Document [original]

Immobilization of Catalysts for the

Asymmetric Transfer Hydrogenation of Aryl

Ketones

vorgelegt von

Diplom-Ingenieur

Jonas Dimroth

aus Berlin

Von der Fakultät II – Mathematik und Naturwissenschaften

der Technischen Universität Berlin

zur Erlangung des akademischen Grades

Doktor der Ingenieurwissenschaften

Dr.-Ing.

genehmigte Dissertation

Promotionsausschuss:

Vorsitzende: Prof. Dr. rer. nat. Karola Rück-Braun

Berichter: Prof. Dr. rer. nat. Reinhard Schomäcker

Berichter: Prof. Dr. rer. nat. Rainer Haag

Tag der wissenschaftlichen Aussprache: 18.8.2011

Berlin 2011

D 83

Acknowledgements

Many people helped me directly or indirectly in completing this work. I am grateful and

indebted to all of them. First of all, I would like to thank Prof. Dr. Schomäcker for the

supervision of this thesis, his great support and commitment, and for the many stimulating

discussions we have had. I am also grateful to Prof. Dr. Haag for acting as a referee and for

his encouraging and valuable support during our collaboration.

I am most deeply indebted to Dr. Uwe Schedler for giving me this opportunity and for his

initiative, participation, and belief in this project. This thesis would not have been possible

without him. I would also like to thank the whole team from PolyAn GmbH, especially

Dr. Thomas Thiele for his permanent support and encouragement as well as Dr. Heike

Matuschewski, Markus Heinrich, and Dr. Christian Heise for their help on various occasions.

For his patience, the expert scientific support, and our helpful discussions I am very thankful

to Prof. Dr. Wessig. I also wish to acknowledge the inspiration from Prof. Dr. Wandrey, who

called my attention to ATH reactions.

Sincere thanks are given to Dr. Juliane Keilitz for the pleasant, motivating, and very fruitful

collaboration as well as to all members of the TC 8 research group for the congenial atmosphere

and for the help I received so many times! For assistance with measurements, the introduction

into the handling of analytical equipment, and other kinds of help, I would like to give special

thanks to Astrid Müller-Klauke, Anke Rost, Benjamin Beck, Christa Löhr, Gabi Vetter, Johnny

Nachtigall, Dr. Kathrin Schneider, Le Anh Thu Nguyen, Dr. Maria Schlangen, Dr. Michael

Schwarze, Dr. Oliver Schwarz, Riny Yolandha Parapat, Dr. Robert Frau, Dr. Sebastian Arndt,

Dr. Silvia Czapla, Thorsten Otremba, and Xiao Xie. The assistance of Steffen Schrettl, Ronny

Scherer, Frank Liebau, Claudia Jung, and Michael Stolarski is also gratefully acknowledged.

Furthermore, I would like to express my gratitude to the Berlin Senate State Department of

Science, Research, and Culture for providing an Elsa-Neumann-Scholarship.

I thank my parents for supporting me in all my endeavors. And I thank Jae-Yun Lee for being

the best companion in life.

Abstract

Asymmetric transfer hydrogenation (ATH) is an operationally simple, safe, and environmen-

tally benign catalytic method for the generation of chiral compounds. The catalysts used are

generally transition metal complexes based on chiral ligands and either rhodium, ruthenium, or

iridium. The high purity requirements on chiral substances, especially when they are used as

intermediates in the production of pharmaceuticals, as well as the high catalyst costs make a

complete catalyst separation essential and the reuse of the catalysts desirable. Thus, in order

to increase the overall efficiency of ATH processes, in this thesis quantitatively separable and

highly reusable catalysts were established via immobilization. New strategies were developed to

synthesize linker-containing rhodium and ruthenium complexes, which could then be attached

to surface-functionalized polymer supports. Functional heterogeneous ATH catalysts were gen-

erated from a modified version of a rhodium(III)-p-toluenesulfonyl-1,2-diphenylethylenediamine

complex with tethered cyclopentadienyl unit immobilized on polymer chips and polymer beads.

The supported catalysts were applied in the asymmetric transfer hydrogenation of aryl ketones

in an aqueous solution of sodium formate, and excellent enantioselectivity and reusability were

achieved. In determining the optimal reaction conditions, the pH of the solution was found to

play a particularly decisive role in determining the activity and reusability of the catalysts, and

a significant improvement was achieved when the reaction was run in an acidic medium. The

results of kinetic experiments indicated that a second-order model describes the enantioselective

conversion of acetophenone to phenylethanol under both basic and acidic conditions. However,

significantly different rates and activation parameters suggested different mechanisms; acid

may accelerate the reaction by changing the mode of the proton transfer. The catalytic system

proved very simple to use and robust, and through upscaling of the reaction it was shown that

there is a high potential for technical application to the ecologically and economically rational

production of enantioenriched building blocks.

Zusammenfassung

Mithilfe der asymmetrischen Transferhydrierung (ATH) lassen sich chirale Substanzen kataly-

tisch auf einfachem, sicherem und umweltfreundlichem Wege herstellen. Übergangsmetallkom-

plexe bestehend aus Rhodium, Ruthenium oder Iridium und einem chiralen Liganden werden

dabei in den meisten Fällen als Katalysatoren eingesetzt. Die hohen Reinheitsanforderungen an

die entstehenden Produkte, insbesondere wenn sie als Zwischenstufen für die Produktion von

Pharmaka Verwendung finden, sowie die hohen Kosten für die Katalysatoren selbst erfordern

deren vollständige Abtrennung aus dem Reaktionsgemisch und machen eine Mehrfachver-

wendung wünschenswert. Mit dem Ziel einer Effizienzsteigerung der ATH sollten in dieser

Arbeit quantitativ abtrennbare sowie vielfach wiederverwendbare Katalysatoren durch Immo-

bilisierung hergestellt werden. Zu diesem Zweck wurden Rhodium- und Rutheniumkomplexe

mit funktionellen Gruppen versehen, über die eine Anbindung an oberflächenfunktionalisierte

Polymerträger möglich war. Funktionsfähige ATH-Katalysatoren konnten so aus einem mo-

difizierten Rhodium(III)-p-toluensulfonyl-1,2-diphenylethylendiamin-Komplex mit fixierter η5-

Tetramethylcyclopentadienyl-Einheit, der an Polymerchips und Polymerpartikeln immobilisiert

wurde, gewonnen werden. Die geträgerten Katalysatoren wurden in der asymmetrischen Trans-

ferhydrierung von Arylketonen in wässriger Natriumformiatlösung eingesetzt und wiesen eine

hohe Enantioselektivität sowie die Möglichkeit zur Wiederverwendung auf. Das Katalysatorsys-

tem wurde eingehend im Hinblick auf optimale Reaktionsbedingungen untersucht, wobei ins-

besondere der pH-Wert als entscheidender Faktor für katalytische Aktivität und Wiederverwend-

barkeit identifiziert wurde; eine deutliche Verbesserung der Leistungsfähigkeit wurde durch die

Wahl eines sauren Reaktionsmediums erzielt. Kinetische Untersuchungen zeigten, dass die enan-

tioselektive Reaktion von Acetophenon zu Phenylethanol sowohl unter basischen als auch unter

sauren Bedingungen durch ein einfaches Modell zweiter Ordnung beschrieben werden kann.

Der in Gegenwart von Säure deutliche Anstieg der Reaktionsgeschwindigkeit bzw. die Änderung

der Aktivierungsparameter ohne Verlust an Enantioselektivität könnte jedoch auf einen verän-

derten Mechanismus hindeuten. Das Katalysatorsystem erwies sich bei einfacher Handhabung

als effizient und robust. Durch Upscaling-Experimente konnte zudem gezeigt werden, dass es

großes Potenzial für die technische Anwendung in einer ökologisch und ökonomisch effizienten

Produktion enantiomerenreiner Komponenten birgt.

Contents

1 Introduction 1

2 Fundamentals and State of the Art 3

2.1 Chirality and Chiral Technology . . . . . . . . . . . . . . . . . . . . . . . . . . 3

2.2 IndustrialRequirements .............................. 4

2.2.1 Catalyst Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

2.2.2 Product Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

2.2.3 Overall Process Requirements . . . . . . . . . . . . . . . . . . . . . . . 7

2.3 Enantioselective Catalytic Hydrogenation of Ketones . . . . . . . . . . . . . . . 9

2.3.1 Asymmetric Hydrogenation . . . . . . . . . . . . . . . . . . . . . . . . 9

2.3.2 Asymmetric Hydroboration . . . . . . . . . . . . . . . . . . . . . . . . 10

2.3.3 Asymmetric Hydrosilylation . . . . . . . . . . . . . . . . . . . . . . . . 10

2.3.4 Asymmetric Biocatalytic Reduction . . . . . . . . . . . . . . . . . . . . 11

2.4 Asymmetric Transfer Hydrogenation . . . . . . . . . . . . . . . . . . . . . . . 11

2.4.1 Historical Background . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

2.4.2 Hydrogen Donor Systems . . . . . . . . . . . . . . . . . . . . . . . . . 12

2.4.3 Catalysts.................................. 13

2.4.4 MechanisticAspects............................ 16

2.4.5 Applications and Industrial Impact . . . . . . . . . . . . . . . . . . . . 20

2.4.6 Immobilization............................... 22

3 Support Materials Applied 27

3.1 Material Choice and Preparation . . . . . . . . . . . . . . . . . . . . . . . . . 27

3.1.1 Molecular Surface Engineering . . . . . . . . . . . . . . . . . . . . . . 28

3.1.2 MaterialProperties ............................ 28

4 Modification of a Ruthenium Catalyst 31

4.1 Background .................................... 31

4.1.1 Homogeneous Applications . . . . . . . . . . . . . . . . . . . . . . . . 31

4.1.2 Immobilization Approaches . . . . . . . . . . . . . . . . . . . . . . . . 32

4.2 InitialStudies ................................... 33

VII

Contents

4.2.1 Preparation and Use of Diphosphine/Diamine-based Catalysts . . . . . . 33

4.2.2 Strategies for Modification . . . . . . . . . . . . . . . . . . . . . . . . 35

4.3 Preparation and Use of the Modified PNNP . . . . . . . . . . . . . . . . . . . 36

4.3.1 SyntheticPathway............................. 36

4.3.2 Attempts at Immobilization . . . . . . . . . . . . . . . . . . . . . . . . 38

4.3.3 ConcludingRemarks............................ 40

5 Immobilization of a Rhodium Catalyst 41

5.1 SyntheticModification............................... 41

5.1.1 LigandPreparation ............................ 41

5.1.2 Catalyst Formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

5.2 CatalyticTesting.................................. 45

5.2.1 GeneralRemarks.............................. 45

5.2.2 InitialExperiments............................. 47

5.2.3 Effect of Temperature and Atmosphere . . . . . . . . . . . . . . . . . . 50

5.2.4 pH Dependency of Activity, Enantioselectivity, and Reusability . . . . . 52

5.2.5 Effect of the Concentrations of Acetophenone and Sodium Formate . . 55

5.3 Kinetic and Mechanistic Investigations . . . . . . . . . . . . . . . . . . . . . . 56

5.3.1 Basic Reaction Conditions . . . . . . . . . . . . . . . . . . . . . . . . . 57

5.3.2 Acidic Reaction Conditions . . . . . . . . . . . . . . . . . . . . . . . . 57

5.3.3 Mechanistic Considerations . . . . . . . . . . . . . . . . . . . . . . . . 60

5.4 PracticalAspects ................................. 65

5.4.1 Application-oriented Experiments . . . . . . . . . . . . . . . . . . . . . 66

5.4.2 Discussion on the Efficiency . . . . . . . . . . . . . . . . . . . . . . . . 67

6 General Conclusion and Outlook 71

7 Experimental 74

7.1 Equipment..................................... 75

7.1.1 Instrumentation .............................. 75

7.1.2 Laboratory Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . 76

7.2 SynthesisProcedures ............................... 76

7.2.1 Synthesis of Ru-PNNP . . . . . . . . . . . . . . . . . . . . . . . . . . 76

7.2.2 Synthesis of tethered Rh(III)-catalyst . . . . . . . . . . . . . . . . . . . 86

7.3 Characterization of Supports and Coupling . . . . . . . . . . . . . . . . . . . . 88

7.3.1 Determination of Amine Loading . . . . . . . . . . . . . . . . . . . . . 88

7.3.2 Immobilization and Determination of Catalyst Loading . . . . . . . . . 89

7.4 CatalysisExperiments............................... 90

VIII

Contents

7.4.1 Homogeneous Catalysts . . . . . . . . . . . . . . . . . . . . . . . . . . 90

7.4.2 Heterogenized Catalysts . . . . . . . . . . . . . . . . . . . . . . . . . . 90

Bibliography 92

Appendix 102

List of Figures

2.1 TheBINAPligand................................. 9

2.2 Ligands successfully used in ATH operations . . . . . . . . . . . . . . . . . . . 15

2.3 Intermediates produced via ATH in scale-up studies . . . . . . . . . . . . . . . 21

3.1 Support materials applied . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

3.2 Schematic illustration of the support materials . . . . . . . . . . . . . . . . . . 29

3.3 Loading with amino groups vs. loading with rhodium . . . . . . . . . . . . . . 30

4.1 Diphosphine/diamine ligand and ruthenium complex . . . . . . . . . . . . . . . 32

4.2 Immobilized ruthenium-PNNP complexes . . . . . . . . . . . . . . . . . . . . . 33

5.1 Wills´s tethered Rh-TsDPEN catalyst . . . . . . . . . . . . . . . . . . . . . . . 41

5.2 SupportedCatalysts................................ 44

5.3 IR spectra as proof of coupling . . . . . . . . . . . . . . . . . . . . . . . . . . 44

5.4 Use of supported catalysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

5.5 Reproducibility of the standard experiment using the PB-supported catalyst . . 48

5.6 Variation of temperature and atmosphere within the flask . . . . . . . . . . . . 51

5.7 Multiple catalyst reuse under different pH conditions . . . . . . . . . . . . . . . 54

5.8 Impact of the concentrations of substrate and hydrogen donor . . . . . . . . . 56

5.9 Influence of the concentration of acetophenone on the suspension of beads . . . 56

5.10 ATH of acp under moderately basic conditions . . . . . . . . . . . . . . . . . . 58

5.11 Eyring plot for basic conditions . . . . . . . . . . . . . . . . . . . . . . . . . . 58

5.12 ATH of acp under acidic conditions . . . . . . . . . . . . . . . . . . . . . . . . 59

5.13 Eyring plot for acidic conditions . . . . . . . . . . . . . . . . . . . . . . . . . . 59

5.14 Ru-TsDPEN species under different conditions . . . . . . . . . . . . . . . . . . 60

5.15 Activity and enantioselectivity over pH . . . . . . . . . . . . . . . . . . . . . . 62

5.16 Alternative transition states under acidic conditions . . . . . . . . . . . . . . . 63

5.17 Technically relevant tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66

List of Schemes

2.1 The CBS reduction reaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

2.2 The MPV reduction and Oppenauer oxidation . . . . . . . . . . . . . . . . . . 12

2.3 Innerspheremechanism.............................. 17

2.4 Concerted outer sphere mechanism . . . . . . . . . . . . . . . . . . . . . . . . 18

2.5 Proposed outer sphere mechanism in aqueous media . . . . . . . . . . . . . . . 19

2.6 Covalent attachment of a TsDPEN derivative to polystyrene supports . . . . . . 24

2.7 Crosspolymerization of a TsDPEN-based ligand monomer . . . . . . . . . . . . 25

4.1 ATH of acetophenone as the benchmark reaction . . . . . . . . . . . . . . . . 32

4.2 Preparation of the unmodified PNNP ligand and ruthenium complex . . . . . . 34

4.3 Preparation of key intermediate 7 . . . . . . . . . . . . . . . . . . . . . . . . . 37

4.4 Preparation of key intermediate 12 . . . . . . . . . . . . . . . . . . . . . . . . 37

4.5 Preparation of the modified Ru-PNNP complex . . . . . . . . . . . . . . . . . 38

4.6 Coupling of modified Ru-PNNP . . . . . . . . . . . . . . . . . . . . . . . . . . 39

4.7 Coupling of modified PNNP . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

5.1 Introduction of hex-5-ynoic acid-tert-butylester — first attempt . . . . . . . . . 42

5.2 Introduction of hex-5-ynoic acid-tert-butylester — second attempt . . . . . . . . 42

5.3 Introduction of methyl adipoyl chloride . . . . . . . . . . . . . . . . . . . . . . 43

5.4 Preparation of the supported modified tethered Rh-TsDPEN catalyst . . . . . . 43

5.5 ATH of aryl ketones in an aqueous medium . . . . . . . . . . . . . . . . . . . 46

5.6 ATH of aryl ketones in formate/water . . . . . . . . . . . . . . . . . . . . . . 55

5.7 Proposed mechanism under different pH conditions . . . . . . . . . . . . . . . 61

List of Tables

2.1 Principles of green chemistry and green engineering . . . . . . . . . . . . . . . 8

2.2 Commonly used hydrogen donor systems . . . . . . . . . . . . . . . . . . . . . 14

4.1 PerformanceofIr-PNNP ............................. 35

4.2 Attempts at immobilization to aminated supports . . . . . . . . . . . . . . . . 40

5.1 Performance of different Rh-TsDPEN variants . . . . . . . . . . . . . . . . . . 47

5.2 Recycling of PC-supported catalyst without intermediate washing . . . . . . . . 49

5.3 Recycling of PC-supported catalyst with intermediate washing . . . . . . . . . . 49

5.4 Scope of tested substrates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

5.5 ATH of acetophenone under different pH conditions . . . . . . . . . . . . . . . 53

XII

Abbreviations and Symbols

acp acetophenone

AH asymmetric hydrogenation

aq aqua

ATH asymmetric transfer hydrogenation

ATR attenuated total reflection

BINAP 2,20-bis(diarylphosphino)-1,10-binaphthyl

Boc tert-butyloxycarbonyl

cat catalyst

CBS Corey–Bakshi–Shibata

CDI 1,10-carbonyldiimidazole

COD cyclooctadiene

Cp cyclopentadienyl

DAD diode array detector

DCC N,N0-dicyclohexylcarbodiimide

DCM dichlormethane

DFT density functional theory

DIC N,N0-diisopropylcarbodiimide

DIPEA diisopropylethylamine

DMAP 4-(dimethylamino)pyridine

DMF dimethylformamid

DMSO dimethyl sulfoxide

DPEN 1,2-diphenylethylenediamine

DSC differential scanning calorimetry

ee enantiomeric excess

EI electron ionization

ESI electron spray ionization

FID flame ionization detector

Fmoc 9-fluorenylmethyloxycarbonyl

FT-IR fourier transform infrared spectrophotometer

GC gas chromatography

XIII

Abbreviations and Symbols

hPlank constant [6.626·10−34 J s]

HOAt 1-hydroxy-7-azabenzotriazole

HOBt hydroxybenzotriazole

HPLC high performance liquid chromatography

HRMS high-resolution mass spectrometry

ICP inductively coupled plasma

IPA 2-propanol

IR infrared

Jcoupling constant [Hz]

krate constant

kBBoltzmann’s constant [1.381·10−23 J K−1]

MPV Meerwein–Ponndorf–Verly

MS mass spectrometry

namount of substance [mol]

NHS N-hydroxysuccinimid

NMR nuclear magnetic resonance

OES optical emission spectroscopy

PB polymer bead

PC polymer chip

PDMS polydimethylsiloxane

PE polyethylene

PEG polyethylene glycol

PNNP N,N0-bis[o-(diphenylphosphino)-benzyl]cyclohexane-1,2-diamine

PP polypropylene

ppm parts per million

PPNCl bis(triphenylphosphoranylidene) ammonium chloride

PS polystyrene

Runiversal gas constant [8.3145 J mol−1K−1]

rpm revolutions per minute

rt room temperature

Sselectivity

S/C substrate to catalyst ratio

SDS sodium dodecyl sulfate

STY space time yield [g L−1h−1]

Tabsolute temperature [K]

TBTU o-(benzotriazol-1-yl)-N,N,N0,N0-tetramethyluronium tetrafluoroborate

TEAF triethylamine/formic acid

XIV

Abbreviations and Symbols

tert tertiary

TFA trifluoroacetic acid

TH transfer hydrogenation

THF tetrahydrofuran

TLC thin layer chromatography

TOF turnover frequency [h−1]

TON turnover number

TS transition state

TsDACH N-(p-toluenesulfonyl)-1,2-diaminocyclohexane

TsDPEN N-(p-toluenesulfonyl)-1,2-diphenylethylenediamine

TTN total turnover number

UV-Vis ultraviolet-visible spectral region

Yyield

Greek Characters

δchemical shift [ppm]

∆G‡Gibb´s energy of activation [kJ mol−1]

∆H‡enthalpy of activation [kJ mol−1]

∆S‡entropy of activation [J mol−1K−1]

λwavelength [nm]

νwavenumber [cm−1]

Φphenyl

1 Introduction

The growing demand for enantiopure chiral compounds, which play an important role in the

fabrication of pharmaceuticals [1], is accompanied by increasing acknowledgement of the “need

for more environmentally acceptable processes in the chemical industry” [2]. Maximal efficiency

and minimal hazard are the criteria for the design of new processes to meet the complex needs

of society and the economy. Generally, catalytic production is favorable over non-catalytic since

catalysts are capable of making chemical reactions more energy-efficient; this is the reason

why catalysis-based manufacturing accounts for about 85 % of all chemical production [3]. An

additional benefit is that the use of asymmetric catalysts in the fabrication of chiral substances

leads to the multiplication of the precious chiral information. For current and future asymmetric

production processes, therefore, the optimization of catalytic methods is of great importance.

Of the several asymmetric catalytic methods suitable for the enantioselective reduction of

prochiral substrates, the asymmetric transfer hydrogenation (ATH) of imines and ketones is

considered particularly useful because the reaction proceeds under mild conditions without the

need for sophisticated equipment and hazardous reactants. Furthermore, intensive investigation

in the last three decades has afforded a broad knowledge about ATH reaction mechanisms, the

use of hydrogen donors, and the development of highly active and enantioselective catalysts.

These catalysts are usually homogeneously applied complexes made of chiral organic ligands and

transition metals; this is a drawback since both components are expensive and many transition

metals are toxic. Thus, for industrial applications a complete removal of the catalyst from the

reaction mixture is required to guarantee metal-free products, and the reuse of the catalyst is

highly desired to reduce costs and improve the environmental impact of the process.

Heterogeneous catalysis suggests an approach to addressing this problem. Heterogeneous

catalysts are separated from the reactants by virtue of their physical state, thus facilitated

product separation and multiple use of each catalytic site over an extended lifetime are pos-

sible. With regard to other important issues, however, homogeneous catalysts are favorable.

They are generally more active and more selective, reaction conditions are usually milder, mass

transfer and diffusion problems are less severe, modification of electronic and steric properties

is possible, and mechanistic investigation is easier [4]. To combine the best features of both

heterogeneous and homogeneous catalysis, researchers often use catalyst immobilization. Of

the various methods of catalyst immobilization, heterogenization via covalent attachment might

be the most often applied. Nevertheless, making a homogeneous catalyst heterogeneous usu-

1 Introduction

ally requires an elaborate modification of the ligand as well as a suitable support and efficient

coupling, and thus increases the catalyst cost contribution to the overall process. Additionally,

in many cases activity of the the catalyst decreases upon immobilization. Hence, the potential

advantages of an immobilized system will not be utilized as long as they do not compensate for

the higher costs. Since industrial processes based on immobilized asymmetric catalysts remain

scarce [5], it seems that this condition has not yet been achieved. Both academia and industry

are therefore called on to search for new routes to efficient immobilized catalytic systems, and

to provide a fundamental understanding of how these systems work.

