which the ¼ in. o.d. sample lines were joined outside the cell with an Ultra-Torr® tee to from a single ¼ in. o.d. deposition line or “merged region” that enters the cell through one of the deposition ports. The length of this merged region was varied from 10-50 cm and was heated as high as 70 °C. Figure 2.4 shows a diagram of the merged jet configuration. During merged jet deposition the unused deposition port was sealed with a ¼ in. o.d. stainless steel rod. All deposition lines were ¼ in o.d. Teflon® FEP tubing and aluminum foil was used to cover the merged region during light sensitive depositions.

19

20

21 Infrared spectra were collected with a Perkin-Elmer Spectrum One FT-IR spectrometer from 400 to 4000 cm-1 at a resolution of 1 cm-1 after ~24 hr of deposition. The matrices were then irradiated by one of two methods prior to collecting another spectrum:

(1) by the H2O/quartz filtered output from a medium-pressure short arc 200 W mercury arc lamp for one hour.

(2) by the output from an incandescent light bulb filtered through a red glass cutoff filter

(Corning red glass filter #2418, transmits λ ≥ 600 nm) and an infrared cutoff filter (Schott infrared glass filter RG1000, transmits λ ≥ 1000 nm) for 30 minutes each.

UV-Vis spectra were collected with a Varian Cary 4000 UV-Vis spectrophotometer from

300 to 900 nm with a spectral band width of 2.00 nm, an average scan time of 0.10 sec, and a data interval of 1.00 nm. These matrices were irradiated using the output from an incandescent light bulb filtered through a red glass cutoff filter (λ ≥ 600 nm) for 15 minute increments. Spectra were collected after a total of 15, 30, and 45 minutes of irradiation.

Reagents:

Samples of 1-hexanal (Acros, 96%), cycloheptene (Acros, 94%), 1,3-Cyclohexadiene (Aldrich,

97%), 1,4-cyclohexadiene (Acros, 97%), and benzene (Aldrich, 99%) were prepared from the vapor above the liquid at room temperature after purification by two freeze–pump–thaw cycles at

77 K.

Stock ferrocene (Fc) (Eastman Kodak) and n-butylferrocene (nBuFc) (Strem Chemicals, 99%) were purified by sublimation. Either, Fc or nBuFc was placed in a sample holder that was heated with a sand bath at a temperature of ~45°C causing vaporization. The sample holder was

22 attached by an Ultra-Torr® tee on the deposition line where the vaporized Fc or nBuFc mixed with pure argon.

Ozone was produced by the Tesla coil discharge of O2 (Wright Brothers). The discharge tube was cooled to 77 K with liquid nitrogen to trap the ozone. Excess O2 was pumped off before the

18 ozone was warmed to room temperature. Isotopically labeled O3 was produced in the same manner from 18O labeled O2 (94%, Cambridge Isotope Laboratories).

Molecular oxygen (Wright Brothers) was used without further purification.

Argon (Wright Brothers) was used as the matrix gas in all experiments, without further purification.

References

(1) Ault, B. S. Infrared spectra of argon matrix-isolated alkali halide salt/water complexes. Journal of the American Chemical Society 1978, 100, 2426.

(2) In Cryodyne Refrigerators: Multiple Uses of Model 22C/350C; CTI-Cryogenics Helix Technology Corporation: Massachusetts, 1995, Rev. C (5/98), pp 1.

23 Chapter 3

Computational Details

The calculations were performed with the Gaussian 09 suite of programs1 using density functional theory (DFT) with the Becke, three-parameter, Lee-Yang-Parr (B3LYP) functional and the 6-311++G(d, 2p) basis set. Optimized geometries, infrared vibrational frequencies, and energies for all ground state species were obtained at this level of theory, which has been shown to be a cost effective way to successfully predict the structures and properties of the species of interest. Charge distribution results are from natural bond orbital (NBO) analyses of the optimized species. In addition, transition states and Intrinsic Reaction Coordinate (IRC) calculations were also done at this level. The IRC calculations were used to confirm the identification of the calculated transition states. The energies of the excited states and the absorption spectra of previously optimized species were calculated using time-dependent density functional theory (TD-DFT). All calculations were performed at the Ohio Supercomputer Center.

References

(1) Gaussian 09, Revision C.01, Frisch, M. J. T., G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, Jr., J. A.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, Ö.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian, Inc., Wallingford CT, 2009.

24 Chapter 4

Infrared Matrix Isolation and Theoretical Study of the Initial Intermediates in the

Reaction of Ozone with Cycloheptene

Introduction:

Cycloheptene adds an additional level of complexity not seen in the simpler alkenes that have been studied to date. Cycloheptene (C7H12) may exist in any one of four conformers, chair- cis, boat-cis, chair-trans and boat-trans.1 These are not of equal energies; the chair-cis and boat- cis forms dominate at room temperature. This leads to four possible conformations of the primary ozonide (POZ). Conformational isomerism for the POZ and the Criegee intermediate

(CI) have not been studied to date, and observation of different conformers in these experiments may provide additional insights into the details of the mechanism of the reaction.

Results and Discussion:

Prior to any codeposition experiments, blank experiments were run on each of the reagents used in this study as well as the possible product 1-hexanal. These blanks served as authentic spectra, allowing for clear identification of product bands as well as an internal standard for the extent of reaction. Blank spectra were compared to gas phase and neat infrared spectra in the literature.2-7 Other than slight matrix shifts, the blank spectra were in good agreement with literature spectra and with blanks run previously in this laboratory. Cyclohexene exists in both boat and chair conformers, with the chair calculated to be a few kcal/mol lower in energy than the boat form.1 Thus, both conformers are expect to be present in equilibrium at room temperature; bands for the slightly less stable boat form8 were noted at 629 and 803 cm-1.

Each blank deposit was also irradiated by the H2O/Pyrex-filtered output of a 200 W Hg arc lamp,

25 transmitting light with λ > 220 nm for 1.0 hours. The only change that was noted in the infrared spectra of these blank experiments was a decrease in the band of O3 upon irradiation due to photodissociation.9

Since the final products of the ozonolysis of cis-cyclohexene (cis-C7H12) are known, stable species, the focus of this study is on the initial intermediates in the reaction sequence. In particular, the primary ozonide (POZ) is believed to form initially, then the Criegee intermediate

(CI), followed by the secondary ozonide (SOZ) as shown in Figure 4.1. Because of conformational possibilities, species calculated include the chair and boat forms of cis- cycloheptene, the chair and boat forms of cis-POZ, two syn-CI conformations, and the chair and boat forms of cis-SOZ (while more conformations of the cis-CI are possible, the two that were calculated were those most likely to form from the POZ in a matrix cage environment).

Calculations were carried out here for these species using the B3LYP hybrid functional and basis sets as high as 6-311++G(d,2p). All of these species optimized to energy minima on their respective potential energy surfaces, with all positive vibrational frequencies. Oxygen-18 isotopic shifts and intensities were also calculated. Calculated vibrational frequencies are tabulated in Tables 4.1 and S4.1, while the relative energetics of the four conformers of parent cycloheptene and the possible initial intermediates in the reaction of cycloheptene with ozone are given in Table 4.2. Figure 4.2 shows the optimized structures of several key intermediate species.

The trans isomer of parent cycloheptene is known10 to be unstable with respect to conversion into the cis isomer at room temperature and should not be present in a matrix deposited from a room temperature sample. This was supported by the calculations performed here which indicated that the trans forms are approximately 29 kcal/mol higher in energy than the cis forms.

26

27

Table 4.1 Band positions and Assignments for the Initial Intermediates in the Thermal Reaction of Ozone with Cycloheptene. 18O exptl. calcd. exptl. calcd. bandsa bandsa,b shift shift assignment 701 692 −36 boat cis-primary ozonide 710 736 −38 chair cis-primary ozonide 857 875 −42 Criegee intermediate configuration 2?? 900 914 −41 −47 Criegee intermediate configuration 1 906 −1 −2 chair cis-primary ozonide 934 943 −15 −19 boat cis-primary ozonide 940 949 −12 −15 chair cis-primary ozonide 984 994 −6 boat cis-primary ozonide 1000 1010 −9 chair cis-primary ozonide 1053 1062 −1 −1 Criegee intermediate configuration 1 1167 −1 1180 1181 −1 −1 Criegee intermediate configuration 2?? 1185 1188 −1 −1 Criegee intermediate configuration 1 1467 1516 0 0 chair cis-primary ozonide 1734 1797 −36 −36, −37 Criegee intermediate configuration 1?, 2? a Frequencies in cm-1. b Calculated at B3LYP/6-311G++(d,2p) level of theory and unscaled.

28

29

Table 4.2 Computed Energies of Species Relevant to the Reaction of Cycloheptene with Ozonea. Isomers of cyclohepteneb (kcal/mol) boat trans-C7H12 29.3 chair trans-C7H12 29.3 boat cis-C7H12 4.4 chair cis-C7H12 0.0

Potential productsc (kcal/mol) chair cis-POZ -55.4 boat cis-POZ -51.6 chair trans-POZ -52.7 boat trans-POZ 14.6 config. 2 syn-CI -73.6 config 1 syn-CI -69.5 config 1 anti-CI -70.4 chair cis-SOZ -99.4 boat cis-SOZ -92.0 chair trans-SOZ -75.5 boat tran-SOZ -74.9

1-hexanal + CO2 -197.5

POZ – primary ozonide CI – Criegee intermediate SOZ – secondary ozonide a Energies computed at the B3LYP/6-311++G(d,2p) level of theory and unscaled. c Energies relative to chair cis-C7H12 + O3. b Energies relative to chair cis-C7H12.

30 Thus, intermediates derived from the trans parent compound are not expected to form. The chair and boat forms were calculated to be only 2–4 kcal/mol different in energy, and are likely both present in parent cis-cycloheptene deposited from a room temperature sample.

Merged jet and twin jet deposition probe somewhat different time and temperature regimes with respect to the mixing and reacting of cycloheptene and O3 prior to deposition onto the cold window. Twin jet deposition allows for only a very brief mixing time on the surface of the condensing matrix surface, at temperatures below room temperature but above the 14 K temperature of the rigid matrix. Merged jet deposition allows for room temperature mixing and flow through the reaction zone before condensation into the matrix. The time available for reaction, on the order of milliseconds, is much longer than in twin jet. For the reaction of cycloheptene with O3, quite different products were seen upon merged jet as compared to twin jet deposition.

In an initial experiment, a sample of Ar/cis-C7H12 = 300 was codeposited with a sample of Ar/O3 = 300 using twin jet deposition. A large number of product bands were observed upon initial deposition of this sample as listed in Table 4.1, demonstrating substantial reactivity in this system. The intensities of these peaks ranged from medium to very weak. This matrix was then annealed to 36 K, re-cooled and an additional spectrum recorded. All of the initial product bands increased in intensity, although not all of the bands increased by the same percentage. Increases ranged from 100% to 300%. Subsequent irradiation of this sample led to a decrease in intensity of some of the initial product bands and an increase in intensity of other product bands. In addition, a few new bands were seen that were not present on initial deposition. Figure 4.3 and

4.4 show representative spectra (Table 4.3).

31

32

33

34 18 A similar set of twin jet experiments were conducted with samples of Ar/c-C7H12 and Ar/ O3

18 18 made from O2 containing approximately 94% O. Upon initial sample deposition, a large

16 number of product bands were observed, similar to the bands with O3, although shifted somewhat in a number of cases. Most 18O product bands could be identified as the counterpart of an 16O band observed above; 18O band positions are listed in Table 4.1 for each product band.

The 18O product bands showed the same behavior with respect to annealing and irradiation as did their 16O counterparts.

This twin jet experiment was repeated a number of times while the sample concentration and annealing temperature were systematically varied. Comparable results were obtained throughout, including the same initial products, the same annealing behavior, and the same responses to irradiation. Sets of the product bands showed distinctly different behavior as a function of initial deposition, annealing, irradiation and isotopic labeling. As a result, the observed bands (including those formed only after annealing and/or irradiation) were sorted into several groups.

The significant growth of many of these product bands upon annealing to 35 K indicated that the barrier to reaction must be very low (3/2RT ∼ 0.1 kcal/mol at 35 K). The activation barrier for the reaction of cycloheptene and ozone has not been experimentally measured but it is anticipated to be similar to the Ea for the reaction of ozone with cycloalkenes such as cyclopentene and cyclohexene, ∼2 kcal/mol.11 Also, the large number of product bands strongly suggests that more than a single product is formed under these conditions. Additionally, the product bands observed in the thermal twin jet experiments were in many cases different from those observed in the merged jet experiments, where most of the products are known or anticipated stable oxidation products.

35 The arguments presented in the preceding paragraph strongly support assignment of the product bands formed in thermal twin jet reaction, including upon annealing, to early intermediates in the reaction of ozone with cycloheptene. Based on the results of previous studies on similar alkenes and the relatively well-established Criegee mechanism, likely initial (or

“early”) intermediates that could be observed in the twin jet experiments include the primary ozonide, Criegee intermediate and the secondary ozonide. These possibilities are compounded by the presence of boat and chair forms of the POZ and SOZ. Further complexity arises from the observation that the ring-opened Criegee intermediate can also adopt multiple conformations. To sort out these numerous possibilities, vibrational frequencies derived from theoretical calculations were essential. These calculations provide the relative energies of all of these possible intermediates, along with the computed vibrational spectra, band intensities and 18O isotopic shifts. This information, along with the intensity ratio changes during annealing, allow for a nearly unique set of assignments for all of the observed product bands.

Two sets of product bands consisting three bands each, (701, 934 and 984 cm-1) and (710,

940 and 1000 cm-1) were present upon initial deposition, increased at the same rate upon annealing (by about 250%) and were all reduced by about 20% upon irradiation. The 934 and

940 cm-1 bands shifted -15 and -12 cm-1, respectively, to lower energy upon 18O substitution, while the 18O counterparts of the 701, 710 and 984, 1000 cm-1 bands could not be identified due to overlap with the strong bands of parent cycloheptene and O3. Even so, it is clear that they have a strong dependence on 18O, as product bands were not observed at 701, 710 or 984, 1000 cm-1 and hence had shifted. All three bands for each set are in regions where primary ozonides are known to absorb strongly.12,13 Calculations for the chair and boat forms of the cis-POZ show that these two isomers should have very similar spectra, shifted by a few wavenumbers from one

36 another. In addition, the three bands calculated to be most intense for the chair-cis-POZ are located at 736, 949 and 1010 cm-1, while for the boat-cis-POZ they come at 692, 943 and 994 cm-1. These calculated positions are quite close to the two observed sets of bands, with the chair- cis-POZ slightly higher in frequency than the boat-cis-POZ. In addition, the 943 and 949 cm-1 bands were calculated to have 18O shifts of -19 and -15 cm-1, in quite reasonable agreement with the observed shifts of -15 and -12 cm-1. Finally, the computed shifts for the 692, 736, 994 and

1010 cm-1 bands would place the experimental 18O band for these modes directly underneath

18 parent cycloheptene and O3 and hence not be observable. Taken all together, the agreement of calculated and experimental bands, 18O shifts, and the growth of these bands upon annealing all support assignment of these two sets of bands to the chair-cis-POZ and boat-cis-POZ. This marks the first report of the observation and characterization of two isomers of the primary ozonide for any alkene.

The reaction of cycloheptene and O3 to form the POZ is calculated to be 50–55 kcal/mol exothermic, a value that is in line with experimental measurements11 on a number of related systems. With this large excess energy, it is possible for the reaction to proceed over the barrier to the ring-opened Criegee intermediate (CI), with a carbonyl oxide (COO) on one end an aldehyde on the other end. The barrier has been estimated to be around 20 kcal/mol for similar systems,12-19 suggesting that unless the argon matrix rapidly deactivates the energetically excited

POZ, reaction to the CI may occur. The infrared spectra of carbonyl oxides are dominated by the

COO antisymmetric stretch near 900 cm-1 while aldehydes are dominated by the C=O stretch in the 1700 cm-1 region. Both should show strong (35–45 cm-1) 18O red shifts. Such characteristic bands were observed here, at 900 and 1729 cm-1 with 18O shifts of -41 and -31 cm-1, respectively.

Calculations for the most likely configuration of the Criegee intermediate (given the rigidity of

37 the surrounding argon matrix) predict the COO and C=O stretching modes to come at 914 and

1797 cm-1 with 18O shifts of -47 and -36 cm-1, respectively. This solid agreement strongly supports identification of the Criegee intermediate formed in the reaction of cis-C7H12 with O3.

Additional product bands at 1053 and 1185 cm-1 showed similar annealing and irradiation behavior to the 914, 1734 cm-1 pair, and based on calculations may also be assigned to this conformer of the Criegee intermediate.

A few weak bands remain to be assigned, most of which had very small 18O shifts, indicating vibrations of the organic framework of the absorbing species. These are in spectral regions that could be assigned to a range of molecules or conformers. These characteristics lead to the conclusion that no definitive assignment is possible for these bands. While the evidence above supports the isolation and identification of the boat- and chair-cis-POZ and at least one conformer of the Criegee intermediate, there is no clear evidence for formation of the secondary ozonide, SOZ. Calculations for the chair and boat forms of the cis-SOZ predict an intense band near 1100 cm-1 with a 10–15 cm-1 18O shift. Such bands and shifts have been observed for several secondary ozonides. However, no bands with these characteristics were observed in the present study, suggesting that the SOZ does not form under these conditions. This may be attributed to the difficulty of closing the ring in the matrix environment once the CI has formed from the POZ by ring-opening.

Irradiation of matrices formed by twin jet deposition followed by annealing led to additional reaction. Bands of the POZ decreased upon irradiation while bands due to the CI grew slightly. In addition, a number of new bands were observed, along with a significant growth in the bands due to CO2 (always present in O3 experiments, including blanks). A likely source for growth of CO2 is from the photo-elimination of CO2 from the CI, yielding 1-hexanal. New peaks

38 produced by irradiation at 1390 cm-1, 1410 cm-1, and 1734 cm-1 matched exactly bands of 1- hexanal, by comparison to an authentic matrix spectrum of this compound. Thus, photodecomposition of the CI to CO2 and 1-hexanal appears to occur under these conditions. In addition, the irradiation of samples containing O3 generates O atoms which may react with neighboring species in the matrix. Additional new bands matched the literature spectrum of 6- heptanal,20,21 indicating O atom reaction with remaining parent cycloheptene, leading to ring opening and formation of 6-heptanal. Additional weak bands formed on irradiation are not as readily identified; several may tentatively be assigned to formation of the SOZ although other assignments are possible.

A series of merged jet experiments was also conducted with samples of Ar/ozone and

Ar/cycloheptene, using a 50 cm merged (reaction) region, held at room temperature. This configuration allows for increased gas phase reaction time for the reactions prior to matrix deposition. In each of the merged jet experiments, the parent bands of both reagents were reduced in intensity compared to twin jet experiments run at the same sample concentrations, indicating that some reaction was occurring. In addition, new product bands were observed throughout the spectrum, the most intense of which were in the carbonyl stretching region.

Figure 4.5 shows a portion of the spectrum from one merged jet experiment while key bands are listed in Table 4.4. These bands were reproduced in several experiments, all with the merged region held at room temperature. Based on comparison to literature spectra most of these bands have been assigned to stable or “late” products of the reaction, namely formaldehyde,

7,22-27 acetaldehyde, CO, 1-hexanal and CO2. These products are not surprising, as most have been noted in the literature as products of the ozone/cycloheptene reaction.28-30 In addition, a few very weak bands of the species identified above in the twin jet experiments were noted.

39

40

Table 4.4 Band Positions and Assignments for the Initial Intermediates in the Merged Jet Reaction of Ozone with Cycloheptene. exptl. bandsa lit. bandsa assignment

656s 656 CO2 dimer + CO2−H2O

663s 663 CO2 857vw Criegee intermediate configuration 2? 900m Criegee intermediate configuration 1 933vw chair cis-secondary ozonide 940vw chair cis-primary ozonide 982w boat cis-primary ozonide 1001m chair cis-primary ozonide 1094vw boat cis-secondary ozonide 1130m 1169w 1168b formaldehyde 1185wm 1246w 1245b formaldehyde 1349w 1349c acetaldehyde 1390m 1390d 1-hexanal 1410m 1410c 1-hexanal 1467vw 1497s 1497b formaldehyde 1708w 1704b formaldehyde (13C) 1729d, 1729c, 1-hexanal, acetaldehyde, formaldehyde, 1735vs 1736b Criegee intermediate config. 1 and 2

2340vs 2340 CO2 dimer + CO2−H2O a Frequencies in cm-1. b Ref. (26). c Ref. (24). d Ref. (7).

41 This is not surprising in view of the fact that some parent c-C7H12 and O3 survived passage through the merged region and were deposited into the argon matrix, where reaction may occur as in the twin jet experiments, leading to traces of the early intermediates.

Conclusions:

The variation of conditions for the deposition of ozone and cycloheptene into argon matrices has led to the observation of both initial reaction intermediates and stable reaction products. Twin jet deposition followed by annealing led to product bands which have been assigned to the chair and boat forms of the cis-primary ozonide and at least one conformer of the

Criegee intermediate. These results provide strong evidence that these ozonolysis reactions follow the mechanism proposed by Criegee in the 1950s. This marks the first observation of these early intermediates for an alkene that demonstrates conformational isomerism, and the first observation of two isomers of the primary ozonide for any alkene. Further, these results also support prior indications that the cis conformer of cycloheptene is substantially lower in energy than the trans form. Irradiation of these matrices led to growth of 1-hexanal and CO2 from photodissociation of the CI, as well as 6-heptanal from O atom addition to parent cycloheptene.

Merged jet deposition, which allows for a longer gas phase reaction time, led to stable, known species.

This approach to the study of reactive, elusive intermediates in the ozonolysis of alkenes and similar reactions may be extended in multiple directions with the current work as a foundation. One interesting question that might be resolved is the nature of the intermediates in the ozonolysis of cyclodienes, where the ozone might attack one or both of the carbon–carbon double bonds. Another system of interest would be the ozonolysis of strained bicyclic alkenes

42 such as α-pinene (which is prevalent in the troposphere in some locales). Will ozone attack the carbon–carbon double bond, or the highly strained four-membered bridging ring? The combination of matrix isolation with a range of deposition approaches, supported by theoretical calculations, is an effective tool to address these questions. Further work is in progress.

