Exploring Catalytic Tellurium- Based Antioxidants

Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology 1423

Synthesis and Evaluation

JIA-FEI POON

ACTA UNIVERSITATIS UPSALIENSIS ISSN 1651-6214 ISBN 978-91-554-9684-5 UPPSALA urn:nbn:se:uu:diva-302475 2016 Dissertation presented at Uppsala University to be publicly examined in B41, BMC, Husargatan 3, Uppsala, Thursday, 27 October 2016 at 09:30 for the degree of Doctor of Philosophy. The examination will be conducted in English. Faculty examiner: John Murphy (University of Strathclyde ).

Abstract Poon, J.-f. 2016. Exploring Catalytic Tellurium-Based Antioxidants. Synthesis and Evaluation. Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology 1423. 72 pp. Uppsala: Acta Universitatis Upsaliensis. ISBN 978-91-554-9684-5.

This thesis is concerned with the synthesis and evaluation of various tellurium-based chain- breaking antioxidants. The purpose is to find novel regenerable compounds with improved radical-trapping capacity. In the first part of this work, we explore the possibilities to incorporate tellurium into tocopherols and aromatic amines. Overall, tocopherols carrying alkyltelluro groups are better radical-trapping agents than the corresponding sulfur- and selenium analogues. Among them, 7-octyltelluro δ-tocopherol showed a ca. 17-fold higher reactivity than recorded for α- tocopherol and much better regenerability. Even longer inhibition times were recorded for the corresponding bis(tocopheryl) tellurides. In the aromatic amine series, diphenyl amines carrying alkyltelluro groups were shown to function as efficient radical-quenchers capable of inhibiting peroxidation for 460 min in the presence of N-acetylcysteine. Thiol-consumption experiments suggested that the long inhibition times are due to efficient quenching of in-situ formed alkoxyl radicals in a solvent cage. In the second part of the thesis, we study how the antioxidant properties are affected by variations in the electron density at tellurium and the number of alkyltelluro substituents in the molecule. Evaluation of a series of aryltelluro phenols carrying electron donating and electron withdrawing groups in the para-position of the aryl moiety suggested that a high electron density at the heteroatom prolonged the inhibition time. Among alkyltelluro phenols, alkyltelluro resorcinols and bis(alkyltelluro) phenols, phenols carrying alkyltelluro groups in both ortho positions were superior when it comes to radical-trapping activity and regenerability.

Keywords: antioxidant, tellurium, selenium, radical-trapping, chain-breaking, ROS, glutathione peroxidase, tocopherol, aromatic amine, oxidative stress

Jia-fei Poon, Department of Chemistry - BMC, Organic Chemistry, Box 576, Uppsala University, SE-75123 Uppsala, Sweden.

ISSN 1651-6214 ISBN 978-91-554-9684-5 urn:nbn:se:uu:diva-302475 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-302475)

To my parents

List of Papers

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

I Singh, V. P.; Poon, J.; Engman, L. Catalytic Antioxidants: Re- generable Tellurium Analogues of Vitamin E. Org. Lett. 2013, 15, 6274–6277.

II Poon, J.; Singh, V. P.; Yan, J.; Engman, L. Regenerable Antiox- idants–Introduction of Chalcogen Substituents into Tocopherol. Chem. Eur. J. 2015, 21, 2447–2457.

III Poon, J.; Yan, J.; Gates, P.; Engman, L. Regenerable Radical- trapping Tellurobistocopherol Antioxidants, Manuscript (2016).

IV Poon, J.; Yan, J.; Singh, V. P.; Gates, P.; Engman, L.; Alkyltel- luro Substitution Improves the Radical-Trapping Capacity of Aromatic Amines. Chem. Eur. J. 2016, 32, 12891–12903.

V Poon, J.; Singh, V. P.; Engman, L. In Search of Catalytic Anti- oxidants–(Alkyltelluro)phenols, (Alkyltelluro)resorcinols, and Bis(alkyltelluro)phenols. J. Org. Chem. 2013, 78, 6008–6015.

Appendix: Supplementary material.

Reprints were made with permission from the respective publishers.

Publications not included in this thesis:

VI Singh, V. P.; Poon, J.; Engman, L. Turning Pyridoxine into a Catalytic Chain-Breaking and Hydroperoxide-Decomposing Antioxidant. J. Org. Chem. 2013, 78, 1478–1487.

VII Singh, V. P.; Poon, J.; Butcher, R.; Engman, L. Pyridoxine- Derived Organoselenium Compounds with Glutathione Peroxi- dase-Like and Chain-Breaking Antioxidant Activity. Chem. Eur. J. 2014, 20, 12563–12571.

VIII Singh, V. P.; Poon, J.; Butcher, R.; Lu, X.; Mestres, G.; Karls- son Ott, M.; Engman, L. Effect of a Bromo Substituent on the Glutathione Peroxidase Activity of a Pyridoxine-like Diselenide. J. Org. Chem. 2015, 50, 7385–7395.

IX Yan, J.; Poon, J.; Singh, V. P.; Gates, P.; Engman, L. Regener- able Thiophenolic Radical-Trapping Antioxidants. Org. Lett. 2015, 17, 6162–6165.

X Kumar, S.; Yan, J.; Poon, J.; Singh, V. P.; Lu, X.; Karlsson Ott, M., Engman, L., Kumar, S. Multifunctional Antioxidants: Re- generable Radical-Trapping and Hydroperoxide-Decomposing Ebselenols. Angew. Chem. Int. Ed. 2016, 55, 3729–3733.

XI Lu, X.; Mestres, G.; Singh, V. P.; Effati, P.; Poon, J.; Engman, L.; Karlsson Ott, M. Selenium- and Tellurium-Based Antioxi- dants for Modulating Inflammation and Osteoblastic Activity, Submitted manuscript (2016).

XII Singh, V. P.; Poon, J.; Yan, J.; Lu, X.; Karlsson Ott, M.; Butch- er, R. J.; Gates, P.; Engman, L. Nitro-, Azo- and Amino Deri- vates of Ebselen. Synthesis, Structure and Cytoprotective Ef- fects, Manuscript (2016).

Contribution Report

The author wishes to clarify his contribution to the papers included in the thesis:

I Participated significantly in the synthesis and evaluation experiments using the two-phase lipid peroxidation system. Contrib- uted to writing of the paper.

II Performed a significant part of the synthesis and evaluation experiments using the two-phase lipid peroxidation system. Contributed to writing of the paper.

III Contributed to the formulation of the research idea and participated extensively in the synthetic work. Performed all evaluation experiments using the two-phase peroxidation system and all thiol-consumption experiments. Contributed to writing of the paper.

IV Performed most of the synthetic work, evaluation experiments using the two-phase lipid peroxidation system and thiol- consumption studies. Contributed to writing of the paper.

V Performed all the synthesis and evaluation experiments using the two-phase peroxidation system. Participated in the GPx- experiments. Contributed to writing of the paper.

Appendix: Contributed to the formulation of the research idea and participated extensively in the synthetic work. Performed most of the evaluation experiments using the two-phase peroxidation system and all thiol-consumption experiments. Contributed to writing of the appendix.

Contents

1. Introduction ...... 1 1.1 Oxidation ...... 1 1.1.1 Oxygen ...... 1 1.1.2 Free Radicals ...... 2 1.1.3 Autoxidation ...... 3 1.1.4 Oxidative Stress ...... 5 1.2 Antioxidants ...... 6 1.2.1 Preventive antioxidants ...... 7 1.2.2 Chain-Breaking Antioxidants ...... 8 1.3 Chalcogens ...... 11 2. Antioxidant Design ...... 13 2.1 Perspectives on Chain-Breaking Antioxidants ...... 13 2.1.1 Hydrogen-Atom Transfer ...... 13 2.1.2 Criteria for effective antioxidants ...... 14 2.1.3 Stoichiometric factor ...... 17 2.2 Perspectives on Preventive Antioxidants ...... 19 2.3 Towards Multifunctional and Regenerable Antioxidants ...... 21 2.3.1 Background ...... 21 2.3.2 Multifunctional Antioxidants ...... 21 2.3.3 Regeneration ...... 22 2.3.4 Aim of the work ...... 23 3. Synthesis ...... 25 3.1 Synthesis of Chalcogen-Containing Tocopherol Antioxidants (Papers I – III) ...... 25 3.2 Synthesis of Tellurium-Containing Aromatic Amine Antioxidants (Paper IV) ...... 29 3.3 Synthesis of Phenolic Antioxidants: Alkyltelluro Phenols, Alkyltelluro Resorcinols and bis(Alkyltelluro) Phenols (Paper V) ...... 35 3.4 Synthesis of Aryltelluro Phenol Antioxidants (Appendix) ...... 37

4. Evaluation ...... 40 4.1 Antioxidant Profile ...... 40 4.2 Evaluation of the Radical-Trapping Capacity and Regenerability ..... 40 4.2.1 The Two-Phase Lipid Peroxidation Model ...... 40 4.2.2 Evaluation of Chalcogen-Containing Tocopherol Antioxidants (Papers I – III) ...... 42 4.2.3 Evaluation of Tellurium-Containing Aromatic Amine Antioxidants (Paper IV) ...... 45 4.2.4 Evaluation of Alkyltelluro Phenols (Paper V) ...... 47 4.2.5 Evaluation of Aryltelluro Phenol Antioxidants (Appendix) ...... 49 4.2.6 Thiol Consumption Assay ...... 51 4.2.7 NAC-consumption during Peroxidation Inhibited by Tellurium-Based Antioxidants ...... 52 4.3 Catalytic Decomposition of Hydrogen Peroxide ...... 55 4.3.1 The Coupled Reductase Assay ...... 55 4.3.2 The Thiol Peroxidase Assay ...... 56 4.3.3 Evaluation ...... 56 5. Mechanistic Considerations ...... 58 6. Concluding Remarks ...... 61 Summary in Swedish ...... 63 Acknowledgement ...... 66 References ...... 68

Abbreviations

a.q. Aqueous AIBN Azobisisobutyronitrile AMVN 2,2'-Azobis(2,4-dimethylvaleronitrile) Ar Aryl ATP Adenosine triphosphate BDE Bond dissociation enthalpy BHA Butylated hydroxyanisole BHT Butylated hydroxytoluene Boc tert-Butyloxycarbonyl Conc. Concentrated DCM Dichloromethane DMF N,N-Dimethylformamide DMSO Dimethylsulfoxide DNA Deoxyribonucleic acid EDG Electron donating group EPR Electron paramagnetic resonance Eq. Equivalent ET Electron transfer EWG Electron withdrawing group GPx Glutathione peroxidase GSH Glutathione HAT Hydrogen-atom transfer HOMO Highest occupied molecular orbital HPLC High-performance liquid chromatography IP Ionization potential LUMO Lowest occupied molecular orbital m-CPBA meta-Chloroperoxybenzoic acid NAC N-Acetylcysteine NADPH Nicotinamide adenine dinucleotide phosphate NBS N-Bromosuccinimide NMMO N-Methylmorpholine N-oxide NMR Nuclear magnetic resonance PCET Proton-coupled electron transfer Ph Phenyl PT Proton transfer r.t. Room temperature

ROS Reactive oxygen species SOD Superoxide dismutase SOMO Singly occupied molecular orbital TFA Trifluoroacetic acid THF Tetrahydrofuran THP Tetrahydropyran UV Ultraviolet

1. Introduction

While evolution has given us antioxidants to protect ourselves from oxidative stress, organic chemists have kept trying to develop new ones with improved antioxidative properties. Today, more than one million tons of antioxidants are consumed annually for different purposes such as nutrition sup- plements, food conservatives and additives to fuels, polymers and other organic materials. Their role is to prevent or delay decomposition in the presence of air. This thesis is concerned with the design, synthesis and evaluation of a series of novel tellurium-based antioxidants with outstanding radical-trapping activity. These compounds are regenerable and multifunctional in the sense that they target both peroxyl radicals and hydroperoxides. Before we start the journey towards the unknown, I would like to take the opportunity to provide the readers with a brief background to the topic.

1.1 Oxidation 1.1.1 Oxygen Undoubtedly, molecular oxygen is the most essential element for living or- ganisms. Ironically, it is also associated with oxidative degradation of organic materials and many pathological conditions. Fortunately, it is kinetically unfavourable for the molecular oxygen in the ground state to react with organic substrates via a two-electron process. In the ground state, dioxygen is a triplet i.e. the two unpaired electrons are in two degenerate singly occupied molecular orbitals (SOMOs). According to the Pauli exclusion principle, an atomic orbital can at most contain two electrons and they must have opposite spins.1 It is therefore unlikely for oxygen to accept an electron pair from any organic substrate since it would require a conversion from a triplet state to a singlet state. Rather, it would be natural for dioxygen to accept a single electron or combine with a carbon-centred radical to form a peroxyl radical. Therefore, oxidation of organic materials is largely occurring via processes involving free radicals.

1.1.2 Free Radicals A free radical is defined as an atom, molecule or ion with unpaired valence electrons.2 Such species are in general highly reactive. Depending on the energy of the SOMO, it can either interact with the highest occupied molecular orbital (HOMO) or the lowest unoccupied molecular orbital (LUMO) of another molecule, hence being either electrophilic or nucleophilic. The SO- MO-HOMO and the SOMO-LUMO interactions are favoured whenever a more stable radical is formed. The concept “unpaired electrons” was already introduced in the first half of the 19th century. Indeed, Bunsen claimed the existence of the dime- thylarsinyl radical in 18423 while Frankland reported the observation of the free ethyl radical in 1850.4 However, the rapid dimerisation of these free radicals made the characterization difficult. For long, many agreed that it was not possible to isolate free radicals and this opinion was prevailing until the discovery of the triphenylmethyl radical (1) by Gomberg. In 1900, Gomberg suggested that 1 was formed when triphenylmethyl chloride was treated with silver metal.5 Later it was shown to be equilibrating with its dimer in solution (Scheme 1). To avoid confusion with the old term radical (the Latin “radix” means root) which was employed by Lavoisier in 1789 to describe a group of atoms with characteristic properties6 or with a modern chemical terminology, functional group, Gomberg introduced the term free radical to denote a species with unpaired electrons.

1 Scheme 1. Gombergs radical 1 in equilibrium with a dimer.

The field of radical chemistry kept a rather low profile for the next 30 years until 1937 when three important publications by Hey7, Flory8 and Kharasch9 appeared. Hey proposed that phenyl radicals could be important intermedi- ates whereas Flory's kinetic study of vinyl polymerization introduced the concept of chain transfer. Kharasch revised the mechanism for chain reactions when he reported the radical addition of hydrogen bromide to olefins. While these publications contributed to some of the most important ideas to the field, free radical chemistry showed its practical potential when radical- polymerization became one of the most important methods to get access to synthetic rubber during the second world war.

As early as in the 1950s, Gerschman et al. proposed that oxygen toxicity is related to free radicals.10 The discovery of superoxide dismutase11 in the 1960s gave a hint as to how our body defends itself against reactive oxygen species (ROS), a collective name for peroxyl radicals, hydroxyl radicals, singlet oxygen, superoxide and hydroperoxides. Harman proposed in 1972 that the mitochondrial production of ROS is a cause of ageing.12 Today, we do not only know that free radicals are relevant for different diseases,13 but also that they are involved in many essential biological processes such as cellular signalling.14 The study of ROS was simplified when methods for detection of free radicals were available. The most significant development came with the electron paramagnetic resonance (EPR) technique in the 1940s, which not only provided a powerful tool to detect and characterize free radicals, but was also together with the rotating sector method15, the only way to determine absolute rate constants for reactions involving free radicals.16 The new idea to spin-label biomolecules by McConnell in 1960s improved our understand- ing of various biological processes involving free radicals and development of spin-labelling methods allowed studied of ROS in the body.17 It is noteworthy that free radicals can be stabilized by different means e.g. by resonance-interactions with electron withdrawing (EWG) or electron donating groups (EDG). Synergistic effects, known as captodative effects, could also be observed when the radical is stabilized by both EWG and EDG groups. In Gomberg's radical 1, the three benzene rings stabilize the radical by resonance effects and induction. Resonance-stabilized free radicals are key to the action of chain-breaking donating antioxidant (vide infra).

1.1.3 Autoxidation Autoxidation brings about the conversion of hydrocarbons (R-H) to the corresponding hydroperoxides (ROOH) via a free radical chain reaction in the presence of oxygen. The reaction can be subdivided into three elemental steps: initiation, propagation and termination (Scheme 2).18 Formation of free radicals are initiated by various environmental factors such as UV-light, heat, metals and ionizing radiation.19 The initiating radical (In•) combines with molecular oxygen at a diffusion-controlled rate to form a peroxyl radical (In-OO•), which in turn abstracts a H-atom from R-H to produce a carbon-centred radical, R• (eq. 1-3). Also, the weak oxygen- oxygen bond (BDE = ca. 45 kcal mol-1)20 of peroxides could be cleaved in the presence of transition metals, UV light or heat (eq. 4). The resulting alkoxyl (RO•) and hydroxyl radicals (HO•) may in turn start new chain reactions by abstracting H-atoms from hydrocarbons.

Scheme 2. The initiation, propagation and termination steps of autoxidation.