It was thus the objective of this thesis to generate new heterogenized catalysts and elucidate

their strengths and weaknesses with regard to a potential application in ATH processes on

laboratory or industrial scales. The aim was to develop an easy to use and efficently recyclable

catalytic system and to achieve a comprehensive understanding of how the reaction conditions

determine catalyst performance. To efficiently achieve these aims, a joint project was established

bringing together expertise from academia and industry. Major parts of the experimental work,

in particular all activities related to surface-functionalization, were done at the laboratories of

PolyAn GmbH, Berlin. Analytics and performance tests were carried out at the Institute of

Chemistry at the Technische Universität Berlin. Substantial support was received from the

Institute of Chemistry and Biochemistry of the Freie Universität Berlin.

This thesis begins with an overview of the basic theoretical aspects and the fundamental de-

velopments in the field of ATH, followed by a description of the approaches used and the results

obtained. First, I briefly describe the support materials which were applied. The modification,

coupling, and experimental application of the transition metal complexes are then discussed

separately for each approach. Finally, the major results and conclusions are reconsidered in a

general discussion. Experimental details are given at the end of the dissertation.

2 Fundamentals and State of the Art

2.1 Chirality and Chiral Technology

Chirality (from the ancient Greek word

 qeÐr

, hand) or handedness is a phenomenon of fun-

damental importance in biology and thus in all life science-related technologies. An object is

denoted chiral when it has a non-superimposable mirror image, just like a human hand. A pair

of mirror image molecules is called a pair of enantiomers (

ânantÐoc

, opposite), with one having

an R-configuration and the other an S-configuration according to the Cahn-Ingold-Prelog rules.

Usually, three forms of chirality are distinguished: central (point), axial, and planar chirality

[6]. Most important in the context of this thesis is central chirality, which is present when four

different residues are attached to a carbon atom.

Different enantiomers of a given molecule can exhibit dramatically different biological ac-

tivities. For example, the S,S-form of ethambutol is used as a tuberculostatic whereas the

R,R-form may cause blindness [7]. Although the relevance of optically pure drugs had been

known for some time, the chiral switch—the substitution of racemic for enantiopure pharmaco-

logical compounds—did not begin until the 1990s when the U.S. Food and Drug Administration

(FDA) decreed new rules for the development of stereoisomeric drugs [8].

The production of pure enantiomers is a challenging task, since in any (conventional) reac-

tion of achiral starting materials to chiral products these products will be formed as racemates,

i. e., 50:50 mixtures of both enantiomers [9]. Currently, chiral pool methods or racemic resolu-

tions are the approaches most commonly applied for the production of chiral compounds [5, 10].

Thus, either enantiopure building blocks provided by nature are used as starting materials, or

the enantiomers produced are separated via crystallization, kinetic resolution, or chromato-

graphy [11, 12]. The first approach, however, is limited due to a limited number of accessible

compounds which are not necessarily suitable, so that many steps with attendant losses in yield

may be required for the synthesis of the desired product. The latter approach, the formation

and separation of racemic products, has the drawback that 50 % waste is generated.

A generally more atom-economic and potentially more cost-efficient method is thus the use

of asymmetric catalysts, which act as multiplicators of their chiral information. The following

section provides an overview of the requirements that those catalysts have to meet to be ap-

plicable to industrial processes. A list of (academically) successful examples of enantioselective

(chemo)catalytic approaches includes hydrogenation, transfer hydrogenation, hydrosilylation,

2 Fundamentals and State of the Art

and hydroboration of unsaturated compounds, epoxidation of allylic alcohols, vicinal hydroxyla-

tion, hydrovinylation, hydroformylation, cyclopropanation, and isomerization of olefins, propy-

lene polymerization, organometallic addition to aldehydes, allylic alkylation, organic halide-

organometallic coupling, aldol type reactions, and Diels-Alder and ene reactions [13].

2.2 Industrial Requirements

Requirements for catalytic chemical fabrication include overall process requirements,product

requirements, and catalyst requirements, all of which are interdependent to some extent and

thus lead to different points of view for how to deal with analogous problems. Enantiopure

products are often subjected to the special conditions of pharmaceutical fabrication as they

are processed to medicinal products in large part [14]. These conditions include, of course,

the general requirements in chemical production, but are usually somewhat broader and more

stringent.

Unfortunately, the definitions of many terms and key figures (performance indicators) which

are employed to describe and compare catalyst or process properties, vary among the fields

in which they are used. Furthermore, there is no general agreement on the conditions under

which those key figures are to be determined. Thus, a reliable comparison is only possible if

the units are known and the reaction conditions are roughly equal. The following definitions

represent the general understanding in the context of organometallic catalysis with a focus on

asymmetric synthesis. A general overview is given, for instance, by Behr [15], Hagen [16], and

Weitkamp and Gläser [17].

2.2.1 Catalyst Requirements

The applicability of a catalyst in an industrial process depends mainly on three properties:

•activity

•selectivity

•stability.

Although a general statement is difficult to make due to the diverse industrial applications with

their individual requirements on the catalysts, according to Hagen “the target quantities should

be given the following order of priority: selectivity > stability > activity” [18]. Catalyst testing

is mostly performed with simplified benchmark substrates, but for example in the production

of pharmaceuticals, highly functionalized starting materials are often employed. Thus, the

substrate scope and the functional group tolerance are also of great importance [19]. Further

important aspects include, of course, economic issues. Although normally more important in

2 Fundamentals and State of the Art

the fabrication of low-priced chemicals, the catalyst costs can become crucial in asymmetric

production processes where chiral metal complexes consisting of expensive noble metals and

often even more expensive chiral ligands are applied. The contribution of the catalyst to the

total cost, however, has to be discussed in the context of the overall process (see below).

Selectivity The selectivity, SP, of a given reaction with more than one possible product is

calculated by the ratio of the amount of the desired product (P) to the amount of substance

of the starting material (A) which has been converted, taking into account the stoichiometric

coefficients of the reactants (Equation 2.2.1). SPrefers to only one specific product and starting

material, as do the conversion (only starting material) and the yield.

SP=nP|νA|

(nA,0 −nA)|νP|(2.2.1)

The conversion, XA, represents the ratio of converted starting material to employed starting

material (Equation 2.2.2). The yield, YP, is defined as the ratio of the amount of the desired

product actually formed to the theoretical maximum amount (Equation 2.2.3). It is thus the

product of conversion and selectivity.

XA=nA,0 −nA

nA,0

(2.2.2)

YP=nP|νA|

nA,0 |νP|=conversion ·selectivity (2.2.3)

There is, however, more than one selectivity in many processes. In the special case of asym-

metric catalysis, where prochiral starting materials are converted to chiral products, the enan-

tioselectivity is of decisive importance. It is quantified by the ratio of the excess amount of the

major enantiomer, e. g., the R-form, to the sum of the amounts of both enantiomers (Equa-

tion 2.2.4). When no further purification is possible, the enantiomeric excess (ee) provided by

the catalyst should be higher than 99% for pharmaceuticals, but since this is rather seldom,

lower ees(>90 %) are usually acceptable. For agricultural chemicals 80 % ee can be sufficient

[14, 20].

ee [%] = R−S

R+S·100 (2.2.4)

Stability The stability is strongly related to the catalyst lifetime, which is limited due to metal

leaching, poisoning, or other types of deterioration resulting from chemical, thermal, or mechan-

ical stress. Although the turnover number (TON) does not represent a direct quantification

of the stability or lifetime of a catalyst, it implicitly provides information thereof by indicating

how productive the catalyst is under the given reaction conditions. The TON is defined as the

2 Fundamentals and State of the Art

ratio of the amount of converted starting material to the amount of catalyst employed. Thus,

it specifies the number of catalytic cycles run by one catalyst site under defined conditions

(Equation 2.2.5).

TON =converted starting material [mol]

amount of catalyst employed [mol] (2.2.5)

For the cost-efficient fabrication of rather cheap mass-marketed products, a TON of more than

20 000 is usually required. The small-scale production of compounds with higher added value

can be profitable at a TON of 1 000 [21]. In a batchwise process in which a certain amount of

the catalyst is repeatedly used, the TON of each batch can be summed to the total turnover

number (TTN, Equation 2.2.6), which indicates the total productivity of the reused catalyst.

TTN =XTON (2.2.6)

Another indicator often used by researchers is the substrate to catalyst ratio (S/C), i. e., the ratio

of the total amount of substrate employed to the amount of catalyst employed. Information

about the catalyst´s productivity, however, is only provided when the conversion or yield is

additionally quoted (which effectively gives the TON).

Activity The catalyst activity is in most cases quantified by the turnover frequency (TOF),

which is defined as the number of molecular reactions or catalytic cycles per unit time (Equation

2.2.7). For instance in enantioselective hydrogenation processes, the TOF at a conversion of

more than 95 % ought to be above 500 h−1for small-scale and above 10000 h−1for large-scale

productions [14, 20].

TOF [h−1] = converted starting material [mol]

amount of catalyst employed [mol] ·time [h] =TON

time [h] (2.2.7)

2.2.2 Product Requirements

In addition to the requirements regarding the enantiomeric purity (>99 % ee for pharmaceu-

ticals, >80 % ee for agrochemicals), the main problem from the point of view of the product

is the contamination with other compounds. Usually, a product purity of more than 99% with

metal residues below 10 ppm is reqired in the production of pharmaceuticals [5]. Thus, in

addition to a high selectivity, the efficent separation of the catalyst is of decisive importance in

the production of fine chemicals and pharmaceuticals.

2 Fundamentals and State of the Art

2.2.3 Overall Process Requirements

The benefit of an (asymmetric) catalytic step within the production of a target molecule should

be judged in the context of the whole process. Today´s industrial fabrication processes have to

meet the following requirements — which are by no means mutually exclusive [22]:

•cost efficiency

•safety

•environmental friendliness.

Cost Efficiency Ideally, a chemical fabrication process provides a high product yield at low

cost [23]. Thus, yield relative to proportional costs is a viable way of determining the efficiency

of a given reaction. A detailed discussion of economic aspects is outside the scope of this thesis;

with regard to the catalyst, however, there is a rule of thumb saying that its cost contribution

to the total cost price of the product should not be more than 5 % [24]. Generally, the lower the

added value of the compounds produced the more important the catalyst´s performance and

cost as well as process optimization. In contrast, the decisive factors for the development of

compounds with very high added value are “the total synthetic features and time issues together

with intellectual property aspects and risks of constraints” [25].

A useful key factor for the determination of the efficiency of a (catalytic) reaction is the

space time yield (STY). It indirectly indicates a part of the investment costs, since it includes

the volume of the reactor. STY is most commonly defined as the amount (specified as mass) of

the desired product, which is formed within a given time in a given reaction volume (Equation

2.2.8).

STY [g L−1h−1] = desired product [g]

total reaction volume [L] ·time [h] (2.2.8)

Safety and Environmental Issues There is an increasing demand for and interest in the

development of sustainable and inherently safe processes in the chemical industry. Anastas and

co-workers formulated 12 principles of green chemistry [26] and green engineering [27], which

have been widely recognized as a useful guideline for more favorable approaches in chemical

fabrication and process design. Some of these principles are redundant in a way; nevertheless,

they are worth mentioning at least in an abbreviated form (see Table 2.1). Major importance

in the context of “green” production is given to the minimization of waste, that is, all matter

produced which is not part of the product at the end of the chemical process. The degree of

waste prevention is expressed by the atom economy or atom efficiency of a given reaction or

process, which—in addition to yield and selectivity —has become “the third element of the

2 Fundamentals and State of the Art

Table 2.1: Principles of green chemistry and green engineering.

Green chemistry Green engineering

1 Waste prevention Inherent rather than circumstantial

2 Atom economy Prevention instead of treatment

3 Less hazardous syntheses Design for separation

4 Designing safer chemicals Maximize efficiency

5 Safer solvents and auxiliaries Output-pulled versus input-pushed

6 Design for energy efficiency Conserve complexity

7 Use of renewable feedstocks Durability rather than immortality

8 Reduce derivatives Meet need, minimize excess

9 Use of catalysis Minimize material diversity

10 Design for degradation Integrate material and energy flows

11 Analysis for pollution prevention Design for commercial “afterlife”

12 Inherently safer chemistry Renewable rather than depleting

triadic goal that any synthetic chemist should seek” [28]. Several key factors are used to make

the atom economy a comparable value. Arguably the most prevalent is the E-factor (Equation

2.2.9), introduced by Sheldon in 1992 [29].

The optimal value of the E-factor is zero, since it represents the ratio of the mass of the

waste generated to the mass of the desired product. Conceptual limitations of this key measure

become evident when one considers that the definition of waste is anything at the end of the

process that is not the desired product; unconverted starting materials, additives, side products,

solvent losses, and even the mass of fuel corresponding to the required energy can be taken into

account. Water, however, is not considered as waste; only the compounds contained in the water

are counted. Thus, substances are either weighted with zero (water) or with one (all others),

not taking into account the real potential for harm of each individual compound. Other authors

have attempted to overcome these limitations by proposing a more sophisticated weighting of

toxicity and the waste-associated hazards of different types of substances [28]. Nevertheless,

when used in addition to other criteria, the E-factor may be a powerful instrument for the

comparison of different pathways which lead to the same product. It seems particularly useful

when the calculation is simplified to the ratio of the mass of raw materials minus the mass of

the desired product to the mass of the desired product, thus excluding all waste streams which

may be difficult to characterize [30].

E-factor =mass of waste

mass of desired product (2.2.9)

2 Fundamentals and State of the Art

2.3 Enantioselective Catalytic Hydrogenation of Ketones

In addition to asymmetric transfer hydrogenation, there are four important catalytic methods to

enantioselectively reduce (unactivated) prochiral ketones to chiral alcohols: asymmetric hydro-

genation, asymmetric hydroboration, asymmetric hydrosilylation, and asymmetric biocatalytic

reduction. All of these methods will be briefly presented in the following, though asymmetric

transfer hydrogenation is discussed in more detail in Section 2.4.

2.3.1 Asymmetric Hydrogenation

The asymmetric hydrogenation (AH) of ketones based on ruthenium(II)-diphosphine catalysts

established proof of concept during the 1980s and is now an industrially established method

[31]. The advancement of AH is inextricably linked to R. Noyori who was decorated for his

contributions to the field with the Nobel Prize in chemistry in 2001 [32]. Noyori and co-workers

developed the 2,20-bis(diphenylphosphino)-1,10-binaphthyl (BINAP) ligand (Figure 2.1), which

is — together with its derivatives — the basis for the most successful catalysts for AH operations

[33, 34]. However, ruthenium(II)-BINAP complexes were initially only used as catalysts for

the reduction of β-keto esters [35], until further progress extended the applicability to the

hydrogenation of simple unactivated ketones such as acetophenone. It was found that the

activity and productivity of AH reactions using an achiral standard catalyst hRuCl2[P(C6H5)3]3i

were remarkably enhanced when the reaction was carried out in the presence of an alkaline base

and ethylenediamine in 2-propanol. In an enantioselective version of this approach, a chiral

BINAP ligand was combined with a chiral diamine ligand in conjunction with ruthenium(II)

and in the presence of KOH [36]. Generally, high TON and TOF values can be achieved in

the asymmetric hydrogenation of ketones (indeed, TON values of AH catalysts are the highest

reported in the context of asymmetric catalytic reductions of ketones). However, the rate is

highly sensitive to the pressure of hydrogen. For example, hydrogenation at an S/C of 500

at 1 bar gave a TOF of 880 h−1, whereas a TOF of 23 000 h−1was obtained at an S/C of

10 000 at about 50 bar [34]. In a more recent study, base-free conditions were successfully

applied when a modified BINAP-based catalyst, trans-RuH(η1-BH4)(binap)(1,2-diamine), was

used [37].

(a) R-BINAP (b) S-BINAP

Figure 2.1: The BINAP ligand.

2 Fundamentals and State of the Art

2.3.2 Asymmetric Hydroboration

The application of boranes to the enantioselective reduction of carbonyls was first de-

scribed in the late 1960s [38], but yielded only poor results until in 1979 Johnson used β-

hydroxysulfoximine with diborane gas [39], and in 1981 Hirao and co-workers used chiral amino

alcohols with BH3·THF [40] to reduce prochiral ketones to chiral alcohols with optical yields

of up to 82 % and 60 %, respectively. These approaches, however, were not catalytic, but the

latter especially was seminal for later developments in the field. Through further studies by the

group of Itsuno and the mechanistic investigations performed by Corey, Bakshi, and Shibata,

the catalytic properties of oxazaborolidines were discovered [41, 42], and a novel analog known

as the CBS-catalyst (after the inventors´ names) was developed [43]. Scheme 2.1 shows the

original CBS system with BH3·THF as the reducing agent, providing nearly quantitative yield

in the transformation of acetophenone to (R)-1-phenylethanol (via boron enolates, which re-

quire acidic workup) with up to 97 % ee. Further investigations led to a wide range of chiral

oxazaborolidine-based catalysts and reducing agents applicable to the highly enantioselective

reduction of different ketones even on an industrial scale [41]. More recently, borohydrides such

as NaBH4and LiBH4were successfully applied to enantioselective hydroboration, and effective

metal catalysts have been developed [44].

Scheme 2.1: The CBS reduction reaction.

2.3.3 Asymmetric Hydrosilylation

There are several ways to use hydrosilanes for the hydrogenation of unsaturated compounds;

enantioselective hydrosilylation reactions of carbonyl and imino groups, however, are catalyzed

by chiral metal catalysts with only few exceptions [45]. Rhodium especially, but also titanium,

copper, zinc and other metals, are applied in combination with chiral ligands containing most

often nitrogen or phosphorus as donor atoms. Different mechanisms are possible for the H-

transfer, depending on the metal center [46]. In either case, a silyl ether is formed when

carbonyls are applied as substrates and has to be subjected to acidic hydrolysis to yield the

desired chiral alcohol. The silane chosen as the source of hydrogen has a strong impact on

the enantiomeric excess of the products in the process. Although highly active and selective

rhodium and copper-catalyzed asymmetric hydrosilylations have been developed, widespread

industrial application has not yet been established [46].

2 Fundamentals and State of the Art

2.3.4 Asymmetric Biocatalytic Reduction

Biocatalytic approaches are distinguished by their often extremely high (enantio)selectivities,

mild reaction conditions, and harmless reagents. Nonetheless, the general instability of biocat-

alysts — which can be isolated enzymes or whole microorganisms — to thermal and mechanical

stresses and pH variations limits their use. Additionally, development cycles are long, modifica-

tions difficult, and the number of commercially available biocatalysts is rather small [47]. Unlike

several chemo catalysts which have a broad range of application (privileged catalysts), biocat-

alysts are usually effective for only one type of substrate [48]. In enzyme-catalyzed reductions

of carbonyl groups, the hydride, which is transferred to the carbonyl C-atom, is delivered by a

coenzyme; the reduced form of nicotinamide adenine dinucleotide (NADH) or the corresponding

phosphate (NADPH) are commonly applied. Various methods and hydrogen sources are in use

for the in situ regeneration of the coenzymes, which is an important issue because NAD(P)H

is too expensive to be stoichiometrically employed [49]. Early biocatalytic ketone reductions

were performed with baker´s yeast (whole cell approach); more recent studies on the reduction

of aromatic substrates such as acetophenone have been performed using both microbial and

enzyme/coenzyme approaches [50, 51].

2.4 Asymmetric Transfer Hydrogenation

Asymmetric transfer hydrogenation (ATH) is the stereoselective “reduction of an organic sub-

strate by transfer of dihydrogen eqivalents from a suitable donor” [52]. Additionally, the general

use of the term ATH implies that a catalytic step is decisive for both hydrogen transfer and

chiral induction. As it is an operationally simple, mild, and efficient method for the generation

of chiral substances from prochiral substrates, it supplements the afore-mentioned methods

and may provide certain advantages. ATH is generally applicable to carbonyls, imines and C-

double bonds, but it is particularly and most successfully used for the synthesis of non-racemic

secondary alcohols from prochiral ketones [53, 54]. This is also the focus of the following

considerations on different aspects of this method.

2.4.1 Historical Background

Hydrogen transfer reactions have been known since 1903 when Knoevenagel and Bergdolt

reported the formation of 1,4-dihydro-dimethyl cyclohexa-2,5-dienecarboxylate by heating a

mixture of dimethyl 1,4-cyclohexanedicarboxylate and dimethyl terephthalate in the presence

of palladium [55]. The application to carbonyls was independently reported by Meerwein and

Schmidt [56], Ponndorf [57], and Verly [58] more than twenty years later (MPV reduction,

after the inventors´ names). They demonstrated that carbonyls can be converted into alcohols

2 Fundamentals and State of the Art

Scheme 2.2: The MPV reduction and Oppenauer oxidation.

in the presence of superstoichiometric amounts of aluminum alkoxides in solution of easily

oxidizable alcohols such as 2-propanol (Scheme 2.2). The reverse reaction was described later

by Oppenauer [59].

The first “asymmetric catalytic” transfer hydrogenation reaction was reported by von Doe-

ring and Young in 1950 [60]. They reduced ketones by using racemic aluminum alkoxides

in an excess of chiral alcohols and provided evidence for the hypothesis that the hydrogen

transfer proceeds via a six-membered transition state (cf. Scheme 2.2). Mitchell, Henbest,

et al. reported the first example of a transition metal-catalyzed transfer hydrogenation us-

ing iridium complexes in 2-propanol [61, 62]. Some years later, Sasson and Blum applied a

dichloro(tristriphenylphosphine)ruthenium(II) complex as catalyst for the transfer hydrogena-

tion of unsaturated carbonyls [63], and subsequently published an extensive and seminal kinetic

study [64]. In this study, they reported an acceleration in the rate of the reduction of unsaturated

compounds in the presence of a base as co-catalyst. Nevertheless for most such reactions, high

temperatures were required until Bäckvall and Chowdhury reported the ruthenium-catalyzed

transfer hydrogenation of ketones under significantly milder conditions [65]. Just as in the field

of asymmetric hydrogenation (see Section 2.3.1), the development of ATH since the early 1990s

is linked to the name Noyori. His group introduced complexes based on (mono-sulfonated) di-

amine and β-aminoalcohol ligands, which still are among the most successful catalysts in ATH,

and provided a number of seminal contributions regarding the mechanistic aspects of this type

of reaction (see below).

2.4.2 Hydrogen Donor Systems

Arguably the most often employed hydrogen donor in ATH reactions is isopropyl alcohol (IPA =

2-propanol), frequently used with an inorganic base or an alkali metal alkoxide as co-catalyst to

promote the increase of the concentration of 2-propoxide and/or to facilitate the generation of

the true catalyst [66, 67]. Other alcohols have also been used, but were less efficient since— at

least for catalysts based on mono-tosylated diamine or amino alcohol ligands — the order of

reactivity has been determined to be 2-propanol > ethanol > methanol [68, 69]. This might be

due to facile decarbonylation of primary alcohols (especially methanol), which causes catalyst

poisoning [65]. Further hydrogen donors include formic acid, mostly applied as an azeotropic

2 Fundamentals and State of the Art

5:2 mixture with triethylamine (TEAF), or formates used in aqueous solutions. The main

characteristics of each hydrogen donor system are summarized in Table 2.2.

