Iodine and Lipase Based Green Oxidation Technology

Home , Iodobenzene dichloride, Peroxymonosulfuric acid

Aleksandra Joanna KOTLEWSKA‐MIERNOWSKA

Iodine and Lipase Based Green Oxidation Technology

PROEFSCHRIFT

ter verkrijging van de graad van doctor aan de Technische Universiteit Delft, op gezag van de Rector Magnificus prof. ir. K. C. A. M. Luyben, voorzitter van het College voor Promoties, in het openbaar te verdedigen op vrijdag 5 november 2010 om 10.00 uur door

Aleksandra Joanna KOTLEWSKA‐MIERNOWSKA

Master of Science, Engineer in Chemical Technology, Warsaw University of Technology (Polen) geboren te Ciechanów, Polen

Dit proefschrift is goedgekeurd door de promotoren: prof. dr. R. A. Sheldon en prof. dr. I. W. C. E. Arends

Samenstelling promotiecommissie: Rector Magnifcus, voorzitter Prof. dr. R. A. Sheldon, Technische Universiteit Delft, promotor Prof. dr. I. W. C. E. Arends, Technische Universiteit Delft, promotor Prof. dr. K. R. Seddon, The Queen’s University of Belfast Prof. dr. J. Martens, Carl von Ossietzky University Oldenburg Prof. dr. ir. R. Orru, Vrije Universiteit Amsterdam Dr. U. Hanefeld, Technische Universiteit Delft Dr. M. Ostendorf, MSD Prof. dr. ir. H. van Bekkum, Technische Universiteit Delft

ISBN: 978‐90‐816123‐1‐9

The research described in this thesis was supported by ACTS‐IBOS

All rights reserved. No part of the material protected by this copyright notice may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording or by any information storage and retrieval system, without written permission from the author.

Printed in the Netherlands

Dla mojego Arka

Cover design by Arkadiusz Kotlewski Contents

Chapter 1 Introduction to sustainable oxidation chemistry 1

Chapter 2 Hypervalent iodine in organic synthesis 19

Chapter 3 Polymer‐attached iodine(III) reagent in selective oxidations 39 Scope of oxidants

Chapter 4 Polymer‐attached iodine(III) reagent in selective oxidations 61 Scope of substrates

Chapter 5 Lipase catalyzed in situ generation of hypervalent iodine 91 reagent for selective alcohol oxidation

Chapter 6 Epoxidation and Baeyer‐Villiger oxidation using hydrogen 109 peroxide and lipase dissolved in ionic liquids

Summary 131

Samenvatting 135

Acknowledgements 139

Curriculum Vitae 143

List of abbreviations

ACTS Advanced Catalytic Technologies for Sustainability program. acid carboxylic acid Adogen 464 methyltrialkyl(C8‐C10)ammonium chloride, phase transfer catalyst Asp aspartate ‐ amino acid ARP‐Pt Pt nanoparticles dispersed in amphiphilic polystyrene‐ polyethylene glycol resin BASF The chemical company: Baden Aniline and Soda Factory BAM Fallhammer The BAM Fallhammer test is used to determine the sensitivity of a given solid (including paste‐like and gel‐ type) substances and liquids to drop‐weight impact of known force BA butanoic acid = butyric acid ‐ [BF4] tetrafluoroborate anion BF3OEt Lewis acid, boron trifluoride etherate [BMIm]+ (1‐butyl‐3‐methylimidazolium) cation BTI = PIFA phenyliodine(III) bis(trifluoroacetate) = (bis(trifluoroacetoxyiodo) benzene) = PhI(OCOCF3)2 BV Baeyer‐Villiger CA Octanoic acid= caprylic acid CaLA Candida antarctica lipase A CaLB Candida antarctica lipase B + + [Choline] = [TMEOA] (CH3)3N CH2CH2OH = N,N,N‐trimethylethanolammonium cation cinnamyl alcohol 2(E)‐3‐phenylprop‐2‐en‐1‐ol citral 3,7‐dimethylocta‐2,6‐dienal = lemonal cl. cross‐linked CLEA cross linked enzyme aggregates CTAB cetyl‐trimethyl ammonium bromide conv. conversion DIB = DAIB = BAIB = PIDA iodobenzene diacetate =iodosobenzene diacetate = (diacetoxyiodo)benzene DMP Dess‐Martin Periodane = 1,1,1‐Triacetoxy‐1,1‐dihydro‐ 1,2‐benziodoxol‐3(1H)‐one DMSO dimethylsulfoxide [dbmim] (4‐diacetoxyiodobenzyl)‐3‐methylimidazolium EDG electron donatng groups [emim]+ 1‐ethyl‐3‐methylimidazolium cation + ‐ Et4N Br tetraethylammonium bromide EtOAc ethyl acetate eq. molar equivalent EWG electron withdrawing groups geraniol 2‐trans‐3,7‐dimethyl‐2,6‐octadiëen‐l‐ol Hammett plot relationship (LFER) linear free‐energy relationship relating reaction rates and equilibrium constants for many reactions HBD hydrogen bond donaing H2O2 hydrogen peroxide His histidine – amino acid [HOPMim]+ 1‐(3‐hydroxypropyl)‐3‐methylimidazolium cation ‐ [HCO2] formate anion IBOS The Integration of Biocatalysis and Organic Synthesis – program IBX 2‐iodoxy benzoic acid ILs ionic liquids Janda Jel polytetrahydrofuran cross‐linked polystyrene resins Koser’s reagent hydroxy(tosyloxy)iodobenzene = HTIB ld. Loading lin linear Macroporous polystyrene polystyrene cross linked with bigger spacer Me methyl group = CH3 MeCN = CH3CN acetonitrile m‐CPBA 3‐chloroperoxybenzoic acid Myrcene 7‐Methyl‐3‐methylene‐1,6‐octadiene = β‐myrcene NaOAc sodium acetate ‐ [NO3] nitrate anion Nov. 435 Novozym 435 – CaLB immobilized on polyacrylic resin immobilized NMO 4‐methylmorpholine‐N‐oxide OH alcohol OTf trifluoromethanesulfonate = triflate = CF3SO3‐ OTs tosylate refers to the anion of p‐toluenesulfonic acid ‐ (CH3C6H4SO3 ). phen 1,10-phenanthroline PhICl2 iodobenzene dichloride PhIF2 iodobenzene difluoridee PhIO iodosobenzene (PhIO)n iodosylarenes = iodosylbenzene PhIX2 iodobenzene dihalogen PhI(OAc)2 = ArI(OAc)2 iodobenzene diacetate =iodosobenzene diacetate = (diacetoxyiodo)benzene PIFA phenyliodo‐ bis(trifluoroacetate) PIPO polymer immobilized piperidinyl oxyl = polymer immobilized TEMPO β‐pinene 1S,5S)‐2,6,6‐trimethylbicyclo[3.1.1]hept‐2‐ene or (1S,5S)‐ 6,6‐dimethyl‐2‐methylenebicyclo[3.1.1]heptane PF6 hexafluorophosphate PTC phase transfer catalyst PS‐I iodopolystyrene, polymer‐supported iodosobenzene = iodine (I) PS‐I(III) polymer oxidant PS‐I(OAc)2 = PS(DAIB) = C2 diacetoxyiodobenzene bound to polystyrene resin = derivative polymer‐attached iodosobenzene diacetate PS‐DBIB = C4 derivative polymer supported iodosobenzene dibutanoate PS‐DHIB = C6 derivative polymer supported iodosobenzene dihexanoate PS‐DCIB = C8 derivative polymer supported iodosobenzene dioctanoate pulverization grinding to receive a fine powder ROP ring opening polymerization ROOH organic peroxide e.g. peracetic acid (Ethaneperoxoic acid) salen 2,2'‐Ethylenebis(nitrilomethylidene)diphenol, = N,N'‐ Ethylenebis(salicylimine) TM Select Fluor (F‐TEDA‐BF4) 1‐(chloromethyl)‐4‐fluoro‐1,4‐ diazoniabicyclo[2.2.2]octane bis(tetrafluoroborate) Ser serine – amino acid SIBX stabilized 2‐iodoxybenzoic acid + ‐ [TEA] [(HOCH2CH2)3NH = triethanolamine cation TEMPO 2,2,6,6‐tetramethyl‐1‐piperidinyloxyl TFA trifluoroacetic acid TGA thermogravimetric analysis (TGA) TLC thin layer chromatography TMS trimethylsilyl = (CH3)3Si TOF turnover frequency TON turnover number TPAP tetrapropylammonium perruthenate UHP urea‐hydrogen peroxide quant. quantitative

CChhaapptteerr 11

Introduction to sustainable oxidation chemistry

1.1 Sustainability in oxidation chemistry 1.1.1 Clean oxidation pathways 1.1.2 Sustainability and fine chemicals 1.1.3 Clean solvent 1.1.4 Recovery and recycling 1.2 Alcohol and alkene oxidation 1.2.1 Alcohol oxidation 1.2.2 Alkene epoxidation 1.3 Lipases 1.4 Scope of the thesis 1.5 References

Abstract “Green chemistry efficiently utilizes (preferably renewable) raw materials, eliminates waste and avoids the use of toxic and/or hazardous reagents and solvents in the manufacture and application of chemical products...”

P. T. Anastas, 1998

Chapter 1

1.1 Sustainability in oxidation chemistry 1.1.1 Clean oxidation pathways Oxidation is a pivotal transformation in organic chemistry. Usually this is the first step in the conversion of oil‐ and natural gas based feedstock to bulk chemicals. 1,2 As primary oxidant in these processes molecular oxygen is employed in combination with a metal catalyst in order to enhance the rate and selectivity of the reaction. Both heterogeneous and homogeneous catalysts are employed on an industrial scale. Industry makes use of gas phase oxidation for the majority of technologies to produce e.g. styrene, formaldehyde, ethylene oxide, acrylonitrile, acrylic acid or maleic anhydride. Nevertheless, liquid phase oxidation is used to produce compounds such as phenol, acetic acid, propylene oxide, benzoic acid, styrene or vinyl acetate. In general hydrogen peroxide and alkyl hydroperoxides are widely applied to produce epoxides for the bulkd an fine chemicals industry. 3,4

Figure 1.1 (A) Classical route (B) new BASF route to citral 5

The use of catalytic oxidations in bulk chemicals manufacture is common practice but in the fine chemicals industry fewer catalytic methodologies are used. Oxidation reactions are generally performed using stoichiometric amounts of inorganic oxidants such as chromiumVI,

Introduction to sustainable oxidation chemistry permanganate, manganese dioxide and periodate, leading to the formation of large quantities of inorganic salts as waste.5 Therefore, there is a pressing need for designing catalytic technologies employing benign oxidants, such as oxygen and hydrogen peroxide, for the production of fine chemicals. One pioneering example is the BASF process for the production of citral (Figure 1.1) via vapour phase oxidation over a silica supported silver catalyst.5 The BASF process uses a silver catalyst to oxidize the intermediate alcohol to the aldehyde. In the chemical process, in contrast, stoichiometric amounts of MnO2 were required. In Table 1.2 a number of single oxygen donors is listed, which can be applied in catalytic oxidations.

Table 1.2 Mass percentage active oxygen content of commonly applied oxidants6,7

Oxidant % Active oxygen co-product

O2 50 (100) H2O, H2O2

H2O2 47 H2O

N2O 36.4 N2

NaClO2 35.6 NaCl

O3 33.3 O2

HNO3 25.4 NOx NaOCl 21.6 NaCl

CH3CO2OH 21 CH3CO2H

t-BuO2H 17.8 t-BuOH

NMO 13.7 C5H11NO NaBrO 13.4 NaBr 5 KHSO 10.5 KHSO4

NaIO4 10 NaI PhIO 7.3 PhI

In principle, oxidations can be performed with a high variety of oxidants. Some of these can be used as such, but many of them need to be activated by a catalyst. However, from an environmental viewpoint, oxidants which produce salts as by‐products need to be eliminated and replaced by more friendly benign technologies.

A commonly applied combination is the transition metal activation of H2O2 and organic peroxide ROOH. Two different mechanisms are possible, involving either oxometal or peroxometal species as the active oxidant (Figure 1.2).

Chapter 1

Figure 1.2 Metal‐based activation of alkylperoxide

Catalysis which involves transition metals such as early transition metals Mo, W, Re, V, Ti, Zr generally proceeds via high‐valent peroxometal complexes according to a Lewis acid mechanism, whereas Ru, Os, or Cr, Mn, Fe participate in oxometal active species.

Jacobsen‐Katsuki enantioselective oxidation employs MnIII (salen) complex with iodosobenzene (PhIO), sodium hypochlorite (NaOCl) or m‐chloroperbenzoic acid (m‐CPBA) as terminal oxidants.8‐10 In this case the true oxidant is most likely a metaloxo MV=O species which donates the oxygen from the salen group. Also other ligands such as porphyrins, related to Fe‐chelating systems, were used in combination with iodosylbenzene as donor.11,12 A downside is the use of chlorinated solvents.13

In short, the area of selective catalytic oxidation still suffers from the requirement of non‐atom efficient oxidants. A quest for sustainable methods based on H2O2 or oxygen as oxidants is still present.

1.1.2 Sustainability and fine chemicals This paragraph deals with oxidation processes from the viewpoint of green chemistry and will highlight selected examples of sustainable processes. In the early nineties the term “Green Chemistry“ has been introduced by P. Anastas from the Environmental Protection Agency ein th USA and can be defined as follows: 14,15

“Green chemistry efficiently utilizes (preferably renewable) raw materials, eliminates waste and avoids the use of toxic and/or hazardous reagents and solvents in the manufacture and application of chemical products...”

A number of guidelines have been postulated which serve as a basis to design truly sustainable processes from starting material to the final product: 15‐19

1. Avoid formation of any waste. 2. Atom efficiency.

Introduction to sustainable oxidation chemistry

3. Elimination of hazardous or potentially dangerous reagents. 4. Benign methodologies have to lead to the formation of the desired product. 5. Auxiliary substances (solvents etc.) preferably should be avoided or innocuous if needed. 6. Energy efficient processes employing mild conditions are preferred. 7. Preferred use of renewables as raw materials. 8. Preferably one‐step synthesis to the desired product. Avoid use of additional protection/deprotection steps. 9. Catalytic rather than stoichiometric procedures. 10. Chemical products should be environmentally friendly and degradable. 11. Development of analytical methodologies for real time analyses are strongly desired. 12. Production processes and used substances should be safe by design. Numerous factors such as solvent, hazardous materials, number of reaction steps, waste formation per kilogram of product, overall yield, etc. have to be taken into account. 20,21 In this paragraph, these factors will be discussed in the context of oxidation chemistry.

1.1.3 Clean solvent To evaluate the sustainability of a process, besides oxidant and catalyst, also the use of solvent has to be considered. In view of the fact that 80% of all organic waste is due to solvent losses, the best solvent is no solvent. 22 Nevertheless, solvent–free technologies are noted to be poor in terms of mass transfer efficiency and rates and are only applied in special cases. Therefore, there is a strong need for green reaction media that meet the demands of excellent solvents, improve phase contact, miscibility, diffusion, size distribution, and allow high reaction rates. 23

Next to conventional solvents such as ethers and aromatics, water as green solvent is beginning to play a more important role. An even bigger impact is expected from non‐conventional solvents such as ionic liquids and supercritical fluids that are finding more and more applications based on their unique properties.24‐28

The ideal solvent for sustainable oxidation processes should be inert to air, does not undergo autoxidation, and does not coordinate to reactants, catalyst or products. Preferably, the solvent should be polar and stabilize polar intermediates. In practice the use of ionic liquids is accompanied by difficulties in the work‐up. More on this topic can be found in chapter 6.

Chapter 1

Generally speaking, formation of waste during any process should be judged according to the E factor. 29, 30 Solvents that are less toxic, less harmful and that can be easily recovered and recycled are the desired ones.

1.1.4 Recovery and recycling Ease of recovery of reagents and catalysts is important for two reasons: (1) complete removal of catalyst is necessary to prevent contamination of the product. (2) It allows the recycling of the catalyst. Heterogenization is a commonly applied procedure that allows easy and efficient regeneration of the catalyst. In addition it will also stabilize and disperse the catalyst. Many different supporting materials and techniques are available, such as silica, alumina, zeolite, carbon and polymers. 31 For example: cross‐linking is a technique that has been applied to heterogenize and stabilize enzymes. 32 In addition to the afore‐mentioned advantages, supports can also reduce the explosive properties of an oxidant or catalyst. In this thesis we have used iodine(III) precursors with a polymeric backbone, as support for an oxidant. Depending on the solvent and the oxidation state, the iodine polymer will precipitate and can be easily recycled via filtration. An iodine(III) oxidant, supported on a polymeric carrier, is produced in situ by oxidation with a peracid. The latter can be generated in situ via enzyme catalyzed reaction of the carboxylic acid with hydrogen peroxide (see chapter 5 for more details).

1.2 Alcohol and alkene oxidation In this paragraph two classes of reactions will be outlined with in more detail: alcohol oxidation and alkene oxidation. A variety of examples will be listed that are representative of the state‐of‐ the‐art in green oxidation chemistry.

1.2.1 Alcohol oxidation Oxidation of primary and secondary alcohols to their corresponding carbonyl compounds is widely used in the fine chemicals industry since aldehydes and ketones are the key raw materials for a wide range of pharmaceutical intermediates, fragrances and flavors. 1,5,33,34 Green chemistry approaches are of pivotal importance here because organic chemistry textbooks still recommend the stoichiometric and toxic i.e. Jones’ reagent (chromium(VI)) as the preferred oxidant. 35,36

For oxidation of alcohols a number of metal‐based catalytic systems have been developed which allow the use of air as the oxidant. In this area, Pd, Pt, Ru, Cu, Au have delivered the best examples of promising catalysts. For this we refer to the excellent reviews of Dijksman et al. 37 and Muzart et al. 38 Two examples that show a wide substrate scope will be treated here. The first one is the example of ruthenium compounds. 39‐41

Introduction to sustainable oxidation chemistry

Aerobic oxidation of activated alcohols such as allylic alcohols is already achieved with high yields under mild conditions using simple ruthenium precursors such as RuCl3 and RuCl2(PPh3)3 + ‐ 42 as catalysts and ionic liquids as solvents [BMIm] [PF6] under aerobic conditions. Tetrapropylammonium perruthenate (TPAP), is known as one of the best homogeneous ruthenium catalysts, albeit using 4‐methylmorpholine N‐oxide (NMO) as the oxidant 33,43,44 Notably, heterogeneous immobilized TPAP showed high performance as well in oxidation as in recycling. 45,46

Another important example is the combination of RuCl2(PPh3)3 with a nitroxyl radical 2,2’,6,6’‐ tetramethylpiperidinyl‐1‐oxy (TEMPO). High selectivity and yields are obtained, and overoxidation does not occur (Figure 1.3). 47,48

Figure 1.3 Ruthenium/TEMPO catalyzed oxidation of alcohols 49,50

The second example of aerobic oxidation of alcohols is the use of Pd(II) as the catalysts.

Palladium‐based catalysts such as PdCl2, Pd(OAc)2 with addition of base have been intensively studied. Thirty years ago, the PdCl2‐NaOAc system was employed in the oxidation of secondary alcohols using ethylene carbonate as solvent. 51

Further on there are similar examples with sodium carbonate as a base and Adogen 464® 52 as Phase Transfer Catalyst (PTC), which in chlorinated media are capable to oxidize diols into their

53 lactones. Another example concerns the use of Pd(OAc)2 in combination with sodium bicarbonate, employing DMSO or pyridine/toluene as reaction medium. In general TOFs in these systems are low. 54‐56 The breakthrough in this area was achieved by ten Brink et al., 57 using a water soluble palladium(II) bathophenanthroline as recyclable and stable catalyst (Figure 1.4). 58

Chapter 1

Figure 1.4 Oxidation of functionalized alcohols in water using palladium(II) bathophenanthroline as catalyst

Additional examples are available for primary alcohols, where the palladium(II) bathophenanthroline complex is used in combination with the nitroxyl radical TEMPO to prevent oxidation of the aldehyde to acid. 58 From recent developments, Pd nanoclusters turned out to be alternative catalysts for secondary alcohol oxidation reaching a turnover frequency (TOF) of 70 per hour. 59

Nowadays many nanocluster catalysts are available, with notable examples for Au and Pt. Gold nanoclusters in aerobic alcohols oxidation afford high activity, with TOF of 960 per hour, albeit at high temperatures (160 oC). 60 Pt nanoparticles dispersed in amphiphilic polystyrene‐ polyethylene glycol resin (ARP‐Pt) have been reported as stable and recyclable heterogeneous catalysts for aerobic oxidation for a variety of alcohols in water. 61

Besides metal‐based oxidation systems for aerobic oxidation of alcohols, the use of homogeneous TEMPO as organocatalyst is well known. In this case a stoichiometric amount of a terminal oxidant e.g. sodium hypochlorite, meta‐chloroperbenzoic acid, sodium bromite,

62 sodium chlorite, oxone or O2/CuCl is required. Homogeneous TEMPO, despite its high performance, still suffers from a lack of recyclability.

A remarkable improvement was achieved by using a recyclable, oligomeric form of TEMPO denoted as Polymer Immobilized Piperidinyl Oxyl (PIPO). 63 In combination with hypochlorite solution, PIPO was applied for alcohol oxidation in the absence of organic solvents. Also in contrast to TEMPO, PIPO did not require bromide as co‐catalyst. PIPO can be employed with oxygen as the oxidant, and with Cu as co‐catalyst. The major limitation of the PIPO‐CuCl system is its inability to oxidize olefinic alcohols, which undergo chlorination of the double bond during the catalysis. The advantage of organocatalysts is their high selectivity and functional group tolerance.

Introduction to sustainable oxidation chemistry

Another class of transition metal–free oxidation procedures is that employing iodine(III) or iodine(V) as efficient recoverable hypervalent reagent. 64‐66 This topic will be studied extensively in this thesis and will be covered in chapters 3 and 4.

1.2.2 Alkene epoxidation Epoxides are key intermediates in organic chemistry. They react easily with a range of nucleophiles, leading to stereospecific formation of C‐C, C‐N, C‐S and C‐O bonds. Epoxides can be readily obtained by oxidation of alkenes. 67,68 One common procedure in fine chemistry is the electrophilic addition reaction of peroxycarboxylic acids ‐ so called peracids ‐ pioneered by Prilezhaev. Commonly, Stoichiometric amounts of m‐CPBA are commonly used in chlorinated solvent (Figure 1.5). 69

Figure 1.5 Alkene epoxidation with peracids

By increasing the electron density of the double bound of the alkene the rate of epoxidation is enhanced. For example, when stoichiometric amounts of peracid are used, the most substituted double bond in dienes will be epoxidized preferably. Terminal alkenes or electron deficient olefins react slowly, due to their low‐electron density. Non‐coordinating chlorinated or aromatic solvents favour the reaction. Epoxidation occurs in a stereospecific manner, which means that if the alkene has the trans‐configuration, this configuration is retained in the product.

The epoxidation of olefins can also be carried out by transition metal elements such as: (Mo, W, V, Ti, Mn, Fe, Ru, Re) and selenium organocatalysts in combination with a variety of oxidants. The field of homogeneous epoxidation catalysts was initiated in the seventies by Mimoun et al., 70 with molybdenum diperoxocomplexes. The exact mechanism is still a subject of debate. One mechanism proposal given by Sharpless et al., 71 focuses on concerted oxygen transfer, while the mechanism proposed by Mimoun et al.,70 emphasizes the stepwise cycloinsertion pathway. This mechanism hence is the basis for all late transition metal based reactions that have been

Chapter 1 published with Ti, V, Mo and W and is related to the peracid mechanism. Other metals follow the oxo‐pathway (Figure 1.2), which is often accompanied by side‐reactions and C‐C cleavage. For a complete overview the reader is referred to many excellent reviews. 13,72

Oxidation and epoxidation of olefins using hydrogen peroxide in combination with FeIII as nanocatalyst, was described by Beller et al. 73 It would however be very interesting to use cheap and simple iron as the catalyst. A homogeneous iron catalyst system using peracetic acid as an oxidant was presented by Dubois et al. 74 They presented promising conversion and selectivity for a wide range of substrates. However, in our hands, the high rates from literature could not be reproduced, and only low yields were obtained. For crystal structure of the intermediate iron species observed in their reaction is denoted in Figure 1.6.

III 4+ Figure 1.6 Active homogeneous epoxidation catalyst of [((phen)2(H2O)Fe )2(µ‐O)]

This example demonstrates the importance of peracid oxidant used in reactions. However, a significant disadvantage when working with peracids, is their explosive properties, which prohibits their application on large scale. In situ generation of peracid would solve this. An elegant method to generate peracid in situ from acid and H2O2, is based on lipase enzyme as catalyst. In order to introduce this topic, first a short introduction to the use of lipases in biocatalysis is given.

1.3 Lipases Lipases are enzymes that belong to the class of serine hydrolases. Their natural function is to catalyze the hydrolysis of triglycerides into the corresponding fatty acids and glycerol (Figure 1.7).

Introduction to sustainable oxidation chemistry

Figure 1.7 Biological reaction of lipases

Lipases have found application in industry mainly due to their high stability in non‐aqueous media. Moreover, lipases accept a wide variety of unnatural nucleophiles (as illustrated in Figure 1.8). 75

R1 OH

O R1 NHXH R1 O 2 NH OH, R 2 H O O 2 R2-OH O NH2NH2, etc.

hydroxylamine, R1 O hydrazine and primary and secondary derivatives O alcohols 2 3 Ser HNR R hydrogen peroxide alkyl amines NH3 H 2O 2 R2 ammonia R1 N R 1 O H R3 O 1 O R NH2 O O

Figure 1.8 Lipases: acyl accepting compounds

Lipases are widely applied in commercial synthesis of optically active compounds; they are used as detergent enzymes, in the pharmaceutical and specialty food industry. They also show high potential in medical application such as in the treatment of digestive disorders and pancreatic diseases. 76

Non‐immobilized lipases are soluble in water, and they remain active in the oil‐water interface. They can easily catalyze hydrolysis of long‐chain, insoluble triglycerides and other insoluble esters of fatty acids. Therefore in contrast to esterases, which are restricted to water soluble substrates, lipases are highly desired because of their versatility. The acyl binding site of lipases has a shape of a shallow groove, with an optimum fit for esters of linear, non‐branched carboxylic acids. The acid moiety should consist of three to twelve carbon units to ensure enough hydrophobic character. In most lipases a lid consisting of one or more alpha helices, covers the catalytic site in the inactive form of the enzyme. During the contact of the lid with the interface, the lid undergoes a conformational rearrangement making the active site

Chapter 1 accessible to the substrate. 77 A recent approach where free lipase was dissolved in ionic liquids

78 and supercritical CO2, proved a high stability of the enzyme in benign media.

The yeast Candida antarctica produces two different lipases A and B (CaLA or CaLB). Both lipases exhibit high stability over a wide pH range, with the pH optimum at 7. Compared to CaLB which is classified as thermostable, CaLA is even extremely thermostable. Both lipases in immobilized form are stable against non‐natural i.e. non‐aqueous conditions. They can be used at 60 ‐ 80 οC without significant loss in activity. 79

Regarding the mechanism of lipases, a catalytic site has been identified known as the Serine‐ Histidine‐Aspartate (Ser‐His‐Asp) triad (Figure 1.9). 80

Figure 1.9 Mechanism of lipase catalyzed acyl transfer within the active site

By nucleophilic attack of the hydroxyl group of a (Ser) residue on an ester moiety, a tetrahedral intermediate is formed, which is assisted by the His and Asp group. As a result the negatively charged carbonyl oxygen is stabilized by hydrogen bonding with the peptide backbone NHs of the so‐called oxyanion hole. 81 In the following step release of the R’OH group enables formation of the acyl enzyme intermediate. Final formation of the product and regeneration of an enzyme is achieved by subsequent reaction of the acyl enzyme intermediate with the nucleophile. In the case of hydrolysis, the nucleophile is water. In the absence of water, any nucleophile can react with the acyl enzyme intermediate, affording a number of useful transformations. 75,82 Also hydrogen peroxide can act as acyl acceptor, leading to the equilibrium formation of peracids from either acids or esters. 83 In situ generation of peracid leads to an oxidation procedure for alkene epoxidation. 84 At the same time the acid is continuously regenerated, without loss of turnover activity in the total cascade process (Figure 1.10).

Introduction to sustainable oxidation chemistry

Figure 1.10 Epoxidation using the lipase/acid/H2O2 methodology

The same lipase driven cascade methodology will also be applied in the Baeyer‐Villiger reaction, where novel media allow homogeneous catalysis to occur. This subject will be studied in more detail in chapter 6 of this thesis.

1.4 Scope of the thesis The PhD work described in this thesis entitled “Iodinee and Lipas Based Oxidation Technology” is the result of collaboration between academia and industry within the Integration of Biocatalysis and Organic Synthesis (IBOS) program and Advanced Catalytic Technologies for Sustainability (ACTS) program. The project was entitled “Catalytic cascade reactions for selective oxidation with benign oxidants”. A characteristic feature of the IBOS program is the active input from the industrial partners, Organon (now MSD) and Givaudan in our case.

The objective of the program was the development of a mild and selective bromide‐free method for the oxidation of pharmaceutical intermediates, which has a high functional group tolerance, and which uses lipase and iodine(III) and hydrogen peroxide as the primary oxidant.

In this way the selective nature of iodine(III) as oxidant could be fully used, while circumventing the need for stoichiometric oxidants.

Outline of the thesis:

In the second chapter an introduction to hypervalent iodine chemistry will be given, followed by examples of synthetic applications.

In chapter 3 the focus is on the oxidant. Different polystyrene‐derived iodine materials are synthesized and characterized. Their oxidation with hydrogen peroxide results in the desired iodine(III) polystyrene materials. The effects of polymer support, solvent and safety aspects are investigated for two representative alcohol oxidations.

Chapter 1

Chapter 4 of the thesis describes the optimized procedure for the selective oxidation of alcohols employing polymeric iodine(III) reagent as an oxidant. Next to the scope of the substrates, which can be oxidized, various new derivatives of the oxidant are introduced. Recycling and recovery of iodine(III) reagent is discussed. In addition chapter 4 presents mechanistic studies on the oxidation process with the polymeric iodine(III) oxidant.

In chapter 5 the possibility to generate iodine(III) in situ, using the H2O2/lipase/acid procedure is studied for selected alcohols. The limitations and future directions of the method are investigated.

Chapter 6 describes a novel solution for chemo‐catalytic cascade for epoxidation and Baeyer‐ Villiger oxidation in new generations of ionic liquids as designer media.

Introduction to sustainable oxidation chemistry

1.5 References

1R. A. Sheldon, J. K. Kochi, Metal–Catalyzed Oxidations of Organic Compounds, Academic Press, New York, 1981. 2G. Franz, R. A. Sheldon, Ullmann’s Encyclopedia of Industrial Chemistry, (Eds.), Wiley‐VCH, Weinheim, 1991, 261. 3R. A. Sheldon, R. A. van Santen, Catalytic Oxidation: Principles and Applications., (Eds.), World Scientific, Singapore, 1995. 4J. A. Horward, Adv. Free‐Radical Chem., 1972, 4, 49. 5R. A. Sheldon, I. W. C. E. Arends, U. Hanefeld, Green Chemistry and Catalysis., (Eds.), Wiley‐ VCH, Weinheim, 2007 6R. A. Sheldon, Dioxygen Activation and Homogeneous Catalytic Oxidation, Elsevier Sci., Publ. Amsterdam, 1991, 573. 7R. A. Sheldon, Top Curr. Chem., 1993, 164, 21. 8W. Zhang, J. L. Loebach, S. R. Wilson, E. N. Jacobsen, J. Am. Chem. Soc., 1990, 112, 2801. 9R. Irie, K. Noda, Y. Ito, N. Matsumoto, T. Katsuki, Tetrahedron Lett., 1990, 31, 7345. 10H. Sasaki, R. Irie, T. Hamada, K. Suzuki, T. Katsuki, Tetrahedron 1994, 50, 11827. 11J. P. Collman, L. Zeng, H. J. H. Wang, A. Lei, J. I. Brauman, Eur. J. Org. Chem., 2006, 2707. 12J. R. Linsay Smith, D. N. Mortimer, J. Chem. Soc. Chem. Commun., 1985, 410. 13S. T. Oyama, “ Mechanism in Homogeneous and Heterogeneous Epoxidation Catalysis”. (Eds.), Elsevier Sci. Publ. Amsterdam, 2008, 26. 14P. T. Anastas, J. C. Warner, (Eds.), Green Chemistry, Theory and Practice, Oxford University Press, Oxford, 1998. 15P. T. Anastas, M. M. Kirchhoff, Acc. Chem. Res., 2002, 35, 686. 16P. T. Anastas, T. C. Williamson, (Eds.), Green Chemistry: Frontiers in Chemical Synthesis and Processes, Oxford University Press, Oxford, 1998. 17P. T. Anastas, I. G. Heine, T. C. Williamson (Eds.), Green Chemical Synthesis and Processes, ACS, Washington DC, 2000. 18P. T. Anastas, L. B. Bartlett, M. M. Kirchhoff, T. C. Williamson, Catal. Today, 2000, 55, 11. 19A. D. Curzons, D. J. C. Constable, D. N. Mortimer, V. L. Cunningham, Green Chem., 2001, 3, 1. 20D. J. C. Constable, A. D. Curzons, Green Chem., 2001, 3, 7. 21R. A. Sheldon, Green Chem., 2005, 7, 267. 22A. S. Matlack, Introduction to Green Chemistry, Marcel Dekker Inc.: New York (NY) USA, 2001. 23P. Walden, Molecular “Weights and electrical Conductivity of several Fused Salts” Bull. Acad. Sci., (St. Petersburg), 1914, 405. 24J. S. Wilkes Green Chem., 2002, 4, 73. 25P. G. Jessop, W. Leitner, Chemical Synthesis using Supercritical Fluids, (Eds.), Wiley‐VCH, Weinheim, 1999. 26G. Brunner, Gas Extraction, Topics in Physical Chemistry. Springer Darmstadt, Germany, 2003, 4. 27E. J. Beckman, J. Supercrit. Fluids, 2004, 28, 121. 28Federal Register, Notices, 1997, 62, 67377, electronic release. 29K. R. Seddon, P. Wasserscheid, Green Chem., Special issue on Green Solvents for Catalysis, (Eds.), W. Leitner, 2005, 5, 99. 30R. A. Sheldon, Green Chem., 2007, 9, 1273. 31I. Yu., Yermakov, B. N. Kuznetsov, V. A. Zakhanarov, Stud. Surf. Sci. Cat., 1999. 32L. Cao, F. van Rantwijk, R. A. Sheldon, Org. Lett., 2000, 2, 1361; P. Lόpez‐Serrano, L. Cao, F. van Rantwijk, R. A. Sheldon, Biotechnol. Lett., 2002, 24, 1379. 33S. V. Ley, J. Norman, W. P. Griffith, S. P. Marsden, Synthesis, 1999, 639.