Supporting Information

Supplemental calculated infrared vibrational frequencies and intensities of possible intermediates are available in Appendix A.

References

(1) Leong, M. K.; Mastryukov, V. S.; Boggs, J. E. Structure and conformation of cyclopentene, cycloheptene and trans-cyclooctene. Journal of Molecular Structure 1998, 445, 149.

(2) S. Kinugasa, K. T., T. Tamura, Liquid Phase Cycloheptene Spectrum; National Institute of Advanced Industrial Science and Technology (AIST).

(3) Silverstein, R. M.; Bassler, G. C.; Morrill, T. C. Spectrometric Identification of Organic Compounds; Wiley, 1991.

(4) Bahou, M.; Schriver-Mazzuoli, L.; Schriver, A. Infrared spectroscopy and photochemistry at 266 nm of the ozone dimer trapped in an argon matrix. The Journal of Chemical Physics 2001, 114, 4045.

(5) Andrews, L.; Spiker, R. C. Argon matrix Raman and infrared spectra and vibrational analysis of ozone and the oxygen-18 substituted ozone molecules. The Journal of Physical Chemistry 1972, 76, 3208.

(6) Neto, N.; di Lauro, C.; Califano, S. Vibrational spectra and molecular conformations of cyclenes—II: Infrared and Raman spectra, normal co-ordinate analysis and conformation of cycloheptene. Spectrochimica Acta Part A: Molecular Spectroscopy 1970, 26, 1489.

(7) Hexanal (Neat IR). In Sigma Library of FT-IR Spectra; I ed.; Vol. 2, pp 609C.

(8) Haines, J.; Gilson, D. F. R. Phase transitions in solid cycloheptene. Canadian Journal of Chemistry 1990, 68, 604.

(9) Leighton, P. Photochemistry of Air Pollution; Academic Press: New York, 1961.

43 (10) Squillacote, M. E.; DeFellipis, J.; Shu, Q. How Stable Is trans-Cycloheptene? Journal of the American Chemical Society 2005, 127, 15983.

(11) Atkinson, R.; Carter, W. P. L. Kinetics and mechanisms of the gas-phase reactions of ozone with organic compounds under atmospheric conditions. Chemical Reviews 1984, 84, 437.

(12) Clay, M.; Ault, B. S. Infrared matrix isolation and theoretical study of the initial intermediates in the reaction of ozone with cis-2-butene. J. Phys. Chem. A 2010, 114, 2799.

(13) Hoops, M. D.; Ault, B. S. Matrix isolation study of the early intermediates in the ozonolysis of cyclopentene and cyclopentadiene: observation of two Criegee intermediates. J Am Chem Soc 2009, 131, 2853.

(14) Bailey, P. S. Ozonation in Organic Chemistry; Elsevier Science, 2012; Vol. 1.

(15) Olzmann, M.; Kraka, E.; Cremer, D.; Gutbrod, R.; Andersson, S. Energetics, kinetics, and product distributions of the reactions of ozone with ethene and 2,3-dimethyl-2- butene. J Phys Chem A 1997, 101, 9421.

(16) Anglada, J. M.; Crehuet, R.; Bofill, J. M. The ozonolysis of ethylene: A theoretical study of the gas-phase reaction mechanism. Chemistry-A European Journal 1999, 5, 1809.

(17) Hendrickx, M. F. A.; Vinckier, C. 1,3-Cycloaddition of Ozone to Ethylene, Benzene, and Phenol: A Comparative ab Initio Study. The Journal of Physical Chemistry A 2003, 107, 7574.

(18) Ljubic, I.; Sabljic, A. Theoretical study of the mechanism and kinetics of gas- phase ozone additions to ethene, fluoroethene, and chloroethene: A multireference approach. J Phys Chem A 2002, 106, 4745.

(19) Rathman, W. C. D.; Claxton, T. A.; Rickard, A. R.; Marston, G. A theoretical investigation of OH formation in the gas-phase ozonolysis of E-but-2-ene and Z-but-2-ene. Physical Chemistry Chemical Physics 1999, 1, 3981.

(20) Hoye, T. R.; Danielson, M. E.; May, A. E.; Zhao, H. Dual Macrolactonization/Pyran–Hemiketal Formation via Acylketenes: Applications to the Synthesis of (−)-Callipeltoside A and a Lyngbyaloside B Model System. Angewandte Chemie International Edition 2008, 47, 9743.

(21) Hon, Y.-S.; Wong, Y.-C.; Chang, C.-P.; Hsieh, C.-H. Tishchenko reactions of aldehydes promoted by diisobutylaluminum hydride and its application to the macrocyclic lactone formation. Tetrahedron 2007, 63, 11325.

(22) Ogawara, Y.; Bruneau, A.; Kimura, T. Determination of ppb-Level CO, CO2, CH4, and H2O in High-Purity Gases Using Matrix Isolation FT-IR with an Integrating Sphere. Analytical Chemistry 1994, 66, 4354.

44 (23) Guasti, R.; Schettino, V.; Brigot, N. The structure of carbon dioxide dimers trapped in solid rare gas matrices. Chemical Physics 1978, 34, 391.

(24) Védova, C. O. D.; Sala, O. Raman and infrared spectra and photochemical behaviour of acetaldehyde isolated in matrices. Journal of Raman Spectroscopy 1991, 22, 505.

(25) Diem, M.; Lee, E. K. C. Photooxidation of formaldehyde in solid oxygen and argon/oxygen matrixes at 12 K. The Journal of Physical Chemistry 1982, 86, 4507.

(26) Khoshkhoo, H.; Nixon, E. R. Infrared and Raman spectra of formaldehyde in argon and nitrogen matrices. Spectrochimica Acta Part A: Molecular Spectroscopy 1973, 29, 603.

(27) Dubost, H. Infrared absorption spectra of carbon monoxide in rare gas matrices. Chemical Physics 1976, 12, 139.

(28) Criegee, R. The Course of Ozonization of Unsaturated Compounds. Record Chem.Progr.(Kresge-Hooker Sci.Lib.) 1957, 18, 111.

(29) Horie, O.; Moortgat, G. K. Decomposition Pathways of the Excited Criegee Intermediates in the Ozonolysis of Simple Alkenes. Atmos Environ a-Gen 1991, 25, 1881.

(30) Fenske, J. D.; Kuwata, K. T.; Houk, K. N.; Paulson, S. E. OH radical yields from the ozone reaction with cycloalkenes. J Phys Chem A 2000, 104, 7246.

45 Chapter 5

Matrix Isolation Study of the Ozonolysis of 1,3- and 1,4-Cyclohexadiene: Identification of

Novel Reaction Pathways

Introduction:

Cyclohexadiene adds an additional level of complexity not seen in the alkenes that have been studied to date. Cyclohexadiene (CHD) has two isomers, the conjugated 1,3-CHD form and the nonconjugated 1,4-CHD form. In addition, cyclohexadiene may react with more than one ozone molecule as a consequence of the presence of two carbon–carbon double bonds.1

Consequently, a larger number of initial reaction products may be observed in the ozonolysis of

1,3- and 1,4-CHD. This mechanism begins with the formation of an initial primary ozonide through addition of the ozone across the double bond followed by C–C and O–O bond rupture to form a Criegee intermediate and a carbonyl-containing species. However, other mechanisms have been proposed.2 Although the findings of solution phase studies of conjugated cyclodienes generally support the Criegee mechanism,3,4 one study noted the formation of a small amount of phenol in the ozonolysis reaction of 1,3-cyclohexadiene (1,3-CHD).3

Results and Discussion:

1,4-Cyclohexadiene + Ozone. After blank experiments of the two parent compounds each alone in argon, including deposition, annealing, and irradiation, were conducted many codeposition experiments with these two reagents were carried out. In an initial twin jet codeposition experiment, a large number of product bands were observed upon initial deposition of the reactants, the most intense of which appeared as a doublet at 682 and 688 cm-1with an

46 absorbance greater than 2.0. Product bands were also noted in the O–H stretching region between

3300 and 3500 cm-1. All of the product bands are listed in Table 5.1.

After the matrix was annealed to 36 K, all of the initial product bands increased in intensity, and all of the bands increased by about the same amount (∼130%). Subsequent irradiation of this sample led to a slight decrease in intensity of all of the initial product bands and the formation of several new weak product bands. Figure 5.1 shows representative spectra.

To further explore this system, this twin jet experiment was repeated a number of times, systematically varying sample concentrations of 1,4-CHD and ozone (O3), as well as the annealing temperature. Comparable results were obtained throughout, including the same initial products, the same annealing behavior and the same responses to irradiation. Finally, a twin-jet codeposition experiment was conducted with 1,4-CHD and O2. No reaction was observed, demonstrating that O3 is the reacting species, not residual O2.

These observations indicate that substantial reaction is occurring between the two parent compounds during the very brief mixing time in twin jet deposition. On the basis of previous studies, one might anticipate formation of a primary ozonide and possibly either a Criegee intermediate or a secondary ozonide. The concentrations were for the most part too low to allow for the possibility of the reaction of two ozone molecules with a molecule of 1,4-CHD, which could lead to additional products. However, the numerous product bands exhibited nearly identical behavior as a function of initial deposition, annealing, and irradiation. This would indicate that only one product is being formed or that all products are forming at the same rate.

This differs from previous studies of the ozonolysis of alkenes, where multiple products were observed and showed differing behaviors with respect to annealing and irradiation.

47

Table 5.1 Band Positions and Assignments for the Products in the Thermal Reaction of Ozone with 1,4-Cyclohexadiene. 18O exptl. calcd. calcd. exptl. bandsa bands1 shiftb shift assignments

479 451 -11 benzene-H2O3

509 516 -20 -19 benzene-H2O3 682 682 0 0 benzene

688 687 0 0 benzene-H2O3 720 POZ?

744 774 -44 -40 benzene-H2O3

825 823 -53 benzene-H2O3 843 POZ? 853 -7 POZ? 877 -2 POZ? 904 POZ?

1039 1040 0 benzene-H2O3

1179 1180 0 0 benzene-H2O3 1290 -4 POZ?

1341 1330 -7 -6 benzene-H2O3 1474 0 POZ?

1481 1481 0 0 benzene-H2O3

1822 1812 0 0 benzene-H2O3 1963 0

3097 3102 0 0 parent & benzene-H2O3

3388 3431 -12 -15 benzene-H2O3

3450 3515 -12 -11 benzene-H2O3 1 band calculated by applied calculated shift due to complexation to the literature matrix bands of parent species a Frequencies in cm-1 b Calculated at the B3LYP/6-311G++(d.2p) level of theory.

48

49 Also, the present experiments differed from previous studies in that strong peaks in the O–H stretching region were present upon initial deposition and increased in intensity after annealing in the present study.5,6

To assist the identification of products in the 1,4-CHD/O3 system, an equivalent set of

18 18 twin jet experiments were conducted with samples of Ar/1,4-C6H8 and Ar/ O3 made from O2 containing approximately 94% 18O. Upon initial sample deposition, a large number of product

16 bands were again observed, quite similar to the bands with O3. It was striking that the product bands separated into two sets, the first of which shifted substantially as a result of 18O substitution whereas the second set did not shift at all. The most intense band, at 688 cm-1, did not shift at all with this isotopic substitution. For the set of bands that did shift, 18O product bands could be identified as the counterparts of most of the 16O bands that had been observed.

18O band shifts are also listed in Table 5.1 for each product band. The 18O product bands showed the same behavior with respect to annealing and irradiation as did their 16O counterparts. These result again different from previous ozonolysis studies of small alkenes where the most intense bands were associated with oxygen atom motions in the products and all showed substantial shifts.

The short time available for reaction in twin jet deposition suggested the observed products in the ozonolysis of 1,4-CHD are likely initial intermediates in the reaction sequence as observed in previous studies. In particular, the primary ozonide is believed to form initially, then the Criegee intermediate followed by the secondary ozonide. However, the product bands, particularly those around 3400 cm-1, indicate that a different reaction pathway than that followed for most alkenes is predominant. The bands between 3300 and 3500 cm-1 are clearly identified as

O–H stretches on the basis of position and 18O shift. Further, the bands at 682, 688, 1179, 1481,

50 and 3097 cm-1, which did not shift upon 18O substitution, all lie very close to the most intense infrared absorptions of benzene, C6H6. The proximity to known bands of benzene and corroborated here by in a blank experiment with benzene, combined with the anticipated lack of any 18O shifts, strongly suggests that a dehydrogenation reaction has occurred leading to benzene formation. Literature studies7,8 of the gas phase reaction of 1,4-CHD with metal atoms and metal ions also led to dehydrogenation. For example, Davis et al. reported that the reaction of Y atoms with 1,4-CHD in crossed molecular beams led to the formation of YH2 and C6H6. This suggests two likely reaction channels, 1,4-CHD + O3 → C6H6 + H2O3 and 1,4-CHD + O3 → C6H6 + H2O

-1 + O2. These channels must account for the set of bands at 509, 744, 1341, 3388, and 3450 cm bands shifted -19, -40, -6, -15, and -11 cm-1, respectively, to lower energy upon 18O substitution.

9 The infrared spectrum of H2O3 was first reported in argon matrixes by Engdahl and Nelander in

2002 whereas the infrared spectra of H2O and the C6H6–H2O complex in argon matrixes are well-known.10 This indicates that both channels are viable possibilities. However, the experimental spectra observed here are not at all consistent with the known spectra of H2O and the C6H6–H2O complex in argon matrixes with respect to number of bands, band locations, and

18 O shifts. Thus, the 1,4-CHD + O3 → C6H6 + H2O3 channel needs examination.

If C6H6 and H2O3 are formed through a dehydrogenation reaction during the matrix deposition process, then it is likely that they may interact and form a complex in the matrix cage.

Therefore, DFT (B3LYP/6-311++g(d,2p)) calculations were carried out for the separated benzene and H2O3 species as well as a benzene–H2O3 complex (BHC). Two conformers of this complex were found to be stable. One conformer has one hydrogen of H2O3 hydrogen bonded to the π electron density on the C6H6 ring in a manner similar to that for complexes of the hydrogen

51 halides with C6H6, whereas the second conformer has both hydrogens interacting with the π electron density, as shown in Figure 5.2.

The calculated energies of these two conformers are similar within computational error.

In addition, the calculated spectra for the two are quite similar, other than a reversal of intensities for the two O–H stretching modes.

Though one could compare the spectrum calculated for the complex with the experimental spectrum, there are always systematic computational errors and perturbations caused by interactions with the argon matrix. These could be reduced by determining the calculated shift of the two subunits in the complex from the calculated positions of the uncomplexed molecules as shown in Tables 5.1 and S5.1. Then, these calculated shifts are applied to the known experimental bands of C6H6 and H2O3 in argon matrixes. This methodology shows clearly that the reaction of 1,4-CHD with O3 to form the C6H6–H2O3 complex occurs and is the dominant, if not exclusive, reaction channel. Specifically, the intense antisymmetric O–O–

-1 -1 O stretching mode of H2O3 is calculated to come at 774 cm (unscaled) with a -44 cm shift with 18O, which is in very reasonable agreement with a strong product band at 744 cm-1 with -40 cm-1 18O shift. Similarly, the antisymmetric and symmetric O–H stretches are calculated to come at 3431 and 3515 cm-1 with 18O shifts of -12 cm-1 each. These compare well to the experimental

-1 -1 values of 3388 and 3450 cm , with shifts of -15 and -11 cm . For the C6H6 subunit in the complex, the most intense band is calculated at 696 cm-1, in good agreement with the very intense product band at 688 cm-1. Table 5.1 shows the complete comparison of experimental bands and shifts with those calculated using this methodology.

52

53 From the above, it is clear that the reaction of O3 with 1,4-CHD leads to dehydrogenation, with the formation of C6H6 and H2O3 which are weakly complexed in the argon cage, as shown in

Figure 5.3. To the best of our knowledge, this is an unprecedented reaction pathway for the ozonolysis of an alkene.

The transition state for the hydrogen elimination from 1,4-CHD to form BHC was calculated and verified with IRC calculations. The transition states were also calculated for the

Criegee mechanism and for other products observed in the spectra. Each transition state was optimized to (TS) Berny and the frequency calculation found one imaginary frequency. Then

IRC calculations were performed on each transition state showing that they connected to the correct reactants and products. The resulting potential energy surface is shown in Figure 5.4 and supports the conclusion that formation of the H2O3–C6H6 complex is an energetically viable pathway. O3 and 1,4-CHD form an initial weak complex with the O3 subunit positioned directly over the ring of 1,4-CHD. From there, the pathway to double hydrogen abstraction to form H2O3 and C6H6 is essentially barrierless. Once these two species are formed within the matrix cage, they interact to form a weak complex in the cage. The reaction to form this complex is over 70 kcal/mol exothermic with respect to the parent species.

It is noteworthy that the barrier to reaction to form the primary ozonide of 1,4-CHD is low as well, perhaps just slightly higher than the barrier to form H2O3 and C6H6. Therefore, DFT calculations were then carried out to determine the stability and vibrational spectrum of the potential intermediates. Each structure was optimized using the B3LYP hybrid functional and basis sets as high as 6-311++G(d,2p) to obtain accurate energies and vibrational frequencies.

Further, because of the presence of two double bonds and results from solution phase studies,1,3,4,11 additional possible intermediates were calculated.

54

55

56 All of these species optimized to energy minima on their respective potential energy surfaces, with all positive vibrational frequencies. 18O isotopic shifts and intensities were also calculated.

Several structural conformations of the intermediates were also calculated and the energies relative to the reactants are given in Table S5.2. Comparing the few weak, unassigned product bands to these calculations, several bands of the primary ozonide are calculated to come near these product bands. However, in most cases the 18O counterparts are not observed and may be obscured by strong parent bands of bands of the C6H6–H2O3 complex in the region. Thus, such assignments are tentative and noted by “POZ?” in Table 5.1. Finally, it is noteworthy that in merged jet experiments, intense bands of H2O3 and C6H6 were observed, indicating that these are gas phase products for this system and not a consequence of the matrix itself. Overall, it is clear that the pathway leading to dehydrogenation and formation of C6H6 + H2O3 dominates the reaction of O3 with 1,4-CHD.

1,3-C6H8 + O3, Twin Jet. After blank experiments of the two parent compounds, including deposition, annealing, and irradiation were conducted, many codeposition experiments with these two reagents were carried out. In an initial twin jet codeposition experiment, a large number of product bands, with intensities ranging from medium to very weak, were observed upon initial deposition as listed in Table 5.2.

In contrast to the 1,4-CHD system, after the matrix was annealed, some of the initial product bands increased in intensity while other bands decreased. Increases ranged from 100 to

200% and decreases ranged from 90 to 60%. Also unlike the 1,4-CHD system, subsequent irradiation of this sample led to a decrease in intensity of some of the initial product bands and an increase in intensity of other product bands. In addition, a few new bands were seen that were not present on initial deposition. Figure 5.5 shows representative spectra.

57

Table 5.2 Band Positions and Assignments for the Initial Intermediates in the Thermal Reaction of Ozone with 1,3- Cyclohexadiene. 18O exptl. calcd. exptl. calcd. bandsa bandsb shift shiftb assignments 441 411 -6 -4 PI-B 476 488 -11 -11 PI-A 639 -3 c 672 666 -1 CI 710 738 -7 -9 CI 723 702 -14 -13 POZ 766 755 -11 -9 POZ 770 772 -9 -7 PI-A 784 790 -12 -8 PI-B 825 c 863 -3 c 876 -1 c 895 -8 c 956 978 -14 -15 POZ 967 951 -29 CI 989 999 -13 CI 1011 1001 -5 -4 POZ 1318 -1 c 1340 -7 c 1365 1407 -5 -8 PI-B 1380 0 c 1740 1806 -40 -36 CI 3443 3664 -15 -12 PI-A 3540 3706 -13 -12 PI-B a Frequencies in cm-1. b Calculated at the B3LYP/6-311G++(d.2p) level of theory and unscaled. c Could be assigned to POZ, PI-A, or PI-B; see text.

58

59 Comparison of the bands observed for this system to the product bands in the reaction of

O3 with 1,4-CHD discussed above clearly established that dehydrogenation of 1,3-CHD does not occur, e.g., that C6H6 and H2O3 are not observed. Certainly, a different reaction pathway (or pathways) is being followed. Further, these results allow the sorting of product bands into groups based on their behavior with respect to initial deposition, annealing, and irradiation. The significant growth of a number of these product bands upon annealing to 35 K indicates that the barrier to reaction for this pathway must be very low (3/2RT ∼ 0.1 kcal/mol at 35 K). Although the activation barrier for the reaction of 1,3-cyclohexadiene and ozone has not been experimentally measured, it is anticipated to be similar to the Ea for the reaction of ozone with

12 cycloalkenes such as cyclopentene and cyclohexene, ∼2 kcal/mol. For the reaction of O3 with cyclopentene, similar growth of product bands upon annealing was observed. Additionally, the product bands observed in the thermal twin jet experiments are in many cases different from those observed in the merged jet experiments, where most of the products were known or anticipated stable oxidation products.5,6,13

The arguments presented in the preceding paragraph strongly support assignment of the product bands formed in thermal twin jet reaction of O3 with 1,3-CHD, including upon annealing, to early intermediates in the reaction of ozone. On the basis of the results of previous studies on similar alkenes and the relatively well-established Criegee mechanism, likely initial

(or “early”) intermediates that could be observed in the twin jet experiments include the primary ozonide, Criegee intermediate, and the secondary ozonide. These possibilities are compounded by the presence of two double bonds meaning that ozone can react with one or both the double bonds, as well multiple structural isomers. Further complexity arises from the observations that

60 the ring-opened Criegee intermediate can form two isomers, because there are two possible positions for the remaining double bond relative to the carbonyl oxide and carbonyl groups.