IPA is a suitable solvent for most substrates and catalysts, and it is easily disposed, eco-

nomical, and (relatively) innocuous [66]. A major drawback of its use in ATH reactions is the

reversibility of the conversion to acetone, affecting both the conversion of the substrate and

the enantiomeric excess of the product. During the course of transformation of the substrate,

the rate of the reverse reaction increases, and the ratio of the enantiomers comes under ther-

modynamic control, effecting erosion of the enantiomeric purity [54]. In order to disable the

reverse reaction, IPA is mostly used in excess (as a source of hydrogen and solvent with low

substrate concentrations); distilling off acetone is a further option. The back reaction is more

easily avoided, however, when formic acid or formates are used as the hydrogen donors, because

carbon dioxide is generated upon hydrogen transfer and can simply be vented. Nevertheless,

the use of formic acid and aqueous solutions of formates is limited due to the instability of

some catalysts in acidic media and water.

However, reactions carried out in water have attracted growing attention in recent years.

Although there is some controversy on this subject [70], using water as the reaction medium is

generally considered as a strategy toward “greener” approaches in chemistry since water is safe,

sustainable, and economical [71, 72]. The insolubility of many organic compounds in water has

often been assumed to be a hindrance for its use; in recent studies, however, a growing number

of rate accelerations and improved selectivities have been reported for reactions performed in

water [73], e. g., the asymmetric transfer hydrogenation of aryl ketones with significantly higher

rates (and slightly lower ees) in HCO2Na/H2O than in TEAF [74]. The first ATH approaches

in aqueous media were reported in 2001. Williams et al. added water to reactions carried out

in IPA with water soluble ruthenium [75] as well as rhodium and iridium [76] catalysts, with

the aim of developing biphasic systems and supported liquid phase catalysts. Chung et al. used

the pure sodium formate/water system as a medium for the enantioselective reduction of aryl

ketones. Additionally, they performed a reuse of the catalyst, which remained in the aqueous

phase when the product was extracted with hexane [77]. An overview of the research in the field

of ATH operations carried out in water has been given very recently by Wu and Xiao [78, 79].

2.4.3 Catalysts

2.4.3.1 Metals

Transfer hydrogenation reactions are — with very rare exceptions — catalyzed either by main

group metal alkoxides or by transition metal complexes. The great advances in asymmetric

approaches in the last 30 years, however, have been made using the latter. The most often

employed transition metal is ruthenium, followed by iridium and rhodium [66]. Despite their

2 Fundamentals and State of the Art

Table 2.2: Commonly used hydrogen donor systems.

2-propanol Formic acid Sodium formate/

water

Solvent 2-propanol organic solvents water

Promotor inorganic base triethylamine none

Drawbacks reversibility of the reaction too harsh for some catalysts solubility of reactants

Indication IPA HCO2H; TEAF when used HCO2Na/H2O

with triethylamine

enormous success in an ever growing number of applications, these three metals have the

drawback of high cost and toxicity. Thus, some research groups are currently investigating

alternatives, and interesting results have been reported.

Furthering the work of Kagan et al. [80], Evans and co-workers reported an asymmetric

version of the MPV reduction using a chiral tridentate samarium complex as catalyst [81]; aryl

methyl ketones were able to be reduced with high enantioselectivity (up to 97 % ee) and high

isolated yields (up to 96 %). List and Yang used a copper-bisoxazoline complex as chiral catalyst

and Hantzsch esters as synthetic NADH-analogous hydrogen donors for the enantioselective

transfer hydrogenation of α-ketoesters in chloroform [82]. This approach, however, provided

ee values above 90 % only at a low temperature (−25 ◦C) with prolonged reaction periods.

Iron complexes have been successfully used by the groups of Beller [83, 84] and Morris [85] (cf.

Section 4.1.1).

2.4.3.2 Ligands

A great number of chiral chelating ligands (chelate from

 qhl

, claw or pincer), several

successful examples of which are depicted in Figure 2.2, have been reported in the context of

ATH. As donor atoms, nitrogen, oxygen, phosphorus, and rarely sulfur or others are employed.

The multidenticity of these ligands — belonging to one of the categories bidentate,tridentate,

or tetradentate — is an important factor which influences the complex stability. Gladiali further

distinguishes between “anionic” and “neutral” ligands, since possessing or not a protonated

donor centre of appropriate acidity, e. g., tosyl-NH in monosulfonated diamines and OH in

amino alcohols (see below) [54], “is crucial for enabling an outer or an inner sphere mechanism

in H-transfer” [66]. However, amino alcohols and mono-sulfonated diamine ligands possess two

protic donor atoms, amine-N/alcohol-O and amine-N/amide-N, respectively. The alcohol and

amide groups of these ligands deprotonate upon metal insertion (to build the precatalyst) and

act as anionic donor units, while deprotonation of the amino groups occurs in a second step

under reaction conditions (to build the active catalyst and during the course of H-transfer,

cf. Section 2.4.4). An important feature of these “anionic” ligands is thus the “NH effect”,

2 Fundamentals and State of the Art

the interaction between amine-H and substrate-O in the ATH of ketones [67]. This effect,

however, has also been observed with “neutral” ligands containing secondary amines, e.g., the

tetradentate diphosphine/diamine (PNNP) ligand (Figure 2.2 c), which yielded a significantly

more active catalyst than the diimine analog [86, 87]. Laue therefore suggested that the active

catalyst contains one deprotonated secondary amine [88], and thus follows a pathway which is

analogous to that of catalysts made from “anionic” ligands [89].

(a) TsDPEN [90] (b) TsDACH [91] (c) PNNP [86] (d) [92]

(e) aminoindanol

[93]

(f) [94] (g) tethered TsDPEN [95] (h) CsDPEN [96]

Figure 2.2: Ligands successfully used in ATH operations.

Complexes made of the p-toluenesulfononyl-diphenylethylenediamine (TsDPEN) ligand (Figure

2.2 a) and ruthenium-η6-arene precursors (Ru-η6-arene-TsDPEN, henceforth referred to as Ru-

TsDPEN) are considered the most versatile catalysts in ATH, providing high enanatioselectivity

with a variety of substrates [54]. Variants of the TsDPEN ligand, however, have given superior

results in certain applications, e. g., the use of diaminocyclohexane (DACH) instead of diphenyl-

ethylenediamine (Figure 2.2 b), the tethering of a tetramethyl cyclopentadienyl moiety, which

acts as an ancillary η5-ligand (tethered TsDPEN, Figure 2.2 g), or the substitution of the tosyl

for a camphorsulfonyl (Cs) moietey (Figure 2.2 h). Some β-aminoalcohol ligands such as 2-

methylamino-1,2-diphenylethanol (Figure 2.2 d) in conjunction with [RuCl2-η6-C6Me6]2showed

a higher activity than Ru-TsDPEN, but lagged behind in terms of enantioselectivity [92]. Exper-

imental and theoretical studies of aminoalcohol-based catalysts have revealed the great impact

of the ancillary arene ligands on the catalytic performance [67, 92, 97, 98]. Thus, both the

activity and the enantioselectivity of the catalyst can be enhanced by substituting H-atoms of

the η5-cyclopentadienyl units — used in conjunction with rhodium(III) and iridium(III) — and

η6-arenes — used in conjunction with ruthenium(II) — for alkyl groups.

2 Fundamentals and State of the Art

2.4.4 Mechanistic Aspects

For metal-catalyzed hydrogen transfer reactions involving carbonyl molecules, two basic mech-

anistic routes have been proposed: the direct hydrogen transfer and the hydridic route [99].

The first has been presumed to occur predominantly with main group metal catalysts; the latter

is presumed for transition metal catalysts and can be further divided into monohydridic and

dihydridic routes [100].

2.4.4.1 Direct Hydrogen Transfer

In catalytic reactions following this pathway, a hydride is transferred directly from the metal-

coordinated reductant to the simultaneously coordinated oxidant via a concerted six-membered

ring transition state, as depicted in Scheme 2.2 (see above). The mechanism has been supported

by experimental as well as computational results for MPV reduction and Oppenauer oxidation

in the classical version applying aluminum alkoxide as a promotor, or using aluminum and other

non-transition metals as catalysts [101–104]. Some examples of transition metal catalyzed

reactions, however, are proposed to follow an analogous pathway [94, 105].

2.4.4.2 Hydridic Route

In contrast to the direct pathway, hydridic routes imply a metal-hydride transition state where

one or two hydrogen atoms are coordinated to the metal before being transferred to the sub-

strate. Accordingly, a division into monohydridic and dihydridic reaction routes has to be made

[100]. Rhodium and iridium catalysts generally favor the mono-hydride pathway, whereas the

mechanism of the hydrogen transfer mediated by ruthenium catalysts depends on the ligands

[106]. Additionally, reactions that follow the monohydridic path can proceed via an inner sphere

mechanism or an outer sphere mechanism, depending on whether the ligand is neutral or an-

ionic, respectively [66], or depending on whether there is an “NH effect” or not (cf. Section

2.4.3.2).

Monohydridic Route A characteristic of metal mono-hydrides is that they are only formed

by carbon-bound hydrogen atoms. Remarkably, these hydrogen atoms keep their identity, i. e.,

when alcohols are used as hydrogen donors, the alcohol´s C-bound H-atom forms the metal

hydride and is then transferred to the carbonyl-C, whereas the alcohol´s O-bound H-atom is

independently transferred to the substrate´s carbonyl-O [107]. In reactions that proceed in the

inner sphere of the catalyst, a metal alkoxide is formed by the hydrogen donor and the catalyst

[53]. Scheme 2.3 depicts as an example the reduction of acetophenone using 2-propanol as the

hydrogen source and a transition metal complex as precatalyst (Sch. 2.3 a; X = anionic ligand,

typically a halide; L = supporting ligand) [69]. After formation of the metal alkoxide (Sch.

2 Fundamentals and State of the Art

Scheme 2.3: Inner sphere mechanism.

2.3 b) via displacement of X by 2-propoxide in a basic medium, successive β-elimination and

elimination of the acetone yield the metal hydride (Sch. 2.3 b→d). Linkage of acetophenone,

followed by migratory insertion, then yields the new metal alkoxide (Sch. 2.3 d→f). The

catalytic cycle is completed by ligand exchange and proton transfer yielding the metal propoxide

(Sch. 2.3 b) and 1-phenylethanol as the product.

In contrast, there is no metal alkoxide formation in reactions that take place in the outer

sphere of the catalyst. The hydrogen transfer instead occurs via a transition state where

the substrate interacts with the catalyst via hydrogen bonds and dipolar interactions, without

coordination to the metal center. A further subdivision of the outer sphere mechanism has

been made by the proposal of a stepwise version in addition to the concerted mechanism, and

by the proposal of a bimetallic pathway.

The concerted outer sphere mechanism — also known as metal-ligand bifunctional mecha-

nism — is generally accepted for the ATH of ketones in the presence of Ru-η6-arene complexes

of mono-sulfonated diamines or β-amino alcohols, combined with an alcoholic hydrogen source

[67, 69, 98, 108, 109]. Scheme 2.4 presents an example for this mechanism, which was first

proposed by Noyori and co-workers in 1997 [68]. For both the formation of the precatalyst

(Sch. 2.4 a) in situ and the formation of the active 16 e complex (Sch. 2.4 b) via HCl-elimination,

a base is required in 2-propanol solution; it is, however, not required as a co-catalyst, since the

amine-N of the ligand acts as a basic center [110]. The transfer hydrogenation of the substrate

catalyzed by the 18 e species (Sch. 2.4 d) involves a concerted transfer of proton and hydride

2 Fundamentals and State of the Art

Scheme 2.4: Concerted outer sphere mechanism.

(Sch. 2.4 e) received from the hydrogen donor (see Sch. 2.4 c). The precatalyst (Sch. 2.4 a) as

well as the active catalyst (Sch. 2.4 b) and the reactive intermediate (Sch. 2.4 d) were isolated,

and the molecular structures were confirmed [68]. A non-concerted but rather stepwise outer

sphere mechanism, however, has been proposed, e. g., by Adolfsson et al. for the use of a ruthe-

nium pseudo-dipeptide complex in the ATH of aryl ketone substrates performed in 2-propanol

in the presence of lithium chloride [111].

Findings for the ATH of acetophenone in an aqueous/organic solution of excess sodium for-

mate catalyzed by Ru-TsDPEN indicated a catalytic cycle similar to the concerted mechanism

in 2-propanol (Scheme 2.5, cf. Scheme 2.4), as reported by Xiao, Liu, and co-workers [112].

When water is present, it acts as the base in the in situ formation of the precatalyst (Sch. 2.5 a).

The addition of sodium formate provides the hydride species (Sch. 2.5 c) presumably via for-

mation of a formato complex (b, species that could not be isolated). The hydrogen transfer

to the substrate probably proceeds via a slightly modified transition state (Sch. 2.5 d); water

significantly accelerates the reaction and — as supported by DFT calculations— is believed to

alter the mode of hydrogen transfer from concerted to stepwise. As water acts as the source of

protons, the reaction medium becomes increasingly basic with increasing conversion. A rate law

2 Fundamentals and State of the Art

Scheme 2.5: Proposed outer sphere mechanism in aqueous media.

consistent with the mechanistic consideratons and experimental kinetic results — a linear rela-

tionship relative to both the catalyst and the substrate concentration in a certain range — has

been developed under the assumption that the coordination of formate and the decarboxylation

are reversible and equilibrated prior to hydrogen transfer (Equation 2.4.1).

r=K1K2k[cat][HCO−

2][acp]

K1K2[HCO−

2]+[CO2][OH−](2.4.1)

Furthermore, the term [CO2][OH−]was expected to be small since emerging carbon dioxide

would react with the hydroxide ions released from the water molecules to form bicarbonate.

Thus, when

[CO2][OH−]<< K1K2[HCO−

is true, Equation 2.4.1 is reduced to Equation 2.4.2.

r=k[cat][acp](2.4.2)

On the basis of this rate law, the rate constant kwas calculated at different temperatures in

order to determine the enthalpy of activation (∆H‡) and entropy of activation (∆S‡) from an

Eyring plot. As can be seen from Equation 2.4.4, which is the linear form of Equation 2.4.3, the

slope of a plot of ln k

Tagainst 1

Tallows for the calculation of ∆H‡, whereas ∆S‡can be calculated

from the intercept [113]. The values of the activation parameters, especially ∆S‡, are routinely used

2 Fundamentals and State of the Art

for mechanistic interpretation. Xiao, Liu, and co-workers reported a ∆H‡of 12.8 kcal mol−1

(53.6 kJ mol−1) and a ∆S‡of −25 cal K−1mol−1(−105 J K−1mol−1) for the conversion of

acetophenone [112]. From the large negative value of ∆S‡, a well-ordered transition state (Sch.

2.5 d) was deduced.

k=kBT

h·e−∆H‡

RT ·e∆S‡

R(2.4.3)

ln k

T=−∆H‡

R·1

T+lnkB

h+∆S‡

R(2.4.4)

Dihydridic Route It has been observed in experiments with deuterated alcohols that hydrogen

atoms transferred in ruthenium-dichloride-catalyzed reactions in the presence of a base lose

their identities [107, 110]. Thus, a further mechanism has been proposed in hydrogen transfer

reactions: the dihydridic route. In this mechanism, both the alcohol´s O-bound and the α-

C-bound H-atoms are transferred to the metal and then transferred from the metal to the

ketone.

2.4.5 Applications and Industrial Impact

Asymmetric transfer hydrogenation has emerged from a method of mere academic interest to

one that is used for the commercial production of chiral compounds [114]. While data about

implemented commercial processes are difficult to obtain, and thus little is known about the

number and details of such processes, the increasing interest in technical applications of this

type of chemical operation is evidenced by the growing number of scale-up studies and tests

of industrially relevant substrates by both academic and industrial researchers [20, 115, 116].

These studies demonstrate that a wide range of highly functionalized substrates can be employed

in ATH operations even on a technical level. At the same time, it is apparent that each process

is unique regarding the applied catalyst, reaction conditions, solvent, hydrogen donor system,

etc., and great effort is required to find the optimal solution to a given problem. In particular,

the transfer of the know-how developed under simplified laboratory conditions to technical

applications remains challenging [19]. The predominant functional group reduced in ATH

reactions on the pilot scale is the carbonyl group; imines, however, have also been employed.

Figure 2.3 shows examples of chiral compounds for the production of pharmaceuticals which

have been synthesized making use of ATH.

•In 2003, Merck researchers reported on the development of an ATH process for the

manufacture of (R)-3,5-bistrifluoromethylphenyl ethanol (Figure 2.3 a), an intermediate

in the synthesis of aprepitant (a pharmaceutical sold under the brand name Emend) [117].

The reaction was performed using a cis-aminoindanol-Ru(p-cymene) complex (cf. Figure

2 Fundamentals and State of the Art

(a) (R)-3,5-bistrifluoro-

methylphenyl ethanol

(b) cis-(2S)-lactam (c) (S)-OPC-41061

Figure 2.3: Intermediates produced via ATH in scale-up studies.

2.2 e, Section 2.4.3.2) as catalyst in the IPA/base system on a multi-10 kg scale. The

chiral alcohol was obtained with an ee of 91 %, which had to be improved by further

purification.

•A process including an ATH step, designed to avoid the resolution of a racemate and

attendant waste, were developed for production of the drug Diltiazem [116]. The key

building block cis-(2S)-lactam (Figure 2.3b) was produced on a multi-liter scale by using

an Ir-CsDPEN catalyst (cf. Figure 2.2 h, Section 2.4.3.2) at an S/C of 2 000 in a biphasic

medium of water/isobutyl acetate with TEAF as the hydrogen source. The product was

obtained with an overall ee of more than 99 % while generating half the waste of the

traditional route.

•Researchers at Otsuka studied the production of the Vasopressin V2 Receptor Antagonist

OPC-41061 (Figure 2.3 c) on the laboratory scale (∼6 g) [118]. They used a Ru-TsDPEN

catalyst (cf. Figure 2.2 a, Section 2.4.3.2) in IPA/KOH, and obtained the product in 99 %

yield and 89 % ee, which was further improved via crystallization. This appraoch proved

more efficient than asymmetric reduction methods using borane reagents.

A further indicator of the increasing commercial interest in ATH reactions is the growing number

of commercially available catalysts. Under the name CATHyT M , Avecia has developed catalyst

kits, which are available from Strem Chemicals Inc. [119] and which contain aminoindanol

and TsDPEN ligands as well as rhodium and iridium precursors to prepare the required ATH

catalyst in situ. The TsDPEN ligand as well as a derivative (FsDPEN) are available from

Takasago [120], and a tethered version of the η6-arene-Ru-TsDPEN complex is available from

Johnson Matthey [121].

2 Fundamentals and State of the Art

2.4.6 Immobilization

As mentioned above, a major drawback of using homogeneously dissolved molecular catalysts

in chemical production processes is their separation from reactants and products after the

reaction has finished. To overcome this problem, a multitude of methods for the immobilization

of catalysts has been developed, mainly in the context of asymmetric reactions [122–127]. To

comprise all approaches, a definition of the term immobilization must be quite wide, for example

“any strategy to facilitate catalyst separation and recycling”. Catalyst immobilization strategies

can be classified, e. g., by the state of the immobilized system under reaction conditions (solid

or dissolved), by the support material applied (organic or inorganic), or by the method of

immobilization (covalent binding, entrapment, micellar embedment, etc.). However, a complete

discussion is outside the scope of this thesis, and two of the major categories —covalently bound

catalysts and non-covalently bound catalysts — were built by considering the number of the

respective applications rather than systematically. The most common techniques are discussed

below, with recent examples in the field of asymmetric transfer hydrogenation referenced for

each.

The starting point for the formation of an immobilized asymmetric catalyst is usually a

chiral ligand or the corresponding metal complex, which has previously proven efficient in non-

supported operations. The covalent binding of catalysts, either via copolymerization or via

covalent attachment (anchoring) to a support, is the most often applied method for cata-

lyst heterogenization [123], but is suitable for building soluble immobilized catalysts as well.

Non-covalent immobilization includes approaches as different as the use of liquid supports,

entrapment in porous solid supports, immobilization via adsorption, and others. Despite the

great variety, approaches other than covalent attachment have gained less attention in ATH

and studies on them are thus less in number.

Although significant advances have been made in the fields of both asymmetric catalysis

and catalyst immobilization, there is apparently no use of immobilized chiral metal complexes

on an industrial scale [3, 5, 123]. Blaser and Pugin have identified a number of requirements

on heterogeneous asymmetric catalysts for commercial use, most of which have so far not been

met [5]. As a result, most immobilized catalysts are not yet competitive compared to alternative

methods such as enantiomer resolution and chiral pool synthesis, or homogeneous catalysis.

The requirements on heterogeneous asymmetric catalysts include the variability (combinatorial

approaches) and cost efficiency of their preparation (additional costs of the immobilized catalyst

should be lower than those for the separation of the homogeneous catalyst or outweighed by

other advantages such as catalyst reuse), their performance (selectivity, activity, productivity

should be comparable or better than of the homogeneous catalyst, 95 % recovery and minimal

metal leaching is required, cf. Section 2.2.2), and their handling (simple separation is required).

Furthermore, the molecular weight of the heterogeneous catalyst should not exceed 10 kDa

2 Fundamentals and State of the Art

per mole of active site. Additionally, it is crucially important that it be available in technical

quantities within the time frame of the process development [5]. In fact, availability is a

problem even for testing on a laboratory scale; at least immobilized ATH catalysts are not yet

commercially available [128] (see Section 5.4.2).

2.4.6.1 Immobilization via Covalent Binding

Independent of the procedure, covalent binding mostly requires the modification of the ligand

structure to introduce at least one functional group (linker), and in some cases a certain

distance between the linker and the catalytically active center has to be assured by a spacer.

Furthermore, the applied support material or co-monomer, the length and flexibility of the

spacer, the catalyst loading, etc. affect the accessibility of the active catalyst center and the

final catalyst performance. In the following, covalent attachment and polymerization approaches

are discussed separately. The first is distinguished from the latter by the existence of the support

prior to the immobilization procedure; that is, a ligand or catalyst is attached to an existing

support in order to form the immobilized catalyst. In case of (co)polymerization, though, the

support is formed from the ligand (and the co-monomers) during the immobilization procedure.

Covalent Attachment Polywka et al. presented an early example for the heterogenization

of ATH catalysts [129]. As depicted in Scheme 2.6, the ligand was prepared by coupling a

modified tosyl moiety (containing a linker) with the chiral diamine backbone. Amino resins

were used as solid supports to which the modified TsDPEN ligand was efficiently attached via

peptide coupling. Metal insertion using a [RuCl2(p-cymene)]2precursor was performed in situ

when the polymer-bound catalyst was used for the ATH of acetophenone in IPA and TEAF

systems. Ee values of up to 99 % were achieved, but the reaction was slow, and the catalyst

was nearly non-reusable.

A similar approach has recently been presented by Somanathan and co-workers [130]. Instead

of a TsDPEN-based ligand, a modified TsDACH ligand (cf. Figure 2.2b, Section 2.4.3.2) was

coupled to aminomethylated polystyrene, and used in conjunction with [RhCl2(Cp*)]2as a

heterogeneous catalyst for the ATH of aryl ketones in water/sodium formate. An enantiomeric

excess of 90 % and nearly complete conversion were observed, and the catalyst was recycled

four times with only a slight drop in activity.