Chapter 1

34M. Hudlicky, Oxidation in Organic Chemistry, ACS, Washington DC, 1990. 35J. Clayden, N. Greeves, S. Warren, P. Wothers, Organic Chemistry, (Eds.), Oxford University Press, Oxford, 2001. 36G. Cainelli, G. Cardillo, Chromium Oxidants in Organic Chemistry., Springer, Berlin, 1984. 37A. Dijksman, I. W. C. E. Arends, R. A. Sheldon, Plat. Met. Rev., 2001, 45, 15. 38J. Muzart, Tetrahedron, 2003, 59, 5789; J. Muzart, Chem. Assian J., 2006, 1, 508. 39T. Naota, H. Takaya, S.‐I. Murahashi, Chem. Rev., 1998, 98, 2599. 40R. A. Sheldon, I. W. C. E. Arends, A. Dijksman, Catal. Today, 2000, 57, 158. 41I. W. C. E. Arends, R. A. Sheldon, Modern Oxidation Methods., J. Bäckvall (Eds.), Wiley‐VCH, Weinheim, 2004, 83. 42A. Wolfson, S. Wuyts, D. E. de Vos, I. F. J. Vancelecom, P. A. Jacobs, Tetrahedron Lett., 2002, 43, 8107. 43I. E. Marko, P. R. Giles, M. Tsukazaki, I. Chellé‐Regenaut, C. J. Urch, S. M. Brown, J. Am. Chem. Soc., 1997, 119, 12661. 44R. Lenz, S. V. Ley, J. Chem. Soc., Perkin Trans. 1, 1997, 3291. 45A. Bleloch, B. F. G. Johnson, S. V. Ley, A. J. Price, D. S. Shephard, A. W. Thomas, Chem. Commun., 1999, 1907. 46M. Paliaro, R. Ciriminna, Tetrahedron Lett., 2001, 42, 4511. 47A. Dijksman, I. W. C. E. Arends, R. A. Sheldon, Chem. Commun., 1999, 1591. 48A. Dijksman, I. W. C. E. Arends, R. A. Sheldon, Plat. Met. Rev., 2001, 45, 15. 49A. Dijksman, I. W. C. E. Arends, R. A. Sheldon, in Supported Catalyses and their Applications. D.C. Sherrington, A. P. Kybett, (Eds.), The Royal Society of Chemistry RSC, Cambridge, 2001, 118. 50A. Dijksman, I. W. C. E. Arends, R. A. Sheldon, Synlett., 2001, 102. 51T. F. Blackburn, J. Schwartz, J. Chem. Soc. Chem. Commun., 1977, 157. 52Adogen 464 is a registered trademark of Ashland Chemical for methyltrialkyl (C8‐C10)‐ ammonium chloride. 53S. Alt‐Mohand, F. Hénin J. Muzart, Tetrahedron Lett., 1998, 39, 6011. 54K. P. Peterson, R. C. Larock, J. Org. Chem., 1998, 63, 3185. 55T. Nishimura, T. Onoue, K. Ohe, S. Uemura, Tetrahedron Lett., 1998, 39, 6011. 56T. Nishimura, T. Onoue, K. Ohe, S. Uemura, J. Org. Chem., 1999, 64, 6750. 57G‐J. ten Brink, I.W. C. E. Arends, R. A. Sheldon, Science, 2000, 287, 1636. 58G‐J. ten Brink, I.W. C. E. Arends, R. A. Sheldon, Adv. Synth. Catal., 2002, 344, 355. 59A. Biffis, L. Minatti, J. Catal., 2005, 236, 405. 60A. Biffis, S. Cunial, P. Spontoni, L. Prati, J. Catal., 2007, 236, 405. 61Y. M. A. Yamada, T. Arakawa, H. Hocke, Y. Uozumi, Angew. Chem. Int. Ed., 2007, 46, 704. 62M. F. Semmelhack, C. R. Schmid, D. A. Cortés, C. S. Chou, J. Am. Chem. Soc., 1984, 106, 3374. 63A. Dijksman, I. W. C. E. Arends, R. A. Sheldon, Chem. Commun., 2000, 271. 64Z‐Q. Z. Y. Liu, H. Luo, L. Chai, S. Qiuju, Tetrahedron Lett., 2007, 48, 3017. 65H. Tohma, T. Maegawa, S. Takizawa, Y. Kita, Adv. Synth. Catal., 2002, 344, 328. 66T. Dohi, J. Pharm. Soc. Japan, 2006, 126, 757. 67E. J. Corey, M. Chaykowsky, J. Am. Chem. Soc., 1962, 84, 867. 68E. J. Corey, M. Chaykowsky, J. Am. Chem. Soc., 1965, 87, 1353. 69Ulmann’s Encyclopedia of Industrial Chemistry, 2007, (Eds.), Wiley‐VCH, electronic release. 70H. Mimoun, I. S. D. Roch, L. Sajus, Tetrahedron, 1970, 26, 37. 71K. B. Sharpless, D. R. Williams, J. M. Townsend, J. Am. Chem. Soc., 1972, 94, 295. 72I. W. C. E. Arends, Angew. Chem. Int. Ed., 2005, 45, 6250. 73G. Anilkumar, B. Bitterlich, F. G. Gelalcha, M. K. Tse, M. Beller, Chem. Commun., 2007, 289. 74G. Dubois, A. Murhpy T. D. P. Stack, Org. Lett., 2003, 5, 2469; M. E. J. Pepels, “Chemo‐ Enzymatic cascade for alkene epoxidation using hydrogen peroxide as a benign oxidant”, MSc project TU Delft, 2005.

Introduction to sustainable oxidation chemistry

75M. de Zoete “Lipase and Protease‐Catalyzed Transformation with Unnatural Acyl Acceptors”, TU Delft, PhD thesis, 1995. 76R. D. Schmid, R. Verger Angew. Chem. Int. Ed., 1998, 37, 1609. 77R. Madeira Lau, “Conversion in Ionic Liquids and Resolution of Amines”, TU Delft, PhD thesis, 2003. 78M. D. Bermejo, M. Montero, E. Saez, L. J. Florusse, A. J. Kotlewska, M. J. Cocero, F. van Rantwijk, C. J. Peters, 2007. Book of abstracts V International Symposium on High Pressure Process Technology and Chemical Engineering Segovia: sn. Proceedings, 2007, 90. 79O. Kirk, M. W. Christensen, Org. Proc. Res. Dev., 2002, 6, 446. 80J. Pleiss, M. Fisher, R. D. Schmid, Chemistry and Physics of Lipids, 1998, 93, 67. 81T. Ema, Curr. Org. Chem., 2004, 8, 1009. 82K. Faber, “Biotranformation in Organic Chemistry”. (Eds.), Springer, Berlin, 2000. 83F. Bjorkling, H. Frykman, S. E. Godtfredsen, O. Kirk, Tetrahedron, 1992, 48, 4587. 84O. Kirk, M. W. Christensen, Organic Proc. Res. Dev., 2002, 6, 446.

CChhaapptteerr 22

Hypervalent iodine in organic synthesis

2.1 Introduction 2.2 PhIO 2.3 PhIX2 2.4 PhI(OAc)2 2.5 Polymer supported iodine(III) 2.6 Iodine(III) reagent and ionic liquids 2.7 Iodine(V) vs. Iodine(III) 2.8 Concluding remarks 2.9 References

Abstract Hypervalent iodine is the oldest known oxidant in organic synthesis and is still the subject of studies. However, due to the explosive properties of iodine(V) there is a significant room for improvement in terms of safety. The use of iodine(III) is preferred over that of iodine(V). This chapter presents an overview of the broad range of possible transformations, using both iodine(III) as well as iodine(V) as reagents.

Chapter 2

2.1 Introduction Polyvalent iodine compounds gained importance when it became apparent that chemical properties and reactivity of iodine(III) compounds are most likely similar to those of heavy metals like (HgII, TlIII, PbIV). The use of the environmentally friendly iodine(III) can thus overcome the use of these toxic metals. Polyvalent iodine species differ in the number of valence electrons (N) that surround the central iodine atom, as well in the number of anions (L) and their chemical nature. We can distinguish up to four characteristic types of polyvalent iodine (Figure 2.1). 1

Figure 2.1 Traditional classification of polyvalent iodine compounds Used code: 8‐I‐2 (8 corresponds to N = number of valence electrons; I=iodine; 2 corresponds to L=number of anions

The first two types correspond to trivalent iodine(III) and the last two belong to the well‐known pentavalent iodine(V) species. In Figure 2.2 the most frequently applied iodine(III) oxidants are presented.

Figure 2.2 Iodine(III) oxidants

Nowadays hypervalent iodine compounds are mainly associated with strong oxidizing reagents. However, they can be used for a wide array of synthetic applications in organic chemistry (Table 2.1).2‐10

Hypervalent iodine in organic synthesis

Table 2.1 Hypervalent iodine reagents in organic synthesis 4

Synthetic applicaion of hypervalent iodine compounds

1 C-C bond formation reactions 2 Intermolecular reaction 3 Cyclizations 4 Transformation of C-C single bond into C=C double bonds 5 C-Heteroatom bond forming reactions 6 Addition reactions to arenes 7 Addition reactions to alkenes 8 Reactions by activation of heteroatom 9 Reactions by activation of carbon atom 10 Oxidation reactions 11 Rearrangements 12 Heteteroatom-heteroatom bond forming reactions 13 Nucleophilic subsitution reactions

2.2 PhIO – iodosobenzene

Iodine(III) oxidants such as iodosoarenes (PhIO)n display a high diversity in terms of functional properties. As demonstrated by numerous spectroscopic studies the structure of iodosobenzene is a polymeric zigzag in which monomeric units of PhI+‐O‐ are linked by intramolecular I‐O secondary bonds.11‐13 The polymeric and amorphous character of

iodosoarenes hampers their solubility in most conventional media. (PhIO)n belongs to the first discovered iodine reagents and can easily undergo useful transformations (Figure 2.3). 14, 15

Figure 2.3 Iodosylarenes as precursor to iodine(III) derivatives

Naked polymeric iodosoarenes are usually applied in the presence of a catalyst. The main role of the catalyst (Lewis acid, bromide anion, transition metal complex, or hydroxylic solvents) is to

depolymerize (PhIO)n and afford active monomeric oxidant in the solution. Moriarty, Prakash and co‐workers reported that iodosobenzene in the oxidation of trimethylsilyl ketene acetals of esters and lactones in methanol as a reaction medium gives the corresponding α‐methoxylated carbonyl compounds in good yields 16, 17 as illustrated in the Figure 2.4.

Chapter 2

Figure 2.4 (PhIO)n as oxidizing agent of trimethylsilyl ketene acetals of esters and lactones

Another interesting application of I(III) from the same authors is the oxidation of dihydropyran, cyclohexene and styrene under mild conditions leading to rearrangement products. Use of water in the dihydropyran oxidation with iodosobenzene gave tetrahydro‐2‐furaldehyde via cationic ring contraction (Figure 2.5). 17

Figure 2.5 (PhIO)n in aqueous oxidation of dihydropyran

As emphasized earlier, the reactivity of iodosoarenes is remarkably improved by the presence of various additives. In the work of Kita, Tohma and co‐workers it has been demonstrated that catalytic amounts of bromide salts or quaternary ammonium bromides improved the efficiency of PhIO. 18

In the oxidation of various sulfides, the best results were obtained using 10 mol% cetyl‐ trimethyl ammonium bromide (CTAB) as the catalyst and non‐coordinating solvents such as: toluene, methylene chloride or hexane, with traces of water (Figure 2.6).17

Figure 2.6 Oxidation of sulfides with (PhIO)n/CTAB

The same authors have reported an efficient alcohol oxidation system, where catalytic amounts of KBr were required to activate polymeric iodosoarene in water (as displayed in Figure 2.7).

Hypervalent iodine in organic synthesis

Figure 2.7 Oxidation of alcohols with (PhIO)n/KBr in water

In the oxidation of primary alcohols, acids were obtained as the only product. 18 Kita and co‐ workers proposed an activation mechanism involving binding of Br‐ to iodine (Figure 2.8).

Figure 2.8 Depolymerization mechanism of the (PhIO)n

Alcohol oxidation in aqueous methanol instead of water as media leads to the formation of methyl esters with good to high yields, albeit using excess oxidant (Figure 2.9). 19, 20

Figure 2.9 Oxidation with PhIO/KBr in aqueous methanol

Besides the direct use of PhIO as oxidant, iodine(III) has also found application as stoichiometric oxidant in combination with metals as catalyst. Remarkably high selectivities were obtained in the asymmetric epoxidation of alkenes using Cr, Fe, Rh, Mn and Cr‐based catalysts and also examples are given for hydroxylation of hydrocarbons. 21‐25

It was shown by the group of Breslow that the 6α carbon in an androstanediol derivative can be hydroxylated using the Mn‐iodine(III) system with high regioselectivity (Figure 2.10). The authors concluded that the regioselectivity in this case was driven by the geometry of the catalyst‐substrate interactions. A disadvantage is that often methylene chloride is applied as reaction solvent. 26‐30

Chapter 2

Figure 2.10 Selective hydroxylation of steroids catalyzed by Mn‐porphyrin using PhIO as oxidant

2.3 PhIX2 – iodobenzene dihalogenated 31 The iodine(III) reagent, PhICl2, was reported already in 1886 by Willgerodt. Iodoaryl halides

ArIX2 (X= Cl, F, Br) can be synthesized from iodosoarenes PhIO with the corresponding acid HX, or via direct halogenation of aryl iodides. A notable example for iodobenzene dichloride, where it is generated in situ is presented in Figure 2.11. In this case a solid state reaction of tetraphenylene to its epoxide was carried out in a hydrochloride treated silica gel. Prior starting the reaction an additional pulverization (grinding) of the powder was applied in order to increase the surface activity. Within a similar approach organic sulfides are converted to sulfonyl chloride or sulfones.32

Figure 2.11 In situ formation of active PhICl2 species

The oxidation of para‐substituted phenol with PhICl2 in the presence of a nucleophile affording 4,4‐bisubstituted cyclohexadienones, is a synthetically useful transformation (Figure 2.12 ). 1,33,

Hypervalent iodine in organic synthesis

Figure 2.12 Oxidation of para‐substituted phenol to 4.4‐bisubstituted cyclohexadienones

35 For bromide derivatives of the general formula PhIBr2 only few examples are known. This is due to their low stability. PhIF2 suffers from difficult and unefficient preparation procedures and its use is hampered by its low stability.1, 20, 36‐39 A possible synthesis involves reaction of iodobenzene dichloride with aqueous solutions of HF and HgO. Nevertheless, the use of iodobenzene difluoride next to iodobenzene dichloride, is well established in the formation of carbon‐heteroatom and heteroatom‐heteroatom bonds. For example selective introduction of fluorine under mild conditions is possible. The fluorination of sulphur or selenium‐substituted esters affords the monofluorinated product in high yields (Figure 2.13).

Figure 2.13 Formation of monofluorinated esters or amides with PhIF2

Another interesting example of fluorination with iodine(III) benzene is presented in the reaction of steroidal dienes, where the fluorinated product is obtained with high regioselectivity and stereoselectivity (Figure 2.14). 40

Figure 2.14 Fluorination of steroidal dienes with ArIF2

Chapter 2

Similarly iodobenzene dichloride can be used for selective chlorination, mostly following an ionic pathway, due to the electrophilic character of iodine. 1 There are numerous examples for

3 the use of PhICl2 in substitutive chlorination at sp ‐carbon of various alkanes, ethers, esters thioesters, ketones, sulfoxides. 3, 5, 41‐51 Interestingly, depending on the reaction conditions,

PhICl2 may follow a radical or an ionic mechanism. In the first case i.e. chlorination of alkenes, products of 1,2‐additions are obtained, while under polar conditions rearrangement products

49 are formed. In the example of chlorination of 1,5‐diketones with PhICl2 this effect can be clearly observed (Figure 2.15). 47

Figure 2.15 Chlorination via two pathways using PhICl2

Similarly to PhIO iodosoarenes, PhIX2 iodoaryl halides are mainly applied in combination with chlorinated solvent, which in the present green development makes them less applicable.

2.4 PhI(OAc)2 iodobenzene diacteate Phenyliodine(III) carboxylates (Figure 2.2) have found broad application in organic synthesis.

PhI(OAc)2 (commonly abbreviated as DIB=DAIB=BAIB=PIDA), and (bis(trifluoroacetoxyiodo) benzene)) PhI(OCOCF3)2 (BTI=PIFA) are the most well‐known compounds of this class. They are cream‐white crystalline solids showing poor stability against air and light and should preferably be stored in the dark and at lower temperatures. Generally, they are prepared by oxidation of iodobenzene with peroxyacetic acid in glacial acetic acid as a solvent. 52,,53 However, there are novel synthetic procedures using sodium periodate as an oxidant, that also lead to the

54 formation of PhI(OAc)2.

ArI(OAc)2 is also the starting material for the synthesis of fluorinated derivatives (BTI) or ArI(OH)OTs)) [hydroxy(tosyloxy) iodo] arenes. 55‐70

The major application of DAIB is in oxidation reactions. 18,71 Piancatelli et al., have reported the use of the nitroxyl radical TEMPO using DAIB as the oxidant for the catalytic oxidation of

Hypervalent iodine in organic synthesis primary alcohols to aldehydes. Hence, the commonly used oxidant for TEMPO, NaOCl, is replaced by DIB, thereby avoiding the use of chlorine (Figure 2.16). 2,72,73

Figure 2.16 Primary alcohols oxidation by DAIB/TEMPO system

(Diacetoxyiodo) benzenes are often used in phenol oxidation. Especially DIB is widely applied in the preparation of various substituted ortho‐ and para‐hydroquinones from phenols or anilines, giving nearly quantitative yields. The best solvent for this reaction is methanol, which is also the alcohol reactant, and the reaction occurs at ambient temperature (Figure 2.17). 57,74,75

Figure 2.17 Oxidation of phenols with PhI(OAc)2

DAIB shows high compatibility in oxidation of sulfides with sulfonamide as co‐reactant to the direct transformation to the corresponding N‐sulfonylsulfimines. 76 Despite the lack of detailed mechanistic studies over the formation of double bonds, a heteroatom attached to the single bond is required. Iodosobenzene diactate showed high performance in the oxidation of trisubstituted pyrazolines to the pyrazoles as depicted in Figure 2.18. 77

Figure 2.18 Oxidation to heteroaromatic compounds by PhI(OAc)2

Carbonyl compounds such as ketones can undergo functionalization in the α‐position. Among known hypervalent iodine reagents (IBX, PhIO), the PhI(OAc)2 shows the best performance when it is used in the presence of strong base and alcohol to obtain α‐hydroxylated dialkylacetals. These can then be hydrolyzed under protic conditions to α‐hydroxy ketones. 16, 78, 79 This approach is widely applied in the functionalization of natural products. 80, 81, 82 For an example see Figure 2.19.

Chapter 2

Figure 2.19 Functionalization of carbonyl compounds in α‐position by PhI(OAc)2

A useful chemical property of polyvalent iodine reagents is their natural ability to react as an electrophile and if necessary to be converted to the leaving group. The latter is successfully applied in a diversity of rearrangement reactions in order to build highly functionalized compounds. As a model example Hoffman‐type rearrangements can be given. Direct cyclization of aromatic amides and formation of heterocyclic compounds can be achieved when a nucleophilic substituent is present in the ortho‐position (Figure 2.20). 83

Figure 2.20 Hoffman‐rearrangements with PhI(OAc)2

Zheng et al. reported an efficient and simple method for the halogenation of 6‐methyluracil derivatives. 84 The halogenation system is based on [bis(acyloxoiodo)benzene] and iodine. Mixing all reactant in appropriate media and stirring at ambient temperature was required to obtain the halogenated product. The reaction occurs via in situ formation of an active acyl hypoiodides species (Figure 2.21). 85

Figure 2.21 Halogenation by PhI(OAc)2 – halogen system under mild conditions

Similarly to iodosoarenes, [bis(diacetoxyiodo) arenes] are used in combination with transition metals such as Mn or Rh and they show promising feature in amination reactions. One of the

Hypervalent iodine in organic synthesis feasible one‐pot systems where polyvalent iodine is used in combination with a manganese porphyrin complex is an azidation reaction of alkenes or amidation of the benzylic position of hydrocarbons i.e. indane (Figure 2.22). 86

Figure 2.22 Activation of C‐H bonds in polyvalent iodine mediated amination

Recently Du Bois and co‐workers reported an example where Rh catalysts together with

PhI(OAc)2 were used for a C‐H insertion reaction, namely the one–pot amination procedure of sulfamate esters and carbamates (Figure 2.23). 87

Figure 2.23 Activation of a C‐H bond in rhodium‐catalyzed cyclization

A promising method developed recently by Kita et al., employs iodine(III) reagent (p‐methoxy‐ (diacetoxyiodo) benzene) and KBr, in direct lactone formation from carboxylic acid via selective C‐H abstraction. 88

2.5 Polymersupported iodine(III) reactions Oxidation by solid‐phase based polyvalent reagents has been developed parallel to the growing desire for a greener chemistry approach. The solid‐phase technique has been introduced in the sixties by Merrifield. 89 Molecules are attached via either ionic or covalent bonds, to organic polymeric or inorganic supports, and undergo a number of reaction steps.

There is a definite need for sustainable processes, based on readily recyclable reagents for the fine chemical, pharmaceutical and agrochemical industry. 2, 90

Active reagents attached to organic, polymeric or inorganic supports show several remarkable advantages. 89

1. Yield of the product is improved due to overcoming separation problems.

Chapter 2

2. Possible reuse or regeneration of polymer‐supported reagent.

3. For processes where reaction proceeds to full completion and with 100% selectivity the process can be automated.

4. For fast reactions a column of the supported‐reagent can be employed, introducing the possibility of flow systems.

5. The lack of solubility and non‐volatility (low or non‐toxicity) of the supporting species makes the overall processes more environmentally acceptable.

The first example of polymer‐supported iodine(III) reagents was reported in seventies. Zupan and Pollak prepared difluorinated iodoaryl residue supported on cross‐linked polystyrene. 91 The chlorinated analogue was synthesized by Hallensleben. 92 The synthesis of polystyrene supported iodosobenzene diaceate was published by Yamada and Okawara. 93 An alternative possible preparation method of iodine(III) diacetoxy fluorous alkyl derivatives such as phenyliodo‐bis(trifluoroacetate) (PIFA), on polystyrene supports has been thoroughly studied (Figure 2.24). 94‐101

Figure 2.24 Examples of polymer‐supported iodine(III) reagents

Nowadays PS‐DAIB, the iodobenzene diacetate bound to polystyrene resin is widely applied as solid reagent supported on polystyrene resin. Poly (diacetoxyiodo) styrene can be easily replaced by its non‐polymeric analogue and in many cases an increase of activity has been observed as reviewed by Togo et al., Ley et al. 10, 102,,103 and Kita et al. demonstrated that PS‐ DAIB‐KBr is a suitable, environmentally benign system for the oxidation of primary and secondary alcohols under mild conditions. They found that water was a suitable solvent for induced oxidation and preferably PS‐DAIB shows excellent stability in the former media in the presence of KBr salt. 104 Recently the group of Kita has reported a metal‐free aqueous oxidation

Hypervalent iodine in organic synthesis of alcohols with the combination of the trivalent iodine oxidant and tetraethylammonium

+ ‐ 105 bromide (Et4N Br ).

Recently the group of Sanford presented a catalytic palladium based system for oxidative C‐H

106 bond functionalization using PS‐I(OAc)2 (Figure 2.25).

Figure 2.25 Pd(OAc)2‐catalyzed ligand directed C‐H acetoxylation with PS‐I(OAc)2

The bottleneck in preparation of PS‐DAIB is the synthesis of iodinated polystyrene, since the known procedures employ toxic or hazardous reagents. 93

This thesis provides a solution to circumvent the use of hazardous chemicals in high concentration via the in situ generation of peracid, followed by in situ formation of iodine(III) oxidant (chapter 5).

2.6 Iodine(III) and ionic liquids Qian et al. applied ion‐supported hypervalent iodine(III) in clean oxidation of alcohols in imidazolium type ionic liquids. 107 The authors claimed that the dehydrogenating properties of

+ ‐ 1‐(4‐diacetoxyiodobenzyl)‐3‐methylimidazolium tetrafluoroborate ([dibmim] [BF4] ) and remaining Br‐ impurities enable smooth formation of carbonyl compounds in primary alcohol oxidation and improve the degree of selectivity (Figure 2.26). The reaction was carried out in 1‐

+ ‐ ethyl‐3‐methylimidazolium tetrafluoroborate ([emim] [BF4] ) and acetonitrile.

Figure 2.26 Clean alcohol oxidation by ion‐supported ionic liquid in ionic liquid/CH3CN mixture

Another example of the use of ionic support is the work of But and co‐workers who applied a combination of a supported nitroxyl radical on Janda Jel as catalyst and PS‐DAIB as oxidant, in the alcohol oxidation (Figure 2.27_A). 108

Chapter 2

A) Simultaneous use of PS DAIB and JJ‐TEMPO for alcohol oxidation

Figure 2.27 B) Macroporous polystyrene supported MP‐DAIB and TEMPO system for alcohol oxidation at ambient temperature

The same group recently reported another (diacetoxyiodo) benzene‐TEMPO system supported on macroporous polystyrene. Macroporous polystyrene has a higher ability to swell during the reaction which may enhance the oxidation performance. Use of solvents such as acetone, acetonitrile, dimethylformamide or water is proposed (Figure 2.27_B).109

2.7 Iodine(V) vs. iodine(III)

One of the first pentavalent iodine compounds to be prepared was iodylbenzene PhIO2 synthesized by Willgerodt in 1900. 110 Numerous pentavalent iodine derivatives have been subsequently prepared but only a few have found practical synthetic applications. The most well known is the Dess‐Martin Periodane reagent (DMP) or its precursor 2‐iodoxy benzoic acid (IBX). The first reported synthesis dates from 1893 by Hartmann & Meyer in Chemische Berichte111 For structures see Figure 2.28.

Figure 2.28 Common pentavalent reagents

Hypervalent iodine in organic synthesis

Dess‐Martin Periodane reagent (DMP) is notably known for its high selectivity in the selective oxidation of alcohols for difficult to oxidize substrates. In addition it tolerates many alcohols with functional groups. In general the synthetic potential of pentavalent iodine compounds is poorly explored compared to trivalent iodine. This is mainly edu to the low stability, difficult handling and high explosive nature of pentavalent iodine reagent. The use and application of iodine(V) is limited to laboratory scale. Strict industrial rules do not permit application of iodine(V) on a larger scale. In view of the current need for sustainable processes iodine(V) has a very narrow future compare to iodine(III) reagents.

The detailed chemistry of organic compounds of pentavalent iodine was reviewed in the book of Varvoglis. „The Organic Chemistry of Polycoordinated Iodine“. 5 Nevertheless, there are ongoing studies on supported iodine(V) reagents. One of them is the work of Chung et al., who reported polymer‐supported IBX‐with amide as spacer for selective oxidation of alcohols. The

. 112 rate of the oxidation was enhanced by additives such as TFA or BF3 OEt. A significant disadvantage however, is the requirement of chlorinated solvents such as chloroform as reaction media (Figure 2.29).

Figure 2.29 Oxidation of alcohols with PS‐IBX‐amide

The iodine(V) reagents can also be used as insoluble polystyrene‐supported analogue, i.e. for alcohol oxidation. 113‐116 In 2003 a stabilized formulation of IBX (SIBX) was reported.117 This stabilized oxidant displays none of the explosive properties of IBX reagent. The reactions result in comparable yields to those of IBX. (Figure 2.30)

Figure 2.30 Oxidation of alcohols with stabilized IBX

Chapter 2

Iodine(V) has been also applied in Ionic liquids. They have been studied as alternative for chlorinated solvents oxidation performed in ionic liquids or ionic liquid/water system (Figure 2.31). 118, 119

Figure 2.31 Alcohol oxidation by IBX or DMP in ionic liquids

Iodine(V) will certainly be the first choice for laboratory scouting experiments due to the very high reactivity potential. However, iodine(III) will be always a choice in terms of future application and scale up purposes.

2.8 Concluding remarks Especially the use of iodine(III) bears great promise in oxidation chemistry. It is a versatile reagent, which can moderate a large variety of reactions. The application of iodine(III) will be substantially facilitated by a mild in situ method based on hydrogen peroxide as oxidant.

Hypervalent iodine in organic synthesis

2.9 References

1P. J. Stang, V. V. Zhdankin, Chem. Rev., 1996, 96, 1123. 2T. Wirth, Topics in Curr. Chem., 2003, 224, 186. 3A. Varvoglis, Synthesis 1984, 709. 4R. M. Moriarty, O. Prakash, Acc. Chem. Res., 1986, 19, 224. 5A. Varvoglis, The Organic Chemistry of Policoordinated Iodine, Academic Press (Eds.), VCH, Weinheim 1992. 6A. Varvoglis, Hypervalent Iodine in Organic Synthesis Academic Press (Eds.), London, 1997. 7R. M. Moriarty, O. Prakash, Adv. Heterocyclic Chem., 1998, 69, 1. 8R. M. Moriarty, O. Prakash, Org. React., 1999, 54, 273. 9T. Wirth, U. H. Hirt, Synthesis, 1999, 1271. 10H. Togo, M. Katohki, Synlett., 2001, 565. 11V. V. Zhdankin, P. J. Stang, Chem. Rev., 2002, 102, 2523. 12C. J. Carmalt, J. G. Crossley, J. G. Knight, P. Lightfoot, A. Martin, M. P. Muldovney, N. C. Norman, A. G. Orpen, J. Chem. Soc., Chem. Commun., 1994, 2367. 13J. V. Carey, P. A. Chaloner, K. R. Seddon, M. Taylor, J. Chem. Res. Synop., 1996, 156. 14N. S. Zefirov, V. V. Zhdankin, S. O. Safronov, A. A. Kaznachaev, Zh. Org. Khim., 1989, 25, 1807. 15V. V. Zhdankin, R. Tykwinski, B. L. Wiliamson, P. J. Stang, N. S. Zefirov, Tetrahedron Lett., 1991, 32, 733. 16R. M. Moriarty, N. Rani, C. Condeiu, M. P. Duncan, O. Prakash, Synth. Commun., 1997, 27, 3273. 17R. M. Moriarty, O. Prakash, M. P. Duncan, R. K. Vaid, N. Rani, J. Chem. Res. Synop., 1996, 9, 432. 18H. Tohma, S. Takizawa, T. Maegawa, Y. Kita, Angew. Chem. Int, Ed., 2000, 39, 1306. 19H. Tohma, S. Takizawa, T. Maegawa, Y. Kita, Synlett., 2003, 723. 20T. Wirth, Angew. Chem. Int. Ed., 2005, 44, 2. 21T. J. McMurry, J. T. Groves In Cytochrome P‐450: Structure, Mechanism and Biochemistry; P. R. Ortiz de Montellano, (Eds.), Plenum Press: New York, 1986, 1. 22R. A. Johnson, K. B. Sharpless In Catalytic Asymetric Synthesis, (Eds.), I. Ojima, Wiley‐VCH, New York 2000, 6A. 23E. N. Jacobsen, M. H. Wu In Comprehensive Asymetric Catalysis: E. N. Jacobsen, A. Pfaltz, H. Yamamoto, (Eds.), Springer, New York, 1999,18.2, 649. 24T. Katsuki, In Catalytic Asymetric Synthesis, (Eds.), I. Ojima, Wiley‐VCH, New York, 2000, 6B. 25K. M. Ryan, C. Bousquet, D. G. Gilheany, Tetrahedron Lett., 1999, 40, 3613. 26S. Mukerjee, A. Stassinopoulos, J. P. Caradonna, J. Am. Chem. Soc., 1997, 119, 8097. 27C. –C. Guo, J. –X. Song, X. –B. Chen, G. –F. Jiang, J. Mol. Catal. A: Chem., 2000, 157, 31. 28M. das Dores Assis, J. R. L. Smith, J. Chem. Soc., Perkin Trans., 2, 1998, 2221. 29R. Breslow, B. Gabriele, J. Yang, Tetrahedron Lett., 1998, 39, 2887. 30J. Yang, R. Weinberg, R. Breslow, J. Chem. Soc., Chem. Commun., 2000, 531. 31C. Willgerodt, J. Prakt. Chem., 1886, 33, 154. 32H. Sohmiya, T. Kitamura, M. Fujita, T. Ando, Chem. Lett., 1992, 891. 33A. Pelter, S. M. A. Elgendy, J. Chem. Soc. Perkin Trans., 1, 1993, 1891. 34A. McKillop, L. Mclaren, R. J. K. Taylor, J. Chem. Soc. Perkin Trans., 1, 1994, 2047. 35R. L. Amey, J. C. Martin, J. Org. Chem., 1979, 44, 1779. 36W. Carpenter, J. Org. Chem., 1966, 31, 2688. 37M. Sawaguchi, S. Ayuba, S. Hara, Synthesis, 2002, 1802. 38M. A. Arrica, T. Wirth, Eur. J. Org. Chem., 2005, 395. 39A. B. Sheremetev, D. E. Dmitriev, S. M. Konkina, Russ. Chem. Bull. Int. Ed., 2004, 53, 1130.

Chapter 2

40J. J. Edmunds, W. B. Motherwell, J. Chem. Soc., Chem. Commun., 1989, 881. 41E. B. Merkushev, Russ. Chem. Rev., 1987, 56, 826. 42T. V. Moskovkina, V. I. Vysotski, Zh. Org. Khm., 1991, 27, 883. 43M. Mikolajczyk, W. H. Midura, S. Grzejszczak, F. Montanari, M. Cinqquini, M. W. Wieczorek, J. K. Wojciechowska, Tetrahedron, 1994, 50, 8053. 44S. Masson, A. Thuillier, Bull, Soc. Chim. Fr., 1969, 4368. 45V. A. Andreev, S. N. Anvilogova, T. I. Pekh, N. A. Belikova, Zh, Org. Khim., 1993, 29, 142. 46A. P. March, V. D. Sorokin, D. Rajagropal, S. G. Bott, Tetrahedron, 1994, 50, 9933. 47R. Poli, M. A. Kelland, J. Organomet. Chem., 1991, 419, 127. 48A. Crispini, M. Ghedini, F. Neve, Inorg. Chim. Acta., 1993, 209, 235. 49N. Burford, J. A Clyburne, D. P. Gates, M. J. Schriver, J. F. Richardson, J. Chem. Soc. Dalton Trans., 1, 1994, 997. 50A. S. Radhakrishna, K. Sivaprakash, B. B. Singh, Synthetic Commun., 1991, 21, 1625. 51A. S. Radhakrishna, K. Sivaprakash, B. B. Singh, Synthetic Commun., 1991, 21, 1473. 52A. J. Kotlewska, R. A. Sheldon, I. W. C. E. Arends manuscript in preparation. 53J. G. Sharefkin, H. Salzman, Organic Syntheses; Wiley, New York, 1973, Collect., 5, 660. 54P. Kazimierczak, L. Skulski, L. Kraszkiewicz, Molecules, 2001, 6, 881. 55A. Varvoglis, Hypervalent Iodine in Organic Synthesis Academic Press: San Diego, 1997. 56A. Varvoglis, Chem. Soc. Rev., 1981, 10. 377. 57H. Tohma, Chem. Abstr., 2000, 133, 222460. 58M. Arisawa, H. Tohma, Y. Kita, Chem. Abstr., 2001, 134, 29586. 59A. Varvoglis, Tetrahedron, 1997, 53, 1179. 60T. Kitamura, Y. Fujiwara, Org. Prep. Proced. Int., 1997, 29, 409. 61T. Muraki, H. Togo, M. Yokoyama, Rev. Heteroatom Chem., 1997, 17, 213. 62T. Kitamura, Chem. Abstr., 1997, 127,121516 63H. Togo, Y. Hoshina, G. Nogami, M. Yokoyama, Chem. Abstr., 1, 1997, 126, 1569. 64R. M. Moriarty, O. Prakash, Adv. Heterocycl. Chem., 1998, 69, 1. 65V. V. Zhdankin, P. J. Stang, Tetrahedron, 1998, 54, 10927. 66T. Muraki, H. Togo, M. Yokoyama, J. Org. Chem., 1999, 64, 2883. 67Y. Kita, M. Egi, T. Takada, H. Tohma, Synthesis, 1999, 885. 68V. V. Grushin, Chem. Soc. Rev., 2000, 29, 315. 69K. Akiba, Chemistry of Hypervalent Compounds, Ed. Wiley‐VCH, New York, 1999, 11. 70E. D. Matveeva, M. V. Proskurnina, N. S. Zefirov, Heteroatom Chem., 2006, 17, 595. 71W. Panhchan, S. Chiampanichayakul, D. L. Snyder, S. Yodbuntung, M. Pohmakotr, V. Reutrakul, T. Jaipetch, Ch. Kuhakarn, Tetrahedron, 2010, 66, 2732. 72A. De Milco, R. Margarita, L. Parlanti, A. Vescovi, G. Pincatelli, J. Org. Chem., 1997, 6, 6974. 73 Z. Chenjie, W. Yunyang, J. Lei Synthetic Commun., 2010, 40, 2057. 74G. Wells, T. D. Bradshaw, P. Diana, A. Seaton, D. F. Shi, A. D. Westwell, M. F. G. Stevens, Bioorg. Med. Chem. Lett., 2000, 10, 513. 75G. Wells, A. Seaton, M. F. G. Stevens, J. Med. Chem., 2000, 43, 1550. 76W. Ou, Z. C. Chen, Synth. Commun., 1999, 29, 4443. 77S. P. Singh, D. Kumar, O. Prakash, R. P. Kapoor, Synth. Commun., 1997, 27, 2683. 78R. M. Moriarty, O. Prakash, M. P. Duncan, Synthesis, 1985, 943. 79R. M. Moriarty, M. P. Duncan, O. Prakash, J. Chem. Soc. Perkin Trans., 1, 1987, 1781. 80S. Yasuda, T. Yamada, M. Hanaoka, Tetrahedron Lett., 1986, 27, 2023. 81R. M. Moriarty, L. S. John, P.C. Du, J. Chem. Soc., Chem. Commun., 1981, 641. 82S. J. Daum, Tetrahedron Lett., 1984, 25, 4725. 83O. Prakash, H. Batra, H. Kaur, P. K. Sharma, V. Sharma, S. P. Singh, R. M. Moriarty, Synthesis, 2001. 84D. P. Cheng, Z. C. Chen, Q. G. Zheng, J. Chem. Res., (S), 2002, 624.