To aid in the identification of products in this system, twin jet experiments were

18 conducted with samples of Ar/1,3-C6H8 and Ar/ O3. Upon initial sample deposition, a large

16 number of product bands were observed, similar to the bands with O3, although shifted somewhat in a number of cases. Most 18O product bands could be identified as the counterpart of a 16O band observed above; 18O band shifts are listed in Table 5.2 for each product band. As anticipated, the 18O product bands showed the same behavior with respect to annealing and

16 irradiation as did their O counterparts. Given the results described above for the reaction of O3 with 1,4-CHD, a search for absorptions of H2O3, C6H6 and the H2O3–C6H6 complexes was made,

-1 particularly for the very intense band of the H2O3–C6H6 complex at 688 cm . Comparison of

Tables 5.1 and 5.2 clearly show that H2O3, C6H6 and the H2O3–C6H6 are not formed in the reaction of O3 with 1,3-CHD.

To sort out the numerous possible products for the O3 + 1,3-CHD system, theoretical calculations were essential. These calculations provide the relative energies of all of these possible intermediates as shown in Figures 5.6 and S5.1 and in Table S5.3, along with the computed vibrational spectra, band intensities, and 18O isotopic shifts as listed in Table S5.4.

First, a number of possibilities could be eliminated due to a serious mismatches between experimental band positions and calculated positions. The possibilities that were initially eliminated included the primary ozonide, Criegee intermediate and secondary ozonide from the trans form of 1,3-CHD as well as those products requiring two O3 molecules (e.g., the double

POZ).

61

62 The former result is anticipated because the trans forms are approximately 7 kcal/mol higher in energy that the cis forms. The latter result is not surprising in that it would be unlikely that two ozone molecules would react with both double bonds simultaneously because both the ozone and cyclohexadiene are very dilute in argon.

This information, along with the intensity ratio changes during annealing, allow for assignment of a number of the product bands to the primary ozonide and Criegee intermediate arising from ozonolysis of one of the carbon–carbon double bonds through a Criegee mechanism as shown in Figure 5.7.

Specifically, intense product bands at 723, 766, 895, 967, 1011, 1340, and 1380 cm-1 were present upon initial deposition, decreased at the same rate upon annealing (by about 70%), and were all reduced by about 50% upon irradiation. The bands at 723, 766, 956, 1011, and 1340 cm-1 shifted -14, -11, -14, -5, and -7 cm-1, respectively, to lower energy on 18O substitution, whereas the 1380 cm-1 band did not shift. The 723, 766, 956, and 1011 cm-1 bands are in regions where primary ozonides are known to absorb strongly.5,6,13 In addition, the 702, 755, 978, 1001, and 1340 cm-1 bands of the primary ozonide were calculated to have 18O shifts of -12, -9, -15, -4, and -2 cm-1, which is in reasonable agreement with the observed shifts of -14, -11, -14, -5, and -7 cm-1. The experimental band at 1380 cm-1 that did not shift is consistent with a C–H bending vibration of the primary ozonide and is in good agreement with the computed primary ozonide band at 1373 cm-1 with no 18O shift. Taken altogether, the agreement of calculated and experimental bands, 18O shifts, and the decrease of these bands upon annealing all support assignment of these bands to the primary ozonide.

63

64 The reaction of cyclohexadiene and O3 to form the primary ozonide is calculated to be about 51 kcal/mol exothermic, a value that is in line with experimental measurements12 on a number of related systems. With this large excess energy, it is possible for the reaction to proceed over the barrier to the ring-opened Criegee intermediate, with a carbonyl oxide (C—O—

O) on one end an aldehyde on the other end. The barrier has been estimated to be around 20 kcal/mol for similar systems,6,12-18 suggesting that unless the argon matrix rapidly deactivates the energetically excited primary ozonide, reaction to the Criegee intermediate may occur. There are many different structural conformations of the Criegee intermediate and further complexity arises in that the ring-opened Criegee intermediate can form two isomers. There are two possible positions for the remaining double bond: in form (CI-A) the carbon–carbon double bond is located at the carbonyl oxide end of the molecule and in the other form (CI-B) this bond is located at the carbonyl end (Figure 5.6b,c). Solution phase studies on the ozonolysis of cyclic

1,3-dienes showed that the CI-A is preferentially formed.1,4 The infrared spectrum is most consistent with the CI-A form. The infrared spectra of carbonyl oxides are dominated by the

C—O—O antisymmetric stretch near 900 cm-1 whereas aldehydes are dominated by the C═O stretch in the 1700 cm-1 region. Both should show strong 18O red shifts. Such characteristic bands were observed here, at 967 and 1740 cm-1 with a −40 cm-1 18O shift for the 1740 cm-1 product band (no 18O counterpart of the 967 cm-1 band was seen due to spectral congestion in the region).

Calculations for the most likely configuration of the Criegee intermediate (given the rigidity of the surrounding argon matrix) predict the C—O—O and C═O stretching modes to come at 951 and 1806 cm-1 with 18O shifts of -28 and -36 cm-1, respectively. This reasonable agreement supports identification of the Criegee intermediate formed in the reaction of 1,3-CHD with O3.

Additional product bands at 672, 710, and 989 cm-1 showed annealing and irradiation behavior

65 similar to that of the 967, 1740 cm-1 pair, and on the basis of calculations may also be assigned to this conformer of the Criegee intermediate.

The entries in Table 5.2 make it clear, however, that identifications of the POZ and CI of

1.3-CHD do not account for all of the observed product bands. In particular, two product bands were observed in the O–H stretching region, bands that cannot be attributed to the POZ or CI. On the basis of the results for the 1,4-CHD/O3 system, above, they also cannot be attributed to H2O3 or the complex of H2O3 with C6H6. As shown in Figure 5.8, detailed exploration of the potential energy surface for this pair of reagents leads to another possibility.

With only a 2 kcal/mol barrier, ozone can interact with carbon C1 and the axial hydrogen of C5 in transition state TS-4 shown in Figure 5.8 en route to the formation of a hydroperoxy

(O–O–O–H) species PI attached to the ring at C1. This barrier is similar to that for the formation of the primary ozonide, suggesting that both pathways should be accessible. This hydroperoxy intermediate is 42 kcal/mol exothermic with respect to the parent species, providing it with sufficient energy to settle into one of several conformers. PI-A and PI-B were identified as two such conformers. The calculated infrared spectra of PI-A and PI-B match up well with several of the as-yet unassigned bands in Table 5.2, in particular with the two bands in the O–H stretching region and their 18O counterparts. All of the remaining unassigned bands are in regions in which the POZ, PI-A, and PI-B all have calculated bands. Although definitive assignments cannot be made due to the lack of observation of 18O counterparts due to band overlap, they can all very likely be assigned to one or more of these intermediate species.

The findings here indicate that for the reactions of 1,3- and 1,4-cyclohexadiene with ozone, reaction mechanisms other than the standard Criegee mechanism are followed.

66

67 One question that arises is why these two alkenes have viable if not dominant pathways available to them that are not present in the wide range of alkenes studied to date. For 1,4-CHD, the answer likely lies with the gain in resonance stabilization energy with the formation of benzene.

The dehydrogenation of 1,4-CHD is known for several reagents for very much the same reason.

Second, for 1,4-CHD, the distance between the axial hydrogens on the two −CH2– groups (C3 and C6) in the ring matches very well the spacing needed to interact with the two terminal oxygen atoms of O3. In contrast, this is not the case for 1,3-CHD, where the two −CH2– groups

(C5 and C6) are adjacent to one another on the ring, and the hydrogens are too close together to

7 match up well with the terminal oxygen atoms of O3. It is noteworthy that the reactions of Y atoms with 1,3- and 1,4-CHD both lead to dehydrogenation, forming C6H6 and YH2. The smaller size of a Y atom compared to O3 may facilitate dehydrogenation of 1,3-CHD. Thus, the pathway to benzene formation is less favorable. A related question is: Why then does the reaction of 1,3-

CHD and O3 form a hydroperoxy species? The answer again appears to be favorable geometry in an initially weakly bonded complex. From there, there is only a small (4 kcal/mol) barrier to a transition state in which one terminal oxygen atom of O3 position interacts with an axial hydrogen on C5 whereas the second oxygen atom spans the ring and interacts with C1. From this transition state, it is energetically very favorable for hydrogen atom abstraction from C5 and

O atom addition to C1, forming the hydroperoxy species PI.

A final question has to do with the branching ratio between the two pathways available in each of these systems (for 1,4-CHD + O3, formation of either the primary ozonide or the C6H6–

H2O3 complex and for 1,3-CHD + O3, formation of either the primary ozonide or the hydroperoxy intermediate). There are probably two factors that contribute to determining the preferred pathway for each system. One is that the angle of incidence or approach of the two

68 parent molecules may determine which pathway is preferred. Second is that the subtle details of the potential energy surface and the transition states determine the pathway. Although the calculated barriers are similar in each case, the uncertainties in the barriers are sufficient that we cannot conclude with certainty which would be preferred pathway. To explore this point further, we looked at the relevant orbitals on ozone and the cyclohexadiene molecules. As expected, the favorable molecular orbital interactions between the ozone and the 1,3- and 1,4-alkenes allow for the concerted reactions to take place. However, this conclusion is qualitative and does not permit a determination from those interactions as to which reaction is favored. Similarly, the calculated charges on the alkenes and the ozone match but this is not sufficient to permit a determination as to which reactions are favored on the basis of charges. In the future, a study of the reactions of ozone with substituted 1,3- and 1,4-cyclohexadienes may allow insights into this question.

Conclusions:

The deposition of ozone and 1,4-CHD into argon matrixes led to the observation of a high yield of the benzene–H2O3 complex through a dehydrogenation reaction. The formation of this complex is not consistent with the mechanism proposed by Criegee. In addition, a small yield of the primary ozonide was likely observed. Within the uncertainties of the calculations, the barriers of the two pathways are comparable; small subtleties in the potential energy surface appear to favor the dehydrogenation reaction.

The ozonolysis reaction of 1,3-CHD led to the formation of both the primary ozonide and

Criegee intermediate, along with two conformers of a novel hydroperoxy intermediate. At the

69 same time, a dehydrogenation reaction to form H2O3 and C6H6 does not occur. For this system, the potential energy surface has similar, low barriers to reaction for the Criegee and hydroperoxy pathways. These findings indicate that structural configuration and position of the double bonds in the ring influence the reaction pathways for the reaction of cyclohexadiene with O3.

Supporting Information

Additional computational results including structures, energies, and vibrational spectra of possible intermediates are available in Appendix B.

References

(1) Park, S. H., T. Diozonides from Coozonolyses of Cyclodienes and Carbonyl Compounds. Bull. Korean Chem. Soc. 2002, 23, 2426.

(2) Bailey, P. S. Ozonation in Organic Chemistry; Elsevier Science, 2012; Vol. 1.

(3) Griesbaum, K.; Jung, I. C.; Mertens, H. Difunctional and Heterocyclic Products from the Ozonolysis of Conjugated C5-C8 Cyclodienes. J Org Chem 1990, 55, 6024.

(4) Wang, Z. Y.; Zvlichovsky, G. Selectivity in Ozonolyses of Cyclic 1,3-Dienes. Tetrahedron Letters 1990, 31, 5579.

(5) Pinelo, L.; Ault, B. S. Infrared matrix isolation and theoretical study of the initial intermediates in the reaction of ozone with cycloheptene. Journal of Molecular Structure 2012, 1026, 23.

(6) Hoops, M. D.; Ault, B. S. Matrix isolation study of the early intermediates in the ozonolysis of cyclopentene and cyclopentadiene: observation of two Criegee intermediates. J Am Chem Soc 2009, 131, 2853.

(7) Schroden, J. J.; Davis, H. F. Reactions of Neutral Gas-Phase Yttrium Atoms with Two Cyclohexadiene Isomers. The Journal of Physical Chemistry A 2012, 116, 3508.

(8) Tews, E. C.; Freiser, B. S. Heteronuclear diatomic transition metal cluster ions in the gas phase. Reactions of CuFe+ with hydrocarbons. Journal of the American Chemical Society 1987, 109, 4433.

70 (9) Engdahl, A.; Nelander, B. The Vibrational Spectrum of H2O3. Science 2002, 295, 482.

(10) Engdahl, A.; Nelander, B. A matrix isolation study of the benzene-water interaction. The Journal of Physical Chemistry 1985, 89, 2860.

(11) Frisch, M. J. T., G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonne Gaussian 09; Revision C.01 ed.; Gaussian, Inc.: Wallingford, CT, 2009.

(12) Atkinson, R.; Carter, W. P. L. Kinetics and mechanisms of the gas-phase reactions of ozone with organic compounds under atmospheric conditions. Chemical Reviews 1984, 84, 437.

(13) Clay, M.; Ault, B. S. Infrared matrix isolation and theoretical study of the initial intermediates in the reaction of ozone with cis-2-butene. J. Phys. Chem. A 2010, 114, 2799.

(14) Olzmann, M.; Kraka, E.; Cremer, D.; Gutbrod, R.; Andersson, S. Energetics, kinetics, and product distributions of the reactions of ozone with ethene and 2,3-dimethyl-2- butene. J Phys Chem A 1997, 101, 9421.

(15) Anglada, J. M.; Crehuet, R.; Bofill, J. M. The ozonolysis of ethylene: A theoretical study of the gas-phase reaction mechanism. Chemistry-A European Journal 1999, 5, 1809.

(16) Hendrickx, M. F. A.; Vinckier, C. 1,3-Cycloaddition of Ozone to Ethylene, Benzene, and Phenol: A Comparative ab Initio Study. The Journal of Physical Chemistry A 2003, 107, 7574.

(17) Rathman, W. C. D.; Claxton, T. A.; Rickard, A. R.; Marston, G. A theoretical investigation of OH formation in the gas-phase ozonolysis of E-but-2-ene and Z-but-2-ene. Physical Chemistry Chemical Physics 1999, 1, 3981.

(18) Ljubic, I.; Sabljic, A. Theoretical study of the mechanism and kinetics of gas- phase ozone additions to ethene, fluoroethene, and chloroethene: A multireference approach. J Phys Chem A 2002, 106, 4745.

71 Chapter 6

Low-energy Photochemistry of Ozone and n-Butylferrocene: A matrix isolation study

Introduction:

Recently a matrix isolation study of ferrocene (Fc) and ozone (O3) was conducted by

Kugel, et al.1 This study used matrix-isolation combined with IR spectroscopy and theoretical calculations to show that a photochemical reaction of O3 with Fc occurs upon irradiation with red light (λ ≥ 600 nm). This low-energy photochemical reaction leads to the production atomic oxygen O(3P) which subsequently reacts with Fc resulting in the formation of an iron cyclopentadienyl ring moiety and ether: (1) a pyran, (2) an aldehyde, or (3) a bidentate cyclic aldehyde with a seven-membered ring including the iron atom. It should be noted that a dark green matrix was observed prior to irradiation and this color was attributed to a charge transfer complex between O3 and Fc (Fc-O3). The present work investigates the photochemical reaction of O3 with n-butylferrocene (nBuFc), and it explores the effect of the n-butyl substituent on the reaction and the observed products.

Results and Discussion:

Blank spectra including deposition, annealing, and irradiation were obtained for both parent compounds in argon (Ar) prior to conducting codeposition experiments. These blanks showed excellent agreement with both literature and calculated spectra.2,3 The experimental IR spectrum of matrix-isolated nBuFc in Ar is given in Figure S6.1 of Appendix C. A list of the calculated infrared vibrational frequencies and intensities for nBuFc is provided in Table S6.1a.

The calculated ring conformation of nBuFc is eclipsed (Figure 6.1a) and a stable staggered conformation could not be located. The calculated eclipsed conformation for nBuFc is consistent

72

73 with that of the experimentally determined gas phase structures of methylferrocene and ethynylferrocene.4,5

Early results of the Ar/O3 and Ar/nBuFc twin jet (TJ) deposition experiments indicated that a photochemical reaction occurs upon exposure to light from the He/Ne calibration laser (λ=

632 nm) or the IR source. The exposure to light from the IR spectrometer resulted in a change of the color of the matrix at the point where the calibrating laser and IR source passed through. This

1 phenomenon was also observed in the Ar/O3/Fc experiments. All further experiments were performed by covering the cell windows during deposition.

Dark Deposition. Product bands were not observed in the dark deposited spectrum of

Ar/nBuFc/O3. However, slightly red shifted ozone absorptions were observed (Figure S6.2). The matrix formed during the dark deposition was also a vivid green color (Figure S6.3a). The colors of the Ar/O3 and Ar/nBuFc matrices formed during the collection of the blanks are white and pale yellow respectively. The green matrix color is theorized to be the result of an nBuFc-O3 charge transfer complex. Similarly, a green matrix was also previously formed in the dark

1 deposition of Ar/Fc/O3, and the green color was attributed to a Fc-O3 charge transfer complex.

The bond lengths ort for the formation of these charge transfer complexes is based on the green color, suggesting strong red and blue absorptions in the visible spectrum, and slightly red-shifted ozone absorptions that are characteristic of perturbed ozone.6 Currently a UV-Vis study is being undertaken and preliminary results support this conclusion. Annealing the matrix to 36 K did not lead to the formation of any product peaks nor any change in color of the matrix.

Photochemical Reaction. When the matrix was irradiated using the output from an incandescent light bulb filtered through a red glass cutoff filter (λ ≥ 600 nm) a number of product bands were observed (Figure 6.2). Diagnostically significant product peaks are located in the

C=O stretching region at 1681 cm-1 and in the C–H stretching region at 2712 cm-1. Several

74

75

76 product peaks were also observed in the C–O–C stretching region. In addition to the formation of product peaks upon irradiation, a change in the in the color of the matrix from vivid green to red-orange was observed. Photographs of the green dark deposited Ar/nBuFc/O3 matrix prior to irradiation with red light and resulting red-orange matrix post irradiation are provided in

Appendix C (Figure S6.3). There are many similarities in the product peaks and color changes

1 observed in both the Ar/nBuFc/O3 and Ar/Fc/O3 spectra after the irradiation with red light.

It has been documented in the literature that excitation of the Chappuis band (420 nm ≤ λ

6,7 ≤ 700 nm) leads to the photodissociation of O3. Furthermore, Kugel, et al. observed isotope

18 16 scrambling between O3 and O2 co-isolated in an Ar matrix after irradiation with red light (λ ≥

1 3 600 nm). The photolysis of O3 with red light (λ ≥ 600 nm) produces atomic oxygen, O( P), and

3 3 molecular oxygen, O2( Σ). The O( P) produced by red photolysis of O3 has up to as much as +23 kcal/mol of excess thermal energy.8 The photochemical products observed here are proposed to

3 be the result of a reaction between O( P) and nBuFc and not those of O3 and nBuFc.

Based on the products observed in the literature for reactions of O(3P) with alkenes, the spectral similarities to the previous study of Ar/Fc/O3, and the diagnostically significant peaks

1,9- observed in the spectrum of irradiated Ar/nBuFc/O3, several possible products were proposed.

13 These proposed products consist of an iron atom with a coordinated n-butylcyclopentadienyl

(nBuCp) or cyclopentadienyl (Cp) ring and either: (1) a pyran, (2) an aldehyde, or (3) a bidentate cyclic aldehyde with a seven-membered ring including the iron atom. However, in the case of nBuFc the butyl group on one of the Cp rings means that there are multiple regioisomers of the pyran (pyran-A–F), aldehyde (ald-A–I), and ring-aldehyde (rald-A–E); as well as a ketone (keto) and ring-ketone (rketo) to consider (Figure 6.3). In addition, two conformational isomers of the aldehyde were also calculated. The two configurations are designated by the position of the

77

78 hydrogen on the aldehyde group relative to the hydrogen or n-butyl group on the adjoining carbon atom. The aldehyde with oxygen atom pointing outwards and the hydrogens (or the n- butyl group and hydrogen) opposite of each other is designated the (E)-aldehyde (Figure 6.4a).

The aldehyde with oxygen atom pointing inward and the hydrogens (or the n-butyl group and hydrogen) on the same side is designated the (Z)-aldehyde (Figure 6.4b). Figure S6.4 shows the

Gaussian 09 optimized structures and energies relative to nBuFc and O(3P). Table S6.1 gives calculated infrared vibrational frequencies, infrared intensities, and 18O-isotopic shifts of these products the in Appendix C.

Products formed from the reaction of O(3P) with the butyl group were not considered, given the presence of the conjugated cyclopentadienyl ring. The literature suggested that alkyl substituent makes the reaction of atomic O with the unsaturated carbons more favorable.13 The reaction of O(3P) with alkenes was reported to occur through the addition of the O atom to a carbon of the double bond.9,12,13 The mechanism proposed and supported by theoretical calculations for the reaction of Fc and O(3P) involves the addition of O(3P) to one of the ring carbons to form a biradical species where the oxygen atom is bound to one of the carbons in the cyclopentadienyl ring. This species then subsequently reacts to form the photo-products observed in the matrix.1 A mechanism proposed for the reaction of nBuFc and O(3P), discussed in detail in the following section is consistent with the formation of a biradical species in which the oxygen atom is bound to one of the ring carbons as the precursor for the products observed in the matrix after irradiation. A natural bond orbital (NBO) analysis of nBuFc shows that the charge density on the ring carbons is not uniform as shown in Figure 6.1b. The ring carbon (C1) with the n-butyl substituent has a significantly lower negative-charge density than the other ring carbons. As a result it is unlikely that the electrophilic oxygen atom would bond to the C1 to

79

80 form a biradical species. Furthermore, in reactions of atomic oxygen with alkenes, the preference for the initial oxygen atom to attack to occur at the least substituted carbon atom has been documented in the literature.13 This makes the formation of the ketone (keto) or ring-ketone

(rketo), Figure 6.3d, resulting from the initial formation of C1 biradical species unlikely.

The calculated infrared vibrational frequencies and 18O-isotopic shifts (Tables S6.2a-f) for each of the nBuFc pyran regioisomers all show substantial similarities to each other. These similarities are also observed when comparing the calculated results of the nBuFc aldehyde regioisomers and structural isomers to each other (Tables S6.3a-r). Given the uncertainty associated with the calculated spectra and 18O-isotopic shifts, the calculated spectra do not allow for the distinction between one the regioisomer of a given type from another regioisomer of the same type based on spectroscopy. However, some of the ring-aldehyde regioisomers’ calculated spectra do show some significant differences that can be used to eliminate some from consideration (Tables S6.4a-e). Both rald-A and -D are calculated to have intense peaks with

18O-isotopic shifts not observed in the irradiated experimental spectra (Tables S6.4a and d).