Van Leuwen, Reek, et al. reported the attachment of a trimethoxysilane-functionalized

aminoalcohol ligand onto silica. The remaining silanol sites on the support were transformed

into alkylsilane sites in a further step [131]. Both batchwise and continuous applications of

the immobilized ligand and the [RuCl2(p-cymene)]2precursor as catalyst were successfully per-

formed in IPA. In a similar approach, the introduction of ethyltrimethoxysilane as a spacer/linker

into the tosyl moiety allowed a TsDPEN derivative to be anchored onto different silica supports.

2 Fundamentals and State of the Art

Scheme 2.6: Covalent attachment of a TsDPEN derivative to aminated polystyrene supports.

The heterogenized ligand in conjunction with [RuCl2(p-cymene)]2was successfully used for the

ATH of aryl ketones in the organic TEAF solutions [132] as well as in aqueous media [133, 134].

Further heterogeneous silica-supported systems were reported in recent years [135–137]; one

remarkable approach was provided by Li et al., who immobilized a Ru-TsDPEN catalyst in a

siliceous mesocellular foam which was magnetized through the grafting of magnetic nanopar-

ticles [138]. The use of cobalt nanoparticles as support for ATH catalysts was reported by

Pericàs and co-workers [139].

An example for attaching a (pre)catalyst to a soluble support was given by Liese et al.

[140]. Their “chemzyme approach” consisted of a PNNP-based ruthenium catalyst (cf. Fig-

ure 2.2 c, Section 2.4.3.2) which was linked to a polysiloxane chain (Section 4). The sol-

ubility of the polymer in 2-propanol was increased by additionally attaching polar tris(2-

methoxyethoxy)vinylsilane, and the catalyst was successfully used in a continuously operated

membrane reactor providing high reusability (see also Section 4.1.2).

(Co)polymerization Chiral polymers had been used for various purposes long before immo-

bilized ATH catalysts gained interest [141]. Arguably the first example of an immobilization

approach for ATH applications was reported by Lemaire et al. in 1994 [142]. Polyamides and

polyureas were made from a chiral diphenylethylenediamine ligand in conjunction with tereph-

thaloyl chloride and bis(1,4-isocyanatophenyl)methane, respectively. Although rather modest,

the overall performance of the insoluble rhodium-polyurea catalyst was better than that of the

non-immobilized rhodium-diamine in the ATH of acetophenone in 2-propanol/KOH.

In a variation on this approach, preformed rhodium complexes were employed for copoly-

2 Fundamentals and State of the Art

Scheme 2.7: Crosspolymerization of a TsDPEN-based ligand monomer.

merization, and molecular imprinting was applied in order to improve the enantioselectivity

[143]. In fact, the templated copolymers yielded better results than the non-templated ones

in terms of the enantioselectivity, but the overall performance was still not satisfactory [144].

Further examples of copolymerization combined with molecular imprinting were reported by

Severin and Polborn, who applied a “styrene derivative” of the TsDPEN ligand [145–148].

The same modified TsDPEN ligand had been previously used by Lemaire for copolymerization

with styrene and styrene/divinylbenzene to get both linear and crosslinked polymers (see Scheme

2.7), respectively, which were used with iridium and ruthenium precursors as catalysts for the

ATH of acetophenone in IPA [149, 150]. This approach was seminal for further developments.

The introduction of hydrophilic pendent groups into the polymer framework allowed the catalyst

to be successfully applied to the ATH of ketones and imines in aqueous reaction media [151–

153].

2.4.6.2 Non-Covalent Immobilization

The significant advantage of many immobilization methods not based on covalent binding is

that a modification of the ligand is not necessarily required. For some of these methods, recent

examples can be found in the context of ATH.

Liquid Supports As mentioned above, Chung et al. reported arguably the first liquid-

supported ATH catalyst [77]. They used a modified proline amide ligand, which was made

water soluble through the introduction of a fluoride group, and obtained promising results,

although the activity dropped slightly upon recycling. In five consecutive runs with 94–95 %

ee, the time until nearly complete conversion was achieved was prolonged from four to seven

hours.

Fan, Gu, Chan, and co-workers reported the ATH of acetophenone mediated by unmodified

2 Fundamentals and State of the Art

Ru-TsDPEN in mixtures of polyethylene glycol (PEG) and different hydrogen donor systems.

The best results in terms of recyclability and reactivity were obtained when a homogeneous

mixture of PEG and sodium formate/water was used as the reaction medium. After each run,

the product was extracted with hexane, while the catalyst remained in the PEG/water phase.

The hydrogen donor was regenerated by the addition of formic acid, and up to 14 recycling

experiments were performed with nearly stable ees(94–96 %) but with increasing reaction

periods [154].

Micelles The first report on the use of surfactants in aqueous ATH was published by Chung

et al. [155]. An unmodified proline-based ligand in conjunction with a ruthenium precursor

was used as the catalyst for the reduction of various aryl ketones in sodium formate/water.

Recycling of the catalyst was possible, but the performance after recovery was rather poor.

Zhu, Deng, and co-workers reported the use of Ir, Rh, and Ru-TsDPEN catalysts in the

presence of micelle-forming surfactants in sodium formate/water [156]. Ru-TsDPEN, embedded

in micelles formed from cationic surfactant cetyltrimethylammonium bromide, was successfully

applied to six consecutive runs of the ATH of acetophenone, in which stable eesof 95 % were

obtained but reaction periods increased (first run: 6 h to conversion >99 %; sixth run: 13 h).

Entrapment The encapsulation of a homogeneous Ru-TsDPEN catalyst in mesoporous sil-

ica via adjustment of the pore entrance size by silylation was demonstrated by Yang, Li, and

co-workers [157]. The supported system was reusable multiple times in ATH reactions of aryl ke-

tones in sodium formate/water, but the reuse suffered from increasing reaction periods, probably

due to the loss of catalyst during the course of recovery. Using acetophenone as the substrate,

the ee was stable at 92–93 % over six runs, the same enantioselectivity as achieved with the

non-encapsulated catalyst. In a very recent study, this approach was significantly enhanced by

tuning the microenvireonment of the catalyts within the nanocage using an amphiphilic silyla-

tion agent [158]. The modified cage showed a highly increased adsorption capacity for water,

benzene, and negatively charged ions. The higher rates at stable eesachieved with this system

were therefore ascribed to the ability of the cage to accumulate the reactants.

3 Support Materials Applied

Appropriate support materials for catalyst heterogenization purposes have to satisfy several

conditions: chemical and mechanical stability, high specific surface area, high density of functio-

nalities as well as accessibility and acceptable costs. Additional requirements depend on the

specific application, for example a high or low hydrophilicity of the surface, or a special geometry.

To achieve the aim of this thesis, the development and study of highly reusable catalysts for

ATH operations, immobilization via covalent attachment to solid supports was the method of

choice. On the one hand covalent bonds provide the strongest type of linkage, thus impeding

catalyst loss, and on the other hand solid supports have known material properties and most

efficiently allow for facile quantitative separation. This required materials with the above-

mentioned properties in addition to functional groups which allow for catalyst immobilization

via standard coupling procedures. Such supports were supplied by PolyAn GmbH, a surface-

technology-focused company located in Berlin. Special needs were met by the joint development

of a new support type.

3.1 Material Choice and Preparation

PolyAn specializes in Molecular Surface Engineering (MSE) and provides custom-made,

molecular-designed material surfaces and boundary layers, equipped with specific functions

[159]. These materials are mainly used as supports for biomolecules for the application in med-

ical diagnostics and for use in membrane-based separation processes. Other applications that

have been studied include the use of surface-functionalized polymer chips (PCs) as supports for

C–C-coupling catalysts [160]. These attempts were considered a starting point, and analogous

(a) Membranes (b) Sinter chip (c) Beads

Figure 3.1: Support materials applied.

3 Support Materials Applied

PCs were chosen for the present study, i. e., aminated membrane slices made of polypropylene

(Figure 3.1 a) and sinter chips made of polyethylene (Figure 3.1 b). Both types of polymer chips

were readily available from PolyAn, but for several experiments which had to be conducted un-

der highly reproducible conditions in a flask with mechanical stirring, another support version,

namely beads, was required (cf. Section 5.2.1). Thus in a joint project, the MSE-based func-

tionalization was adapted for ultra high molecular weight polyethylene (UHMW-PE) particles of

m in diameter, generously provided by Ticona (Figure 3.1 c, micro particles henceforth de-

noted as polymer beads, PBs). While the details of the PolyAn technology cannot be published,

some aspects of the material properties are discussed in the following.

3.1.1 Molecular Surface Engineering

Molecular Surface Engineering denotes the design of the chemical properties of material surfaces

by utilizing copolymerization techniques (grafting from). The method is applicable to a wide

range of base materials with different morphologies, whether inorganic like glass or organic like

artificial polymers, and whether with planar or porous structures etc.

Hence, the applied base materials are modified by grafting a “matrix” with terminal func-

tional groups. The advantage of MSE over other techniques is that the matrix is covalently

bound to the base material, and that the chosen functional sites are mono-fractional. Both

the functional groups as well as further properties of the matrix, e. g., the hydrophilicity–

hydrophobicity balance, can be chosen from various options or ranges [159]. Additionally,

due to the three-dimensional structure of the grafted matrices, comparatively high loadings of

accessible functional groups are provided.

3.1.2 Material Properties

The materials applied in this study were composed of polypropylene (PP) or polyethylene

(PE) base materials and the grafted polymer matrix, which contained both hydrophilic and

hydrophobic compounds as well as terminal primary amines as functionalities (Figure 3.2).

The amino groups were Boc-protected for chemical stability during storage. The balance of

hydrophilicity and hydrophobicity of the polymer chips provided a sort of amphiphilic character,

which had a positive effect on the reaction rate when the materials were used as catalyst

supports for the reaction of organic substrates in polar media (see Section 5.2.2.1). Thus,

further optimization of the matrix properties were not considered necessary, and beads were

functionalized in an analogous manner using an adapted process.

The supports were manufactured with different loadings of amino groups, usually in the range

of 10–30

mol g−1. The number of functional groups on the surface is related to the thickness

of the grafted matrix, which can be controlled in the MSE process. A study of analogous

polymer materials showed that the amount of detectable functional groups per surface area

3 Support Materials Applied

Figure 3.2: Schematic illustration of the support materials.

follows a saturation curve when plotted versus the layer thickness, since with increasing layer

thickness there is a growing number of inaccessible groups. Reasonable values for the matrix

thickness are in the range of 10 nm to 300 nm as determined via interferometry and other

methods [161].

Accurately determining of the number of (accessible) amines is challenging, and different

methods will provide different results. A reasonable method to quantify them is the coupling

of (cleavable) compounds which absorb in the ultraviolet (UV) and visible (Vis) range, which is

carried out as follows: after coupling of the specific compound, the supports are washed, the UV

active part is cleaved, and finally the concentration, which is equal to the effectively accessed

amino groups, is determined via UV absorption measurement. Another method is to use an

accurately determined amount of the coupling reagent and measure the fraction which has not

coupled to the surface, again via UV-Vis or fluorescence analysis. For this, different compounds

(Fmoc-β-Ala-OPfp [Fmoc], ortho-phthaldialdehyde [OPA], Ponceau S) were applied, and a

systematic deviation was found, which might primarily be due to the molecular size of the

respective compounds. The same was observed when the amount of metal determined via

inductively coupled plasma mass spectrometry or optical emission spectroscopy (ICP-MS/ICP-

OES) of digested PB and PC supports was compared to the number of amino groups previously

determined using Fmoc as the reference compound. Figure 3.3 shows that there is a linear

relationship in the investigated range, but that there is less coupling of the catalyst complex

(see Section 5.1.2) than Fmoc coupling. The Fmoc/UV-Vis test was used as the standard

procedure to determine of the amine loading, taking into account that absolute values might

be somewhat deceptive. ICP-based analysis, on the other hand, was used to determine the

catalyst loading, assuming that the amount of metal found is equal to the amount of catalyst.

Consequently, it became evident that a certain number of free amino groups were still present

after catalyst immobilization. Tests of functionalized materials (carboxyl function) have revealed

that the number of transformed groups is much smaller than that of non-transformed [161].

The free amines might have an additional effect on the material properties, but this (potential)

3 Support Materials Applied

Figure 3.3: Loading with amino groups vs. loading with rhodium.

effect could not be investigated in the present study. However, the mechanical stability of

the polymer beads, which were required to ensure homogeneous suspension through vigorous

mechanical stirring, was proven via optical determination of the size distribution. No significant

deviation was observed in the particle size of unused beads and beads employed in an ATH

experiment performed at 1 200 rpm and 60 ◦C.

4 Modification of a Ruthenium Catalyst

ATH ligands or (pre)catalysts which meet the requirements for immobilization are not com-

mercially available. In academia, however, a number of pathways have been developed for

the modification of various ATH catalysts with the aim of immobilization. Thus, the man-

ufacturing of ligands or catalysts appropriate for immobilization could be managed either by

following one of the routes described in the literature or by developing a new route. As described

above, for the purpose of covalent attachment the ligand has to possess a suitable functional

group and should provide high complex stability through multidenticity. A tetradentate diphos-

phine/diamine ligand, which has widely been used in the context of ATH, was chosen as a

candidate for immobilization. Since none of the prior attempts at modification of this ligand

were considered appropriate for the covalent attachment to aminated supports, a new strategy

for the introduction of a linker was required.

4.1 Background

4.1.1 Homogeneous Applications

Diphosphine/diamine ligands had been studied for some time [162–164] before the N,N0-bis[2-

(diphenylphosphino)-benzyl]cyclohexane-1,2-diamine (PNNP) ligand (Figure 4.1 a) and its di-

imine derivative were used in conjunction with ruthenium(II) (Figure 4.1 b) as (pre)catalysts for

ATH reactions by Gao, Ikariya, and Noyori in 1996 [86]. The Ru-PNNP catalyst provided 97 %

ee and 93 % yield (TON = 186, TOF = 27 h−1) in the asymmetric transfer hydrogenation of

acetophenone in IPA/base. Both the rate and enantioselectivity provided by the diimine-based

version were significantly lower. Promising results, however, were obtained with cationic PNNP-

based rhodium(I) complexes with different anions in the IPA/base medium [165]. Nevertheless,

in terms of the overall performance of the ATH reaction, the neutral Ru-PNNP complex was

favorable over the ionic rhodium version [166].

PNNP has also been used for ATH operations in conjunction with rhodium and iridium pre-

cursors in 2-propanol without the addition of a base, and high yields and ee values were obtained

with various aryl ketone substrates [167]. Morris et al. reported on the successful asymmetric

transfer hydrogenation of ketones performed in the 2-propanol/base system using an iron com-

plex containing the diimine version of PNNP as ligand [85]. More recently, this approach was

4 Modification of a Ruthenium Catalyst

(a) (R,R)-PNNP (17) (b) (R,R)-Ru-PNNP (18)

Figure 4.1: Diphosphine/diamine ligand and ruthenium complex.

applied to imines as substrates by Beller and co-workers [84]. The PNNP ligand has also been

used in aqueous reaction media; Gao et al. reported the in situ generation of a PNNP-based

iridium catalyst and its application to the ATH of various aryl ketones in sodium formate/water

in the presence of phase transfer catalysts such as bis(triphenylphosphoranylidene) ammonium

chloride (PPNCl) [168]. Further applications of complexes on the basis of diphosphine/diamine

or diphosphine/diimine ligands for catalytic operations include asymmetric epoxidation and ox-

idation reactions [169–171].

Generally, the R-enantiomer of the product is formed in ATH reactions of prochiral aryl

ketones when the S,S-version of the PNNP-based catalyst is applied, whereas the S-enantiomer

is obtained when the R,R-catalyst is used. Acetophenone is usually employed as a benchmark

substrate (see Scheme 4.1).

Scheme 4.1: ATH of acetophenone as the benchmark reaction.

4.1.2 Immobilization Approaches

Two pathways have been established to immobilize complexes based on 18 via covalent at-

tachment; in both cases, soluble polymers were used as supports. Gao et al. presented an

immobilized version which is characterized by simple preparation, since no modification is re-

quired to link one of the amines of the complex with poly acrylic acid (PAA, see Figure 4.2 a)

[172]. The performance of the supported catalyst, however, was rather poor, displaying signif-

icantly prolonged reaction times compared to the free analog, decreased ee values, and poor

reusability.

The other example was presented by Liese et al. [140]. Their “chemzyme approach” consisted

of a modified version of 18 which was linked to a polysiloxane chain (Figure 4.2 b). The

solubility of the polymer in 2-propanol was increased by additionally attaching polar tris(2-

methoxyethoxy)vinylsilane. A total of 12 steps were required for the preparation, and the

4 Modification of a Ruthenium Catalyst

(a) PAA-(R,R)-Ru-PNNP (b) “PDMS”-(R,R)-Ru-PNNP

Figure 4.2: Immobilized ruthenium-PNNP complexes.

resulting catalyst was successfully used in a continuously operated membrane reactor providing

high reusability. A detailed description of the system performance was given, with reported

STY values of up to 578 g L−1d−1(24 g L−1h−1), a TTN of 2 630, an ee of 91 %, and a TOF

of 0.22 min−1(13 h−1) for the ATH of acetophenone in IPA/base. Furthermore, a detailed

kinetic study of this system was reported by Greiner and co-workers [173].

4.2 Initial Studies

The high catalyst stability and reusability reported for the “chemzyme approach” as well as

the high versatility of the PNNP ligand indicated by various studies— not only in the field of

ATH — were encouraging signs that this system could be made applicable for the attachment to

solid supports. Initial studies were undertaken in order to identify the strengths and weaknesses

of different PNNP-based catalysts as well as potential strategies for synthetic modification.

4.2.1 Preparation and Use of Diphosphine/Diamine-based Catalysts

4.2.1.1 Ruthenium-Diphosphine/Diamine

The “original” PNNP ligand is prepared in a straightforward way by coupling two equivalents of

2-(diphenylphosphino)benzaldehyde (9) with either the R,R- or the S,S-version of diaminocyclo-

hexane followed by subsequent reduction [165], as depicted in Scheme 4.2. Metal insertion using

dichlorotetrakis(dimethyl sulphoxide)ruthenium(II) as precursor yields the Ru-PNNP complex in

a further step [86]. In this study, only the R,R-version of the ligand (17) and the catalyst (18)

was prepared. 18 was applied as homogeneous catalyst for the conversion of acetophenone to

(S)-1-phenylethanol in the 2-propanol (hydrogen donor and solvent)/potassium propan-2-olate

(co-catalyst) system. In addition to the studies of Gao and co-workers [86, 166] (cf. Section

4.1.1), a detailed description of this catalyst including thermodynamic and mechanistic consid-

4 Modification of a Ruthenium Catalyst

Scheme 4.2: Preparation of the unmodified PNNP ligand and ruthenium complex.

erations has been carried out by Laue [88]. He pointed out that a high enantiomeric excess is

contrary to a high yield (cf. Section 2.4.2), resulting in a decreased ee of 92 % (intermediate

maximum ee: 96 %) when the reaction is run to 98 % yield. Testing 18 confirmed a maximum

excess of (S)-1-phenylethanol of 95–96 % and revealed the high sensitivity of the catalyst to

moisture and air. Thus, a “blow off“ of acetophenone to increase the final ee and the reaction

rate was tested by applying an argon overflow in an open flask. Even though this approach was

effective, it is, of course, not an optimal answer to the problem.

4.2.1.2 Iridium-Diphosphine/Diamine

In order to test further applications, the PNNP ligand was used in conjunction with an iridium

precursor as catalyst for ATH operations performed in aqueous media. Compound 17 and

[IrHCl2(COD)]2were dissolved in toluene or tetrahydrofurane (THF) and stirred for a moderate

period of time before the hydrogen donor (sodium formate/water) and the substrate were added.

The reaction could also have been performed via in situ generation of the catalyst in water, but

an organic solvent was preferred since in water the catalyst tended to adhere to the stirrer tool.

As reported by Gao and co-workers, the performance of the system without the addition of

phase transfer catalysts was rather poor [168]. By employing excess surfactants, it was tested if

an improvement in rate and enantioselectivity could be achieved. Furthermore, immobilization

was attempted using the formation of micellar emulsions. As shown in Table 4.1, the anionic

surfactant sodium dodecyl sulfate (SDS) almost completely inhibited the reaction whereas the

non-ionic surfactant Lutensol XA 60 indeed enhanced rate and ee. Nevertheless, these initial

attempts did not achieve the results reported by Gao et al. (Table 4.1, Entry 4).

Due to the rather low enantiomeric excess achieved with Ir-PNNP in sodium formate/water,

the potential for application as an immobilized catalyst in ATH operations was limited. As

pointed out in Section 2.2.1, eesin asymmetric catalytic processes should be >90 %, but might

be acceptable if they exceed 80 %; less than 70 % ee, however, is too low. The focus of im-

mobilization attempts therefore lies with the ruthenium catalyst. Nevertheless, an immobilized

non-enantioselective catalyst might be interesting for other applications, and optimization of

the reaction conditions might further improve the overall performance of the system.

4 Modification of a Ruthenium Catalyst

Table 4.1: Performance of Ir-PNNP.

Reaction conditions: acp 1.2 mmol, HCO2Na 1.8 mmol, water 20 mL, (R,R)-PNNP

0.03 mmol, [IrHCl2(COD)]20.013 mmol, toluene 2–3 mL, 50 ◦C.

Entry Additive Time Conversion eea

[h] [%] [%]

1 – 24 70 30

2 SDS (1 g) 8 <5 n. d.

3 Lutensol XA 60 (1 g) 24 80 54

4bPPNCl (5 mol%) 47 99 62

aexcess of the S-enantiomer, n. d. = not determined

btaken from [168]; acp 0.25 mmol, HCO2Na 1.25 mmol, water 2 mL, (R,R)-PNNP 0.0055 mmol,

[IrHCl2(COD)]20.0025 mmol, 60 ◦C

4.2.2 Strategies for Modification

To achieve the aim of covalent attachment of the ligand or catalyst to the aminated supports

(cf. Chapter 3), a modified, linker-containing version of the PNNP ligand had to be prepared.

As a basis for the decision on how to implement the linker via a new synthetic route, the

following considerations were made.

•The carboxyl group is the suitable linker which corresponds to the amino-functionalized

supports since coupling can be performed via standard amide formation procedures.

•The linker has to be introduced into the ligand as far as possible from the catalytic center.

•A certain distance between catalyst and support should be provided by a spacer.

•The compounds required for modification should be commercially available to keep the

number of synthetic steps as few as possible.

Due to the striking results achieved with the polysiloxane-supported catalyst by Liese et al.

(cf. Figure 4.2 b), the 5-position of compound 8and 9was considered the suitable site for

the introduction of a linker/spacer moiety. The number of applicable, commercially available

compounds, however, was limited, and 2-bromo-5-iodobenzoic acid (3) was chosen as the most

promising candidate. Thus, apart from the transformation of the carboxyl to an aldehyde group

as well as the introduction of diphenylphosphine in 2-position (cf. Scheme 4.2), the coupling of

a C4–C6chain carrying a carboxyl group would have to be performed as the key step.

Knochel, Cahiez, et al. reported the formation of organozinc halides which could be selec-

tively coupled with aromatic iodides even in the presence of further functional groups [174].