Hypervalent iodine in organic synthesis

85A. Kryska, L. Skulski, J. Chem. Res., (S), 1999, 590. 86J. –L. Liang, J. –S. Huang, X. –Q. Yu, N. Zhu, C.‐M. Che, Chem. Eur. J., 2002, 8, 1563. 87C. G. Espino, J. Du Bois, Angew. Chem., 2001, 113, 618. 88T. Dohi, N. Takenaga, A. Goto, A. Maruyama, Y. Kita, Organic Lett., 2007, 9, 3129. 89P. Hodge, D. C. Sherrington, Polymer‐supported Reaction in Organic Synthesis, (Eds.), Wiley & Sons, New York, 1980. 90F. Z. Doerwald, Organic Synthesis on Solid Phase,(Eds.), Wiley‐VCH, Weinheim, 2000. 91M. Zupan, Collect. Chech. Chem. Commun., 1977, 42, 266. 92M. L. Hallensleben, Angew. Macromol. Chem., 1972, 27, 223. 93Y. Yamada, M. Okawara, Makromol. Chem., 1971, 152, 153. 94L. M. Yagupolski, I. I. Maletina, N. V. Kondratenko, V. V. Orda, Synthesis, 1978, 835. 95G. M. Loudon, A. S. Radhakrishna, M. R. Almond, J. K. Blodgett, R. H. Boutin, J. Org. Chem. 1984, 49, 4272. 96Ch. J. Kuehl, J. T. Bolz, V. V. Zhdankin, Synthesis, 1995, 312. 97Ch. Rocaboy, J. A. Gladysz, Chem. Eur. J., 2003, 9, 88. 98V. Tesevic, J. A. Gladysz, Green. Chem., 2005, 7, 833. 99V. Tesevic, J. A. Gladysz, J. Org. Chem., 2006, 71, 7433. 100M. Celik, C. Alp, B. Coskun, M. Serdar, Gueltekin, M. Balci, Tetrahedron Lett., 2006, 47, 3659. 101T. Dohi, M. Ito, N. Yamaoka, K. Morimoto, H. Fujioka, Y. Kita, Tetrahedron, 2009, 65, 10797. 102S. V. Ley, A. W. Thomas, H. Finch, J. Chem. Soc. Perkin Trans., 1, 1999, 669. 103S. V. Ley, O. Schucht, A. W. Thomas, P. J. Murray, J. Chem. Soc. Perkin Trans., 1, 1999, 1251. 104H. Tohma, T. Maegawa, S. Takizawa, Y. Kita, Adv. Synth. Catal., 2002, 344, 328. 105N. Takenaga, A. Goto, M. Yoshimura, H. Fujioka, T. Dohi, Y. Kita, Tetrahedron Lett., 2009, 50, 3227. 106E. W. Kalberer, S. R. Whitfield, M. S. Sanford, J. Mol. Catal. A. Chemical, 2006, 251, 108. 107W. Qian, E. Jin, W. Bao, Y. Zhang, Angew. Chem. Int. Ed., 2005, 44, 952. 108T. Y. S. But, Y. Tashino, H. Togo, P. H. Toy, Org. Biomol. Chem., 2005, 3, 970. 109Y. Shang, T. Y. S. But, H. Togo, P. H. Toy, Synlett., 2007, 1, 67. 110H. J. Lucas, E. R. Kennedy in Organic Syntheses 1995, collected work incl. Willgerodt 1900, Wiley New York, III, 482.S 111C. Hartmann, V. Meyer, Chem. Ber. 1893, 26, 1727; B. D. Dess, J. C. Martin, J. Org. Chem., 1983, 48, 4155. 112W‐J. Chung, D‐K. Kim, Y‐S. Lee, Synlett., 2005, 14, 2175. 113N. N. Reed, M. Delgato, K. Herefors, B. Clapham, K. J. Janda, Bioorg. & Medical Chem. Lett., 2002, 12, 2047. 114Z. Lei, C. Denecker, S. Jegasothy, D. C. Sherrington, N. K. H. Slater, A. J. Sutherland, Tetrahedron Lett., 2003, 44, 1635. 115Z. Liu, Z‐C. Chen, Q‐G. Zheng, Organic Lett., 2003, 5, 3321. 116W‐J. Chung, D‐K. Kim, Y‐S. Lee, Tetrahedron Lett., 2003, 44, 9251. 117A. Ozanne, L. Pouysegu, D. Depernet, B. Francüois, S. Quideau, Organic. Lett., 2003, 5, 2903. 118J. S. Yadav, B. V. S. Reddy, A. K. Basak, A. Venkat Narasaiah, Tetrahedron, 2004, 60, 2131. 119G. Karthikeyan, P. T. Perumal, Synlett., 2003, 14, 2249.

CChhaapptteerr 33

Polymerattached iodine(III) reagents in selective alcohols oxidation. Scope of oxidants.

3.1 Introduction 3.2 Results and Discussion 3.2.1 The choice of the polymeric support (cross‐linked vs. linear) 3.2.2 Effect of carbon chain length on the reactivity of the polymeric oxidants 3.2.3 Safety aspects of the oxidants 3.2.4 Solvent effect on iodine(III) oxidants 3.3 Conclusions 3.4 Experimental details 3.5 References

Abstract Hypervalent iodine(III) oxidants can be conveniently synthesized and applied in corresponding polystyrene supported derivatives. They are excellent oxidants for the selective oxidation of alcohols and do not require additives or catalysts. Benzylic alcohol could be oxidized within 5 h at 70 oC in a variety of organic solvents resulting in 100% benzaldehyde. The effect of polystyrene support (linear versus cross‐linked) on the oxidation performance was studied. Peracids varying in chain length were used to generate iodine(III), resulting in oxidant polymers with different stabilities, safety properties and activities.

Chapter 3

3.1 Introduction The selective oxidation of alcohol moieties in highly functionalized molecules is a pivotal transformation in the preparation of pharmaceuticals. 1 The recently developed large variety of methodologies based on catalytic amounts of metal complexes and molecular oxygen are often not suitable for this specific task,. 2, 3 They are hampered by the deactivation of the metal complexes by e.g. amine functionalities or simply not active.

For this reason the Dess‐Martin reagent, 4 or its precursor the IBX reagent, 5 are often applied in the laboratory as reliable, selective and effective reagents for the oxidation of these functionalized molecules. However, these iodine(V) reagents are potentially explosive and, hence, are not suitable for use on a large scale. Some examples of the catalytic use of iodine(V) have been reported using oxone (peroxymonosulfate) as the stoichiometric oxidant. 6, 7 In this case aliphatic primary alcohols are oxidized to the corresponding carboxylic acids.

Recently, iodine(III) based reagents have emerged as potential alternatives to oxidations with iodine(V) reagents. The use of iodine(III) opens the way to much safer oxidative transformations and rearrangements, and the application of these compounds is still increasing. 8 In the oxidation of alcohol functionalities with iodine(III), in all reported cases, the iodine(III) is accompanied by large amounts of TEMPO or Br‐ as co‐oxidants. 9, 10 In the first case the real oxidant is the oxoammonium ion, which is indeed an excellent oxidant for converting alcohols into the corresponding aldehydes, ketones or acids 11 but this leaves the full oxidation potential of iodine(III) itself unexplored. On the other hand, the co‐use of bromide has the major disadvantage i.e. in the reaction mixture unwanted brominated side‐products are formed. Kita and co‐workers investigated the nature of the active oxidant in the I(III)/Br‐ system, and concluded that a I(III)/Br‐type species was most likely.12

Ley and co‐workers reported that iodine(III) could oxidize benzylic alcohols into the corresponding aldehydes without the use of co‐oxidants, however requiring excess reagents and long reaction times. 13, 14 Unfortunately authors did not present any other examples.

In this chapter the synthesis of various polystyrene supports was described. The effect of the chain length of the peracids used to activate iodine was investigated. The stability of the polymeric oxidant as well as its physical behavior in organic solvents was studied. The focus in this chapter is on the development of the polymeric oxidant, using benzylic alcohol as a reference compound. The substrate scope and mechanism of the reaction will be described in chapter 4.

Polymersupported iodine(III) in selective oxidation of alcohols. Oxidants

3.2 Results and discussion

3.2.1 The choice of the polymeric support (crosslinked vs. linear) Synthesis of the oxidant followed by IR spectroscopy

Our approach was to synthesize a variety of iodine(III) polymers, with varying polystyrene backbones length and compare them in the oxidation of benzyl alcohol. Polymer‐supported iodine(III) oxidant was synthesized in a two step process. First, iodination of polystyrene was carried out aiming at a nearly quantitative iodination degree. The next step was to load iodine with “active oxygen” via oxidation with in situ generated peracetic acid (Figure 3.1).

Figure 3.1 Scheme for the synthesis of polymeric iodine(III) oxidant, PS‐DAIB = polystyrene‐ iodosobenzene diacetate

Iodination of polystyrene was performed in acetonitrile with 1‐(chloromethyl)‐4‐fluoro‐1,4‐ diazoniabicyclo[2.2.2]octane bis(tetrafluoroborate) known under the commercial name of

TM 15 Select Fluor (F‐TEDA‐BF4) and molecular iodine at milder than generally applied conditions. Select FluorTM is recognized as a modern electrophilic fluorinating reagent and a mediator for selective direct iodination (Figure 3.2). The synthstic procedure is reported in the experimental part.

Figure 3.2 Structure of Select Fluor TM

Infrared analysis was used for the qualitative evaluation of the reaction progress and identification of the desired oxidant. In Figure 3.3 a representative example of infrared spectra of: (A) starting compound (PS), and (B) the following intermediate iodine(I) (PS‐I) are presented.

As starting compound linear polystyrene with a Mw of 250,000 g/mol was used.

Chapter 3

78.6 75 2336.91

1745.47 1541.28 70 1869.45 840.65 1942.38 1801.72 979.65 621.03

964.02 65 3430.12 1181.24 1328.13 1154.38 1583.07 1068.88 906.09 60

3082.07 55 2849.15 1028.12 %T 538.02 3059.78 1601.21 50

45 3025.54 2921.93 756.19 1492.80 40 1452.20 1384.35

35 696.77

31.1 4000.0 3000 2000 1500 1000 450.0 cm-1 (A) Characteristic frequencies: mono‐substituted phenyl rings C‐H bands: 697 and 756 cm‐1

68.7 65 1779.36 629.68 60 1895.12 1091.08 959.22 1561.27 1303.98 55 1353.55 1182.92 940.96 3051.16 781.76 3014.91 1636.39 751.69 50 1584.72 699.50 2848.27 1384.83 715.56 45 1447.93 538.28 1062.07 %T 40 3434.16

35 2920.93 1403.67

25 816.47 1004.79 1481.90 20 17.5 4000.0 3000 2000 1500 1000 450.0 cm-1 (B) Characteristic frequencies: para‐substituted phenyl rings C‐H bands: 816 and 1004 cm‐1

75.6

70 458.66

65 611.27

1047.54

60 539.09

%T 666.75 3024.61 1481.03 1451.20 1003.70 820.98 55 1183.38

2923.37 1580.29 1364.23 50 1405.96 702.32 1705.15 1384.48 1638.54 1268.38 3412.44 45 762.62

41.4 4000.0 3000 2000 1500 1000 450.0 cm-1 (C) Characteristic frequencies: C‐O; CO‐O; C= O bands: 1638 and 1705 cm‐1

Figure 3.3 (A, B, C) IR analysis : formation of iodine(III) oxidant. (A) Starting compound, linear polystyrene (PS), (B) Iodinated polystyrene (PS‐I) iodination degree 98.5%, (C) Oxidized PS‐I with peracetic acid (PS‐ DAIB)

Synthesis of iodinated polystyrene

In total four iodine‐polystyrenes were synthesized and their analysis data are given in Table 3.1. All batches of PS‐I were synthesized according to method A. The Select Fluor TM method was only applied on a small scale to demonstrate its viability as synthesis method. As can be seen the method for producing PS‐I results in a material that has a maximum iodine loading. In

Polymersupported iodine(III) in selective oxidation of alcohols. Oxidants general the iodination degree determined by quantitative 13C NMR was in excellent agreement with the elemental analysis. The insolubility of cross‐linked polystyrene limits its analysis to IR and the elemental analysis only. In this case a maximum iodine loading of 67% was obtained.

Table 3.1 Analysis of iodinated polystyrenes

C C H H I I I degree I degree found calc. found calc. found calc. (el. anal.) (NMR)

% % % % % % % %

PS‐I‐crossa‐ 49.0 41.7 2.6 3.5 44.9 55.2 67 ‐ li k d PS‐I linearb 42.5 41.7 3.1 3.5 55.3 55.2 100 98.5

PS‐I linearb 43.6 41.7 3.2 3.5 55.6 55.2 100 91.5

PS‐I linearb 42.0 41.7 3.1 3.5 55.7 55.2 100 98.0 As starting material polystyrene cross‐linked with 2% divinyl was used.b) As starting material PS with Mw of 250,000 was used.

Comparison of polystyrene supports in oxidation

The different iodopolystyrene supports were oxidized with peracetic acid as described in section 4. A representative spectrum of PS‐DAIB is given in Figure 3.3 and clearly reveals the presence of esters bonds. The oxidation reactivity of the supported hypervalent iodine(III) reagent was compared for in total four different polystyrenes as support materials illustrated in Figure 3.4.

Oxidant polymeric support

PS cross linked 100-200 mesh 1 % - Cl-divinyl (1)

Mw = 350,000

commercial oxidant

PS cross linked 200-400 mesh 2 %-Cl- divinyl (2)

PS linear Mw = 51,000 (3)

PS linear Mw = 250,000 (4)

Figure 3.4 Iodine(III) polystyrene supports used in this study

Chapter 3

The commercially available polymer‐bound p‐phenoxy iodosobenzene diacetate (1), containing cross‐linked polystyrene Mw= 350,000 as a support, was compared with cross‐linked 2%‐Cl‐ divinyl polystyrene based iodosobenzene diacetate, which is known from the literature to be an efficient oxidant (2), and two linear based oxidants containing linear polystyrene Mw= 51,000 (3) and linear polystyrene Mw= 250,000 (4) as support. Initial screening revealed that all tested oxidants were active in the oxidation of benzyl alcohol and 1‐octanol at 70 oC in ethyl acetate. For results see Table 3.2.

Table 3.2 Comparison of different polystyrene supports in primary alcohol (benzyl alcohol and octan‐1‐ol) oxidation with 1 eq. of polymer‐supported iodobenzene diacetate iodine(III) as sole oxidant a

Entry Polymeric support I (III) loadingb Yield benzaldehyde Yield octanal

1 Commercial PS (Figure 3.4) 0.33 mmol/g 40 % (7 h)

2 PS cross‐linked 2.06 mmol/g 72 % (24 h) 39 % (8 h)

3 PS linear Mw 51,000 1.53 mmol/g 100 % (24 h) 98 % (24 h)

4 PS linear Mw 250,000 3.73 mmol/g 97 % (5 h) 99 % (5 h) a: Conditions: 2 mmol alcohol at 70 oC in ethyl acetate using 1 eq. of iodobenzene diacetate polymer– attached oxidant. b: Determined by iodometric titration.

As can be seen, the oxidants synthesized in this project displayed much higher activity than the commercial oxidant. The oxidant with the cross‐linked support showed low activity (entry 2, Table 3.2). A plausible explanation is that the accessibility of active iodine is limited when the oxidant is supported on a cross‐linked polymer; also the iodination degree was already lower in these cases (67% for the cross‐linked 2%‐Cl‐divinyl 200‐400 mesh polystyrene). This is probably also the reason why the commercially available I(III) polymer is lower in activity. In contrast with linear supports an efficient and reactive oxidant is obtained. In addition the cross‐linked PS‐ DAIB oxidant showed a tendency towards overoxidation. After 24 h of reaction the formation of benzoic acid is observed, while in the case of oxidant supported on a linear support (entries 3 and 4, Table 3.2), after 5 h the reaction is completed with more than 97% selectivity towards benzaldehyde as the only product.

Therefore, in subsequent experiments, oxidized linear iodopolystyrene was used as the oxidant of choice for alcohol. In addition the linear polystyrene with Mw of 250,000 gave better results with respect to loading and activity (see Table 3.2).

3.2.2 Effect of carbon chain length on the reactivity of the polymeric oxidant

In order to explore the full potential of the polymeric iodine(III) oxidant our next aim was to understand the nature of carbon chain length not only on the reactivity of the polymer based

Polymersupported iodine(III) in selective oxidation of alcohols. Oxidants iodine(III) oxidants but predominantly on the safety performance. The only derivative described in the literature is iodosobenzene diacetate supported on a cross‐linked polystyrene, cl‐PS‐ DAIB. The general way to synthesize this oxidant is to first iodinate polystyrene with molecular iodine followed by oxidation of the latter with peracetic acid. e Th growing need for environmentally friendlier and safer procedures was a reason to investigate the use of the more stable longer chain percarboxylic acids, such as perbutanoic, perhexanoic or peroctanoic acids. Longer chain iodine(III) derivatives will exhibit possibly less explosive properties. Thus, in

addition to the PS‐DAIB (C2), three new derivatives (C4, C6, C8) were synthesized as shown in Figure 3.5.

O O C H O O PS PS I 3 PS PS I I I O O O CH OCH O CH3 O 3 O 3 O O O O O C H 3

H3C 3CH H3C

PS - DAIB = C 2 PS -DBIB = C4 PS-DHIB = C6 PS -DCIB = C8

Figure 3.5 Postulated structures of polymeric iodine(III) oxidants with different chain lengths

IR, iodometric titration and elemental analysis were used to evaluate the structure of the synthesized iodine(III) polymers. The data are summarized in Table 3.3.

The elemental analysis data cannot easily be matched to one structural formula. Already for the first oxidant, for which almost all iodine(I) was oxidized to iodine(III) it is apparent that very little extra carbon was added to the polymer. Perhaps, our oxidant has a structure which more closely resembles iodosoarene, instead of acetoxylated iodoarene. We postulate that the

structure is most close to that of dihydroxy iodobenzene with a structural formula of C8H9IO2 (calculated data C 36,4%; O 12.1%; H 3.4%; I 48.1%). However, this is inconsistent with the infrared data that clearly indicate the presence of the carbonyl stretch at around 1700 cm‐1. For

‐1 comparison, the non‐polymeric PhI(OAc)2 exhibits a typical carbonyl stretch around 1650 cm (carbonyl stretch in DAIB). We therefore assume that a certain percentage of the iodine is diacetoxylated, which also explains the presence of extra oxygen (O/I ratio = 2.5/1).

Further studies are required to unravel the difference between IR results and elemental analysis data. For the hypervalent polymer oxidants synthesized with longer peracids a similar discrepancy is found. They seem to be a mixture of iodinated polystyrene (C8H7In), dihydroxy

Chapter 3 iodostyrene and alkoxylated iodostyrene. The fact that the C/I ratio is between 9.1 and 9.6 indicates extra carbon is incorporated in line with the observed ester bonds in the infrared. It is also apparent that the increasing length of the oxidant (C2 up to C8) generally makes it more difficult to reach active site (oxidation percentage varies from 97 to 33%). It is not clear why the alkoxy chain does not stay attached to the oxidant. It could be that during work‐up and washing with water, hydrolysis could take place. Overall we managed to synthesize a range of polymeric iodine(III) materials, with distinct differences in active oxygen loading and structure.

Table 3.3 Analysis data of the polymeric iodine(III) oxidants

% PS‐DAIB PS‐DBIB PS‐DHIB PS‐DCIB

C found 34.7 36.1 39.2 41.8

O found 15.2 16.1 11.1 8.3

H found 2.9 3.2 3.4 3.6

I found 48.7 41.8 43.4 48.8

I(III) from titration 97% 67% 100% 33%

C/I found 7.6 9.1 9.6 9.2 (mol/mol)

O/I found 2.5 3.1 2.0 1.4 (mol/mol)

3.2.3 Safety aspects of the polymeric iodine(III) oxidants From the viewpoint of industrial application, safety and ease of handling are of key importance. It is known from the literature that soluble hypervalent iodine reagents have a high enthalpy of decomposition, which severely hampers their large scale application. Hence, we have performed thermodynamic analyses and thermal decomposition studies on dry samples of the various polymeric iodine(III) oxidants.

Thermogravimetric Analysis TGA

Thermogravimetric analysis (TGA) describes the loss of mass under heating of a dry sample of the oxidant within a described period of time. Oxidants supported on cross‐linked and linear supports were. tested The decomposition profiles provide information on the stability of the polymeric oxidant (Figure 3.6). In the design for the oxidant the aim was to elevate the decomposition temperature as much as possible.

Polymersupported iodine(III) in selective oxidation of alcohols. Oxidants

Figure 3.6 Thermogravimetric analysis results of the oxidants

o The C2 = lin‐PS‐DAIB derivative decomposed rapidly at an average temperature of 160 C. In contrast, for the C4 derivative (lin PS‐DBIB) and C8 derivatives (lin and cl PS DCIB) of the oxidant the decomposition temperature was 170 oC and the decomposition process proceeded gently.

Based on the decomposition trajectory and temperature the C2 derivative (linear and cross‐ linked support) appeared to be the least thermally stable, 5% of weight was lost within the first minute of heating, while the linear C4 and C8 oxidants decomposed slowly without any sudden change in weight.

Test for thermal stability (Setaram C80)

The linear C2, C4 and C8 derivatives of the polymeric iodine(III) oxidant gave evidence of an o exothermic effect upon heating at onset temperatures below 250 C. Overall the C2 derivative showed the largest exothermic effect. The effect with the lowest onset temperature was recorded for the C8 derivative of the oxidant. For the results see Table 3.4. In general for a safe process temperature on the basis of the onset temperature, a safety margin between 50 oC and 100 oC is often applied. 16 In the case of trivalent polymeric iodine oxidants a safety margin

o temperature of 70 C was chosen. Taking this margin into account, linear PS‐DCIB (C8 derivative) o is recommended to be used at a maximum temperature of 54 C, the C2 derivative PS‐DAIB at o o 78 C and the C4 derivative PS‐DBIB at 82 C.

Chapter 3

Table 3.4 Thermal stability of the linear oxidants

Oxidant abbrev. Onset temp. Enthalpy ∆Tad * º C J / g º C

PS-DAIB C2 derivative 152 1307 654

PS-DBIB C4 derivative 148 898 449

PS-DCIB C8 derivative 124 782 391

* The calculations were performed using estimation for the Cp value of 2 J/ g K

Impact sensitivity studies

The impact sensitivity test showed that the C2 derivative appeared to be the most sensitive and gave the largest sounds effect upon impact with 40 J. The sample holder in this case showed complete discoloration of the sample to black, indicating full decomposition of the sample. For the C4 and C8 oxidants only slight discolouration was observed on the sample surface when an impact of 40 J or 7.5 J was applied. Upon impact of 40 J, the PS‐DBIB (C4 derivative) gave smoke production and upon 7.5 J discoloration of the sample surface up to 40%. Similarly, smoke was observed for the C8 derivative upon impact of 40 J and approximately 30% of the sample surface discoloured to black upon impact of 7.5 J. For results see Table 3.5.

Table 3.5 Results impact sensitivity tests (BAM Fallhammer)

Impact Impact Impact Oxidant abbreviated sensitivity 40 J sensitivity 7.5 J sensitivity 2 J

PS-DAIB C2 derivative positive negative negative

PS-DBIB C4 derivative positive positive approx. 40 % discolouration (black) negative

PS-DCIB C8 derivative positive positive approx. 30 % discolouration (black) negative

In conclusion, based on the results of the three tests, the linear PS‐DAIB oxidant exhibited the most rapid decomposition according to TGA, combined with a relatively high onset temperature and the lowest impact sensitivity. Overall oxidants C4 and C8 are more stable and thus safer to handle. These results are in line with the active oxygen loading for the oxidants, which ranged from 3.7 mmol/g (C2) to 1.71 mmol/g for C4, to 1.3 mmol/g for C8.

Polymersupported iodine(III) in selective oxidation of alcohols. Oxidants

3.2.4 Solvent effect on iodine(III) oxidants

We studied acetonitrile, ethyl acetate and toluene as reaction media for alcohol oxidations with the polymer‐supported trivalent iodine reagents. The physical and chemical properties of these solvents are compared in Table 3.6.

Table 3.6 Physicochemical properties of the selected solvents

Solvent key properties acetonitrile ethyl acetate toluene

Viscosity at 20 oC 0.36 cps 0.44 cps 0.6 cps

Boiling point 81-82 oC77 oC111 oC

Polarity (10.2 water) 5.8 4.4 2.4

Log P -0.22 0.53 2.66

Dielectric constant 37.5 6.02 2.38

Dipole moment 3.44 1.78 0.36

In water the oxidation of a primary alcohol affords the corresponding carboxylic acid as the sole product. However, we were interested in the selective oxidation of primary alcohols to the corresponding aldehydes. Therefore, we turned our attention to other polar media. Ethanol as a reaction medium gave very poor conversions and yields, presumably because it is itself an alcohol substrate. Other polar solvents such as acetonitrile and ethyl acetate performed remarkably well in alcohol oxidation. Among apolar solvents, such as n‐hexane, toluene, methyl tert‐butyl ether, the only promising medium was toluene, where an interesting enhancement in the oxidation of primary alcohols was observed. This subject is elaborated further in chapter 4 of this thesis.

Particle size distribution

With the three best solvents for alcohol oxidation we performed additional tests with regard to their influence on the particle size distribution and rate of sedimentation of the polymeric oxidants at different time intervals. Sedimentation data for all linear supported oxidants indicated that after 5 min or 2 h of sedimentation the best size distribution is obtained in ethyl acetate and acetonitrile (Figure 3.7).

Chapter 3

PS‐DBIB C4 – derivative, sedimentation for 5 min PS‐DBIB C4 – derivative, sedimentation for 2 h 1.Toluene, 2. Acetonitrile, 3. Ethyl acetate 1.Toluene, 2. Acetonitrile, 3. Ethyl acetate

PS‐DCIB C8 – derivative sedimentation for 5 min PS‐DCIB C8 – derivative sedimentation for 2 h 1.Toluene, 2. Acetonitrile, 3. Ethyl acetate 1.Toluene, 2. Acetonitrile, 3. Ethyl acetate

Figure 3.7 Sedimentation experiments for linear PS‐DBIB and linear PS‐DCIB oxidants in reaction media

Sediment suspensions were analyzed under an optical microscope, where the average size of particles and clusters in the organic solution can be defined (Figure 3.8).

In toluene the particles are In ethyl aceate the particles form In acetonitrile the size of the slightly larger than in acetonitrile agglomerates. The size of particles (agglomerates) is or ethyl acetate. agglomerates is between 20‐30 around 20 microns. microns. The size of the single particle is about 1 micron.

After sonification for 1h in After sonification for 1h in After sonification for 1h in toluene. EtOAc. MeCN.

Figure 3.8 Optical microscope pictures of linear PS‐DCIB oxidant dispersed in reaction media

Polymersupported iodine(III) in selective oxidation of alcohols. Oxidants

The size of the agglomerated particles in all solvents was in the range of 10‐40 microns. Obviously the smaller the particle size, the better the size distribution, thus reducing possible mass transfer limitations (substrate‐oxidant) during the reaction. In toluene a high tendency towards cluster formation was observed. This might explain the very fast sedimentation process. However, this observation was not in line with oxidation results in which toluene was an excellent solvent for alcohol oxidation.

Table 3.7 Solvents and oxidants overview in benzyl alcohol oxidation

% conv. of benzyl alcohol to benzaldehyde

solvent ethyl acetate oxidant abbrev. 1h 3h 5h

PS-DAIB C2 derivative 45 56 76

PS-DBIB C4 derivative 19 82 quant.

PS-DHIB C6 derivative 69 82 quant.

PS-DCIB C8 derivative 40 40 82

solvent acetonitrile oxidant abbrev. 1h 3h 5h

PS-DAIB C2 derivative 50 53 62

PS-DBIB C4 derivative 33 81 quant.

PS-DHIB C6 derivative 57 77 75

PS-DCIB C8 derivative 12 23 56

solvent toluene oxidant abbrev. 1h 3h 5h

PS-DAIB C2 derivative 19 23 73

PS-DBIB C4 derivative 27 39 96

PS-DHIB C6 derivative 46 55 72

PS-DCIB C8 derivative 10 16 72

Conditions: 0.16 mmol Benzyl alcohol, 0.16 mmol oxidant, 0.05 mmol of dodecane as internal standard, 70 oC, stirring in 1 mL of solvent.

To improve sedimentation and/or size distribution pre‐sonification was applied to the samples of oxidants. The size distribution was improved by a factor of about 2 (as depicted in Figure 3.8). The effect of pre‐sonication on the oxidation of model substrates was investigated and turned out to be negligible. In Table 3.7 an overview is given of the oxidation of benzyl alcohol in de different solvents. These results were obtained after sonification of the I(III) polymers.

We also investigated the oxidation of benzyl alcohol in the solvent mixtures: toluene/MeCN and MeCN/EtOAc (see Table 3.8) but no improvements were observed. Indeed, the rate of oxidation was lower. After 5 h of reaction full conversion was observed only with the C4 oxidant

Chapter 3 derivative, whereas for the other oxidants more time was required to complete the reaction.

When the oxidation was performed in one solvent a higher rate was observed, e.g. with the C4 oxidant benzyl alcohol was completely converted to benzaldehyde after 5 h using acetonitrile (compare Table 3.7 vs. Table 3.8).

Table 3.8 Oxidation of benzyl alcohol in solvents mixture under optimized conditions

MeCN/Toluene Conv.%

Benzyl alcohol 1h 3h 5h 24h PS-DAIB 41 54 65 quant. PS-DBIB 22 35 quant. quant. PS-DHIB 28 41 60 quant. PS-DCIB 17 31 88 98

MeCN/EtOAc Conv.%

Benzyl alcohol 1h 3h 5h 24h PS-DAIB 2 14 39 53 PS-DBIB 20 21 21 46 PS-DHIB 41 50 71 83 PS-DCIB 21 36 71 quant.

Conditions: 0.16 mmol Benzyl alcohol, 0.16 mmol oxidant, 0.05 mmol of dodecane as internal standard, 70 oC, stirring in 1 mL of solvent.

The general conclusion can be drawn that neither pre‐sonification, nor the use of solvent mixtures was able to influence the oxidation rate. We conclude that all three solvents reported here are suitable for oxidation with iodine(III) polymers. The differences observed vary depending on oxidant and will be presented in more detail in chapter 4.

3.3 Conclusions In summary, iodine(III) oxidants anchored to both cross‐linked and linear polystyrenes as supports were compared in the oxidation of benzyl alcohol and 1‐octanol. It turned out that iodine(III) polymers are excellent oxidations for the selective oxidation of benzyl alcohol in organic solvents, without the need for co‐catalysts such as Br‐. 10 Oxidants based on cross‐linked supports showed lower activity compared to linear polymeric oxidants. Furthermore, longer carbon chain derivatives of polymeric oxidants were analyzed in terms of safety properties and their oxidation reactivity. Based on the trajectory decomposition and temperature the C2 derivative (linear and cross‐linked support) appeared to be the least thermally stable, while the linear C4 and C8 oxidants decomposed slowly. Overall oxidants C4 and C8 – which have active oxygen loadings below 1.7 mmol/g ‐are more stable and thus safer to handle. The solvent

Polymersupported iodine(III) in selective oxidation of alcohols. Oxidants

(acetonitrile, ethyl acetate or toluene) had only a minor influence on the oxidation of the model substrate. Pre‐sonification of the polymeric oxidant did not enhance the performance of the oxidants.

3.4 Experimental details Materials: Alcohols were purchased from Sigma‐Aldrich and used without prior purification. Polystyrene cross linked: Copolymer 98% styrene + 2% divinylbenzene 200‐400 mesh Fluka A.G, Polystyrene linear Mw = 350,000, Mw = 51,000 Sigma Aldrich, required purification, Potassium iodide – Acros p.a., Sodium thiosulfate heptahydrate –Baker p.a., Potassium bromide – Acros IR grade TM p.a., Molecular iodine‐Sigma‐Aldrich p.a., I2O5 from Sigma‐Aldrich, Select Fluor from Sigma‐ Aldrich. Solvents: technical grade were from Baker.

Elemental analysis The iodine content was determined by Instrumental Neutron Activation Analysis (INAA) Apparatus: the “Hoger Onderwijs Reactor” was used as a source of neutrons. The number of neutrons (neutronflux) produced by this reactor measures 1017 neutrons s‐1 cm‐1. The gamma spectrometer uses a Germanium semiconductor as detector and a computer controlled sample changer for optimal usage of the apparatus. The systematical error is estimated 5%. Duplo measurements were performed. INAA is a method for qualitative and quantitative elemental analysis of approx. 60 elements in a solid, liquid or gaseous sample. The method is particularly suitable for measuring trace elements. The method is based on the conversion of stable nuclides into radioactive nuclides (radionuclides) by irradiation with thermal neutrons (neutrons of low energy). These radionuclides stabilize by decline to stable nuclei by transmission of the radioactive Γ‐radiation. The energy ‐of Γ radiation provides information on the radionuclide and the number of pulses (intensity) at certain energy is proportional to the concentration of the element in the sample. The sample requires freeze‐drying to remove any presence of water. The sample (10 to 20 mg) is placed in a polyethylene capsule.

The C, H, O, analyses: Apparatus: EA1108 Elemental Analyser from Carlo Erba Instruments. The experimental error is up to 5%. 10 mg of sample is placed in a tin capsule and brought in a o combustion chamber temperature of 1030 C. When the sample enters the chamber O2‐gas is injected into the helium carrier gas. The temperature in the combustion chamber increases up to 1800 oC by the presence of tin and the exothermic reaction, and the whole sample is fully combusted. The next step is catalytic oxidation followed by reduction of the combustion gases to CO2, H2O N2 and SO2 respectively. These gases are separated on a chromatographic column and detected with a thermal conductivity detector (TCD). The samples were measured independently in duplicate. The standard error is 10‐20%.

Infrared spectroscopy IR spectra were recorded on a PerkinElmer Spectrum One FT‐IR Spectrometer. KBr film pellets were the carrier for solid samples.

Nuclear Magnetic Resonance (NMR) analysis: 13C NMR 100 Hz spectra were recorded on a Bruker AC 400 or Varian Inova VXR‐ 400S spectrometer using TMS as an external standard. Samples were dissolved in CDCl3 as solvent.

Chapter 3

Gas chromatography analysis Analysis were carried out with GC Varian Star 3400 instrument equipped with a polar CP WAX 52 CB 50 m*0.53 mm*2.0 µm column, T max = 250 oC. As internal standard for all oxidation reactions n‐dodecane was used. Column temperature profile: 65 deg (2 min), rate 10 deg/min to 230 deg (9.5 min). Injector temperature profile: 85 deg (2 min), rate 15 deg/min to 250 deg (16 min)

Stability of trivalent iodine compounds

Thermal stability Seteram C80 A sample of approx 90 mg of the material was heated in the Seteram C80 from 30 oC to 250 oC with a heating ramp of 0.5 K/min. The sample was held at 250 oC for 8 h and subsequently cooled to room temperature. During the run the heat flow was recorded.

Thermogravimetric analysis TGA Thermogravimether analyzer Perkin Elmer TGA7. The samples were heated from 25 till 250°C with a heating rate of 10°C/min in air. During the measurement the loss of sample weight was recorded.

The impact sensitivity BAM Fallhammer test, which measures sensitivity to mechanical force. A sample of 40 mm3 of the material was initially exposed to an impact of 40 J six times, by dropping a weight of 10 kg from a height of 0.4meters on a sample enclosed in a stainless steel holder. In case at least one of these impact resulted in an effect (smoke, spark or explosion) the test was repeated with an impact of 7.5 J (a weight of 5 kg from a height of 0.15 meters. In case at least one of these impacts resulted in an effect, the test was repeated with an impact 2 J (a weight of 1kg from a height of 0.2 meters).