Figure 6.2 show spectra of the TJ-dark-codeposited matrix of O3 and nBuFc in argon after red irradiation and Table 6.1 lists the product peaks and assignments. It is likely that multiple regioisomers of each species are present in the matrix. The assignments of product peaks are made to a group of regioisomers and not to a specific member of that group. However, for comprehensibility of the table only the calculated vibrational frequencies and 18O-isotopic shifts of the regioisomer that best matches the assigned peak are given in Table 6.1.

81

Table 6.1 Band Positions and Assignments for the Products from the Twin Jet Deposition of Ozone with n-Butylferrocene upon Irradiation with Light of λ ≥ 600 nm. 18O calcd exptl exptl bandsa calcd bandsb shiftb shift assignments 551 (m) 559 -6 -7 pyran 781 (m) 755 0 -1 ring-aldehyde 846 sh (vw) 841 0 0 aldehyde 889 (m) 894 -4 -6 pyran 927 (s) 956 -7 -8 pyran 945 (vw) 961 -11 -13 pyran 959 (vw) -4 1078 (m) 1091 -5 -4 pyran 1135 (vw) -8 1140 (m) 1160 -4 -5 pyran 1164 (vw) 1199/1196 -2/0 -1 pyran/aldehyde 1360 (vw) 1330 -1 0 ring-aldehyde 1372 (vw) 1339 -2 -3 ring-aldehyde 1440 (m) 1455/1465 0 0 aldehyde/ring-aldehyde 1512/1471 & 1471 (w) 0 0 pyran/ring-aldehyde 1474 1583 (m) 1608 -6 -6 ring-aldehyde 1680 (vs) 1747 -35 -27 aldehyde 2712 (m) 2902 0 0 aldehyde a Frequencies in cm-1 b Calculated at the B3LYP/6-311G++(d,2p) level of theory and unscaled.

82 Reaction Mechanism. The proposed photochemical reaction between nBuFc and O3 starts with production of ground state atomic oxygen with up to as much as +23 kcal/mol of excess thermal energy by photolysis at λ ≥ 600 nm, as already established in the previous section.8 Based on the products observed in the spectra after irradiation, and literature for reactions of O(3P) with alkenes9-12 a possible general reaction of O(3P) and nBuFc is shown in

Scheme 1:

Scheme 1. Proposed general scheme for the (λ ≥ 600 nm) photochemical reaction between

O(3P) and nBuFc.

(1) O(3P) + nBuFc → nBuFc–O (biradical) → nBuFcO (pyran)

(2) O(3P) + nBuFc → nBuFc–O (biradical) → nBuFcO (aldehyde)

(3) nBuFcO (pyran) → nBuFcO (aldehyde)

(4) nBuFcO (aldehyde) → nBuFcO (ring-aldehyde)

The O(3P) can then react with nBuFc to form a pyran (1) and an aldehyde (2). The aldehyde can then subsequently form the ring-aldehyde (4). Reaction (4) is based on the presence of infrared peaks assigned to the ring-aldehyde after red photolysis (λ ≥ 600 nm) which are not significantly present after infrared photolysis (λ ≥ 1000 nm) (Figure 6.5). Reaction (3) is the conversion of the pyran through a ring opening to form the aldehyde. Although the average energy of the aldehyde is 11 kcal/mol higher than the average energy of the pyran reaction (3) could still be possible given the excess energy available from the initial exothermic formation of pyran.

Figure 6.6a gives the reaction potential energy diagram for the formation of pyran-A, reaction (1) and ald-A, reactions (2) and (3). The calculated structures corresponding to potential energy diagrams are given in Figure 6.6b.

83

84

85

86 The calculated structures suggest that the transition states are accessible by irradiation at λ ≥ 600 nm and that the reactions (2) and (3) involve a stable but unobserved biradical [2T→2S] and an epoxide [3S]. The diagram starts with addition of O(3P) to carbon 2 (C2) to form a triplet van der Waals complex (T-vdW-C2), [1T]. The calculations show that these species can react in an essentially barrierless exothermic reaction to form a triplet biradical (T-BR-C2), [2T]. This triplet biradical can undergo an intersystem crossing to form the corresponding singlet biradical

(S-BR-C2), [2S]. The S-BR-C2, [2S], can rearrange to yield an epoxide (S-epox-A), [3S]. The epoxide can undergo further rearrangement to form the pyran (pyran-A) from reaction (1). It should be noted that although a stable S-epox-A could not be located, evidence for its transitory presence comes from the IRC calculated for the transition of S-BR-C2 to pyran-A. This is likely due to a low activation energy barrier for the formation of pyran-A. The energies and structures given for [3S] and [ts3S] in Figure 6.6 are derived from the IRC calculation and the IRC is provided in Figure S6.5. In addition to forming pyran-A, S-BR-C2, [2S], can also rearrange to yield an aldehyde (S- ald-A), [5S], by reaction (2), and this rearrangement has an activation barrier of 11.3 kcal/mol. Figure 6.6a also shows that pyran-A can undergo a ring opening to form ald-A by reaction (3), and that this reaction has an activation energy barrier of 21.0 kcal/mol. The mechanism proposed for reaction of nBuFc and O(3P) is complicated by the multiple regioisomers of each product. The full reaction mechanism proposed when O(3P) reacts with C2 to form (T-vdW-C2) is given in Figure 6.7. It proposes that S-BR-C2, [2S], can also form pyran-B, ald-B, and rald-B. These pathways indicate that reactions (1) – (4) and the regioisomers of products may be connected by an unobserved singlet biradical. rald-A was not listed as a product since it was previously established to not be present in the matrix based on its calculated IR spectrum.

87

88 The formation of the photochemical products’ remaining regioisomers are proposed to follow the same reaction mechanism discussed above; which starts with addition of O(3P) to a carbon in either ring of nBuFc to form a triplet van der Waals complex (T-vdW). However, the formation of a T-vdW complex involving the C1 carbon is unlikely given that this ring carbon is calculated have a significantly lower charge density than the other ring carbons; as already discussed (Figure 6.1b). As a consequence the formation of resulting ketone (keto) and ring- ketone (rketo) would be unlikely. Although the formation of the keto and rketo cannot be totally eliminated as they could be formed from pyran-A via reactions (3) and (4) they would likely not be significant products. The mechanism of the reaction resulting in the formation of regioisomers involving cyclopentadienyl ring carbons (C6-C10) of nBuFc (pyran-D –F, ald-E –I, and rald-E) are proposed to follow the same mechanism as that given for Fc and O(3P).1

Conclusions:

There is strong evidence that upon irradiation with λ ≥ 600 nm the initial step in the photochemical reaction of nBuFc and O3 is photodissociation of O3. This yields a highly reactive

3 3 3 atomic oxygen, O( P), and molecular oxygen, O2( Σ). O( P) reacts with nBuFc to produce the photochemical products observed. The assigned photochemical products consisting of an iron atom with a coordinated n-butylcyclopentadienyl or cyclopentadienyl ring and either: (1) a pyran, (2) an aldehyde, or (3) a bidentate cyclic aldehyde with a seven-membered ring including the iron atom. We were unable to assign the specific regioisomers and conformational isomers of each product present due to similarities the calculated spectra. It is likely that multiple regioisomers of each product are present in the matrix and the potential energy diagram of the reaction mechanism shows multiple paths to some products. Furthermore, n-butyl group may have some effect on the formation of some regioisomers by altering the charge distribution of

89 ring carbons in nBuFc. A reaction of O3 with a ferrocene that has a substituent group that strongly affects charge distribution, such as an amine group, merits further investigation. Lastly, the green matrix formed during the initial dark deposition of O3 with nBuFc is theorized to be the result of an nBuFc-O3 charge transfer complex.

Supplementary Information:

Supplemental spectra, figures and data tables are available in Appendix C.

References

(1) Kugel, R. W.; Pinelo, L. F.; Ault, B. S. Infrared Matrix-Isolation and Theoretical Studies of the Reactions of Ferrocene with Ozone. J Phys Chem A 2014.

(2) Butylferrocene (97%) FT-IR spectrum. In The Aldrich Library of FT-IR spectra; Aldrich: 1997; Vol. 3, pp 4569.

(3) Butylferrocene FT-IR spectrum Liquid Film (SDBS-NO=10729). National Institute of Advanced Industrial Science Technology.

(4) Margolis, D. S.; Tanjaroon, C.; Kukolich, S. G. Measurements of microwave spectra and structural parameters for methylferrocene. Journal of Chemical Physics 2002, 117, 3741.

(5) Subramanian, R.; Karunatilaka, C.; Keck, K. S.; Kukolich, S. G. The gas phase structure of ethynylferrocene using microwave spectroscopy. Inorganic Chemistry 2005, 44, 3137.

(6) Bahou, M.; SchriverMazzuoli, L.; Schriver, A.; Chaquin, P. Structure and selective visible photodissociation of the O-3:Br2 and O-3:BrCl complexes: An infrared matrix isolation and ab initio study. Chemical Physics 1997, 216, 105.

(7) Bahou, M.; SchriverMazzuoli, L.; CamyPeyret, C.; Schriver, A. Photolysis of ozone at 693 nm in solid oxygen. Isotopic effects in ozone reformation. Chemical Physics Letters 1997, 273, 31.

(8) Wagman, D. D.; American Chemical Society; American Institute of Physics; United States; National Bureau of Standards The NBS tables of chemical thermodynamic properties : selected values for inorganic and C1 and C2 organic substances in SI units; ACS and AIP for NBS, 1982.

(9) Boocock, G.; Cvetanović, R. J. Reaction of oxygen atoms with benzene. Canadian Journal of Chemistry 1961, 39, 2436.

90 (10) Cvetanović, R. J. Evaluated chemical kinetic data for the reactions of atomic oxygen O (3P) with unsaturated hydrocarbons. Journal of physical and chemical reference data 1987, 16, 261.

(11) Cvetanović, R. J. Relative rates of reactions of oxygen atoms with olefins. The Journal of chemical physics 1959, 30, 19.

(12) Messaoudi, B.; Mekelleche, S.; Mora-Diez, N. Theoretical study of the complex reaction of O(3P) with cis-2-butene. Theoretical Chemistry Accounts 2013, 132, 1.

(13) Min, Z.; Wong, T.-H.; Su, H.; Bersohn, R. Reaction of O(3P) with Alkenes: Side Chain vs Double Bond Attack. The Journal of Physical Chemistry A 2000, 104, 9941.

91 Chapter 7

Charge Transfer Complexes and Photochemistry of Ozone with Ferrocene and n-

Butylferrocene: A UV-Vis matrix isolation study

Introduction:

Infrared argon (Ar) matrix isolation studies of the reactions of ozone (O3) with ferrocene

(Fc) and of O3 with n-butylferrocene (nBuFc) resulted in the discovery of a photochemical reaction that can be initiated with red light (λ ≥ 600 nm).1,2 Prior to irradiation with red light, vivid green matrices were formed. These green matrices were proposed to be the results of charge transfer complexes forming between O3 and Fc or between O3 and nBuFc. This conclusion was based primarily on the green color of the matrix, suggesting strong red and blue absorptions in the visible spectrum, and slightly red-shifted O3 infrared absorptions. These

3 slightly red-shifted ozone infrared absorptions are characteristic of perturbed O3. Other than the perturbed O3 bands, no other product peaks were observed in the infrared spectra prior to irradiation.

The present work uses ultraviolet-visible spectroscopy to investigate the possible formation of O3 complexes with Fc and with nBuFc during dark deposition. In addition, the formation of O2 complexes with Fc and with nBuFc was also investigated. The O2 contamination resulted from O2 remaining after the O3 samples were purified and from the decomposition of

4 some O3 prior to deposition. This study was done using the low-temperature technique of matrix isolation combined with UV-Vis spectroscopy and theoretical calculations. Qualitative results about products present during dark deposition were obtained using difference spectra.

Theoretical calculations were used to characterize the absorptions and investigate the molecular orbitals involved.

92 Results and Discussion:

Before co-deposition experiments were performed using O3 or O2 and Fc or nBuFc in Ar,

UV-Vis spectra of: Ar, Ar/O2, Ar/O3, Ar/Fc, and Ar/nBuFc matrices were collected (Figures

S7.1b-f in Appendix D). The experimentally collected Fc, nBuFc, O3, and O2 blanks are

5-10 consistent with what is reported in the literature. The dark TJ-deposition of both Ar/O3/Fc and Ar/O3/nBuFc resulted in the formation of green matrices as previously seen in the FT-IR

1,2 experiments. Difference spectra were used to identify the Ar/O3/Fc and Ar/O3/nBuFc dark deposition product peaks present in the spectra. The resulting spectra from 300 to 900 nm, plotted in Figures 1, show several product bands. The difference spectrum of Ar/O3/Fc has a broad product peak with a λmax at ~765 nm, a second product peak with a λmax at ~361 nm, and a weak shoulder at ~313 nm (Figure 7.1a). The difference spectrum of Ar/O3/nBuFc has a broad product peak with a λmax at ~816 nm, a second product peak with a λmax at ~382 nm, and a weak shoulder at ~328 nm (Figure 7.1b). The spectrum of Ar/O2/Fc has a product band with a peak at

λmax = ~366 nm as shown in Figure 7.1c. The spectrum of Ar/O2/nBuFc is shown in Figure 7.1d and has a product band with a large peak at λmax = ~388 nm.

Based on the result of the previous infrared studies, complexes of O3 with Fc and O3 with nBuFc were investigated as possible products.1,2 There are a number of possible configurations for the charge transfer complexes of O3 with Fc (Fc-O3) and O3 with nBuFc (nBuFc-O3).

Calculated configurations of representative charge transfer (CT) complexes considered here are shown in Figure 7.2. In one CT complex the O3 is over one of the rings (Fc-O3A or nBuFc-

O3A). In a second configuration O3 spans the two Cp rings with Fe in the middle (Fc-O3B or nBuFc-O3B). The natural bond orbital (NBO) analysis of the ground state Fc-O3A, Fc-O3B, nBuFc-O3A and nBuFc-O3B complexes showed an increased charge density on the terminal

93

94

95

96 oxygens of O3 and a decreased charge density on the iron of Fc or nBuFc as compared to the uncomplexed species. The optimized ground state structures showed increased oxygen-oxygen bond lengths of O3 in the charge transfer complexes (Fc-O3 and nBuFc-O3) as compared to those of uncomplexed O3 (Table 7.1). The increased oxygen-oxygen bond lengths of O3 in the charge transfer complexes is supported by the observation of slightly red-shifted O3 infrared absorptions seen in the experimental dark deposition FT-IR spectra from the earlier studies of Ar/Fc/O3 and

1,2 Ar/nBuFc/O3. The π*-orbital (LUMO) of O3 is partial populated in the complexes. In addition, as a result of the O2 contamination the van der Waals complexes of O2 with Fc (Fc-O2) and O2 with nBuFc (nBuFc-O2) were also considered (Figure 7.3). A summary of natural population analysis of the species of interest are provided in Table S7.1 of Appendix D. Since a spin flip is unlikely during the formation of these complexes, TD-DFT calculations of singlet only excited states were done for the O3 complexes, O3, Fc, and nBuFc; triplet only TD-DFT calculations were done for O2 and the O2 complexes. Figures 7.4 and 7.5 show the difference spectra from

300 to 900 nm for Ar/Fc/O3, Ar/Fc/O2, Ar/nBuFc/O3, and Ar/nBuFc/O2, as well as the calculated electronic transitions of the complexes.

The difference spectrum of Ar/O3/Fc has a broad product peak with a λmax at ~765 nm, a second product peak with a λmax at ~361 nm, and a weak shoulder at ~313 nm (Figure 7.1a).

Based on the TD-DFT calculations, the broad product peak with a λmax at 765 nm was assigned to Fc-O3A and Fc-O3B. The second product peak with λmax at 361 nm and the weak shoulder at

313 nm were assigned to the complex of Fc-O2. These assignments were based on the TD-DFT calculations and the difference spectrum of Ar/O2/Fc, which has a peak at λmax = ~366 nm

(Figure 7.1c).

97

Table 7.1 Calculateda Oxygen– Oxygen Bond Lengths of O3 (O1— O2—O3).

Uncomplexed O3 (Å) O1—O2 1.25606 O3 O2—O3 1.25606

Charge-transfer complexes (Å) O1—O2 1.26281 Fc-O3A O2—O3 1.26435

O1—O2 1.26788 Fc-O3B O2—O3 1.26794

O1—O2 1.26411 nBuFc-O3A O2—O3 1.26753

O1—O2 1.26888 nBuFc-O3B O2—O3 1.26904 a Computed at the B3LYP/6- 311++G(d,2p) level of theory.

98

99

100

101 Figure 7.4 shows the difference spectra from 300 to 900 nm for Ar/Fc/O3 and Ar/Fc/O2, as well as the calculated electronic transitions for Fc-O3A, Fc-O3B, and Fc-O2 complexes.

The difference spectrum of Ar/O3/nBuFc has a broad product peak with a λmax at ~816 nm, a second product peak with a λmax at ~382 nm, and a weak shoulder at ~328 nm (Figure

7.1b). Based on the TD-DFT calculations the broad product peak with a λmax at 816 nm was assigned to nBuFc-O3A and nBuFc-O3B. The weak shoulder at 328 nm can also be assigned to nBuFc-O3B and nBuFc-O2. The second product peak with λmax at 383 nm was assigned to nBuFc-O2. These assignments were based on the TD-DFT calculations and the difference spectrum of Ar/O2/nBuFc; which has a product band with a large peak at λmax = ~388 nm (Figure

7.1d). Figure 7.5 shows the difference spectra from 300 to 900 nm for Ar/nBuFc/O3 and

Ar/nBuFc/O2, as well as the calculated electronic transitions for nBuFc-O3A, nBuFc-O3B, and nBuFc-O2 complexes.

The experimental UV-Vis spectra for both Ar/Fc/O3 and Ar/nBuFc/O3 after irradiation with red light show a significant decrease in the peaks at λmax≈ 765 and 816 nm respectively

(Figures 7.6a and 7.6b). It is the broad charge-transfer product bands at 765 and 816 nm that are absorbing the light in this range (λ ≥ 600 nm). It should also be noted that the Fc-O2 and nBuFc-

O2 do not exhibit any absorption in this region and are not involved in the photochemical reaction at λ ≥ 600 nm (Figures 7.6c and 7.6d). The major photochemical products previously assigned using IR spectroscopy consist of an iron atom with a coordinated n- butylcyclopentadienyl or cyclopentadienyl ring and either: pyran or an aldehyde.1,2 Figure 7.7 shows the optimized structures for the major photochemical products.

102

103

104

105 Figures 7.8a and 7.8b show UV-Vis difference spectra of the Ar/Fc/O3 and Ar/nBuFc/O3 matrices after 45 min of red irradiation minus the dark deposited spectrum. Based on the TD-

DFT calculated electronic transitions for the previously assigned photochemical products are not inconsistent with the observed UV-Vis spectra after the matrices have been irradiated with red light (λ ≥ 600 nm) (Figures 7.8c-f).1,2

The band at 765 nm was assigned to Fc-O3A and Fc-O3B and a review of the relevant molecular orbitals shows (Figure 7.9) that this peak is primarily due to electronic transitions from two of the n-orbitals on ozone to the π*-orbital on ozone. Table 7.2 lists the calculated absorption bands, the corresponding excitation energies, and the molecular orbitals involved that have been assigned to the experimental peak at 765 nm for each species. The band at 816 nm was assigned to nBuFc-O3A and nBuFc-O3B and a review of those relevant molecular orbitals

(Figure 7.10) shows that that peak is also primarily due to electronic transitions from the n- orbitals on ozone to π*-orbital on ozone. Table 7.3 list the calculated absorption bands, the corresponding excitation energies, and the molecular orbitals involved that have been assigned to the experimental peak at 816 nm for each species.

These transitions all involve the excitation of ozone non-bonding electrons to the π*- orbital associated with ozone. It has been shown in the literature that O3 has two very weak and broad absorption bands from 420 nm to 1048 nm; known as the Chappuis (420 nm ≤ λ ≤ 700 nm)

4,11,12 and Wulf (700 nm ≤ λ ≤ 1048 nm) bands. The spectrum of O3 that we collected (Figure

S7.1b) did not have detectable Chappuis or Wulf bands. However, these finding are not inconsistent with the literature given the experimental conditions. The Chappuis band is more than three orders of magnitude weaker and the Wulf band is about five orders of magnitude weaker than the Hartley band.7,13

106

107

108

109

Table 7.2 Assignment of Calculateda Absorption Bands, the Corresponding Excitation Energies, and Oscillator Strength of Fc-O3A and Fc-O3B to the Ar/Fc/O3 Experimental Peak at λmax ≈ 765 nm. excitation absorption energy oscillator transition (nm) (kcal/mol) strength n (I) → π* Fc-O A 804 35.6 0.0280 3 n (II) → π*

n (I) → π* Fc-O B 821 34.8 0.0010 3 n (II) → π* a Calculated at the B3LYP/6-311G++(d,2p) level of theory and unscaled.

110

111

Table 7.3 Assignment of Calculateda Absorption Bands, the Corresponding Excitation Energies, and Oscillator Strength of nBuFc-O3A and nBuFc-O3B to the Ar/nBuFc/O3 Experimental Peak at λmax ≈ 816 nm.

excitation absorption energy oscillator transition (nm) (kcal/mol) strength n (I) → π* 978 29.2 0.0271 nBuFc-O A 3 n (II) → π* 799 35.8 0.0001

n (I) → π* 929 30.8 0.0008 nBuFc-O B 3 n (II) → π* 834 34.3 0.0006 a Calculated at the B3LYP/6-311G++(d,2p) level of theory and unscaled.