Encouraged by these results, it was attempted to insert zinc into bromo-hexanoic acid tert-

butyl ester and couple the intermediate zinc species with the model compound 1-bromo-4-

4 Modification of a Ruthenium Catalyst

iodobenzene. Unfortunately, the formation of the organozinc halide did not succeed, and an-

other strategy had to be applied. The selective coupling of a further C6-acid ester, namely

5-hexynoic acid tert-butyl ester (2), at the iodo functionality of 1-bromo-4-iodobenzene was

effectively performed via Sonogashira coupling. Hence, the strategy for the preparation of the

modified PNNP ligand entailed the coupling of an alkynyl acid with 2-bromo-5-iodobenzoic

acid or a derivative, which after further steps would provide the linker-containing analog to

compound 9.

4.3 Preparation and Use of the Modified PNNP

4.3.1 Synthetic Pathway

The synthetic route was divided into three major sections, the first of which was considered to

be the most challenging and accompanied by considerable losses. Almost all reactions required

strictly anhydrous conditions and an inert atmosphere. Scheme 4.3 shows the synthetic steps

required for the preparation of the first key intermediate, the “modified diphenylphosphinobenz-

aldehyde” (7). Since the designated linker — the carboxyl group of 5-hexynoic acid (1) — might

have hindered further synthetic steps and work-up procedures, a protective group was required.

The introduction of the easily cleavable “tert-butyl” group into 1was achieved via Steglich

esterification to yield 2. Commercially available compound 3was reduced to the correspond-

ing alcohol 4by treatment with borane dimethyl sulfide complex. Both intermediates 2and

4were linked via a palladium-catalyzed Sonogashira coupling reaction, yielding compound 5.

By treating 5with Dess–Martin periodinane, aldehyde 6was formed, which in the next step

was coupled with diphenylphosphine to form key compound 7. Due to losses on all stages, an

overall yield of 7of 16.5 % (not including 3→4) was obtained.

Using compound 8, the coupling with diphenylphosphine to yield 9resulted more efficient

than the reaction of 6under analogous conditions (Scheme 4.4). A new procedure, however, had

to be developed for the monosubstitution of (R,R)-diaminocyclohexane (10), since the (quite

sophisticated) procedure reported by Laue could not be followed successfully [88]. A yield of

72 % of compound 11 was achieved when dichloromethane (DCM) was used as solvent for the

first step — monosubstitution of dimaniocyclohexane yielding the intermediate (mono)imine at

reduced temperature, in high dilution, and with careful addition of 9— and methanol for the

second step — reduction by treatment with sodium borohydride — in a comparatively simple

procedure. The next step, the preparation of the second key intermediate (12), was performed

at room temperature by adding 7to a solution of 11 in DCM to form the imine, which again

was subsequently reduced by treatment with sodium borohydride in methanol.

Cleavage of the tert-butyl ester (12) by treatment with trifluoroacetic acid (TFA) yielded

the desired “modified PNNP” ligand (13), which was transformed to the ruthenium complex

4 Modification of a Ruthenium Catalyst

Scheme 4.3: Preparation of key intermediate 7.

Scheme 4.4: Preparation of key intermediate 12.

4 Modification of a Ruthenium Catalyst

(14) via reaction with dichlorotetrakis(dimethyl sulphoxide)ruthenium(II) (Scheme 4.5). Thus,

10 synthetic steps were required to prepare the modified Ru-PNNP complex so that it was ready

for immobilization. Although the average yield was ∼70 %, the overall yield of 14 was only

about 8 % (calculation based on the longest linear sequence of seven steps, thus three further

steps are not included).

Scheme 4.5: Preparation of the modified Ru-PNNP complex.

Metal insertion into compound 12 instead of acidic cleavage gave the “protected modified Ru-

PNNP“ complex (16), which was used to test the catalytic activity in homogeneous applications,

avoiding potential inhibiting effects of the carboxyl group. Reactions performed under equal

conditions indeed showed that modified (16) and unmodified Ru-PNNP (18) performed almost

equally as catalysts in the ATH of acp. Thus, the structural modification — the triple bond

especially could have been suspected to have an inhibiting effect — had no impact on the

catalytic center.

4.3.2 Attempts at Immobilization

Due to the more promising results of the ruthenium catalyst, the main focus of the immo-

bilization attempts lay with the generation of a solid polymer-supported Ru-PNNP system.

Generally, two pathways could be applied to build the supported catalyst: either coupling of the

modified complex 14 (Scheme 4.6) or coupling of the modified ligand 13 (Scheme 4.7) with

subsequent metal insertion. While the first approach would allow for better control (unspecific

metal–surface interaction could be largely avoided), the latter would provide a more flexible

platform, permitting not only the use of ruthenium, but, in principle, also iridium or rhodium

either for ATH operations or for other reactions. A limitation of the latter approach, however,

derives from the fact that the “pure” ligand is much more sensitive to oxidation than the com-

4 Modification of a Ruthenium Catalyst

Scheme 4.6: Coupling of modified Ru-PNNP.

plex. Hence, coupling and storing have to be performed even more carefully since an attached

oxidized species cannot be separated.

The coupling of carboxylic acids with amine compounds is preceded by activation of the

carboxyl group. A great number of activation agents have been developed in the con-

text of peptide coupling in organic synthesis, each of which is applicable to specific opera-

tions [175]. Coupling to the polymer chips — in this case membrane slices were used — was

tested with model reagent 3-(4-methoxyphenyl)propionic acid, activated by o-(benzotriazol-1-

yl)-N,N,N0,N0-tetramethyluronium tetrafluoroborate (TBTU) in a dimethyl formamide (DMF)

solution of diisopropylethylamine (DIPEA) at room temperature. Analysis via infrared (IR)

spectroscopy indicated the successful attachment.

Scheme 4.7: Coupling of modified PNNP.

Thus, initial attempts at the coupling of compound 14 to generate PB and PC-supported ver-

sions of the ruthenium complex 15 were performed using TBTU under the same conditions.

The supports were stained slightly yellow after the procedure, but a clear indication of a co-

valent attachment by IR analysis was not obtained, and the supports were not catalytically

active. As displayed in Table 4.2, the application of further coupling reagents was tested: 1,10-

carbonyldiimidazole (CDI), N,N0-dicyclohexylcarbodiimide (DCC), N,N0-diisopropylcarbodiimide

(DIC), either with or without an additive such as N-hydroxysuccinimide (NHS), hydroxybenzo-

triazole (HOBt), or 1-hydroxy-7-azabenzotriazole (HOAt), all without success. Inter-molecular

interactions between the carboxylic group and the metal center were considered a possible ex-

planation for the unsuccessful coupling. The second pathway — linkage of the PNNP ligand

(13) and subsequent metal insertion — therefore became the focus.

The insertion of ruthenium into the PNNP ligand (both original and modified version) was

performed in toluene. The polymer supports, however, were not stable in this solvent (even

4 Modification of a Ruthenium Catalyst

Table 4.2: Attempts at immobilization to aminated supports.

General conditions: anhydrous and inert, excess of DIPEA over specified amino

groups, room temperature, 16–24 h.

Entry Compound Support Amine loading Coupling reagent/

[

mol g−1]additive/solvent

1 Ru-PNNP PB 3.8 TBTU/DMF

2 Ru-PNNP PC 8.6 TBTU/DMF

3 Ru-PNNP PC 8.6 DCC/NHS/DMF

4 Ru-PNNP PC 8.6 DCC/HOBt/DMF

5 Ru-PNNP PC 8.6 DIC/HOAt/DMF

6 PNNP PB 20 TBTU/DMF

7 PNNP PC 14 CDI/DMF

8aPNNP - - DCC/NHS/THF

aproof of NHS-ester formation attempted

less under reflux conditions), and another solvent was required to prove that this strategy was,

in principle, valid. Using 17 as model ligand, isopropylic alcohol was found to be an effective

alternative to toluene, although it has a significantly lower boiling point. Nevertheless, all

attempts at linking 13 to one of the supports were unsuccessful; IR analysis did not indicate any

attachment, nor did the insertion of ruthenium or the use of these supports in conjunction with

the iridium precursor lead to a catalytically active system. At this point, it was assumed that the

failure of coupling was due to unsuitable procedures. This could not be definitively determined,

however, and finally the search for more appropriate coupling reagents and conditions was

abandoned.

4.3.3 Concluding Remarks

A modified version of the Ru-PNNP catalyst was prepared via a new 10-step convergent syn-

thesis route. Although many attempts were made, the attachment to surface-functionalized

polymer supports via amide formation reaction failed. A significant improvement in the overall

yield of less than ∼8 % in the preparation of compound 14 was not achieved, and upscaling

was considered incalculable, so only small amounts could be produced. Even if an appropriate

coupling procedure had been developed, the low overall yield was considered a major hindrance

for the detailed study of the catalyst properties and optimization within a reasonable time and

cost frame. Thus, taking into account the aim of this work, the decision was made to focus

on the use of the second catalyst system which represents the latest developments in the field

(Chapter 5). Further considerations on the efficiency of immobilized (PNNP-based) catalysts

are presented in Chapter 6.

5 Immobilization of a Rhodium Catalyst

The search for another complex which was suitable for immobilization and use as catalyst in the

ATH of prochiral substrates, resulted in the tethered tetramethyl cyclopentadienyl rhodium(III)

p-toluenesulfonyl-1,2-diphenylethylenediamine complex (henceforth tethered Rh-TsDPEN, Fig.

5.1, cf. section 2.4.3) which was presented by Wills et al. in 2005 [95]. It was shown to be a

highly enantioselective, active, and versatile catalyst for the transfer hydrogenation of various

ketones in TEAF, and from reports on the use of analogous catalysts, an excellent performance

in aqueous media was anticipated [176, 177]. The special feature of the ligand is the connection

(tether) between the coordinating η5-cyclopentadienyl moiety and the chiral backbone, which

contains two electron-donating nitrogen atoms. The resulting multidenticity provides a high

complex stability. Moreover, tethered versions of TsDPEN-based catalysts were demonstrated

to have significant advantages in terms of reaction rate and versatility over the untethered

variants for various substrates [178].

Figure 5.1: Wills´s “tethered Rh-TsDPEN” catalyst.

5.1 Synthetic Modification

5.1.1 Ligand Preparation

Analogous to the synthetic modification of the PNNP ligand (cf. Chapter 4), the introduction of

hex-5-ynoic acid tert-butyl ester as the protected spacer/linker moiety via Sonogashira coupling

reaction was chosen to create a modified version of the tethered TsDPEN ligand, ready for

attachment to aminated supports. For this a halide functionality at any site of the ligand was

required. Since the 4-position of the phenylsulfonyl moiety was favored for the introduction,

toluenesulfonyl was substituted for bromobenzene-sulfonyl. Two strategies for the preparation

5 Immobilization of a Rhodium Catalyst

of the ligand were employed, differing in the order of the synthetic steps. One consisted in first

coupling the diamine with bromobenzene-sulfonyl chloride to obtain the “bromo-functionalized

TsDPEN” (Scheme 5.1 b), which afterwards was coupled with hex-5-ynoic acid tert-butylester

via palladium-catalyzed Sonogashira cross coupling (Sch. 5.1 c).

Scheme 5.1: Introduction of hex-5-ynoic acid-tert-butylester— first Attempt.

Unfortunately, this approach did not lead to the desired product. The “bromo-functionalized

TsDPEN” (Sch. 5.1 b) was obtained in high yields, but cross coupling with the alkynyl in

the second step was not achieved. This was ascribed to coordination of the diamine to the

palladium catalyst, which led to inhibition of the latter. The reversed order, first the cross

coupling of bromobenzene-sulfonyl chloride with hex-5-ynoic acid tert-butyl ester and then the

reaction with the chiral diamine, failed in the very first step (Scheme 5.2).

Scheme 5.2: Introduction of hex-5-ynoic acid-tert-butylester— second attempt.

An applicable procedure for the synthesis of a linker-containing version of the tethered TsDPEN

ligand was finally developed by Keilitz and Haag [179]. According to an approach by Deng et al.,

mono Boc-protected aminosulfonyl-diphenylethylenediamine was used as the starting compound

for further modification [180]. The introduction of the (methyl-protected) carboxylic linker was

accomplished through the reaction of the “amino-functionalized TsDPEN” with methyl adipoyl

chloride (Scheme 5.3). In four further steps, the desired linker-containing ligand (20, Scheme

5.4) was obtained in ∼40 % overall yield (calculated on the basis of the seven linear steps of

the convergent route).

5.1.2 Catalyst Formation

The insertion of rhodium into the ligand could either be performed before the ligand was

coupled to the support (Scheme 5.4, path A) or afterwards (path B). In order to keep the

number of synthetic steps on the solid phase as small as possible and to avoid unspecific

binding of rhodium onto the surface, the coupling of the whole complex (path A) was favored.

5 Immobilization of a Rhodium Catalyst

Scheme 5.3: Introduction of methyl adipoyl chloride.

Nevertheless, the rhodium insertion could not be performed in methanol, as was done with the

unmodified ligand by Wills and co-workers [95], because of the resulting esterification of the

linker. Aprotic polar solvents were tested to carry out this reaction, and tetrahydrofuran (THF)

proved the most convenient. Methanol, however, could be used as solvent for the insertion

in path B, or when the preparation of the methyl-protected variant (23) was sought, e.g.,

for homogeneous applications of the catalyst. The resulting NMR spectra of 21 were more

complex than those of the methyl-protected version. Characterization of the desired substance

was therefore performed via high resolution mass spectrometry and infrared spectroscopy (see

Section 7.2.2).

Scheme 5.4: Preparation of the supported modified tethered Rh-TsDPEN catalyst (22).

After the successful preparation of compound 21, the carboxylic linker was activated with TBTU

in DMF at room temperature, according to standard amide coupling procedures. The supports

had to be prepared via acidic cleavage of the Boc protecting groups in HCl/2-propanol. Thus,

5 Immobilization of a Rhodium Catalyst

(a) Membrane (b) Sinter Chip (c) Beads

Figure 5.2: Supported Catalysts.

diisopropylethylamine (DIPEA) was added as a base to dissolve the hydrochlorides generated.

Activated 21 dissolved in DIPEA/DMF was shaken together with the deprotected supports for

20 hours, washed with organic solvents, and dried under reduced pressure. Scheme 5.4 depicts

the preparation of the S,S-version of the immobilized catalyst. The same procedure applying

the antipodal diamine yielded the R,R-version; the S,S-version, however, was prepared in larger

scale and was hence used exclusively in this study.

Figure 5.3: IR spectra as proof of coupling.

Successful coupling was indicated by the bright orange color of the supports after the procedure

was finished (Figure 5.2). Additionally, IR analysis of the catalyst-loaded PCs yielded phenyl

and sulfonamide signals at 1593/1496 cm−1and (1375/)1157 cm−1, respectively (Figure 5.3).

The amount of catalyst coupled to the surface was considered consistent with the amount of

rhodium which was quantified via ICP analysis (cf. Section 3.1.2). Samples were therefore

prepared via microwave digestion in aqueous HNO3/HCl solution. It was observed that the

5 Immobilization of a Rhodium Catalyst

amount of DIPEA added during the coupling procedure had a significant influence on the

amount of catalyst attached to the surface. Based on the number of amino groups determined,

four to six equivalents of DIPEA provided a catalyst loading that was more than twice as high

as with two equivalents.

5.2 Catalytic Testing

5.2.1 General Remarks

Both PC-supported versions (membrane slices and sinter chips) as well as the PB-supported

version of the modified tethered Rh-TsDPEN catalyst were used in ATH reactions. Generally,

tests to elucidate basal properties such as substrate scope or reuseability were performed with

PCs, since the handling was easier and an ordinary reaction set-up consisting of a standard

flask with a magnetic stirrer could be used. Hence, PCs could even be applied to parallel

experiments. The PCs were simply added to the reaction mixture and taken out for washing

and reuse using forceps without any loss of support material (Figure 5.4 a). Because of the

higher mechanical stability, the sinter chip version was favored over membranes, particularly

for testing the reusability. However, several drawbacks of the 2-dimensional supports made the

application of beads necessary:

•The functionalization of the support materials as well as the immobilization of the catalyst

was performed batchwise. As a result, all batches differed slightly in amine and catalyst

loading. The advantage of beads was that a major blend, providing a large amount of

highly reproducible samples, could be made from various batches.

•A related problem was the catalyst metering, which was less precise when using PCs,

even when they stemmed from one batch. Usually slices of equal geometric surface, e.g.,

1×1 cm2, were applied for one series experiment, but an equal amount of immobilized

catalyst was not necessarily provided, neither by this method nor by weighing the PCs,

due to the deviations of the specific surface. This problem, however, was less severe since

deviations were found to be small.

•Nevertheless, reproducible reaction conditions were a problem, since both PBs and PCs

have a low density. Through vigorous mechanical stirring, the beads were homogeneously

suspended in aqueous solutions, whereas the chips could only be used with magnetic

stirring, and floated on the water surface. A more precise adjustment of the stirring speed

and the achievement of a homogeneous suspension, which guaranteed highly reproducible

conditions, was only possible when beads were applied.

5 Immobilization of a Rhodium Catalyst

(a) PC-supported catalyst. (b) PB-supported catalyst.

Figure 5.4: Use of supported catalysts.

ATH experiments with the PB-supported catalyst (22 a) were performed in a jacketed 100-

mL flask equipped with a mechanical stirrer, a thermostat, an insight temperature sensor,

and an optional argon sweep gas installation (Figure 5.4 b). As depicted in Scheme 5.5, the

focus of all experiments performed with the supported tethered Rh-TsDPEN catalysts lay on

the enantioselective reduction of prochiral ketones to chiral alcohols and the use of water as

a potentially “green” reaction medium [71, 72]. The ATH of acetophenone performed in an

aqueous solution of sodium formate at 40 ◦C was used as the benchmark reaction for the testing

of the PC and PB-supported versions of the catalyst.

In contrast to the PNNP-based catalyst, by applying the S,S-version of the tethered Rh-

TsDPEN the S-enantiomer of the product was formed in excess. Both enantiomeric excess and

conversion were determined via GC analysis. For the determination of the conversion, samples

were taken from the reaction mixture with a syringe, beads (when used) were filtered off, and

the samples were diluted in ethanol and injected into the gas chromatograph. Usually, the ee

was only determined after the reaction was finished. Two apparatus equipped with different

chiral columns were used for this analysis, and slightly different ee values were obtained, varying

between 98 % and 99 %.

Scheme 5.5: ATH of aryl ketones in an aqueous medium.

5 Immobilization of a Rhodium Catalyst

5.2.2 Initial Experiments

5.2.2.1 ATH of Acetophenone

In order to guarantee complete wetting of the supports without exceeding the usual range of the

substrate to catalyst ratio (S/C), a relatively high dilution was required in all experiments. The

PC-supported catalyst (22 b) was used at a substrate concentration of 0.1 mol L−1with an S/C

between 200 and 450. Under these conditions, 22 b provided a higher conversion than both

the unsupported analog (23) and a soluble hyperbranched PEG-supported version within the

same time frame (Table 5.1) [179]. Under the standard conditions for the use of catalyst beads

(see Section 5.3), however, no comparison between the PB and the non-supported catalyst was

possible. Apparently, the substrate concentration of 0.04 molL−1was too low, and after a

time the catalyst tended to adhere to the stirrer tool. These results are remarkable because

a significant limitation of the rate due to mass diffusion effects should have been expected

from a water/oil/solid triphase reaction. However the opposite effect, a rate enhancement,

was observed when the immobilized catalyst was used. This acceleration is presumably derived

from an accumulation effect of the support material which causes higher concentrations of the

reactants in the microenvironment of the catalyst (cf. Section 2.4.6.2).

Table 5.1: Performance of different versions of the modified tethered Rh-TsDPEN.

Reaction conditions: acp 1 mmol, HCO2Na 5 mmol, water 10 mL, S/C ≈430, 40 ◦C.

Entry Catalyst Time Conversion eea

[h] [%] [%]

1 soluble PEG-supported 4 18 98

2 PC-supported (22 b) 4 57 99

3 non-supported (23) 4 37 98

aexcess of the S-enantiomer

Generally, the results obtained with the PC-supported catalyst in organic media were far inferior

to those obtained in water. In pure TEAF, the supports were tinged black, and the catalyst

was almost inactive. When TEAF was diluted with DCM, a somewhat higher conversion was

achieved, but it was still not in the range of that obtained with the sodium formate/water

system; under otherwise equal conditions, the TEAF/DCM medium provided a conversion of

<10 % in six hours whereas in HCO2Na/H2O almost full conversion was achieved. The reaction

proceeded even more slowly in pure 2-propanol. These results may support the above-mentioned

suggestion of a potential accumulation effect. If the substrate (at a low concentration) is

completely dissolved in the organic solvent and not accumulated in the environment of the

catalyst, the reaction may become diffusion-controlled. The addition of methanol as a co-

solvent to the sodium formate/water system, however, slightly enhanced the rate and did not

5 Immobilization of a Rhodium Catalyst

Figure 5.5: Reproducibility of the standard experiment.

Reaction conditions: acp 0.04 M, HCO2Na 0.4 M, total vol-

ume of water/HCO2Na/acp 100 mL, PB-supported catalyst 2 g

(≈0.113 mM Rh), 40 ◦C, 1 200 rpm, argon overflow.

affect the enantioselectivity.

In contrast to the Ru-PNNP catalyst, no induction phase was observed with the immobilized

tethered Rh-TsDPEN catalyst in aqueous media. This may indicate that the formation of the

active species proceeds rapidly compared to the rate-determining step, i. e., the hydrogen trans-

fer [112]. When the bead-supported catalyst (22 a) was applied under the standard conditions,

high reproducibility of the conversion over time was achieved. In four independent experiments,

the maximum standard deviation was about 3 % (Figure 5.5). Other conditions, however, pro-

vided less reproducible results (see below). A significant difference between reactions performed

in air and in an argon atmosphere (closed flasks without gas disengagement in either case) was

not observed.

5.2.2.2 Recycling

Catalyst recycling was performed either with or without washing the supports with water and/or

organic solvents after each run. A high decrease in activity was observed within the first three

cycles when the supports were repeatedly used in the standard sodium formate/water system

without washing or further treatment (Table 5.2). This decrease is not fully explained by the

metal leaching, which was determined via ICP-MS analysis. A leaching of 56 ppm (rhodium per

total amount of substrate), which corresponds to approximately 2.5 % of the initial rhodium

content, was observed after the first run. After two runs, approximately 4 % of the rhodium had

5 Immobilization of a Rhodium Catalyst

Table 5.2: Recycling of PC-supported catalyst without intermediate washing.

Reaction conditions: acp 1 mmol, HCO2Na 5 mmol, water 10 mL, PC-supported

catalyst, S/C ≈430, 40 ◦C.

Run Time Conversion eeaLeaching

[h] [%] [%] [ppm]b[%]c

1 8 100 98 56 2.5

2 16 100 98 35 1.6

3 8 53 98 14 0.6

4 16 68 98 n. d. n. d.

5 8 55 98 n. d. n. d.

6 16 59 98 12 0.6

aexcess of the S-enantiomer

bdetected amount of rhodium per total amount of substrate, n. d. = not determined

cdetected amount of rhodium per initial amount of catalyst, n. d. = not determined

leached into the liquid phase, but the conversion dropped by nearly 50 % in the third run (cf.

entry 1 and 3, Table 5.2). In further runs, the amount of rhodium loss was significantly lower.

Thus, unspecifically bound 21 or rhodium is assumed to have washed away during the first

runs. The ee values were determined to be 98 % in all runs, and thus proved to be independent

of the conversion in the investigated range.

Table 5.3: Recycling of PC-Supported catalyst with intermediate washing.

Reaction conditions: acp 1 mmol, HCO2Na 5 mmol, water 10 mL, PC-supported

catalyst, S/C ≈430, 40 ◦C.