Iodometric titration to determine oxidant loading In a 30 mL colourless glass jar, covered with aluminum foil and stopper, approximately 0.1 mmol of sample was weighed. Then 15 mL of an acetic acid/chloroform 2:1 solution and 1.5 mL of 40% KI (60 g in 100 mL of water) solution was added. Solution was stirred at RT for 20 minutes in the dark. The blank became light yellow and the sample became dark brown. The formed iodine was immediately titrated with 0.1 M solution of sodium thiosulfate Na2S2O3 x 7H2O. As an end point of titration discolouration from purple to transparent was taken. Efficient and stable stirring during titration is highly recommended.

Pre‐treatment of crude linear polystyrene Crude pellets were dissolved in methylene chloride; Dissolved crude polystyrene was slowly precipitated by adding to a stirred methanol (proportion 1:150). The white solid started to form. Solid was filtrated through a Buchner funnel. The pure polystyrene was then dried in a vacuum for several days at RT.

Synthesis

Preparation of iodinated polystyrene_Method A) Initially synthesis was adapted from a previously published procedure 17. Polystyrene (20 g) was dissolved in 200 mL nitrobenzene and iodine (20 g) and iodine pentoxide (8 g) were added. Next, 40 mL of 50% H2SO4 and 40 mL CCl4 were introduced and the mixture was heated at 105 οC under reflux conditions with efficient stirring for 7 days. After the reaction was completed, 100 mL of methylene chloride was added and precipitation of the iodinated polystyrene was

Polymersupported iodine(III) in selective oxidation of alcohols. Oxidants performed in 1500 mL of methanol. Precipitation was repeated minimal two times until the dark violet colour disappeared. A light cream solid powder was obtained in nearly quantitative yield.

Preparation of iodinated polystyrene_Method B) The optimized Select Fluor procedure only for demonstration on small scale 15: To a solution of linear polystyrene (7.5 mmol) in 75 mL of MeCN were added corresponding molar amounts of iodine (11.25 mmol) and Select FluorTM (11.25 mmol) and the reaction mixture was stirred at 65 oC for 48 h. After the reaction a crude solid could be filtered off. The crude solid was then dissolved in dichloromethane and precipitated in MeOH. Washing‐precipitation procedure was 13 repeated 3 times until the purple colour had disappeared. Quantitative C NMR (CDCl3, ppm, 100 MHz), δ 144.30‐128.44(5C, C‐Ar, CH2), 91.26(C, C‐I, C‐Ar), 44.55‐40.23 (2C, aliph, CH2)

Oxidation of iodinated polystyrene with peracetic acid under solvent free conditions General procedure17,18 To a cooled (0 οC) solution of acetic anhydride (1.54 mol), aqueous 30% hydrogen peroxide (0.39 mol) was added dropwise over a period of one hour. The resulting peracetic acid solution was brought to ambient temperature and 0.034 mol (8 g) of iodinated polystyrene was introduced to the reaction mixture. The temperature was increased to 50 οC, and the mixture was stirred overnight. The precipitation of an insoluble creamwhite solid indicated formation of the product. Filtration and washing with diethyl ether and water was performed to purify the oxidant.

GC scale oxidation of alcohols with polymeric trivalent oxidant 0.16 mmol alcohol was dissolved in 1 mL of organic solvent (MeCN or EtOAc, or toluene), 0.05 mmol of dodecane as internal standard was added. As oxidant, 1 eq. of oxidized PS‐DAIB (as determined by titration) was introduced after the first t= O sample was withdrawn, then by heating and intense stirring the reaction was started. The progress of the reaction was followed by withdrawing samples of 100 microliters from the reaction mixture. Samples were diluted with 1 mL of diethyl ether. After the reaction was complete (1 to 8 h) the oxidant was filtered off using a glass pipette filled with cotton. The clear solution was then injected to GC.

Oxidation of iodinated polystyrene with peracetic acid under solvent‐free conditions. 17,18 Preparation of PS‐ DAIB (C2 derivative)‐(adopted method from several references). A polymer‐attached iodosobenzene diacetate oxidant was prepared following a two step procedure. In the first step formation of peracetic acid was performed and in the second step oxidation of iodinated polystyrene took place. To a cooled (0 οC ice bath) solution of acetic anhydride (1.54 mol), aqueous 30% hydrogen peroxide (0.39 mol) was added dropwise over a period of one hour. The resulting peracetic acid solution was brought to ambient temperature and 0.034 mol (according to the iodination degree) of iodinated polystyrene was introduced to the reaction mixture. The temperature was increased to 50 οC and the mixture was stirred overnight. The precipitation of an insoluble cream white solid indicates formation of the product. Filtration and washing with diethyl ether and water was performed to purify the oxidant. Cream white polymer‐supported iodosobenzene diacetate PS‐DAIB was obtained as a fine powder, in 73% yield. The quality check via IR analysis of the solid oxidant was performed on KBr film. Characteristic bands CO‐O, CO‐O‐alkyl, C‐O were found (1698 cm ‐1, 1578 cm‐1, 1266 cm‐1). Iodometric titration: 3.73 mmol oxidant/g. This indicates that 97% of the iodine was oxidized (based on an iodine content of 48.7%).

Chapter 3

Preparation of PS‐DBIB (C4 derivative) The synthesis of polymer‐supported iodobenzene dibutyrate involves two steps. In the first step the synthesis of perbutyric acid was carried out according to the known recipes.18 To a cooled (0 οC) mixture of butyric acid (0.2 mol), 25 mL of 95% sulfuric acid and 25 mL of water were slowly added. The colour of the solution should remain light yellow. Next an aqueous 30% hydrogen peroxide solution (0.4 mol) was added dropwise over a period of one hour. The reaction mixture was stirred for an additional 12 h at ambient temperature. Subsequently, several volumes of ice‐water were added to the reaction mixture, which was slowly extracted with diethyl ether. Inorganic acid was removed in the aqueous phase. To the resulting perbutanoic acid solution, 0.04 mol of iodinated polystyrene (90.14% iodination degree) was introduced. The second reaction step was carried out under reflux conditions overnight. The resulting yellow powder was filtered off via a Bűchner funnel. Additional washing with several volumes of diethyl ether was done. The wet product was dried under vacuum at room temperature. 11,5 g of yellow‐white polymer‐attached iodobenzene dibutanoate (PS‐DBIB) as dry fine powder was obtained, with the yield of 66%.

The qualitative IR analysis of the solid oxidant was performed on KBr film. Characteristic bands CO‐O, CO‐O‐alkyl, C‐O were found (1628 cm ‐1, 1580 cm‐1, 1182 cm‐1). Iodometric titration indicated that 1,71 mmol/g (68%) of iodinated polystyrene were oxidized. (Theoretical 2.53 mmol/g).

IR spectrum of PS–DBIB 77.2 75

65 3047.63 1090.20 1046.44 3410.70 1580.10 1280.30 1628.88 1450.12 %T 60 538.38

1182.18 55 1480.61 2921.71 1404.39 701.47 1003.68

50 816.92 1384.25

44.8 4000.0 3000 2000 1500 1000 450.0 cm-1 Preparation of PS‐DHIB (C6 derivative) The preparation of the polymer‐attached iodosobenzene dihexanoate involves two synthesis steps. In the first step peroxyhexanoic acid was obtained in an inorganic acid catalyzed reaction following the known procedure.17,18 To a cooled (0 οC) solution of hexanoic acid (0.2 mol), 90 mL of 50% sulfuric acid (45 mL of 98% H2SO4 in 45 mL of water) were slowly added. The colour of the solution should remain light yellow. Next an aqueous 30% hydrogen peroxide solution (0.4 mol) was added dropwise over a period of one hour. The reaction mixture was stirred for an additional 12 h at ambient temperature. Subsequently, several volumes of ice‐water were added to the reaction mixture, which was slowly extracted with methyl t‐butyl ether. Inorganic acid was removed in the aqueous phase. To the resulting peroxyhexanoic acid organic solution, 0.02 mol of iodinated

Polymersupported iodine(III) in selective oxidation of alcohols. Oxidants polystyrene (98% iodination degree) was introduced. The second reaction step was carried under reflux conditions overnight. The resulting yellow powder was obtained by filtration via a Bűchner funnel. Additional washing with several volumes of diethyl ether was done. The wet product was dried under vacuum at room temperature. 6,5 g of yellow‐white polymer‐attached iodosobenzene dihexanoate (PS‐DHIB) as dry fine powder was obtained. The qualitative IR analysis of the solid oxidant was performed on KBr film. Characteristic bands CO‐O, CO‐O‐alkyl, C‐O were found (1706 cm ‐1, 1581 cm‐1, 1183 cm‐1). Iodometric titration indicated a loading of 3.5 mmol/g. iodine (elemental analysis) = 43.4%. This corresponds to 100% loading of I(III).

IR spectrum of PS–DHIB 66.5 65 1900.25

1581.99 50 1105.11 1062.10 1450.95 539.59 %T 45 3052.24

3407.79 1183.40 40 2856.03 1706.71 1405.98

35 1482.23

30 1004.88 818.11 765.42 2926.34

24.9 4000.0 3000 2000 1500 1000 450.0 cm-1

Preparation of PS‐DCIB (C8 derivative) The preparation of the polymer‐attached iodosobenzene dioctanoate involves two synthesis steps. In the first step peroxyoctanoic acid was obtained in an inorganic acid catalyzed reaction following to the known procedure.17,18 To a cooled (0 οC) solution of caprylic acid (0.2 mol), 40 mL of 50% sulfuric acid (20 mL of 98% H2SO4 in 20 mL of water) were slowly added. The colour of the solution should remain light yellow. Next an aqueous 30% hydrogen peroxide solution (0.4 mol) was added dropwise over a period of one hour. The reaction mixture was stirred for an additional 12 h at ambient temperature. Peroxyoctanoic acid was formed as a white solid. Subsequently, several volumes of ice‐water were added to the reaction mixture, which was slowly extracted with diethyl ether. Inorganic acid was removed in the aqueous phase. To the resulting organic solution 0.04 mol of iodinated polystyrene 10 g (90.45% iodination degree) was introduced The second reaction step was carried out under reflux conditions overnight. The resulting yellow powder was obtained by filtration via a Bűchner funnel. Additional washing with several volumes of diethyl ether was done. The wet product was dried under vacuum at room temperature. 11 g of yellow‐white polymeric product with recovery of ca 45% was obtained.

The qualitative IR analysis of the solid oxidant was performed on KBr film. Characteristic bands CO‐O, CO‐O‐alkyl, C‐O were found (1687 cm‐1, 1582 cm‐1, 1182 cm‐1). Iodometric titration indicates oxygen loading 1.3 mmol/g of the polymer and elemental analysis indicated that 35% of iodinated segments were successfully oxidized with esters groups.

Chapter 3

IR spectrum of PS‐DCIB developed on KBr film 75.8

1582.85 1103.62 60 1359.93 3050.09 1267.13 1061.55 3398.40 1644.05 1225.38 %T 1687.68 1182.67 538.10 1450.09 55

50 2851.40 1404.75 817.29 1481.68 45 2921.86 1004.48

757.46 39.4 4000.0 3000 2000 1500 1000 450.0 cm-1

Calculation of Iodination degree ‐ Based on elemental analysis

I = 44.9% per 100 g of polymer; PS =A; MA= 104 g/mol; PS‐I = B; MB = 230 g/mol; I = C; Mc = 127g/mol I 9.44 Then,  a mol;  35.0 mol of iodinated segments; a x MB = d g of iodinated segments C 127 0.35 x 230 = 81.31 g of iodinated segments; 100‐d = e g of non‐iodinated segments; 100‐81.31 = 18.68 g of non‐iodinated segments e 68.18 x100  f mol of non‐iodinated segments  179.0 mol of non‐iodinated segments M A 104 f 1796.0 Finally, Non‐iodinated degree is:  N ; Non‐iodinated degree is x100 =  ef  35.01796.0 33% Iodinated degree is 100‐N= ID; Iodinated degree is 100‐33 = 67%

Calculation of Iodination degree from quantitative 13 C NMR

Two methods: A and B Area integration ∑ area C‐arom 16259.48 ∑ area C‐I 3192.52

Method A Ratio of C‐arom/C‐I should be 1:5 in theory; Actual ratio is integrated surfaces: 1:5.093 in experiment ∑ area C‐arom‐ ∑ area C‐I = D (difference); 16259.48‐3192.52 = 13066.96; ∑ area C‐I x 5 = P (possible C‐arom value )

Polymersupported iodine(III) in selective oxidation of alcohols. Oxidants

3192.52 x5 = 15962.6; ∑ area C‐arom – P = R (total C‐arom); 16259.48‐15962.6 = 296.88 R V V (one C‐arom value); 296.88 / 6 = 49.48; %100  Nx Degree of non 6carbons   IC iodinated segments (49.48/ 3192.52) x 100%= 1.54% Degree of non iodinated segments 100‐ N = ID iodination degree; 100‐1.54 = 98.45%

Method B ∑ area C‐arom 16259.48 + ∑ area C‐I 3192.52 If everything was iodinated: theoretical value ∑ area C‐arom 16259.48 / 6 = 3242; Actual value is ∑ area C‐I 3192.52. Then, ∑ area (C‐I 3192.52/ 3242) x 100% = 98.45% iodination degree.

Chapter 3

3.5 References

1M. Hudlicky,. Oxidations in Organic Chemistry, ACS, Washington DC, 1990. 2R. A. Sheldon, I. W. C. E. Arends, U. Hanefeld, Green Chemistry and Catalysis, (Eds.), Wiley‐VCH, Weinheim, 2007. 3R. A. Sheldon, I. W. C. E. Arends, A. Dijksman, Catal. Today, 2000, 57, 158. 4B. D. Dess, J. C Martin, J. Am. Chem. Soc., 1991, 113, 7277. 5T. Wirth, Angew. Chem. Int. Ed., 2001, 40, 2812 and references cited therein. 6A. P. Thottumkara, M. S. Browsher, T. K. Vinod, Org. Lett., 2005, 7, 2933. 7A. Schulze, A. Giannis, Synthesis 2006, 257. 8Volumes contributed to hypervalent iodine chemistry: T. Wirth, Ed. Top. Curr. Chem., 2003, 224; V. V. Zhdankin, P. J. Stang, Chem. Rev., 2002, 102, 2523; R. M. Moriarty, J. Org. Chem., 2005, 70, 2893; T. Wirth, R. D. Richardson, Angew. Chem. Int. Ed., 2006, 45, 4402; N. Takenaga, A. Goto, M. Yoshimura, H. Fujioka, T. Dohi, Y. Kita, Tetrahedron Lett., 2009, 50, 3227; T. Dohi, Chem. Pharm. Bull., 2010, 58, 135. 9Combinations of iodine(III) and TEMPO: K. Sakuratani, H. Togo, Synthesis, 2003, 21; Y. Tashino, H. Togo, Synlett., 2004, 2010; T. Y. S. But, Y. Tashino, H. Togo, P.H. Toy, Org. Biomol. Chem., 2005, 3, 970; W. Qian, E. Jin, W. Bao, Y. Zhang, Tetrahedron, 2006, 62, 556; Y. Shang, T. Y. S. But, H Togo, P. H. Toy, Synlett., 2007, 67; Z. Chenjie, W. Yunyang, J. Lei, Synthetic Commun., 2010, 40, 2057; For a catalytic version see: C. I. Herrerias, Y. Zhang, C‐J. Li, Tetrahedron Lett., 2006, 47, 13. 10Combinations of iodine(III) and Br‐: H. Tohma, S. Takizawa, T. Maegawa, Y. Kita, Angew. Chem. Int. Ed., 2000, 39, 1306; H. Tohma, T. Maeagawa, S. Takizawa, Y. Kita, Adv. Synth. Catal., 2002, 344, 328; W. Qian, E. Jin, W. Bao, Y. Zhang, Angew. Chem. Int. Ed., 2005, 44, 952; For a catalytic version see: R. Mu, Z. Liu, Z. Yang, L. Wu, Z.‐L. Liu, Adv. Synth. Catal., 2005, 347, 1333; T. Dohi, Y. Kita, Chem. Commun., 2009, 2073. 11A. E. J. de Nooy, A. C. Besemer, H. van Bekkum, Synthesis, 1996, 10, 1153. 12H. Tohma, T. Maeagawa, S. Takizawa, Y. Kita, Adv. Synth. Catal., 2002, 344, 328. 13S. V. Ley, A. W. Thomas, H. Finch, J. Chem. Soc., Perkin Trans., 1, 1999, 669. 14H. Togo, G. Nogami, M. Yokoyama, M. Synlett., 1998, 534. 15S. Starber, P. Kralj, M. Zupan, Synthesis, 2002, 11, 1513. 16R. Nomen, HarsBook. Harsnet (Thematic Network on Hazard Assessment of Highly Reactive Systems), 2006. 17Synthesis of polymeric iodine(III) oxidant PS‐DAIB: Y. Yamada, M. Okawara, Die Makromoleculaire “LXIII Synthesis and reactions of functional polymers”, 1972 152, 153; H. Togo, G. Nogami, M. Yokoyama, Synlett., 1998, 534; G‐P. Wang, Z‐Ch. Chen, Synth. Commun., 1999, 29, 2859; S. Abe, K. Sakuratami, H. Togo, J. Org. Chem., 2001, 66, 6174; X. Huang, Q. Zhu, Synth. Commun., 2001, 31, 111; I. R. Baxendale, S. V. Ley, M. Nessi, C. Piutti, Tetrahedron, 2002, 58, 6285; For regeneration and reuse see: H. Togo, G. Nogami, M. Yokoyama, Synlett., 1998, 534; Ch. A. Briehn, T. Kirschbaum, P. Bäuerle, J. Org. Chem., 2000, 65, 352; X. Huang, Q. Zhu, Synth. Commun., 2001, 31, 111; X. Huang, Q. Zhu, Y. Xu, Synth. Commun., 2001, 31, 2823. 18Synthesis of peracids catalyzed by acid: W. E. Parker, C. Ricciuti, C. L. Ogg, D. Swern, J. Am. Chem. Soc., 1955, 77, 4037; S. Effkemann, U. Pinkernell, R. Neuműller, F. Schwan, H. Engelhardt, U. Karst, Anal. Chem., 1998, 70, 3857; I. R. Baxendale, S. V. Ley, M. Nessi, C. Piutti, Tetrahedron, 2002, 58, 6285; H. Togo, G. Nogami, M. Yokoyama, Synlett., 1998, 534.

CChhaapptteerr 44

Polystyreneattached iodine(III) reagents in the selective oxidation of alcohols. Scope of substrates.

4.1 Introduction 4.2 Results and Discussion 4.2.1 Oxidation of benzylic alcohols 4.2.2 Oxidation of aliphatic and cyclic alcohols 4.2.3 Oxidation of allylic alcohols 4.2.4 Oxidation of functionalized alcohols (incl. heterocyclic alcohols) 4.2.5 Oxidation of steroidal alcohols 4.2.6 Less reactive, unreactive alcohols, non‐selective oxidations 4.2.7 Oxidation of selected substrates with iodine(III) in ionic liquids 4.2.8 Mechanistic studies 4.2.9 Regeneration and recycling of the oxidant 4.3 Conclusions 4.4 Experimental details 4.5 References

Abstract Polymer‐attached iodine(III) oxidants were shown to be effective in the selective oxidation of various primary and secondary alcohols to the corresponding aldehydes and ketones, in organic solvents. The less reactive steroidal alcohols also underwent facile oxidation leading to nearly quantitative yields and high selectivity within a short reaction time. The method avoids the formation of waste and the catalyst can be easily separated by filtration and recycled. A two‐step mechanism for the reaction was postulated based on competition studies and a Hammett plot study for the oxidation of substituted benzylic alcohols. Results from the competition experiments contribute to the mechanistic studies.

Chapter 4

4.1 Introduction In the preceding chapter a variety of polymer‐attached iodine(III) oxidants were presented.

Namely polymer‐attached iodosobenzene diacetate PS‐DAIB (C2 derivative), dibutanoate PS‐

DBIB (C4 derivative), dihexanoate PS‐DHIB (C6 derivative) and dioctanoate PS‐DCIB (C8 derivative). Properties such as reactivity and their general performance in alcohol oxidations were drafted. We could see that the thermal stability and the onset of decomposition of the various derivatives differed. Also differences are expected in oxidation performance. From a safety point of view derivative C4, C6, or C8 of the oxidant appeared to be safe in handling while the C2 derivative was the least stable in thermal or impact sensitivity experiments. From the preceding chapter we have learnt that linear polymer‐attached iodosobenzene dialkanoate oxidants are able to oxidize alcohols without any need of additional reagents.

We here wish to report for the first time the full potential of iodine(III) itself as an excellent oxidant of alcohols, applicable to a wide range of alcohols and using 1.0 equivalent of oxidant. In literature, the use of hypervalent iodine as oxidant is usually accompanied by bromide salts as additive 1, or TEMPO as catalyst. 2

For this purpose, we used new iodine(III) derivatives supported on linear styrene polymer. It has been shown that polymeric iodine reagents are very stable and much more user‐friendly compared to their soluble analogues. 3 The oxidation by polymeric iodine(III) is visualized in Figure 4.1.

CH3 O PS I O O

O CH3 R OH R O

Figure 4.1 Oxidation of alcohols using PS‐DAIB as the sole reagent

In addition, the polymeric iodine(III) reagent (C2 derivative) can be easily regenerated after use 4 by oxidation with H2O2 in acetic anhydride. Results demonstrate its selectivity and the potential to use a variety of organic solvents. No excess iodine is required and the reagent can be fully regenerated and reused at least two times afterwards. 5

This chapter describes the oxidation of a variety of substrates. A broad range of alcohols, such as aliphatic, benzylic, allylic and cyclic, were tested in an optimized oxidation system. The influence of double bonds in unsaturated alcohols, functional groups and heteroatoms will also be discussed. In order to get more insight in the mechanism competition and initial rate studies

PSattached iodine(III) reagents in sel. oxidation. Scope of substrates. were carried out. In addition a Hammett linear energy relationship was constructed for benzylic alcohol oxidation.

4.2 Results and Discussion

In order to draw the general performance of the commonly recognized C2 derivative (PS‐DAIB) iodine(III) oxidant a number of oxidation results are presented. Based on the results obtained, the optimal reaction conditions were designed and further applied with other derivatives of iodine(III) oxidants. In Table 4.1 the results for PS‐DAIB are depicted.

Table 4.1 Primary and secondary alcohol oxidation using PS‐DAIB as the oxidant

Yields of aldehyde (%) entry alcohol ethyl acetate acetonitrile toluene toluene/100°Cb

primary alcohols

1 OH 95 92 97

2 OH 93 93 99 90

OH 3 94 95 97

4 OH 90 86 98

OH 5 100c 67c 75c 97

OH 6 72c 79c 70c 92

7 78c 89c 80c O OH secondary alcohols OH 8 98 70 85 91 OH

9 66 50 48 60

10 33 41 51 51 OH OH 11 55 54 57

12 56 60 58 50 OH OH 13 68 89 95 95 OH

14 97 88 70

15 90 97 92

aConditions: 0.16 mmol alcohol; 0.16 mmol. PS-DAIB, 1 ml solvent, 70°C, 5h; b Time 2 h; c Time 4 h

Chapter 4

We found that the reactivity of the iodine(III) oxidant in organic solvents is generally enhanced at higher temperatures (results of the last column, Table 4.1) and with efficient vigorous stirring. Oxidations were carried out at a range of temperatures: room temperature, 40, 50, 70, 80, 100 oC (results not shown). At ambient temperature the rate of benzyl alcohol oxidation is very slow. On heating the rate of oxidation increased and at 100 oC most of the alcohols are converted to the corresponding carbonyl compound in 1 hour or less. Considering safety aspects and future large scale application an optimum reaction temperature of 70 oC was chosen.

We believe that the ultimate strength of the method will be in the conversion of functionalized alcohols. In Table 4.2 a variety of functionalized alcohols is converted with high selectivity.

Table 4.2 Oxidation of functionalized alcohols with linear PS‐DAIB oxidant in acetonitrile at 70 °C

entry alcohol Yield of aldehyde (%) Time 1 3‐(hydroxymethylpyridine) 94 4 h 2 3‐(hydroxymethylpyridine)* quant. 40 min 3 p‐chlorobenzylalcohol quant. 24 h 4 p‐chlorobenzylalcohol * 97 40 min 5 m‐iodobenzyl alcohol 96 5 h 6 m‐fluorobenzyl alcohol 98 5 h 7 furfuryl alcohol 78 5 h 8 p‐methoxybenzyl alcohol quant. 8 h 9 2,5‐dimethoxybenzyl alcohol 90 24 h 10 citronellol 88 5 h 11 Cinnamyl alcohol 78 5 h Conditions: 0.16 mmol alcohol, 0.16 mmol linear PS‐DAIB (loading 3.73 mmol/g), 70 οC, 1 mL solvent. Yield based on GC using n‐dodecane as internal standard during the reaction. *Reaction temperature was 100 οC.

Our interest was to evaluate the performance of the potentially safer derivatives of the iodine(III) oxidants. Below systematic studies for oxidation using the previously synthesized C4,

C6 and C8 derivatives of PS‐DAIB are presented.

4.2.1 Benzylic alcohols In general benzylic alcohols are more reactive than aliphatic ones. We observed that using 1 eq. of oxidant, without any additional reagent, all benzylic alcohols undergo smooth oxidation to the corresponding carbonyl products with 100% selectivity within a reaction time from 1 up to 5 h. Toluene, ethyl acetate or acetonitrile were used as solvents.

PSattached iodine(III) reagents in sel. oxidation. Scope of substrates.

Oxidation using PS‐DBIB (C4‐derivative), PS‐DHIB (C6 derivative) and PS‐DCIB (C8 derivative)

The results of oxidations of several benzylic alcohols with polymer‐attached iodosobenzene dibutanoate are shown in Table 4.3.

Table 4.3 PS‐DBIB in oxidation of substituted benzyl alcohols

Applied oxidant PS-DBIB abbreviated as C4 derivative

substituent(s) solvent time (h) conv. (%)

-H CH3CN 3 96.3 toluene 3 quant. EtOAc 1 96.4

p -methoxy CH3CN 3 quant. Toluene 1 99.4 EtOAc 24 76

m -methoxy CH3CN 5 quant. Toluene 3 99.7 EtOAc 3 quant.

Conditions: 0.16 mmol alcohol, 0.16 mmol oxidant PS‐DBIB (loading 1.71 mmol/g), 1 mL of solvent, 70 oC. Conversions based on GC results using 0.05 mmol dodecane as internal standard. Selectivities were > 99% in all cases.

PS‐DBIB exhibited the best reactivity in benzylic alcohol oxidation in toluene as solvent. The desired aldehyde was obtained within 1 to 3 h. The short reaction time indicated that the C4 oxidant showed high reactivity. Electron donating substituents, such as methoxy in para position (p‐OCH3), had a rather minor influence on the reaction rate. Therefore, we conclude that electronic effects are not a serious rate limiting factor.

Similarly, only relatively minor substituent effects were observed with the C6 and C8 derivatives in the oxidation of a series of substituted benzylic alcohols as shown in Table 4.4.

Chapter 4

Table 4.4 The influence of substituents in benzyl alcohol oxidation with PS‐DHIB and PS‐DCIB.in acetonitrile

Applied oxidants: PS-DHIBa PS-DCIBb

substrates conv. alcohol conv. alcohol (%) (%) after 5h after 5h EDG benzyl alcohol 66 78 p -methoxybenzyl alcohol 77 71 m -methoxybenzyl alcohol 49 31 2,5-dimethoxybenzyl alcohol quant. 66 3,4-dimethoxybenzyl alcohol 65 54 p -methylbenzyl alcohol 87 - EWG p -(trifluoromethyl)benzyl alcohol 36 66 m -iodobenzyl alcohol 78 71 p -fluorobenzyl alcohol 46 50 p -chlorobenzyl alcohol 61 -

a Oxygen loading 3.52 mmol/g; b Oxygen loading 1.29 mmol/g; EDG – electron donating groups, EWG‐ electron withdrawing groups

Both for electron donating as well as for electron withdrawing groups reasonable conversions were obtained: 49% or higher in most cases. The highest activity was obtained with 2.5‐ dimethoxybenzyl alcohol, a quantitative conversion using PS‐DHIB as oxidant. Overall the use of PS‐DHIB and PS‐DCIB resulted in slightly lower primary benzylic alcohols conversions compared to PS‐DBIB. e. g. for p‐methoxybenzyl alcohol 100% conversion (3 h, CH3CN) was obtained with PS‐DBIB, while for PS‐DHIB and PS‐DCIB conversions of 77% and 71% were found respectively.

Results obtained with the secondary benzylic alcohol, 1‐phenylethanol, and the isomeric primary alcohol, 2‐phenyethanol, are shown in Table 4.5. The former was selectively converted to acetophenone after more than 5 h, while the latter needed ca. 3 h to be converted to phenylacetaldehyde as the only product. The C4 oxidant was much more active than the C8 oxidant in the oxidation of the secondary alcohol, while both C4 and C8 derivatives were equally effective with the primary alcohol.

PSattached iodine(III) reagents in sel. oxidation. Scope of substrates.

Table 4.5 Primary vs. secondary alcohol oxidation with either PS‐DBIB or PS‐DCIB (separate experiments)

Applied oxidants: PS-DBIBa PS-DCIBb

alcohol solvent conv. alcohol conv. alcohol (%) (%) 2-phenylethanol CH CN quant. (5h) 97 (5h) OH 3 toluene 91 (3h) 95 (5h) EtOAc 97 (3h) 99 (3h)

1-phenylethanol

OH CH3CN 90 (5h) 38 (5h) toluene 74 (5h) 40 (5h) EtOAc 52 (5h) -

a PS‐DBIB 1.71 mmol/g; b PS‐DCIB 1.29 mmol/g

4.2.2 Aliphatic and cyclic alcohols Aliphatic alcohols

Long chain aliphatic alcohols were tested under optimized conditions using 1 eq. of PS‐DBIB or PS‐DCIB in toluene, acetonitrile and ethyl acetate as reaction media. The results, after 5 hours of the reaction, are shown in Figure 4.2. In general, they were less reactive compared to benzylic alcohols. Most of the primary aliphatic alcohols required a minimum of 5 h to undergo oxidation to aldehydes. No linear tendency was observed for the reactivity of alcohols in the range C6‐C12. Solvent effects were generally minor but the fastest transformation was observed in toluene, and the slowest reactions were observed for acetonitrile (Figure 4.2).

Chapter 4

Figure 4.2 Oxidation of primary aliphatic alcohols with PS‐DBIB polymeric oxidant (reaction time 5 h)

The corresponding oxidations of aliphatic primary alcohols with the C8 derivative are shown in

Figure 4.3. The C8 derivative generally showed a lower activity compared to the C4 derivative and little difference was observed with increasing chain length of the alcohol. After 5 hours of the reaction the average conversion was about 30‐50%. When reactions were allowed to run for 24 hours complete conversion was observed for nearly all the tested alcohols (Figure 4.3).

Figure 4.3 Primary aliphatic alcohols oxidation with PS‐DCIB oxidant in acetonitrile

PSattached iodine(III) reagents in sel. oxidation. Scope of substrates.

In the case of the less reactive secondary aliphatic alcohols such as 2‐hexanol or 3‐hexanol up to 2‐nonanol or 3‐nonanol the success of oxidation under optimized conditions (1 eq. oxidant PS‐ DBIB, 70 0C, and organic solvent) is strongly dependent on both temperature and thechoice of organic solvent. With acetonitrile as solvent the oxidation rate is comparable to that of primary aliphatic alcohols and ca. 98% conversion is obtained within 5 hours. The reactivity of different secondary alcohols e.g. as with 2‐octanol vs. 3‐octanol is very similar. Both alcohols are fully converted to their keto products after 24 h of reaction. The rate of oxidation is very similar and after 3 h of reaction ca. 60% substrate is converted when PS‐DBIB oxidant in acetonitrile was used. This illustrates the high potential of the polymeric C4 oxidant, which is able to oxidize less exposed hydroxyl groups. The reactivity remains constant with increasing chain length in the case of the secondary aliphatic alcohols. In contrast, the rate of reaction for secondary alcohols with the C8 derivative is rather low. After 5 hours of the reaction in acetonitrile only ca. 30% conversion was obtained, while when the C4 derivative is applied under the same conditions the reaction was nearly complete. The summarized results of secondary alcohol oxidations with either C4 or C8 oxidant are presented in Table 4.6.

Table 4.6 Secondary aliphatic alcohol oxidations to ketones with either C4 or C8 polymeric iodine(III) oxidants

Applied oxidants: PS-DBIB PS-DCIB

alcohol structure conv. alcohol conv. alcohol (%) (%) after 5h after 5h

2-hexanol quant.a 30.5 OH

2-octanol 95.8 36.7 OH

3-octanol 92.9 36.3 OH

2-nonanol 93.2 33.8 OH

3-nonanol 95.9 31.8 OH

aConversion reached within 3 h. For conditions see Table 4.3, solvent is acetonitrile.

Chapter 4

Cyclic alcohols

Cyclohexanol was chosen as a representative alicyclic alcohol substrate. The highest reaction rate was obtained when polystyrene‐attached iodosobenzene diacetate PS‐DAIB (C2 derivative) was used. The reaction was complete within 5 h, while for PS‐DBIB or PS‐DCIB after 3 h of reaction the best result was obtained in acetonitrile, giving 78% of conversion. Alkyl‐substituted cyclohexanols displayed low activities which we attribute to the more crowded and less exposed hydroxyl groups being less accessible to the iodine(III) oxidant (C8 derivative). The results are summarized in Table 4.7.

Table 4.7 Polystyrene‐attached iodine(III) reagents reactivity in cyclic alcohol oxidations

Applied oxidants: PS-DAIB PS-DBIB PS-DCIB

alcohol solvent conv. alcohol conv. alcohol conv. alcohol (%) (%) (%) cyclohexanol

CH3CN quant. (5h) quant.(5h) 38.0 (5h) OH toluene 88.3 (5h) 78.5 (5h) 42.1 (5h) EtOAc 95.4 (5h) 47.3 (5h) 89.2 (5h)

2,6-dimethylcyclohexanol

OH CH3CN quant. (48h) 76.6 (48h) 58.0 (48h)

2-methylcyclohexanol

CH3CN 49.9 (5h) n.m n.m OH toluene 62.0 (24h) n.m n.m EtOAc 71.2 (24h) n.m n.m

4.2.3 Oxidation of allylic alcohols

Primary allylic alcohols, such as geraniol and cinnamyl alcohols, were tested in the oxidation with polymeric iodine(III) (C2‐C8) derivatives (Figure 4.4) under optimized conditions.

OH OH

geraniol cinnamyl alcohol

Figure 4.4 Allylic substrates

The oxidation rate for these alcohols varies with the oxidant used. For example geraniol was

o o quantitatively converted to citral within 1 h (100 C, C2, derivative or 70 C, C8 derivative), when

PSattached iodine(III) reagents in sel. oxidation. Scope of substrates. either polystyrene‐attached iodosobenzene diacetate or dioctanoate was used as the oxidant. For other oxidants longer reaction times were required. When the reaction temperature was increased from 70 oC to 100 oC nearly complete transformation (97%) of geraniol to citral was obtained after only 40 min with PS‐DAIB in acetonitrile. Since very fast reaction is more difficult to control on a lab scale it was decided to follow the reaction performance at 70 oC.

In the oxidation of cinnamyl alcohol oxidants with PS‐DHIB and PS‐DCIB 3 h and 5 h reaction times were required, respectively, for quantitative conversion to cinnamyl aldehyde in 100% selectivity. In contrast, with PS‐DHIB as an oxidant, geraniol was transformed to citral only after 24 h. For summarized results see Table 4.8.

Table 4.8 Allylic alcohol oxidations with polystyrene‐attached iodine(III) in organic solvents

yield of aldehyde (%) allylic alcohol solvent time 1h 3h 5h 24h

Applied oxidant PS-DAIB

geraniol CH3CN 22 69 79, 9(6h) quant. geraniol EtOAc 37 55 85, 9(6h) quant. o geraniol at 100 C CH3CN 97(40min)

cinnamyl alcohol CH3CN 29 70 79 quant. cinnamyl alcohol EtOAc 29 61(4h) 72 quant.