In biological systems, free radicals are generated by different enzymes (NADPH oxidase21, xanthine oxidase22, cytochrome P45023 and nitric oxide synthase24). They also leak out as by-products from the electron transport chain in the mitochondria25. On the other hand, azo-initiators such as azobisisobutyronitrile (AIBN; 2) and 2,2'-azobis (2,4-dimethylvaleronitrile) (AMVN; 3) are commonly used by organic chemists for the generation of carbon-centred radicals.

Upon heating, they undergo homolytic cleavage to form nitrogen and two carbon-centred radicals. Thermal decomposition of azo-compounds occurs at a well-defined rate and follows first-order kinetics. A wide selection of azo-compounds with different properties (hydrophilic, lipophilic and am- phiphilic) are available. Alternatively, initiation of free radical reactions can be performed photochemically. During the propagation step, the carbon-centred radical (R•) reacts with dioxygen at a close to diffusion-controlled rate (k = 2 – 5·109 M-1s-1)26 to form a peroxyl radical (eq. 5). As a chain-carrying species in the autoxidation process, the in-situ formed peroxyl radical propagates the reaction by abstracting a hydrogen-atom from hydrocarbons in the system (eq. 6) at a -1 -1 27 relatively slow rate (kprop = 62 M s for linoleic acid at 37 °C) to produce ROOH and another R•.

Termination occurs by radical-radical combination of any two of ROO•/R• (eq. 7 – 9), and occurs with rate constants ranging from 106 to 109 M-1s-1.27, 28 However, due to the low concentration of ROO• in the system, the termination is not competitive when the system contains a high concentration of R-H. Also, reactions 8 and 9 are unlikely to contribute to any termination in the presence of dioxygen (reaction 5 is so fast). Thus, to prevent the exponential growth of new chains, one would need to interfere with the propagation step. Indeed, supplementation of antioxidants has long been a strategy to prevent degradation of man-made materials.

1.1.4 Oxidative Stress Autoxidation in biological membranes is called lipid peroxidation and it is known to be associated with oxidative stress. The term "oxidative stress" refers to conditions where production of ROS exceeds the amounts that the antioxidant defence systems can handle. Overproduction of ROS in biological systems can be triggered by air pollutants, bacteria, UV-light, drugs or toxic food. The highly reactive ROS-compounds can be damaging to DNA, sugars, lipids and proteins29 and for long ROS were regarded as the major cause of various diseases, such as cardiovascular disease,30 cancer,31 chronic lung diseases32 and aging.33 However, studies by MacCord and Fridovich in the late 60's revealed that oxidative processes involving ROS were ongoing even at normal physiological conditions,34 suggesting that free radicals are in fact important participants in many biological processes. Indeed, accumulated evidence indicates that pathophysiological disorder is not only caused by an excess of ROS but also by insufficient generation of the same. Mechanistically, lipid peroxidation products are formed via H-atom abstraction from the lipid, reaction with oxygen and H-atom abstraction by the resulting peroxyl radical. Due to isomerization, a mixture of lipid hydroperoxides (5-9) with varying (E)- and (Z)- configurations are produced from linoleic acid (Figure 1). Due to the rapid β-fragmentation of the peroxyl 6 -1 -1 35 radical 4, which occurs with a rate constant kβ = 2.4 x 10 M s , the non- conjugated lipid hydroperoxide 5 could only be trapped by good antioxidants 6 -1 -1 36 such as α-tocopherol (kinh = 3.2 x 10 M s ). On the other hand, β- fragmentation of conjugated hydroperoxyl radicals occurs at a much slower rate. The rate of the β-fragmentation is also affected by the polarity of the solvent as well as hydrogen-bonding. Consequently, the distribution of oxidation products will vary depending on the conditions.35 Lipid peroxidation will cause severe biological dysfunction. The lipids are not only important building blocks for the eukaryotic cell membranes37 but also have a function in cell signalling38 and to act as energy storage units39. Furthermore, it is also well known that cytotoxic compounds40 such as α,β-unsaturated aldehydes could be formed by oxidative cleavage of the unsaturated hydroperoxides. Today, lipid peroxidation is associated with a

number of diseases such as atherosclerosis,41 cancer,42 diabetes,43 lung inju- ry,44 Alzeimer’s45 and Parkinson’s disease.46

Figure 1. Peroxidation of linoleic acid.

1.2 Antioxidants Antioxidants are substances that can delay, prevent or remove oxidative damage to a target molecule.47 Traditionally, they are divided into chain- breaking antioxidants and preventive antioxidants, depending on how they act to remove ROS.

1.2.1 Preventive antioxidants Preventive antioxidants retard the formation of ROS that could initiate autoxidation. They do so by reducing hydroperoxides or preventing them to fall apart to form reactive radicals (metal complexing agents and UV- absorbers). In biological systems, enzymatic antioxidants such as superoxide dismutases (SOD), catalases and glutathione peroxidases (GPx) have an important role as preventive antioxidants. SOD catalyses the dismutation of superoxide anion to oxygen and hydrogen peroxide, which in turn, is reduced further to water in the presence of catalases or glutathione peroxidases. SODs are divided into four categories, based on the metal ion (iron, manganese, copper or nickel) at the active center where the dismutation of superoxide anions is taking place.48 Mechanisti- cally, the catalytic cycle involves a sequential reduction-oxidation step of the metal. The process is essentially diffusion-controlled. Catalases are known to dismutate hydrogen peroxide to water and oxygen. Since the first report in 1900 by Loew,49 more than 300 different catalases have been identified.50 Although the overall process is simple, the action of the catalases is rather complicated. Initially, an Fe(III) at the active center is oxidized by hydrogen peroxide to form a Fe (IV) species where one of the electrons is delocalized into the heme group. By accepting an electron from a second molecule of hydrogen peroxide, iron is reduced back to the trivalent state. This closes the catalytic cycle. GPx and mimics thereof have attracted a lot of attention from organic chemists. From a biological point of view, the ubiquitous glutathione (GSH) protects us from oxidative stress by reducing hydrogen peroxide to water. However, the uncatalyzed process is rather slow. Therefore, Nature gave us the GPx to catalyze the reaction. GPx is a general name for an enzymatic family first discovered in 1957.51 It contains eight isoenzymes (GPx1– 8). Structurally, only GPx1– 4 and GPx6 are selenoproteins while the rest of the family members contain a cysteine at the active site. Since GPx1 – 4 showed a much higher GPx-activity than the other GPx members, it has been specu- lated that the selenocysteine in the catalytic center plays a vital role in the reduction of hydrogen peroxide. Both experiment52 and theory53 suggest that a selenol reacts with hydrogen peroxide to form selenenic acid, which is then reduced, via a selenosul- fide, by two equivalents of GSH (Scheme 3).

Scheme 3. Catalytic cycle for the action of the glutathione peroxidases.

1.2.2 Chain-Breaking Antioxidants Chain-breaking antioxidants have been extensively used in various materials.54 As the term suggests, they interfere with the chain reaction by scav- enging peroxyl radicals. The chain-breaking antioxidants are further classi- fied into accepting and donating antioxidants.

Scheme 4. Radical-trapping by a chain-breaking accepting antioxidant.

The chain-breaking accepting antioxidants are themselves free radicals and they can compete with oxygen for the carbon-centred radicals in the reaction described in eq. 5. Non-radical products are formed and, thus, the chain reaction is "broken". Nitroxides are known to act as chain-breaking accepting antioxidants and traditionally, diphenyl amines are used as their precursors (Scheme 4).

Scheme 5. Radical-trapping by a chain-breaking donating antioxidant (phenol).

As compared to chain-breaking accepting antioxidants, the chain-breaking donating ones seem to have wider applicability both in nature and in the industry. Chain-breaking donating antioxidants (AH) quench peroxyl radi-

cals (ROO•) by transferring a hydrogen atom to them. A resonance- stabilized, unreactive, radical A• is formed, which can no longer propagate autoxidation (Scheme 5). Nature gave us vitamins C and E for protection against oxidative stress whereas organic chemistry provided us with series of Nature-inspired hindered phenols. Vitamin E is recognized as the most reactive lipophilic chain-breaking antioxidant in man. The vitamin E family consists of eight members: four tocopherols (10-13) and four tocotrienols.55 Structurally, tocopherols are similar to the tocotrienols. While tocopherols carry a saturated phytyl chain, the side chains of the corresponding tocotrienols are triply unsaturated. Among tocopherols, α-tocopherol is the most reactive. It is known to quench two peroxyl radicals per molecule before it is converted into an inactive quinone form.

R1 HO

R3 O R2 10 α-Tocopherol R1 = R2 = R3 = Me 11 β-Tocopherol R1 = R2 = Me; R3 = H 12 γ-Tocopherol R1 = H; R2 = R3 = Me 13 δ-Tocopherol R1 = R3 = H; R2 = Me

Structurally, chain-breaking antioxidants are often phenols or aromatic amines. Ascorbic acid (14), which is commonly known as vitamin C, is an example of a non-phenolic antioxidant. It was discovered in the late 1920s by Albert Szent-Györgyi,56 and synthesis development in the mid 1930s57 made it available at a reasonable price. Because of the many hydroxyl groups, 14 is a water soluble chain- breaking antioxidant. For long, it has been known to act as a peroxyl-radical quencher. Vitamin C exists as ascorbate at physiological pH and it quenches peroxyl radicals by electron/proton transfer (Scheme 6).

ROO ROOH HO H pK = 4.2 HO H O a HO H O O O O HO HO O HO

HO OH O OH O O 14

Scheme 6. The radical-trapping action of ascorbic acid.

In 1941, Golumbic and Matill observed that the inhibition period for vitamin E in vitro could be extended in the presence of ascorbic acid.58 Later on, Tappel suggested that the ascorbate could act as a reducing agent towards the tocopheroxyl radical, hence regenerating the tocopherol (Scheme 7). This is one of the first examples of a catalytic antioxidant.

Scheme 7. The regeneration of α-tocopherol in the presence of ascorbic acid.

There are numerous man-made antioxidants where the structures have been inspired by Nature. Hindered phenols, e.g. butylated hydroxyanisole (BHA; 15 and 16) and butylated hydroxytoluene (BHT; 17) are arguably the most important examples of man-made chain-breaking antioxidants for the modern society. Indeed, these synthetic antioxidants are supplemented to products such as petroleum, oil, rubber and food.59 Structurally, they are all reminiscent of α-tocopherol. However, due to the simplified structure, the reactivity towards peroxyl radicals is lower. Since 1947, BHA has been used to stabilize fat food. Today, BHA is not only used as a food preservative under the name E320, but also as a common additive to palm oils and soy bean for preservation purposes.60 In fact, BHA is a mixture of two constitutional isomers (15 and 16). Among them, 16 is generally known to have a higher radical-trapping capacity.

OH OH OH

OMe OMe

15 16 17

The synthesis of BHT was patented in the US in 1947.61 BHT was mainly employed as a chain-breaking antioxidant for plastics, oil and food.59, 62 In- deed, BHT was approved as a stabilizer for food already in 195463 and was further confirmed to be safe for this purpose in 1977.64 Today, BHT is mainly used as a preservative for food containing high amounts of fat.

1.3 Chalcogens The term "chalcogen" was first coined by the German chemist Werner Fischer in the 1930s to denote the chemical elements in the group 16 (oxygen, sulfur, selenium, tellurium and polonium) of the periodic table.65 Liter- ally the name chalco-gen means "copper-born", from the ancient Greek words khalcós (Copper) and genēs (born). As the name is reminiscent of the halogens (salt-former), the nomenclature was widely used among many others in the field and in textbooks. In 1938, the term was officially approved by the Committee for the Reform of Inorganic Chemical Nomenclature.66 In comparison with oxygen, which we have previously discussed, the remaining members of the chalcogen family behave, chemically and physical- ly, rather different. Not surprisingly, oxygen is often not included in discussions regarding chalcogen chemistry; neither is the radioactive polonium. Since elemental sulfur occurs naturally, there are many records worldwide on the use of sulfur for different purposes. For example, sulfur was mentioned in the bible 15 times in the name of brimstone67 and has been used in China since as early as the 6th century B.C.68 However, it was not until 1777 that sulfur was recognized as a chemical element by Antoine Lavoisier.69 Organosulfur compounds such as gluathione and S-nitrosothiols70 play important roles in biological systems. The antibiotic penicillin is also a good example. Selenium was discovered by Jöns Jacob Berzelius in 1817 and the name originates from the Greek word selènè, meaning moon.71 Initially, selenium was recognized as a toxin, but further studies revealed the essential role of selenium for the immune system,72 thyroid hormone metabolism73 and intra- cellular redox regulation.74 While an intake of selenium in excess would lead to selenium poisoning, selenium deficiency is reported to be associated with various diseases such as Kashin-Beck disease75 and Keshan disease.76 In-

terestingly, the UGA-codon, which is usually a stop-codon, is also coding for selenocysteine. The selenium analogue of cysteine is incorporated into 25 selenoproteins77 in man, including the GPx-enzyme discovered in 1973.78 The discovery of tellurium was first reported by Franz-Joseph Müller von Reichenstein in 1782.79 While examining the gold-containing ores of Tran- sylvania, he obtained a sample of an unknown metal which he initially thought was antimony. Realizing that the properties of the newly discovered metal were not identical to any known elements, he gave it the name metal- lum problematicum (Problematic metal). The finding did not stir much attention until 1798, when Martin Heinrich Klaproth performed a more detailed investigation of this metal and confirmed the discovery of a new element. He decided to name it Tellurium after the Greek word tellus, meaning earth. Although the first organotellurium compound was reported by Friedrich Wöhler already in 1840 (diethyl telluride),80 the development of tellurium chemistry was rather slow. Today, the major applications of tellurium are in the metallurgy, rubber and electronics industries. Furthermore, cadmium telluride solar panels have shown some potential for electricity production.81 Biologically, tellurium is not essential to humans. However, tellurocysteine and telluromethionine have been found in fungi.82 In general, heavy chalcogens are readily oxidized. Thiolates, selenolates and tellurolates upon oxidation give disulfides, diselenides and ditellurides, respectively. Sulfides, selenides and tellurides are converted to the corresponding oxides (sulfoxide, selenoxide and telluroxide). Reduction of chalcogen is also a facile process. Chalcogen compounds are therefore useful in applications where facile redox-cycling is required.

2. Antioxidant Design

2.1 Perspectives on Chain-Breaking Antioxidants 2.1.1 Hydrogen-Atom Transfer Mechanistically, phenols and aromatic amines transfer H-atoms to quench reactive peroxyl radicals. The resulting phenoxyl/aminyl radical is resonance-stabilized, hence uncreative enough not to initiate new chains by abstracting hydrogen atoms from the substrate. After reaction with another molecule of a peroxyl radical they turn into non-radical products. The HAT from phenols and aromatic amines is generally recognized to be the key step in quenching of peroxyl radicals, however, other mechanisms such as proton-coupled electron transfer (PCET) have also been proposed (Scheme 8). A hydrogen-atom is transferred from the phenol to the oxygen- atom bearing the unpaired electron in HAT whereas in PCET, a proton is transferred from the phenol to the lone pair of the oxygen-atom while simul- taneously an electron moves from a lone pair in the phenol to the orbital of the oxygen-atom bearing the unpaired electron (Figure 2). Whether the reaction occurs via HAT or PCET depends on the energy barrier of the reaction pathway which is affected by solvent and pH.83

Scheme 8. Direct hydrogen-atom transfer and proton-coupled electron transfer.

The efficiency of hydrogen-atom transfer is described by the inhibition rate constant, kinh. It should be several orders of magnitude larger than the propagation constant, kprop, for efficient inhibition of autoxidation.

Figure 2. The mechanism of PCET.

It is noteworthy that HAT occurs more efficiently between two heteroatoms than between one heteroatom and carbon.84 For example, whereas the bond dissociation enthalpy (BDE) is similar for the O-H bond of phenol and the C-H bond of toluene, the reactivity of the hydrocarbon is much lower. The explanation is not clear. Zavitsas introduced the concept of triplet repulsion.85 He proposed that the H-atom transfer between two atoms could be described by a model like X↑H↓Y↑, where the X and Y radicals have paral- lel spins while the H atom has an opposite spin. Since both the X and Y atoms carry a spin of the same configuration, there will be a repulsive force between them and this force increases when the distance becomes shorter. The oxygen atom is more electronegative than carbon, hence, more of the spin density would be localized there. This would reduce the spin repulsion. On the other hand, Mayer et al. suggested that PCET is probably involved in the reaction between two heteroatoms while a direct hydrogen-atom transfer mainly occurs between carbon and heteroatoms.86 It has also been suggested that a pre-organized hydrogen-bonded complex lowers the energy barrier for the HAT to occur.87 Since this is only possible for heteroatoms, the HAT between carbon and heteroatoms or between two carbon atoms is in general less efficient.

2.1.2 Criteria for effective antioxidants Among phenolic chain-breaking antioxidants, some compounds quench per- 6 oxyl radicals more efficiently than others. For example 10 (kinh = 3.2 x 10 -1 -1 88 M s in benzene at 25°C) is almost 300-fold more reactive than 17 (kinh = 1.1 x 104 M-1s-1 in chlorobenzene at 25°C).90 To understand why, it is neces- sary to have a closer look to the factors that influence the kinh. The reactivity of a phenolic antioxidant is determined by the rate-limiting 89 H-atom transfer. BDEO-H — the energy that is required to cleave the O-H bond, is one of the main factors that influences the HAT and could be described as the energy difference between the phenoxyl radical and the parent phenol (Figure 3).