Run Time Conversion eea

[h] [%] [%]

1 4 57 99

2 4 51 98

3 4 44 99

4 4 43 99

5 16 79 99

6 4 27 98

7 4 23 98

aexcess of the S-enantiomer

A decrease in catalytic activity was also observed in recycling experiments in aqueous sodium

formate with intermediate washing and drying (Table 5.3). In a reaction that was run for

four hours to a conversion of 57 % in the first run, a drop of 6 % and 7 % (referred to 100 %

conversion) was observed in the second and third run, respectively. After a fourth run with

almost identical conversion, the fifth run was performed overnight, achieving 79 % conversion

within 16 hours. The subsequent sixth run only achieved 27 % conversion within four hours.

5 Immobilization of a Rhodium Catalyst

Again, ee values were stable throughout the experiment. From these results, it was concluded

that washing the supports did not have a significant effect on the catalyst reusability. High losses

of activity were observed in either case when the reaction was run to high conversion. Thus,

a catalyst deactivation due to byproducts of some kind was considered a possible explanation

and investigated in further experiments (see Sections 5.2.3 and 5.2.4). The use of argon as

protective gas in closed flasks could again not improve the performance.

5.2.2.3 Substrate Scope

The versatility of the catalyst was tested in ATH reactions with different aryl ketones as sub-

strates. High values of the enantiomeric excess were achieved in all cases, ranging from 87 %

to 98 % (Table 5.4). In contrast, the conversion within six hours reaction time differed more

strongly, presumably due to different solubilities in water and the functional matrix of the sup-

port material. This was definitely observed with 40-nitroacetophenone (Entry 2), a solid with

low solubility in water.

Table 5.4: Scope of tested substrates.

Reaction conditions: substrate 1 mmol, HCO2Na 5 mmol, water 10 mL, PC-

supported catalyst, S/C ≈430, 40 ◦C.

Substrate

Entry Indication R1R2Time Conversion eea

[h] [%] [%]

1 acp H H 6 97 98

2 - 40-NO2H 6 36 87

3 - 40-Cl H 6 71 95

4 - 40-OMe H 6 79 97

5 - H Me 6 58 98

aexcess of the S-enantiomer

5.2.3 Effect of Temperature and Atmosphere

For the determination of temperature and atmosphere effects on the ATH of acetophenone,

the PB-supported catalyst was used. A normal correlation of temperature and conversion rate

was observed in experiments conducted between 10 ◦C and 40 ◦C in the closed 100-mL flask

without applying any protective gas or sweep gas. Hence, within this range an increase in

temperature effected an increase in rate. At 50 ◦C the reaction started faster than at 40 ◦C but

was slower at the end; at 60 ◦C a significantly lower initial rate was observed, and the reaction

leveled off at about 40 % conversion (Figure 5.6 a). The retarding effect on the conversion rate

at temperatures above 40 ◦C presumably derives either from inhibition due to interaction with

byproducts which emerge only at higher temperatures, or from catalyst decomposition.

5 Immobilization of a Rhodium Catalyst

(a) Temperature variation.

Reaction conditions: acp 0.04 M, HCO2Na 0.4 M, PB-

supported catalyst 1.5 g (∼0.105 mM Rh.), 1 200 rpm

(b) Atmosphere variation.

Reaction conditions: acp 0.04 M, HCO2Na 0.4 M,

PB-supported catalyst 2 g (∼0.113 mM Rh), 40 ◦C,

1 200 rpm.

Figure 5.6: Variation of temperature and atmosphere within the flask.

Reuse experiments indicated that decomposition was less probable, because the employment of

catalyst beads taken from two experiments, one conducted at 40 ◦C, the other conducted at

60 ◦C, gave similar results in terms of conversion and ee in the corresponding runs at 40 ◦C

after recycling. Nevertheless, the rate was significantly lower than in the first run at 40 ◦C.

Two effects that decreased the catalyst activity had to be distinguished:

•a temperature-related effect which was reversible by washing the supports, and

•another effect which was not reversible (at least not by washing the supports, see above).

However, it is known that CO2emerges as a byproduct during the reaction, and a CO2insertion

[181] as well as the inhibiting effect of CO2headspace gas in ATH reactions performed in

aqueous solutions of sodium formate [112] have been reported. Furthermore, the generation of

CO from formic acid in ATH reactions has been noted [116, 182].

In order to remove the emerging gases, an argon overflow was applied. As a result, the

conversion rate at 60 ◦C was significantly higher than without gas disengagement, and full

conversion was successfully achieved (Figure 5.6 a). Nevertheless, apart from a trend toward

higher reaction rates at temperatures between 50 ◦C and 70 ◦C when employing the sweep gas,

fully consistent and reproducible data could not be generated. This was ascribed to unstable

and insufficient gas disengagement, and therefore a gas entrainment impeller was used for more

efficient dispersion of argon into the reaction mixture. Unfortunately, the catalyst beads could

not be suspended homogeneously under these conditions because the buoyancy was too strong.

Reliable data in experiments above 40 ◦C could therefore not be obtained.

For a more detailed study of the generation of carbon monoxide and carbon dioxide, samples

taken from the gas phase in the flask during a reaction at 60 ◦C were analyzed. A significant

5 Immobilization of a Rhodium Catalyst

amount of CO could not be detected; only the generation of CO2from sodium formate was

proven by GC analysis even when no substrate was present. The impact of carbon dioxide

on the reaction was investigated by comparing the conversion in experiments conducted at

40 ◦C with argon overflow, in a closed flask, and in a CO2atmosphere (CO2sweep gas),

respectively. Surprisingly, an increasing amount of CO2in the reactor effected an increase in

reactivity rather than inhibiting the reaction (Figure 5.6 b). This was presumed to derive from

the reduced solution basicity due to solubilized carbon dioxide, and in fact pH values at the end

of the reactions were significantly lower when no gas disengagement was carried out (closed

reactor: pH 8.6) or when CO2was used as sweep gas (pH 7.1), compared to experiments with

argon overflow (pH 9.4).

Although it could not be confirmed via GC analysis, it is most likely that the inhibition at

temperatures above 40 ◦C is due to the generation of CO. The carbon monoxide is believed to

coordinate quickly to the metal center, inhibiting the catalytic activity and hence the production

of additional CO. Blacker and Thompson, who studied the use of a (non-tethered) Rh-TsDPEN

catalyst in the TEAF system on a larger scale, stated that the amount of carbon monoxide

produced was small and that coordination to the catalyst was reversible [116, 183]. This

would confirm the above findings and assumptions. However, a temperature dependency of the

assumed production of CO has not been mentioned in the literature (to the best of the authors´

knowledge); only the need for carbon dioxide removal, which is apparently not required when

22 is used as catalyst, has been stated [116].

5.2.4 pH Dependency of Activity, Enantioselectivity, and Reusability

Both enantioselectivity and activity in ATH reactions catalyzed by transition metal complexes

based on TsDPEN or its derivatives have been shown to be pH-dependent [184, 185]; a few

studies have been conducted to elucidate the reasons for this [112, 183]. In order to investi-

gate the influence of pH on the catalyst performance and to determine the optimal reaction

conditions, the initial solution pH was adjusted to values between 1 and 12 by applying sodium

formate, formic acid, and mixtures of both, or adding NaOH(aq) or HCl(aq) to further increase

or decrease the pH, respectively (see Table 5.5). For these tests, the sinter chip immobilized

version of the catalyst was used due to its easier handling compared to the micro particles.

In almost all experiments the pH increased with increasing conversion. Only when the initial

pH was adjusted to 9.5 by adding a small amount of NaOH (∼1µmol) to the sodium for-

mate/water solution was a decrease of the pH to a value of 8.5 observed during the reaction

(Table 5.5, Entry 9).

The optimal initial pH in terms of activity was found to be between 3 and 4, giving values of

about 4 to 5 at the end of the reactions (Entries 3 to 5). By performing analogous experiments

following the more sophisticated approach of applying the PB-supported version of the catalyst,

5 Immobilization of a Rhodium Catalyst

Table 5.5: ATH of acetophenone under different pH conditions.

Reaction conditions: acp 1 mmol, HCO2Na or HCO2H/HCO2Na 10 mmol, water

10 mL, PC-supported catalyst, S/C ≈200, 40 ◦C.

2 h 4 h

Entry HCO2C/HCO2Na pH Conv. TOFapH Conv. eebLeachingc

[mmol] [%] [h−1] [%] [%] [%]

1 (+ HCl)10/0 1.0 0 0 1.1 3 11 2.4

2 10/0 1.7 61 62 2.1 83 75 n. d.

3 7/3 3.1 89 86 3.7 >99 97 n. d.

4 5/5 3.5 91 88 4.3 >99 98 n. d.

5 3/7 3.9 94 94 5.1 >99 98 1.9

6 1/9 4.5 81 80 7.0 >99 98 n. d.

7 0.03/9.97 5.6 67 68 8.2 93 98 n. d.

8 0/10 7.5 61 60 8.4 81 97 2.2

9 0/10 (+ NaOH) 9.5 59 56 8.5 79 98 n.d.

10 0/10 (+ NaOH) 12.3 8 9 12.3 10 90 1.8

11d5/5 3.5 92 99 4.4 >99 98 n. d.

12e5/5 3.5 61 64 4.1 92 97 n. d.

aFor the calculation of TOF values, the concentration of applied catalyst was determined from the mass of

each sinter slice.

bexcess of the S-enantiomer

cdetected amount of rhodium per initial amount of catalyst, n. d. = not determined

dusing catalyst chip from entry 1

eusing catalyst chip from entry 10

it was confirmed that 3.9 was the optimal pH value rather than 3.5. Maximal ees(98 %) were

achieved at initial pH values between 3.5 and 5.6 (Entries 4 to 7).

The reaction was extremely slow at both very low and very high pH levels, and also lower ees

were obtained. In order to investigate a potential deactivation effect from strongly acidic and

strongly basic media, the catalyst chips which had been used under those conditions (Entry 1,

pH 1 and Entry 10, pH 12.3, respectively) were recovered and applied in moderately acidic

HCO2H/HCO2Na/H2O mixtures (Entries 11 and 12, respectively, pH 3.5). It was found that

the strongly acidic conditions, which inhibited the conversion in the first run, had no significant

influence on the catalyst performance in the second run when the pH was adjusted to 3.5;

values of conversion and ee were in agreement with those obtained with the non-pretreated

catalyst or even somewhat better (Entry 11 cf. Entry 4). The catalyst chip which had been

used in the strongly basic HCO2Na/NaOH/H2O mixture (pH 12.3) also showed an increased

activity at a pH of 3.5, but it lagged behind that of the non-pretreated catalyst (Entry 12 cf.

Entry 4). This indicates that the inhibition under acidic conditions is completely reversible,

whereas in basic media it is only partly reversible.

Increased metal leaching under basic conditions was considered a possible explanation for

5 Immobilization of a Rhodium Catalyst

Figure 5.7: Multiple catalyst reuse under different pH conditions.

Reaction conditions: acp 0.1 M, HCO2Na or HCO2H/HCO2Na

0.5 M, water 2.5 mL, PC-supported catalyst, S/C ≈200, 40 ◦C,

reaction time 2 h per run.

these results. Thus, the rhodium content of the solutions of four experiments performed under

different pH conditions was analyzed via ICP-OES analysis. The amount of rhodium in the

solutions was found to be between 1.8% and 2.4 % of the initial amount on the catalyst

chips. Since a correlation between leaching and solution pH was not evident (cf. Entries 1,

5, 8, and 10), the reduced reactivity under basic conditions is more likely to derive from an

inactive catalyst species [186]. The increase of the pH value, likely due to the generation of

hydroxide ions from formate, is accompanied by an increasing conversion of acp. An irreversible

or only partly reversible inhibiting effect from OH−ions is presumed to explain both the reduced

reusability when the reaction was run to high conversion in HCO2Na/H2O (cf. Section 5.2.2.2)

and the low reactivity under strongly basic conditions.

The recyclability was further tested under different pH conditions. The data points shown in

Figure 5.7 represent conversion (full lines) and ee (dotted lines) of three series of experiments,

each with eight consecutive runs performed under equal conditions. The initial pH values were

adjusted to 2.1, 3.9, and 7.4. After two hours reaction time, the catalyst chips were taken

out, washed, dried, and used for the next run, while conversion and ee were determined via

GC-MS analysis. The best performance in terms of recyclability was achieved at an initial

pH of 3.9, providing a TTN of 1 255 over eight runs, with the catalyst still active. This

good reusability under moderately acidic conditions was accompanied by the highest conversion

and enantioselectivity (constant ∼98 % ee). The total rhodium leaching over eight runs was

5 Immobilization of a Rhodium Catalyst

Scheme 5.6: ATH of aryl ketones in formate/water.

determined to be around 5%. In the acidic mixtures, the conversion increased from the first

to the second run and declined in the subsequent runs. This increase was ascribed to the basic

conditions of the coupling procedure which may have affected the catalyst performance in the

first run.

Taking into account the results obtained from varying the temperature and atmosphere

and the increase in pH observed during the course of the reaction, a more complete picture

of the process can be drawn (Scheme 5.6). Thus, besides the reduction of acetophenone

to phenylethanol, in which simultaneously carbon dioxide and hydroxide are generated from

formate and water, there are probably two side reactions occurring. One is the decomposition

of formate in the presence of water to molecular hydrogen and carbon dioxide, and the other

is decarbonylation, which is assumed to occur only at temperatures above 40 ◦C when 22 is

used as the catalyst. These side reactions can only indirectly diminish the yield of the desired

product (catalyst poisoning by carbon monoxide, as discussed), since the aryl ketone substrate

is not involved.

5.2.5 Effect of the Concentrations of Acetophenone and Sodium Formate

The linear relationship between rate and substrate concentration was shown to be valid up to

about 0.1 mol L−1of acp when 1.5 g of the PB-supported catalyst were applied (Figure 5.8 a).

In this concentration range the beads were homogeneously or quasi-homogeneously suspended

(Figure 5.9 a). At higher substrate concentrations, though, the reaction rate decreased due to

bead agglomeration within the emulsified acp. The size of the agglomerates increased with

increasing concentration and decreasing rate. At an acp concentration of about 0.14 mol L−1

firm “balls” of about 0.75 cm in diameter were formed, and almost no conversion occurred

(Figure 5.9 b). The inhibition due to aggregation impeded the investigation of a substrate

inhibition effect on the molecular level, which has been observed with the unsupported variant

5 Immobilization of a Rhodium Catalyst

at acp concentrations above 1.3 mol L−1[112].

(a) Variation of the concentration of acp.

Reaction conditions: constant 10/1 ratio of HCO2Na

to acp, PB-supported catalyst 1.5 g (∼0.105 mM Rh),

40 ◦C, 1 200 rpm.

(b) Variation of the concentration of HCO2Na.

Reaction conditions: constant substrate concentration

0.04 M, PB-supported catalyst 2 g (∼0.113 mM Rh),

40 ◦C, 1 200 rpm, argon overflow.

Figure 5.8: Impact of the concentrations of substrate and hydrogen donor.

Figure 5.8 b shows the conversion of acp over time at different hydrogen donor to substrate

ratios. It is apparent that the acceleration of the reaction due to an increasing excess of sodium

formate follows a saturation curve. Thus, the enhancement of the reaction rate brought about

by the doubling of the amount of sodium formate — from a hydrogen donor to substrate ratio

of 5 to a ratio of 10 —is rather small. Ratios above 10 were therefore not investigated.

(a) Even suspension at lower concentra-

tions.

(b) Aggregation of beads at higher con-

centrations.

Figure 5.9: Influence of the concentration of acetophenone on the suspension of beads.

5.3 Kinetic and Mechanistic Investigations

A systematic series of experiments were performed in order to determine the kinetic parameters,

taking into account the above-mentioned results that reveal the rate of the ATH reaction under

different conditions. As introduced in Section 2.4.4, intensive experimental and theoretical

5 Immobilization of a Rhodium Catalyst

investigations of the asymmetric transfer hydrogenation of acetophenone in aqueous media were

reported by Liu, Xiao et al. [112]. A first-order rate dependence with respect to the substrate

as well as to the catalyst concentration was proposed and experimentally confirmed for an acp

concentration of up to 1.3 mol L−1and excess sodium formate using a non-tethered η6-arene

ruthenium(II)-p-toluenesulfonyl-1,2-diphenylethylenediamine complex in one-phase water/DMF

solution [74, 112]. This rate dependence (Equation 2.4.2) served as the starting point for

the present study: the second-order model was applied to fit the experimentally determined

concentration of acp over time. All experiments were exclusively performed using the bead-

supported version of the catalyst (22 a).

5.3.1 Basic Reaction Conditions

In the standard experiment, 1.5 g of the PB-supported catalyst with a rhodium content of

7µmol g−1was used, and a tenfold excess of sodium formate (0.4 M) over acetophenone

(0.04 M) in neat water was applied. The flask was kept closed except when samples were

taken, and no gas disengagement or protective gas was used. A conversion of more than 99 %

with an enantiomeric excess of 98 % was achieved after six hours (Figure 5.10a). The solution

pH increased during the reaction from about 7.8 to 8.6. Kinetic and activation parameters were

calculated from averaged results of two experiments performed under equal conditions.

Despite the apparent differences between 22 a and the catalytic system used by Liu and co-

workers [112], the model provided good agreement with the experimental results (Figure 5.10 b).

Accordingly, a linear dependence of the reaction rate on the acp concentration (Figure 5.10 c)

as well as a linear dependence of the reaction rate on the amount of catalyst (Figure 5.10 d)

were found in the investigated range. From these experiments, a value of 1.9±0.2 L mol−1s−1

was calculated for the rate constant k.

The activation parameters were calculated from experiments conducted between 10 ◦C and

40 ◦C, since at higher temperatures disturbing effects impeded reliable results (see above). Fig-

ure 5.11 depicts the Eyring plot of the investigated temperature range, from which an activation

enthalpy ∆H‡= 71 ±1 kJ mol−1and an activation entropy ∆S‡=−15 ±2 J mol−1K−1were

obtained. The values of kand ∆S‡differ significantly from those found for the homogeneously

catalyzed ATH reaction [112]; a discussion follows in Section 5.3.3.

5.3.2 Acidic Reaction Conditions

A comparison of the conversion profiles of the standard experiment and an analogous experiment

performed at pH 3.9 (initial value) demonstrated again the significantly higher reaction rate

under acidic conditions (Figure 5.12a). Furthermore, the agreement of computed data and

the experimentally determined concentration curve indicated that the first-order dependency

of substrate and reaction rate (Equation 2.4.2) was still valid (Figure 5.12 b). Under the

5 Immobilization of a Rhodium Catalyst

(a) Conversion of acp and ee of (S)-1-phenylethanol. (b) Experimentally determined and computed concen-

tration of acp.

Figure 5.10: ATH of acp under moderately basic (standard) conditions.

Reaction conditions: acp 0.04 M, HCO2Na 0.4 M, PB-supported catalyst 1.5 g (∼0.105 mM

Rh), 40 ◦C, 1 200 rpm, total reaction volume 100 mL.

Figure 5.11: Eyring plot for basic conditions.

5 Immobilization of a Rhodium Catalyst

acidic conditions, a value of 3.7 ±0.3 L mol−1s−1for the rate constant kwas computed. This

is approximately twice as high as the value determined under the standard conditions. The

solution pH increased during these experiments from 3.9 to about 4.6.

From an Eyring plot for temperatures between 10 ◦C and 40 ◦C at an initial pH of 3.9, a

value of 81 ±1.5 kJ mol−1for ∆H‡and a value of +24 ±5 J mol−1K−1for ∆S‡was determined.

The entropy of activation was thus significantly higher than that obtained under the standard

conditions; possible mechanistic interpretations are discussed in the following section.

(a) Conversion of acp.

(Standard conditions as in Fig. 5.10).

(b) Experimentally determined and computed concen-

tration of acp.

Figure 5.12: ATH of acp under pH-optimized conditions.

Reaction conditions: acp 0.04 M, HCO2H 0.12 M, HCO2Na 0.28 M, PB-supported catalyst 1.5 g

(∼0.105 mM Rh), 40 ◦C, 1 200 rpm, total reaction volume 100 mL.

Figure 5.13: Eyring plot for acidic conditions.

5 Immobilization of a Rhodium Catalyst

5.3.3 Mechanistic Considerations

The second-order kinetics found for the ATH of acetophenone in sodium formate/water solution

catalyzed by 22 are in agreement with the findings for the homogeneous untethered Ru-TsDPEN

catalyst [112]. The intensive investigations on the ruthenium-catalyzed reaction carried out by

Liu and co-workers afforded a well-founded mechanistic concept which explains the variations

of the reactivity and enantioselectivity under different reaction conditions (cf. Scheme 2.5

in Section 2.4.4.2). According to this concept the hydrogen transfer proceeds via a highly

organized transition state (TS) in which the substrate is coordinated to the metal hydride and

the nitrogen proton at the same time (Figure 5.14 a). Under acidic conditions, ring-opening of

the chelate ligand due to protonation of the sulfonamide nitrogen is presumed to complicate

the formation of the transition state, leading to the observed decrease in reaction rate and

enantioselectivity (Figure 5.14 b). These suggestions were supported by a spectroscopic study

of the complex under acidic conditions. The low reactivity under basic reaction conditions,

however, is explained by an inactive catalyst species formed through the coordination of a

hydroxide ion (Figure 5.14 c).

(a) TS under neutral conditions. (b) TS under acidic conditions. (c) Inactive species under

basic conditions.

Figure 5.14: Ru-TsDPEN species under different conditions.

Because of the concordant rate laws of both catalytic systems under neutral and moderately

basic conditions as well as the decline in catalyst activity under strongly basic conditions ob-

served in both cases, an analogous mechanism as proposed for Ru-TsDPEN seems likely for

the (immobilized) tethered Rh-TsDPEN catalyst (cf. the middle part of Scheme 5.7 and Figure

2.5). Thus, despite the heterogenization, the different metal and the structural differences, the

catalyst is believed to operate in the same way as the homogeneous non-tethered ruthenium

analog when those conditions are applied. Nevertheless, in contrast to the Gaussian-like dis-

tribution of the TOF around pH 7 which has been found for ruthenium, rhodium, and iridium

catalysts based on the TsDPEN ligand [184, 186, 187], the activity of the tethered catalyst

applied here peaked at pH 3.9, and an almost constantly high ee was obtained in a pH range

between 3.5 and 9.5 (Figure 5.15).

Under the assumption that the structural differences of the ligand do not cause a significant

5 Immobilization of a Rhodium Catalyst

Scheme 5.7: Proposed mechanism under different pH conditions.

Remarks: L might be H2O, which would lead to a +1 charge, or HCO−

2; dashed bonds are drawn

either when undefined interactions between the reactants in the transition state are meant or

when there is no certainty about the bond — likewise the atoms/molecules are put in parentheses.