112 There has been a considerable amount of debate over electronic transitions of these bands.12-14

The only transitions that were calculated using TD-DFT from 420 nm to 1048 nm for uncomplexed O3 are at 536 and 602 nm and have oscillator strengths of less than 0.0001. These absorptions at 536 and 602 nm correspond to the transition from the n-orbital I and n-orbital II to the unoccupied π*-orbital respectively (Figure 7.11 and Table 7.4). Experimental results from the literature show that the Chappuis band’s two most intense features are located at 568 and 597 nm.7 The Chappuis band has been reported in the literature to have measured oscillator strength of ~0.000032 and this would not be inconsistent with our calculated value of < 0.0001. It has been shown experimentally that excitation of the Chappuis and Wulf bands with λ ≤ 600 nm

3 causes the photodissociation of O3 resulting in the formation of ground state O( P) and

3 15,16 3 O2( Σ). Subsequent reactions of O( P) with Fc and nBuFc have been documented in previous infrared and computational studies of the photochemical reactions of O3 with Fc and O3 with

1,2 nBuFc. These findings coupled with the stronger bands observed at λmax≈ 765 and 816 nm in the Ar/Fc/O3 and Ar/nBuFc/O3 spectra indicate that the formation of the charge transfer complex lowers the excitation energy and increases the oscillator strength of the n → π* transitions allowing the π*-orbital of O3 to be populated more readily is when red light (λ ≥ 600 nm) is absorbed (Tables 7.2-7.4).

113

114

Table 7.4 Calculateda Absorption Bands, the Corresponding Excitation Energies, and Oscillator Strength of O3. excitation absorption energy oscillator transition (nm) (kcal/mol) strength n (I) → π* 602 47.5 0.0000 O 3 n (II) → π* 536 53.3 0.0000 a Calculated at the B3LYP/6-311G++(d,2p) level of theory and unscaled.

115 Conclusions:

The computational data and experimental UV-Vis spectra provide strong support for the formation of charge transfer complexes of O3 with Fc and of O3 with nBuFc. The combined results of both the infrared and UV-Vis studies provide strong evidence that the formation of these charge transfer complexes affects how readily O3 undergoes photodissociation when irradiated with low-energy red light.1,2,7,10 In addition, the calculated UV-Vis spectra of the photochemical products assigned using infrared spectroscopy are consistent with those observed in the experimental UV-Vis spectra. The reactions of O3 with other metallocenes may also yield promising results.

Supporting Information

Supplemental computational and experimental information is available in Appendix D.

References

(1) Kugel, R. W.; Pinelo, L. F.; Ault, B. S. Infrared Matrix-Isolation and Theoretical Studies of the Reactions of Ferrocene with Ozone. J Phys Chem A 2014.

(2) Pinelo, L.; Gudmundsdottir, A. D.; Ault, B. S. In Ozone and n-butylferrocene: Fascinating Photochemistry Unpublished Material.

(3) Bahou, M.; SchriverMazzuoli, L.; Schriver, A.; Chaquin, P. Structure and selective visible photodissociation of the O-3:Br2 and O-3:BrCl complexes: An infrared matrix isolation and ab initio study. Chemical Physics 1997, 216, 105.

(4) Horváth, M.; Bilitzky, L.; Hüttner, J. Ozone; Elsevier: Amsterdam; New York, 1985.

(5) Armstron.At; Smith, F.; Elder, E.; Mcglynn, S. P. Electronic Absorption Spectrum of Ferrocene. Journal of Chemical Physics 1967, 46, 4321.

116 (6) Sohn, Y. S.; Hendrickson, D. N.; Hart Smith, J.; Gray, H. B. Single-crystal electronic spectrum of ferrocene at 4.2°K. Chemical Physics Letters 1970, 6, 499.

(7) Vaida, V.; Donaldson, D. J.; Strickler, S. J.; Stephens, S. L.; Birks, J. W. A Reinvestigation of the Electronic-Spectra of Ozone - Condensed-Phase Effects. J Phys Chem-Us 1989, 93, 506.

(8) Krupenie, P. H. The Spectrum of Molecular Oxygen. Journal of Physical and Chemical Reference Data 1972, 1, 423.

(9) Rabie, U. M. Intra- and intermolecular charge transfer: twin themes and simultaneous competing transitions involving ferrocenes. Spectrochimica acta. Part A, Molecular and biomolecular spectroscopy 2009, 74, 746.

(10) Tarr, A. M.; Wiles, D. M. Electronic Absorption Spectra and Photodecomposition of Some Substituted Ferrocenes. Canadian Journal of Chemistry 1968, 46, 2725.

(11) Orphal, J. A critical review of the absorption cross-sections of O3 and NO2 in the ultraviolet and visible. Journal of Photochemistry and Photobiology A: Chemistry 2003, 157, 185.

(12) Minaev, B.; Agren, H. The Interpretation of the Wulf Absorption-Band of Ozone. Chemical Physics Letters 1994, 217, 531.

(13) Grebenshchikov, S. Y.; Schinke, R.; Qu, Z. W.; Zhu, H. Absorption spectrum and assignment of the Chappuis band of ozone. J Chem Phys 2006, 124, 204313.

(14) Woywod, C.; Stengle, M.; Domcke, W.; Flöthmann, H.; Schinke, R. Photodissociation of ozone in the Chappuis band. I. Electronic structure calculations. The Journal of Chemical Physics 1997, 107, 7282.

(15) Kuze, H.; Sato, T.; Kambe, T.; Hayashida, S.; Tatsumi, Y. Isotope-selective photodissociation of ozone molecules induced by infrared laser irradiation. Chemical Physics Letters 2008, 455, 156.

(16) Bahou, M.; SchriverMazzuoli, L.; CamyPeyret, C.; Schriver, A. Photolysis of ozone at 693 nm in solid oxygen. Isotopic effects in ozone reformation. Chemical Physics Letters 1997, 273, 31.

117 Appendix A Supporting Information for: “Infrared Matrix Isolation and Theoretical Study of the Initial Intermediates in the Reaction of Ozone with Cycloheptene”

Table S4.1 Calculateda Frequenciesb and Intensitiesc for the Possible Intermediates from the Ozonolysis of Cycloheptene

Chair cis- Boat cis- Criegee Criegee Chair cis- Boat cis- primary primary Intermediate Intermediate secondary secondary ozonide ozonide Configuration 1 Configuration 2 ozonide ozonide

IR IR IR IR IR IR Freq. Freq. Freq. Freq. Freq. Freq. Int. Int. Int. Int. Int. Int. 375 2 55 0 17 2 363 8 342 4 107 0 415 0 146 0 43 1 435 1 424 2 151 3 429 7 183 3 70 13 507 9 488 2 226 1 504 3 256 1 76 5 544 2 539 8 249 1 507 0 292 2 84 0 664 3 615 2 265 1 698 1 309 2 101 2 737 6 689 7 300 2 701 2 351 0 167 3 778 2 719 1 405 3 717 1 413 4 238 1 823 19 759 1 426 1 736 28 496 5 274 1 848 9 808 4 498 5 808 4 524 2 307 3 876 84 813 3 514 2 844 5 558 0 338 12 889 2 859 8 636 8 864 3 692 29 381 2 911 3 867 14 715 1 906 8 701 1 447 1 992 3 888 5 750 8 916 1 750 2 527 21 997 9 911 13 772 2 949 16 773 5 578 10 1011 26 953 41 807 13 992 2 808 3 744 5 1036 4 979 24 811 1 1010 36 818 3 764 2 1059 0 994 20 834 0 1031 3 837 3 809 3 1102 9 1006 14 851 5 1034 1 898 6 852 6 1140 14 1045 2 855 28 1054 0 907 6 885 9 1181 6 1108 19 905 5 1089 2 943 31 914 91 1237 14 1116 82 941 63 1110 0 979 0 927 7 1262 2 1124 50 950 36 1177 1 994 29 968 53 1293 3 1137 24 1004 2 1206 2 1010 4 994 2 1315 9 1185 9 1016 23 1224 0 1024 1 1023 22 1335 7 1251 1 1044 3 1226 3 1048 2 1062 22 1343 1 1262 2 1100 86 1293 1 1070 0 1091 8 1347 2 1298 4 1120 8 1301 1 1125 0 1099 1 1367 8 1301 3 1125 13

118 1326 1 1140 10 1152 3 1376 6 1318 2 1134 23 1329 2 1199 0 1188 7 1402 3 1355 6 1196 0 1366 0 1210 2 1240 3 1426 4 1362 6 1222 15 1375 1 1219 0 1251 6 1462 6 1385 2 1260 9 1383 0 1294 0 1291 6 1479 8 1386 4 1288 10 1390 3 1294 1 1300 0 1494 1 1392 12 1291 7 1394 0 1312 0 1328 1 1498 0 1394 8 1315 1 1394 2 1326 2 1336 4 1514 9 1403 2 1337 10 1419 2 1333 0 1363 21 1550 41 1412 7 1348 0 1493 3 1350 1 1386 6 1797 228 1482 2 1377 6 1495 5 1378 0 1389 11 2873 95 1485 4 1386 2 1497 4 1386 3 1399 7 3005 10 1488 14 1391 8 1505 8 1395 2 1420 4 3013 7 1495 8 1397 5 1516 14 1395 5 1469 1 3019 7 1500 12 1405 7 2999 9 1403 0 1478 5 3022 22 3009 12 1413 3 3003 3 1489 0 1492 5 3027 21 3018 24 1489 0 3005 17 1497 2 1498 11 3035 30 3020 3 1494 1 3008 30 1505 10 1504 9 3060 35 3024 34 1500 11 3026 50 1512 9 1560 0 3077 1 3030 29 1509 8 3033 2 1533 13 1805 240 3088 17 3033 46 1528 17 3036 13 3005 9 2858 111 3096 18 3048 51 3004 7 3045 54 3016 18 2996 2 3167 6 3050 47 3011 32 3053 38 3018 33 2998 7 3055 24 3024 85 3057 51 3042 54 3006 18 3067 7 3031 26 3071 32 3042 8 3010 24 3069 49 3036 10 3077 42 3050 15 3020 14 3085 37 3040 24 3053 34 3042 16 3045 24 3054 3 3052 44 3064 29 3056 37 3056 34 3087 12 3067 11 3095 22 3092 31 3085 32 3114 6 3100 21 3098 41 3132 7 3111 28 a Calculated at the B3LYP/6-311G++(d,2p) level of theory and unscaled. b Frequencies in cm-1. c Intensities in km mol-1.

119 Appendix B Supporting Information for: “Matrix Isolation Study of the Ozonolysis of 1,3- and 1,4-Cyclohexadiene: Identification of Novel Reaction Pathways”

120 Table S5.1 Experimental and Literature Bands of Benzene, H2O3, and BHC

H2O3 H2O3 BHC Calc. Exptl. Benzene Benzene Exptl. Calc. Calc. Shift for BHC Exptl. Freq. Calc. Freq. Freq. Freq. Freq. BHC Freq. (cm-1)a (cm-1)c (cm-1)b (cm-1)c (cm-1)c (cm-1) (cm-1) 346 380 378 -2 344

412 410 -1

412 412 0

387 427 491 64 451

509 528 535 7 516

622 621 -1

622 622 0

679 688 696 8 687

717 711 -5

776 781 778 -2 774

865 867 3

867 872 5

821 932 934 2 823

994 989 -5

996 992 -4

1010 1009 -2

1017 1013 -5

1025 1024 -1

1037 1059 1058 -1 1036

1040 1059 1059 0 1040

1173 1176 3

1197 1198 1

1197 1199 2

1336 1334 -2

1347 1392 1374 -17 1330

1383 1382 -1

1359 1397 1419 22 1381

1480 1511 1509 -2

1481 1511 1510 -1 1479

1633 1630 -3 1478

1633 1631 -2

1812

3040 3153 3162 9 3049

3047 3162 3171 8 3055

3075 3163 3172 9 3084

3078 3179 3186 7 3085

3095 3179 3186 7 3102

3100 3189 3195 6 3106

121 3530 3734 3635 -99 3431

3530 3739 3724 -15 3515

a Spoliti, M.; Cesaro, S.; Grosso, V. The I.R. Spectrum of C6H6 and C6D6 in Inert Gas Matrices at 15 K. Spectrochimica acta. 1976, 32A, 145-147. b Engdahl, A.; Nelander, B. The Vibrational Spectrum of H2O3. Science 2002, 295, 482-483. c Calculated at the B3LYP/6-311G++(d,2p) level of theory.

122 Table S5.2 Computeda Energies of Species Relevant to the Reaction of 1,4-Cyclohexadiene with Ozone

Potential Products (kcal/mol) ozone & 1,4-CHD 3

1,4-CHDO3 0 TS-1 0.5 POZ -52 TS-2 -39 CI -62 TS-3 -59 SOZ -98 TS-4 0.4 BHC -73 a Energies calculated at the B3LYP/6-311G++(d,2p) level of theory.

123 Table S5.3 Computeda Energies of Species Relevant to the Reaction of 1,3-Cyclohexadiene with Ozone Potential Products (kcal/mol) Ozone & 1,3-CHD →SOZ 5 Ozone & 1,3-CHD →PI 7

1,3-CHDO3 0 TS-1 0.1 POZ-A -51 POZ-B -51 POZ-C -51 POZ-D -51 POZ-E -44 POZ-F -44 TS-2 -37 CI-A -70 CI-B -66 TS-3 -59 SOZ-A -94 SOZ-B -93 SOZ-C -93 TS-4 2 PI-A -42 PI-B -42 I-A -50 I-B -87 I-C -86 I-D -108 I-E -112 I-F -178 a Energies calculated at the B3LYP/6- 311G++(d,2p) level of theory.

124 Table S5.4 Calculateda Frequenciesb and Intensitiesc of Intermediates Assigned in the Reaction of 1,3-CHD and Ozone. POZ-A CI-A 16O 18O 16O 18O IR IR IR IR Freq. Freq. Shift Freq. Freq. Shift Int. Int. Int. Int. 71 1 69 1 -2 16 2 15 2 0 185 1 183 1 -2 56 3 54 3 -2 259 1 255 1 -4 66 3 65 2 -1 328 0 318 0 -11 98 12 96 11 -2 386 5 373 5 -12 172 5 167 4 -4 421 0 417 0 -4 198 4 193 4 -5 474 3 468 3 -6 207 2 203 2 -4 513 4 508 4 -5 313 6 305 6 -8 673 3 649 6 -24 364 3 361 3 -2 702 24 678 20 -24 475 1 463 1 -12 714 20 690 14 -24 483 2 479 1 -3 744 6 738 3 -6 540 17 531 17 -9 755 14 746 19 -9 666 42 665 42 -1 810 5 807 5 -3 738 41 730 52 -9 888 7 857 1 -31 775 21 770 18 -5 903 3 886 9 -16 859 14 857 17 -2 948 6 935 10 -13 951 94 922 85 -29 978 37 963 47 -15 962 4 960 2 -2 1001 22 997 13 -4 992 2 992 5 0 1009 6 1005 1 -4 999 52 986 38 -13 1026 2 1018 1 -7 1024 15 1023 12 -1 1069 3 1069 3 0 1072 13 1068 8 -4 1074 1 1073 1 -2 1085 9 1084 5 -1 1125 3 1124 3 -1 1100 3 1097 6 -3 1179 3 1178 3 0 1207 9 1206 9 -1 1222 1 1222 1 -1 1253 1 1251 2 -2 1255 3 1255 2 0 1278 20 1277 22 -1 1302 5 1300 5 -2 1310 8 1310 8 0 1343 5 1341 6 -2 1338 13 1318 11 -21 1348 0 1348 0 0 1370 16 1369 18 -1 1362 7 1361 7 -1 1425 3 1418 2 -7 1373 7 1373 6 0 1450 5 1449 6 -2 1406 2 1405 2 0 1458 10 1456 8 -2 1423 1 1423 1 0 1473 11 1473 10 0 1483 9 1483 9 0 1509 2 1504 2 -5 1503 7 1503 7 0 1642 48 1641 47 -1

125 1703 4 1703 4 0 1806 249 1770 238 -37 2997 20 2997 20 0 2870 87 2870 87 0 3016 8 3016 8 0 2998 7 2998 7 0 3034 40 3034 40 0 3010 10 3010 10 0 3045 8 3045 8 0 3027 4 3027 4 0 3056 27 3056 27 0 3086 12 3086 12 0 3091 21 3091 21 0 3142 1 3142 1 0 3145 7 3145 7 0 3157 6 3157 6 0 3172 14 3172 14 0 3179 4 3179 4 0 a Calculated at the B3LYP/6-311G++(d,2p) level of theory and unscaled. b Frequencies in cm-1. c Intensities in km mol-1.

126 Appendix C Supporting Information for: “Low-energy Photochemistry of Ozone and n-Butylferrocene: A matrix isolation study”

127

128

129

130

131

132

133

134

135 Table S6.1a-hh Calculateda Infrared Vibrational Frequenciesb and Intensitiesc of n- Butylferrocene and Proposed Products a Calculated at the B3LYP/6-311G++(d,2p) level of theory and unscaled. b Frequencies in cm-1. c Intensities in km mol-1.

Table S6.1a) n-butylferrocene Freq. IR Int. Freq. IR Int. Freq. IR Int. 17 0 863 0 1409 0 35 0 877 5 1416 2 55 0 903 0 1432 0 72 0 906 0 1448 0 122 0 907 1 1449 0 129 0 908 0 1487 1 162 1 931 0 1492 0 212 2 942 6 1502 3 243 0 1005 1 1505 7 255 0 1019 7 1506 0 299 0 1020 10 1516 8 312 0 1041 11 2998 7 373 1 1056 0 3005 9 381 0 1061 8 3015 2 430 7 1069 1 3015 34 475 17 1070 0 3018 62 479 20 1077 1 3033 2 494 16 1084 0 3056 36 603 0 1121 6 3075 66 604 0 1132 14 3079 44 605 0 1223 0 3207 1 640 0 1250 3 3211 4 719 1 1267 0 3213 0 732 3 1282 2 3214 0 794 0 1283 0 3221 8 819 6 1311 0 3228 10 821 0 1332 1 3228 10 828 72 1357 0 3233 6 841 9 1376 0 3240 1 846 0 1380 0 851 0 1392 3 852 5 1394 0

136 Table S6.1b) pyran-A

Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 32 0 0 850 0 0 1388 0 0 42 0 0 853 2 0 1393 2 0 55 0 0 866 4 -1 1399 3 -1 75 0 0 888 13 0 1417 2 0 127 0 0 909 2 -1 1433 3 0 135 0 -1 911 1 0 1452 1 0 151 1 -1 918 0 0 1460 1 0 210 1 0 927 4 -3 1485 2 0 234 0 0 935 2 -1 1490 1 0 248 0 -1 961 34 -11 1493 1 0 270 0 -2 1001 4 -3 1498 6 -1 298 1 -1 1013 29 -2 1504 6 0 348 3 -4 1019 12 -1 1505 2 0 353 4 -2 1027 20 -8 1517 8 0 368 4 -2 1029 17 0 3000 8 0 397 3 0 1055 0 0 3006 3 0 433 14 -2 1074 0 0 3014 36 0 441 24 0 1076 0 0 3018 17 0 477 23 -1 1091 26 -5 3019 45 0 544 3 -5 1107 8 -1 3036 2 0 559 11 -6 1120 5 0 3058 32 0 602 7 -4 1135 9 0 3076 64 0 605 1 0 1160 14 -4 3080 42 0 608 1 -1 1221 1 -1 3163 15 0 690 2 -2 1232 1 0 3168 2 0 737 3 0 1283 1 0 3179 17 0 782 4 -3 1287 0 0 3189 11 0 799 0 0 1297 0 -1 3213 0 0 819 45 0 1313 1 -1 3217 1 0 822 7 -1 1333 1 0 3230 8 0 833 16 -1 1358 1 0 3231 7 0 850 4 0 1384 1 0 3243 2 0

137 Table S6.1c) pyran-A Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 19 0 0 851 0 0 1385 1 0 50 0 0 852 2 0 1389 1 0 56 0 0 862 8 -1 1393 2 0 73 0 0 894 15 -4 1415 1 0 129 0 0 904 5 -1 1450 1 0 136 0 0 912 10 -2 1453 1 0 153 1 -2 918 2 0 1457 1 0 197 4 -1 925 2 0 1461 1 0 232 0 0 933 1 0 1480 1 0 244 2 -1 956 29 -7 1493 1 0 264 1 -2 988 19 -11 1503 2 0 294 2 -1 1001 5 0 1503 7 0 349 1 -1 1019 8 0 1516 6 0 359 2 -1 1029 8 0 1521 18 0 393 9 -5 1044 6 -1 2998 16 0 401 10 -1 1054 0 0 3007 8 0 424 12 -2 1075 10 -3 3017 38 0 456 20 -2 1075 0 0 3018 14 0 491 18 -1 1076 10 -1 3024 39 0 514 10 -2 1114 1 -1 3045 2 0 600 4 -4 1119 13 0 3066 11 0 603 2 -1 1136 10 0 3076 72 0 607 6 -2 1159 23 -6 3078 49 0 620 1 -1 1199 21 -2 3167 7 0 660 6 -2 1224 2 0 3170 2 0 736 2 0 1270 2 0 3182 21 0 787 17 -13 1287 1 0 3199 10 0 796 0 0 1288 0 0 3214 0 0 817 19 0 1320 2 0 3217 0 0 832 6 0 1334 1 0 3230 8 0 835 42 0 1360 2 0 3231 8 0 846 12 0 1381 5 -2 3243 2 0