EWG O + H O + H

EWG = Ph, CN, NO2 etc EDG O + H

BDE (O-H)

EDG OH

OH EWG OH

Figure 3. The bond dissociation enthalpy (BDE) is affected by electron-withdrawing and electron-donating groups.

For good antioxidants it is required that the BDEO-H is much lower than the -1 89 BDEO-H of the hydroperoxide formed, which is around 88 kcal mol . The unfavourable thermodynamics explain why phenol (BDE = 87 kcal mol-1) is not a good chain-breaking antioxidant. Functional groups when introduced into phenols can alter the BDEO-H. In 1963, Ingold and co-workers studied the capacity of different para- substituted phenols to inhibit autoxidation of styrene.90 In general, electron- donating substituents lower the BDE of the phenolic O-H while electron- withdrawing ones have the opposite effect. The lower the BDE, the more efficient is the HAT. The effect depends also on the location of the substituent. The maximum effect is observed for the para-position followed by ortho whereas substitution in the meta-position has only a marginal effect. Actually, the substitution will affect the stability of the phenoxyl radical as well as the parent phenol and the combined effect will determine the influence on the BDEO-H. The parent phenol is destabilized by electron-donating groups and stabilized by electron-withdrawing ones. However, while the phenoxyl radical is stabilized by electron-donating groups the influence of electron-withdrawing groups is rather small (Figure 3).

Scheme 9. Due to the low ionization enthalpy, 18 is not stable in the presence of oxygen.

While the methoxy group is commonly introduced in the para-position in order to lower the BDEO-H and increase the reactivity of chain-breaking antioxidants, the dialkylamino group, which is in fact a better electron donor, could not be used. This is because the strongly electron donating amine would lower the ionization potential (IP) of the compound, and, thus, favour a direct electron transfer to the dioxygen. For example, 18 is known to slowly decompose in air at room temperature (Scheme 9).88

One would expect that phenol 20 is more reactive than 19 due to the electron-donating methoxy substituent in the para position. Surprisingly, the experimental data showed that 20 is less reactive than 19 by a factor of three.88 When one of the methyl groups in the meta position was removed (compound 21), the reactivity increased considerably. It was proposed that the rate-enhancement was the result of a stereoelectronic effect. Since the stabilization is strongly dependent on the alignment of the oxygen sp3-orbitals holding the lone pairs and the ring system, the orientation of the substituents will be the key to understand the influence on the BDE. In a similar vein, Ingold et al. compared the BDE of alkoxyphenols and the corresponding benzannulated cyclic ethers in the mid-80s.88 For simple phenols, the methoxy group is rotated away from the plane of the ring system with a dihedral angle of 89 degrees in 20 between the p-orbital of the heteroatom and the π-orbitals of the ring system. This results in a less stabilized phenoxyl radical, and thus a higher BDEO-H. On the other hand, when the heteroatom is part of a five- or six-membered ring (22 and 23), the orbitals of the oxygen overlap much better with the π-system of the ring, which results in a lower BDEO-H (Figure 4).

Figure 4. Dihedral angles in compounds 20, 22 and 23.

As mentioned above, the strategy to introduce electron donating groups into the ortho/para positions of phenols would only be successful as long as the ionization enthalpy of the compound is high enough that electron transfer to oxygen is not possible. By the turn of the millennium, Pratt and co-workers showed a way to circumvent the IP-problem by introduction of heteroatoms into the phenolic ring.91 Pyridine and pyrimidine analogues 24-27 had a high IP whereas the BDEO-H was almost in the same range as recorded for the parent phenolic compound. The reactivity of dialkylamine substituted pyridinols and pyrimidinols towards peroxyl radical was up to 88 times higher than recorded for α-tocopherol! Importantly, these compounds are stable in air.

OH OH HO HO

N N NN NN N N C16H33

NMe2 NMe2 24 25 26 27

Antioxidant-derived phenoxyl radicals are usually rather stable. On rare oc- casions, such as in low density lipoprotein92 they have been found to propagate peroxidation. Steric hindrance would make them even less reactive. Thus, introduction of methyl and tert-butyl groups into phenol caused a drastic lowering in the rate constants 2 kt (eq 7) for termination (Figure 5).

O O O O

-1 9 7 8.6 x 103 2kt (s )= 2.6 x 10 4.5 x 10 0.1

Figure 5. Rate constants (2 kt) for termination of phenoxyl radicals

Finally, it is important to remember that the inhibition rate constant, kinh, for phenols is strongly solvent-dependent and hydrogen bonding solvents have a tendency to slow down the H-atom transfer.93

2.1.3 Stoichiometric factor The stoichiometric factor n is another important parameter for chain- breaking antioxidants. It is defined as the number of peroxyl radicals that each molecule of the antioxidant can trap. For instance, α-tocopherol traps

two peroxyl radicals before it is converted to tocopherol quinone 28, hence n = 2 (Scheme 10). 94

Scheme 10. α-Tocopherol traps two peroxyl radicals.

It would be nice if an antioxidant could trap a larger number of peroxyl radicals. This is indeed the case for aromatic amines which could be regenerated at higher temperatures.

Scheme 11. Proposed mechanism for the catalytic action of diphenylamine at elevated temperature.

Aromatic amines have since long been used as additives to fuels and petroleum products.95 In 1978 a stoichiometric number of 41 was recorded for diphenylamine in paraffin oil at 130 °C.96 The high n-value remained a mys- tery until 1995, when Korcek and co-workers proposed a catalytic cycle where the antioxidant could be regenerated (Scheme 11).97 Initially, diphenylamine performs a HAT to the peroxyl radical to form an aminyl radical 29. Then it traps another molecule of peroxyl radical to form an unstable peroxyl dihenyl amine 30, which in turn decomposes to form the

nitroxide 31. Radical-radical combination of the nitroxide and substrate- derived alkyl radical forms the N-alkyl hydroxyl amine 32, which upon heating fragmentises to regenerate diphenylamine. Therefore, the catalytic cycle is only operative at elevated temperature. At room temperature, the n-value of diphenylamine is only 2. A revised mechanism for regeneration has been recently proposed for similar compounds.98

2.2 Perspectives on Preventive Antioxidants Since the first report99 by Sies et al. in 1984 that ebselen (33) could mimic the action of the GPx-enzymes, there has been an ongoing search for better catalysts for the thiol-induced reduction of hydroperoxides to water. Ebselen has been subjected to various structural modifications, either by incorpora- tion of new functional groups or by modification of the core structure.100 Other organoselenium compounds, such as diaryl diselenide 34,101 diaryl selenide 35102 and dialkyl selenide 36103 also show GPx-like activity.

According to the accepted mechanism for the action of ebselen, the Se-N bond is cleaved by glutathione (GSH) to form the selenenyl sulfide 37. Then, a second glutathione attacks on sulfur to yield a selenol 38, which is oxidized to selenenic acid 39 in the presence of peroxide. Finally, cyclization of the selenenic acid closes the catalytic cycle. Overall, two molecules of thiol are consumed in order to reduce one molecule of hydrogen peroxide to water (Scheme 12). Recently, a slightly revised catalytic cycle was proposed by Mugesh et al. involving a seleninic acid as intermediate.104 Depending on the nature of the peroxide and thiol used, the rate-limiting step of the cycle could vary.105

Scheme 12. Proposed mechanism for the catalytic action of ebselen.

The preventive antioxidant action of the GPx enzyme relies on facile redox cycling (oxidation / reduction) of selenium. Tellurium is even more readily oxidized than the lighter chalcogens. Indeed, several tellurium-containing GPx-mimics have been reported. In 1992, the tellurapyrylium dye 40 was described by Detty and co-workers.106 Later on, various diorganyl tellurides (e.g. 41)107 and diorganyl ditellurides (e.g. 42) were reported to catalyze hydroperoxide reduction.108

Te Te t-Bu t-Bu OH TePh PF6 Te C C C Te H H H t-Bu t-Bu 40 41 42

The catalytic cycles proposed for the tellurium-based compounds are similar to those for the corresponding organoseleniums. Redox cycling between Te(II)- and Te(IV)-species is shown in Scheme 13 for compound 41.

Scheme 13. Proposed mechanism for the catalytic GPx-activity of a diorganyl telluride.

2.3 Towards Multifunctional and Regenerable Antioxidants 2.3.1 Background There are obvious drawbacks with conventional chain-breaking antioxidants. As outlined above, chain-breaking phenols and aromatic amines convert peroxyl radicals by HAT and PCET to the corresponding hydroperoxides. Such species have to be removed by preventive antioxidants before they fall apart and initiate new chains. Another disadvantage with most of the phenolic antioxidants is that they act on a stoichiometric basis. They only quench two peroxyl radicals before they are gone and turned into waste. Clearly, this is not sustainable. Thus, there is a need to develop novel chain-breaking antioxidants that are multifunctional in the sense that they could quench both peroxyl radicals and hydroperoxides and regenerable in a manner that they quench multiple peroxyl radicals per molecule (n >> 2).

2.3.2 Multifunctional Antioxidants A multifunctional antioxidant would in principle result if one could introduce a H2O2-decomposing chalcogen functionality into a radical-trapping antioxidant. In 2001, Engman et. al. studied the antioxidant profile of a series of chalcogen-containing phenols 43.109 Later on, alkyltelluro-pyridinol- (44)110and BHA-analogues (45)111 were reported. It was found that these compounds are three-fold more reactive than α-tocopherol towards peroxyl

radicals in a two-phase lipid peroxidation system (vide infra). Also, the alkyltelluro-pyridinol showed a GPx-like activity 22-fold higher than recorded for diphenyl diselenide, which is a common reference compound.

The high reactivity of the alkyltelluro phenols and pyridinols towards peroxyl radicals was surprising, but it was not until 2013 that a mechanism was proposed (Scheme 14).112 Initially, an oxygen-atom is transferred from the peroxyl radical to the chalcogen. The resulting alkoxyl radical 46, in a solvent cage would then produce an alcohol ROH by HAT from the phenol to yield a telluroxide/phenoxyl radical species 47. Thus, the compounds are, at the same time, chain-breaking and hydroperoxide-decomposing.

Scheme 14. Proposed mechanism for the trapping of peroxyl radicals by ortho- alkyltelluro phenols.

2.3.3 Regeneration In biological membranes, α-tocopherol is known to be regenerable in the presence of ascorbate while the GPx-enzymes use GSH as a stoichiometric reducing agent to recycle. We were very pleased to find that alkyltelluro phenols are also regenerable by thiols in a simple two-phase system. The thiol has a dual role in the regeneration. It reduces the telluroxide to telluride and the phenoxyl radical to phenol (Scheme 15). In total, three equivalents of thiol are consumed per cycle to regenerate the alkyltelluro phenol. While the reduction of the telluroxide probably follows the same mechanism as for the GPx mimics, the regeneration of the phenol is less clear. It has been proposed to occur via PCET in a two-phase chlorobenzene / water system but a HAT is believed to be more favourable in homogeneous solution.112

OH Te R ROO H2O + 3/2 RSSR

3RSH

RO O O OH O Te Te R R

ROH Solvent cage

Scheme 15. Proposed catalytic cycle for the chain-breaking and hydroperoxide- decomposing action of ortho-alkyltelluro phenols.

The thiol/alkyltelluro phenol system would seem useful also in a biological setting where glutathione could serve as a stoichiometric reducing agent.

2.3.4 Aim of the work In this thesis, we designed, synthesized and evaluated a series of tellurium- based antioxidants for the following reasons:

To find novel compounds with improved antioxidative properties (Papers I – IV) Tocopherols and aromatic amines are known to be the most important antioxidants in Nature and for industrial applications. It seems therefore logical to incorporate tellurium into these compounds in order to see if the radical- trapping capacities as well as the regenerabilities could be further improved. Within the tocopherol series (Papers I – III), we have introduced alkyltelluro groups into β-, γ- and δ-tocopherols, and compared these novel antioxidants with their corresponding thio- and seleno- analogues. Furthermore, we also studied the antioxidative properties of a small series of bis(tocopheryl) tellurides. In the aromatic amine series (Paper IV), we prepared and evaluated a structurally varied library of alkyltelluro compounds, including primary anilines, N-alkyl anilines as well as diphenyl amines.

To understand which factors are important for improving the antioxidative properties (Papers V and Appendix) We learned from previous studies that ortho-substituted alkyltelluro phenols are better antioxidants than their corresponding para-susbtituted ones and

that high lipophilicity is essential for efficient regeneration. Both the tellurium and the hydroxyl group are important for the inhibition of peroxidation. Curious to see how the radical-trapping capacity and the regenerability is influenced by the arrangement and the number of the alkyltelluro- and hydroxyl substitutents in the phenol, we compared a series of alkyltelluro phenols, alkyltelluro resorcinols and bis(alkyltelluro) phenols (Paper V). How does the electron density at tellurium affect reactivity and regenerability of the antioxidant? We prepared a small library of aryltelluro phenols with electron-withdrawing and electron-donating groups attached para to the tellurium-atom in order to study this. Aryltelluro phenols with benzannulated cyclic ethers were included in order to increase the substituent effect of the oxygen (Appendix).

3. Synthesis

3.1 Synthesis of Chalcogen-Containing Tocopherol Antioxidants (Papers I – III) It occurred to us that it would be interesting to try to improve the radical- trapping capacity and the regenerability of tocopherol by introduction of alkyltelluro substituents into β-, γ- and δ-tocopherol. Also, for reference purposes, we decided to synthesize the corresponding selenium and sulfur analogues. δ-Tocopherol was the only tocopherol that could be purchased for a reasonable price. Bromination with Bu4NBr3 occurred selectively at position 5 and gave compound 48 in excellent yield. Initially, the phenol was THP- protected (49) before lithiation and addition of dioctyl ditelluride. The target compound 51b was obtained in reasonable yield (64%) by treatment of 50 with trifluoroacetic acid (Scheme 16).

Scheme 16. (a) Bu4NBr3, DCM, r.t. (b) 3,4-dihydropyran, cat. HCl, DCM, r.t. (c) (i) 2 eq. t-BuLi, THF, -78 °C; (ii) Dioctyl ditelluride; (iii) aq. NH4Cl. (d) TFA, DCM.

Later on, we found that the protection of the phenolic OH was unnecessary. Treatment of 48 with t-BuLi followed by addition of elemental tellurium and air-oxidation afforded the ditelluride 52. Subsequent reduction with sodium borohydride and alkylation of the in situ-generated tellurolate with butyl- and octyl bromide afforded 51a-b in fair yields (Scheme 17).

Scheme 17. (a) (i) 3 eq. t-BuLi, THF, -78°C; (ii) Te powder, r.t., overnight; (iii) Air- oxidation. (b) (i) NaBH4, EtOH; (ii) 1-Bromobutane / 1-Bromooctane.

β-Tocopherol (53) was needed and it was prepared from δ-tocopherol following a literature procedure by Netscher et al.113 Treatment of δ-tocopherol with the Mannich reagent obtained from paraformaldehyde and morpholine, followed by a reductive cleavage of the benzylamine (using NaBH4 and NaOH) provided 53 in good yield. Bromination of 53 afforded 54. By using a similar strategy as described in Scheme 17, bromophenol 54 was converted to alkyltelluro-substituted β-tocopherols 55a-b (Scheme 18).

Scheme 18. (a) Paraformaldehyde, moropholine, 80 °C. (b) NaOH, NaBH4, i-BuOH, 120 °C. (c) Bu4NBr3, DCM, r.t. (d) (i) 3 eq. t-BuLi, THF, -78 °C; (ii) Te powder, r.t., overnight; (iii) Air-oxidation; (iv) NaBH4, EtOH; (v) 1-Bromobutane / 1- Bromooctane.

On the other hand, the synthesis of 5-alkyltelluro substituted γ-tocopherols 59 was less straightforward (Scheme 19). γ-Tocopherol (12) was prepared by a 6-step synthetic route reported by Salvatori et al.114 Bromination of α- tocopherol, to give labile 56, followed by acetylation provided 57 in good yield. Stepwise oxidation of the bromomethyl group with NMMO and NaClO2 converted 56 to carboxylic acid 58. Hydrolysis under basic conditions followed by decarboxylation at elevated temperature afforded γ- tocopherol (12). Finally, the 5-butyltelluro- and octyltelluro substituted γ- tocopherols 59 were prepared in moderate yields by using the synthetic approach described in Scheme 17.

CH2Br CH2Br HO HO AcO a b O C H O C16H33 16 33 O C16H33 56 57 (74% for two steps)

CO H CHO CO2H 2 AcO AcO HO c d e O C H O C16H33 O C16H33 16 33

73% 58 (100%) 84% Br Te-R HO HO HO f g h

O C16H33 O C16H33 O C16H33

12 (85%) 100% 59a R = Butyl (60%) 59b R = Octyl (42%) Scheme 19. (a) Br2, hexane, r.t. (b) Ac2O, AcOH, DCM, cat. H2SO4, r.t., overnight. (c) NMMO, MeCN, r.t., overnight. (d) NH2SO3H, NaClO2, 4-dioxane/H2O, r.t. (e) 2 M KOH in MeOH, 50 °C. (f) 170°C, solvent-free, 3h. (g) Bu4NBr3, DCM, r.t. (h) (i) 3 eq. t-BuLi, THF, -78°C; (ii) Te powder, r.t., overnight; (iii) Air-oxidation; (iv) NaBH4, EtOH; (v) 1-Bromobutane / 1-Bromooctane.