Applied oxidant PS-DBIB

geraniol CH3CN 57 quant. geraniol EtOAc 64 quant. geraniol Toluene 77 99.7 quant

Applied oxidant PS-DHIB

geraniol CH3CN 41 60 69 76

cinnamyl alcohol CH3CN 50 quant.

Applied oxidant PS-DCIB

geraniol CH3CN 34 61 79 100 geraniol Toluene 97 quant.

cinnamyl alcohol CH3CN 47 69 quant.

Conditions: 0.16 mmol of substrate, 0.16 mmol of oxidant (corresponding to active oxygen), 0.05 mmol of dodecane as internal standard, 1mL of solvent , 70 oC selectivities > 99% in all cases. Loadings of the oxidants: PS–DAIB 3.73 mmol/g; PS‐DBIB 1.71 mmol/g; PS‐DHIB 3.52 mmol/g; PS‐DCIB 1.29 mmol/g

2‐cyclohexene‐1‐ol was chosen as an example of a cyclic allylic alcohol. The PS‐DBIB oxidant showed overall the best reactivity. The initial rate of the oxidation rate was much higher than that of cyclohexanol under optimized conditions and the olefinic double bond remained untouched. 2‐cyclohexene‐1‐one was formed in 96% yield in 1 hour compared to 35% yield of

Chapter 4 cyclohexanone in 1 h under the same conditions. Good results were obtained for 2‐ cyclohexene‐1‐ol in acetonitrile (1 h, 96%), as well as ethyl acetate or toluene (quantitative yield in 3 h) (Figure 4.5).

100

yield of ketone [ %] 0135 time [h]

cyclohexanone 2-cyclohexenone

Figure 4.5 Reactivity of cyclohexanol vs. 2‐cyclohexene‐1‐ol in oxidative transformations with PS‐DBIB as sole oxidant in acetonitrile

4.2.4 Oxidation of functionalized alcohols (incl. heterocyclic alcohols)

Oxidation of alcohols containing various functionalities such as heteroatoms and double bonds, in general result in low selectivities or sometimes no reaction. Heterocyclic alcohols and steroidal alcohols are often very difficult to oxidize. In contrast, polymer‐attached iodine(III) oxidants proved to be highly effective in the oxidation of these types of alcohols. Examples of heterocyclic substrates are shown in Figure 4.6. The reactions were followed by TLC.

(A)

(B)

Figure 4.6 Oxidation of heterocyclic alcohols with iodine(III) reagent. All derivatives were tested in either acetonitrile or toluene or ethyl acetate.

PSattached iodine(III) reagents in sel. oxidation. Scope of substrates.

In the case of substrate from equation A) (Figure 4.6), the rate of oxidation was the highest in toluene, where after 5 h the desired product was obtained in quantitative conversion. For other tested solvents (ethyl acetate, acetonitrile) the reaction was complete only after 24 h. For the second tested heterocyclic alcohol from equation B) (Figure 4.6) the oxidation was complete within 5 h, in all tested solvents. The reactivity for all polymeric iodine based oxidants was also the same. High selectivities were observed for both substrates.

4.2.5 Oxidation of steroidal alcohols

Potentially less reactive or highly satirically hindered steroidal alcohols were smoothly and selectively oxidized to their carbonyl products when polystyrene‐attached iodine(III) oxidant was applied. Tested substrates are shown in the Figure 4.7.

(A)

O OH O OH H H H H H

H H H H H H OH O O

3α-hydroxy-5α-H-pregnanolone nandrolone 11α-hydroxynordione

(B)

O O O O H H H H H

H H H H H H O O O

5α-H-3,20-pregnanedione nordione 11-ketonordione

Figure 4.7 Steroidal alcohol substrates (A) and their products (B)

All three test substrates were oxidized between 3 and 24 h of reaction time under initially optimized conditions. There was a negligible difference in reactivity of the different iodine(III) oxidants. A simple work‐up procedure, namely filtration of the oxidant, and solvent evaporation were the only operations required to obtain pure steroidal ketone. Progress of oxidation was

Chapter 4 followed by TLC. With 3α‐hydroxy‐5α‐H‐pregnanolone and nandrolone (entry 1 and 2, Table 4.9), isolated yield and selectivity of the products were confirmed with NMR and GC.

Table 4.9 Steroids oxidation using PS‐DBIB – Examples of gram scale application

Entry Substrate Product yield Sel. time oxidant solvent (%) (%) (h) derivative

1 3α-hydroxy-5α-H-pregnanolone 5α-H-3,20-pregnanedione 97* 99.8 5 PS-DBIB* CH3CN

2 nandrolone nordione 96* 99.3 3-5 PS-DBIB* CH3CN

311α-hydroxynordione 11-ketonordione 99% conv. 100** 8-24 all all

* isolated yields, PS‐DBIB loading 1.71 mmol/g; ** selectivity based on TLC

We conclude that polymer‐attached iodine(III) oxidants are excellent reagents for the oxidation of biologically active steroidal alcohols to the corresponding ketones with high (> 99%) selectivity on a gram scale. Use of acetonitrile was the most beneficial in order to isolate the product in a simple manner. In addition, other solvents such as toluene and ethyl acetate were also used as reaction media in GC‐scale experiments.

The successful oxidation of steroid alcohols prompted us to investigate the oxidation of other poorly reactive alcohols under oxidation conditions as described below.

4.2.6 Less reactive, unreactive alcohols, nonselective oxidations

Notwithstanding the success of polystyrene‐attached iodine(III) oxidants with a wide variety of alcohol substrates there remained a small group of alcohols which do not undergo selective oxidation and even in some cases no oxidation is observed (Figure 4.8). Aliphatic alcohols or diols containing heteroatoms (O, S) were totally unreactive. Aliphatic diols containing atoms (Br, Cl) and SH groups undergo oxidation but this led to the formation of several difficult to identify products.

PSattached iodine(III) reagents in sel. oxidation. Scope of substrates.

OH SH Br SH OH Cl OH Br OH

1,4-dithioerythritol 2,3-dibromopropanol 1-chloro-3-hydroxypropane

OH OH O S OH S O OH OH

2,2-thiodiethanol 2-(2-methoxyethoxy)ethanol 3-methylthio-1,2-propanediol

O OH OH OH OH

OH OH

lactic acid 1,2-octanediol 1,5-hexanediol

OH OH OH OH

1,5-pentanediol 1,10-decanediol

Figure 4.8 Unreactive or difficult substrates

4.2.7 Oxidation of benzyl alcohol with iodine(III) in ionic liquids Ionic liquids have emerged in recent years as interesting reaction media in various chemistry and bio‐related disciplines (in chapter 6 more attention is given to ionic liquids as reaction solvents). 6 For example, organic pentavalent iodine reagents (IBX and DMP) showed promising alcohol oxidation activity (often faster reaction) when hydrophilic or hydrophobic imidazolium

+ ‐ + ‐ + ‐ based ionic liquids such as: [BMIm] [Cl] , [BMIm] [BF4] and [BMIm] [PF6] , respectively, were used as reaction solvents compared to conventionally used DMSO, DMF, EtOAc or water. 7,8

We have studied the use of several novel ionic liquids with strongly coordinating, oxidation inert anions such as nitrate in the oxidation of benzyl alcohol (for ionic liquids structures see

+ ‐ Figure 4.9). In our study we have also included more classical non‐coordinating [BMIm] [BF4] + ‐ and [BMIm] [PF6] .

Chapter 4

‐ ‐ ‐ ‐ Anions: nitrate [NO3] , glycolate [HOCH2CO2] , tetrafluoroborate [BF4] , hexafluorophosphate [PF6] , ‐ formate [HCO2]

Figure 4.9 Ionic liquids used as solvents for benzyl alcohol oxidation

The results using PS‐DBIB as oxidant for the oxidation of benzyl alcohol in ionic liquids are shown in Table 4.10.

Table 4.10 Benzyl alcohol oxidation in ionic liquids using PS‐ DBIB as oxidant (70 oC)

Benzyl alcohol entry ILs cation anion % conv. remarks

- 1 HOPMIm NO3 79 (24h) dissolution of the oxidant 2 HOPMIm glycolate - no reaction 3 BMIm nitrate 80 (3h) formation of acid 4 choline nitrate* 48 (3h) - 5BMImBF4 16 (3h) -** 6BMImPF6 24 (3h) - 7TEAHCO2 12 (24h) dissolution of the oxidant ***

* 50 v/v of acetonitrile; ** hydrophobic ionic liquid was used; *** oxidant was dissolved only at reaction temperature 70 oC

+ ‐ Oxidation in [HOPMIm] [NO3] (entry 1, Table 4.10) gave reasonable conversion of benzyl alcohol to aldehyde (selectivity > 99%). Oxidative transformation in other ionic liquids gave modest results, or resulted in lower selectivities with overoxidation to acids (entry 3, Table 4.10 30% of acid already formed). Interestingly dissolution of the oxidant PS‐DBIB was observed during the course of the reaction (entry 1 and entry 7, Table 4.10) at 70 oC.

PSattached iodine(III) reagents in sel. oxidation. Scope of substrates.

4.2.8 Mechanistic studies Competition experiments

We performed competition experiments by allowing a 1:1 mixture of primary and secondary alcohol to react with a sub‐stoichiometric amount of PS‐DAIB (polystyrene‐attached iodosobenzene diacetate) in both toluene and acetonitrile as solvents. The results are presented in Table 4.11

Table 4.11 Competition reaction of primary versus secondary alcohols using PS‐DAIB as oxidant

Entry substrate yield (%) o o in acetonitrile, C 2 oxidant derivative 5h k1 /k2

1 benzyl alcohol vs. 1-phenylethanol 47/4 15.6

2 1-octanol vs . 2-octanol 31/15 2.3

3 1-nonanol vs. 2-nonanol 29/14 2.3

yield (%) o o In toluene, C 2 oxidant derivative 5h k1 /k2

4 benzyl alcohol vs. 1-phenylethanol 48/28 2.0

5 1-octanol vs. 2-octanol 25/13 2.1

6 1-nonanol vs. 2-nonanol 23/9 2.8

Conditions: 0.16 mmol oxidant PS‐DAIB (loading 3.73 mmol/g), 0.08 mmol primary alcohol, and 0.08 mmol of secondary alcohol, solvent 1 mL, 0.05 mmol internal standard (n‐dodecane), temperature 70 °C

We observed that primary alcohols were oxidized faster in all cases. Separate experiments were performed to follow the initial time course of the reaction (see Figure 4.10). These studies reveal that the latter difference between benzylic alcohol and 1‐phenylethanol oxidation only occurs in a later stage of the reaction. The initial rates are comparable.

Chapter 4

(A)

% 16 product 12 carbonyl 8 the

4 yield

0 5 10203060

time [min]

benzaldehyde acetophenone

(B) 25 %

20 product 15

10 carbonyl

the 5

Yield 0 5 10 20 30 60

Time [min] nonanal 2‐nonanone

Conditions: alcohol: oxidant 1 eq. 0.08 mmol of an alcohol, oxidant 0.04 mmol of PS‐DAIB in respect to its loading (loading 3.73 mmol/g), 0.05 mmol internal standard, 0.5 mL acetonitrile, temperature 70 oC

Figure 4.10 The initial oxidation rate in the competition experiments. (A) Benzyl alcohol vs. 1‐ phenylethanol; (B)1‐nonanol vs.2‐nonanol

Hammett plot and proposed mechanism

In order to identify the electronic effect for the oxidation of benzylic alcohols by polymer‐ attached iodosobenzene dihexanoate (PS‐DHIB) we constructed a Hammett plot. The logarithm of the initial oxidation rate of various substituted benzyl alcohols using C6 derivative of hypervalent iodine(III) oxidant was plotted against the σ‐value of the para or meta— substituents, resulting in a linear relationship (Figure 4.11).

PSattached iodine(III) reagents in sel. oxidation. Scope of substrates.

0.4 m-CH3

0.2 log(kx/kH) p-CH3

p-OCH3 H 0 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 -0.2 sigma

-0.4 p-F

-0.6

m-Cl p-CF -0.8 3 m-I

-1 ρ = -1.78 x

-1.2

Figure 4.11 Hammett plot for the oxidation of substituted benzyl alcohols with polystyrene‐attached

iodosobenzene dihexanoate (PS‐DHIB), (C6 derivative, loading 3.53 mmol/g)

The relatively large negative slope of ρ = ‐1.78 indicates that a large accumulation of electrons is required in the transition state. This is consistent with hydrogen elimination being rate limiting. Recently for Pd‐nanoclusters in ferritin a similar ρ value (ρ = ‐ 1.65) was obtained at 80 oC. 9 The oxidation rate increases with electron donating groups (EDG) and decreases with electron withdrawing groups (EWG). A negative ρ value is also consistent with electrons being removed from the aromatic ring in the rate determining step.

We postulate the following two step mechanism (Figure 4.12). In the first step the alcohol coordinates to the electrophilic iodine(III) and one acid molecule is displaced and released to the solution. In the second step β (beta) proton elimination is accompanied by reduction of iodine(III) to iodine(I) and formation of a carboxylate anion. Based on the Hammett plot relationship we suggest that the second step is rate limiting.

Chapter 4

O O

O O R' R

PS I I H H O R' O

R O O + R OH R' OH

I +2 + R O R' OH Figure 4.12 Proposed oxidation mechanism

4.2.9 Regeneration and recycling of the oxidant It is recognized from this work and from literature that the polystyrene‐supported iodine(III) reagent can be easily regenerated and recycled. 5 We were interested in the performance of both oxidants with C2 and longer alkyl chain (C4 derivative) in the recycling tests. The iodine reagent was first removed from the reaction mixture by filtration, then dissolved in dichloromethane and precipitated with methanol. Regeneration was performed by treating the

o reduced reagent with freshly prepared organic peracid (peracetic C2 or perbutyric C4) at 50 C for 12 h according to the standard procedure (experimental section) and the thus regenerated oxidant was applied for 3 cycles in the various alcohol oxidations. In Table 4.12 the results of the recycle experiments are summarized.

PSattached iodine(III) reagents in sel. oxidation. Scope of substrates.

Table 4.12 Regeneration and reuse of the polymeric oxidant PS‐DAIB over 3 cycles

yield (%) Entry substrate product 1st run 1st recycle 2nd recycle 3rd recycle

1benzyl alcohol benzaldehyde 97 96 87 91

21‐octanol octanal 93 92 89 90

31‐hexanol hexanal 92 92 78 80

4 geraniol citral 80 *7776 ‐

5 cyclohexanol cyclohexanone 99 99 96 97

62‐hexanol 2‐hexanone 98 98 82 83

*as the only product, 100% selectivity towards all tested substrates

Conditions: 2.4 mmol alcohol, 2.4 mmol oxidant PS‐DAIB (loading 3.73 mmol/g), solvent 15 mL acetonitrile, temperature 70 oC Run 1 and 1st recycle Isolated yields are given. Following recycles GC yields were reported due to reduced amount of the oxidant .

The recycling experiments demonstrated that the polymer‐attached iodosobenzene diacetate (PS‐DAIB) could be recycled and reused at least three times. In the subsequent runs, 1‐octanol (entry 2, Table 4.12) and cyclohexanol (entry 5, Table 4.12) showed nearly identical oxidation reactivity compared to the initial activity.

In addition we tested whether reactivation with perbutyric acid – thus oxidation with polystyrene‐attached iodosobenzene dibutyrate (PS‐DBIB) was equally effective (Table 4.13).

Chapter 4

Table 4.13 Regeneration and reuse of the polymeric iodine(III) oxidant PS‐DBIB(C4 derivative)

yield (%) Entry substrate product 1st run 1st recycle

1benzyl alcohol benzaldehyde 96 95

2 p ‐methoxybenzyl alcohol p ‐methoxybenzaldehyde quant. quant.

32‐phenylethanol phenylacetaldehyde 91 71

4 geraniol citral quant. 95

51‐hexanol haxanal 88 78

61‐nonanol nonanal 76 75

71‐phenylethanol acetophenone 90 67

8 cyclohexanol cyclohexanone quant. 58

92‐hexanol 2‐hexanone quant. 71

10 2‐nonanol 2‐nonanone 93 79

Selectivity > 99.9% for all tested substrates.

As presented in Table 4.13 the polymer‐attached iodosobenzene dibutyrate C4 derivative (PS‐ DBIB) could be recycled for an additional run. However, in contrast to PS‐DAIB, the recovery of the polymer was only 50% after the first run. At present is not yet clear what the reasons for lower recovery are. PS‐DBIB tends to agglomerate faster than other oxidants.

4.3 Conclusions A wide variety of alcohols, including heterocyclic and steroidal alcohols, can be successfully oxidized to the corresponding carbonyl compounds in high selectivity using polymer‐attached iodine(III) oxidants. Activated alcohols such as allylic alcohols underwent rapid oxidation with complete conversion in 1 hour or less with all the polymeric iodine(III) reagents. Primary and secondary alcohols afforded aldehydes and ketones, respectively, in quantitative selectivities. Thus iodine(III) can act as a sole oxidant for alcohols, in contrast to many studies reported in literature where bromide or TEMPO is used as cocatalysts in the reaction. 1,2

PSattached iodine(III) reagents in sel. oxidation. Scope of substrates.

The optimum procedure involved the use of 1eq. of polymer‐attached iodine(III) dialkanoate under the following conditions: solvent: acetonitrile, toluene or ethyl acetate, reaction temperature 70 oC, vigorous stirring, reaction time 40 min to5 h (substrate specific).

It was shown that the PS‐DBIB oxidant, with active oxygen loading of 1.7 mmol/g, was the most active oxidant in the oxidation of hydrophobic and benzylic alcohols. Overall all oxidants appeared to be highly active for secondary, acyclic or cyclic alcohols oxidation. The solvent (acetonitrile, ethyl acetate or toluene) had only a minor effect, which can be substrate specific. Efficient agitation and higher temperatures had a positive effect. Regeneration and recycling of the polymeric oxidant was shown to be feasible. The regenerated C2 oxidant was recycled three times without significant loss of activity.

A two‐step mechanism is postulated comprising initial substitution of the acetoxy group by the alcohol followed by elimination of a second molecule of carboxylic acid concomitant with reductive elimination of iodine(III) to iodine(I) and release of the carbonyl product. This was in line with results of competition experiments and initial rate measurements that revealed that primary alcohols react faster than secondary alcohols. The lower rates with secondary alcohols are probably due to increased steric hindrance.

4.4 Experimental General The organic solvents (p. a.) were used as received. Polystyrene was purchased from Sigma– Aldrich as crude product, thus additional purification was required. Alcoholic substrates were purchased as high purity > 99% from Sigma‐Aldrich or Acros. Steroids and functionalized heterocyclic alcohols were donated by Organon (now MSD).

Nuclear Magnetic Resonase (NMR) analysis 1 H NMR 400 Hz or 300 MHz and 13 C NMR 100 Hz or 300 Hz spectra were recorded on a Bruker AC 400 or Varian Inova VXR‐400S spectrometer using TMS as an external standard. Samples were dissolved either in CDCl3, CD2Cl2, D2O,or t‐BuOD as solvent.

Gas chromatography analyses GC were carried out with GC Varian Star 3400 instrument equipped with a polar CP WAX 52 CB 50 m*0.53 mm*2.0 µm column, T max=250 oC. As internal standard for all oxidation reactions n‐ dodecane was used. Column temperature profile: 65 deg (2 min), rate 10 deg/min to 230 deg (9.5 min). Injector temperature profile: 85 deg (2 min), rate 15 deg/min to 250 deg (16 min). GC sample preparation. For experiments performed at 70 oC samples each of 100 µl were taken from reaction mixture after 1 h, 3 h, 5 h, 24 h, 48 h and diluted to 1 mL of diethyl ether. Before injection on GC an additional filtration through a Pasteur pipette filled with cotton was often required to remove any solid particles.

Steroids TLC methods: Plate: Merck Silica F254, detection UV 254 and 366 nm after treatment with 10% H2SO4 in EtOH.

Chapter 4

11α‐hydroxynordion, Eluent methylene chloride/acetone 7:3; Rf values: substrate: 0.44, product: 0.75, Nandrolone, Eluent: toluene/ethyl acetate 1:1 or n‐heptane/acetone 2:1 5‐α‐H‐3,20‐pregnanedione Eluent: toluene/ethyl acetate 2:1

Oxidations Gram scale experiments: Alcohol oxidation 3.2 mmol of alcohol was dissolved in 20 mL of acetonitrile and 1 eq. of PS‐ DAIB added. After reaction was complete the oxidant was filtered off using a Bűchner funnel and the solvent was evaporated. The obtained product was purified via Kugel‐Rohr distillation yielding pure carbonyl product with high yields. Purity of obtained products was analyzed by GC and 1H NMR study. The NMR data are given for the reactions performed on gram scale (end of this paragraph).

GC scale oxidation of alcohols with polymeric trivalent oxidant 0.16 mmol alcohol was dissolved in 1 mL of organic solvent (MeCN or EtOAc, or toluene), 0.05 mmol of n‐dodecane as internal standard was added. All compounds were prior weighted. 1 eq. of PS‐DAIB in respect to its loading was introduced after the first t= 0 min sample was withdrawn, then by heating and intense stirring the reaction began. The progress of the reaction was followed by withdrawing a 100 µl sample from the reaction mixture and dissolving it then in 1 mL of diethyl ether. After the reaction was complete (1 to 8 h) the oxidant was filtered off using a glass pipette filled with cotton. The clear solution was then injected on GC.

Initial rate experiments alcohol : oxidant 1: 0.5 ratio 0.08 mmol of an alcohol was dissolved into 0.5 mL of acetonitrile (MeCN), 0.05 mmol of dodecane as internal standard was added. All compounds were prior weighted. Subsequently oxidant 0.04 mmol of PS‐DAIB in respect to its loading (ld. 3.73 mmol/g) was introduced after the first t=0 h sample was withdrawn, then by heating and intense stirring the reaction began. The progress of the reaction was followed by withdrawing samples each of 100 µl from the reaction mixture and dissolving in 1 mL of diethyl ether. Samples were taken after 5, 10, 20, 30 and 60 min of reaction. Following oxidant was filtrated off via a cotton filled pipette and transparent solution was injected on GC for further analysis.

Competition experiments primary and secondary alcohols 1 equivalent mixture of primary (0.08 mmol) and secondary alcohol (0.08 mmol) were dissolved in 1 mL of acetonitrile or toluene. Subsequently 50.0 mmol internal standard (n‐dodecane) was added. Aside required amount of 0.08 mmol oxidant PS‐DAIB ld. 3.73 mmol/g, powder was weighed. Reaction was started by introducing the oxidant to reaction at temperature of 70 oC under vigorous stirring. The progress of the reaction was followed by withdrawing samples of 100 μl from reaction mixture and dissolving it in 1 mL of diethyl ether. Samples were taken after 1 h, 3 h, 5 h and 24 hours of reaction.

Hammett plot relationship (LFER) The Hammett plot, linear free energy equation was plotted according to the formula:

kX log   kH  (sigma ) the substituent constans  (rho) the reaction constans

PSattached iodine(III) reagents in sel. oxidation. Scope of substrates.

If  (rho) gives (+) values then rate increases with EWG and decreases with EDG The reaction rate was calculated according to the formula:

Csub ][ k  ln t0 Csub  Cprod ])[]([ t0 xht

where C sub (t= 0) –concentration of alcohol at t=0 h C prod (t= xh)‐ concentration of product at t = xh, x=1 h, 3 h....etc k‐ rate of the reaction

Table 4.14 Tested alcohols and their  values

substrate group 

Benzyl alcohol -H 0.00

p -methoxy benzyl alcohol p -OCH3 -12.00

m-methoxy benzyl alcohol m-OCH3 0.115

m-methylbenzyl alcohol m-CH3 -0.070

p -methylbenzyl alcohol p -CH3 -0.140

p -trifluoromethyl benzyl alcohol p -CF3 0.530 m-iodobenzyl alcohol m-I 0.352 m-fluorobenzyl alcohol m-F 0.337 p -fluorobenzyl alcohol p -F 0.150 m-chlorobenzyl alcohol m-Cl 0.370

Regeneration and recycling procedure To 25 mL round bottom two‐neck flask a 2.4 mmol of alcohol and was dissolved in 15 mL of acetonitrile The temperature was raised to 70 oC. To start reaction subsequently 1 eq. of oxidant (C2 derivative PS‐DAIB ld. 3.73 mmol/g) or (C4 derivative PS‐DBIB ld 1.71 mmol/g) was introduced. Vigorous stirring (magnetic bar) was applied. After reaction was completed firstly the oxidant was filtered off using a Bűchner funnel and remaining mixture was analyzed on GC after prior dilution in diethyl ether. Collected oxidant was then undertaken for regeneration procedure. First reduced oxidant was dissolved in dichloromethane and precipitated back into methanol. Amount of dichloromethane was adjusted to the amount of oxidant collected and kept low. The obtained solid was dried in vacuo at room temperature overnight. Sample of dry reduced oxidant was analyzed for IR. The filtrate recovery after reaction was about 58% of initially used oxidant. In the following step collected fine powder was re‐oxidized at 50 oC for 12 h with freshly formed carboxyperacid using procedures reported in chapter 3. The oxidation loading of the obtained oxidant as cream‐yellow powder was then checked and IR was recorded to confirm the presence of desired carbonyl groups.

Products from gram scale oxidation

Benzaldehyde 1 13 H NMR (400 MHz, CDCl3) δ 9.967 (1H, d, J =1.6 Hz, CHO, H‐6); 7.837‐7.435 (5H, m, arom); C NMR (100 MHz; CDCl3) δ 192.33 (CHO, C‐7); 136.41 (C, arom C‐1); 134.41 (CH, arom C‐6), 130.07 (CH, arom C‐2) 129.66 (CH, arom C‐3) 128.98 (CH, arom C‐4); 128.45 (C, arom C‐5).

Chapter 4

Hexanal 1 H NMR (400 MHz, CDCl3) δ 9.762 (H, s, H‐6); 2.425 (2H, m, H‐5); 1.638 (2H, m, H‐4); 1.333 (5H, 13 m ); 0.905 (4H, m); C NMR (100 MHz; CDCl3) δ 202.84 (C, CHO, C‐2); 43.93 (C, aliph, ‐CH2‐CH2‐ CHO); 31.41 (C, aliph, CH2, C‐4); 22.50 (C, aliph, CH2, C‐3); 21.85 (C, aliph, CH2, C‐5); 13.90 (C, aliph, CH3, C‐6).

Octanal 1 13 H NMR (400 MHz, CDCl3) δ 9.70 (1H, CHO); 2.42 (H, m); 1.262 (10H, m); 0.883 (3H, m, H‐8) C NMR (100 MHz; CDCl3) δ 203.06 (C, CHO) 43.95 (C, ‐CH2‐CH2‐CHO); 29.69 (C, ‐CH2‐CH2‐CHO); 28.10 (C, ‐CH2‐); 25.64 (C, ‐CH2‐); 22.15( C, ‐CH2); 14.08 (C, CH3‐CH2).

3‐pyridinecarboxaldehyde 1 H NMR (400 MHz, CDCl3) δ 10.18 (1H, CHO); 9.124 (1H, ‐N‐CH); 8.87 (1H, HC‐N); 8.22 (1H, d, J = 13 7.6, CH‐CH); 7.565 (1H, d, J = 4.4 CH‐CH); C NMR (100 MHz; CDCl3) δ 191.06 (C, CHO); 154.66 (C, HC‐N); 151.87(C, ‐N‐CH); 135.85(C, HC‐C‐); 131.44(C, ‐C‐CHO); 124.13 (C, ‐HC‐CH ). p‐anisylaldehyde 1 H NMR (400 MHz, CDCl3) 9.87 (1H, CHO); 7.835(2H, d, J= 8.8 HC‐C‐CHO ); 7.002 (2H, d, J = 8.4 13 HC‐C‐CH3); 3.87 (3H, CH3) δ C NMR (100 MHz; CDCl3) δ 190.82 (C, CHO); 164.63 (C, C‐CH3 ); 131.97 (2C, ‐HC‐C‐CHO); 129.94 (C, ‐HC‐C‐CHO); 114.33 (C, HC‐C‐O‐CH3); 55.57 (C, CH3).

Steroids–isolated yield

Nordione 1 H NMR (300 MHz, CD2Cl2, ppm) δ 5.815 (1H, s, =CH2CO); 2.349‐2.71 (H, m); 2.115‐1.963 (H, m, CH2); 1.328‐1.29 (H, m ); 0.043 (3H, s, CH3); 13 C NMR (100 MHz; CD2Cl2, ppm) δ 220.099 (C, H2C‐C=O); 199.307(C, ‐CH‐C=O); 166.073 (C, CH2‐ C=CH2); 124.773 (C, ‐C=CH); 54.373‐13.825 (C, CH2); Yield 96%

5‐α‐H‐3,20‐pregnanedione 1 H NMR (300 MHz, CD2Cl2, ppm) δ 0.643 (3H, s, CH3); 1.185 (3H, s, CH3); 2.105 (3H, s, COCH3); 0.820‐1.006 (1H, m ); 1.201‐1.248 (1H, m);1.320‐1.459 (1H, m); 1.616–1.776 (1H, m); 2.018‐ 2.083 (2H, m); 2.124‐2.285 (2H, m), 2.300‐2.418 (1H, m); 2.523‐2.584 (2H, m); 13 C NMR (100 MHz; CD2Cl2, ppm) δ 209.129 (C, CHCOCH3); 211.273 (C, CH2COCH2); 11.390‐ 63.862 (C, CH2, CH); Yield 97%

Synthesis of several ionic liquids

+ ‐ [HOPMIm] [NO3] 1‐(3‐hydroxypropyl)‐3‐methylimidazolium nitrate was prepared in two steps procedure. 10 In the first step 1‐(3‐hydroxypropyl)‐3‐methylimidazolium chloride [HOPMIm]+[Cl]‐ was prepared by reacting N‐methylimidazole with 3‐chloropropanol according to the recipe. 1‐methylimidazole (34.75 g; 0.42 mol) was slowly added to the freshly pre‐distilled 1‐chloro‐3‐ propanol (55.05 g, 0.58 mol) under constant flow of nitrogen atmosphere. An inert gas conditions were required to avoid decomposition of 1‐methylimidazole. Reaction was let for stirring in a three neck round‐bottom flask equips with reflux condenser at 60 oC for 4 days to form ionic liquid. Long reaction time is required to receive nearly quantitative conversion. After completion reaction mixture was extracted (3x 50 mL) with 50 mL of ethyl acetate to remove excess of unreacted 3‐chloropropanol. The upper, organic phase was decantated and the lower, viscous [HOPMIm]+[Cl]‐ ionic liquid phase was taken to rotary evaporator to remove remaining ethyl acetate. Was obtained 72 g of [HOPMIm]+[Cl]‐ giving yield of 97%

PSattached iodine(III) reagents in sel. oxidation. Scope of substrates.

1 H NMR (300 MHz, D2O, t‐BuOD ppm) δ 8.99 s1H (N‐CH=N), 7.66; 7.59 d2H (CH=CH), 4.78 s1H (D2O; OH), 4.37; 4.35; 4.33 t2H (CH2‐CH2‐OH), 3.95 s3H (CH3‐N), 3.65; 3.63; 3.61 t2H (N‐CH2‐CH2‐ CH2‐OH), 2.13; 2.11; 2.10; 2.07; 2.062; 2.06; 2.04 m2H (‐CH2‐CH2‐CH2OH); 13 C NMR (100 MHz, D2O, t‐BuOD ppm) δ 32.42; 35.41; 48.16; 58.36; 123.71; 137.02

In the second step an anionic exchange was done using a 400 mL ion exchange column (capacity approx 0.4mol exchange sites), filled with 250 g Dowex 1x8 200 anion exchange resin. Saturated with Cl‐ resin was rinsed with 1L of Milli‐Q water. A 1M solution of sodium nitrate pH 6.5 adjusted with fum. HNO3 was introduced to the column to replace the chlorine ions with nitrate ‐ NO3 . The presence of chlorine was checked using silver chromate titration subsequently*. Following 0.2 M solution of the [HOPMIm]+[Cl]‐ionic liquid in Milli–Q water was slowly passed through the column where anionic exchange took place. Eluted fraction was collected and controlled for presence of chlorine. Subsequently column was rinsed with additional 1L of Milli‐ + ‐ Q water. Eluted aqueous solution of [HOPMIm] [NO3] was concentrated on a rotavap to remove water. Obtained colourless viscous ionic liquid was extensively dried in vacuo over phosphorous pentoxide for at least 48 hours. Chlorine test indicated that near quantitative exchange was obtained. Thus viscous ionic liquids contained < 30 ppm Cl‐ according to the silver chromate test. 1 H NMR (300 MHz, D2O, t‐BuOH ppm) δ 8.71 s1H (N‐CH=N), 7.48; 7.42 d2H (CH=CH), 4.68 s1H (D2O; OH), 4.30; 4.28; 4.26 t2H (CH2‐CH2‐OH), 3.88 s3H (CH3‐N), 3.62; 3.60; 3.58 t2H (N‐CH2‐CH2‐ CH2‐OH), 2.13; 2.11; 2.10; 2.08; 2.062; 2.06; 2.04 m2H (‐CH2‐CH2‐CH2OH); 13 C NMR (100 MHz, D2O, t‐BuOD ppm) δ 31.12; 37.16; 48.00; 59.45; 123.80; 137.64

*Chlorine test Solutions of silver nitrate AgNO3 (1 mM) and potassium chromate K2CrO4 (5g/L) in milli‐Q water were prepared according to standard titration’s recipe.11 A single drop of the potassium chromate solution was added to a sample (few drops) of eluted from the column solution of ionic liquid. Secondly dropwise titrant (AgNO3) was carefully added. Appearance of a red colour Ag2CrO4 precipitate (positive result, solution free of chloride) upon addition of a first drop of titrant indicates that the concentration of Cl– is essentially less than 600 ppm. An appearance of a green colour indicates formation of AgCl, thus confirming still a high content of Cl‐ anions (negative result).

+ ‐ + ‐ Choline nitrate [HOCH2CH2Me3N] [NO3] = [TMEOA] [NO3] Preparation of cholide nitrate required only an ionic exchange from chloride to nitrate. An aqueous solution 0.2 M of a commercial choline chloride was passed through the exchange column saturated with 1M aqueous solution of NaNO3. The elute fraction was tested for the presence of any chlorine and water was removed on a rotavap. Choline nitrate appear as a white‐crystaline solid at room temperature. Pure highly hydrophilic choline nitrate was dried and stored in vacuo dessicator over phosphorous pentoxide for additional 48 h. 100% exchange took place.

‐ + ‐ Triethanolamine formate [(HOCH2CH2)3NH][HCO2] = [TEA] [HCO2] To a cooled (ice‐bath) 0.2 mol, 26 mL of triethanolamine (Fluka 98.5% GC [102‐71‐6]), 0.2 mol (7.5 mL) of 98‐100% formic acid was added dropwise. Yellowish solid was brought to room temperature; melting point was 60 oC. Solution of ionic liquid was let for stirring at 60 oC for 5 hours. Resulting yellow viscous ionic liquid was dried in an exiccator over P2O5. Obtained hydrophilic ionic liquid has a Mp = of 60 oC.

+ ‐ 12 [BMIm] [NO3] was synthesized and characterized according to the known recipe.