138 Table S6.1d) pyran-B Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 21 0 0 848 4 0 1387 0 0 42 0 0 851 1 0 1395 2 0 49 0 0 852 1 0 1406 2 0 76 0 0 896 5 -3 1413 8 -3 121 0 -1 906 3 0 1416 2 0 126 0 0 908 1 0 1443 0 0 168 0 0 914 2 -1 1451 1 0 189 0 -3 916 1 0 1460 1 0 238 2 -1 932 1 -2 1486 1 0 239 0 0 949 77 -13 1493 1 0 276 0 -2 970 5 -14 1504 2 0 311 1 0 1005 1 0 1504 7 0 348 6 -3 1019 8 0 1516 7 0 363 0 0 1027 2 -1 1529 19 0 367 1 -1 1029 15 0 2999 9 0 418 16 0 1056 0 0 3005 2 0 425 20 -2 1073 0 0 3015 40 0 443 27 -1 1076 0 0 3017 3 0 507 3 -4 1094 1 -1 3019 55 0 518 8 -1 1120 6 0 3034 1 0 541 17 -4 1135 10 0 3057 34 0 604 0 0 1144 7 -2 3077 63 0 605 0 0 1169 32 -9 3080 42 0 658 0 -3 1225 1 0 3159 2 0 686 5 -3 1230 0 0 3160 10 0 734 3 0 1285 2 0 3177 33 0 776 3 -4 1287 0 0 3181 3 0 796 0 0 1294 0 -1 3213 0 0 812 1 -1 1322 0 -1 3217 1 0 817 8 0 1335 1 0 3230 8 0 818 58 0 1361 0 0 3231 8 0 841 3 0 1384 0 0 3243 2 0

139 Table S6.1e) pyran-C Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 18 0 0 857 1 0 1393 3 0 44 0 0 861 4 -3 1406 5 -2 53 0 0 881 7 -1 1410 1 0 80 0 0 907 0 0 1416 2 0 124 0 -1 908 13 -6 1433 1 0 129 0 -1 913 21 -2 1436 1 0 161 0 -1 914 6 -1 1459 0 0 204 1 0 933 0 0 1490 1 0 239 0 0 944 3 -2 1493 0 0 254 0 0 956 42 -7 1500 9 -1 271 0 -3 964 7 -13 1504 4 0 315 0 0 1006 1 0 1505 7 0 347 8 -4 1016 25 -1 1510 3 0 357 7 -3 1037 4 0 1517 8 0 369 1 -1 1043 8 0 2999 5 0 416 17 0 1055 0 0 3006 15 0 429 15 -1 1066 6 0 3016 6 0 439 14 0 1075 2 0 3017 32 0 484 15 0 1089 0 0 3020 50 0 515 10 -1 1099 14 -4 3036 6 0 554 13 -9 1120 6 0 3060 28 0 605 0 0 1143 6 -2 3076 68 0 642 1 -3 1171 20 -7 3079 43 0 644 0 0 1225 0 0 3167 5 0 647 1 -3 1252 2 0 3169 2 0 717 1 0 1270 0 0 3183 24 0 735 3 0 1284 2 0 3183 13 0 795 0 0 1313 0 0 3200 6 0 816 8 0 1316 1 -1 3208 1 0 818 7 0 1332 1 0 3214 4 0 821 39 0 1360 0 0 3223 8 0 848 6 0 1391 0 0 3234 4 0

140 Table S6.1f) pyran-D Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 33 1 0 857 1 -1 1394 3 0 45 0 0 863 9 -3 1404 4 -2 54 0 0 884 4 0 1407 1 0 81 0 0 907 3 0 1416 2 0 127 0 0 907 5 0 1433 1 0 132 0 0 909 25 -9 1443 0 0 147 1 -2 915 5 0 1460 0 0 206 1 -1 934 0 0 1490 1 0 237 0 0 941 9 -1 1493 0 0 255 0 0 953 53 -7 1501 9 -1 284 1 -2 965 5 -14 1504 4 0 311 0 0 1005 1 0 1505 7 0 341 7 -4 1015 21 -1 1509 1 0 361 3 -3 1037 4 0 1517 8 0 368 4 0 1048 9 0 2999 8 0 407 16 0 1055 0 0 3006 5 0 429 14 -1 1063 5 0 3016 32 0 460 14 0 1075 0 0 3017 4 0 476 13 0 1087 0 0 3018 64 0 516 13 -1 1100 14 -4 3034 1 0 553 15 -9 1120 6 0 3057 34 0 607 0 0 1142 9 -2 3076 66 0 640 0 -2 1172 21 -7 3080 43 0 644 2 -1 1225 0 0 3167 4 0 648 1 -3 1256 1 0 3169 2 0 716 1 0 1270 0 0 3181 22 0 736 3 0 1285 3 0 3183 15 0 797 0 0 1314 0 0 3200 6 0 814 24 0 1316 1 -1 3209 2 0 818 6 0 1333 1 0 3216 3 0 829 4 0 1360 0 0 3224 8 0 838 22 0 1389 0 0 3237 4 0

141 Table S6.1g) pyran-E Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 21 1 0 856 19 -2 1393 3 0 39 0 0 857 0 0 1405 6 -2 51 0 0 872 9 -2 1415 1 0 73 0 0 907 2 0 1418 1 0 124 0 0 907 3 0 1432 0 0 124 0 0 911 45 -9 1435 1 0 168 0 0 915 1 0 1460 0 0 182 1 -3 932 0 0 1488 1 0 243 0 0 947 2 0 1492 0 0 253 1 0 957 52 -7 1500 7 -1 296 1 -1 966 6 -15 1504 4 0 309 0 0 1005 2 0 1505 7 0 345 7 -6 1015 23 -1 1510 3 0 349 6 0 1036 4 0 1516 9 0 370 2 -2 1038 7 0 2998 8 0 408 17 -1 1056 0 0 3006 8 0 422 10 0 1062 4 0 3015 3 0 466 13 0 1083 2 0 3016 31 0 468 19 0 1086 0 0 3018 62 0 514 9 -1 1098 12 -4 3034 2 0 552 15 -9 1121 6 0 3057 34 0 604 0 0 1143 7 -2 3076 65 0 638 1 -2 1172 23 -7 3080 43 0 644 0 -1 1224 0 0 3168 4 0 647 1 -3 1248 2 0 3169 0 0 716 1 0 1269 0 0 3183 33 0 733 3 0 1282 2 0 3184 5 0 795 0 0 1313 0 0 3200 6 0 811 0 0 1316 1 -1 3208 3 0 814 30 0 1333 1 0 3210 3 0 833 5 0 1357 0 0 3225 6 0 844 5 0 1387 1 0 3238 4 0

142 Table S6.1h) (E)-ald-A Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 38 0 0 846 21 0 1389 1 0 44 1 0 852 0 0 1393 2 0 59 1 -1 855 2 0 1411 2 -6 70 1 -1 869 8 0 1416 2 0 79 2 -1 905 2 0 1433 5 0 128 1 0 921 1 0 1453 2 0 135 1 -2 926 3 0 1463 1 0 154 1 -1 933 4 0 1483 1 0 182 3 -1 940 9 0 1491 0 0 209 5 0 958 5 0 1498 4 0 235 0 0 976 1 -2 1502 2 0 250 1 0 1006 2 0 1504 7 0 285 5 -1 1022 7 0 1515 7 0 307 2 -3 1026 5 0 1712 404 -35 341 1 -2 1032 17 0 2907 37 0 370 4 0 1055 0 0 2999 4 0 386 4 -1 1065 14 0 3003 30 0 407 16 0 1075 0 0 3017 37 0 414 14 -1 1080 0 0 3023 8 0 463 11 0 1119 5 0 3030 25 0 512 10 0 1123 5 0 3047 16 0 555 2 -3 1138 7 0 3074 18 0 599 4 -4 1201 19 0 3078 48 0 601 4 -1 1220 28 0 3079 47 0 607 0 0 1236 2 0 3091 25 0 668 3 -1 1284 1 0 3144 16 0 735 4 0 1289 0 0 3165 10 0 759 7 -1 1300 9 0 3217 1 0 800 2 0 1334 0 0 3225 0 0 819 5 0 1348 11 0 3235 3 0 830 38 0 1362 11 0 3238 4 0 837 22 0 1385 3 0 3249 1 0

143 Table S6.1i) (Z)-ald-A Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 33 0 0 851 1 0 1390 2 0 47 0 0 853 4 0 1393 2 0 60 0 0 862 9 0 1414 8 -5 68 1 -1 880 33 -1 1417 1 0 82 1 -1 903 3 0 1437 6 -1 114 1 -2 919 1 0 1453 2 0 131 0 0 922 0 0 1464 1 0 156 1 -1 926 4 0 1486 10 0 178 6 0 931 3 0 1492 1 0 197 0 -1 955 4 0 1499 34 -1 208 7 0 994 1 -2 1503 6 0 235 0 0 1003 1 0 1504 2 0 254 2 0 1022 8 0 1516 7 0 284 2 -3 1028 3 0 1741 293 -35 296 2 -1 1033 20 0 2856 99 0 355 1 -1 1054 4 0 2998 8 0 371 4 0 1059 15 0 3004 4 0 392 6 -1 1075 0 0 3014 39 0 406 11 0 1079 0 0 3016 3 0 448 20 -1 1118 7 0 3019 49 0 497 9 -1 1121 1 0 3037 8 0 540 8 -1 1138 7 0 3060 27 0 598 1 0 1160 0 0 3077 61 0 606 0 0 1210 18 0 3080 42 0 640 7 -4 1234 4 0 3091 23 0 678 1 -3 1281 1 0 3138 19 0 733 5 0 1289 0 0 3170 10 0 734 6 0 1303 10 0 3217 0 0 794 2 0 1333 0 0 3222 0 0 813 7 0 1354 2 0 3234 4 0 822 38 0 1360 33 0 3237 5 0 842 28 0 1386 3 0 3247 1 0

144 Table S6.1j) (E)-ald-B Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 27 1 0 852 0 0 1388 1 0 45 0 0 860 4 0 1388 1 0 51 1 -1 881 19 0 1417 3 0 69 1 -1 902 3 0 1429 1 -6 73 2 -1 920 1 0 1453 4 0 111 0 -1 925 2 0 1455 35 0 120 1 -1 932 4 0 1464 1 0 142 2 0 939 1 0 1484 9 0 164 2 -1 966 23 0 1493 1 0 234 1 0 982 11 -2 1498 10 -1 240 3 0 993 2 0 1504 3 0 249 0 0 1020 10 0 1505 5 0 276 4 0 1022 7 0 1517 8 0 314 2 -2 1028 17 0 1721 466 -33 333 1 0 1031 10 0 2902 43 0 350 1 0 1053 1 0 2965 34 0 376 0 0 1074 0 0 3002 16 0 403 15 0 1079 1 0 3014 16 0 443 5 0 1093 9 0 3020 37 0 455 10 -5 1117 8 0 3024 15 0 479 11 0 1137 8 0 3040 12 0 535 9 -2 1155 16 0 3061 21 0 540 9 -5 1196 34 0 3077 28 0 602 1 0 1229 2 0 3079 71 0 606 1 0 1264 2 0 3082 43 0 688 5 -1 1272 8 0 3147 10 0 739 3 0 1288 0 0 3161 14 0 762 0 -2 1313 2 0 3216 1 0 805 0 0 1331 1 0 3223 1 0 825 37 0 1344 4 0 3233 3 0 841 25 0 1368 4 -2 3236 3 0 852 0 0 1386 5 0 3247 1 0

145 Table S6.1k) (Z)-ald-B Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 24 0 0 852 0 0 1392 2 0 37 0 0 856 9 0 1415 5 -2 54 1 -1 867 8 -3 1420 7 -4 67 0 0 880 36 0 1425 35 0 78 3 -2 902 9 0 1453 3 0 109 3 -2 910 15 0 1461 36 -1 115 2 -1 918 1 0 1464 6 0 139 1 0 920 2 0 1483 29 0 147 3 0 936 2 0 1492 3 0 194 4 -1 953 18 0 1500 36 0 219 4 -3 991 3 0 1504 7 0 236 0 0 1014 6 -1 1504 7 0 246 1 0 1022 7 0 1516 10 0 275 2 0 1028 29 0 1740 264 -33 314 1 -1 1030 4 0 2856 152 0 343 1 0 1051 1 0 2959 32 0 374 1 0 1074 0 0 2998 30 0 382 5 0 1078 1 0 3019 23 0 429 6 -1 1087 14 0 3019 23 0 459 7 -1 1116 6 0 3034 6 0 472 19 0 1137 10 0 3043 26 0 535 8 0 1147 23 0 3073 0 0 601 1 0 1182 10 -1 3079 28 0 606 1 0 1227 2 0 3079 41 0 635 15 -3 1268 1 0 3082 78 0 711 1 0 1275 6 0 3113 15 0 734 3 0 1288 0 0 3148 20 0 805 1 0 1314 2 0 3216 0 0 815 33 0 1333 1 0 3221 1 0 830 19 -1 1349 6 0 3232 4 0 837 15 -1 1386 5 0 3235 4 0 850 1 0 1388 3 0 3246 1 0

146 Table S6.1l) (E)-ald-C Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 39 1 0 853 1 0 1388 1 0 45 0 0 854 1 0 1393 2 0 52 0 0 869 7 0 1404 1 -5 68 3 -2 899 1 0 1417 1 0 75 1 -1 914 7 0 1428 27 -1 122 0 -1 923 1 0 1453 2 0 133 1 -1 926 21 0 1464 1 0 140 0 -1 930 1 0 1489 1 0 193 2 -1 939 2 0 1493 1 -1 207 4 0 947 7 0 1495 14 -1 241 0 0 986 9 -2 1505 5 0 251 0 -1 1003 2 0 1505 3 0 301 6 0 1022 6 0 1516 9 0 313 3 -1 1032 11 0 1723 471 -34 337 0 -2 1055 1 0 2895 38 0 361 2 0 1058 1 0 3001 4 0 377 1 0 1075 0 0 3008 19 0 404 17 0 1080 2 0 3019 13 0 415 4 -1 1086 22 0 3019 27 0 450 20 0 1114 15 0 3023 40 0 468 11 -1 1129 24 0 3040 6 0 534 11 -7 1138 6 0 3064 18 0 599 2 -1 1187 14 0 3078 37 0 603 1 0 1225 18 0 3080 59 0 607 0 0 1236 7 0 3082 41 0 641 4 0 1284 0 0 3143 6 0 736 4 0 1289 0 0 3162 11 0 742 9 0 1307 1 0 3216 0 0 799 0 0 1334 1 0 3224 1 0 822 4 0 1346 3 0 3234 2 0 830 45 0 1362 2 0 3236 3 0 846 20 0 1386 4 0 3248 1 0

147 Table S6.1m) (Z)-ald-C Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 42 0 0 855 12 -1 1392 5 0 49 1 -1 855 0 0 1396 2 0 52 0 0 864 1 -3 1417 30 -5 75 3 -2 876 6 0 1420 1 0 104 0 0 894 37 0 1440 34 -3 119 2 -2 912 10 0 1455 2 0 131 0 0 932 0 0 1467 1 0 140 1 0 935 3 0 1490 1 0 186 4 -2 935 8 0 1494 3 0 194 9 -1 943 3 0 1498 57 0 205 2 -1 977 8 -1 1505 8 0 240 1 -1 1003 3 0 1506 2 0 261 0 0 1025 6 0 1518 10 0 302 3 -1 1035 8 0 1747 282 -34 319 3 -1 1039 5 -1 2834 150 0 353 0 -1 1054 1 0 2982 11 0 373 1 0 1080 0 0 2991 2 0 382 6 0 1083 5 0 2999 15 0 420 1 0 1085 25 0 3000 19 0 440 28 0 1113 18 0 3002 58 0 471 9 0 1124 12 0 3020 2 0 558 12 -2 1138 7 0 3042 20 0 599 1 0 1189 10 0 3058 11 0 607 0 0 1232 2 0 3059 87 0 643 5 0 1239 7 0 3061 42 0 733 31 -3 1285 1 0 3081 15 0 745 3 0 1295 0 0 3145 11 0 754 7 0 1308 1 0 3193 0 0 804 0 0 1336 1 0 3198 1 0 829 6 0 1359 3 0 3210 4 0 836 34 0 1377 6 0 3213 4 0 852 16 0 1389 2 0 3224 1 0

148 Table S6.1n) (E)-ald-D Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 33 0 0 852 1 0 1389 0 0 39 1 0 856 0 0 1392 2 0 43 0 0 869 16 0 1415 10 -1 69 1 -1 887 14 0 1417 1 0 74 2 -2 899 3 0 1435 0 -5 118 0 -1 918 2 0 1452 2 0 126 1 -1 920 1 0 1464 1 0 138 1 -1 933 2 0 1489 0 0 169 3 0 938 3 0 1491 15 -1 212 2 0 966 28 0 1493 1 0 241 0 0 982 6 -2 1505 4 0 248 0 0 1004 5 0 1505 4 0 309 3 -1 1021 6 0 1517 8 0 332 2 -2 1030 3 0 1725 490 -34 351 3 -1 1041 25 0 2895 42 0 367 0 0 1056 1 0 3001 4 0 371 1 0 1072 10 0 3005 4 0 412 14 0 1075 2 0 3014 41 0 417 16 -1 1080 1 0 3019 14 0 453 9 -2 1091 4 0 3020 40 0 476 16 -2 1120 11 0 3040 11 0 503 9 -5 1137 7 0 3063 23 0 598 1 0 1186 42 0 3078 63 0 607 0 0 1224 2 0 3082 42 0 617 2 0 1254 3 0 3137 13 0 693 4 -2 1279 2 0 3149 11 0 735 2 0 1290 0 0 3156 7 0 736 4 -1 1308 1 0 3215 1 0 797 0 0 1332 1 0 3224 1 0 826 37 0 1354 1 0 3234 2 0 835 5 0 1364 10 -1 3237 3 0 840 23 0 1385 4 0 3249 1 0

149 Table S6.1o) (Z)-ald-D Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 35 0 0 852 2 0 1394 3 0 39 0 0 855 1 0 1407 49 -1 41 1 0 865 14 0 1416 1 0 74 2 -1 874 24 0 1425 1 -6 77 1 -1 892 39 -1 1439 14 0 112 2 -2 904 0 0 1453 2 0 121 0 0 920 0 0 1465 1 0 132 3 -1 928 1 0 1489 1 0 169 3 -1 933 9 -1 1491 67 -1 187 2 -2 938 5 0 1494 1 0 217 3 0 961 29 0 1503 7 0 236 0 0 1004 4 0 1505 1 0 254 2 0 1022 8 0 1517 8 0 299 4 -4 1027 0 -1 1747 266 -34 327 0 0 1032 9 0 2859 157 0 364 1 0 1053 5 0 3000 14 0 366 1 0 1056 5 0 3006 7 0 391 1 -1 1076 0 0 3016 23 0 406 13 0 1080 1 0 3019 26 0 458 12 0 1093 4 0 3021 36 0 482 18 0 1119 11 0 3043 6 0 565 20 -2 1138 8 0 3064 22 0 599 1 0 1173 15 -1 3078 65 0 607 0 0 1224 2 0 3081 43 0 626 1 0 1254 7 0 3111 16 0 722 3 0 1280 2 0 3135 17 0 735 3 0 1290 0 0 3154 8 0 797 0 -3 1310 1 0 3216 0 0 800 9 -2 1335 1 0 3222 1 0 821 39 0 1355 1 0 3234 4 0 831 7 0 1385 4 0 3236 4 0 837 20 0 1389 4 0 3248 1 0

150 Table S6.1p) (E)-ald-E Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 20 2 0 857 1 0 1394 1 0 47 1 0 860 2 0 1411 2 -1 50 2 -1 891 16 0 1416 2 -3 73 0 0 897 14 0 1418 4 -2 84 2 -2 907 2 0 1436 8 -1 122 0 0 924 1 0 1439 1 0 131 1 -2 933 0 0 1483 9 -1 151 0 -1 938 2 0 1491 1 0 189 3 0 949 3 0 1495 1 0 211 4 0 967 12 0 1504 7 0 236 0 0 979 8 -2 1505 4 0 255 0 0 996 20 0 1510 3 0 309 5 -1 1005 2 0 1518 9 0 312 4 0 1045 5 0 1717 409 -33 336 3 -2 1055 0 0 2900 39 0 347 2 -1 1062 1 0 2999 1 0 376 1 0 1069 9 0 3007 29 0 401 9 0 1077 1 0 3016 42 0 429 26 -1 1081 18 0 3018 6 0 448 5 0 1091 0 0 3029 22 0 479 4 -3 1121 7 0 3039 14 0 508 12 -1 1188 31 0 3067 4 0 529 8 -6 1213 24 0 3075 75 0 601 1 0 1224 0 0 3079 42 0 640 2 0 1253 1 0 3090 30 0 666 4 -1 1271 0 0 3148 11 0 722 1 0 1286 1 0 3161 16 0 734 3 0 1312 0 0 3163 5 0 780 9 -1 1331 1 0 3213 1 0 792 0 0 1359 9 -1 3220 3 0 824 13 0 1363 1 0 3228 4 0 835 31 0 1391 3 0 3238 2 0

151 Table S6.1q) (Z)-ald-E Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 24 0 0 855 2 -2 1402 43 0 31 0 0 859 3 -1 1414 4 -1 42 2 -1 871 25 0 1417 1 -2 77 0 0 881 40 -1 1420 6 -3 84 2 -2 895 10 0 1435 1 0 111 3 -2 907 1 0 1446 6 -1 124 0 0 919 1 0 1485 74 -1 147 2 0 933 0 0 1488 1 0 185 3 -2 934 12 0 1493 0 0 190 5 0 948 3 0 1504 3 0 209 5 0 954 19 0 1504 7 0 238 0 0 1004 1 -1 1512 3 0 254 0 0 1005 3 0 1517 10 0 290 2 -3 1042 6 0 1750 297 -34 318 1 0 1045 4 -1 2853 133 0 327 7 0 1056 0 0 3000 1 0 377 0 0 1069 7 0 3007 27 0 390 5 -1 1079 8 0 3018 35 0 408 10 0 1081 17 0 3018 11 0 446 4 0 1089 1 0 3027 28 0 482 13 -1 1120 7 0 3039 10 0 502 16 0 1172 14 -1 3066 15 0 596 2 -1 1213 13 0 3076 73 0 606 12 -2 1223 0 0 3080 42 0 642 0 0 1251 0 0 3090 27 0 714 3 -1 1272 0 0 3113 13 0 725 8 -1 1283 1 0 3149 20 0 733 2 0 1312 0 0 3165 9 0 793 0 0 1333 1 0 3212 0 0 810 28 -1 1359 0 0 3217 4 0 815 9 0 1392 2 0 3226 5 0 842 6 -2 1393 4 0 3238 2 0