We also wanted to introduce an alkyltelluro group into the 7-position of δ- tocopherol. However, the corresponding 7-bromo compound was not readily available. Hence, an approach involving dibromination, followed by selective debromination was proposed. Selective debromination of 60 in position 5 was achieved by in situ-generated Na2Te to afford 61 in moderate yield. Butyltelluro and octyltelluro δ-tocopherols 62 were obtained by our standard lithiation procedure (Scheme 20).

Scheme20. (a) Br2, DCM, reflux. (b) Na2Te, EtOH, reflux, 3 days. (c) (i) t-BuLi, THF, -78°C; (ii) Te powder, r.t., overnight; (iii) Air-oxidation. (iv) NaBH4, EtOH; (v) 1-Bromobutane / 1-bromooctane.

By replacing the elemental tellurium for selenium and sulfur in Schemes 18 and 20, the S- and Se-analogues 55c-f of β-tocopherol and the S- and Se-

analogues 62c-f of the 7-substituted δ- tocopherols were obtained in low to moderate yields (19-61%).

Electrophilic aromatic substitution was used to synthesize the remaining S- and Se-containing γ- and 5-substituted δ-tocopherols. δ-Tocopherol was converted to 51c-f in the presence of K2S2O8 and the corresponding dialkyl disulfide or dialkyl diselenide. Using the chemistry described in Scheme 18, compounds 51 were aminomethylated in position 7 (compound 63) and reduced to give 59 (Scheme 21). Yields were only rather low.

Scheme 21. (a) R2X2, K2S2O8, TFA, r.t. (b) Morpholine, paraformaldehyde, 80°C. (c) NaBH4, i-BuOH, 120°C.

We were also curious to see if we could link two tocopherols to the same tellurium atom. Lithium-halogen exchange in 48 with t-BuLi, followed by addition of TeCl4 afforded 64 in low yield after reductive workup (Scheme 22).

C16H33 O Br HO (i)-(iii) OH Te

O C16H33 HO 48 O C16H33

64 (24%) Scheme 22. (i) 3 eq. t-BuLi, THF, -78°C; (ii) 0.5 eq. TeCl4, r.t., overnight; (iii) aq. Na2SO3.

Similarly, compounds 65 and 66 were synthesized from 54 and 61, respectively. The isolated yields were again not impressive.

For reference purposes, we also prepared 67 and 68 by using a similar strategy.

3.2 Synthesis of Tellurium-Containing Aromatic Amine Antioxidants (Paper IV) We were curious to see if the radical-trapping capacity of aromatic amines could be improved by introduction of alkyltelluro substituents. Since the valence of the nitrogen is higher than for the oxygen, structural variations would be more easily achieved.

2-Bromoaniline was efficiently methylated by treatment with 1 eq. of n- BuLi followed by the addition of methyl iodide to give 69. Further treatment with t-BuLi, followed by addition of elemental tellurium powder and air- oxidation afforded the ditelluride 70. Finally, addition of NaBH4 and the appropriate alkyl bromide provided 71a-c in moderate yields (Scheme 23).

Scheme 23.(a) (i) 1 eq. n-BuLi, THF, -78 °C; (ii) MeI, r.t. (b) (i) 3 eq. t-BuLi, THF, - 78°C; (ii) Te powder, r.t., overnight; (iii) Air-oxidation; (iv) NaBH4, EtOH; (v) 1- Bromobutane / 1-Bromooctane / 1-Bromohexadecane.

By following the same procedure, compound 72 was prepared from 2- bromo-N-hexadecylaniline.

Diarylamines were accessed by condensation of 2-bromoaniline with 2- cyclohexen-1-one to provide 73 in moderate yield. By following the same approach as described in Scheme 23, 74a-b were obtained in reasonable yields (Scheme 24).

Scheme 24. (a) I2, p-TsOH, DMSO, 90 °C. (b) (i) 3 eq. t-BuLi, THF, -78°C; (ii) Te powder, r.t., overnight; (iii) Air-oxidation; (iv) NaBH4, EtOH; (v) 1-Bromobutane / 1-Bromohexadecane.

Diarylamines 75 and 76 were prepared in a similar way.

We also decided to incorporate an extra nitrogen atom into the aniline scaffold (Scheme 25). The commercially available 3-cyano-4,6-dimethyl-2- pyridinol was allowed to react with tetrabutylammonium bromide and phos- phorous pentoxide to afford 77, which was hydrolysed to 78 under acidic conditions. A subsequent Hoffman rearrangement converted the amide to amine 79 in good yield. Insertion of the octyltelluro functionality was carried out following the procedure in Scheme 23 to afford 80 in moderate yield. In order to access 82, N-methylated aminopyridine 81 was prepared prior to installation of the alkyltelluro moiety.

Scheme 25. (a) Bu4NBr, P2O5, toluene, reflux. (b) H2SO4, 100 °C. (c) Br2, KOH, H2O. (d) (i) 4 eq. t-BuLi, THF, -78°C; (ii) Te powder, r.t., overnight; (iii) Air- oxidation; (iv) NaBH4, EtOH; (v) 1-Bromooctane. (e) (i) n-BuLi, THF, -78 °C; (ii) MeI, r.t. (f) (i) 3 eq. t-BuLi, THF, -78°C; (ii) Te powder, r.t., overnight; (iii) Air- oxidation; (iv) NaBH4, EtOH; (v) 1-Bromooctane.

We also explored ortho-lithiation for the introduction of alkyltelluro substituents. Boc-protection of aniline was performed in the presence of NaH and Boc anhydride to give 83 in good yield. The N-Boc served as an ortho- directing group during lithiation of 83 with t-BuLi. Subsequent addition of finely ground tellurium powder, followed by air-oxidation caused the formation of ditelluride 84, which was treated with NaBH4 and an alkyl bromide to yield 85a-b. Deprotection using TFA afforded 86a-b (Scheme 26).

Scheme 26. (a) (i) NaH, THF; (ii) Boc2O, overnight. (b) (i) 2 eq. t-BuLi, THF, - 78°C; (ii) Te powder, r.t., overnight; (iii) Air-oxidation. (c) (i) NaBH4, EtOH; (ii) 1- Bromobutane / 1-Bromohexadecance. (d) TFA, DCM, r.t.

The chemistry described in Scheme 26 was applicable to the preparation of 87a-b using 1,2,3,4-tetrahydroquinoline as a starting material.

In order to investigate the effect of an alkyltelluro group placed further away from the N-H, we decided to synthesize the benzylic amine 89. Aniline was treated with 2-bromobenzaldehyde and the condensation product reduced with NaBH4 to afford 88. This served as a precursor of compound 89 via lithium-halogen exchange followed by addition of dibutyl ditelluride. (Scheme 27).

Scheme 27. (a) (i) 2-Bromobenzaldehyde, Et2O; (ii) NaBH4, MeOH. (b) (i) 3 eq. t- BuLi, THF, -78°C; (ii) Dibutyl ditelluride, r.t., overnight.

Curious to see if anilines carrying two alkyltelluro groups would show improved radical-trapping activity, we prepared a series of bis-alkyltelluro anilines. Sequential treatment of 1-bromo-2-(diethoxymethyl)benzene with t- BuLi and elemental tellurium powder followed by air-oxidation provided a ditelluride intermediate which was reduced with NaBH4 and the resulting tellurolate was alkylated to yield 90a-b. Under acidic conditions, the acetal group was transformed into the corresponding aldehyde to afford 91a-b in excellent yields. The following reductive amination afforded 92a-b in low to moderate yields (Scheme 28).

Scheme 28. (a) (i) 2 eq. t-BuLi, THF, -78°C; (ii) Te powder, r.t., overnight; (iii) Air- oxidation; (iv) NaBH4, EtOH; (v) 1-Bromobutane / 1-Bromohexadecane. (b) Conc. HCl (c) (i) 86a / 86b; (ii) NaBH3CN.

The N-Boc protected 2-bromoaniline served as a starting material for 94, carrying two aromatic amines flanked by octyltelluro groups. Treatment with NaH and 1,4-dibromobutane under reflux condition provided 93 which was reacted with t-BuLi and dioctyl ditelluride to obtain the target molecule (Scheme 29).

Scheme 29. (i) NaH, THF; (ii)1,4-Dibromobutane, reflux. (b) (i) 4 eq. t-BuLi, THF, - 78°C; (ii) Dioctyl ditelluride, r.t., overnight.

Bromination of 2-cyclohexen-1-one with bromine provided 95 in rather poor yield. By applying the synthetic strategy described in Scheme 24, 97 was obtained in low yield from the symmetrical diarylamine 96 (Scheme 30).

Scheme 30. (a) Br2, Et3N, DCM. (b) 2-Bromoaniline, I2, p-TsOH, DMSO, 90 °C. (c) (i) 5 eq. t-BuLi, THF, -78°C; (ii) Te powder, r.t., overnight; (iii) Air-oxidation. (iv) NaBH4, EtOH; (v) 1-Bromohexadecane.

In order to study the influence of electronic effects on the antioxidative capacity, we prepared 98a-c via lithium-halogen exchange and nucleophilic substitution using diphenyl ditellurides carrying 4-H, 4-OMe and 4-CF3 substituents as electrophiles. The isolated yields varied from 39 to 46% (Scheme 31).

Scheme 31. (a) (i) 3 eq. t-BuLi, THF, -78°C; (ii) Diphenyl ditelluride / Bis(4- methoxyphenyl) ditelluride / Bis(4-trifluoromethylphenyl) ditelluride.

To rule out the possibility that an N-alkoxylamine could be involved in the quenching of the peroxyl radical, we decided to synthesize 103. Treatment of 2-bromoaniline with n-BuLi followed by addition of 3-bromo-1-propene yielded 99 which was N-methylated in a similar fashion to afford 100. In the presence of m-CPBA, N-oxidation of 100 occurred, and the resulting N- oxide was rearranged to give 101. Hydrogenation with palladium on char- coal afforded 102 in moderate yield. Lithium-halogen exchange followed by addition of dioctyl ditelluride gave the target molecule 103 (Scheme 32).

Scheme 32. (a) (i) n-BuLi, THF, -78 °C; (ii) 3-Bromo-1-propene; (iii) aq. NH4Cl (b) (i) n-BuLi, THF, -78 °C; (ii) MeI. (c) m-CPBA, CHCl3. (d) Pd/C, H2, EtOAc. (e) (i) n-BuLi, THF, -78°C; (ii) Dioctyl ditelluride.

Also, for reference purposes, we prepared the N,N-dimethylated aniline 104 by methylation of 2-bromoaniline. Compound 104 was then lithiated with t- BuLi and treated with tellurium powder. Following air-oxidation, the resulting ditelluride was allowed to react with NaBH4 and 1-bromooctane to afford 105 (Scheme 33).

Scheme 33. (a) K2CO3, MeI, MeCN, reflux. (b) (i) 2 eq. t-BuLi, THF, -78°C; (ii) Te powder, r.t., overnight; (iii) Air-oxidation. (iv) NaBH4, EtOH; (v) 1-Bromooctane.

3.3 Synthesis of Phenolic Antioxidants: Alkyltelluro Phenols, Alkyltelluro Resorcinols and bis(Alkyltelluro) Phenols (Paper V) A series of alkyltelluro phenols, alkyltelluro resorcinols and bis(alkyltelluro) phenols were prepared to find out about the influence of the arrangement and the number of OH- and alkyltelluro groups on antioxidant capacity. A strategy involving lithium-halogen exchange followed by addition of the corresponding dialkyl ditelluride provided convenient access to the alkyltelluro phenols (Scheme 34). By treatment of 2-bromophenol with 3 eq. of t-BuLi followed by addition of the corresponding dialkyl ditelluride, 2- alkyltelluro phenols 106a-c were obtained in low yields (21 – 31%).

Scheme 34. (i) 3 eq. t-BuLi, THF, -78 °C; (ii) Dibutyl ditelluride / Dioctyl ditelluride / Dihexadecyl ditelluride; (iii) aq. NH4Cl.

The corresponding O-methylated analogues 107a-c (prepared for reference purposes) were obtained in better yields using the same strategy.

Alkyltelluro resorcinols 110a-c and the corresponding O-alkylated analogues 112a-c were prepared from resorcinol. Bromination afforded tri- bromoresorcinol (108) in excellent yield. This was in turn debrominated selectively with Na2SO3 to give 109. Lithium-halogen exchange with t-BuLi generated a trianion in situ, which was allowed to react with dibutyl-, dioctyl- and dihexadecyl ditelluride to afford the alkyltelluro resorcinols 110a-c, respectively. In order to increase the lipophilicity of 110a-c, we decided to O-alkylate one of the phenolic groups. Treatment of 109 with K2CO3 and 1- bromooctane gave the O-octylated 111 in low yield due to competing di-O- alkylation. By using the chemistry for the preparation of compounds 106, the O-octylated compound 111 served as the precursor for 112a-c (Scheme 35).

Scheme 35. (a) 3 eq. Br2, CHCl3, r.t. (b) Na2SO3. (c) (i) 4 eq. t-BuLi, THF, -78 °C; (ii) Dibutyl ditelluride / dioctyl ditelluride / dihexadecyl ditelluride; (iii) aq. NH4Cl. (d) K2CO3, 1-bromooctane, DMF, 80°C (e) (i) 3 eq. t-BuLi, THF, -78 °C; (ii) Dibu- tyl ditelluride / Dioctyl ditelluride / Dihexadecyl ditelluride. (iii) aq. NH4Cl.

As an alternative strategy to increase the lipophilicity, without affecting the two phenolic OHs, two methyl groups were introduced into the aromatic ring of the alkyltelluro resorcinol 110. Dibromination of 1,3-dimethoxybenzene with bromine afforded 113 in good yield. This was treated with t-BuLi followed by addition of methyl iodide to yield 114. De-O-methylation with BBr3, followed by treatment of 115 with perbromide afforded the resorcinol

analogue 116. Following the protocol in Scheme 34 with slight modifications, tellurides 117a-c were prepared in low yields (Scheme 36).

Scheme 36. (a) Br2, DCM, r.t. (b) (i) 4 eq. t-BuLi, THF, -78 °C; (ii) MeI. (c) BBr3, DCM, -78 °C. (d) Bu4NBr3, DCM, r.t.; (e) (i) t-BuLi, THF, -78 °C; (ii) Dibutyl ditelluride / Dioctyl ditelluride / Dihexadecyl ditelluride; (iii) aq. NH4Cl.

Bis(alkyltelluro) phenols 119 were accessed by selective ortho-bromination of phenol in the presence of t-BuNH2 followed by a lithium-halogen exchange in 118 and addition of the appropriate dialkyl ditelluride (Scheme 37).

Scheme 37. (a) Br2, t-BuNH2, toluene, DCM. (b) (i) 5 eq. t-BuLi, THF, -78°C; (ii) Dibutyl ditelluride / Dioctyl ditelluride / Dihexadecyl ditelluride. (iii) aq. NH4Cl.

3.4 Synthesis of Aryltelluro Phenol Antioxidants (Appendix) To study the influence of electronic effects on the radical-trapping capacity and regenerability of tellurium-containing phenols, a series of compounds carrying phenyltelluro groups with electron donating and electron withdrawing groups in position 4 were prepared. Initially, we tried to introduce aryltelluro groups by treatment of 2- bromophenol with t-BuLi followed by addition of diphenyl ditelluride, bis(4- methoxyphenyl) ditelluride and bis(4-trifluoromethylphenyl) ditelluride. However, the yields of 120a-c were unsatisfactory (<15%). Later, we were pleased to find a better strategy involving THP-protection of the starting

material prior to lithiation (Scheme 38). Finally, THP-deprotection with para-toluenesulfonic acid gave 120a-c in moderate to good yields (42 – 70%).

Scheme 38. (a) 2,3-dihydropyran, pyridinium p-toluenesulfonate, DCM. (b) (i) 2 eq. t-BuLi, THF, -78°C; (ii) Diphenyl ditelluride / Bis(4-methoxyphenyl) ditelluride / Bis(4-trifluoromethylphenyl) ditelluride; (iii) aq. NH4Cl. (c) TsOH, DCM/MeOH (1:1).

Inspired by Ingold (p 16), compounds 121 and 122a-b were prepared following the same approach as described in Scheme 38. Due to a stereoelectronic effect, oxygen would be expected to be more strongly electron-donating in a benzannulated tetrahydropyran or tetrahydrofuran than in a free-rotating methoxy group. On the other hand, an opposite effect was expected when two methyl groups were installed next to the methoxy group since they force it to adopt a conformation where the oxygen p-orbitals are roughly perpen- dicular to the aromatic π-orbitals.

The ditelluride 124 was prepared in good yield (89%) by bromination of 2- methoxy-1,3-dimethyl benzene followed by lithium-halogen exchange, addition of tellurium powder and air oxidation (Scheme 39).

Scheme 39. (a) NBS, MeCN. (b) (i) 2 eq. t-BuLi, THF, -78°C; (ii) Te powder, r.t., overnight; (iii) Air-oxidation.

In a similar way, diaryl ditelluride 125 was synthesized in moderate yield (67%).