Chapter 4

4.5 References

1Combinations of iodine(III) and Br‐: H. Tohma, S. Takizawa, T. Maegawa, Y. Kita, Angew. Chem. Int. Ed., 2000, 39, 1306; H. Tohma, T. Maeagawa, S. Takizawa, Y. Kita, Adv. Synth. Catal., 2002, 344, 328; W. Qian, E. Jin, W. Bao, Y. Zhang, Angew. Chem. Int. Ed., 2005, 44, 952; For a catalytic version see: R. Mu, Z. Liu, Z. Yang, L. Wu, Z.‐L. Liu, Adv. Synth. Catal., 2005, 347, 1333. 2Combinations of iodine(III) and TEMPO: K. Sakuratani, H. Togo, Synthesis, 2003, 21, Y. Tashino, H. Togo, Synlett., 2004, 2010; T. Y. S. But, Y. Tashino, H. Togo, P.H. Toy, Org. Biomol. Chem., 2005, 3, 970; W. Qian, E. Jin, W. Bao, Y. Zhang, Tetrahedron, 2006, 62, 556; Y. Shang, T. Y. S. But, H. Togo, P. H. Toy, Synlett., 2007, 67; Z. Chenjie, W. Yunyang, J. Lei, Synthetic Commun., 2010, 40, 2057. 3Volumes contributed to hypervalent iodine chemistry: T. Wirth, Top. Curr. Chem., 2003, 224; V. V. Zhdankin, P. J. Stang, Chem. Rev., 2002, 102, 2523; R. M. Moriarty, J. Org. Chem., 2005, 70, 2893; T. Wirth, R. D. Richardson, Angew. Chem. Int. Ed., 2006, 45, 4402; N. Takenaga, A. Goto, M. Yoshimura, H. Fujioka, T. Dohi, Y. Kita, Tetrahedron Lett., 2009, 50, 3227. 4 H. Tohma, T. Maeagawa, S. Takizawa, Y. Kita, Adv. Synth. Catal., 2002, 344, 328. 5Regeneration and reuse: H. Togo, G. Nogami, M. Yokoyama, Synlett., 1998, 534; Ch. A. Briehn, T. Kirschbaum, P. Bäuerle, J. Org. Chem., 2000, 65, 352; X. Huang, Q. Zhu, Synth. Commun., 2001, 31, 111; X. Huang, Q. Zhu, Y. Xu, Synth. Commun., 2001, 31, 2823. 6Ionic liquids as reaction media: For synthesis and catalysis: T. Welton, Chem. Rev., 1999, 99, 2071; For transition metal catalysis: P. Wasserscheid, W. Keim, Angew. Chem. Int. Ed., Engl, 2000, 39, 3772; For catalytic reaction: R. A. Sheldon, Chem. Commun., 2001, 2399; For biocatalysis: F. van Rantwijk, R. A. Sheldon, Trends in Biotechnology, 2003, 21, 131. 7J. S. Yadav, B. V. S. Reddy, A. K. Basak, A. Venkat Narsaiah, Tetrahedron, 2004, 60, 2131. 8Z. Liu, Z‐Ch Chen, Q‐G Zhang, Organic Lett., 2003, 5, 3321; G. Karthikeyan, P. T. Perumal, Synlett., 2003, 14, 2249. 9S. Kanbak‐ Aksu, “Protein‐Metal Hybrids as Catalysts for Selective Oxidations”, TU Delft, PhD thesis, 2010, 78. 10J. Fraga‐Dubreuil, J. P. Bazureau Tetrahedron, 2003, 59, 6125. 11Chloride measurements by Argentometric Method Standard Methods (Eds.), Method 4500‐ Cl‐B, online resources. http://.environmentalet.org/env1221/titrations.htm 12J. S. Wilkes, M. J. Zaworotko, J. Chem. Soc., Chem. Commun., 1992, 965.

CChhaapptteerr 55

Lipase catalyzed, in situ generation of hypervalent iodine reagent for selective alcohol oxidation

5.1 Introduction 5.2 Results and Discussion 5.2.1 Feasibility studies of one‐pot cascade for oxidation of alcohols 5.2.2 In situ formation of C4 and C8 polymer‐attached iodine(III) derivative 5.2.3 Three‐step cascade for secondary alcohols 5.2.4 Catalytic cascade with steroids 5.3 Conclusions 5.4 Experimental data 5.5 References

Abstract This chapter covers studies on in situ generation of an active polymer‐attached iodine(III) oxidant, using hydrogen peroxide in the presence of lipase and a carboxylic acid. The in situ generated iodine(III) oxidant can be used directly for selective oxidation of alcohols. Herein, we highlight the limitations and future directions of the method.

Chapter5

5.1 Introduction The development of chemo‐enzymatic cascade reactions, where both enzymes and organic or metal‐based catalysts are employed is in line with the aims of Green Chemistry. 1 The integration of catalytic steps, rather than two separate reactions, can circumvent intermediate work‐up procedures, and thus save chemicals and prevent waste. Apart from combinations of metal‐based catalysts 2, examples of the combination of enzymes and chemo‐catalysts have been reported. 3‐5

For oxidation methods employed in the fine chemical industry there is a definite need to replace salt‐ and heavy metal waste producing procedures with atom efficient technologies. 6 Cascade‐based oxidation procedures, where oxidants are generated in situ, could contribute to the use of more benign oxidants in this area. Two promising examples involving biocatalysis have been published. One example is a chemo‐enzymatic cascade where the enzyme laccase is used in combination with the stable nitroxyl radical, TEMPO, and air for the oxidation of primary alcohols to aldehydes. 7 In this example in situ formed oxoammonium cation acts as the oxidant. Another example is the epoxidation of alkenes where in situ formed peracid acts as the oxidant. 8 In the latter example the peracid is formed by lipase‐catalyzed esterification of hydrogen peroxide (see chapter 6).

In this chapter, the aim is to further extend the possibilities of iodine(III) as a versatile oxidant by generating it in situ during the reaction. 9 Peracid is the oxidant of choice in this case (peracid + iodine(I)  iodine(III) + acid) starting from iodine(I). The peracid can also be generated in situ from the acid and hydrogen peroxide according to known procedures. In this way, an enzyme/organocatalyst cascade can be envisaged as shown in scheme 5.1.

Scheme 5.1 Enzyme/organocatalyst cascade based on iodine(I) as catalyst for selective alcohol oxidation

The enzyme employed in this procedure is the robust enzyme Candida antarctica lipase B (CaLB) that is known to catalyze reactions in organic solvents. 10 This procedure provides a chloride‐ and bromide‐free method for e.g. the selective oxidation of alcohols, based on hydrogen peroxide as primary oxidant and a carboxylic acid, lipase and iodine(I) as catalysts. As

Lipase catalyzed, in situ generation of hypervalent iodine reagent iodine(I) precursor the iodinated polystyrene copolymer was used. This iodopolymer was previously shown to be a stable and easy to handle iodine carrier for studies under stoichiometric conditions. 11

Our aim was thus to investigate the possibility to generate iodine(III) oxidant in situ, using the

H2O2/lipase/acid procedure for the oxidation of alcohols. In this chapter two oxidant derivatives

C4 and C8 will be applied. The strength of this methodology will be in the oxidation of alcohols that cannot be oxidized by currently available, metal‐based, catalytic oxidation methods employing air or hydrogen peroxide as oxidant, (Pd, Pt, Au, W). 12, 13 These methods are usually unsuitable for sterically demanding and coordinating substrate that block the metal site. Therefore, this iodine(I) cascade using hydrogen peroxide as oxidant will be specifically tested for the oxidation of steroidal alcohols.

5.2 Results and discussion 5.2.1 Feasibility studies of onepot cascade for oxidation of alcohols Preliminary studies

The proposed catalytic cascade procedure, (as shown in Scheme. 5.1) comprises of three individual oxidation steps. First, the perhydrolysis of the carboxylic acid is catalyzed by Novozym 435 (an immobilized form of CaLB). In the second step an in situ oxidation of polystyrene‐ attached iodine precursor by the peracid, to form active polymeric iodine(III), takes place. Third, the selective oxidation of the alcohol by the polystyrene‐attached iodine(III) oxidant leads to the formation of the observed carbonyl products.

In a preliminary study, to a reaction mixture consisting of benzyl alcohol, butanoic acid, Novozym 435 and iodinated linear polystyrene in organic solvent (acetonitrile), 30% aqueous hydrogen peroxide was added stepwise to initiate perhydrolysis of acid. However, benzyl butanoate ester was observed as the only product (60% yield), as shown in Figure 5.1. We conclude that esterification of alcohols catalyzed by CaLB, rather than perhydrolysis of hydrogen peroxide, is the dominant process under these conditions.

Figure 5.1 Undesired reaction in one‐pot cascade

Chapter5

Furthermore, a series of reference experiments was carried out with cyclohexanol as model substrate for secondary alcohols, and the results are listed in Table 5.1.

Table 5. 1 Series of reference experiments for oxidation of cyclohexanol / cyclohexanone in one‐pot cascade.

Entry substrate substrate Butyric acid Ethyl butyrate Nov. 435 30% H202 obtained butyrate ester in 1 mL of MeCN mmol mmol mmol mg mmol mmol

1 cyclohexanone* 0.16 0.32 - - 0.32 * 2 cyclohexanol 0.16 0.32 - - - 0.013 3 cyclohexanol 0.16 0.32 - 10 0.32 0.038 4 cyclohexanol 0.16 - 0.32 10 0.32 0.008 5 cyclohexanol 0.16 - 0.32 10 - 0.063

*3% conversion of cyclohexanone was obtained

The first observation is that cyclohexanone is not stable in the presence of H2O2 (entry 1, Table 5.1). Apparently Baeyer–Villiger oxidation takes place, which could be assisted by perbutyric acid). With cyclohexanol and an excess of butanoic acid a Fischer esterification proceeds (entry 2, Table 5.1). Similarly, esterification competes effectively with perhydrolysis when a lipase is present, (entry 3, Table 5.1). In entry 3, Table 5.1: about 24% of the butyrate ester was formed. With cyclohexanol, ethyl butyrate and lipase a transesterification takes place (entries 4, 5, Table 5.1). A comparison of entry 4 with entry 5 shows that perhydrolysis becomes the main reaction in the presence of hydrogen peroxide.

Thus, reaction of the acid with the alcohol appears to be favoured over perhydrolysis with hydrogen peroxide in the presence of CaLB. As a result low or no peracid, and thus no formation of the polystyrene‐attached iodine(III) take place. Hence, it was obvious that introduction of all reagents at the same time would not lead to formation of iodine(III). Therefore, we turned our attention to a stepwise approach. In the first step, peracid is generated in the absence of both iodine and alcohol. Subsequently, removal of the enzyme and utilization of non‐reacted hydrogen peroxide and subsequent addition of iodine(I) to the peracid mixture will result in iodine(III) formation. In the third step, the introduction of the alcohol will lead to carbonyl compound formation (Figure 5.2).

Lipase catalyzed, in situ generation of hypervalent iodine reagent

Step 1. Synthesis of percarboxylic acid cat. by CaLB

O cat. CaL B O H O + + 2 2 ROH ROOH H2O 5 h, 50 oC

R ‐ CH3, C3H7, C7H15 1. Filtration through basic alumina 2. Introduction of the PhI

Step 2. Oxidation of iodine(I) to iodine(III) O R acid / peracid O PS I PS I 12 - 15 h 50 oC O O R iodine(I) = iodinated linear polystyrene iodine(III) = polymeric oxidant PS ‐ polystyrene 1. Introduction of an alcohol

Step 3. Alcohol oxidation

I(III) oxidant, OH org. solvent O

R' R" (H) 3‐24 h 70 oC R' R" (H)

R', R" ‐ aliphatic, Ph

Figure 5.2 Schematic representation of three step, one‐pot cascade for oxidation of alcohols to aldehydes using iodine(III) as catalyst

In the following section, the results of these individual steps are presented in more depth.

5.2.2 In situ formation of C4 and C8 polystyreneattached iodine(III) derivate Two steps are required in order to accomplish the formation of polymer‐attached iodine(III) reagent (Scheme 5.2). Successful perhydrolysis of acid and hydrogen peroxide allows for oxidation of iodine(I) to iodine(III). In this first step perhydrolysis of the carboxylic acid is catalyzed by CaLB. This is a well documented biotransformation. 14

Our starting point was to use a molar ratio of H2O2: acid 2:1 according to the work of de Zoete 8 o et al. In this case up to 50% of H2O2 could be converted to peracid at 50 C. In our case we used a 30% aqueous solution of hydrogen peroxide instead of 50%. In order to rationalize the

Chapter5 parameters, a series of experiments were performed. The optimum reaction temperature for perhydrolysis and iodine activation step was chosen as 50 oC. This temperature choice is based on previous studies (chapter 3 of this thesis). 8, 15 We also decided to use formation of polymer‐ attached iodine(III) oxidant as a measure of the efficiency of step 1 of the cascade.

The formation of both butyric PS‐DBIB (C4) and caprylic PS‐DCIB (C8) derivatives of the polymer‐ attached iodo oxidant were studied (see Scheme 5.2).

Scheme 5.2. First two steps of the cascade: formation of iodonated(III) polystyrene oxidant

First, acids (butyric or caprylic), hydrogen peroxide and lipase (Novozym 435) were mixed and the formation of peracid was followed by TLC for 5 hours. Second, the lipase was removed by filtration and iodinated polystyrene was added to the reaction mixture to form the desired oxidant (overnight). This step took about 12‐15 h at 50 oC. Disappearance of the peracid was taken as an end‐point of the reaction. The isolated polystyrene‐attached oxidant was analyzed by redox titration and the results are presented in Table 5.2.

Table 5.2 Formation of iodine(III) polymer in a two‐step cascade

Entry set of conditions* BA 30 % H2O2 PS‐Immolyield % efficiency based

mmol mmol mmol oxidant PS‐DBIB PS‐DBIB on H2O2 %

1 1 0.32 0.64 0.16 0.11 71 34 2 2 0.32 0.32 0.16 0.09 59 28 3 3 0.32 0.24 0.16 0.05 29 16 4 4 0.32 0.16 0.16 0.02 12 6

Entry set of conditions* CA 30 % H2O2 PS‐I mmol % efficiency based

mmol mmol mmol oxidant PS‐DCIB PS‐DCIB on H2O2 %

5 1 0.32 0.64 0.16 0.05 33 16 6 2 0.32 0.32 0.16 0.04 27 13

*Parameters which were not changed: 10 mg CaLB and 1 mL of MeCN

Lipase catalyzed, in situ generation of hypervalent iodine reagent

As can be seen in Table 5.2, in all cases oxidation of iodine(I) to iodine(III) occurred. Using 2‐4 equivalents of hydrogen peroxide (entry 1, 2, Table 5.2) with respect to butyric acid, 60‐70% of the iodine(I) could be oxidized to iodine(III). When starting with a ratio of butyric acid to hydrogen peroxide of 2:1 (entry 4, Table 5.2), only minimal formation of the iodine(III) oxidant was observed. On increasing the absolute amount of hydrogen peroxide, a gradual increase of PS‐DBIB was observed. The maximum efficiency of hydrogen peroxide was 34% (entry 1, Table 5.2).

While conducting these experiments ite becam clear that it was necessary to remove the enzyme beads and residual peroxide before adding the iodine polymer. This is mainly due to the fact that the enzyme beads and the polymer tended to form clusters, which could not be separated. In other words, it was not possible to shift the equilibrium of the acid‐hydrogen peroxide mixture by adding iodinated polymer gradually to the reaction. In a separate experiment it was validated that iodinated polystyrene is inert upon treatment with hydrogen peroxide.

The presence of the carbonyl group (‐C=O at 1730 cm‐1), indicates the formation of iodine(III) oxidant. Several cascade samples of 3 mg each (1 h, 5 h, 24 h) were isolated and analyzed using IR spectroscopy.

Chapter5

80.6

75 1894.87 629.94 3051.09 1645.00 3434.40 1561.54 1353.13 781.90 70 3014.50 1584.85 1304.17 1102.68 752.27

700.03 65 1183.21 715.49 2848.43 1729.35 538.54 60 1448.37 1062.36

55 %T 1403.87 50 2920.35

40 816.56

35 1481.98 1004.90

28.3 4000.0 3000 2000 1500 1000 450.0 cm-1 1h C=O 1729 cm-1

74.4

70 2586.07 630.03 65 1894.90 3432.03 1561.76 3051.02 1304.04 60 1645.96 960.49 1353.27 941.88 782.04 55 3014.49 1584.92 1103.31 755.28 700.38

50 2848.12 1146.40 1183.60 715.34 45 537.92

%T 1062.41 40 1448.39

35 1728.98 1403.84

30 2920.65

20 1004.89 816.40 15 1482.02

10.9 4000.0 3000 2000 1500 1000 450.0 cm-1 5h

69.5

65 1896.57 629.98 60

50 1603.65 964.13 1585.63 3313.85 45 1537.42 760.25 539.07 714.14 40 2850.65 1385.85 702.25

%T 35 1657.26 1062.94 1404.30 30 1270.26 25

1449.73 817.34 20 1194.22 15 2926.76 1145.59 1482.40 1005.06 10 1730.24

4.4 4000.0 3000 2000 1500 1000 450.0 cm-1 24 hours

Figure 5.3 The IR spectroscopic studies of iodine(III) oxidant PS‐DBIB (polymer‐attached iodosobenzene dibutanoate) formation during the cascade procedure

As presented in Figure 5.3, we were able to monitor oxidation of iodinated polystyrene (PS‐I) to the desired polymer‐attached iodosobenzene dibutanoate (PS‐DBIB) oxidant in time. Obtained

Lipase catalyzed, in situ generation of hypervalent iodine reagent

IR spectra clearly show that after 5 hours the formation of the polymeric oxidant is still ongoing and therefore longer reaction times are required. With octanoic acid the same conditions (entry 1 and 2, Table 5.2) were used. In this case, the yield of polymer‐attached iodine(III) oxidant reached only 27‐33%. This is most likely due to the more bulky ester‐residues located on the iodine (see also chapter 3 of this thesis). Furthermore, the use of octanoic acid is favoured over butyric acid because of the absence of irritating odour.

Summarizing, in general two and preferably four equivalents of hydrogen peroxide with respect to iodine were sufficient to generate polymer loadings suitable for oxidation. These observations provide us with the proof of principle for the first two steps of the cascade. In the next section, the subsequent oxidation of alcohol will be probed.

5.2.3 Threestep cascade for secondary alcohols In a three‐step cascade (Figure 5.2) the following alcohols were used as model substrates: cyclohexanol, 3‐nonanol, and 1‐phenylethanol. Results are given in Table 5.3. After reacting cyclohexanol for 5 h at 50 oC, around 15% of cyclohexanone was formed. Using 4 instead of 2 equivalents of hydrogen peroxide the yield of ketone increased only to 19%. With 3‐nonanol and 1‐phenylethanol, respectively, 11% and 10% ketone was formed after 5 h at 50 °C (entries 3, 5, Table 5.3). Therefore, we increased the temperature of the third step (alcohol oxidation) to 70 oC (based on studies described in chapters 3 and 4). It is worth mentioning that preliminary experiments revealed that at 70 oC neither perhydrolysis nor oxidant formation (cascade steps 1 and 2) were observed. At 70 oC percarboxylic acid formation is modest and the stability of formed peracid stability is insufficent. The in situ pre‐formation of the oxidant with lipase and acid was still carried out at 50 oC (for results see Table 5.3).

Chapter5

Table 5.3 Three‐step cascade for oxidation of secondary alcohols

entry substrate condition set eq. ratio GC yield ketone [%]

OH/Acid/H2O2/PS‐I Cascade: all steps at 50 o Creaction time* 1 h3 h5 h

1 cyclohexanol set A 1/2/2/1 3.0 10 15 2set B 1/2/4/1 6.0 13 19

33‐ nonanol set A 1/2/2/1 2.7 9.0 11.4 4set B 1/2/4/1 7.0 11.8 14

51‐phenylethanol set A 1/2/2/1 4.4 8.0 10 6set B 1/2/4/1 9.0 16 25

Cascade: last (3rd step) at 70 oCreaction time* 1 h3 h67 h

7 cyclohexanol set B: 1/2/4/1 69 76 80

Conditions: Set A: alcohol 0.16 mmol, butyric acid 0.32 mmol, 30% H2O2 0.32 mmol, PS‐I 0.16 mmol, lipase 10 mg, 1mL MeCN Set B: alcohol 0.16 mmol, butyric acid 0.32 mmol, 30% H2O2 0.64 mmol, PS‐I 0.16 mmol, lipase 10 mg, 1mL MeCN *Note that, the reaction time given for the third cascade step (from the addition of alcohol).

Using the optimized procedure from section 5.2.2 (2 eq. butyric acid and 4 eq. H2O2) 3 h after the addition of cyclohexanol we could reach 76% conversion. The actual amount of ketone formed is in agreement with the amount of iodine(III) that was formed under these conditions as described in the previous section (0.12 mmol ketone formed in entry 7, Table 5.3 corresponds to 0.11 l mmo of PS‐DBIB in entry 1, Table 5.2). In other words, the amount of alcohol oxidized at 70 °C under these conditions, can be taken as a measure of the amount of iodine oxidized.

In conclusion we have shown that the oxidation of cyclohexanol can be carried out using a three step cascade reaction applying hydrogen peroxide as the terminal oxidant, and Novozym 435 and iodinated polystyrene as catalysts. The latter can be conveniently recovered after the reaction. The limitations of the one‐pot cascade procedure were also investigated.

5.2.4 Catalytic cascade with steroids Previously chapter 4 shows that polymer ‐attached iodine(III) is an excellent oxidant for the selective oxidation of steroidal alcohols to the corresponding ketones. For this reason we

Lipase catalyzed, in situ generation of hypervalent iodine reagent decided to apply the catalytic cascade for the oxidation of more complex steroidal substrates. Three model substrates were chosen to be tested in the optimized three‐step‐cascade procedure (Figure 5.4). Earlier, in chapter 4 we demonstrated that with 3α‐hydroxy‐5α‐H‐ pregnanolone and nandrolone, it was possible to selectively obtain carbonyl products on a gram scale using iodine(III) as oxidant.

(A) O OH O OH H H H H H

H H H H H H OH O O

3α-hydroxy-5α-H-pregnanolone nandrolone 11α-hydroxynordione

(B) O O O O H H H H H

H H H H H H O O O

5α-H-3,20-pregnanedione nordione 11-ketonordione

Figure 5.4 Steroidal substrates (A) and corresponding oxidation products (B)

Steroid oxidation in the three‐step cascade was performed using optimized conditions based on previously derived cyclohexanol oxidation procedure. Alcohol (0.16 mmol), butyric acid (0.32 mmol), Novozym 435 (10 mg), PS‐I (0.16 mmol), and 30% aq. hydrogen peroxide (0.32 or 0.64 mmol) in 1 mL of acetonitrile were introduced in a three‐step. Results are presented in Table 5.4. The final oxidation step was carried out for 5 h at 70 oC.

Chapter5

Table 5.4 Cascade oxidation of steroidal alcohols

condition set eq. ratio substrate product GC yield

OH/Acid/ H2O2/ PS‐I%

set A 1/2/2/1 3α-hydroxy-5α-H-pregnanolone 5α-H-3,20-pregnanedione 52 set B 1/2/4/1 3α-hydroxy-5α-H-pregnanolone 5α-H-3,20-pregnanedione 67

set A 1/2/2/1 nandrolone nordione 56 set B 1/2/4/1 nandrolone nordione 69

set A 1/2/2/1 11α-hydroxynordione 11-ketonordione 39 set B 1/2/4/1 11α-hydroxynordione 11-ketonordione 61

Conditions: Set A: alcohol 0.16 mmol, butyric acid 0.32 mmol, PS‐I 0.16 mmol, 30% H2O2 0.32 mmol, Lipase 10 mg, 1 mL MeCN Set B: alcohol 0.16 mmol, butyric acid 0.32 mmol, PS‐I 0.16 mmol, 30% H2O2 0.64 mmol, Lipase 10 mg, 1 mL MeCN Steps 1 and 2 were performed at 50 oC and 3rd step at 70 oC (filtration step: enzyme and hydrogen peroxide utilizations)

The results are in agreement with previous data, namely around 70% maximum yield of iodine(III) oxidant could be generated under these conditions, which corresponds to the final amount of ketone as the only product.

In order to turn this procedure into a synthetic 100% yield method further tuning of the reaction parameters and increase of polymer loading is necessary. Increasing the mixing speed and therefore improving the mass transfer might result in slightly higher yields. However, instead of further optimization of the three‐step cascade the original concept of the one‐pot reaction was tested. This concept could overcome the limitations of the present system: i.e. the unfavorable equilibrium composition of the H2O2/acid/peracid mixture which necessitates high amounts of hydrogen peroxide. We decided to investigate the influence of esterification for steroids separately. It is reasonable to assume that steroids do not act as substrates in the lipase‐catalyzed esterification reaction of acid, due to their steric hindranes. 3α‐hydroxy‐5α‐H‐ pregnanolone (0.16 mmol), which is not soluble in acetonitrile at ambient temperature or at 50 oC, was added to the reaction mixture, consisting of butyric acid (0.32 mmol), and Novozym 435 (10 mg). The reaction mixture was stirred for 5 hours at 50 oC in the absence of hydrogen peroxide and iodinated polystyrene. No esterification occurred and substrate recovery was nearly quantitative, indicating that steroids do not serve as substrate in the lipase catalyzed esterification.

100

Lipase catalyzed, in situ generation of hypervalent iodine reagent

Therefore the three‐step cascade was carried out without an intermediate filtration step, in which removal of the enzyme and unreacted hydrogen peroxide took place. The same two sets of conditions were applied. Results are listed in Table 5.5. The formation of ketone under one‐ pot cascade conditions shows the proof of concept.

Table 5.5 One‐pot, three step cascade process for steroidal alcohol oxidation

condition set eq. ratio substrate product GC yield

OH/acid/ H2O2/ PS‐I%

set A 1/2/2/1 3α-hydroxy-5α-H-pregnanolone 5α-H-3,20-pregnanedione 31 set B 1/2/4/1 3α-hydroxy-5α-H-pregnanolone 5α-H-3,20-pregnanedione 39

set A 1/2/2/1 nandrolone nordione 38 set B 1/2/4/1 nandrolone nordione 43

set A 1/2/2/1 11α-hydroxynordione 11-ketonordione 22 set B 1/2/4/1 11α-hydroxynordione 11-ketonordione 27

Conditions: Set A: alcohol 0.16 mmol, butanoic acid 0.32 mmol, PS‐I 0.16 mmol, 30% H2O2 0.32 mmol, Lipase 10 mg, 1 mL MeCN Set B: alcohol 0.16 mmol, butanoic acid 0.32 mmol, PS‐I 0.16 mmol, 30% H2O2 0.64 mmol, Lipase 10 mg, 1 mL MeCN Steps 1 and 2 at 50 oC and 3rd step at 70 oC (no filtration step).

Overall 39% of the 3α‐hydroxy‐5α‐H‐pregnanolone was converted to 5α‐H‐3,20‐pregnanedione after 5 hours in the last step. With nandrolone as substrate a slightly better yield was obtained (43%), whereas 11α‐hydroxynordione gave only 27% of keto product. Longer reaction times were not evaluated, because previous results for cyclohexanol indicated that the reaction took place in the first few hours. In general, the use of more hydrogen peroxide led to better yields.

The main difference between Table 5.4 (with the filtration of enzyme beads) and Table 5.5 is that enzyme beads and unreacted hydrogen peroxide were present throughout the cascade. We confirmed that chemical incompatibility was not a problem (cascade examples displayed in Table 5.5). We presume that the second step of the cascade‐oxidation of iodinated polystyrene beads is slowed down by the presence of the enzyme. We suggest that deposition of the enzyme on the polymer‐attached iodine(III) limits the accessibility of the latter. We found that overall 40% less ketone was formed under these conditions.

101

Chapter5

5.3 Conclusions The in situ generation of the polymeric iodine(III) oxidant, by reaction with percarboxylic acid formed in situ by lipase catalyzed perhydrolysis of the corresponding carboxylic acid, proved to work under certain conditions. For simple alcohols such as cyclohexanol, in the one‐pot cascade, esterification rather than the desired oxidation to carbonyl compounds took place. The three‐step procedure suppresses the undesired side reactions. In addition, the step‐wise reactant additions eliminate mass transfer limitations by physically separating the enzyme from the polymer‐attached iodine catalyst. Furthermore, two different temperature regimes had to be applied: 50 oC for the peroxidation and I(III) formation and 70 o C for the alcohol oxidation step. The hydrogen peroxide concentration with respect to acid should be in the range of 2‐4 eq. We found that the bulky steroidal substrates, due to steric hindrance did not undergo CaLB catalyzed esterification under the applied conditions. This meant that, in contrast to simple alcohols, steroidal alcohols could be oxidized to the corresponding carbonyl compounds using a simplified, one‐pot cascade process. In practice such a cascade process could be performed as a “tea bag” batch or as a continuous process.

5.4 Experimental data General The organic solvents (p.a.) were used as received. Polystyrene was purchased from Sigma– Aldrich as crude product. Additional purification of polystyrene was carried out. Alcoholic substrates, ketones acids, internal standard n‐dodecane (99%) with high purity > 99% were purchased from Sigma‐Aldrich or Acros. Steroids were donated by MSD (formerly Organon, currently MSD). The enzyme Novozym 435 (CaLB adsorbed on an acrylic carrier) was kindly provided by Novozymes (Bagsvaerd, Denmark);

Synthesis Cascade procedure in three‐step Condition Set A To a GC vial, an acid (0.32 mmol) (caprylic or butyric) and 1 mL of acetonitrile were added. 10 mg of Novozym 435 was added and the mixture was brought to the temperature of 5O oC under vigorous stirring. Next, 30% aqueous hydrogen peroxide (0.32 mmol) was introduced in five portions at intervals of 1 h. After 5 h the reaction was stopped, and the enzyme was filtered off (through a pipette filled with basic alumina and cotton) and iodinated polystyrene (0.16 mmol, 0.0368 g) was added. Reaction was carried out overnight at 50 oC. Absence of peracid was taken as an end‐point of the iodine(III) oxidant formation. In the last step the alcohol (0.16 mmol) was added to the reaction vial. The temperature was raised to 70 oC. Alcohol oxidation was allowed to run for an additional 5 h. The final mixture was separated from polymeric oxidant and extracted according to the work up procedure.

Condition Set B The amount of peroxide was increased to 4 eq. instead of 2 eq. (condition Set A), while other parameters remain the same. Note that each reaction was run in duplo or triple. Produced oxidant was analyzed for the loading and IR (carbonyl bands) as a separate set of experiments.

102

Lipase catalyzed, in situ generation of hypervalent iodine reagent

Cascade procedure in one step All reactants: alcohol 0.16 mmol, acid 0.32 mmol, lipase 10 mg, iodinated polystyrene 0.16 mmol and 1 mL of acetonitrile were introduced in one step. 0.64 mmol of hydrogen peroxide was added at time intervals of one hour.

Additional pre‐treatment of crude linear polystyrene Crude polystyrene pellets were dissolved in methylene chloride and slowly precipitated by mixing with methanol (volume ratio 1:10. Precipitation of a white solid was observed. The solid was then obtained by filtration through a Buchner funnel. (The pure polystyrene was dried in a vacuo for several days at RT). Iodination of Polystyrene Iodinated polystyrene was prepared according to the procedure described in chapter 3. Iodination degree: 98.45%

Analyses Gas chromatography analysis Analyses were carried out on a GC Varian Star 3400 instrument equipped with a polar CP WAX 52 CB 50 m*0.53 mm*2.0 µm column, T max = 250 oC As internal standard for all oxidation reactions n‐dodecane was used. Column temperature profile: 65 oC (2 min), rate 10 o C/ min to 230 oC (9.5 min), total time 30 min Injector temperature profile: 85 oC (2 min), rate 15 o C/ min to 250 oC (16 min) For steroids Analyses were carried out on a GC Varian Star 3400 or a 3600 instrument equipped with a non‐ polar CP Sil‐5 CB 50 m*0.53 mm column df=2.0 µm , T max = 300 oC Column temperature profile: 65 oC (1 min), rate 15 o C/ min to 280 oC (5 min) Injector temperature profile: 85 oC (1 min), rate 20 o C/ min to 290 oC (10 min)

GC sample preparation For experiments performed at 70 oC samples each of 100 µl were taken from the reaction mixture and diluted with 1 mL of diethyl ether. Before injection on GC an additional filtration through basic alumina was required. In this way hydrogen peroxide was removed. Prior to analysis MgSO4 was added to remove any water.

TLC Steroids TLC methods: Plate: Merck Silica F254, detection UV 254 and 366 nm after treatment with 10% H2SO4 in EtOH. 11α‐hydroxynordion Eluent methylene chloride/acetone 7 : 3,% Rf values: substrate: 0.44, product: 0.75, Nandrolone Eluent: toluene/ethyl acetate 1:1 or n‐heptane/acetone 2:1 5‐alpha‐H‐3,20‐pregnanedione Eluent: toluene/ethyl acetate 2:1

Titration Iodometric titration In a 30 mL colourless glass jar, covered with aluminium foil and stopper, approximately 0.1 mmol of dry sample was weighed. Then 15 mL of an acetic acid/chloroform 2:1 solution and 1.5 mL of 40% KI (60 g in 100 mL of water) solution were added. The solution was stirred at RT for 20 min. in the dark. The blank became light‐yellow and the sample became dark brown. The formed iodine was immediately titrated with 0.1 M solution of sodium thiosulfate Na2S2O3 x 7

103

Chapter5

H2O. As an end‐point of titration discoloration from purple to transparent was taken. Efficient and stable stirring during titration is highly recommended.

104

Lipase catalyzed, in situ generation of hypervalent iodine reagent

5.5 References

1R. A. Sheldon, I. W. C. E. Arends, U. Hanefeld, Green Chemistry and Catalysis., (Eds.), Willey‐ VCH, Weinheim, 2007. 2A. Dijksman, I. W. C. E. Arends, R. A. Sheldon, Chem. Commun., 2000, 271. 3M. C. de Zoete, F. van Rantwijk, R. A. Sheldon, Catal. Today, 1994, 22, 563. 4I. W. C. E. Arends, Y – X. Li, R. Ausan, R. A. Sheldon, Tetrahedron, 2006, 62, 6659. 5R. M. Lau, M. J. Sorgedrager, G. Carrea, F. van Rantwijk, F. Secundo, R. A. Sheldon, Green Chem., 2004, 6, 483. 6I. W. C. E. Arends, R. A. Sheldon, Modern Oxidation Methods., J. Bäckvall, (Eds.), Wiley‐VCH, Weinheim, 2004. 7M. Fabbrini, C. Galli, P. Gentili, D. Macchitella, Tetrahedron Lett., 2001, 7551. 8M. de Zoete “Lipase and Protease‐Catalyzed Transformation with Unnatural Acyl Acceptors”, TU Delft, PhD thesis, 1995, M. C. de Zoete, F. van Rantwijk., R. A. Sheldon, Catal. Today, 1994, 22, 563. 9P. J. Stang and V. V. Zhdankin, Chem. Rev., 1996, 96, 1123. 10O. Kirk, and M.W. Christensen, Organic Process Research & Development, 2002, 6, 446. 11H. Togo, G. Nogami, M. Yokoyama, Synlett., 1998, 534. 12G.‐J. ten Brink, I. W. C. E. Arends, R. A. Sheldon, Science, 2000, 287,1636. 13Y. Ishii, K. Yamawaki, T. Ura, H. Yamada, T. Yoshida, M. Ogawa, J. Org. Chem., 1988, 53, 3587. 14F. Bjorkling, H. Frykman, S. E. Godtfredsen, O. Kirk, Tetrahedron, 1992, 48, 4587. 15Chapter 3, this thesis.

105

CChhaapptteerr 66

Epoxidation and BaeyerVilliger oxidation using hydrogen peroxide and lipase dissolved in ionic liquids

6.1 Introduction 6.1.1 Biotransformation in ionic liquids 6.1.2 Epoxidation and Baeyer‐Villiger oxidation 6.2 Results and Discussion 6.2.1 Alkene epoxidation with enzymatically generated peroxy acids in ionic liquids 6.2.2 Ketone Baeyer‐Villiger oxidation in ionic liquids 6.2.3 Lipases solubility, stability and catalytic activity in ionic liquids 6.3 Conclusions 6.4 Experimental data 6.5 References

Abstract Candida antarctica lipase B dissolves in hydrogen bond donating ionic liquids, thereby enabling homogeneous biocatalysis under semi‐anhydrous conditions. Using this starting point, epoxidation and Baeyer‐Villiger oxidation of olefins and (cyclic) ketones, respectively, was performed in ionic liquids, using a lipase‐catalyzed cascade and hydrogen peroxide as terminal oxidant. Excellent conversions and selectivities were obtained using this biocatalytic system, which opens the possibility for continuous operation.

Chapter 6

6.2 Introduction Although ionic liquids (ILs) were discovered at the beginning of the 20th century, their potential as reaction media and extraction solvents remained unexplored for a long time. 1 Ionic liquids possess advantageous physical properties compared to commonly used organic solvents, such as negligible vapour pressure (although distillation might be possible), low flammability, good recyclability, high thermal stability, and excellent of solubilizing properties. 2‐5 Their tunable properties, such as polarity, miscibility, hydrophilicity or hydrophobicity have earned ionic liquids the distinction of designer solvents. 6‐11

Most importantly, the use of ionic liquids can result in much greener processes, because recyclability and optimized physical properties can lead to smart process design. 12‐17 For instance, ionic liquids can be used as reaction solvent in combination with extraction with

18‐22 supercritical carbon dioxide (scCO2) to improve separation.