152 Table S6.1r) (E)-ald-F Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 19 2 0 842 21 0 1394 3 0 30 1 0 858 0 0 1404 4 0 56 1 -1 880 12 0 1415 1 -3 72 0 0 895 18 0 1417 6 -2 84 3 -2 906 1 0 1436 9 -1 124 0 0 918 0 0 1445 1 0 132 1 -2 934 0 0 1484 8 -1 156 1 -1 940 0 0 1490 1 0 173 3 -1 944 3 0 1494 1 0 214 5 0 965 17 0 1503 13 0 240 0 0 981 7 -2 1505 8 0 256 1 0 996 21 0 1508 1 0 304 4 0 1006 1 0 1517 9 0 316 1 -1 1048 10 0 1717 409 -33 333 2 -1 1055 0 0 2900 44 0 348 5 -1 1060 3 0 3001 1 0 384 2 0 1062 2 0 3007 26 0 407 14 0 1074 0 0 3017 44 0 437 8 0 1081 21 0 3019 2 0 440 19 -2 1093 0 0 3025 25 0 478 3 -2 1120 6 0 3041 15 0 505 10 -2 1189 34 0 3068 4 0 531 8 -5 1212 27 0 3075 73 0 603 1 0 1225 1 0 3079 42 0 643 0 0 1257 2 0 3090 28 0 665 3 -1 1271 0 0 3150 12 0 716 1 0 1287 2 0 3162 15 0 734 3 0 1315 0 0 3165 5 0 779 7 -1 1333 1 0 3212 1 0 796 0 0 1356 8 -1 3222 1 0 825 26 0 1364 1 0 3227 3 0 831 2 0 1392 1 0 3240 3 0

153 Table S6.1s) (Z)-ald-F Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 17 1 0 849 3 -2 1400 26 0 38 0 0 859 1 -1 1406 13 0 52 0 0 866 33 0 1416 2 -1 78 3 -2 880 47 -1 1421 6 -5 80 0 0 884 6 0 1444 0 0 114 2 -2 907 1 0 1446 7 -1 128 0 0 919 1 0 1483 73 -1 154 1 -1 934 6 0 1491 3 0 173 7 -1 935 3 0 1494 0 0 193 1 -2 945 4 0 1504 7 0 217 8 0 954 20 0 1504 8 0 236 0 0 1004 2 -1 1509 4 0 255 0 0 1006 3 0 1518 9 0 291 2 -4 1044 2 -1 1749 321 -34 312 1 0 1048 12 0 2856 130 0 325 3 0 1056 0 0 3001 7 0 379 3 0 1062 5 0 3009 12 0 392 4 -1 1075 0 0 3018 35 0 428 12 -1 1080 22 0 3019 3 0 443 5 0 1089 0 0 3022 46 0 479 8 0 1120 7 0 3039 5 0 497 15 0 1169 13 -1 3063 23 0 598 6 -1 1211 15 0 3077 69 0 609 8 -1 1225 0 0 3081 40 0 645 0 0 1256 2 0 3090 25 0 712 5 0 1271 0 0 3116 14 0 722 6 -1 1284 2 0 3149 18 0 736 2 0 1314 0 0 3168 9 0 797 0 0 1333 1 0 3212 1 0 808 16 -1 1359 0 0 3221 1 0 829 6 0 1392 10 0 3226 4 0 839 19 -1 1394 4 0 3241 3 0

154 Table S6.1t) (E)-ald-G Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 13 2 -1 857 2 0 1394 2 0 39 0 0 862 17 0 1415 1 -2 56 0 0 889 1 0 1418 4 -3 73 3 -2 898 17 0 1419 3 0 76 0 0 908 2 0 1436 10 -1 123 0 0 930 1 0 1439 1 0 133 0 -1 934 1 0 1483 11 -1 147 2 -1 938 4 0 1489 1 0 180 3 -1 948 2 0 1493 0 0 221 3 0 968 18 0 1504 5 0 240 0 0 982 9 -2 1505 7 0 255 1 0 995 22 0 1509 3 0 299 7 0 1006 1 0 1517 8 0 312 2 -1 1049 8 0 1724 478 -34 329 2 -2 1055 0 0 2898 41 0 353 3 -1 1061 1 0 3000 5 0 384 0 0 1068 6 0 3007 16 0 413 11 0 1080 2 0 3017 6 0 416 23 0 1081 24 0 3018 33 0 455 7 -2 1086 1 0 3021 45 0 466 5 -2 1121 6 0 3037 6 0 507 12 -2 1189 36 0 3061 27 0 531 8 -5 1213 22 0 3076 65 0 601 1 0 1223 0 0 3081 41 0 644 1 0 1253 0 0 3090 27 0 665 3 -1 1272 0 0 3148 11 0 716 3 0 1284 2 0 3159 14 0 734 3 0 1313 0 0 3163 7 0 778 7 -1 1333 1 0 3209 2 0 795 0 0 1358 8 -1 3221 1 0 824 7 0 1359 1 0 3229 2 0 828 28 0 1388 3 0 3241 1 0

155 Table S6.1w) (Z)-ald-G Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 21 1 0 852 15 -1 1403 55 0 42 0 0 856 4 0 1416 1 -3 54 1 0 868 8 -1 1419 4 -2 75 2 -2 881 56 -1 1420 3 -1 79 0 0 887 6 0 1438 1 0 108 2 -2 907 1 0 1447 7 -1 130 0 0 922 2 0 1485 83 -1 145 1 0 933 8 0 1490 0 0 179 8 -2 935 6 0 1493 0 0 189 1 -1 946 4 0 1504 4 0 222 8 0 956 25 0 1505 8 0 239 0 0 1003 3 -1 1510 4 0 255 1 0 1007 1 0 1517 8 0 290 2 -2 1043 1 -1 1749 306 -34 313 2 -1 1046 12 0 2858 143 0 318 4 0 1057 0 0 3000 4 0 381 1 0 1067 6 0 3007 19 0 395 6 -1 1081 27 0 3017 2 0 416 10 0 1082 2 0 3018 37 0 454 3 0 1087 1 0 3022 41 0 466 14 -1 1121 6 0 3037 7 0 497 16 0 1174 14 -1 3061 25 0 599 1 0 1213 11 0 3076 67 0 606 13 -2 1224 0 0 3080 42 0 646 1 0 1252 0 0 3090 24 0 714 5 0 1271 0 0 3115 13 0 722 7 -1 1285 1 0 3149 19 0 735 2 0 1313 0 0 3163 9 0 796 0 0 1333 1 0 3209 1 0 806 10 -1 1360 0 0 3218 2 0 820 26 0 1389 2 0 3227 3 0 845 1 -2 1394 2 0 3239 3 0

156 Table S6.1x) (E)-ald-H Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 20 2 -1 855 1 0 1393 2 0 37 0 0 863 6 0 1411 1 -1 54 0 0 887 26 0 1416 3 -2 70 3 -2 897 7 0 1418 5 -2 77 0 0 908 1 0 1435 6 -1 125 0 0 925 1 0 1435 4 0 130 1 0 933 0 0 1483 12 -1 136 1 -2 940 1 0 1488 1 0 190 3 -1 950 6 0 1493 0 0 228 7 0 964 14 0 1504 3 0 240 0 0 981 8 -2 1505 6 0 256 1 0 994 25 0 1515 2 0 295 3 0 1006 2 0 1518 8 0 309 1 0 1043 7 0 1724 464 -34 330 2 -2 1056 0 0 2899 40 0 348 4 0 1060 1 0 2999 7 0 383 1 0 1071 5 0 3006 6 0 407 21 0 1080 5 0 3016 29 0 435 12 0 1081 23 0 3017 5 0 456 8 -3 1090 0 0 3019 61 0 464 6 0 1121 6 0 3035 2 0 505 12 -2 1188 31 0 3058 32 0 531 8 -5 1212 30 0 3077 64 0 607 0 0 1226 0 0 3081 41 0 639 2 0 1250 0 0 3088 28 0 664 3 -1 1273 0 0 3143 10 0 716 0 0 1284 1 0 3158 15 0 734 3 0 1316 0 0 3163 8 0 779 7 -1 1333 1 0 3210 2 0 796 0 0 1357 8 -1 3220 1 0 823 13 0 1361 1 0 3230 2 0 844 23 0 1392 0 0 3244 2 0

157 Table S6.1y) (Z)-ald-H Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 15 1 0 854 5 0 1402 42 0 34 0 0 858 29 -1 1412 6 0 47 1 0 866 15 0 1415 1 -2 70 0 0 881 48 -1 1419 7 -4 80 3 -2 896 5 0 1435 1 0 109 2 -2 908 2 0 1445 6 -1 119 1 0 924 1 0 1484 85 -1 128 1 0 931 2 0 1487 1 0 180 6 -2 933 11 0 1492 0 0 196 8 0 950 6 0 1503 6 0 234 0 0 953 16 0 1503 3 0 236 6 0 1003 3 -1 1515 3 0 253 0 0 1005 1 0 1518 7 0 289 2 -3 1042 3 0 1750 320 -34 301 2 0 1043 9 0 2858 136 0 317 1 0 1055 0 0 2999 6 0 382 2 0 1069 5 0 3005 4 0 387 3 -1 1079 22 0 3015 39 0 428 10 0 1081 7 0 3017 4 0 458 3 0 1087 1 0 3019 56 0 464 15 0 1120 6 0 3033 2 0 501 16 0 1171 11 -1 3056 34 0 602 11 -2 1211 17 0 3076 64 0 610 3 -1 1225 0 0 3081 41 0 638 3 0 1250 0 0 3090 24 0 704 3 -1 1272 0 0 3108 15 0 729 8 -1 1283 2 0 3146 19 0 733 3 0 1312 0 0 3169 8 0 790 0 0 1331 1 0 3210 2 0 808 8 -1 1357 0 0 3216 2 0 836 10 -1 1390 2 0 3229 3 0 849 10 -2 1392 2 0 3243 2 0

158 Table S6.1z) (E)-ald-I Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 24 2 -1 847 42 0 1393 2 0 39 0 0 858 2 0 1413 3 0 44 2 -1 885 6 0 1416 3 -1 77 2 -2 893 18 0 1420 7 -4 79 1 0 907 1 0 1436 8 -1 123 0 0 929 1 0 1444 2 0 125 1 -1 934 1 0 1484 13 -1 149 1 -1 939 2 0 1490 2 0 177 4 0 946 3 0 1494 0 0 234 6 0 968 13 0 1502 9 0 237 0 0 984 8 -2 1504 8 0 255 0 0 994 24 0 1507 2 0 293 4 0 1004 3 0 1517 9 0 310 2 0 1049 8 0 1724 504 -34 330 2 -2 1055 0 0 2900 41 0 356 6 -1 1060 1 0 3000 8 0 380 1 0 1065 4 0 3008 10 0 397 19 0 1081 15 0 3018 19 0 430 8 0 1082 13 0 3018 15 0 456 11 -4 1087 1 0 3021 54 0 473 7 0 1120 9 0 3038 3 0 508 9 -2 1189 40 0 3061 27 0 530 12 -5 1213 24 0 3077 67 0 604 0 0 1224 0 0 3081 41 0 639 0 0 1254 2 0 3087 27 0 665 3 -1 1271 0 0 3145 10 0 719 1 0 1284 2 0 3160 15 0 735 3 0 1314 0 0 3163 6 0 778 6 -1 1334 1 0 3216 1 0 796 0 0 1359 1 0 3220 2 0 824 2 0 1361 10 -1 3232 2 0 837 11 0 1389 4 0 3244 1 0

159 Table S6.1aa) (Z)-ald-I Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 23 1 -1 850 7 -3 1403 48 0 34 0 0 858 5 0 1412 8 -1 43 1 -1 867 12 -1 1417 4 -4 76 0 -1 882 64 0 1418 5 -2 81 3 -2 886 2 0 1444 1 0 109 2 0 907 1 0 1445 6 -1 121 0 0 922 1 0 1485 92 -1 137 2 -1 933 5 0 1490 1 0 177 3 -1 934 1 0 1493 0 0 186 4 -2 944 8 0 1503 5 0 240 1 0 956 25 0 1505 8 0 241 13 0 1002 3 -1 1508 1 0 255 0 0 1004 2 0 1517 9 0 299 3 -2 1043 3 -1 1750 324 -34 301 2 -2 1049 9 0 2857 148 0 311 1 0 1055 0 0 3000 7 0 380 1 0 1065 6 0 3008 15 0 386 4 0 1079 8 0 3018 3 0 429 11 0 1081 21 0 3018 34 0 458 8 -1 1087 0 0 3022 47 0 469 11 0 1120 8 0 3038 5 0 499 17 0 1171 14 -1 3061 26 0 601 8 -1 1213 12 0 3077 67 0 605 4 -1 1224 0 0 3081 41 0 642 1 0 1255 2 0 3087 24 0 714 7 -1 1271 0 0 3115 14 0 724 4 -1 1284 3 0 3147 18 0 734 2 0 1313 0 0 3167 8 0 795 0 0 1333 1 0 3214 1 0 809 6 -1 1358 0 0 3220 2 0 823 6 -1 1388 6 0 3229 4 0 841 34 0 1393 3 0 3242 2 0

160 Table S6.1bb) rald-A Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 18 0 0 846 3 0 1395 4 -1 27 0 0 851 0 0 1417 32 -10 44 0 0 857 7 0 1420 1 -1 76 0 0 902 1 0 1440 97 -4 84 0 0 916 5 -1 1451 2 0 120 0 0 919 1 0 1471 2 0 123 0 0 920 4 0 1481 95 -2 138 0 -1 923 1 0 1487 1 0 147 1 0 963 3 0 1494 1 0 182 1 -1 974 1 0 1504 3 0 247 0 0 1012 0 0 1506 8 0 252 0 -1 1020 14 0 1517 4 0 302 0 -1 1027 0 0 1550 16 0 307 0 -3 1035 10 0 1610 16 -3 327 6 -1 1057 1 0 2993 6 0 353 12 -1 1077 0 0 2998 3 0 360 6 -6 1078 1 0 3009 67 0 387 6 -3 1101 8 0 3010 3 0 408 1 -2 1129 2 -1 3018 41 0 456 1 -2 1134 0 0 3022 0 0 478 1 -7 1141 10 0 3031 29 0 557 7 0 1191 8 -3 3047 39 0 582 2 -6 1251 0 0 3068 56 0 593 2 0 1269 3 -1 3076 63 0 601 0 -1 1290 0 0 3081 47 0 685 1 -8 1292 1 -1 3089 40 0 722 11 0 1308 126 -8 3104 41 0 746 11 0 1325 0 0 3215 0 0 757 1 -1 1338 0 0 3217 0 0 794 0 0 1364 6 0 3231 3 0 812 40 0 1385 0 0 3237 3 0 840 18 0 1389 2 -1 3246 1 0

161 Table S6.1cc) rald-B Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 15 0 0 851 1 0 1391 3 0 36 0 0 856 7 0 1411 12 -9 56 0 0 872 9 -3 1414 7 0 65 0 0 891 0 0 1452 2 0 84 0 0 908 1 0 1465 40 0 101 0 0 919 1 0 1471 51 0 134 0 0 921 4 0 1474 65 0 161 2 -2 942 6 -2 1491 7 -3 172 1 -1 956 2 -2 1493 18 -1 196 1 -2 985 1 0 1502 4 0 219 0 0 987 1 -1 1504 8 0 243 0 0 1023 13 0 1515 3 0 284 1 -2 1024 2 0 1589 19 0 305 2 -2 1035 11 0 1608 66 -6 311 0 -3 1048 0 0 2984 6 0 331 8 -3 1070 1 0 2997 5 0 355 11 -1 1079 0 0 3007 36 0 365 7 -3 1081 0 0 3016 27 0 422 2 -2 1098 17 -1 3017 24 0 428 3 -2 1113 4 -1 3033 71 0 489 2 -5 1143 10 0 3040 16 0 554 4 -1 1151 6 -3 3060 20 0 581 2 -3 1243 2 0 3074 67 0 593 2 0 1261 11 -4 3078 44 0 597 0 0 1287 18 -1 3097 10 0 660 1 -3 1291 0 0 3123 34 0 725 4 0 1317 0 0 3159 28 0 755 24 0 1330 40 -1 3215 0 0 794 9 0 1339 86 -2 3219 1 0 819 47 0 1364 16 -2 3230 3 0 837 13 0 1385 19 -2 3242 2 0 845 1 0 1389 4 0 3258 2 0

162 Table S6.1dd) rald-C Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 21 0 0 844 8 -1 1389 6 0 30 0 0 846 4 0 1413 38 -8 42 0 0 852 1 0 1417 3 0 64 0 0 856 1 0 1435 3 -3 72 0 0 901 0 -5 1452 1 0 114 0 0 911 1 0 1467 8 0 127 0 0 917 0 0 1470 67 -1 143 1 0 921 2 0 1491 2 0 170 1 -4 939 0 0 1496 1 0 187 0 0 989 1 0 1505 7 0 241 0 0 997 4 0 1508 3 0 246 0 -2 1003 4 0 1518 7 0 275 0 -2 1022 13 0 1573 49 -7 325 2 0 1034 12 0 1592 156 -1 343 12 -2 1052 1 0 3001 12 0 364 10 -3 1077 0 0 3010 13 0 371 9 -2 1078 1 0 3018 37 0 402 0 -1 1086 1 -2 3020 4 0 414 2 -4 1115 25 0 3025 30 0 420 5 -1 1141 11 0 3027 65 0 462 1 -8 1151 4 -1 3030 57 0 593 3 0 1181 5 -5 3042 9 0 596 0 0 1237 3 -1 3063 18 0 602 2 -3 1257 7 -2 3070 42 0 652 5 -1 1282 1 0 3077 84 0 691 4 -2 1290 0 0 3081 47 0 735 3 0 1307 12 0 3140 24 0 763 7 -3 1329 79 -3 3215 0 0 798 1 0 1331 22 0 3216 1 0 803 5 0 1352 6 0 3231 3 0 812 40 0 1383 0 0 3238 3 0 837 28 0 1389 1 0 3246 1 0

163 Table S6.1ee) rald-D Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 25 0 0 846 0 0 1390 3 0 28 0 0 853 2 0 1390 2 0 51 0 -1 855 0 0 1415 2 0 61 0 0 868 0 0 1439 63 -7 74 0 0 910 1 0 1453 1 0 115 0 0 919 2 -1 1468 0 0 139 0 -2 920 1 0 1487 0 0 141 1 0 921 5 0 1493 1 -1 145 1 0 952 7 0 1500 2 -3 210 0 -3 981 0 -1 1504 8 0 223 1 -1 996 1 0 1507 1 0 239 0 0 1005 1 0 1516 8 0 249 5 -3 1022 13 0 1569 36 0 323 3 0 1034 9 -2 1599 53 -4 347 3 -2 1035 5 -1 2996 1 0 365 3 0 1057 1 0 3002 13 0 371 6 -5 1077 0 0 3010 44 0 380 6 -2 1078 1 0 3015 7 0 394 0 -1 1113 8 -2 3017 52 0 477 1 -6 1119 10 0 3019 54 0 500 1 -3 1141 10 0 3034 11 0 569 7 -3 1159 7 -2 3039 59 0 593 2 0 1233 5 0 3056 30 0 598 0 0 1278 3 -2 3075 61 0 611 7 -2 1281 2 -3 3078 48 0 715 1 -2 1290 0 0 3106 20 0 735 1 0 1309 10 0 3151 27 0 792 4 -1 1324 89 -2 3216 0 0 800 15 -1 1331 10 0 3217 0 0 812 33 0 1354 4 0 3231 3 0 832 18 -3 1372 63 -6 3238 3 0 841 19 0 1384 0 0 3246 1 0

164 Table S6.1ff) rald-E Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 19 0 0 839 31 0 1403 14 -9 27 0 0 855 2 0 1410 19 -1 33 0 0 879 3 0 1417 3 0 71 0 0 902 0 -5 1450 1 0 82 0 0 906 1 0 1466 84 -2 120 0 0 912 6 -1 1487 3 -3 127 1 0 918 1 0 1490 2 0 148 1 0 933 0 0 1494 1 0 172 0 -4 943 3 0 1503 10 0 193 1 -1 976 0 0 1505 7 0 236 1 -2 992 0 0 1508 2 0 243 0 0 1003 4 0 1518 9 0 258 0 0 1031 2 0 1581 31 0 307 2 0 1053 8 0 1598 67 -5 325 1 -3 1054 1 0 3000 5 0 358 7 -2 1064 7 0 3008 18 0 368 10 -2 1077 1 0 3018 4 0 380 9 -5 1088 0 0 3018 42 0 401 0 -1 1099 6 -5 3023 38 0 458 1 -6 1119 12 0 3032 36 0 467 2 0 1224 0 0 3035 57 0 560 7 0 1235 2 -5 3038 6 0 586 2 -5 1259 2 0 3063 21 0 595 1 0 1272 0 0 3074 50 0 635 3 0 1283 2 0 3076 72 0 699 40 -1 1290 1 -2 3080 42 0 706 1 -4 1313 1 0 3122 27 0 720 1 0 1323 139 -3 3155 29 0 734 2 0 1330 1 0 3212 0 0 795 1 0 1358 0 0 3220 3 0 819 21 0 1389 0 -1 3228 3 0 834 4 -1 1390 2 0 3239 2 0

165 Table S6.1gg) keto Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 13 1 0 850 0 0 1389 8 0 28 1 0 853 0 -1 1391 1 0 33 1 0 856 8 -1 1406 89 0 64 1 0 867 9 -1 1417 5 0 86 1 -1 890 16 -1 1446 7 0 120 0 0 911 5 0 1453 2 0 126 1 0 921 0 0 1463 1 0 140 1 -2 924 1 0 1477 88 0 175 4 0 936 0 0 1485 14 0 190 9 -2 945 9 -2 1493 1 0 238 0 0 960 23 0 1504 0 0 261 2 -3 1003 1 0 1504 7 0 287 0 -2 1022 4 0 1517 8 0 300 3 -2 1030 6 0 1733 189 -33 360 1 -1 1035 22 0 2998 2 0 361 1 0 1038 58 -1 3004 32 0 383 8 0 1054 2 0 3018 37 0 410 2 -3 1076 3 0 3024 12 0 426 7 -1 1077 33 0 3029 28 0 476 16 -1 1080 17 0 3052 4 0 500 16 -1 1119 4 0 3073 2 0 592 17 -3 1138 8 0 3078 80 0 600 3 0 1158 21 0 3080 47 0 607 1 0 1186 69 -2 3091 24 0 678 2 -1 1213 14 0 3115 13 0 716 4 -2 1241 12 -1 3147 20 0 737 3 0 1275 7 0 3167 9 0 777 3 -2 1290 0 0 3217 0 0 799 4 -1 1312 2 0 3222 1 0 817 5 -1 1333 1 0 3234 5 0 822 37 0 1349 12 0 3236 4 0 841 22 0 1385 4 0 3248 1 0

166 Table S6.1hh) rketo Freq. IR Int. Shift Freq. IR Int. Shift Freq. IR Int. Shift 23 0 0 844 1 0 1384 0 0 30 0 0 850 3 0 1389 11 0 31 0 0 852 1 0 1417 5 0 56 0 -1 856 2 0 1438 134 -3 62 0 0 889 1 0 1452 1 0 117 0 0 914 0 -4 1467 0 0 134 1 0 915 0 0 1486 5 0 145 1 0 920 3 0 1493 12 -4 176 0 -4 957 0 -4 1494 36 -1 199 0 0 974 9 0 1503 2 0 220 0 -3 984 0 0 1508 8 0 243 4 -1 1022 10 -5 1512 10 0 276 0 -1 1022 10 0 1577 51 -3 293 3 -3 1032 1 0 1592 146 -3 330 1 -2 1034 11 0 3004 29 0 344 3 0 1071 0 0 3014 10 0 380 2 0 1077 0 0 3020 22 0 399 14 -1 1078 1 0 3031 36 0 427 0 0 1101 19 -1 3034 11 0 486 1 -5 1118 0 -1 3036 40 0 507 0 -2 1141 10 0 3049 8 0 592 3 0 1193 39 -1 3073 57 0 598 0 -1 1229 4 -1 3078 35 0 607 4 -8 1254 2 -6 3080 46 0 656 2 -4 1282 6 0 3085 42 0 693 37 -1 1290 0 0 3118 32 0 725 6 -2 1299 12 -2 3159 23 0 752 3 0 1317 36 -1 3214 0 0 791 2 0 1330 43 0 3216 1 0 809 36 0 1344 126 0 3231 3 0 824 12 -3 1376 1 0 3237 3 0 836 21 0 1381 14 0 3246 1 0

167 Appendix D Supporting Information for: “Charge Transfer Complexes and Photochemistry of Ozone with Ferrocene and n- Butylferrocene: A UV-Vis matrix isolation study”

168

169 a Table S7.1a-j Natural Population Analysis Summaries for O3, O2, Ferrocene, n- Butylferrocene, Charge Transfer and van der Waals Complexes a Calculated at the B3LYP/6-311G++(d,2p) level of theory.