5 Immobilization of a Rhodium Catalyst

Figure 5.15: Activity and enantioselectivity over pH.

change in the basicity of the nitrogen donor atoms (particularly in the pKbof the sulfonamide-

N), protonation and ring-opening will occur with the tethered rhodium complex in an acidic

medium as proposed for the analogous ruthenium complex. However, as long as only one of

the nitrogen atoms is protonated — it is most likely that the anionic tosyl-N will be protonated

first — the catalyst might be enabled by the tether to keep its steric orientation. An opened ring

with a linked cyclopentadienyl moiety might possibly facilitate the formation of the transition

state instead of impeding it. These considerations are supported by the higher entropy of

activation, which suggests a more flexible transition state under moderately acidic conditions

than under basic reaction conditions. Additionally, the tether between the chiral backbone and

the cyclopentadienyl moiety provides a plausible explanation for the complete reversibility of

the catalyst inhibition under strongly acidic conditions after recovery; even when both nitrogen

atoms are protonated and an inactive species is formed, the ligand is connected to the central

metal via η5-coordination (see Scheme 5.7 j). The strength of the η5-coordination has been

demonstrated in a study presented by Perutz et al., where the analogous, but unmodified

and non-tethered, Rh-TsDPEN catalyst dissolved in methanol was shown to form triply bridged

dimers of rhodium-tetramethylcyclopentadienyl h[{Cp∗Rh}2(µ-H)(µ-Cl)(µ-HCOO)]+iupon the

addition of 10 equivalents of formic acid [183]. Thus, the connection to the chiral backbone is

released under those conditions whereas the connection to the aryl moiety is not. Consequently,

the tethered aryl moiety is assumed to prevent metal leaching and to facilitate reversion to the

chelate complex when conditions are less acidic. Generally, strongly acidic conditions adjusted

by the addition of HCl(aq) might effect also inhibition due to chloride coordination to the

central metal, but the presence of chloride anions was not mandatory for inhibition of the

tethered catalyst, since the activity also decreased without adding HCl(aq) when the ratio of

formic acid to sodium formate exceeded 3:7 (cf. Table 5.5, Section 5.2.4).

5 Immobilization of a Rhodium Catalyst

(a) TS with ring-opening

and participation of wa-

ter.

(b) TS with ring-opening and

participation of hydronium.

with participation of hy-

dronium.

Figure 5.16: Alternative transition states under acidic conditions.

All attempts to prove the existence of a protonated species via spectroscopic methods, however,

were unsuccessful. NMR spectra of dissolved complex 21 in the presence of different equivalents

of TFA in CDCl3were difficult to interpret, but suggested complete decomposition rather than

protonation of a nitrogen atom. The difficulties of NMR measurements in this context have

been stated in the literature [112]; according to the results reported there, a more sophisticated

approach is required to prove the protonation of the catalyst. However, infrared spectra of 21,

which were recorded in the presence of formic acid, also did not provide any reliable evidence

of a protonated species.

Nevertheless, some further considerations may be useful in discussing whether ring-opening

due to protonation provides a satisfactory explanation for the values of the activation parameters

determined from the kinetic experiments. Ring-opening under acidic conditions might provide a

transition state which is less ordered than that under basic conditions. For the assessment of the

entropy of activation, however, the degrees of freedom of the reactants prior to the formation

of the activated complex also have to be discussed. Protonation under acidic conditions which

leads immediately to ring-opening would provide a “flexible” catalyst which loses “freedom”

upon formation of the transition state. Thus, freedom in absolute terms would not be gained.

Nonetheless, protonation of the tosyl-N does not inevitably lead to the release of coordination

to the central metal and attended ring-opening [112, 188]. It is possible that the coordinative

bond between rhodium and the protonated tosyl-N is not substituted by coordination of a

formate anion or a water molecule until the activated complex is formed. In this case, the loss

of degrees of freedom due to the interactions between substrate and catalyst in the transition

state might be compensated to some extent by the break of the bond between rhodium and

tosyl-N. The resulting transition state is depicted in Figure 5.16 a; further options as indicated

in Scheme 5.7 (h) and Figure 5.16 b/c are discussed below.

Hence, it remains questionable if an increase in entropy, as indicated by the positive sign of

∆S‡under acidic conditions, rather than a (possibly incomplete) compensation of a decrease

may be caused through ring-opening. The absolute value and sign of ∆S‡are calculated from

an Eyring plot and depend on the values of the rate constant kat the respective temperatures.

5 Immobilization of a Rhodium Catalyst

The determination of kusing immobilized catalyst 22 entails the problem that the exact con-

centration of substrate in the environment of the catalyst is unknown. An accumulation effect

is assumed to occur (see section 5.2.2.1), but no quantitative statement can be given at this

time. The calculation of kis therefore based on the total concentration of acp in the reaction

mixture, which is quite low. The presumed accumulation of acp in the immediate environment

of the catalyst thus leads to a value of kwhich is probably too high (cf. [189]). The slope in

the Eyring plot, which represents the activation enthalpy, is not affected as long as the accu-

mulation of acp is not temperature-dependent. In contrast, the activation entropy is somewhat

erroneous, though probably only the absolute values, since accumulation is assumed to occur to

the same extent under basic as well as under acidic conditions. Thus, ∆∆S‡≈38 J mol−1K−1

is assumed to be correct. The inhomogeneous concentration of the substrate could explain the

great difference to the activation parameters found by Xiao, Liu, and co-workers [112]; using the

homogeneous Ru-TsDPEN catalyst for the ATH of acetophenone (under conditions comparable

to the “standard reaction conditions”), a ∆H‡of 53.6 kJ mol−1and a ∆S‡of −105 J K−1mol−1

were found (cf. Section 2.4.4.2). The difference of about −95 J K−1mol−1in the entropy of

activation might be best explained by the accumulation effect. In conclusion, a reliable state-

ment that the entropy of activation increases upon the formation of the transition state under

acidic conditions cannot be made, but an activated complex, which is less hindered than under

basic conditions, is strongly suggested.

Apart from ring-opening as a reason for the higher entropy of activation under acidic condi-

tions, a general explanation for the observed rate acceleration may be acid catalysis. Results of

DFT calculations, which were made to elucidate the role of water in the hydrogen transfer me-

diated by Ru-TsDPEN (in a non-acidc medium), showed that the hydrogen bonding interaction

between the amine-H and the carbonyl-O was smaller than without water and that the acti-

vation energy was lowered (water made the transition state more “reactant-like” in accordance

with the Hammond postulate) [112]. Hydronium ions might intensify this effect, given that the

formation of the transition state is not impeded for other reasons. Thus, either protonation

promotes the hydrogen transfer or does not interfere with it under moderately acidic conditions

(Figure 5.16 a/b), or it does not occur (Figure 5.16 c).

In the first case rate acceleration due to the addition of acid might be in addition to a

potential increase in flexibility due to ring-opening (Figure 5.16 b); in the latter it provides the

only explanation for a decrease in ∆G‡. The mechanism of the proton transfer might even be

completely altered so that the hydronium delivers the proton instead of the catalyst´s amine.

The amine-H, however, would still be necessary to interact with the carbonyl-O to get the

substrate in the right position. It would have to be clarified to which extent the proton transfer

may proceed prior to the formation of the transition state. If a transition state as depicted

in Figure 5.16 c is true, single protonation would be a sufficient explanation for inhibition,

5 Immobilization of a Rhodium Catalyst

and ring-opening as an explanation for a (relative) increase in ∆S‡would have to be rejected.

Determination of the pKbvalues of the unmodified tethered Rh-TsDPEN might provide a first

indication.

A further effect that has to be taken into account is solvation, especially when dealing

with charged reactants. In this context, it might be useful to discuss the ancillary ligand (L

in Scheme 5.7, Figure 5.14 and Figure 5.16) that will saturate the central rhodium ion if

ring-opening occurs. The group of Süss-Fink and others have proven the existence of aqua

complexes formed from Ru-TsDACH via hydrolysis [185, 190]. Hence, Xiao et al. suggested

that water is coordinated to the central metal of Ru-TsDPEN when protonation under acidic

conditions leads to an opened ring [79]. However, as they could not isolate an aqua complex,

the formate anion was considered as alternative ancillary ligand [112]. Accordingly, L is either

H2O (leading to a +1 charge of the complex) or HCO−

2. From the above results, however, it

seems implausible that L is water when hydronium participates in the transition state at the

same time (Figure 5.16 b). This situation would lead from a +1-charged catalyst to a +2-

charged activated complex with loss of “freedom” due to electrostriction [191]. Additionally, in

case of L being water and the presence of a pre-activated, +1-charged substrate, the frequency

of collisions would be reduced according to the kinetic theory of collisions. In summary, a

definite statement on the mechanism cannot be made on the basis of the data obtained thus

far, and further steps have to be made to provide evidence for one of the mechanistic options

considered or even for an entirely new catalytic cycle. These steps include:

•determination of the pKbof unmodified tethered Rh-TsDPEN catalyst (Figure 5.1) and

21 to identify the pH conditions that lead to protonation and to elucidate differences

caused by the structural modification

•determination of the rate of conversion of acp when using the unmodified tethered Rh-

TsDPEN catalyst to shed light on the impact of the support

•NMR experiments to elucidate the complex structure in the presence of acids analogous

to those described in the literature [112], preferably using the unmodified tethered Rh-

TsDPEN catalyst

•DFT calculations to prove the possibility of an enantioselective hydrogen transfer by an

“opened” catalyst and to determine the energy barrier for when water is substituted for

hydronium.

5.4 Practical Aspects

The experimental results discussed in the preceding sections provide data and information about

the specific properties of the system developed here which are required for optimal use on a

5 Immobilization of a Rhodium Catalyst

technical level. However, from an industrial point of view, several further aspects might be

of importance. Studying these aspects does not necessarily provide additional answers to the

question of how the catalytic system works, but rather to the question of how well it works.

All results which are considered important in this context are summarized at the end of this

chapter.

5.4.1 Application-oriented Experiments

(a) Performance after storage.

Reaction conditions: acp 0.04 M, HCO2Na 0.4 M,

PB-supported catalyst 2 g (∼0.113 mM Rh), 40 ◦C,

1 200 rpm.

(b) Variation of stirring speed.

Reaction conditions: acp 0.04 M, HCO2Na 0.4 M, PB-

supported catalyst 1.5 g (∼0.105 mM Rh.), 40 ◦C.

Reaction conditions: acp 0.04 M, HCO2H 0.12 M,

HCO2Na 0.28 M, PB-supported catalyst 1 g

(∼0.13 mM Rh), 40 ◦C, 1 200 rpm, total reaction vol-

ume 100 mL (+ HOAc 0.42 M, NaOAc 0.12 M).

(d) Upscaling.

Reaction conditions: acp 0.32 M, HCO2H 0.96 M,

HCO2Na 2.24 M, PB-supported catalyst 3 g

(∼0.39 mM Rh), 40 ◦C, 1 200 rpm, total reaction vol-

ume 100 mL; standard conditions as in Fig. 5.10.

Figure 5.17: Technically relevant tests.

The high stability of 22 and analogous catalysts in reactions performed in water and air was

demonstrated in this and other studies [176]. It was found that a period of about four months

of storage in air in the refrigerator did not affect the catalyst performance (Figure 5.17 a).

It was thus concluded that durability does not pose a problem with the supported tethered

5 Immobilization of a Rhodium Catalyst

Rh-TsDPEN catalyst.

On a technical scale, stirring contributes considerably to the cost of a given process since

it influences the dimensioning of the reactor as well as the energy consumption. Reducing the

stirring speed may thus be a useful approach to reduce costs. As discussed previously, the

aggregation of beads leads to a decrease in the rate and reproducibility of the reaction. A high

stirring speed of 1 200 rpm was therefore applied in all experiments. Under certain conditions,

however, the stirring speed can be significantly reduced without affecting the performance of

the reaction. Figure 5.17b shows that there is no significant difference in the conversion of

acp over time in the standard experiment performed at 1 200 rpm and at 600 rpm. Whether

the suspension is homogeneous or not, though, also depends on the number of beads and

the concentration of the substrate. Hence, the optimal stirring speed has to be determined

individually for each approach.

The pH was found to be a decisive parameter for the performance of the catalyst, and

the activity (TOF) peaked at an initial pH of about 4. Since the basicity of the reaction

medium increased with increasing conversion, the effect of an additional buffer with a higher

capacity than HCO2H/HCO2Na was tested. Through the addition of small amounts of acetate

buffer, the increase of the pH was reduced; when an initial pH of 3.9 was set, a value of 4.2

was determined at the end of the reaction instead of 4.6 (conversion >99 %). The reaction

rate, however, was reduced rather than improved (Figure 5.17c). This might be ascribed to

a disfavorable interaction between acetate and catalyst. Consequently, a more suitable buffer

system might be capable of improving the reaction, but it also might be challenging to find the

optimal system at the optimal concentration. Since the expected improvement was marginal,

no further tests on this matter were carried out.

However, the efficiency of the reaction can be improved and transferred to a more technically

relevant level if the concentrations of catalyst and substrate as well as the S/C are increased.

The use of additional beads in the solution was limited, since a homogeneous dispersion was

required to ensure reproducible conditions and a fast conversion. At the same time, aggregation

occurred mainly at a higher substrate concentration when a smaller number of beads was

employed. A compromise was reached using beads with higher loading (13 µmol g−1instead of

7µmol g−1) while increasing the amount by a factor of two (from 1.5 g to 3 g), whereas the

concentration of acetophenone was increased by a factor of eight to 0.32 M (Figure 5.17 d).

Under these conditions, little aggregation occurred at a stirring speed of 800 rpm, and an STY

of 10 ±1 g L−1h−1with an average TOF of 200 ±10 h−1was achieved.

5.4.2 Discussion on the Efficiency

It was demonstrated that the immobilized Rh(III)-catalyst (22) developed in this study can be

applied in ATH reactions of various aryl ketones (cf. Section 5.2.2.3); all performance tests,

5 Immobilization of a Rhodium Catalyst

however, were conducted exclusively using acetophenone as the substrate. Thus, the following

considerations on the strengths and weaknesses of the catalytic system and its potential for an

industrial application are based on this benchmark compound, taking into account the industrial

requirements discussed in Section 2.2 and Section 2.4.6.

Generally, the selectivity of ATH reactions is 100 %, because there is no generation of side

products from the prochiral substrate. Since in sodium formate/formic acid/water complete

conversion can be achieved, the theoretical yield is also 100%. The enantioselectivity achieved

with 22 is about 98 % ee of 1-(S)-phenylethanol under optimal reaction conditions (40 ◦C,

pH 3.9). This is in the range of the best results found in the literature and significantly exceeds

the minimum requirement of 90 % ee in pharmaceutical production processes. Furthermore, a

TON of at least 1 000 is required for the small scale fabrication of high added value products.

This requirement is met by catalyst reuse which provided a TTN of over 1 200 in eight consec-

utive runs with a total loss of metal of about 5 %. Upscaling experiments, however, revealed

that even a single run performed at a higher S/C can provide a TON of more than 800. For

the TOF, which should be higher than 500 h−1at 95 % conversion, a value of 200 h−1(at

>99 % conversion) was achieved, hence there is a need for improvement. Both TOF and TON

may be enhanced by the application of a higher substrate concentration; the TOF will also

increase when a higher catalyst concentration is applied in addition to a higher substrate con-

centration. The problem of bead aggregation will therefore have to be adressed; the addition

of a polar organic co-solvent such as methanol may be a starting point. An increased substrate

concentration will also improve the STY, which was about 10 g L−1h−1.

A drawback of the functionalized support materials is the the low surface area, which limits

the loading with amino groups (15–30 µmol g−1) and consequently the catalyst loading. A

molecular weight of the heterogeneous catalyst of about 10 kDa per mole of active site cannot

be achieved when these materials are applied. Particles that are porous and smaller in diameter

are required if this is truly a decisive criterion. One advantage of the catalytic system present

here (22), however, is its robustness. The stability to air as well as the high mechanical stability

allow for the use of simple reaction and filtration setups, and the insensitivity to carbon dioxide

makes gas disengagement unnecessary.

Nevertheless, the question of efficiency is not settled. Catalyst immobilization — in particular

heterogenization — is generally associated with increased effort; a statement on the advantage

of an immobilized catalyst over its free analog should therefore consider the potentially higher

preparation costs. The established key figures which are used to compare catalyst and process

performances, however, do not include such factors. This might be for various reasons: costs

(as a dimension of “effort”) are usually difficult to determine, most often do not play a role in

scientific publications, are kept secret in industry, etc. If however they are known, they should

be considered by using a modified productivity indicator, which could be referred to as “catalyst

5 Immobilization of a Rhodium Catalyst

cost efficiency” (CCE, Equation 5.4.1), borrowing a term coined by Campbell [192].

Since the catalyst developed in this study will be commercially available from Strem Chemi-

cals Inc., this key figure may be used to decide whether an application could be advantageous or

not. Of course, the overall process is not taken into account, and further considerations, e. g.,

on the effort and costs required for separation, will have to be made. If data on the catalyst

costs are not available, the “catalyst efficiency” (CE) could be estimated using Equation 5.4.2.

CCE [mol e−1] = converted starting material [mol]

amount of catalyst employed [mol] ·costs of catalyst [emol−1](5.4.1)

CE =converted starting material [mol]

amount of catalyst employed [mol] ·yield

number of synthetic steps (5.4.2)

This latter approach may be oversimplified as it does not differentiate among the synthetic steps,

but it does consider the resources employed (number of synthetic steps) and the yield obtained

in the preparation of the catalyst. For example, three linear steps are required to prepare the

unmodified tethered Rh-TsDACH complex (Figure 5.1) when chiral DACH is considered the

starting compound (not taking into account the preparation of the tether). Presuming a yield

of 90 % in the first step (preparation of TsDACH), the overall yield should be about 15 %

[95, 177]. Analogous calculation of the overall yield in the preparation of 22 gives about 11 %

in nine steps, not taking into account the synthesis of the tether and the preparation of the

support. When an S/C of 200 is applied and complete conversion in a single run is achieved with

the homogeneous Rh-TsDACH catalyst, and a TTN of 1 250 is achieved with the supported

catalyst under otherwise comparable conditions, CE values of 10 for Rh-TsDACH and 15 for 22

are obtained. Although not all preparation “costs” and application advantages are taken into

account, this might at least demonstrate that heterogenization is becoming a serious alternative

to homogeneous catalytic systems.

Apart from economic aspects, catalytic transfer hydrogenation reactions offer a number

of benefits over alternative methods according to the principles of green chemistry and green

engineering, e. g., the use of safe solvents and auxiliaries as well as safe processes. Nevertheless,

ATH reactions are usually not considered atom economical [28]. This is mainly because the

hydrogen source itself is transformed into “waste” during the process. The determination of

the E-factor or another key figure which assesses the atom economy by taking into account

the waste generated is challenging (see Section 2.2.3). An estimation of the E-factor following

the simplified approach, i. e., by only taking into account the reagents and the desired product

(not taking into account the catalyst, solvent, energy etc.) gives a value of about 0.7 when

the reaction is performed with a stoichiometric amount of sodium formate, but 5.7 when 10

equivalents are applied and non-transformed sodium formate is considered as waste. analogous

5 Immobilization of a Rhodium Catalyst

data of alternative methods are required to compare these values; a thorough determination of

the ecological impact, however, should consider the overall process.

6 General Conclusion and Outlook

Asymmetric transfer hydrogenation is becoming a more commonly applied industrial method.

The intensive research performed during the last 20–30 years has afforded substantial progress,

including the development of highly efficient catalysts applicable to different reaction media

and different hydrogen donors. Not only are the operational aspects of this method persuasive

(high safety and simplicity), but the overall performance has become competitive compared to

other approaches. Additionally, many attempts have been made to immobilize ATH catalysts to

address the problem of efficient catalyst separation and reuse, and promising results have been

obtained. Only very few studies in this field, however, have focused on application-oriented

data and kinetic parameters.

The aim of this work was the preparation and study of immobilized catalysts for the ATH

of prochiral ketones. Heterogenization of outstanding homogeneous catalysts was chosen as

the immobilization strategy, and kinetic studies were performed with the goal of obtaining a

fundamental understanding of the operating mode of the catalytic systems. Two transition

metal complexes were subjected to synthetic modification in order to introduce linkers suitable

for the covalent attachment to solid polymer supports. These supports were prepared via

Molecular Surface Engineering, i. e., a “functional matrix” containing (protected) primary amines

was grafted from either polypropylene membranes or polyethylene sinter chips and beads.

In a first attempt, a diphosphine/diamine (PNNP) ligand was modified via a new nine-step

convergent synthesis process to introduce a carboxylic linker [86]. For the purpose of ATH

operations, the ligand could either be used with ruthenium(II) in isopropyl alcohol solutions or

with iridium(III) — after in situ generation— in aqueous media [168]. However, all attempts

at covalent attachment of both the ligand and the corresponding ruthenium complex failed.

Another attempt was made with the linkage of a modified rhodium(III) complex based on a

variant of the monotosylated diphenylethylenediamine (TsDPEN) ligand, which features a linked

η5-cyclopentadienyl unit [95]. A carboxylic linker was introduced in a joint project with the

Haag Research Group, and after metal insertion and successful attachment to the supports, a

new immobilized catalyst, the supported tethered Rh-TsDPEN (22), was applied in the ATH of

aryl ketones. Comparison of the catalytic performance in the sodium formate/water system in

air showed that the enantioselectivity in the transfer hydrogenation of acetophenone and related

substrates was as high as or even better than reported for the unmodified analogous catalysts

in the literature [79, 177]. In terms of reaction rate, the polymer-supported catalyst (22) was

6 General Conclusion and Outlook

found to be superior even in comparison to the monomeric analog (23) in the concentration

range investigated. This is ascribed to the properties of the support material; an accumulation of

the substrate in the microenvironment of the catalyst is likely to occur, leading to an enhanced

rate at low substrate concentrations. A successful reuse after simple filtration was achieved

when an acidic mixture of sodium formate and formic acid in water was used.

Kinetic investigation revealed that the conversion of acetophenone to phenylethanol follows

a first-order dependency of the reaction rate on both the substrate and the catalyst concentra-

tion. It was therefore suggested that the operating mode of the immobilized tethered catalyst

is similar to that proposed for untethered metal-arene-TsDPEN catalysts in the moderately

basic reaction medium sodium formate/water [112]. However, considerable differences in the

catalyst performance compared to previously reported results were noted when the reaction

conditions were changed: increased activity and reusability without loss of enantioselectivity of

the immobilized system were observed in the acidic HCO2H/HCO2Na/H2O medium, but the

second-order rate law was found to still be valid. Thus, the optimal pH for this catalyst is

considerably below that reported for the untethered analogs (pH ∼4 and pH ∼7, respectively)

[184, 186, 187]. Additionally, emerging carbon dioxide was found to accelerate the reaction

rather than inhibiting it, due to its buffer capacity. These findings are presumably ascribable

to the structural superiority of the immobilized complex, effected by the tether between chiral

backbone and cyclopentadienyl ligand. Comparison of the activation parameters of reactions

carried out in moderately basic and moderately acidic media revealed a significant difference

in the entropy of activation (∆S‡

acidic −∆S‡

basic ≈38 J mol−1K−1). The underlying effect of

the rate enhancement under acidic conditions is thus presumed to be an altered mechanism

in which — apart from possible solvation effects — either ring-opening or acid catalysis or both

provide a “less ordered” transition state. If ring-opening occurs, the tether might provide an

explanation why this catalyst, in contrast to untethered variants, is capable of keeping its steric

orientation. Further investigations including the spectroscopic analysis of the homogeneous

analog as well as DFT calculations are required to elucidate these considerations. A deeper

understanding of the mechanism may help in developing even more efficient catalysts in the

future.