The so‐called miscibility switch, in which the IL and scCO2 form one homogeneous phase under the reaction conditions and divide into two easily separable phases on lowering the pressure and/or temperature, has also been reported. 23 From a toxicity point of view, however, the term “green solvents” has to be handled with care. 24‐27

Three generations of ionic liquids can nowadays be distinguished. Ionic liquids based on pyridinium, pyrolidinium, or ammonium cations are recognized as the first generation (Figure 6.1). The second generation is formed by phosphonium, imidazolium, and guanidinium cations in combination with specific either strongly or weakly coordinating anions (Figure 6.1). The third generation of ionic liquid refers to the task specific ionic liquids. 28 In addition ionic liquids that originate from renewable resources are under development.29 Anions play a crucial role in the

‐ ‐ physical properties of ionic liquids. Both weakly coordinating anions (BF4 , PF6 ) as well as strongly coordinating anions such as nitrate are used.

108

Epoxidation and BaeyerVilliger oxidation using H2O2 and lipase

Cations:

R1 R2 R N + 1 + R + R4 6 C R N P 3 R2 N N R3 R R 1 5 R4

Pyridinium Phosphonium Guanidinium

R1 R + 4 NN+ + N R R N R2 1 2 R3 R1 R2 Ammonium Imidazolium Pyrrolidinium

Anions:

- NO3 2------SO BF PF Tf N Cl AcO 4 CF3COO 4 6 2 Strongly Weakly coordinating coordinating

Cl- chloride AcO- acetate

- - NO3 nitrate BF4 tetrafluoroborate

2- - SO4 sulfate PF6 hexafluorophosphate

- - - CF3CO2 trifluoroacetate Tf2N bis(trifluoromethanesulfonyl)imide N(CF3SO2)2

Figure 6.1 Commonly used combination of 1st and 2nd generation’s ionic liquids

6.1.1 Biotransformations in ionic liquids The pioneering work of Klibanov and co‐workers, describing the use of hydrolytic enzymes in anhydrous organic solvents, opened the possibility to perform biotransformations under non‐ natural conditions. 30‐32. In 2000 ionic liquids were introduced into the enzyme catalysis field. 33,34 Lipases are robust enzymes that exhibit excellent tolerance towards organic solvents, while maintaining their hydrolytic and synthetic activity. 35,36 Several reactions involving acyl group transfer by lipases have been studied in ILs 37 such as alcoholysis, 38,39 ammoniolysis, 36 perhydrolysis in non‐aqueous media 40‐47 and perhydrolysis in ionic liquids.36,38,48 The number of conducted studies and reviews where lipases are employed in biocatalytic transformations in ionic liquids is still growing. 49,50

Apart from the activity, the solubility of lipase in ionic liquids has also been studied and shown to be largely influenced by the anion.51 It was found that dissolution of free Candida antarctica lipase B (CaLB) occurs in imidazolium based ionic liquids containing alkylsulphate, nitrate or lactate anions. Although dissolution is accompanied by activity loss, for these cases between 40

109

Chapter 6

and 75% of the original activity could be recovered upon rehydration. The topic of lipase stability in ionic liquids is a subject of ongoing studies. 52‐54

6.1.2 Epoxidation and BaeyerVilliger oxidation Epoxides and lactones are valuable intermediates both in the polymer and fine chemical industries. Classically, synthesis is performed by peracid oxidation of the starting olefins and ketones (Scheme 6.1). The pioneering reactions in this field, named after their inventors, are the Prileshaev epoxidation, and Baeyer‐Villiger (BV) reaction. For the latter case the oxidation of menthone using Caro’s acid (peroxymonosulfuric acid) was originally described. 55, 56

Scheme 6.1 Prileshaev epoxidation and Baeyer‐Villiger oxidation using organic peracids

Stoichiometric use of peroxycarboxylic acids, generally carried out in chlorinated hydrocarbon solvents is nowadays unattractive from a safety and environmental point of view. Many catalytic epoxidation methods based on hydrogen peroxide or oxygen as oxidants have been reported. These commonly involve transition metals catalyst such as Mo, Mn, Ru, Ti, Cr, ,Fe Zr, and W. 57 For BV‐oxidation, catalytic procedures employing hydrogen peroxide in the presence of Lewis acids or dioxygen in combination with transition metal complexes such as: Pt, Pd, Mo or Re, have been developed. Also Se(IV) complexes are known to catalyze Baeyer‐Villiger oxidation as well as epoxidation using hydrogen peroxide as the oxidant. 58 However, the versatility and wide applicability of the peracid approach is unequalled by the metal catalyzed methods, where functional group tolerance is often a problem.

In our approach to extend the full potential of the peracid approach, we have studied biocatalytic methods to generate peracids in situ. In this way the major disadvantages, namely handling of large amounts of peracid, and the non‐catalytic use of acid, can be avoided. It overcomes the need for peracid transportation and makes the process safer to operate. Lipases are known to catalyze the reaction of carboxylic acids with hydrogen peroxide to afford the corresponding peracids. 59 In previous examples in the literature this method was conducted in an organic solvent. In general, Novozym 435 was the lipase of choice, and in situ formed peroctanoic acid was applied for the oxidation. In this way, epoxidation of 1‐octene can be performed in acetonitrile or toluene as the organic solvent. 60,61 This method has also been

110

Epoxidation and BaeyerVilliger oxidation using H2O2 and lipase

applied to fatty acid esters, without adding extra solvent or acid. 62 The full cascade is shown in Scheme 6.2.

O O O R2 HOOH R O OH R1 R2 1 or Nov 435 O O H O 2 OOH R R 1 2 R1 R2

step 1 step 2a step 2b Scheme 6.2.Chemo‐enzymatic cascade for oxidation with peracid

To improve this technology there is a need for reaction media that allow epoxidation to occur with high selectivity, while maintaining high performance of the enzyme. In order to control the cost, the enzyme would have to be recycled and reused many times. Notably, in contrast to organic solvents enzymes can be stored as suspensions in ILs for longer periods of time. 51 It has

+ ‐ already been shown that the lipase cascade with hydrogen peroxide works in [BMIm] [BF4] as ionic liquid. 36 However, in this case the system is heterogeneous in nature and recycling and ease of operation is not straightforward. We therefore embarked on a study to extend the lipase cascade in homogeneous lipase‐ILs reaction solutions, based on hydrogen‐bonding ionic liquids.

Apart from epoxidation, the lipase in situ method could also be useful for Baeyer‐Villiger oxidation. The lipase‐cascade has been investigated for Baeyer‐Villiger‐oxidations in organic solvents such as ethyl acetate and toluene. 63‐65 The use of ILs has been studied for

chemocatalytic BV‐oxidations with H2O2, and was shown to be beneficial in this case, e.g. for homogeneous Pt(II) complexes the stability of catalyst improved in hydrophobic ionic liquid‐

66 67 water systems. Also heterogeneous catalysts such as methyltrioxorhenium CH3ReO3 , or Sn‐ 68 + ‐ β‐zeolite were active in hydrophilic [BMIm] [BF4] for BV‐oxidation. Therefore, we decided to study the chemoenzymatic reaction as well in ionic liquids.

Our aim was to study the lipase‐driven cascade for BV‐oxidation in a selected number of hydrogen bond donating ionic liquids that are expected to dissolve lipase. The effects of the ionic liquids on the epoxidation with the in situ generated peracid (step 1, Scheme 6.2) as well as on the activity and stability of lipase in the perhydrolysis of hydrogen peroxide (step 2, Scheme 6.2) were studied.

111

Chapter 6

6.2 Results and discussion 6.2.1 Alkene epoxidation with enzymatically generated peroxy acids in ionic liquids The aim of lipase‐driven cascade reactions is to generate in situ peroxyacids. The enzyme catalyzes perhydrolysis of acid to peracid with hydrogen peroxide, and in a second reaction the peracid oxidizes the olefin to the epoxide. Previously the viability of this chemo‐enzymatic

+ ‐ 36 epoxidation cascade in ionic liquids was shown using [BMIm] [BF4] . In our approach, we compared the use of first generation ionic liquids, with that of hydrogen bond donating and

+ ‐ + ‐ + ‐ strongly coordinating ILs, namely [HOPMIm] [NO3] , [TEOA] [NO3] , and [TMEOA] [NO3] that are expected to dissolve the lipase (Figure 6.2). 51 Nitrate anions were chosen because of their stability under oxidative conditions. Hydrogen‐bond donating cations were chosen in order to stabilize and dissolve the lipase. The previously reported 1‐(3‐hydroxypropyl)‐3‐methyl imidazolium cation has been shown to be a versatile and stable cation for various applications. 69, 70 In addition, the very accessible and easy to prepare tetra‐ and trialkylammonium ions were used, namely choline and triethanolammonium. The dissolution of lipase in these solvents was confirmed in separate studies, using hydrolysis of triacetin as the assay.

+ - + - N N OH NO3 N N X = NO3 (2a)

- BF4 (2b) + - + - [HOPMIm] [NO3] (1) [BMIm] [X] (2) - PF6 (2c)

OH OH

+ + N N OH OH H

+ - + - [TMEOA] [NO3] (3) [TEOA] [NO3] (4)

+ ‐ [HOPMIm] [NO3] 1‐(3‐hydroxypropyl‐3‐methylimidazolium nitrate (1) + ‐ [BMIm] [NO3] 1‐butyl‐3‐methyl imidazalium nitrate (2a) + ‐ [BMIm] [BF4] 1‐butyl‐3‐methyl imidazolium tetrafluoroborate (2b) + ‐ [BMIm] [PF6] 1‐butyl‐3‐methyl imidazolium hexafluorophosphate (2c) + ‐ [TMEOA] [NO3] trimethylethanolammonium nitrate = choline nitrate (3) + - [TEOA] [NO3] triethanolammonium nitrate (4) Figure 6.2 Structures, names and abbreviations of ionic liquids used in this study

Initially we investigated the epoxidation of three reference olefins, cyclohexene, cyclooctene and styrene, at room temperature in the presence of Novozym 435 (carrier‐adsorbed CaLB). Conditions were chosen according to the work of de Zoete et al. 60 Peroctanoic acid was chosen as an oxidant since lipophilic linear carboxylic acids are known to be the best substrates for lipase. At room temperature a reaction time of 24 hours was required to convert all hydrogen

112

Epoxidation and BaeyerVilliger oxidation using H2O2 and lipase

peroxide. Blank reactions (no peroxide, no lipase) were performed and revealed that in the absence of these reactants no oxidation occurred. Results are shown in Figure 6.3.

100 %

80 epoxide 60 of

40 yield

20 [HOPmim][NO3] 1 [Bmim][BF4] 2b 0 [Bmim][PF6] 2c

100 % 80 yield

60 oxide

20 cyclohexene

0 [Bmim][BF4] [HOPmim][NO3] [choline][NO3] [TEOA][NO3] 2b 1 3 4

Conditions: cyclohexene 1.48 mmol, 50% aq. hydrogen peroxide 2.6 mmol (5x added with intervals of 1 h), caprylic acid 0.2 mmol, enzyme CaLB (Novozym 435, 10 mg) ionic liquid 1 mL, 50 oC, 5 h.

Figure 6.3 Lipase‐driven epoxidation in ionic liquids at 50 oC

For cyclohexene and cyclooctene, clear differences were observed upon changing the ionic liquid from the non‐coordinating imidazolium‐based ILs type solvents, to the coordinating and

+ ‐ hydrogen bond donating [HOPMIm] [NO3] 1 solvent. With cyclohexene the yield of epoxide + ‐ reached 70% in the case of the latter solvent which is 15% higher than for [BMIm] [BF4] 2b. For cyclooctene, the difference was even larger, and 40% increase in epoxide yield could be observed upon changing the ionic liquid from 1 to 2b. For comparison, additional experiments

+ ‐ were carried out in hydrophobic [BMIm] [PF6] 2c. In hydrophobic solvents a biphasic system is present during the epoxidation, which, due to mass transfer limitation, decreases the rate of

+ ‐ epoxidation. In addition, in [BMIm] [PF6] CaLB will stay on the carrier whereas in the nitrates it

113

Chapter 6

will dissolve. In the case of cyclohexene only 50% of the epoxide was obtained and for cyclooctene the yield of epoxide was as low as 38%.

Styrene was the most difficult substrate to oxidize under the cascade conditions and a maximum yield of 40% was obtained in all cases. Better results were obtained for cyclohexene and cyclooctene, where influence of ionic liquids on the reaction outcome could be seen.

Further experiments were performed at higher temperatures. Besides increasing the 2nd reaction step, a beneficial effect on the viscosity, and hence, on the mass transfer is also expected: The viscosity of the ionic liquid decreases with an increase in temperature. In this series of experiments a broader range of ionic liquids could be studied and choline nitrate 3 and

+ ‐ o [TEOA] [NO3] 4 (melting points of respectively 40 and 50 C) were tested as well.

The increase in temperature indeed improved the rate of product formation. With 1 as the solvent, the yield for cyclohexene oxide went up from 70% at RT after 24 h, to 86% at 50 oC

+ ‐ after 5 h. For choline nitrate and [BMIm] [BF4] 70‐75% yield of cyclohexene oxide was observed. The use of solvent 4 resulted in a relatively low yield of 52%. When these cascade reactions were performed in organic solvents, 60 such as acetonitrile, toluene or n‐pentane, only 25‐40% of epoxide was obtained after 5 h (also for cyclohexene). Thus the use of ionic liquid results in significantly faster epoxidation.

Besides Novozym 435, CaLB CLEA, an immobilized version of CaLB was also employed in the reaction. Similar conversions were obtained at RT in the presence of these CaLB preparations, in cyclohexene epoxidations with solvents 1, 2b and 2c under the conditions of Figure 6.3. This is an important observation because this leads to the conclusion that the reaction rate is not influenced whether lipase is dissolved or not (CaLB CLEA will not dissolve in hydrophilic ILs). Together with the observation that different substrates result in large differences in product formation, we can safely conclude that the rate‐determining‐step of the cascade must be the 2nd step i.e. the epoxidation of olefin with peracid.

6.2.2 BaeyerVilliger oxidation in ionic liquids Our final goal was to study the performance of our set of coordinating hydrogen‐bond‐donating (HBD) ionic liquids in the lipase‐driven Baeyer‐Villiger (BV) oxidation as illustrated in Scheme 6.3.

114

Epoxidation and BaeyerVilliger oxidation using H2O2 and lipase

Scheme 6.3 Baeyer‐Villiger chemo‐enzymatic oxidation in ionic liquids

In order to compare the efficiencies of different procedures, the conditions used by Olivio et al.65 served as a starting point to explore the potential of ionic liquids as reaction media. In particular, more acid (2 eq.) was employed, which will lead to a higher concentration of peracid in solution. The results are given in Table 6.1.

Table 6.1 Substrate screening in BV lipase mediated oxidation in 2nd generation ionic liquids

(%) Yield *

+ ‐ + ‐ + ‐ Entry Substrate Product(s) [HOPMIm] [NO3] 1 [choline] [NO3] 3 [TEOA] [NO3] 4

O 19O 91837 O

O O 24O 56254

O 38O 36612

O O 49OH 82980

O O O 75 28 26

5 O H O 23 nd nd

O O 6141019

Conditions: Ketone or aldehyde (0.5 mmol), 50% aq. H2O2 (1 mmol), caprylic acid (1 mmol), 25 mg Nov. 435. T = 50 oC, time 5 h, Selectivity 99% for entries 1‐4 for entry 6, minor formation (3%) of p‐ methylphenol is detected* Yield after 5 h.

115

Chapter 6

The best results were obtained with cyclic ketones, aliphatic ketones and aliphatic aldehydes. The selectivity was found to be around 99% for all tested substrates. Only with 4‐heptanone

+ ‐ two products (acid and ester) were observed. In general, solvent 1 [HOPMIm] [NO3] gave the best results. Cyclopentanone underwent complete conversion in this solvent, while cyclohexanone gave only 45% lactone yield. The quaternary choline nitrate led to better results and a lactone yield of 62%. In all three HBD ionic liquids the ring opening polymerization (ROP) was not observed for the BV‐product of cyclohexanone, despite the presence of CaLB. Menthone resulted in 83% of lactone using 1 as ionic liquid.

The aliphatic aldehyde octanal, was nearly quantitatively converted into acid in 1 as the medium. With benzaldehydes, this cascade approach is not very effective: A maximum of 19% acid was observed with p‐methylbenzaldehyde as the substrate, this time with 4 as solvent. This is probably due to competition of two pathways: the other BV‐pathway will lead to minor amounts up to 3% p‐methylphenol as product.

In conclusion, we have shown for the first time that the lipase‐driven BV‐oxidation in ionic liquids is widely applicable to cyclic ketones as well as aliphatic ketones and aldehydes. In particular solvent 1 turns out to promote the reaction, allowing it to proceed within 5 h at 50 oC. The fact that in this case enzyme is present as a stable homogeneous phase (vide supra) will open up the possibility to operate this reaction in a continuous mode. Further experiments are underway to demonstrate the sustainable process design of the Baeyer‐Villiger reaction.

When comparing our results with cyclohexanone and cyclopentanone with those of Olivio and co‐workers, 65 our methodology is considerably faster compared to reactions in organic solvents such as ethyl acetate, and acetonitrile. For cyclohexanone in ethyl acetate using urea‐hydrogen peroxide (UHP) as the oxidant typically 80% of lactone was observed after 6 days, while we observed a non‐optimized yield of 62% already after 5 hours in 3 choline nitrate. The differences in performance between the different ILs led us to believe, that further optimization is possible when exploring a larger structural diversity in these ILs. With cyclopentanone, the effect was even more pronounced: We obtained a straightforward 99% yield of the lactone after 5 h, using solvent 1. In contrast, when using acetonitrile, under rather similar conditions, but using the more expensive UHP as oxidant, 73% yield of lactone was observed after 24 hours at 60 oC.

Our results lead us to believe that coordinating, hydrogen bond donating solvents have a beneficial influence on the Baeyer‐Villiger rearrangement. This can readily be envisaged based

116

Epoxidation and BaeyerVilliger oxidation using H2O2 and lipase

on the mechanism, which is given in Figure 6.4. Rearrangement of the so‐called Criegee‐ intermediate is generally rate determining, and will be facilitated by proton‐donating solvents.

Figure 6.4. Mechanism of the Baeyer‐Villiger reaction with peracids

6.2.3 Lipase solubility, stability and hydrolytic activity in ionic liquids Studies on enzyme dissolution in ionic liquids revealed that ionic liquids that form strong hydrogen bonds such as alkylmethylimidazolium [AlkylMIm]+, with nitrate, acetate or dicyanimide as anion dissolve Candida antarctica Lipase B free enzyme. In this case the dissolution process was accompanied by nearly complete or temporary activity loss. 51 On the basis of IR spectroscopy, this activity loss was attributed to interference with the H‐bonds that maintain the protein secondary and tertiary structure. 71

In this study, we embarked on the screening of a novel class of ionic liquids, namely OH‐bond containing ILs. These solvents are proposed to provide a balance of mild hydrogen bond‐ accepting and donating properties. Their H‐bond donating properties could possibly contribute to the stability of enzymes in solution. 51

+ ‐ The range of solvents studied is presented in Figure 6.1 and consists of [HOPMIm] [NO3] 1, + ‐ + ‐ [BMIm] [NO3] 2a, choline nitrate 3 and triethanolammonium nitrate [TEOA] [NO3] 4. All these ionic liquids are hydrophilic, and for comparison the ionic liquid with weakly coordinating anion

+ ‐ 2b [BMIm] [BF4] was included in the study.

We thus wanted to investigate whether immobilized lipase is stable in these solvents, and whether activity can be maintained, either immobilized or dissolved. As activity assay the hydrolysis of triacetin was used (Scheme 6.4). As immobilized formulations of CaLB, both

117

Chapter 6

Novozym 435 (Nov. 435) as well as cross‐linked enzyme aggregates of lipase CaLB CLEA were tested. 72

Scheme 6.4 Natural reaction: Hydrolysis of triacetylglycerol (triacetin)

In Figure 6.5 the activity is given for Novozym 435 in the range of solvents studied. As reference phosphate buffer was used. The batch of CaLB immobilized on polyacrylic resin (Nov. 435) used in the activity assay had an activity of 388 U/g.

800

700

600

500

400 U/g

300

200

100

0 Nov 435 in HOPMIm NO3 BMIm NO3 BMIm BF4 choline NO3 TEOA NO3 phosphate 1 2a 2b 3 4 buffer

Figure 6.5 Activity of Novozym 435 in hydrophilic ionic liquids using triacetin hydrolysis as assay

In the case of ionic liquids 2a, 1, and 4 the enzyme displayed a twofold activity enhancement compared to Novozym 435 in phosphate buffer. Thus, in ionic liquids with the strongly

+ ‐ + ‐ coordinating nitrate anion 2a [BMIm] [NO3] and 1 [HOPMIm] [NO3] the activity is enhanced + ‐ markedly. In contrast, 2b [BMIm] [BF4] which contains only a weakly coordinating anion gave comparable activity to Novozym 435, thus underlining the importance of nitrate. In the case of tertiary ammonium based ionic liquids, with the Novozym 435 activity is significantly better in 4 [TEOA]+, compared to 3 (choline cation). Although both salts contain free alcohol groups (three ethanolic groups for 4, compared to one ethanolic group in 3) and nitrate as counter anion, 4 was a much better solvent, and thus triethanolammonium a much better cation.

118

Epoxidation and BaeyerVilliger oxidation using H2O2 and lipase

Dissolution of Novozym 435 in ionic liquids

Previous tests revealed that ILs could dissolve free CaLB and could also desorb (leach) CaLB from the polymeric support in Nov. 435. In order to study the stability of immobilized lipase, the same activity assay was applied to the supernatant of reactions conducted in ionic liquids. In this case the solution of CaLB with ionic liquids was filtered after incubation for 1 h in order to remove the resin. Next the standard activity test was applied (Figure 6.6).

900 800 700 600 500

U/g 400 300 200 100 0 Nov 435 in HOPMIm BMIm NO3 BMIm BF4 choline NO3 TEOA NO3 phosphate NO3 1 2a 2b 3 4 buffer

Figure 6.6 Activity of filtered Novozym 435 solution in ionic liquid (triacetin hydrolysis assay)

+ ‐ Roughly the same activity (96‐100%) was observed with 2a [BMIm] [NO3] and 1 + ‐ [HOPMIm] [NO3] compared to the starting heterogeneous preparation. In other words the enzyme dissolved and remained active in the homogeneous IL‐solution. This indicates that the activity previously observed was completely due to dissolved enzyme. It is therefore safe to

+ ‐ conclude that most of the enzyme was desorbed from the support. For A]4 [TEO [NO3] 49% of + ‐ the activity was still present, indicating significant dissolution of enzyme as well. [BMIm] [BF4] 2b with its weakly coordinating anion, was not very effective in dissolving lipase CaLB. Only 18% of total activity was found in the hydrolytic test. In the case of 3 [choline]+[nitrate]‐, hardly any dissolution occurred (2‐5% of activity was observed).

CaLB CLEA

The technique of Cross Linked Enzyme Aggregates is another way of immobilizing enzyme and allows for a broad application of enzymes under a variety of conditions. 73 The same series of experiments was applied to CaLB CLEA. In the activity assay for our CLEA‐preparation in phosphate buffer, an activity of 1001 U/g was measured. Filtered solutions of CaLB CLEA in our range of ionic liquids displayed however very low activity (about 8% of activity in aqueous

119

Chapter 6

solution). This demonstrates that CaLB CLEA is stable in a range of ionic liquids. Subsequently, CaLB CLEA was dispersed and its activity measured as previously described. (Figure 6.7).

2000 1800 1600 1400 1200 1000 U/g 800 600 400 200 0 CLEA‐CALB in CLEA‐CALB in CLEA‐CALB in CLEA‐CALB in CLEA CALB in CLEA CALB in phosphate HOPMIm NO3 BMIm NO3 BMIm BF4 choline NO3 TEOA NO3 buffer 1 2a 2b 3 4

Figure 6.7 CaLB CLEA’s activity in dispersed ILs

Analogous to Novozym 435, an increase in activity could be detected using ionic liquids, rather than phosphate buffer as the reaction medium. In this case, a different order was observed, reflecting the activity of heterogeneous CaLB in these solvents. In this case the hydrophobic 2a gave almost a doubling in activity, followed by the hydrophilic 2b and 4. The previously found

+ ‐ high performance of 1 [HOPMIm] [NO3] as solvent could not be repeated here. Apparently, this solvent is more effective in stabilizing CaLB under homogeneous conditions.

Stability and ageing of the enzymes in ILs.

The stability and ageing of CaLB dispersed in ILs was studied. Both Novozym 435 and CaLB CLEA

+ ‐ + ‐ were kept in ILs (1 [HOPMIm] [NO3] , 3 choline nitrate, 4 [TEOA] [NO3] ) continuously for three months at ± 5 oC. Afterwards, the solutions were subjected to the standard activity test. It turned out that for all ionic liquids 1, 3, 4 no change in the hydrolytic activity was observed.

Notably, the catalytic activity of the enzyme for other (anhydrous) reactions, either dispersed or dissolved in ionic liquid, will probably differ from the activity measured in the standard hydrolysis test. In a previous study, the transesterification of ethyl butanoate with n‐butanol was applied as a test reaction to evaluate the activity of CaLB in a range of ionic liquids. 51

This conclusion contrasts with the results obtained in hydrolysis reactions. For the

+ ‐ transesterification reaction, both hydrophilic 2b [BMIm] [BF4] , as well as hydrophobic 2c + ‐ [BMIm] [PF6] were excellent solvents. This result points towards the beneficial influence of anions with a low coordinating character. 51,74

120

Epoxidation and BaeyerVilliger oxidation using H2O2 and lipase

Comparison of CaLB CLEA and Novozym 435 in the epoxidation reaction

For epoxidation reactions, run at room temperature, the use of Novozym 435 was compared to that of CaLB CLEA (Table 6.2). Epoxidation of cyclooctene was studied in this case. Nearly identical results were obtained for all solvents when using CaLB CLEA instead of Nov. 435 as the source of lipase (see Table 6.3). Note that in the case of CaLB CLEA 2.5 x as much units of lipase were present (1000 vs. 400 units/g in sample). This, together with the observation that

+ ‐ dissolution of enzyme (as for Nov. 435 in 1 [HOPIm] [NO3] ) does not influence the outcome of the reaction, leads to the conclusion that the rate‐determining step of the cascade reaction, is not influenced by lipase and is therefore most likely step 2 i.e. the epoxidation of olefin with peracid (see Scheme 6.2).

Table 6.2. CaLB CLEA vs. Nov. 435 in chemo‐enzymatic epoxidation of cyclooctene in ionic liquids after 24 h

(%) yield of cyclooctene oxide

ILs [HOPMIm][NO3] 1 [BMIm][BF4] 2b [BMIm][PF6] 2c

Novozym 435 69 39 38 CaLB CLEA 72 41 47

Conditions: cyclooctene 1.48 mmol, 50% aq. hydrogen peroxide 2.6 mmol (5x added with intervals of 1 h), caprylic acid 0.2 mmol, enzyme CaLB (Nov. 435 or CaLB CLEA, both 10 mg) ionic liquid 1 mL, RT, 24 h.

6.3 Conclusions In this study we demonstrated the beneficial influence of coordinating hydrogen‐bond‐donating ionic liquids on the lipase‐driven epoxidation and Baeyer‐Villiger oxidation. The effect is two‐ fold: (1) The second step in the cascade, namely the oxidation of substrate by peracid is significantly accelerated by these solvents, which readily can be operated at 50 oC within 5 h. The rearrangement of the Criegee‐intermediate in the BV‐oxidation seems to be facilitated by protonated solvents. (2) The enzyme is dissolved and highly stable in these coordinating‐HBD‐ solvents. This opens the possibility to recycle and reuse the enzyme‐IL as a whole, and allowing for continuous operation.

121

Chapter 6

With regard to the Baeyer‐Villiger reaction, a wide range of substrates could be oxidized. Especially cyclic ketones and aliphatic ketones and aldehydes are prone to oxidation. The fact that different ionic liquids give different results again confirms their design properties: for every substrate a different IL can be designed for the best results. Thus, the application of a new generation of hydrogen‐bond‐donating ionic liquids is a step forward in the search for alternative environmentally benign solutions in biocatalytic transformations.

6.4 Experimental data General Cyclopentanone (99%), cyclohexene (99.5%), cyclooctene (95%), styrene 98% and, phosphorous pentoxide were purchased from Fluka, Caprylic acid (99.5%), triacetin (99%), n‐decane (99%), n‐ dodecane (99%), 50% aqueous solution of hydrogen peroxide, p‐methylbenzaldehyde (99%), 4‐ heptanone (99%), menthone (99%) and, octanal (99%) were purchased from Acros, Activated basic aluminium oxide (for column chromatography, 50‐200 micron), ion exchange resin Dowex 1X8, N‐methylimidazole, 1‐chloro‐3‐hydroxypropane (97%), cyclohexanone (99%), cyclooctanone (99%), cyclooctene oxide 98% and cyclohexene oxide (98%) were purchased from Sigma–Aldrich. Solvents used for extraction were p.a. grade. 1‐Butyl‐3‐methylimidazolium + ‐ 70 nitrate ([BMIm] [NO3] ) was synthesized and characterized as described. Novozym 435 (CaLB adsorbed on an acrylic carrier) was kindly donated by Novozymes (Bagsvaerd, Denmark); CaLB CLEA was donated by CLEA Technologies.

Analysis

Nuclear Magnetic Resonance (NMR) analysis 1 H 300 Hz and 13 C 100 Hz or 300 Hz spectra were recorded on a Bruker AC 400 and Varian Inova VXR‐400S spectrometers using TMS as an external standard and chemical shift were expressed in ppm. Samples were dissolved either in CD3OD, CDCl3 as solvent.

Gas chromatography analysis Analysis were carried out with a GC Varian Star 3400 or 3600 instrument equipped with a non‐ polar CP Sil‐5 CB 50 m*0.53 mm column df=2.0 µm , T max = 300 oC The internal standard with epoxidation reactions was, n‐decane, in Baeyer‐Villiger oxidations n‐ dodecane was used. Column temperature profile for cyclohexene: 50 oC, hold time 1 min, rate 15 oC /min until 240 oC hold time 1 min Injector temperature profile: 70 oC (2 min), rate 15 oC /min to 250 oC

Column temperature profile for cyclooctene, styrene: 65 oC, hold time 1 min, rate 10 oC/min until 165 oC hold time 4 min Injector temperature profile: 85 deg (2 min), rate 15 oC/min to 250 oC

GC sample preparation For experiments performed at RT or 50 oC samples of 50‐100 µl were taken from reaction mixture after 5 h or 24 h. To all reaction samples 500 µl of diethyl ether or methyl t‐butyl ether were added for extraction since the GC column does not tolerate the presence of ionic liquid. Sample was first vortexes for 30 sec. and then centrifuged. To the upper layer MnO2 was added to decompose any peroxide and filtered off. Also anhydrous MgSO4 was added to remove

122

Epoxidation and BaeyerVilliger oxidation using H2O2 and lipase

water. An additional 500 µl of diethyl ether was added during filtration. Before injection on GC an additional filtration through cotton and basic γ‐Al2O3 was required.

Activity assay for CaLB or CaLB CLEA The hydrolytic activity was measured using a triacetin assay. The hydrolysis of triacetin (1,2,3‐ triacetoxypropane, 380 µl, 0.1M) in NaH2PO4 buffer (0.1 M, 19.62 mL) in the presence of 10 or 50 mg CaLB at pH 7.5 and RT was monitored by titration of the formed acetic acid with 0.1 M NaOH solution using an automated titrator. One unit (U) will liberate one mol of acetic acid per min.

Synthesis

GC scale chemo‐enzymatic epoxidation of cyclohexene in ILs Cascade conditions: RT, cyclohexene 1.5 mmol, caprylic acid 0.2 mmol, n‐decane (internal standard (IS) 0.12 mmol), solvent (1 mL), CaLB (10 mg), 50% hydrogen peroxide was added in 5 portions at 1 h intervals (total 2.6 mmol). Extraction of the product was done with either diethyl ether or methyl‐tert‐butyl ether. Reaction was monitored with GC and the resulting sample compared with the referential product. GC scale chemo‐enzymatic B‐V oxidation of cyclooctanone in ILs Novozym 435 (25 mg) was first added to 1 mL of ionic liquid to form a gel‐like homogeneous liquid. Next the ketone (0.5 mmol), n‐dodecane (IS, 0.5 mmol), and octanoic acid (1 mmol) were o added. The reaction mixture was heated to 50 C. To start the reaction 50% H2O2 (in total 1 mmol) was added in 4 portions at 1 h intervals (4x15.5 µl). The reactions were carried out in a thermostatted shaker. The first sample was withdrawn 1 h after the total amount of peroxide had been added). Representative sample of 100 µl were taken from the reaction mixture and extracted with total of 1 mL of diethyl ether using above described GC work‐up procedure. The clear solution was then injected on GC.

Synthesis of ionic liquids + ‐ [HOPMIm] [NO3] (1) 1‐(3‐Hydroxypropyl)‐3‐methylimidazolium nitrate was prepared in a two step procedure. In the first step 1‐(3‐hydroxypropyl)‐3‐methylimidazoliumchloride [HOPMIm]+[Cl]‐ was prepared by reacting N‐methylimidazole with 3‐chloropropanol according to a published procedure.70 1‐ Methylimidazole (34.75 g; 0.42 mol) was slowly added to freshly pre‐distilled 1‐chloro‐3‐ propanol (55.05 g, 0.58 mol) in a three neck round‐bottom flask equipped with a reflux condenser under a constant flow of nitrogen to avoid decomposition of 1‐methylimidazole. The mixture was stirred at 60 oC for 4 days while the flow of nitrigen was maintained. The long reaction time was required to drive the reaction to completion. After completion, the reaction mixture was extracted three times with ethyl acetate (50 mL) to remove any excess of 3‐ chloropropanol. The upper, organic, phase was decanted ande th lower, viscous [HOPMIm]+[Cl]‐ ionic liquid phase was concentrated by rotary evaporation. In total 72 g of [HOPMIm]+[Cl]‐ was obtained corresponding with a yield of 97%. 1 H NMR (300 MHz, CD3OD, TMS ppm) δ 8.99 s1H (N‐CH=N), 7.66; 7.59 d 2H (CH=CH), 4.78 s 1H ; OH), 4.37; 4.35; 4.33 t 2H (CH2‐CH2‐OH), 3.95 s 3H (CH3‐N), 3.65; 3.63; 3.61 t 2H (N‐CH2‐CH2‐CH2‐ OH), 2.13; 2.11; 2.10; 2.07; 2.062; 2.06; 2.04 m 2H (‐CH2‐CH2‐CH2OH); 13 C NMR (100 MHz, CD3OD, TMS ppm) δ 32.42; 35.41; 48.16; 58.36; 123.71; 137.02

123

Chapter 6

In the second step an anion exchange was performed using a 400 mL ion exchange column (capacity approximately 0.4 mol exchange sites), filled with 250 g Dowex 1X8 200 anion exchange resin. The Cl‐ saturated resin was rinsed with Milli‐Q water (1L). A 1 M solution of ‐ sodium nitrate (pH 6.5 adjusted with fuming HNO3) was introduced to the column to replace Cl ‐ 75 with NO3 . The presence of chloride was checked using silver chromate titration. A 0.2 M solution in Milli–Q water of [HOPMIm]+[Cl]‐ from step 1 was slowly passed through the column where anionic exchange took place. The eluted fraction was collected and controlled for the presence of chloride. Subsequently, the column was rinsed with additional of Milli‐Q water (1L). + ‐ The thus eluted aqueous solution of [HOPMIm] [NO3] was concentrated by rotary evaporation. The obtained colourless viscous ionic liquid was extensively dried in vacuo over P2O5 for at least 48 h. A chloride test with silver chromate indicated < 30 ppm Cl‐, consistent with a near‐ quantitative exchange. 1 H NMR (300 MHz, CDCl3 ,ppm) δ 8.71 s 1H (N‐CH=N), 7.48; 7.42 d 2H (CH=CH), 4.68 s 1H (D2O; OH), 4.30; 4.28; 4.26 t 2H (CH2‐CH2‐OH), 3.88 s 3H (CH3‐N), 3.62; 3.60; 3.58 t 2H (N‐CH2‐CH2‐CH2‐ OH), 2.13; 2.11; 2.10; 2.08; 2.062; 2.06; 2.04 m 2H (‐CH2‐CH2‐CH2OH); 13 C NMR (100‐300 MHz, CDCl3 ppm) δ 31.12; 37.16; 48.00; 59.45; 123.80; 137.64

Chloride test75 Solutions of silver nitrate AgNO3 (1mM) and potassium chromate K2CrO4 (5g/L) in Milli‐Q water were prepared as described. A single drop of the potassium chromate solution was added to a sample (a few drops) of ionic liquid solution eluted from the column. Secondly the titrant (AgNO3) was carefully added drop wise. The appearance of a red coloured Ag2CrO4 precipitate (positive result, solution free of chloride) upon addition of a first drop of titrant indicates that the concentration of Cl– is essentially less than 600 ppm. An appearance of a green colour indicates formation of AgCl, thus confirms still a high content of Cl‐ anions (negative result).