Table S7.1a) Summary of Natural Population Analysis for O3 Natural Natural Population Atom No Charge Core Valence Rydberg Total O 1 -0.1357 1.99989 6.10997 0.02580 8.13566 O 2 0.27132 1.99984 5.68142 0.04742 7.72868 O 3 -0.1357 1.99989 6.10997 0.02580 8.13566 * Total * 0.00000 5.99963 17.90136 0.09901 24.00000

170 Table S7.1b) Summary of Natural Population Analysis of O2 Natural Natural Population Atom No Charge Core Valence Rydberg Total O 1 0.00000 1.99990 5.96592 0.03418 8.00000 O 2 0.00000 1.99990 5.96592 0.03418 8.00000 * Total * 0.00000 3.99981 11.93184 0.06835 16.00000

171 Table S7.1c) Summary of Natural Population Analysis for Fc Natural Natural Population Atom No Charge Core Valence Rydberg Total Fe 1 -0.35231 17.99427 8.29669 0.06135 26.35231 C 2 -0.19160 1.99873 4.16912 0.02375 6.19160 H 3 0.22688 0.00000 0.77046 0.00266 0.77312 C 4 -0.19166 1.99873 4.16918 0.02375 6.19166 C 5 -0.19173 1.99873 4.16924 0.02376 6.19173 H 6 0.22689 0.00000 0.77045 0.00266 0.77311 C 7 -0.19165 1.99873 4.16917 0.02375 6.19165 H 8 0.22687 0.00000 0.77047 0.00266 0.77313 C 9 -0.19158 1.99873 4.16910 0.02375 6.19158 H 10 0.22687 0.00000 0.77047 0.00266 0.77313 H 11 0.22686 0.00000 0.77048 0.00266 0.77314 C 12 -0.19156 1.99873 4.16909 0.02374 6.19156 H 13 0.22687 0.00000 0.77047 0.00266 0.77313 C 14 -0.19166 1.99873 4.16917 0.02375 6.19166 C 15 -0.19173 1.99873 4.16924 0.02376 6.19173 H 16 0.22687 0.00000 0.77047 0.00266 0.77313 C 17 -0.19167 1.99873 4.16919 0.02375 6.19167 H 18 0.22688 0.00000 0.77046 0.00266 0.77312 C 19 -0.19161 1.99873 4.16914 0.02375 6.19161 H 20 0.22689 0.00000 0.77045 0.00266 0.77311 H 21 0.22688 0.00000 0.77046 0.00266 0.77312 * Total * 0.00000 37.98156 57.69297 0.32547 96.00000

172 Table S7.1d) Summary of Natural Population Analysis of nBuFc Natural Natural Population Atom No Charge Core Valence Rydberg Total Fe 1 -0.34311 17.99376 8.28651 0.06284 26.34311 C 2 -0.18912 1.99873 4.16627 0.02412 6.18912 H 3 0.22607 0.00000 0.77114 0.00279 0.77393 C 4 -0.18707 1.99866 4.16559 0.02282 6.18707 C 5 -0.18911 1.99873 4.16626 0.02412 6.18911 H 6 0.22584 0.00000 0.77171 0.00246 0.77416 C 7 -0.01368 1.99862 3.98223 0.03283 6.01368 H 8 0.22606 0.00000 0.77115 0.00279 0.77394 C 9 -0.18707 1.99866 4.16559 0.02282 6.18707 H 10 0.22584 0.00000 0.77170 0.00246 0.77416 C 11 -0.19211 1.99873 4.16957 0.02380 6.19211 H 12 0.22625 0.00000 0.77106 0.00269 0.77375 C 13 -0.19499 1.99873 4.17282 0.02344 6.19499 C 14 -0.19320 1.99873 4.17070 0.02377 6.19320 H 15 0.22640 0.00000 0.77101 0.00259 0.77360 C 16 -0.19211 1.99873 4.16958 0.02380 6.19211 H 17 0.22667 0.00000 0.77074 0.00258 0.77333 C 18 -0.19320 1.99873 4.17070 0.02377 6.19320 H 19 0.22625 0.00000 0.77106 0.00269 0.77375 H 20 0.22667 0.00000 0.77075 0.00258 0.77333 C 21 -0.38256 1.99918 4.36733 0.01604 6.38256 H 22 0.19923 0.00000 0.79763 0.00314 0.80077 H 23 0.19923 0.00000 0.79762 0.00314 0.80077 C 24 -0.36229 1.99925 4.34956 0.01347 6.36229 H 25 0.18897 0.00000 0.80837 0.00266 0.81103 H 26 0.18898 0.00000 0.80836 0.00266 0.81102 C 27 -0.37680 1.99929 4.36537 0.01213 6.37680 H 28 0.18435 0.00000 0.81292 0.00273 0.81565 H 29 0.18435 0.00000 0.81292 0.00273 0.81565 C 30 -0.56843 1.99934 4.56097 0.00812 6.56843 H 31 0.19823 0.00000 0.79999 0.00178 0.80177 H 32 0.19272 0.00000 0.80521 0.00208 0.80728 H 33 0.19272 0.00000 0.80521 0.00208 0.80728 * Total * 0.00000 45.97785 81.61762 0.40452 128.00000

173 Table S7.1e) Summary of Natural Population Analysis for Fc-O3A Natural Natural Population Atom No Charge Core Valence Rydberg Total Fe 1 -0.32023 17.99421 8.26410 0.06193 26.32023 C 2 -0.18889 1.99873 4.16626 0.02390 6.18889 H 3 0.22843 0.00000 0.76891 0.00266 0.77157 C 4 -0.19942 1.99874 4.17203 0.02866 6.19942 C 5 -0.18889 1.99873 4.16619 0.02397 6.18889 C 6 -0.19212 1.99873 4.16845 0.02494 6.19212 H 7 0.22844 0.00000 0.76891 0.00265 0.77156 H 8 0.23096 0.00000 0.76643 0.00261 0.76904 C 9 -0.18938 1.99873 4.16688 0.02376 6.18938 H 10 0.22866 0.00000 0.76870 0.00265 0.77134 C 11 -0.19135 1.99874 4.16882 0.02379 6.19135 C 12 -0.18909 1.99873 4.16659 0.02377 6.18909 H 13 0.22942 0.00000 0.76794 0.00264 0.77058 C 14 -0.18709 1.99874 4.16473 0.02362 6.18709 H 15 0.22865 0.00000 0.76870 0.00265 0.77135 C 16 -0.19156 1.99874 4.16901 0.02381 6.19156 H 17 0.22991 0.00000 0.76746 0.00263 0.77009 H 18 0.22934 0.00000 0.76802 0.00264 0.77066 H 19 0.23093 0.00000 0.76649 0.00258 0.76907 C 20 -0.18884 1.99873 4.16508 0.02503 6.18884 O 21 -0.16177 1.99989 6.13811 0.02377 8.16177 O 22 0.25311 1.99984 5.70104 0.04600 7.74689 O 23 -0.16913 1.99989 6.14576 0.02348 8.16913 H 24 0.23990 0.00000 0.75554 0.00456 0.76010 * Total * 0.00000 43.98118 75.59016 0.42866 120.00000

174 Table S7.1f) Summary of Natural Population Analysis of Fc-O3B Natural Natural Population Atom No Charge Core Valence Rydberg Total Fe 1 -0.30951 17.99430 8.25120 0.06401 26.30951 C 2 -0.18911 1.99874 4.16688 0.02348 6.18911 H 3 0.22958 0.00000 0.76782 0.00259 0.77042 C 4 -0.19097 1.99873 4.16674 0.02550 6.19097 C 5 -0.18873 1.99874 4.16620 0.02379 6.18873 C 6 -0.18625 1.99873 4.16373 0.02378 6.18625 H 7 0.22901 0.00000 0.76833 0.00266 0.77099 H 8 0.22992 0.00000 0.76743 0.00265 0.77008 C 9 -0.18875 1.99874 4.16622 0.02379 6.18875 H 10 0.22903 0.00000 0.76832 0.00266 0.77097 C 11 -0.18626 1.99873 4.16374 0.02378 6.18626 C 12 -0.18913 1.99875 4.16690 0.02349 6.18913 H 13 0.22991 0.00000 0.76743 0.00265 0.77009 C 14 -0.19093 1.99873 4.16671 0.02549 6.19093 H 15 0.22959 0.00000 0.76782 0.00259 0.77041 C 16 -0.18847 1.99875 4.16435 0.02537 6.18847 H 17 0.23809 0.00000 0.75875 0.00316 0.76191 H 18 0.23261 0.00000 0.76395 0.00344 0.76739 H 19 0.23282 0.00000 0.76372 0.00346 0.76718 C 20 -0.18844 1.99875 4.16431 0.02539 6.18844 O 21 -0.18154 1.99989 6.15906 0.02259 8.18154 O 22 0.24106 1.99984 5.71197 0.04713 7.75894 O 23 -0.18154 1.99989 6.15906 0.02259 8.18154 H 24 0.23800 0.00000 0.75880 0.00320 0.76200 * Total * 0.00000 43.98133 75.58942 0.42924 120.00000

175 Table S7.1g) Summary of Natural Population Analysis of nBuFc-O3A Natural Natural Population Atom No Charge Core Valence Rydberg Total Fe 1 -0.30071 17.99370 8.24310 0.06392 26.30071 C 2 -0.19708 1.99866 4.16614 0.03229 6.19708 H 3 0.23986 0.00000 0.75640 0.00374 0.76014 C 4 -0.18369 1.99867 4.16227 0.02275 6.18369 C 5 -0.18379 1.99873 4.15938 0.02568 6.18379 C 6 -0.18685 1.99873 4.16432 0.02379 6.18685 H 7 0.23360 0.00000 0.76353 0.00287 0.76640 H 8 0.22796 0.00000 0.76926 0.00278 0.77204 C 9 -0.19242 1.99873 4.16988 0.02381 6.19242 H 10 0.22989 0.00000 0.76760 0.00250 0.77011 C 11 -0.18983 1.99874 4.16735 0.02375 6.18983 C 12 -0.18889 1.99874 4.16639 0.02376 6.18889 H 13 0.22883 0.00000 0.76862 0.00255 0.77117 C 14 -0.19177 1.99874 4.16944 0.02359 6.19177 H 15 0.23004 0.00000 0.76738 0.00259 0.76996 C 16 -0.19245 1.99874 4.17022 0.02349 6.19245 H 17 0.22855 0.00000 0.76890 0.00255 0.77145 H 18 0.22903 0.00000 0.76840 0.00257 0.77097 C 19 -0.01389 1.99862 3.97880 0.03647 6.01389 O 20 -0.18096 1.99990 6.15841 0.02266 8.18096 O 21 0.24782 1.99984 5.70716 0.04519 7.75218 O 22 -0.16662 1.99989 6.14291 0.02381 8.16662 H 23 0.22741 0.00000 0.77014 0.00244 0.77259 C 24 -0.38342 1.99918 4.36822 0.01602 6.38342 H 25 0.19886 0.00000 0.79794 0.00320 0.80114 H 26 0.20168 0.00000 0.79526 0.00307 0.79832 C 27 -0.36584 1.99925 4.35332 0.01327 6.36584 H 28 0.19653 0.00000 0.80044 0.00304 0.80347 H 29 0.18818 0.00000 0.80944 0.00238 0.81182 C 30 -0.37569 1.99929 4.36437 0.01203 6.37569 H 31 0.18351 0.00000 0.81381 0.00268 0.81649 H 32 0.18554 0.00000 0.81182 0.00264 0.81446 C 33 -0.56883 1.99934 4.56137 0.00812 6.56883 H 34 0.19815 0.00000 0.80008 0.00177 0.80185 H 35 0.19219 0.00000 0.80573 0.00208 0.80781 H 36 0.19508 0.00000 0.80279 0.00213 0.80492 * Total * 0.00000 51.97748 99.51056 0.51196 152.00000

176 Table S7.1h) Summary of Natural Population Analysis of nBuFc-O3B Natural Natural Population Atom No Charge Core Valence Rydberg Total Fe 1 -0.29677 17.99383 8.23291 0.07003 26.29677 C 2 -0.01439 1.99863 3.97843 0.03733 6.01439 C 3 -0.18692 1.99873 4.16313 0.02506 6.18692 C 4 -0.18420 1.99867 4.16327 0.02226 6.18420 C 5 -0.18326 1.99873 4.16089 0.02364 6.18326 H 6 0.22796 0.00000 0.76955 0.00249 0.77204 H 7 0.22929 0.00000 0.76804 0.00267 0.77071 C 8 -0.18922 1.99874 4.16697 0.02350 6.18922 H 9 0.22852 0.00000 0.76887 0.00261 0.77148 C 10 -0.18762 1.99873 4.16539 0.02350 6.18762 C 11 -0.19236 1.99875 4.17066 0.02294 6.19236 H 12 0.22985 0.00000 0.76764 0.00252 0.77015 C 13 -0.19239 1.99873 4.16823 0.02543 6.19239 H 14 0.22914 0.00000 0.76836 0.00251 0.77086 C 15 -0.18908 1.99875 4.16504 0.02528 6.18908 H 16 0.23770 0.00000 0.75910 0.00320 0.76230 H 17 0.23379 0.00000 0.76267 0.00354 0.76621 H 18 0.23132 0.00000 0.76552 0.00316 0.76868 C 19 -0.18439 1.99867 4.16137 0.02434 6.18439 O 20 -0.18509 1.99989 6.16323 0.02197 8.18509 O 21 0.23878 1.99985 5.71476 0.04661 7.76122 O 22 -0.18489 1.99989 6.16281 0.02218 8.18489 H 23 0.23794 0.00000 0.75901 0.00306 0.76206 C 24 -0.56864 1.99934 4.56121 0.00809 6.56864 H 25 0.19300 0.00000 0.80494 0.00206 0.80700 H 26 0.19923 0.00000 0.79902 0.00175 0.80077 H 27 0.19353 0.00000 0.80442 0.00206 0.80647 C 28 -0.37680 1.99929 4.36544 0.01207 6.37680 H 29 0.18531 0.00000 0.81202 0.00267 0.81469 H 30 0.18471 0.00000 0.81261 0.00268 0.81529 C 31 -0.36170 1.99925 4.34941 0.01303 6.36170 H 32 0.18975 0.00000 0.80782 0.00243 0.81025 H 33 0.19105 0.00000 0.80654 0.00241 0.80895 C 34 -0.38438 1.99918 4.36922 0.01598 6.38438 H 35 0.20000 0.00000 0.79681 0.00319 0.80000 H 36 0.20123 0.00000 0.79552 0.00325 0.79877 * Total * 0.00000 51.97767 99.51081 0.51152 152.00000

177 Table S7.1i) Summary of Natural Population Analysis of Fc-O2 Natural Natural Population Atom No Charge Core Valence Rydberg Total Fe 1 -0.3523 17.9942 8.2967 0.0614 26.3523 C 2 -0.1915 1.9987 4.1690 0.0237 6.1915 H 3 0.2270 0.0000 0.7704 0.0027 0.7730 C 4 -0.1915 1.9987 4.1691 0.0237 6.1915 C 5 -0.1914 1.9987 4.1690 0.0237 6.1914 H 6 0.2270 0.0000 0.7704 0.0027 0.7730 C 7 -0.1914 1.9987 4.1690 0.0237 6.1914 H 8 0.2270 0.0000 0.7704 0.0027 0.7731 C 9 -0.1914 1.9987 4.1690 0.0237 6.1914 H 10 0.2270 0.0000 0.7704 0.0027 0.7731 H 11 0.2269 0.0000 0.7704 0.0027 0.7731 C 12 -0.1931 1.9987 4.1700 0.0243 6.1931 H 13 0.2270 0.0000 0.7704 0.0027 0.7730 C 14 -0.1919 1.9987 4.1688 0.0244 6.1919 C 15 -0.1919 1.9987 4.1688 0.0244 6.1919 H 16 0.2272 0.0000 0.7702 0.0026 0.7728 C 17 -0.1928 1.9987 4.1697 0.0244 6.1928 H 18 0.2272 0.0000 0.7702 0.0026 0.7728 C 19 -0.1928 1.9987 4.1696 0.0244 6.1928 H 20 0.2271 0.0000 0.7703 0.0027 0.7729 H 21 0.2271 0.0000 0.7703 0.0027 0.7729 O 22 0.0009 1.9999 5.9654 0.0338 7.9992 O 23 0.0009 1.9999 5.9654 0.0338 7.9991 * Total * 0.0000 41.9813 69.6226 0.3960 112.0000

178 Table S7.1j) Summary of Natural Population Analysis of nBuFc-O2 Natural Natural Population Atom No Charge Core Valence Rydberg Total Fe 1 -0.34112 17.99374 8.28424 0.06315 26.34112 C 2 -0.19228 1.99873 4.16975 0.02380 6.19228 H 3 0.22634 0.00000 0.77102 0.00264 0.77366 C 4 -0.19323 1.99873 4.17078 0.02373 6.19323 C 5 -0.19511 1.99873 4.17296 0.02342 6.19511 H 6 0.22668 0.00000 0.77075 0.00256 0.77332 C 7 -0.19328 1.99873 4.17083 0.02372 6.19328 H 8 0.22628 0.00000 0.77112 0.00260 0.77372 C 9 -0.19227 1.99873 4.16976 0.02377 6.19227 H 10 0.22671 0.00000 0.77073 0.00256 0.77329 H 11 0.22639 0.00000 0.77097 0.00263 0.77361 C 12 -0.18724 1.99866 4.16567 0.02291 6.18724 H 13 0.22598 0.00000 0.77152 0.00250 0.77402 C 14 -0.18866 1.99873 4.16610 0.02383 6.18866 C 15 -0.01697 1.99861 3.98184 0.03653 6.01697 H 16 0.22616 0.00000 0.77107 0.00277 0.77384 C 17 -0.19044 1.99873 4.16694 0.02477 6.19044 C 18 -0.18871 1.99866 4.16639 0.02366 6.18871 H 19 0.22606 0.00000 0.77120 0.00274 0.77394 H 20 0.22620 0.00000 0.77124 0.00257 0.77380 O 21 0.00390 1.99990 5.96267 0.03354 7.99610 O 22 -0.00181 1.99990 5.96753 0.03438 8.00181 C 23 -0.38766 1.99918 4.37203 0.01645 6.38766 H 24 0.20097 0.00000 0.79613 0.00291 0.79903 H 25 0.19862 0.00000 0.79830 0.00309 0.80138 C 26 -0.36357 1.99925 4.35099 0.01333 6.36357 H 27 0.18857 0.00000 0.80899 0.00244 0.81143 H 28 0.19643 0.00000 0.80095 0.00262 0.80357 C 29 -0.37716 1.99929 4.36503 0.01285 6.37716 H 30 0.19169 0.00000 0.80559 0.00272 0.80831 H 31 0.18438 0.00000 0.81286 0.00276 0.81562 C 32 -0.57521 1.99934 4.56747 0.00839 6.57521 H 33 0.19955 0.00000 0.79869 0.00177 0.80045 H 34 0.19141 0.00000 0.80683 0.00177 0.80859 H 35 0.19243 0.00000 0.80555 0.00202 0.80757 * Total * 0.00000 49.97762 93.54450 0.47788 144.00000

179