The diaryl ditelluride 126 required for the preparation of 122b was synthesized by a modified Clemmensen reduction of chroman-4-one followed by treatment with TeCl4 at 120 °C and reduction with sodium borohydride (Scheme 40).

Scheme 40. (a) Zn, AcOH, 100 °C. (b) (i) TeCl4, 120 °C; (ii) NaBH4, EtOH.

For reference purposes, we also prepared 128 via a 3-step synthetic route. Treatment of 5-bromo-2,3-dihydrobenzofuran with magnesium afforded a Grignard reagent which in turn was allowed to react with trimethylborate. Hydrolysis, followed by addition of peracetic acid triggered a rearrangement to phenol 127. Bromination of 127 followed by treatment as described in Scheme 38, gave 128 in low yield (Scheme 41).

Scheme 41. (a) (i) Mg, I2, THF; (ii) B(OCH3)3, THF; (iii) 2 M HCl; (iv) AcOOH, Et2O. (b) Bu4NBr3, DCM. (c) See Scheme 38.

In a similar way, another reference compound 129 was synthesized using 2-bromo-4-methoxyphenol as a starting material.

4. Evaluation

4.1 Antioxidant Profile High reactivity towards peroxyl radicals and hydrogen peroxide is essential for the novel multifunctional antioxidants that we develop. A two-phase lipid peroxidation model was used to investigate the radical-trapping capacity and the regenerability of the novel antioxidants whereas the coupled reductase assay and the thiol peroxidase assay were employed for studying the catalytic hydroperoxide-decomposing properties. Furthermore, a thiol concentration assay was developed to monitor the consumption of thiol in the aqueous phase of the two-phase lipid peroxidation model.

4.2 Evaluation of the Radical-Trapping Capacity and Regenerability 4.2.1 The Two-Phase Lipid Peroxidation Model The regenerability and the radical-trapping efficiency of the novel antioxidants were assessed in a rapidly stirred two-phase (water and chlorobenzene) lipid peroxidation system, which is reminiscent of a biological membrane (Figure 6).115 The aqueous phase contains N-acetylcysteine as a co- antioxidant. In the chlorobenzene, linoleic acid is present as an oxidizable substrate, together with the antioxidant to be evaluated. AMVN was added to the rapidly stirred two-phase system at 42 °C in order to initiate the peroxidation of linoleic acid. The lipid layer is sampled every 10 minutes and analyzed by HPLC for the formation of conjugated diene (UV-absorption at 234 nm). The concentration of the conjugated diene was then plotted against time. A representa- tive peroxidation plot using α-tocopherol as an antioxidant is shown in Fig- ure 7. As could be seen, peroxidation is efficiently inhibited for ca. 100 min. Then, it proceeds with a higher, uninhibited rate.

Figure 6. The two-phase lipid peroxidation model.

Two parameters, Rinh and Tinh, were determined in order to compare the reactivity and regenrability of antioxidants tested. Rinh is defined as the slope of the line during the inhibited period. A small value of Rinh indicates efficient quenching of peroxyl radicals. Eventually, when all the antioxidant is consumed, the Rinh-value increases to the uninhibited rate.

4000 3500 α-Tocopherol 3000 2500 2000 1500 1000 500

Conjugated Diene (µM) 0 0 100 200 300 400 500 Time (min)

Figure 7. Peroxidation plot recorded for α-tocopherol.

Tinh refers to the time period that the peroxidation is inhibited. It is recorded as the intersection of the lines corresponding to inhibited and uninhibited peroxidation. This parameter reflects the regenerability of the antioxidant. The novel antioxidants were evaluated in the presence/absence of aqueous-phase NAC. α-Tocopherol was used as a reference compound. As re- flected by the similar Tinh- and Rinh-values obtained with and without NAC, α-tocopherol is not regenerable in the two-phase model.

4.2.2 Evaluation of Chalcogen-Containing Tocopherol Antioxidants (Papers I – III) A series of tocopherols carrying alkylthio-, alkylseleno- and alkyltelluro groups in the ortho-position were evaluated in the two-phase lipid peroxidation model (Table 1). α-, β-, γ- and δ-Tocopherol were used as references. The similar Rinh- and Tinh-values obtained for the four tocopherols in the presence/absence of NAC suggest that they are not regenerable in our assay. Whether NAC is present or not, the alkylthio (51c, 51d, 55c, 55d, 59c, 59d, 62c and 62d) and alkylseleno (51e, 51f, 55e, 55f, 59e, 59f, 62e and 62f) tocopherols were unimpres- sive when it comes to radical-trapping efficiency and inhibition time. It may be that the oxygen-transfer mechanism is not operative with sulfur compounds and less efficient with organoseleniums. Therefore, quenching of peroxyl radicals occurs predominately by HAT. Hydrogen-bonding of the OH group to the chalcogen would increase the BDEO-H. However, the effect is rather small and this is probably not the main reason for the rather high -1 peroxidation rate (34 < Rinh < 134 μM h ) observed. Also, the steric hindrance from the alkylthio- and alkylseleno-substituents in the ortho-position would slow down the HAT. The similar inhibition times recorded for these compounds with and without NAC would also suggest that they are not regenerable in our assay. The antioxidant profiles for the alkyltelluro tocopherols (51e, 51f, 55e, 55f, 59e, 59f, 62e and 62f) are significantly different. In the absence of -1 NAC, the tellurium derivatives of the tocopherols (Rinh = 4 – 48 μM h ) clearly outperformed the corresponding alkylthio- and alkylseleno-analogues with regard to the radical-trapping capacity. In most of the cases, they trap radicals as efficiently as their parent tocopherols. Interestingly, compound 59a and 59b were much more reactive than δ-tocopherol. Tocopherols are known to trap two peroxyl radicals before they turn into non-radical products, i.e. n = 2.0. The inhibition times (96 – 174 min) recorded for the alkyltelluro tocopherols in the absence of NAC suggested that their stoichiometric numbers are similar or higher (55b, 62a and 62b). In contrast, the alkylthio- and alkylseleno analogues showed n-values in the range of 0.5 – 1.5. -1 In the presence of NAC, Rinh decreased (0.7 < Rinh < 5.0 μM h ) significantly as compared to the same experiment without thiol. Much longer inhibition times (342 – 630 min) were also recorded. Alkyltelluro tocopherol 62b, the best antioxidant in the series, showed a ca. 18-fold lower Rinh and a ca. seven-fold longer Tinh than recorded for δ-tocopherol. The antioxidative capacity of the rest of the organotelluriums decreased as follows: 55 > 59 > 51. Furthermore, in all series the more lipophilic octyltelluro compound inhibited peroxidation for longer time than the corresponding butyl analogue.

Table 1. Inhibited Rates of Conjugated Diene Formation (Rinh) and Inhibition Times (Tinh) in the Presence and Absence of NAC (1 mM). With NAC Without NAC

a b a b Antioxidant Rinh Tinh Rinh Tinh -1 -1 (40 μM) (µM h ) (min) (µM h ) (min)

51a R = Te-Butyl 3 ± 1 342 ± 7 48 96 R 51b R = Te-Octyl 1.3 ± 0 458 ± 6 34 135 HO 51c R = S-Butyl 75 ± 3 59 ± 1 80 68 51d R = S-Octyl 68 ± 4 71 ± 4 82 74 O C16H33 51e R = Se-Butyl 77 ± 1 54 ± 5 106 55 51f R = Se-Octyl 53 ± 2 91 ± 2 62 74

55a R = Te-Butyl 1.6 ± 0 504 ± 10 16 115 55b R = Te-Octyl 5.0 ± 1 591 ± 10 39 175 55c R = S-Butyl 89 ± 3 35 ± 1 74 19 55d R = S-Octyl 94 ± 3 33 ± 1 93 33 55e R = Se-Butyl 77 ± 5 67 ± 2 59 96 55f R = Se-Octyl 104 ± 6 25 ± 1 80 116

59a R = Te-Butyl 1.0 ± 0 435 ± 7 4.2 96 59b R = Te-Octyl 2.2 ± 1.2 495 ± 6 4.0 116 59c R = S-Butyl 104 ± 6 53 ± 1 131 39 59d R = S-Octyl 82 ± 6 55 ± 2 109 47 59e R = Se-Butyl 131 ± 4 41 ± 1 78 44 59f R = Se-Octyl 59 ± 6 40 ± 7 69 29

62a R = Te-Butyl 0.7 ± 0 509 ± 7 15 170 62b R = Te-Octyl 1.4 ± 1 630 ± 8 26 174 62c R = S-Butyl 83 ± 2 45 ± 2 114 62 62d R = S-Octyl 93 ± 3 52 ± 2 66 47 62e R = Se-Butyl 34 ± 3 85 ± 3 45 138 62f R = Se-Octyl 111 ± 5 52 ± 3 37 122

α-tocopherol 25 ± 1 97 ± 5 28 ± 2 109 ± 2 β-tocopherol 28 ± 1 105 ± 3 28 ± 1 119 ± 2 γ-tocopherol 31 ± 1 104 ± 3 27 ± 3 119 ± 3 δ-tocopherol 30 ± 1 85 ± 3 33 ± 4 93 ± 3 a Rate of peroxidation during the inhibited phase (uninhibited rate ca. 425 µM h-1). Errors correspond to ± SD for triplicates. b Duration of the inhibited phase of peroxidation. Reactions were monitored for 680 min. Errors correspond to ± SD for triplicates.

A series of bis-tocopheryl tellurides 64-66 were also evaluated in the two- phase lipid peroxidation model (Table 2). They turned out to be excellent

-1 radical-quenchers (1 < Rinh < 4 µM h ) with a better regenerability (Tinh = 635 – 742 min) in the presence of NAC than those recorded for the alkyltelluro toocopherols. Among them, 66 inhibited peroxidation for ca. 8-fold longer than δ-tocopherol. When the phytyl chain as well as the methyls in positions 2 and 8 were removed (68), the Tinh decreased to 544 min. This could be a lipophilicity effect. Further simplification (removal of the two remaining phenolic substituents; 67) shortened the inhibition time to 362 min.

Table 2. Inhibited Rates of Conjugated Diene Formation (Rinh) and Inhibition Times (Tinh) in the Presence and Absence of NAC (1 mM) in the Two-Phase System With NAC Without NAC

a b a b Antioxidant Rinh Tinh Rinh Tinh (40 μM) (µM h-1) (min) (µM h- (min) 1 )

Ar-Te-Ar

64 2 ± 1 651 ± 10 8 170

65 2 ± 0 635 ± 12 7 165

66 4 ± 1 742 ± 13 21 214

67 2 ± 1 362 ± 7 50 11

68 1 ± 1 544 ± 4 25 92

25 ± 1 97 ± 5 28 ± 2 109 ± 2 α-Tocopherol aRate of peroxidation during the inhibited phase (uninhibited rate ca. 479 µM h-1). Errors correspond to ± SD for triplicates. bInhibited phase of peroxidation. Reactions were monitored for 680 min. Errors correspond to ± SD for triplicates.

4.2.3 Evaluation of Tellurium-Containing Aromatic Amine Antioxidants (Paper IV) The alkyltelluro substituted aromatic amines prepared were also assayed in the two-phase lipid peroxidation model (Tables 3 and 4). In the absence of NAC, the lipid peroxidation was almost not inhibited. Residual amount of linoleic acid hydroperoxide that is always present in the commercial sample of linoleic acid is likely to oxidize tellurides to the corresponding telluroxides. It is well-known that telluroxides are much poorer chain-breaking antioxidants than the corresponding tellurides.110b Indeed, the Rinh-values obtained for the alkyltelluro compounds were often smaller than -1 recorded (Rinh = 479 µM h ) in the control experiment. This suggests that the peroxidation is retarded by the antioxidant. As seen above, thiol could change the situation to the better. Thus, the li- -1 pid peroxidation was effectively inhibited (Rinh = 2 – 9 µM h ) when NAC was added. Also, longer Tinh-values were recorded as a result of antioxidant- recycling. Among aryl amines, the alkyltelluro anilines 86 and their pyridine analogues 80 and 82 showed inhibition times (87 – 137 min) much shorter than recorded for the phenols. N-alkylated anilines (71a-c, 72 and 87a-b) carrying alkyltelluro substituents performed better (Tinh = 148 – 342 min) but the longest inhibition times were recorded with diphenyl amines. Compound 74b inhibited the peroxidation for 461 min or four-fold longer than α- tocopherol. There is a trend that Tinh for the aromatic amines increases as the alkyl chain attached to tellurium is prolonged (compounds 86 and 87). The reason for this is still not clear. We hypothesize that the more lipophilic antioxidants are entirely distributed into the chlorobenzene phase where the quenching of peroxyl radicals is taking place. Also, the longer alkyl chain may provide extra stabilization and prevent leakage of alkoxyl radicals from the solvent cage. The distance between the N-H and the alkyltelluro moiety seems to be an important factor for efficient regeneration. Indeed, the ortho-alkyltelluro substituted aromatic amine 74b (Tinh = 461 min) inhibited the lipid peroxidation for ca. three-fold longer than the corresponding para-substituted analogue 76 (Tinh = 152 min). Similarly, the N-benzyl aniline 89 (Tinh = 132 min) was less regenerable than 74a (Tinh = 237 min). Previously, we observed better regenerability for phenols carrying two alkyltelluro groups. However, this was not the case with aromatic amines. The dialkyltelluro-functionalized aromatic amines (92, 94 and 97; Table 4) often showed shorter inhibition times than recorded for the corresponding parent compounds (94 vs 71b and 97 vs 74b).

Table 3. Inhibited Rates of Conjugated Diene Formation (Rinh) and Inhibition Times (Tinh) in the Presence and Absence of NAC (1 mM) in the Two-Phase System With NAC Without NAC a b a b Antioxidant Rinh Tinh Rinh Tinh (40 μM) (μM h-1) (min) (μM h-1) (min) 71a R1 = Me, R2 = Butyl 5 ± 1 204 ± 5 393 0 71b R1 = Me, R2 = Octyl 5 ± 1 218 ± 7 449 0 71c R1 = Me, R2 = Hexadecyl 3 ± 1 219 ± 4 434 0

72 R1 = R2 = Hexadecyl 7 ± 2 300 ± 6 544 0

86a R1 = H, R2 = Butyl 4 ± 1 87 ± 7 533 0

86b R1 = H, R2 = Hexadecyl 4 ± 2 137 ± 10 444 0

98a R1 = R2 = Ph 7 ± 1 346 ± 9 348 0

98b R1 = Ph, R2 = C6H4-4-OMe 7 ± 3 410 ± 11 441 0

98c R1 = Ph, R2 = C6H4 -4-CF3 44 ± 7 279 ± 6 311 0

74a R1 = H, R2 = Butyl 2 ± 1 237 ± 1 401 0

74b R1 = Hexadecyl, R2 = H 2 ± 1 461 ± 10 393 0 75 R = Hexadecyl, R = OMe 2 ± 1 444 ± 6 356 0 1 2

76 R = Te-Hexadecyl 6 ± 1 152 ± 2 290 0

80 R = H 6 ± 2 95 ± 5 466 0

82 R = Me 4 ± 0 106 ± 0 532 0

87a R = Butyl 4 ± 1 148 ± 1 385 0

87b R = Hexadecyl 2 ± 1 342 ± 4 235 0

89 4 ± 1 132 ± 2 247 0

103 R = O-Pr 27 ± 5 134 ± 8 198 0

105 R = Me 90 ± 4 54 ± 5 470 0

α-Tocopherol 25 ± 1 97 ± 5 28 ± 2 109 ± 2 aRate of peroxidation during the inhibited phase (uninhibited rate ca. 479 µM h-1). Errors correspond to ± SD for triplicates. bInhibited phase of peroxidation. Reactions were monitored for 680 min. Errors correspond to ± SD for triplicates.

The results with diphenyl amines 98a-c suggest that a higher electron density at tellurium serves to improve regenerability. Also, an electron-withdrawing 4-CF3-C6H4 group (98c) caused a substantial increase in the inhibited rate of peroxidation (Rinh was 6-fold higher than recorded for the parent compound 98a). It could be that this compound quenches peroxyl radical primarily by HAT. For compounds such as 103 and 105, lacking an N-H, inhibition times were rather short (Tinh = 54 and 134 min, respectively) and the inhibited rates of peroxidation were among the highest recorded in the series.

Table 4. Inhibited Rates of Conjugated Diene Formation (Rinh) and Inhibition Times (Tinh) in the Presence and Absence of NAC (1 mM) in the Two-Phase System With NAC Without NAC a b a b Antioxidant Rinh Tinh Rinh Tinh (40 μM) (μM h-1) (min) (μM h-1) (min)

92a R = Butyl 8 ± 3 135 ± 11 254 0

92b R = Hexadecyl 3 ± 1 147 ± 4 269 0

94 9 ± 3 159 ± 11 460 0

R R H N 97 R = Hexadecyl 3 ± 0 420 ± 7 243 0

4.2.4 Evaluation of Alkyltelluro Phenols (Paper V) A series of alkyltelluro phenols, alkyltelluro resorcinols and bis(alkyltelluro) phenols were evaluated in the two-phase lipid peroxidation model (Table 5). In the absence of NAC, most of the compounds were poor inhibitors (Rinh > 100 μM h-1) of the lipid peroxidation. This is again because the organotelluriums are rapidly oxidized by the trace-amounts of linoleic acid hydroperoxide that is always present in the commercially available linoleic acid.