+ ‐ + ‐ Choline nitrate [HOCH2CH2Me3N] [NO3] = [TMEOA] [NO3] (3) Preparation of choline nitrate required only an ion exchange from chloride to nitrate. An aqueous solution (0.2 M) of commercial choline chloride was passed through the exchange column saturated with 1 M aqueous solution of NaNO3 as described above. The eluted fraction was tested for the presence of chloride and water was removed by rotary evaporation. Choline nitrate appeared as a white‐crystaline solid at room temperature. Pure highly hydrophilic choline nitrate was dried and stored in vacuo over P2O5 for at least 48 h. 100% exchange took place.

+ ‐ Triethanolammonium nitrate [TEOA] [NO3] (4) Similar to choline nitrate, the preparation of ionic liquid 4 required an ion exchange from chloride to nitrate anion. For this purpose 0.1M of commercial triethanolamine hydrochloride 99% (18,6 g in 1l milli‐Q water) was passed through the exchange dowex resin column saturated with 1M aqueous solution of sodium nitrate according to the described procedure. The eluted fraction was tested for the presence of chloride and water was removed by rotary evaporation. + ‐ The obtained 20.4 g of [TEOA] [NO3] appeared as a cream‐white crystalline solid at room o + ‐ temperature with a melting point of 50 C, The highly hydroscopic [TEOA] [NO3] was dried and stored in vacuo over P2O5 for at least 48 h.

124

Epoxidation and BaeyerVilliger oxidation using H2O2 and lipase

6.5 References

1P. Walden, „ Molecular Weights and Electrical Conductivity of Several Fused Salts.” Bull. Acad. Sci., (St Petersburg) 1914, 405. 2M. J. Earle, K. R. Seddon, Pure Appl. Chem., 2000, 72, 1391. 3M. J. Earle, J. M. S. S. Esperanca, M. A. Gilea J. N. Canongia Lopes, L. P. N. Robelo, J. W. Magee, K. R. Seddon, J. A. Wildegren, Nature, 2006, 439, 831. 4M. Smiglak, W. M. Reichert, J. D. Holbrey, J. S. Wilkes, L. Sun, J. T. Thrasher, K. Kirichenko, S. Singh, A. R. Katritzky, R. D. Rogers, Chem. Commun., 2006, 2554. 5Q. Liu, M. H. A. Janssen, F. van Rantwijk, R. A. Sheldon, Green Chem., 2005, 7, 39. 6K. R. Seddon, J. Chem. Technol. Biotechnol., 1997, 68, 351. 7M. Freemantle, Chem. Eng. News 2000, 15, 37. 8J. G. Huddleston A. E. Visser, W. Matthew Reichert, H. D. Willauer, G. A. Broker, R. D. Rogers, Green. Chem., 2001, 3, 156. 9J. D. Holbrey, K. R. Seddon, J. Chem. Soc., Dalton Trans., 1999, 1, 2133. 10H. Olivier‐Bourbigou, L. Magna, J. Mol. Catal. A, Chem., 2002, 182, 419. 11T. Welton, Coord. Chem . Rev., 2004, 248, 2459. 12T. Welton, Chem. Rev., 1999, 99, 2071. 13J. G. Huddleston, H. D. Willauer, R.P. Swatloski, A. E. Visser, R. D. Rogers, Chem, Commun., 1998, 1765. 14P. Wasserscheid, W. Keim, Angew. Chem. Int. Ed., 2000, 39, 3772. 15R. A. Sheldon, Chem. Commun., 2001, 2399, 16R. D. Rogers, K. R. Seddon, Ionic liquids, Industrial applications to Green Chemistry (Eds.), (series 818) ACS Symposium, 2002. 17P. Wasserscheid, T. Welton, 2002, Ionic liquids in Synthesis, (Eds.), Wiley‐VCH. 18F. Liu, M. B. Abrams, R. T. Bakker, W. Tumas, Chem. Commun., 2001, 5, 433. 19D. Ballivet‐Tkatchenko, M. Picquet, M. Solinas, G. Francio, P. Wasserscheid, W. Leitner, Green Chem., 2003, 5, 232. 20K. Tominaga, Y. Sasaki, Chem. Lett., 2004, 33, 14. 21K. Tominaga, Catal. Today, 2006, 115, 70. 22A. M. Scurto, W. Leitner, Chem. Commun., 2006, 3681. 23M. D .Bermejo, A. J. Kotlewska, L. J. Florusse, M. J. Cocero, F. van Rantwijk, C. J. Peters, Green Chem., 2008, 10, 1049. 24K. M. Dochterty, C. F. Kulpa, Green Chem., 2005, 7, 185. 25C. Pretti, C. Chiappe, D. Pieraccini, M. Gregari, F. Abramo, G. Monni, L. Intore, Green Chem., 2006, 238. 26R. J. Bernot, E. E. Kennedy, G. A. Lamberti, Environ. Toxicol. Chem., 2005, 24, 87. 27N. Gathergood, P. J. Scammells, M. T. Garcia, Green Chem., 2006, 8, 156. 28J. H. Davis, Chem. Lett., 2004, 33, 1072. 29G. Imperato, B. Koenig, C. Chiappe, Eur. J. Org. Chem., 2007, 1049. 30A. Zaks, A. M. Klibanov, Science, 1984, 224, 1249. 31A. Zaks, A. M. Klibanov, Proc. Natl. Acad. Sci., U. S. A. 1985, 82, 3192. 32A. M. Klibanov, Chemtech., 1986, 354. 33A. J. Carmichael, K. R. Seddon, J. Phys. Org. Chem., 2000, 13, 591. 34S. G. Cull, J. D. Holbrey, V. Vargas‐Mora, K. R. Seddon, J. Biotechnolol. Bioeng., 2000, 69, 227. 35M. C. de Zoete, A. C. Kock‐van Dalen, F. van Rantwijk, R. A. Sheldon, Biocatalysis, 1995, 10, 307. 36R. M. Lau, F. van Rantwijk, K. R. Seddon, R. A. Sheldon, Org. Lett., 2000, 2, 4189. 37T. Itoh, E. Akasaki, K. Kudo, S. Shirakami, Chem. Lett., 2001, 262.

125

Chapter 6

38R. A. Sheldon, R. M. Lau, M. J. Sorgedrager, F. van Rantwijk, K. R. Seddon, Green. Chem., 2002, 4, 147. 39S. H. Schofer, N. Kaftzik, P. Wasserscheid, U. Kragl, Chem. Commun., 2001, 425. 40S. E. Godtfredsen, O. Kirk, F. Bjőrkling, L. B. Christensen, WO 91/06574 (31. 10. 1990). 41F. Bjőrkling, S. E. Godtfredsen, O. Kirk, J. Chem. Soc., Chem. Commun., 1990, 1301. 42F. Bjőrkling, H. Frykman, S. E. Godtfredsen, O. Kirk, Tetrahedron, 1992, 48, 4587. 43K. Adelhorst, F. Bjőrkling, S. E. Godtfredsen, O. Kirk, Synthesis, 1990, 112. 44O. Kirk, F. Bjőrkling, S. E. Godtfredsen, WO 91/04333 (01.04.1991). 45M. C de Zoete, F. van Rantwijk, L. Maat, R. A. Sheldon, Recl. Trav. Chim., Pays‐Bas, 1993, 112, 462. 46O. Kirk, M. W. Christensen, T. Damhus, S. E. Godtfredsen, Biocatalysis, 1994, 11, 65. 47M. R. Gen. Klaas, S. Warwel, J. Mol. Catal., A, 1997, 117, 311. 48F. van Rantwijk, R. M. Lau, R. A Sheldon, Trends in Biotechnology, 2003, 21, 131. 49S. Cantone, U. Hanefeld, A. Bosso, Green. Chem., 2007, 9, 954. 50F. van Rantwijk, R. A. Sheldon, Chem. Rev., 2007, 107, 2757. 51R. M. Lau, M. J. Sorgedrager, G. Carrea, F. van Rantwijk, F. Secundo, R. A. Sheldon, Green Chem., 2004, 6, 483. 52P. Lozano, T. De Diego, D. Carrié, M. Vaultier, J. L. Iborra, Chem. Commun., 2002, 7, 692. 53M. T. Reetz, W. Wiesenhőfer, G. Franciό, W. Leitner, Adv. Synth. Catal., 2003, 345, 1221. 54P. Lozano, T. De Diego, M. Larnicol, M. Vaultier, J. L Iborra, Biotechnol. Lett., 2006, 28,1559. 55A. Baeyer, V. Villiger, 1899, "Einwirkung des Caro'schen Reagens auf Ketone “ Ber. 32, 3625. doi:10.1002/cber.189903203151. 56A. Baeyer, V. Villiger, 1900, "Ueber die Einwirkung des Caro'schen Reagens auf Ketone" Ber. 33, 858. doi:10.1002/cber.190003301153. 57I. W. C. E. Arends, Angew. Chem., 2006, 45, 38, 6250. 58G‐J. ten Brink, B. C. M. Fernandes, M. C. A. van Vliet, I. W. C. E. Arends, R. A. Sheldon, J. C. S. Perkin Trans., 1 2001, 224; G.‐J. ten Brink, J.‐M. Vis, I. W. C. E. Arends, R. A. Sheldon, J. Org. Chem., 2001, 66, 2429. 59O. Kirk, F. Bjorkling, S. E. Godtfredsen, US Pat. 5541092, 1996, Novo Nordisk. 60M. C. de Zoete, F. van Rantwijk., R. A. Sheldon, Catal. Today, 1994, 22, 563. 61M. A. Moreira, T. B. Bitencourt, M. G. Nascimento, Synth. Commun., 2005, 35, 2107. 62M. R. G. Klaas, S. Warwel, J. Am. Oil Chem. Soc., 1996, 73, 1453. 63C. Lemoult, P. F. Richardson, S. M. Roberts, J. Chem. Soc., Perkin Trans. 1, 1995, 89. 64B. K. Pchelka, M. Gelo‐Pujic, E. Guibé‐Jampel, J. Chem. Soc., Perkin Trans. 1, 1998, 2625. 65M. Y. Rios, E. Salazar, H. F. Olivo, Green Chem., 2007, 9, 459; M. Y. Rios, E. Salazar, H. F. Olivio, J. Mol. Catal. B: Enzymatic, 2008, 54, 61. 66V. Conte, B. Floris, P. Galloni, V. Mirruzzo, A. Scarso, D. Sordi, G. Strukul, Green Chem., 2005, 7, 262. 67R. Bernini, A. Coratti, G. Fabrizi, A. Goggiamani, Tetrahedron Lett., 2003, 44, 8991. 68S. P. Panchgalle, U. R. Kalkote, P. S. Niphadkar, P. N. Joshi, S. P. Chavan, G. M. Chaphekar, Green Chem., 2004, 6, 308. 69A. J. Walker, N. C. Bruce, Chem. Commun., 2004, 2570; Tetrahedron, 2004, 60, 561. 70J. S. Wilkes, M. J. Zaworotko, J. Chem. Soc., Chem. Commun., 1992, 965. 71F. van Rantwijk, F. Secundo, R. A. Sheldon, Green Chem., 2005, 8, 282. 72O. Kirk, M. W. Christensen, Org. Proc. Research & Develop., 2002, 6, 446. 73P. Lόpez‐Serrano, L. Cao, F. van Rantwijk, R. A. Sheldon, Biotechnolog. Lett., 2002, 24, 1379. 74F. van Rantwijk, F. Secundo, R. A. Sheldon, Green Chem., 2006, 8, 282. 75“Chloride measurements by Argentometric Method” Standard Methods (Eds.), Method 4500‐ Cl‐B, online resources. http://.environmentalet.org/env1221/titrations.htm

126

SSuummmmaarryy

Summary

Chapter 1. Introduction to sustainable oxidation chemistry. At the start of the thesis an introduction is given towards the need for sustainable oxidation technologies. Examples are given of recent methods for the oxidation of alcohols. Despite their greenness, their dependency on e.g. rare transition metals and limitations in terms of substrate scope narrows their window of application. Therefore the need for a clean, safe and multipurpose oxidation method is ongoing.

Chapter 2. Hypervalent iodine in organic synthesis In this chapter an introduction to hypervalent iodine chemistry is given and a number of examples for synthetic application of hypervalent iodine(III) and iodine(V) reagents are presented. These examples demonstrate that there is a large potential, especially for application of iodine(III) reagents, due to their versatility.

Chapter 3. Polymerattached iodine(III) reagent in selective oxidations. Scope of oxidants. In this chapter polymer bound iodine(III) reagents are introduced. In total four different polymer‐attached iodosobenzene dialkanoate derivatives (C2, C4, C6 and C8) were successfully synthesized. An extensive safety study was conducted, that revealed that the linear C4 and C8 oxidants were the most stable and easy to handle oxidants with loadings of 1.71 and 1.29 mmol I(III)/g polymer respectively.

These oxidants act as active oxidants in the direct oxidation of benzyl alcohol to benzaldehyde, without the need for additives. The influence of the polystyrene backbone was studied. It turned out that the iodine(III) oxidant anchored to linear polystyrene performs far better than its cross‐linked counterpart.

An alternative greener procedure to the commonly applied iodination method was successfully developed. It circumvents the need for chlorinated solvents commonly used in iodination of polystyrene.

Chapter 4. Polymerattached iodine(III) reagent in selective oxidations. Scope of substrates. This chapter reveals a broad array of alcoholic substrates which can undergo selective oxidation to carbonyl products when polymeric iodine(III) reagent is applied as sole oxidant. Most importantly the development of the optimal oxidation procedure was presented. It turned out that, opposite to what was reported earlier, the catalytic use of inorganic salt such as potassium bromide is not necessary to allow for selective formation of carbonyl products.

128

Summary

It is shown that the different polymer‐attached iodine(III) oxidants slightly differ in their activities towards alcoholic substrates. The C4 derivative was the most active in the oxidation of hydrophobic and benzylic alcohols. The C6 derivative gave the best yield of carbonyl products

with steroidal and substituted benzylic alcohols. And finally the C2 or C8 derivatives are promising for oxidation of secondary, aliphatic and cyclic alcohols. For activated allylic alcohols a rapid and selective oxidation towards the corresponding aldehydes and ketones was observed in all cases. Additionally, the alcohol oxidation with hypervalent iodine(III) oxidant could be carried out in ionic liquid as the reaction solvent.

It was shown that iodine(III) reagents could be recovered and regenerated for at least three runs, with minor loss in activity. A mechanism was postulated for the iodine(III) mediated oxidation of alcohols. This two‐step reaction is supported by a Hammett plot study and initial rate measurements.

Chapter 5. Lipase catalyzed, in situ generation of hypervalent iodine(III) reagent for selective alcohol oxidation. Chapter 5 describes a novel approach for the in situ synthesis and direct application of iodine(III) oxidants for alcohol oxidation. It is shown that in situ generation of the polymeric iodine(III) oxidant by reaction with percarboxylic acid, which is also formed in situ by lipase catalyzed perhydrolysis of the corresponding carboxylic acid, proved to work under certain conditions.

We disclose the limitations of the one‐pot procedure for simple alcohols. For substrates such as cyclohexanol and benzyl alcohol, the esterification reaction dominates over the desired transformations. Hence we produced carbonyl products using a three‐step approach. In this way the undesired side reactions were suppressed. In addition, the step‐wise reactant additions eliminate mass transfer limitations by physically separating the enzyme from the polymer‐ attached iodine catalyst. We have shown that two temperatures regimes were required: 50 oC for the peroxidation and iodine(III) formation and 70 oC for the alcohol oxidation step. Hydrogen peroxide was needed in excess for this procedure (2‐4 eq.).

Importantly, we found that bulky steroidal substrates, due to steric limitations, did not undergo CaLB catalyzed esterification under the applied conditions. Thus oxidation of the alcoholic group could be performed with iodine(III) in the presence of lipase and acid. This finally resulted in a

one‐pot technology for the oxidation of steroidal alcohols using the H2O2‐acid‐iodine(III) cascade in a two‐temperature regime.

129

Summary

A beneficial influence of coordinating hydrogen‐bond‐donating ionic liquids could be demonstrated for the lipase‐driven epoxidation and Baeyer‐Villiger oxidation. The effect is two‐ fold: (1) The second step in the cascade, namely the oxidation of substrate by peracid is significantly accelerated by these solvents under mild conditions. The rearrangement of the Criegee‐intermediate in the BV‐oxidation seems to be facilitated by protonated solvents. (2) The enzyme is dissolved and highly stable in these coordinating‐Hydrogen Bond Donating‐ solvents. This opens the possibility to recycle the enzyme‐ILs as a whole and allowing for continuous operation.

Chapter 6 Epoxidation and BaeyerVilliger oxidation using hydrogen peroxide and lipase dissolved in ionic liquids. In this chapter the application of ionic liquids as reaction solvent for lipase–driven transformations was pursued. Two transformations: epoxidation and Baeyer‐Villiger oxidation were conducted in a chemo‐enzymatic two‐step cascade approach. Novozym 435 and CaLB CLEA immobilized versions of CaLB were applied for cascade reactions in hydrogen‐bond‐ donating ionic liquids. We found that dissolution of enzyme in the case of Novozym 435 occurred in ionic liquids, without influencing the rate of the reaction. It was shown that enzyme dissolution resulted in a stable homogeneous phase, which opens up the possibility to operate this reaction in a continuous mode.

With regard to the Baeyer‐Villiger reaction, a wide range of substrates could be oxidized. Especially cyclic ketones and aliphatic ketones and aldehydes are prone to oxidation. The fact that various ionic liquids gave different results confirmed their design properties. For every substrate a suitable ionic liquid can be designed in order to lead to the best results. Thus, the application of a new generation of hydrogen‐bond‐donating ionic liquids is a step forward in the search for alternative environmentally benign solutions in biocatalytic transformations.

130

SSaammeennvvaattttiinngg

Samenvatting

Hoofdstuk 1. Inleiding tot duurzame oxidatie chemie. Het proefschrift begint met een inleiding tot de noodzaak van technologieontwikkeling op het gebied van duurzame oxidaties. Dit wordt gedemonstreerd aan de hand van een aantal voorbeelden van recentelijk ontwikkelde methoden voor de oxidatie van alkoholen. Ondanks hun groene karakter zijn deze afhankelijk van zeldzame overgangsmetalen en hun beperkte substraat scope beperkt hun toepasbaarheid. Dientengevolge is er nog altijd behoefte aan een schone, veilige en algemeen toepasbare oxidatiemethode.

Hoofdstuk 2. Hypervalent jodium in organische synthese. In dit hoofdstuk wordt een introductie gegeven op jodiumchemie en worden een aantal voorbeelden van de synthetische toepassing van hypervalent jodium(III) en jodium(V) reagentia gegeven. Deze voorbeelden laten zien dat er vanwege hun veelzijdigheid een groot potentieel schuilt in de toepassing van met name jodium(III) verbindingen.

Hoofdstuk 3. Polymeer gebonden jodium(III) reagentia in selectieve oxidaties. Scope van oxidanten. Dit hoofdstuk gaat over polymer gebonden jodium(III) reagentia. In totaal werden vier verschillende polymeer gebonden jodosylbenzeendialkanoaten (C2, C4, C6 en C8) gesynthetiseerd. Een uitvoerige veiligheidsstudie werd uitgevoerd waaruit bleek dat de lineaire

C4 en C8 oxidanten het meest stabiel en handelbaar waren en respectievelijk 1.71 en 1.29 mmol jodium(III)/g polymeer bevatten.

Deze oxidanten treden op als actieve oxidanten in de directe oxidatie van benzylalcohol tot benzaldehyde zonder tussenkomst van andere additieven. De invloed van de polystyreen hoofdketen werd bestudeerd en het bleek dat een aan lineair polystyreen verankerde jodium(III) oxidant veel beter presteerde dan de gecrosslinkte variant.

Tot slot werd er een groenere alternatieve procedure voor de jodering van polystyreen ontwikkeld zonder gebruik te maken van gechloreerde oplosmiddelen die daar doorgaans voor worden gebruikt.

Hoofdstuk 4. Polymeer gebonden iodium(III) reagentia in selectieve oxidaties. Scope van substraten. Dit hoofdstuk beschrijft een breed scala aan alcohol substraten die onder invloed van polymeer gebonden iodium(III) selectief worden geoxideerd tot carbonylverbindingen. Het belangrijkste was het optimaliseren van de oxidatiemethode. In tegenstelling tot wat eerder was beschreven,

132

Summary bleek het gebruik van een katalytische hoeveelheid anorganisch zout zoals kaliumbromide overbodig om selectief de carbonylverbindingen te verkrijgen.

De diverse polymeer gebonden jodium(III) oxidanten bleken enigszins te verschillen in reactiviteit tot de alcohol substraten. De C4 variant was het meest actief in de oxidatie van hydrofobe en benzylische alcoholen. De C6 variant gaf de hoogste opbrengst aan carbonylproducten van sterolen en gesubstitueerde benzylische alcoholen. Tot slot bleken de C2 en C8 varianten veelbelovend voor de oxidatie van secundaire, alifatische en cyclische alcoholen. Met alle varianten werd een snelle en selectieve oxidatie van geactiveerde allylische alcoholen tot de overeenkomstige aldehyden en ketonen waargenomen. Bovendien verliep de oxidatie van alcoholen met hypervalente jodium(III) oxidanten ook in ionische vloeistoffen als oplosmiddel.

De jodium(III) reagentia bleken ten minste drie keer te kunnen worden teruggewonnen en geregenereerd met een klein activiteitsverlies. Een gepostuleerd mechanisme voor de oxidatie van alcoholen door jodium(III) bestaande uit twee stappen werd ondersteund met een Hammett plot en vastgestelde initiële reactiesnelheden.

Hoofdstuk 5. Lipase gekatalyseerde in situ regeneratie van hypervalente jodium(III) reagens voor selectieve alcohol oxidatie. Hoofstuk 5 gaat over een nieuwe benadering van de in‐situ synthese en gelijktijdig gebruik van jodium(III) oxidanten voor alcohol oxidatie. De in situ vorming van de jodium(III) oxidant door reactie met eveneens in‐situ gevormd peroxycarbonzuur door de lipase gekatalyseerde perhydrolyse reactie, bleek onder bepaalde condities een werkbaar concept.

De beperkingen van de één‐pot synthese van simpele alcoholen worden ook uiteengezet. Voor substraten zoals cyclohexanol en benzylalcohol was de ongewenste verestering de overheersende reactie. Om die reden werd een drie‐staps reactie aangewend om de carbonylverbindingen te verkrijgen waarbij de ongewenste nevenreacties konden worden onderdrukt. Bijkomend voordeel van de stapsgewijze synthese was dat diffusielimitaties uitbleven door het enzym gescheiden te houden van de polymeer gebonden jodium(III) katalysator. Beide reacties hadden een ander temperatuuroptimum; 50 °C voor de peroxidatie en vorming van jodium(III) en 70 °C voor de alcohol oxidatie. Een overmaat waterstofperoxide (2‐4 equivalenten) was hierbij noodzakelijk.

Een belangrijke ontdekking was dat sterolen bij deze reactiecondities door hun sterische hindering niet werden veresterd door CaLB. De oxidatie van de hydroxylgroep door jodium(III)

133

Samenvatting

in aanwezigheid van lipase en een carbonzuur verliep daardoor wel. Dit resulteerde uiteindelijk in een één‐pot reactie voor de oxidatie van sterolen door de H2O2‐carbonzuur‐jodium(III) cascadereactie bij twee temperaturen.

Hoofdstuk 6. Epoxydatie and BaeyerVilligeroxidatie met waterstofperoxide en lipase opgelost in ionische vloeistoffen. Dit hoofdstuk gaat over het gebruik van ionische vloeistoffen als oplosmiddel voor lipase gekatalyseerde reacties. Twee reacties, een epoxidatie en een Baeyer‐Villigeroxidatie, werden uitgevoerd in een chemoenzymatische twee‐staps cascade opzet. Twee geïmmobiliseerde CaLB varianten, Novozym 435 en CaLB CLEA, werden toegepast in cascadereacties in waterstofbrug‐ donerende ionische vloeistoffen. Bij gebruik van Novozym 435 bleek het geadsorbeerde enzym op te lossen in de ionische vloeistof zonder dat dit de reactiesnelheid beïnvloedde. Aangetoond kon worden dat er daadwerkelijk sprake was van homogene enzymkatalyse wat de mogelijkheid geeft tot het uitvoeren van deze reactie in een continue opzet.

Het kon worden aangetoond dat coördinerende waterstofbrug‐donerende ionische vloeistofen de lipase gedreven epoxidatie en Baeyer‐Villigeroxidatie bevorderde. Het effect was tweezijdig. Ten eerste werd de tweede stap in de cascadereactie, te weten de oxidatie van het substraat door het peroxyzuur, aanmerkelijk versneld in deze oplosmiddelen bij milde condities. De omlegging van het Criegee‐intermediair in de Baeyer‐Villigeroxidatie lijkt te worden vergemakkelijkt door geprotoneerde oplosmiddelen. Ten tweede was het lipase opgelost in deze coördinerende waterstofbrug‐donerende ionische vloeistof en zeer stabiel. Dit staat de mogelijkheid toe tot hergebruik van het enzym‐ionische vloeistof systeem als geheel, alsmede een continu procédé.

Wat de Baeyer‐Villigerreactie betreft, konden een breed scala aan substraten worden geoxideerd. Met name cyclische ketonen en alifatische ketonen en aldehyden waren gevoelig voor oxidatie. Het feit dat verschillende ionische vloeistoffen verschillende resultaten gaven, bevestigde dat hun eigenschappen ontwerpbaar zijn. Voor elk substraat kan een geschikte ionische vloeistof worden ontworpen voor het beste resultaat. De toepassing van een nieuwe generatie waterstofbrug‐donerende ionische vloeistoffen is dus een stap voorwaarts in de zoektocht voor alternatieve milieuvriendelijke oplossingen op het gebied van biotransformaties.

134

Acknowledgements

Acknowledgements For the first time I arrived to Delft for a short weekend visit in autumn 2002. I was truly charmed with a beauty of this old town and back then I did not even realize that I would return 1.5 year later to stay for much longer than a weekend. The PhD time was for me not only an amazing adventure ‐ it turned out to be the period of shaping my personality, gaining new experience and meeting many important, interesting and most of all helpful people. Therefore I would like to write couple of words to express my gratitude to everyone thate mad it happened.

First of all I would like to acknowledge my promoter Roger Sheldon. Roger thank you a lot for your sharp eye, humor, story about “Catharina Bloemen”, constructive criticism and the opportunity to carry on iodine research. Additionally I would like to thank you for changing my life and introducing to the company world.

Isabel Arends, the woman in science, I would like to acknowledge for your enthusiastic supervision, support and trust. Isabel, thank you for all valuable discussions, suggestions, feedback and most of all for being more than a supervisor to me.

Fred van Rantwijk, I am grateful for your always open office door, introduction to ionic liquids‐ lipase world and the time spent on fruitful discussions.

Mieke vander Kooij, next to my promoters you were always helping to solve all my important and unimportant PhD student problems. Mieke thank you a lot for you never‐ending support.

Mieke Jacobs, for hacking the door to get my proms send out.

Leen Maat, I would like to thank you very much for a large number of students you have arrange for me – even when none were available. These young people not only extended the number of my hands but also taught me a difficult art of supervision. Thank you for your kindness, openness, “Papirus”, pancake‐ and Chinese‐lunches.

Kristina and Joop, you have played an important role from NMR analyses point of view. Kristina thank you for help with all the NMR.

Ulf and Herman I would like to thank you for feedback while reviewing the draft of this thesis.

Ben Norder, for explosively exciting Thermogravimetric Analyses.

135

Acknowledgements

Delia van Rij, for the elemental analyses.

Valerie Butslelaar, for the microscopic pictures.

Prof. Martens, who greatly contributed to the status quo of me as a chemist. I am remaining grateful for giving me the opportunity to work in your team in Oldenburg as IAESTE exchange student.

Our industrial partners “Wizards from Oss”: Martin Ostendorf, Marcel Schreuder Goedheijt and Gerjan Kemperman thank you for all the support, discussion during our meetings. I would like to thank Martin for his feedback while reviewing the draft of this thesis.

Otto, I would like to thank you for help in the calculations on the iodination degree. Your experience in the polymer science is irreplaceable.

Lola, thank you for our fruitful collaboration, which resulted in two publications. I was happy to exchange my experience on ionic liquids and gain yours on super critical CO2 and phase equilibrium measurements.

My students: Petra, Marlieke, Marina, Marta, David, Helena and Dick I owe you a huge hartelijk dank, muchos gracias, grazie mille. With you I discovered my team player and leadership skills. I would like to thank you for your fantastic work and the contribution to my Dutch, Spanish, Catalan and Italian. I wish you all the best in your scientific and life adventure.

John, I would like to thank for your curiosity in all of the projects within our BOC club. You always are open to pass ideas.

Mike, I would like to thank you for a lot of things without mentioning all of them one by one ;‐) Your scientific support, chats about life and above all your true friendship were and still are very important to me.

My dear aquarium mates: Anne, Menno, Dani (Bokkito), Sander, Remi, H‐P and Hai‐Jin, Myriam ‐ we eare th survivors! I have to admit that I spent a part of my scientific life in unusual office having a lot of fun. Thank you for every minute. Anne, Menno, I am sorry that to make our/my stay there enjoyable I had to reorganized the entire office without asking you :‐P

Anne, thank you for your patient supervision during my first year and friendship you gave me. Menno, thank you for male‐fish‐Buggi although he almost became a mother. I enjoy very much work with you in CLEA. Thank you for sharing cleaning obsessions and understanding obsession.

136

Acknowledgements

Pedro, thank you for all the chats about horoscope, stars and for male‐fish‐Szara – He is still a male. I know that you will remember my SUPER and very catalytic side.

CLEA guys: Roger, Menno, Mike, Ivo ‐ guys thank you for this company experience

Maria and Mapi ‐ girls thank you for the Spanish and Valentia flavors you introduced. Daniela – thank you for “La Traviata” experience. Marco Nicastro, thank you for “medium ‐ steak“ in Italian style. Boesekenzaal–coffee table community: Andrea, Hilda, Elsa, Silvia, Paloma, Luuk, Hacking, van Vliet, Bruno, Helene, Isabel Coreia, Toni, Paco, Paolo, Scoob, Lars Veum, Ton, Chretien, Caroline, Lola, Angel, Eliane, Luigi, Fabien, Linda, Loesje, Inga, Ksenia, Aida, Joel, Tsune, Atsushi, Lara, Selvedin, Marco Casola, Monica Antonio, Jeroen, Aurelie, Sander and Dani for the help while cleaning the aquaria, Adrie, Lars van der Mee, Remco for calling me “moeder van de chemie”, showing how things work, and Marteen for data retrieve, Nunspeet meeting IBOS‐mates: Robert and Gerard. And all I did not mention you are all present in my memories, photos and stories.

PhD time wouldn’t be such a fun without people that were always around me sharing all up and downs on the way. A special thanks goes to my outside work friends: Sonia&Daniel for many evenings, Sundays afternoon, renovations, parties, Porto, Port and more.

Sonia, for all female chat and thank you that I can always count on you.

Agata&Erik for pierogi‐party, faworki‐party, game‐party and a lot of fun and even more drinks, Seda&Suha for the Turkish kitchen specialties, Mike, Dani‐ the Spanish football supporter, Plamen, Wilco, Sanna, Paolo&Cerane, Zeynep&Cem, Hayley&Egor, Telma&Casimiro, Chris, Maruti, Alwin&Yvette, Steven&Rene, Gosia&Marcus for delicious cuisine, party events, for being around and more that is hard to list here; Gerard&Betsy for great sailing times on the Foxy Lady ‐ I hope one day beside enjoying the sun I could, get more involved in sailing. Dear neighbors: Evelin&Jeroen with Tijl, Ria&Jan with Cindy, Lieske&Natalio with kids ‐ thank you for a nice time together and taking us to the neighborhood.

There is also a place for the polish colleagues: Gośka, Leszek, Zbyszek, Andrzej, Sylwia ‐ wielkie dzięki. Tysiące podziękowań kieruje do moich przyjaciół z dawnych studenckich czasów: Asia i Waldek, Gosi i Kubie, Krzyśkowi z Basią, Maćkowi z Kasią, Ani z Manim, Ani (Pelegrini), Ani S. za pamięć, szalone wypady, wesela, wakacje i miejsce w waszych domach. Emi, Bali, Oti, Pavel for all crazy meetings in Praha, Gyor and Delft.

Wujkowi Włodkowi, dzięki Tobie moja przygoda z chemią na serio sie zaczęła. Dziękuję.

137

Acknowledgements

I would like to also say special thank you to my whole big family that was always supporting me during my entire life. Kochani – Rodzice: mamusiu Emilio i tatusiu Jerzy, Justynko z Tomkiem, Krzysiu z Arturem, Juleńko, Mikusiu, Aniu, Gabi ‐ chciałabym wam wszystkim serdecznie podziękować za wszystko co dla mnie zrobiliście. Bez was wiele rzeczy w całym moim życiu nie byłoby możliwe. Chciałabym również serdecznie podziękować całej rodzinie mojego Arka – Rodzicom oraz Madgdzie z Damianem. Dziekuje, za wiarę we mnie i miejsce w nowym domu.

Kochanym babciom: Basi, Gieni, Zosi oraz kochanym dziadkom: Tadkowi i Henrykowi. Wszyscy byliście od samego początku doskonałą odskocznią od doktoratu i niezastąpioną podporą.

My four amazing sisters: Justyna, my twin‐Krzysia, Ania, Gabi thank you a lot for giving me a lot of support despite the distance we live now. Siostrzyczki, dzięki wielkie za to ze jesteście najlepszymi siostrami na świecie.

Last but not least, I would like to thank my beloved husband, my Arek, my best friend, my crazy photographer, my explosive expert, who is the motivation and the inspiration in my journey through the life. You are the reason for everything I am doing. We make a great team that can accomplish all our projects. And you know me ‐ I do love to make plans with you.

Dziękuję

Aleksandra

138

Curiculum vitae

Curiculum vitae Aleksandra Joanna Kotlewska ‐ Miernowska was born on August 22nd, 1977 in Ciechanów, Poland. She was raised in Ostrowiec Świętokrzyski, where she obtained her mature exam in 1996. The same year she started studies at the Technical University of Radom at the Material Science department. One year later she moved to Warsaw, where she continued the Chemical Technology studiese at th Warsaw University of Technology. In 2001 as a member of the International Association for the Exchange of Students for Technical Experience (IAESTE) she joined the Physics at Solid State department at the Chemnitz University of Technology in Germany. In 2002 she was working at Chemistry department as IAESTE exchange student at Carl‐von Ossietzky University in Oldenburg, Germany. In 2002 she graduated her Master of Science and Engineer diploma with specialization in homogeneous catalysis and metalorganic chemistry. In 2004 she started her PhD work in the group of prof. Roger Sheldon at the department of the Biocatalysis and Organic Chemistry at Delft University of Technology in The Netherlands. Her assignment was to investigate Catalytic cascade reactions for selective alcohol oxidation. The results of this research are described in this thesis. In 2008 she started to work for CLEA Technology in Delft as a young researcher. From 2008 she is working in the R&D department at Hexion Specialty Chemicals in Europort (Rotterdam).

139