Table 5. Inhibited Rates of Conjugated Diene Formation (Rinh) and Inhibition Times (Tinh) in the Presence and Absence of NAC (1 mM) in the Two-Phase Model System With NAC Without NAC

a b a b Antioxidant Rinh Tinh Rinh Tinh (40 μM) (µM h-1) (min) (µM h-1) (min)

106a R = Butyl 5.4 ± 1.2 228 ± 3 508 0 106b R = Octyl 2.0 ± 0.8 220 ± 10 444 0 106c R = Hexadecyl 2.3 ± 0.9 223 ± 5 441 0

107a R = Butyl 66 ± 4 87 ± 7 446 0 107b R = Octyl 64 ± 5 94 ± 6 461 0 107c R = Hexadecyl 53 ± 5 99 ± 5 466 0

110a R = Butyl 3.0 ± 0.2 92 ± 8 155 50 110b R = Octyl 2.1 ± 1.2 90 ± 3 166 50 110c R = Hexadecyl 2.5 ± 0.8 90 ± 1 128 50

112a R = Butyl 1.9 ± 0.8 198 ± 1 449 0 112b R = Octyl 2.4 ± 1.1 242 ± 6 442 0 112c R = Hexadecyl 1.7 ± 0.5 235 ± 7 437 0

117a R = Butyl 4.7 ± 0.6 89 ± 5 85 90 117b R = Octyl 3.0 ± 1.1 98 ± 2 100 130 117c R = Hexadecyl 3.0 ± 0.5 99 ± 3 97 140

119a R = Butyl 2.5 ± 0.5 396 ± 12 396 0 119b R = Octyl 1.6 ± 0.7 > 410 422 0 119c R = Hexadecyl 3.6 ± 1.2 > 410 411 0

α-Tocopherol 25 ± 1 97 ± 5 28 ± 2 109 ± 2 aRate of peroxidation during the inhibited phase (uninhibited rate ca. 650 µMh-1). Errors correspond to ±SD for triplicates. bInhibited phase of peroxidation. Reactions were monitored for 410 min. Errors correspond to ± SD for triplicates.

In the presence of NAC, the lipid peroxidation was effectively inhibited (Rinh = 2 – 5 μM h-1) by the alkyltelluro phenol derivatives. Indeed, most of them -1 outperformed α-tocopherol (Rinh = 24 μM h ) when it comes to radical-

trapping capacity. The anisole analogues 107a-c inhibited the peroxidation less effectively (53 – 66 μM h-1) since the HAT is not possible. We hypothesize that peroxyl radicals are quenched by a mechanism involving electron transfer from the tellurium. Larger variations in the inhibition times were seen for the alkyltelluro phenols. The simple ortho-substituted phenols 106a-c inhibited peroxidation for 220 – 228 min whereas the inhibition for resorcinol derivatives 110a-c and 117a-c only lasted for 89 – 99 min. Since an increase in the length of the aliphatic chain (from butyl to hexadecyl) did not significantly increase the inhibition time, lipophilicity is not a critical parameter for regeneration of these compounds. Regeneration probably occurs at the interphase between the aqueous and the chlorobenzene layer. We speculate that the 1,3- disubstituted benzene part of the compound is buried into the aqueous phase, thus, the quenching of peroxyl radicals in the chlorobenzene layer becomes less effective. Indeed, when one of the hydroxyl groups were O-alkylated, the inhibition time increased (112a-c, Tinh = 198 – 242 min). The bis(alkyltelluro) phenols 119a-c were the most regenerable compounds in this series with inhibition times exceeding 400 min. This could be the result of a statistical or a lipophilicity effect.

4.2.5 Evaluation of Aryltelluro Phenol Antioxidants (Appendix) Whereas some compounds (120b, 121, 122 and 129) without NAC in the -1 aqueous phase inefficiently (38 < Rinh< 77 µM h ) inhibited peroxidation for -1 short times (14 < Tinh < 74 min) compound 128 (Rinh = 8 µM h ) turned out to be a more efficient radical-quencher than α-tocopherol. The rather long inhibition period (Tinh = 131 min) suggests that it would have a stoichiometric factor n > 2 (Table 6). In the presence of NAC, all compounds outperformed α-tocopherol when it comes to reactivity and regenerability. The Rinh-values for 122a and 128 were only a fraction (< 2%) of those recorded with α-tocopherol. Com- pounds 122a and 128 inhibited the lipid peroxidation for 571 min and 609 min, respectively. Overall, a high electron density at the tellurium seems to improve both the radical-trapping capacity and regenerability. This is clearly seen for compounds 120 and 122a (122a > 120b > 120a > 120c). An electron- withdrawing group has the opposite effect. The electron-donating effect of the alkoxy group could be further increased by minimizing the dihedral angle between the p-orbital of the oxygen atom and the aromatic π-system. This would explain the better antioxidant characteristics for 122a as compared to 120 and 121. A similar effect was also observed when electron- donating groups were introduced para to the phenolic O-H. As compared to 120a, the inhibition times for 128 and 129 were prolonged by 284 min and

78 min, respectively, and the inhibited rates of peroxidation were substan- tially reduced.

Table 6. Inhibited Rates of Conjugated Diene Formation (Rinh) and Inhibition Times (Tinh) in the Presence and Absence of NAC (1 mM) in the Two-Phase System With NAC Without NAC a b a b Antioxidant Rinh Tinh Rinh Tinh (40 μM) (μM h-1) (min) (μM h-1) (min)

120a 4 ± 1 325 ± 8 556 0

120b 1 ± 0.6 379 ± 3 69 42

120c 15 ± 4 310 ± 9 565 0

121 5 ± 1 349 ± 9 633 0

122a 0.4 ± 0 571 ± 3 77 34

122b 1 ± 1 363 ± 10 69 14

128 0.3 ± 0.3 609 ± 9 8 131

129 1 ± 0.6 403 ± 6 38 74

4.2.6 Thiol Consumption Assay We also felt it would be interesting to study the NAC-consumption during a peroxidation experiment. The aqueous phase of the two-phase lipid peroxidation system was therefore sampled every 30 minutes and the remaining thiol was allowed to react with bis-4-pyridyl disulfide (Aldrithiol-4TM). The concentration of pyridine-4-thiol was determined spectroscopically at 324 nm and used as a measure of the NAC-concentration (Scheme 42).

TM Scheme 42. NAC reacts with Aldrithiol-4 to form pyridine-4-thiol.

The rate of thiol consumption was obtained by plotting the concentration of NAC against time. The consumption of NAC was low (37 µM h-1) in a control experiment in the two-phase system.

1000

800 ) M

μ 600

NAC ( 400

200

0 0 50 100 150 200 Time (min)

Figure 8. NAC-consumption during a peroxidation experiment with α-tocopherol as an antioxidant. After 140 minutes, telluride 105 was added.

Indeed, the thiol was consumed at a similarly low rate when non-regenerable α-tocopherol was present. However, addition after some time (140 min) of an antioxidant carrying a tellurium substituent (compound 105) resulted in a drastic increase in the NAC-consumption (Figure 8). Hydroperoxides formed as a result of AMVN-decomposition are likely to react slowly with NAC. However, organotelluriums can mimic the action of the GPx-enzymes and catalyze this process. The rapid consumption of NAC after 140 min is likely due to catalyzed decomposition of alkylhydroperoxide that has accumulated in the system (Figure 8).

4.2.7 NAC-consumption during Peroxidation Inhibited by Tellurium-Based Antioxidants In general, the rate of NAC-consumption was found to be inversely related to the inhibition time recorded in the two phase lipid peroxidation assay (Ta- bles 7, 8 and 9).

Table 7. NAC-Consumption in the Aqueous Phase during Peroxidation Inhibited by Symmetrical Diaryl Tellurides Antioxidant Rate of NAC-Consumption (40 µM) (µM h-1)a

Ar-Te-Ar

64 131 ± 4

65 125 ± 10

119 ± 13 66

67 164 ± 4

68 150 ± 10

α-Tocopherol 33 ± 4 a Errors correspond to ± SD for triplicates. The NAC-consumption rate was 37 ± 8 µM h-1 in the absence of antioxidants and 27 ± 5 µM h-1 with only NAC in the aqueous phase.

In the bis-tocopherol series, the NAC-consumption with compounds 64-66 was low (119 – 131 µM h-1). Indeed, the consumption rates were lower than recorded for most of the aryltelluro phenols (Table 7) and alkyltelluro substituted aromatic amines (Table 8). We hypothesize that the low consumption of NAC is the result of efficient quenching of alkoxyl radicals in the solvent cage. This will minimize the initiation of new chains and spare the thiol re-

ducing agent which would otherwise be consumed unnecessarily to reduce telluroxide to the corresponding telluride. Diaryl tellurides 67 and 68 consumed NAC at a slightly higher rate.

Table 8. NAC-Consumption in the Aqueous Phase during Peroxidation Inhibited by Aryltelluro Phenols Antioxidant Rate of NAC-Consumption (40 µM) (µM h-1)a

120a R = H 153 ± 4 120b R = OMe 153 ± 4

120c R = CF3 155 ± 2

121 152 ± 11

122a n = 1 146 ± 12

122b n = 2 149 ± 3

129 144 ± 2

128 104 ± 10

α-Tocopherol 33 ± 4 a Errors correspond to ± SD for triplicates. The NAC-consumption rate was 37 ± 8 µM h-1 in the absence of antioxidants and 27 ± 5 µM h-1 with only NAC in the aqueous phase.

Except for compound 128 (104 μM h-1), most of the diaryl tellurides consumed the NAC at a rate around 150 µM h-1 (Table 8). This is in accord with the long inhibition time recorded in the two-phase system. So far, this is the lowest value we have recorded in this assay. As compared to 122b, carrying

a six-membered benzannulated cyclic ester, the five-membered analogue 122a showed a slightly slower rate of NAC-consumption.

Table 9. NAC-Consumption in the Aqueous Phase During Peroxidation Inhibited by Tellurium-Containing Aromatic Amines Antioxidant Rate of NAC- (40 µM) Consumption (µM h-1)a

71b R1 = H , R2 = Me, R3 = Te-Octyl 222 ± 18

86b R1 = R2 = H, R3 = Te-Hexadecyl 458 ± 27

103 R1 = OPr, R2 = Me, R3 = Te-Octyl 500 ± 16

105 R1 = Me, R2 = Me, R3 = Te-Octyl 538 ± 26

74b R1 = Te-Hexadecyl, R2 = R3 = H 146 ± 13

75 R1 = Te-Hexadecyl, R2 = H, R3 = OMe 150 ± 14

76 R1 = R3 = H, R2 = Te-Hexadecyl 321 ± 22

98a R1 = Te-Ph, R2 = R3 = H 160 ± 5 98b R1 = Te-C6H4-OMe, R2 = R3 = H 126 ± 10 98c R1 = Te-C6H4-CF3, R2 = H 188 ± 8

82 736 ± 18

87b 146 ± 13

89 R1 = Te-Butyl, R2 = H 586 ± 21

94 305 ± 14

97 R = Te-Hexadecyl 187 ± 8

α-Tocopherol 33 ± 4 a Errors correspond to ± SD for triplicates. The NAC-consumption rate was 37 ± 8 µM h-1 in the absence of antioxidants and 27 ± 5 µM h-1 with only NAC in the aqueous phase.

Among the alkyltelluro aromatic amines, pyridine 82 (736 µM h-1), aniline 86b (458 µM h-1) and benzyl aniline 89 (586 µM h-1) consumed the NAC very rapidly (Table 9). None of these compounds could inhibit peroxidation for longer than 140 min. In contrast, compounds 74b, 75 and 98b that inhibited the lipid peroxidation for more than 400 minutes consumed NAC at relatively low rates (≤ 150 µM h-1). In all cases the time corresponding to depletion of the thiol fell short of Tinh. This suggests that the availability of the thiol reducing agent would be the limiting factor for the inhibition time. Interestingly, compounds with two alkyltelluro substituents showed a higher rate of thiol-consumption than the mono-functionalized aromatic amines. Presumably, the di-functionalized catalyst cannot cope with all alkoxyl radicals generated and there is a considerable leakage from the solvent cage.

4.3 Catalytic Decomposition of Hydrogen Peroxide 4.3.1 The Coupled Reductase Assay Organotellurium compounds are known since long to mimic the GPx- enzymes. To compare the GPx-activity of novel tellurium-containing antioxidants, the coupled reductase assay116 was used. The assay was carried out in a buffered solution at pH 7.5 containing hydrogen peroxide, glutathione, glutathione reductase and the compound to be investigated.

Scheme 43. The coupled reductase assay.

In the presence of hydrogen peroxide, a Te (II)-compound is oxidized to a Te (IV)-species which in turn is reduced by glutathione (Scheme 43). The reduction of the resulting disulfide (GSSG) is catalysed by glutathione reductase and brought about by NADPH. The consumption of NADPH could be monitored by UV-spectroscopy at 340 nm. Ebselen (33), capable of reducing hydrogen peroxide at a rate of 3 μM / min, was used as a reference.

4.3.2 The Thiol Peroxidase Assay Tomoda and co-workers117 reported an assay for GPx-activity where the initial formation of diphenyl disulfide was monitored spectroscopically at 304 nm in methanol containing hydrogen peroxide, the catalyst and thiophe- nol (Scheme 44). For this assay, diphenyl diselenide was used as a reference compound.

Scheme 44. The thiol peroxidase assay.

4.3.3 Evaluation -1 Alkyltelluro phenol 106a (ν0 = 22 µM min ) reduced hydrogen peroxide 7- fold faster than ebselen when evaluated in the coupled reductase assay (Ta- ble 10). O-methylation of the phenol reduced its reactivity (107a vs 106a). This could be explained by hydrogen-bonding which brings the peroxide closer to the chalcogen centre in the phenol. Butyltelluro resorcinol 110a showed a similar reactivity as 106a, however, when one of the hydroxyl- groups was O-alkylated (112), the compound became inactive. Installation of two extra methyl groups into the aromatic scaffold resulted in a decreased -1 reactivity (117a; ν0 = 6 µM min ). When two butyltelluro groups were introduced into the same phenol, the GPx-activity was reduced to 3 µM min-1. Compounds carrying longer aliphatic chains were either inactive or insoluble. For solubility reasons, the thiol peroxidase assay seemed more suitable for evaluating the GPx-activity of these compounds. As compared to the -1 reference, diphenyl diselenide (ν0 = 1.2 µM min ), alkyltelluro phenols 106 showed a 20-fold higher GPx-activity. In line with the results from the coupled reductase assay, a decreased GPx-activity for the O-methylated ana- -1 logues 107 (ν0 = 9 µM min ) was recorded. Resorcinol analogues 110 (ν0 = 80 µM min-1) showed the highest GPx-activities among the compounds tested. Resorcinols 117 were also quite active. For the phenols 119, carrying two alkyltelluro substituents, we observed that longer alkyl chains caused a significant decrease in the GPx-activity (butyl ≈ octyl > hexadecyl). Although only a small fraction of the organotellurium antioxidants described in this thesis were evaluated for their GPx-like activities, we feel confident that most of them would catalyse reduction of hydroperoxides in the presence of thiol reducing agents.

Table 10. Initial Rates for Reduction of H2O2 in the Coupled Reductase and Thiol Peroxidase Assays. -1 a -1 b Antioxidant ν0 (µM min ) ν0 (µM min )

106a R = Butyl 22 ± 3 28.0 ± 2.2 106b R = Octyl 11 ± 2 24.9 ± 2.4 106c R = Hexadecyl Inactive 24.0 ± 1.0

107a R = Butyl 10 ± 2 9.4 ± 2.6 107b R = Octyl Inactive 9.7 ± 0.5 107c R = Hexadecyl Inactive 9.5 ± 2.2

110a R = Butyl 21 ± 2 80.7 ± 1.2 110b R = Octyl 5 ± 2 80.2 ± 2.9 110c R = Hexadecyl Inactive 79.8 ± 0.4

112a R = Butyl Inactive 28.8 ± 2.0 112b R = Octyl Inactive 31.0 ± 0.9 112c R = Hexadecyl Inactive 31.9 ± 1.0

117a R = Butyl 6 ± 1 70.1 ± 2.7 117b R = Octyl 6 ± 2 44.3 ± 1.3 117c R = Hexadecyl Inactive 44.4 ± 1.1

119a R = Butyl 3 ± 1 58.4 ± 2.0 119b R = Octyl Inactive 63.5 ± 1.5 119c R = Hexadecyl Insoluble 31.7 ± 3.0

Ebselen 3 ± 2

Ph2Se2 1.2 ± 0.5 a GPx-activities determined by using the coupled reductase assay. Assay conditions: phosphate buffer (100 mM), pH 7.5, with EDTA (1 mM), H2O2 (0.8 mM), NADPH (0.20 mM), glutathione reductase (1 unit/mL), and catalyst (20 μM). b GPx-activities determined by using the thiol peroxidase assay.

5. Mechanistic Considerations

To account for the improved radical-trapping capacity and regenerability of alkyltelluro phenols, an unconventional mechanism involving oxygen-atom transfer (OAT) from the peroxyl radical to the tellurium, followed by hydrogen-atom transfer (HAT) from the the phenolic O-H to the alkoxyl radical was proposed (Scheme 15). As shown in Scheme 45, we propose a similar mechanism for alkyltelluro aromatic amines. The mechanism is supported by an early pulse radiolysis experiment118 which showed that O-atom transfer from peroxyl radicals to tellurium is a facile process (k = 108 M-1s-1). - We have shown that the ortho-alkyltelluro compound 74b (Rinh = 2 µM h 1 ) is a more efficient radical-quencher than the para-analogue 76 (Rinh = 6 µM h-1). In the para-substituted compound, the alkoxyl radical formed is far away from the N-H. Thus, HAT is more difficult and there is a risk that alkoxyl radicals diffuse out of the solvent cage and propagate the chain- reaction.

Scheme 45. Proposed mechanism for the radical-trapping activity and the regeneration of alkyltelluro aromatic amines.

Regeneration of the tellurium-containing aromatic amines and phenols would require consumption of three equivalents of thiol. The reduction of telluroxide in the presence of thiols is proposed to involve reductive elimina- tion of a tellurium (IV)-dithiolate 130.119 An alternative pathway, involving nucleophilic attack of thiol on a tellurium (IV)-hydroxide thiolate 131 has also been proposed (Scheme 46).

Scheme 46. Proposed mechanisms for the reduction of tetravalent to divalent Te.

On the other hand, the mechanism for reduction of aminyl/phenoxyl radical to amine/phenol is not obvious. Since the S-H bond is much stronger than the phenolic O-H bond, HAT is unlikely to occur. The regeneration may instead involve PCET, wherein the electron-transfer from NAC to the phenoxyl/aminyl radical would be expected to be rate limiting. However, the observation that α-tocopherol is not regenerable by NAC but becomes so when an alkyltelluro group is introduced into the molecule, suggests that the heteroatom is somehow involved in the regeneration.

Scheme 47. Tentative mechanism for the regeneration of a diarylamine antioxidant.

We have also considered another mechanism for the regeneration, involving ET from the tellurium to the phenoxyl/aminyl radical as exemplified for a diarylamine in Scheme 47. Intramolecular ET from the large Te would be favourable as compared to any intermolecular reaction. Judging from one- electron reduction potentials of the redox couples Ph-OH•+/Ph-OH, + + 120 121 122 Ph2NH• /Ph2NH and PhTeMe• /PhTeMe (1.24 V , 1.24 V and 0.97 V vs NHE, respectively), the reaction would be thermodynamically favoured. The radical cation formed is known to disproportionate, yielding half an equivalent each of a Te (II) and a Te(IV) species, the latter reducible by thiol. Regeneration is completed after proton abstraction.

6. Concluding Remarks

In this thesis, we have demonstrated that ortho-alkyltelluro phenols and aromatic amines are efficient radical-quenchers. Also, they can reduce hydroperoxides. In the presence of thiol reducing agents they can serve, in a catalytic fashion, both as chain-breaking and preventive antioxidants.

1. When alkyltelluro groups were introduced into β-, γ- and δ-tocopherols, both the radical-trapping efficiency and the regenerability were significantly improved. Remarkably, bis-(tocopheryl) tellurides showed an inhibition period eight-fold longer than recorded with α-tocopherol.

2. The radical-trapping efficiency and the regenerability increased when alkyltelluro groups were incorporated into aromatic amines. Diphenyl amines carrying alkyltelluro groups showed the most impressive antioxidant properties. Thiol-consumption experiments suggested that efficient quenching of the alkoxyl radical in the solvent cage is key to long- lasting antioxidants.

3. Introduction of a second alkyltelluro groups into alkyltellurophenols significantly increased the inhibition time whereas introduction of a second hydroxyl group had the opposite effect.

4. Evaluation of phenols bearing electron donating and electron withdrawing alkyltelluro groups showed that catalysts with a high electron density at the tellurium atom inhibited peroxidation for a longer time and quenched peroxyl radicals more efficiently.

To explain the improved radical-trapping efficiency and the long inhibition times, we have proposed an unconventional mechanism involving OAT from the peroxyl radical to the tellurium followed by HAT from the O-H/N-H to the resulting alkoxyl radical in a solvent cage. Reduction of the resulting telluroxide phenoxyl / aminyl radical is brought about by NAC. As compared with conventional antioxidants, our organotellurium compounds have many advantages: They are multifunctional in the sense that they are both chain-breaking and peroxide-decomposing. They are better quenchers of peroxyl radicals than most other antioxidants and, most im-

portantly, they are regenerable by mild reducing agents such as thiols. The demand for "green" and "sustainable" antioxidants will undoubtedly increase in the future. In this perspective our antioxidants would seem very attractive. Though, more information concerning their toxicity would be needed. The few studies we have performed so far do not indicate any alarming toxicity. For example, bis-alkyltelluro phenol 119a showed minimal toxicity in a pre- osteoblast MC3T3 cell-line.123

Summary in Swedish

Syre — ett av de viktigaste ämnena för mänskligt liv — har ironiskt nog även skadliga effekter för såväl material som organismer. Reaktiva syreföre- ningar (ROS; ett samlingsnamn för väteperoxid, peroxylradikaler, superoxi- der m.m.) produceras vid cellens metabolism. Trots att ROS är involverade i olika biologiska processer, utgör de ett hot mot det biologiska systemet när halten i cellen blir för hög. Peroxylradikaler är en av de farligaste förening- arna och de har förknippats med olika sjukdomar såsom cancer, hjärt- och kärlsjukdomar. Intresset för att utveckla antioxidanter — föreningar som skyddar oss mot ROS — är därför stort. Antioxidanter klassificeras som preventiva eller kedjebrytande, beroende på hur de påverkar autoxidationen. Preventiva antioxidanter förhindrar bildandet av fria radikaler genom att förstöra dess prekursorer. Seleninnehål- lande enzymer som glutationperoxidas fungerar på detta sätt och har en vik- tig roll i kroppens försvarssystem. Kedjebrytande antioxidanter, å andra si- dan, fångar upp fria radikaler i systemet och därmed bryts den kedjereaktion som fria radikaler kan ge upphov till. Vitamin E är den mest reaktiva kedjebrytande antioxidant som förekommer naturligt. Med vitamin E som förebild togs de välkända konserveringsmedlen BHT och BHA fram. Antioxidanter tillsätts olja, bränsle och diverse organiska material såsom polymer för att öka deras livslängd. Idag används årligen miljontals ton antioxidanter i olika sammanhang. Eftersom de traditionella antioxidanterna används i stökiomet- risk mängd, d.v.s. för varje molekyl peroxylradikal krävs det en motsvarande mängd antioxidant, ger det upphov till stora mängder avfall. Detta är ur ett ekonomiskt och miljömässigt perspektiv ofördelaktigt. Den här avhandlingen rör design, framställning och utvärdering av antioxidanter som innehåller tellur. Fördelen med dessa antioxidanter är att de fungerar katalytiskt och är regenererbara i närvaro av co-antioxidanter såsom N-acetylcystein (NAC). Dessutom är dessa antioxidanter multifunktionella d.v.s. de är både kedjebrytande och peroxidförstörande. Vi har funnit att fenoler som innehåller en alkyltellurogrupp i ortho-position ger en markant ökning vad gäller dess förmåga att bryta kedjereaktionen. Detta kan förklaras med en mekanism där peroxylradikalerna först överför en syreatom till tell- luren. Den bildade alkoxylradikalen i lösningsmedelburen abstraherar sedan en väteatom från den närliggande hydroxygruppen och därmed bildas en resonanstabiliserad fenoxylradikal (Figur 1).

Figur 1. Mekanismen

Syftet med forskningsarbetet som presenteras i denna avhandling är att ta fram nya regenererbara antioxidanter med kedjebrytande egenskaper samt öka vår förståelse för hur dessa tellurinnehållande antioxidanter fungerar.

I den första delen av avhandlingen har vi studerat de kedjebrytande egen- skaperna och regenererbarheten hos tellurinnehållande vitamin E-föreningar (I). Vi har visat att dessa föreningar motverkar autoxidation signifikant bättre än motsvarande svavel- och selenanaloger. Föreningarna är överlägsna vitamin E både vad gäller den radikalfångande förmågan och regenererbarheten. Tellurobistokoferoler, d.v.s. tellurföreningar som har två tokoferoler bundna till heteroatomen har en åtta gånger längre inhiberingstid än vitamin E.

Vi har också framställt och studerat tellurinnehållande aromatiska aminer med olika struktur. Difenylaminer med en alkyltelluro grupp i ortho-position (III) har gett de bästa resultaten. Med hjälp av tiolförbrukningsexperiment har vi visat att de föreningar som fördröjer peroxidationen längst förbrukar

NAC mycket långsammare än föreningar med kort inhiberingstid. Vi tror att inhiberinstiden påverkas av hur effektivt väteatomen överförs till alkoxylradikalen i lösningsmedelsburen. I den andra delen av avhandlingen försöker vi få en bättre bild av hur antioxidanterna fungerar. Syftet är att identifiera de faktorer som påverkar den kedjebrytande egenskapen och regenererbarheten hos de kalkogeninnehål- lande antioxidanterna. Bland annat har vi studerat:

1) Hur antioxidantegenskaperna påverkas av antalet alkyltellurogrupper på aromatringen: Utvärdering av alkyltellurofenoler, alkytelluroresorcinoler och bis(alkyltelluro) fenoler visar att ett ökat antal alkyltellurogrupper bidrar till en längre inhiberingstid medan ett ökat antal hydroxylgrupper leder till motsatt effekt.

2) Hur regenererbarheten påverkas av elektrontätheten på telluratomen: Studier av diaryltellurider med elektrondonerande och elektrondragande grupper para till heteroatomen visar att de förstnämnda är mer regenererbara. Föreningen IV inhibiberar autoxidation cirka 6 gånger längre än vitamin E.

Acknowledgement

I would like express my sincere gratitude to all the people that made my journey as a PhD student joyful, exciting and unforgettable. First of all, I want to thank my supervisor, Prof. Lars Engman for the opportunity to carry out research under your supervision. Thanks for a wonder- ful five-year journey with encouragement, support and patience! I am very glad for the freedom and trust you give me to try around with new ideas in the lab. Most importantly, your peaceful personality truly inspires me how to handle things in research and beyond. I would also like to thgank my co- supervisor Dr. Peter Dinér for the help and enthusiasm you showed during these years and the time during my undergraduate study. Good luck with the research at KTH. Thanks to all the current and past members in the LE group. Specially Dr. Vijay P. S. for all the discussions about chemistry and life. I have learned much from you! Jiajie Y., thanks for all the fun and excitement in the lab and outside. It was really fun to travel around with you. Dr. Khadijeh B., your "nothing is impossible" attitude is really encouraging. Arno, I am still amazed for the number of thesis/books you read every month. Thanks for the nice time and for all the German expressions you taught me. Good luck with your PhD study in Italy! Carsten, thanks for all the scientific and non- scientific discussions in the lab. Big thanks to the colleagues at BMC! The AGHGs: Hao H., thanks for the nice time as a good colleague and flat mate during these time, and for the trainings over the years to spicy food! Xiao H., your stories always give me a big laugh and keep me charged. Michael N., Dr. Magnus B., Anna L. and Sandra O., thanks all of you for the nice atmosphere and the unforgettable teaching times, and, of course Prof. Adolf G. for all the help with NMR and being the "one to find" when the NMR instruments are on strike. Prof. Hele- na G., thanks for all your wise suggestions and advices over the years. The JS/RenFuels gang: Dr. Maxim G. thanks for all the interesting discussions and the help with chemicals. Dr. Alban C., your stories in the lunch room always kept me entertained. Dr. Supaporn S., thanks for all the help and interesting discussions. Anon B., I am always amazed for your passions to life and the amount of photos you took! Dr. Alex P., Dr. Christian D., Dr. Joakim L., Dr. Johan V., Pemikar (Cherry) S., Dr. Sunisa A. and their leader Dr. Joseph S., a great thank you for the help during these years. The LTP group: Karthik D., I really like your philosophy about life and thanks for

being such a good company during the thesis-writing period, good luck with yours! Carina S. and Dr. Emilien D., thanks for all the help and for keeping me activated, and Dr. Luke P. for your help in chemistry and a great course in Organometallics. The JK crew: Jie, your stories and views on life are always entertaining and inspiring. Thanks, Lina and Prof. Jan K. for the kind atmosphere. Also, I would like to express my thanks to Dr. Huan M. for the countless positive energies you are spreading. Dr. Laura M. S., you are not only my first lab mate during my undergraduate studies but also the one that's always full with happiness and excitement. Dr. Alba E. D., your suggestions and view on life is really encouraging. Prof. Thomas N., thanks for the unforgettable teaching periods we spent together in the undergraduate course lab. Dr. Rikard E., Emile H., Alexandra B., Thilak E., Derar S., Dr. Mohit T. and many more that I forgot to mention, thank you all during these years! Many thanks to Gunnar S., for the delivering of chemicals and un- countable bottles of hexanes, and Bo F., for all the technical support. To the ones at Ångström: many thanks to the EB-group members, specially Dr. Julien A., for the great times and for introducing me to the French kitchen and many other nerdy things! Daniel K., thanks for being such a good company and for all the fun and excitements over the years. Don't let the measurements get you! Ruisheng X., your super-productivity and passions for chemistry always amazed me. Dr. Eszter B., thanks for the great lectures and a great teaching period. The HO team: Dr. Raffaello P., it was great to have you in the office and lab. Thanks for all the interesting discussions from benzene to Greek culture. I am still looking forward to your "book" about separation without running columns. Rabia A., I am grateful for all the help and interesting conversations, and of course for letting me try some interesting dishes from Pakistan. Kjell J., thanks for the interesting conversations from aromaticity to aroma in food. Alexandra D., I would remember you as the one that is always happy. Dr. Henrik O., thanks for all interesting talks that keep us exciting over the years. Many thanks to those had enriched my life outside the lab: Jiajie, Hao, Daniel, Rabia, Xiao, Jie, Raffaello and Kjell for all the dinners, movie nights and trips we had made together. I am very glad that I survived those -40°C- nights with many of you in Kiruna! Many thanks to my old friends, Branko, Robin, Mathias, Fredrik, Katriann, Darshi and Dag, just to name few of them. You made me what I become today. Dr. Vijay P. S., Jiajie, Dr. Raffaello P., Prof. Helena G. and Prof. Lars E. Thanks for proofreading my thesis and giving me invaluable suggestions. The thesis would probably not be finished without your efforts. Also, I would like to acknowledge Liljewalchs Stiftelse and Kungliga Vetenskapsakademin for sponsoring the trips to Prague and Japan. At last, I would like to express a big thank you to my parents and relatives for always being there for me!

References

1. Massimi, M. Exclusion Principle (or Pauli Exlusion Principle), in Compendium of Quantum Physics, ed. Greenberger, D.; Hentschel, K.; Weinert, F. Springer, 2009, ISBN 978-3540706267. 2. Halliwell, B; Gutteridge, J. M. C. Free Radicals in Biology and Medicine (4th ed.), Oxford University Press, Oxford, 2007, ISBN 978-0198568681. 3. Seyferth, D. Organometallics 2001, 20, 1488-1498. 4. Frankland, E. Q. J. Chem. Soc. 1850, 2, 263-296. 5. Gomberg, M. J. Am. Chem. Soc. 1900, 22, 757-771. 6. Rozantsev, E. G.; Loshadkin, D. V. Des. Monomers Polym. 2001, 4, 281-300. 7. Hey, D. H.; Waters, W. A. Chem. Rev. 1937, 21, 169-208. 8. Flory, P. J. J. Am. Chem. Soc. 1937, 59, 241-253. 9. Karasch, M. S.; Englemann, H.; Mayo, F. R. J. Org. Chem. 1937, 2, 288-302. 10. Gerschman, R.; Gilbert, D. L.; Nye, S. W.; Dwyer, P.; Fenn, W. O. Science 1954, 119, 623-626. 11. (a) McCord, J. M.; Fridovich, I. J. Biol. Chem. 1968,243, 5753-5760; (b) McCord, J. M.; Fridovich, I. J. Biol. Chem. 1969, 244, 6049-6055. 12. Harman, D. J. Am. Geriatr. Soc. 1972, 20, 145-147. 13. Valko, M.; Leibfritz, D.; Moncol, J.; Cronin, M. T. D.; Mazur, M.; Telser, J. Int. J. Biochem. Cell Biol. 2007, 39, 44-84. 14. Pacher, P.; Beckman, J. S.; Liaudet, L. Physiol. Rev. 2007, 87, 315-424. 15. (a) Burnett, G. M.; Melville, H. W. Nature 1945, 156, 661; (b) Bartlett, P. D.; Swain, C. G. J. Am. Soc. Chem. 1945, 67, 2273-2274. 16. Walling, C. J. Chem. Educ. 1986, 63, 99-102. 17. Stone, T. J.; Buckman, T.; Nordio, P. L.; McConnell, H. M. Proc. Natl. Acad. Sci. USA 1965, 54, 1010-1017. 18. Howard, J. A. Homogeneous Liquid-Phase Autoxidations, in Free Radicals, ed. Kochi, J. K. John Wiley & Sons; New York, 1973, ISBN 978-0471497028. 19. Lalevée, J.; Fouassier, J. P. Overview of Radical Initiation, in Encyclopedia of Radicals in Chemistry, Biology and Materials, Wiley-Blackwell, 2012, ISBN 978-0470971253. 20. Simmie, J. M.; Black, G.; Curran, H. J.; Hinde, J. P. J. Phys. Chem. A 2008, 112, 5010-5016. 21. Nicholson, S.; Leslie, D. E.; Efandis, T.; Fairley, C. K.; Gust, I. D. J. Virol. Methods. 1991, 33, 311-317; (b) Lambeth, J. D. Nat. Rev. Immunol., 2004, 4, 181-189. 22. Kuppusamy, P.; Zweier, L. J. Bio. Chem. 1989, 264, 9880-9884. 23. Ortiz de Montellano, R. R. Trends Pharmacol. Sci. 1989, 10, 354-359. 24. Porasuphatana, S.; Tsai, P.; Rosen, G. M. Comp. Biochem. Physiol. C: Pharma- col. Toxicol. 2003, 134, 281-289. 25. Barja, G. J. Bioenerg. Biomembr. 1999, 31, 347-366. 26. Maillard, B.; Ingold, K. U.; Scaiano, J. C. J. Am. Chem. Soc. 1983, 105, 5095- 5099.

27. Horward, J.; Ingold, K. U. Can. J. Chem. 1967, 45, 793-802. 28. (a) Pozdeeva, N. N.; Denisov, E. T. Kinet. Catal. (Engl. Transl.), 2010, 51, 211- 218; (b) Howard, J. A. Adv. Free Radic. Chem. 1972, 4, 49-174; (c) J. A. How- ard, Ingold, K. U. Can. J. Chem. 1965, 43, 2729-2736. 29. Siddiqui, T.; Zia, M. K.; Ali, S. S.; Rehman, A. A.; Ahsan, H.; Khan, F. H. Arch. Physiol. Biochem. 2015, 122, 1-7. 30. Sugamura, K.; Keaney, J. F. Jr. Free Radic. Biol. Med. 2011, 51, 978-992. 31. Liou, G.-Y.; Storz, P. Free Radic. Res. 2010, 44, 479-496. 32. Park, H. S.; Kim, S. R.; Lee, Y. C. Respirology, 2009, 14, 27-38. 33. Liochev, S. I. Free Radic. Biol. Med. 2013, 60, 1-4. 34. McCord, J. M.; Fridovich, I. J. Biol. Chem. 1969, 244, 6049-6055. 35. Yin, H.; Xu, L.; Porter, N. A. Chem. Rev. 2011, 11, 5944-5972. 36. Burton, G. W.; Ingold, K. U. Acc. Chem. Res. 1986, 19, 194-201. 37. van Meer, Gerrit; Voelker, D. R.; Feigenson, G. W. Nat. Rev. Mol. Cell Biol. 2008, 9, 112-124. 38. Fernandis, A. Z.; Wenk, M. R. Curr. Opin. Lipidol. 2007, 18, 121-128. 39. Brasaemle, D. L. J. Lipid Res. 2007, 48, 2547-2559. 40. Ayala, A.; Muñoz, M. F.; Argüelles, S. Oxid. Med. Cell Longev. 2014, 2014, Article ID 360438, 31 pages. 41. Berliner, J. A.; Heinecke, J. W. Free Radic. Biol. Med. 1996, 20, 707-727. 42. (a) Hammad, L. A.; Wu, G.; Saleh, M. M.; Klouckova, I.; Dobrolecki, L. E.; Hickey, R. J.; Schnaper, L.; Novotny, M. V.; Mechref, Y. Rapid. Commun. Mass Spectrom. 2009, 23, 863; (b) Wu, R. P.; Hayashi, T.; Cottam, H. B.; jin, G.; Yao, S.; Wu, C. C. N.; Rosenbach, M. D.; Corr, M.; Schwab, R. B.; Carson, D. A. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 7479-7484. 43. Silverstein, R. L.; Febbraio, M. Sci. Signalling 2009, 2, re3. 44. Imai, Y. Kuba, K.; Neely, G. G.; Yaghubioan-Malhami, R.; Perkmann, T.; van Loo, G.; Ermolaeva, M.; Veldhuizen, R; Leung, Y. H.; Wang, H.; Liu, H.; Sun, Y.; Pasparakis, M.; Kopf, M.; Mech, C.; Bavari, S.; Peiris, J. S.; Slutsky, A. S.; Akira, S.; Hultqvist, M.; Holmdahl, R.; Nicholls, J.; jiang, C.; Binder, C. J.; Penninger, J. M. Cell 2008, 133, 235-249. 45. Nonas, S.; Miller, I.; Kwakitinarong, K.; Chatchavalvanich, S.; Gorshkova, I.; Bochkov, V. N.; Leitinger, N. Nataajan, V.; Garcia, J. G.; Birukov, K. G. Am. J. Respir. Crit. Care Med. 2006, 173, 1130-1138. 46. Simonian, N. A.; Coyle, J. T. Annu. Rev. Pharmacol. Toxicol. 1996, 36, 83-106. 47. Halliwell, B. Biochem. Soc. Trans. 2007, 35, 1147-1150. 48. Abreu, I. A.; Cabelli, D. E. Biochim. Biophys. Acta 2010, 1804, 263-274. 49. Loew, O. U.S. Dept. of Agri. Repts. 1900, 56, 5-57. 50. Chelikani, P., Fita, I.; Loewen, P. C. Cell. Mol. Life Sci. 2004, 61, 192-208. 51. Mills, G. C. J. Bio. Chem. 1957, 229, 189-197. 52. Flohé, L.; Loschen, G.; Gunzler, W. A.; Eichele, E. Z. Physiol. Chem. 1972, 353, 987-999. 53. Prabhakar, R.; Vreven, T.; Morokuma, K.; Musaev, D. G. Biochemistry 2005, 44, 11864-11871. 54. Brydson, J. A. Plastics Materials (6th ed.), Elsvier; 1995, ISBN 978- 0750618649. 55. Bramley, P. M.; Elmadfa, I.; Kafatos, A.; Kelly, F. J.; Manios, Y.; Roxborough, H. E.; Schuch, W.; Sheehy, P. J. A.; Wagner, K, -H. J. Sci. Food Agric. 2000, 80, 913-938. 56. Szent-Györgyi, A. Biochem. J. 1928, 22, 1387-1409. 57. Cameron, E.; Pauling, L.; Leibovitz, B. Cancer Res. 1979, 39, 663-681. 58. Golumbic, C.; Matill, H. A. J. Am. Chem. Soc. 1941, 63, 1279-1280.

59. Stuckey, B. Handbook of Food Additives, 1972, 185-221. (d) Dacre, J. C. Bio- chem. J. 1961, 78, 758-766. 60. a) Sander, D.; Chang, V.; Ota, A.; Nomura, N. Proc. Hawaii. Entomol. Soc. 1992, 31, 65-72. (b) Augustin, M.; Berry, S. J. Am. Oil Chemist's Soc. 1983, 60, 1520-1523. 61. Stecher, P. O'Neil, The Merck Index, Merck Research Laboratories, Rahway, New Jerey, USA, 1968, 179. 62. Hilton, J. Can. Veterinary J. 1989, 30, 682-684. 63. Hui, Y. H. Handbook of Food Science, Technology and Engineering (Vol. 2) CRC Press Taylor & Francis Group; 2005, ISBN 978-0849398483. 64. Yehye, W. A.; Rahman, N. A.; Ariffin, A.; Hamid, S. B. A.; Alhadi, A. A.; Kadir, F. A.; Yaeghoobi, M. Eur. J. Med Chem. 2015, 101, 295-312. 65. Fischer, W. J. Chem. Educ. 2001, 78, 1333. 66. Jorissen, W. P.; Bassett, H.; Damiens, A.; Fichter, F.; Remy, H. J. Am. Chem. Soc. 1941, 63, 889-897. 67. Greenwood, N. N.; Earnshaw, A. Chemistry of the Elements (2nd Ed.), Oxford: Butterworth-Heinenmann, 1997, ISBN 978-0750633659. 68. Zhang, Y. Isis, 1986, 77, 487-497. 69. Lavoisier, A. Element of Chemistry, Dover Publications Inc, 1965, ISBN 978- 0486646244. 70. Zhang, Y.; Hogg, N. Free Radic. Biol. Med. 2005, 38, 831-838. 71. Weeks, M. E. J. Chem. Educ. 1932, 9, 474-485. 72. Hoffmann, P. R.; Berry, M. J. Mol. Nutr. Food Res. 2008, 52, 1273-1280. 73. Köhrle, J. Thyroid. 2005, 15, 841-853. 74. Trachootham, D.; Lu, W.; Ogasawara, M.; Valle, N. R.; Huang, P. Antioxid. Redox. Signal. 2008, 10, 1343-1374. 75. Moreno-Reyes, R.; Mathieu, F.; Boelaert, M.; Begaux, F.; Suetens, C.; Rivera, M. T.; Nève, J.; Perimutter, N.; Vanderpas, J. Am. J. Clin. Nutr. 2003, 78, 137- 144. 76. Loscalzo, J. N. Engl. J. Med. 2014, 370, 1756-1760. 77. Gromer, S.; Eubel, J. K.; Lee, B. L.; Jacob, J. Cell. Mol. Life Sci. 2005, 62, 2414-2437. 78. Flohe, L.; Günzer, W. A.; Schock, H. H. FEBS Lett. 1973, 32, 132-134. 79. Weeks, M. E. J. Chem. Educ. 1932, 9, 474-485. 80. Wöhler, F. Liebigs Ann. Chem. 1840, 35, 111-115. 81. Basol, B. M.; McCandless, B. J. Photon. Energy. 2014, 4, 040996. 82. Ramadan, S. E.; Razak, A. A.; Ragab, A. M.; El-Meleigy, M. Biol. Trace Elem. Res. 1988, 20, 225-232. 83. Litwinienko, G.; Ingold, K. U. J. Org. Chem. 2005, 70, 8982-8990. 84. DiLabio, G. A.; Ingold, K. U. J. Am. Chem. Soc. 2005, 127, 6693-6669. 85. (a) Zavitsas, A. A.; Melikian, A. A. J. Am. Chem. Soc.1975, 97, 2757-2763; (b) Zavitsas, A. A.; Chatgilialoglu, C. J. Am. Chem. Soc. 1995, 117, 5794-5795. 86. Mayer, J. M.; Hrovat, D. A.; Thomas, J. L.; Borden, W. T.; J. Am. Chem. Soc. 2002, 124, 11142-11147. 87. Ingold, K. U.; Pratt, D. A. Chem. Rev. 2014, 114, 9022-9046. 88. Burton, G. W.; Doba, T.; Gabe, E.; Hughes, L.; Lee, F. L.; Prassad, L.; Ingold, K. U. J. Am. Chem. Soc. 1985, 107, 7053-7065. 89. Lucarini, M.; Pedulli, G. F. Chem. Soc. Rev. 2010, 39, 2106-2119. 90. Howard, J. A.; Ingold, K. U. Can. J. Chem. 1963, 41, 2800-2806. 91. (a)Wijtmans, M.; Pratt, D. A.; Brinkhorst, J.; Serwa, R.; Valgimigli, L.; Pedulli, G. F.; Porter, N. A. J. Org. Chem. 2004, 69, 9215-9223; (b) Valgimigli, L.; Bri-

gati, G.; Pedulli, G. F.; DiLabio, G. A.; Mastragostino, M.; Arbizzani, C.; Pratt, D. A. Chem. Eur. J. 2003, 9, 4997-5010. 92. Bowry, V. W.; Ingold, K. U. Acc. Chem. Res. 1999, 32, 27-34. 93. Valgimigli, L.; Banks, J. T.; Ingold, K. U.; Lusztyk, J. J. Am. Chem. Soc. 1995, 117, 9966-9971. 94. Liebler, D. C.; Kaysen, K.; Kennedy, T. A. Biochemistry, 1989, 28, 9972-9777. 95. Klemchuk, P. P. Ullmann's Encyclopedia of Industrial Chemistry, Wiley-VCH Verlag: Weinheim, 2000, ISBN 978-3527329434. 96. Bolsman, T. A. B. M.; Blok, A. P.; Frijns, J. H. G. Recl. Trav. Chim. Pays-Bas 1978, 97, 310-312. 97. Jensen, R. K.; Korcek, S.; Zinbo, M.; Gerlock, J. L. J. Org.Chem. 1995, 60, 5396-5400. 98. Haidasz, E. A.; Shah, R.; Pratt. D. A. J. Am. Chem. Soc. 2014, 136, 16643- 16650. 99. Müller, A.; Cadenas, E.; Graf, P.; Sies, H. Biochem. Pharmacol. 1984, 33, 3235–3239. 100. (a) Parnham, M. J.; Graf, E. Prog. Drug Res. 1991, 36, 9-47. (b) Parnham, M. J.; Leyck, S.; dereu, N.; Winkelmann, J.; Graf, E. Adv. Inflam. Res. 1985, 10, 397-400. (c) Parmha,. M. J.; Graf, E. Biochem. Pharmacol. 1987, 36, 3095- 3102. 101. Wilson, S. R.; Zucker, P. A.; Huang, R.-R C.; Spector, A. J. Am. Chem. Soc. 1989, 111, 5936-2939. 102. Lamani, D. S.; Bhowmick, D.; Mugesh, G. Org. Biomol. Chem. 2012, 10, 7933- 7943. 103. Back, T. G.; Moussa, Z. J. Am. Chem. Soc. 2002, 124, 12104-12105. 104. Sarma, B. K.; Mugesh, G. Chem. Eur. J., 2008, 14, 10603-10614. 105. Bhowmick, D.; Mugesh, G. Org. Biomol. Chem. 2015, 13, 10262-10272. 106. Detty, M. R. Organometallics, 1992, 11, 2147-2156. 107. Engman, L.; Stern, D.; Frisell, H.; Vessman, K.; Berglund, M.; Ek, B.; Anders- son, C. M.; Bioorg. Med. Chem. 1995, 3, 1255-1262. 108. Ren, X.; Xue, Y.; Zhang, K.; Liu, J.; Luo, G.; Zheng, J.; Mu, Y.; Shen, J. FEBS Lett. 2001, 507, 377-380. 109. Malmström, J.; Jonsson, M.; Cotgreave, I. A.; Hammarström, L.; Sjödin, M.; Engman, L. J. Am. Chem. Soc. 2001, 123, 3434-3440. 110. (a) Kumar, S.; Johansson, H.; Kanda, T.; Engman, L.; Müller, T., Jonsson, M.; Pedulli, G. F.; Petrucci, S.; Valgimigli, L. Org. Lett. 2008, 10, 4895-4898; (b) Kumar, S.; Johansson, H.; Kanda, T.; Engman, L.; Müller, T., Bergenudd, H.; Jonsson, M.; Pedulli, G. F.; Amorati, R..; Valgimigli, L. J. Org. Chem. 2010, 75, 716-725. 111. Johansson, H.; Shanks, D.; Engman, L.; Amorati, R.; Pedulli, G. F. Valgimigli, L. J. Org. Chem. 2010, 75, 7535-7541. 112. Amorati, R.; Valgimigli, L.; Dinér, P.; Bakhtiari, K.; Saeedi, M.; Engman, L. Chem. Eur. J. 2013, 19, 7510-7522. 113. Netscher, T.; Mazzini, F.; Jestin, R. Eur, J. Org. Chem. 2007, 7, 1176-1183. 114. Mazzini, F., Netscher, T.; Salvadori, P. Tetrahedron 2005, 61, 813-817. 115. Shanks, D.; Amorati, R.; Fumo, M. G.; Pedulli, G. F., Valgimigli, L.; Engman, L. J. Org. Chem. 2006, 71, 1033-1038. 116. Wendel, A. Methods Enzymol. 1981, 77, 325-333. 117. Iwaoka, M.; Tomoda, S. J. Am. Chem. Soc. 1994, 116, 2557-2561. 118. Engman, L.; Persson, J.; Merényi, G.; Lind, J. Organometallics 1995, 14, 3641- 3648. 119. You, Y.; Ahsanm, K.; Detty, M. R. J. Am. Chem. Soc. 2003, 125, 4918-4927.

120. Fuchigami, T.; Atobe, M.; Inagi, S. Fundamentals and Applications of Organic Electrochemistry: Synthesis, Materials and Organic Electrochemistry, John Wiley & Sons, 2014, ISBN 978-1118653173. 121. Jonsson, M.; Wayner, D. D. M.; Lusztyk, J. J. Phys. Chem. 1996, 100, 17539- 17543. 122. Engman, L.; Persson, J.; Andersson, C. M.; Berglund, M. J. Chem. Soc. Perkin Trans. 1994, 12, 3174-3182. 123. Unpublished results. Lu, X., Mestres, G., Singh, V. P., Effati, P., Poon, J., Engman, L., Karlsson Ott, M. 2016.

Acta Universitatis Upsaliensis Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology 1423 Editor: The Dean of the Faculty of Science and Technology

A doctoral dissertation from the Faculty of Science and Technology, Uppsala University, is usually a summary of a number of papers. A few copies of the complete dissertation are kept at major Swedish research libraries, while the summary alone is distributed internationally through the series Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology. (Prior to January, 2005, the series was published under the title “Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology”.)

ACTA UNIVERSITATIS UPSALIENSIS Distribution: publications.uu.se UPPSALA urn:nbn:se:uu:diva-302475 2016