The accumulation of the substrate is presumed to derive from a better solubility of the

substrate in the outer layer of the support than in the solvent. This is only the case when a

polar solvent is used. The significantly reduced rate observed when the reaction was run in

organic media instead of water might be a further indication, although a possible involvement

of water in the reaction mechanism also has to be taken into account. However, substrate

accumulation would be a viable way of enhancing the rate of ATH reactions carried out in

IPA, since these reactions are mostly performed at high dilution to shift the equilibrium. The

development of the PNNP-based catalyst was indeed focused on the 2-propanol system, but

6 General Conclusion and Outlook

in light of the above findings, the effectiveness of this approach is called into question. As

long as the support does not attract the substrate to at least the same extent as the solvent

(and to a higher extent than the product), one of the main drawbacks usually associated with

heterogenization — a strong decrease in catalyst activity— would be inevitable. Since even

homogeneously applied Ru-PNNP shows a relatively low activity, the effort of modification and

immobilization would not be worth the cost, neither economically nor ecologically. In principle,

the polarity and hydrophilicity/hydrophobicity balance of the surface of the supports can be

adjusted via Molecular Surface Engineering, but the range of adjustments is, of course, limited.

Thus, the immobilization strategy applied here seems to be inefficient for catalysts that can

only be used in medium polar or nonpolar media.

Nonetheless, the application of the immobilized Rh(III)-catalyst (22) to ATH reactions of

aryl ketones in aqueous media provided promising results, though there is certainly plenty of

room for optimization. As this is the first report of catalyst immobilization together with kinetic

investigation applying a tethered version of the TsDPEN ligand as well as the first time that an

immobilized ATH catalyst will be commercially available, it will hopefully be a starting point

for further studies. Future work will include both the optimization and deeper understanding

of the system developed as well as research into further aspects, particularly with respect

to environmental issues of an industrial application. The use of recyclable catalysts already

complies with several principles of green chemistry and green engineering. To further improve

the ecological impact, the problems of a cycle of matter and a commercial “afterlife” will

possibly be addressed in a joint project with an industrial partner. The commercial application

of a continuously operated reaction system is currently being tested. Furthermore, an ATH

process in combination with a membrane-based separation of the organic products from the

aqueous reaction medium is envisaged. Since analogous ruthenium complexes have been found

to catalyze different reactions in addition to ATH, e. g., the asymmetric Michael-Addition [193,

194], the range of applications could be further tested. A combined chemo-/biocatalytical

application is already under investigation at Delft University of Technology.

7 Experimental

Unless preparative details are provided, all reagents were commercially available from standard

suppliers and used without further purification. Compounds known in the literature are ref-

erenced where appropriate in the procedure. Reactions which required anhydrous and inert

conditions were conducted in oven-dried apparatus and in an atmosphere of argon (5.0) using

standard Schlenk techniques and dry solvents; addition of matter or sampling was routinely

performed in a counter current flow of argon in such reactions. Column chromatography was

performed with silica gel (60 mesh, 0.06–0.2 mm particle size). Synthetic work-up and washing

procedures were performed with deionized water, catalytic testing was performed using water

purified by reverse osmosis. Thin layer chromatography (TLC) was used to monitor reactions

where appropriate. Visualization of the plates was by 254 nm UV light and/or iodine staining.

Nuclear magnetic resonance (NMR) spectra were recorded at room temperature (rt), and

chemical shifts (δ), measured in parts per million (ppm), are reported relative to the tetra-

methylsilane signal (δ= 0.00 ppm, external standard for 1H- and 13C-NMR measurements)

and the phosphorous acid signal (δ= 0.00 ppm, external standard for 31P-NMR measure-

ments), respectively. Multiplicities are denoted as singlet (s), doublet (d), doublet of doublets

(dd), triplet (t), pentet (p), and multiplet (m), and where required prefixed br (broad). Infrared

(IR) spectra were recorded either from pure solids or from thin films of solutions in chloroform

using attenuated total reflection (ATR) technique. IR absorptions are reported in wavenum-

bers (ν), measured in cm−1. High resolution mass spectrometry (HRMS) was performed by

electrospray ionization (ESI), either via direct injection or coupled with high performance liquid

chromatography (HPLC). The calculation of the “mass to elementary charges ratio” (m/z) as

well as the interpretation of the high resolution mass spectra were performed with Xcalibur

Qual Browser, version 2.0.7.

The determination of the conversion in catalysis testing was performed via gas chromato-

graphy (GC) using a flame ionization detector (FID) or a coupled mass spectrometer (MS)

equipped with an electron ionization (EI) device. The GC-MS system was calibrated for the

determination of the conversion of acetophenone to 1-phenylethanol, which was calculated by

dividing the peak area of the ion fragments produced from acp by the sum of the ion fragments

of both compounds. Computational data were obtained by the use of Madonna software,

version 8.3.14.

7 Experimental

7.1 Equipment

7.1.1 Instrumentation

•GC (analysis of the headspace gas in catalysis experiments): Shimadzu GC-2014 gas

chromatograph equipped with ValcoPLOT molsieve, HayeSep D and HayeSep Q columns,

and an FID as well as a thermal conductivity detector.

•GC (analysis in catalysis experiments): Varian CP-3800 equipped with an FID detector,

aVarian CP-8400 AutoSampler, test column (CP-Sil B CB; 15 m x 0.25 mm x 0.25 µm),

or chiral column (CP-Chirasil-Dex CB; 25 m x 0.25 mm x 0.25 µm).

•GC-MS (analysis in catalysis experiments): Hewlett Packard HP 6890 Series gas

chromatography system equipped with a HP 5975 Mass Selective Detector using electron

ionization (EI) at 70 eV and a Gerstel MPS 2L auto sampler. A Machery-Nagel Optima

Wax column (30 m x 0.32 mm x 0.5 µm) was used for determination of the conversion,

and for determination of the enantiomeric excess a Supelco Beta DEX 110 column (30

m x 0.25 mm x 0.25 µm) was used.

•GC-MS (analysis in synthesis): Hewlett Packard HP 6890 Series/HP 5975 Mass Selective

Detector as specified above, equipped with an HP-5MS GC column (30 m x 0.25 mm x

0.25 µm).

•HRMS: Thermo Scientific Orbitrap LTQ XL (spray voltage 5 kV, source temperature

275 ◦C, solvent methanol + 0.1 % formic acid, flow rate 200 µL min−1); HPLC conditions:

Agilent Eclipse XDB-C18 column (4.6 x 150 mm, 5 µm), eluent 1 (H2O + 0.025 %

HCO2H), eluent 2 (MeOH + 0.025 % HCO2H), flow 1.0 mL min−1.

•ICP-OES: Varian 715-ES.

•ICP-MS: Thermo Fisher Element 2.

•FT-IR: Perkin Elmer Spectrum one with Universal ATR Sampling Accessory.

•Microwave Digester: CEM Discover SP-D.

•NMR (1H, 13C): Bruker Avance 400.

•NMR (31P): Bruker AC-F 200.

•UV-Vis: Hitachi U-3410 Spectrophotometer.

7 Experimental

7.1.2 Laboratory Equipment

•Mechanical stirrer: Heidolph RZR 2051 control.

•Thermostat: LAUDA Ecoline Staredition RE 312.

•Vacuum pump: ILMVAC diaphragm pump.

•Magnetic stirrer: HEIDOLPH MR Hei-Standard magnetic stirrer with heating equipped

with temperature control EKT Hei-Con.

•Temperature sensor: testo 720.

•pH meter: HANNA Instruments pH 211.

•Reaction vessel: Rettberg jacketed 100-mL flask.

•Rotary evaporator: Heidolph Laborota 4000 equipped with Vacuubrand MZ 2C dia-

phragm vacuum pump.

•Reverse osmosis water purification system: Millipore Elix 5.

•Ion exchange system: SD 2800.

7.2 Synthesis Procedures

7.2.1 Synthesis of Ru-PNNP

7.2.1.1 Hex-5-ynoic acid tert-butyl ester (2)

Different from the procedure used by Bartoli et al. [195], the introduction of the tert-butyl pro-

tecting group into the linker moiety was performed via Steglich esterification [196]. 5-Hexynoic

acid (10 mmol, 1.12 g), tert-butanol (30 mmol, 2.22 g), and 4-(dimethylamino)pyridine

(DMAP; 0.6 mmol, 0.073 g) were dissolved in 10 mL of dichloromethane (DCM). N,N0-

Dicyclohexylcarbodiimide (DCC; 11 mmol, 2.27 g) was carefully added while the flask was

cooled in ice water. After 4 h of stirring at room temperature, the reaction was completed

(GC-MS control), and precipitated dicyclohexylurea (DCU) was filtered off. The filtrate was

concentrated under reduced pressure and purified via column chromatography (eluent: DCM)

7 Experimental

to yield a colorless oil (Y2= 70 %).

1H-NMR (400 MHz, CDCl3): δ(ppm) = 2.31 (t, 2 H, CH2COOtBu, J = 7.4 Hz), 2.21 (dt, 2

H, CH2Calkynyl , J = 2.6 Hz, J = 7.0 Hz), 1.93 (t, 1 H, Calkynyl H, J = 2.7 Hz), 1.77 (p, 2 H,

CH2CH2CH2, J = 7.2 Hz), 1.41 (s, 9 H, CH3).

13C-NMR (100 MHz, CDCl3): δ(ppm) = 172.3 (COOtBu), 83.3 (Calkynyl CH2), 80.1

[C(CH3)3], 68.8 (Calkynyl H), 34.1 (CH2COOtBu), 28.0 (3 C, CH3), 23.7 (CH2CH2CH2), 17.7

(CH2Calkynyl ).

HRMS (ESI): m/z calculated for C10H17O2([M +H]+) = 169.12231 (100%), 170.12566

(10.82 %); m/z found = 169.12181 (100 %), 170.12517 (9.10 %).

GC-MS (EI): m/z = 153.0 ([M −CH3]+), 112.0 ([M −C4H8]+), 95.0 ([M −C4H9O]+), 67.0

([M −C5H9O2]+), 57.1 ([C4H9]+).

7.2.1.2 2-Bromo-5-iodophenylmethanol (4)

Borane dimethyl sulfide complex (24 mmol, 1.94 g) was added over 5 min to a solution of

2-bromo-5-iodobenzoic acid (10 mmol, 3.27 g) in THF (40 mL). After stirring for 4 h at room

temperature, no further conversion was observed (TLC control). Water (25 mL) was added in

order to hydrolyze remaining borane components, and the mixture was stirred for 15 min. After

extraction with DCM, the organic phase was dried over MgSO4, and the solvent was removed

under reduced pressure. The product, a white solid, was further purified by recrystallization

from hexane; the overall yield was 80 %.

1H-NMR (400 MHz, MeOH-d4): δ(ppm) = 7.85 (d, 1 H, CφH, J = 2.1 Hz), 7.47 (dd, 1 H,

CφH, J = 2.2 Hz, J = 8.3 Hz), 7.26 (d, 1 H, CφH, J = 8.3 Hz), 4.58 (s, 2 H, CH2OH).

13C-NMR (100 MHz, MeOH-d4): δ(ppm) = 142.7 (CφCH2OH), 137.2–133.6 (3 C, CφH),

120.9 (CφBr), 91.9 (CφI), 62.5 (CH2OH).

GC-MS (EI): m/z = 313.9/311.9 ([M]+), 232.9 ([M −Br]+).

7 Experimental

7.2.1.3 2-Bromo-5-(hex-5-ynoic acid tert-butyl ester)-phenylmethanol (5)

Under inert and anhydrous conditions, compound 4(10 mmol, 3.13 g) and compound 2

(12 mmol, 2.02 g) were dissolved in THF (30 mL). While the solution was beeing stirred

at room temperature, PdCl2(PPh3)2(0.3 mmol, 0.211 g), CuI (0.5 mmol, 0.095 g), and

triethylamine (25 mmol, 3.5 mL) were added in succession. After 5 h no further conversion

was observed (TLC control). Precipitated triethylamine hydroiodide was filtered off, and the

filtrate was concentrated by solvent evaporation under reduced pressure. The residue was

dissolved in DCM and washed with a saturated aqueous solution of NH4Cl. The organic phase

was then dried over MgSO4, and solvents were removed under reduced pressure. Column

chromatography (eluent: DCM/hexane, 1/1) was performed for purification yielding the

product in 62 % yield as a yellow-orange oil of moderate viscosity.

1H-NMR (400 MHz, CDCl3): δ(ppm) = 7.51 (d, 1 H, CφH, J = 2.1 Hz), 7.43 (d, 1 H,

CφH, J = 8.2 Hz), 7.16 (dd, 1 H, CφH, J = 2.1 Hz, J = 8.2 Hz), 4.69 (d, 2 H, CH2OH,

J = 5.6 Hz), 2.44 (t, 2 H, CH2COOtBu, J = 7.0 Hz), 2.38 (t, 2 H, CH2Calkynyl , J = 7.5

Hz), 2.23 (t, 1 H, OH, J = 6.0 Hz), 1.87 (p, 2 H, CH2CH2CH2, J = 7.2 Hz), 1.45 (s, 9 H, CH3).

13C-NMR (100 MHz, CDCl3): δ(ppm) = 172.6 (COOtBu), 139.8 (CφCH2OH), 132.4–131.7

(3 C, CφH), 123.5 (CφBr), 121.6 (CφCalkynyl ), 90.4 (Calkynyl CH2), 80.4 (Calkynyl Cφ), 80.3

[C(CH3)3], 64.7 (CH2OH), 34.5 (CH2COOtBu), 28.1 (3 C, CH3), 24.0 (CH2CH2CH2), 18.9

(CH2Calkynyl ).

HRMS (ESI): m/z calculated for C17H22BrO3([M +H]+) = 353.07468 (100 %), 354.07804

(18.4 %), 355.07264 (97.3%), 356.07599 (17.9 %); m/z found = 353.07416 (100%),

354.07739 (13.8 %), 355.07211 (95.8 %), 356.07538 (13.4 %).

7 Experimental

7.2.1.4 2-Bromo-5-(hex-5-ynoic acid tert-butyl ester)-benzaldehyde (6)

Equivalent amounts (10 mmol) of compound 5(3.53 g) and Dess–Martin periodinane (4.24 g)

[197] were dissolved in DCM (30 mL). Stirring at room temperature for 10 h gave a brownish

mixture which turned transparent by addition of 1.3 M sodium hydroxide solution (40 mL).

After extraction with diethyl ether, the organic phase was dried over MgSO4and concentrated

under reduced pressure. Further purification via column chromatography (eluent: DCM)

resulted in a clear yellow oil in 68% yield.

1H-NMR (400 MHz, CDCl3): δ(ppm) = 10.30 (s, 1 H, CHO), 7.89 (d, 1 H, CφH, J = 2.2

Hz), 7.56 (d, 1 H, CφH, J = 8.3 Hz), 7.42 (dd, 1 H, CφH, J = 2.2 Hz, J = 8.3 Hz), 2.46

(t, 2 H, CH2COOtBu, J = 7.0 Hz), 2.39 (t, 2 H, J = 7.4 Hz, CH2Calkynyl ), 1.88 (p, 2 H,

CH2CH2CH2, J = 7.2 Hz), 1.45 (s, 9 H, CH3).

13C-NMR (100 MHz, CDCl3): δ(ppm) = 191.2 (CHO), 172.4 (COOtBu), 137.7 ( CφH), 133.8

(CφH), 133.3 (CφCHO), 132.8 (CφH), 125.7 (CφBr), 124.2 (CφCalkynyl ), 92.0 (Calkynyl CH2),

80.5 (Calkynyl Cφ), 79.2 [C(CH3)3], 34.4 (CH2COOtBu), 28.1 (3 C, CH3), 23.9 (CH2CH2CH2),

18.8 (CH2Calkynyl ).

HRMS (ESI): m/z calculated for C17H20BrO3([M +H]+) = 351.05903 (100 %), 352.06239

(18.4 %), 353.05699 (97.3%), 354.06034 (17.9 %); m/z found = 351.05844 (100%),

352.06168 (15.6 %), 353.05634 (95.8 %), 354.05963 (13.4 %).

7.2.1.5 5-(Hex-5-ynoic acid tert-butyl ester)-2-(diphenylphosphino)benzaldehyde (7)

7 Experimental

Under inert and anhydrous conditions, compound 6(10 mmol, 3.51 g) was dissolved in toluene

(20 mL). Pd(PPh3)4(0.07 mmol, 0.081 g), dissolved in 10 mL of toluene, diphenylphosphine

(13 mmol, 2.42 g), and triethylamine (15 mmol, 2.1 mL) were added successively. The reaction

mixture was refluxed for 3 h, then precipitated triethylamine hydrobromide was filtered off.

The filtrate was washed with an aqueous solution of NH4Cl and brine. After separation from

the aqueous phase, the organic phase was dried over MgSO4and concentrated under reduced

pressure. The product was obtained in 56 % yield after column chromatography (eluent:

DCM) as an viscous, bright yellow oil.

1H-NMR (400 MHz, CDCl3): δ(ppm) = 10.43 (d, 1 H, CHO, J = 5.3 Hz), 7.97 (m, 1 H,

CφH), 7.45–7.24 (m, 11 H, CφH), 6.89 (m, 1 H, CφH), 2.48 (t, 2 H, CH2COOtBu, J =

7.0 Hz), 2.40 (t, 2 H, CH2Calkynyl , J = 7.4 Hz), 1.89 (p, 2 H, CH2CH2CH2), 1.45 (s, 9 H, CH3).

13C-NMR (100 MHz, CDCl3): δ(ppm) = 191.10 (d, 1 C, CHO, 3J = 18.4 Hz), 172.4

(COOtBu), 140.4 (d, 1 C, CφCHO, 2J = 26.7 Hz), 138.3 (d, 1 C, CφP, J = 14.6 Hz), 136.1

(CφH), 135.9 (d, 2 C, CφP, J = 9.5 Hz), 134.1–133.8 (5 C, CφH), 133.7 (d, 1 C, CφH, J =

3.7 Hz), 129.2–128.8 (6 C, CφH), 124.9 (CφCalkynyl ), 92.0 (Calkynyl CH2), 80.4 (Calkynyl Cφ),

79.9 (C(CH3)3O), 34.4 (CH2), 28.1(3 C, CH3), 24.0 (CH2), 18.9 (CH2).

31P-NMR (80 MHz, CDCl3): δ(ppm) = -13.13.

HRMS (ESI): m/z calculated for C29H30O3P ([M +H]+) = 457.19271 (100 %), 458,19606

(31.4 %), 459,19942 (4.8 %); m/z found = 457.19468 (100 %), 458.19800 (30.2 %), 459.20123

(4.1 %).

7.2.1.6 2-(Diphenylphosphino)benzaldehyde (9)

2-(Diphenylphosphino)benzaldehyde 9is commercially available from different suppliers.

Nevertheless, it was synthesized in an analogous manner as 7under inert and anhydrous

conditions. 2-Bromobenzaldehyde (10 mmol, 1.85 g) was dissolved in 20 mL of toluen.

Pd(PPh3)4(0.07 mmol, 0.081 g), dissolved in 10 mL of toluene, diphenylphosphine (13 mmol,

7 Experimental

2.42 g), and triethylamine (15 mmol, 2.1 mL) were added successively. The reaction mixture

was refluxed for 3 h, then precipitated triethylamine hydrobromide was filtered off. The filtrate

was washed with an aqueous solution of NH4Cl and brine. After separation from the aqueous

phase, the organic phase was dried over MgSO4and concentrated under reduced pressure.

Recrystallisation from methanol and further purification via column chromatography (eluent:

DCM) yielded a bright yellow solid as the product (Y9= 80 %).

1H-NMR (400 MHz, CDCl3): δ(ppm) = 10.52 (d, 1 H, CHO, J = 5.4 Hz), 7.97 (m, 1 H,

CφH), 7.7–7.46 (m, 2 H, CφH), 7.40–7.24 (m, 10 H, CφH), 6.98 (m, 1 H, CφH).

13C-NMR (100 MHz, CDCl3): δ(ppm) = 191.6 (d, 1 C, CHO, 3J = 19.6 Hz), 141.2 (d,

1 C, CφCHO, 2J = 26.6 Hz), 138.5 (d, 1 C, CφP, J = 14.8 Hz), 136.14 (d, 2 C, CφP, J

= 9.9 Hz), 134.2–133.6 (6 C, CφH), 130.6 (d, 1 C, CφH, J = 3.9 Hz), 129.1–128.7 (7 C, CφH).

31P-NMR (80 MHz, CDCl3): δ(ppm) = -13.39.

HRMS (ESI): m/z calculated for C19H16OP ([M +H]+) = 291.09333 (100 %), 292.09668

(20.5 %), 293.10004 (2.0 %); m/z found = 291.09265 (100 %), 292.09589 (20.0 %), 293.09918

(1.6 %).

7.2.1.7 (R,R)-N,N0-2-(diphenylphosphino)benzyl-cyclohexane-1,2-diamine (11)

The reaction was performed in a different manner than described by Laue [88]. A vigorously

stirred solution of diaminocyclohexane (10 mmol, 1.14 g) in DCM (100 mL) was cooled in an

ice bath (temperature between 3 ◦C and 10 ◦C), the reaction being carried out under inert and

anhydrous conditions. Over 5 h, a solution of compound 9(9 mmol, 2.61 g) in DCM (50 mL)

was added dropwise. After complete addition, the mixture was further stirred for 1.5 h keeping

the temperature beneath 10 ◦C. DCM was then removed under reduced pressure, and the

residue was dissolved in methanol (100 mL). The solution was again cooled in an ice bath,

sodium borohydride (10 mmol, 0.38 g) was carefully added, and the mixture was stirred for

7 Experimental

20 h at room temperature. In order to quench the reaction, water (20 mL) was added. After

extraction with DCM, the organic phase was dried over MgSO4. Column chromatography

(eluent: CHCl3/MeOH/NH3, 95/5/1) and concentration under reduced pressure yielded a

clear, highly viscous, yellow oil (Y11 = 72 %).

1H-NMR (400 MHz, CDCl3): δ(ppm) = 7.62 (m, 1 H, CφH), 7.51 (m, 1 H, CφH), 7.4–7.2

(m, 10 H, CφH), 7.15 (m, 1 H, CφH), 6.88 (m, 1 H, CφH), 4.10 (dd, 1 H, CH2N, J = 1.5

Hz, J = 12.9 Hz), 3.91 (dd, 1 H, CH2N, J = 2.0 Hz, J = 12.9 Hz), 2.27 (m, 1 H, CHN),

2.09–2.00 (m, 2 H, CHN/CH2), 1.97 (br s, 3 H, NH/NH2), 1.85 (m, 1 H, CH2), 1.69–1.59

(m, 2 H, CH2), 1.27–1.03 (m, 3 H, CH2), 0.88 (m, 1 H, CH2).

13C-NMR (100 MHz, CDCl3): δ(ppm) = 145.2 (d, CφCH2, J = 24.1 Hz), 137.0 (d, 1 C, CφP,

J = 8.6 Hz), 136.9 (d, 1 C, CφP, J = 8.5 Hz), 135.7 (d, 1 C, CφP, J = 13.3 Hz), 134.0–133.7

(5 C, CφH), 129.3 (d, 1 C, CφH, J = 5.4 Hz), 129.1–128.5 (7 C, CφH), 127.2 (CφH), 63.4

(CHNH), 55.4 (CHNH2), 49.6 (d, 1 C, CH2N, J = 21.6 Hz), 35.2 (CH2), 31.36 (CH2), 25.3

(CH2), 25.2 (CH2).

31P-NMR (80 MHz, CDCl3): δ(ppm) = -17.85.

HRMS (ESI): m/z calculated for C25H30N2P ([M +H]+) = 389.21411 (100 %), 390.21747

(27.0 %); m/z found = 389.21329 (100 %), 390.21631 (28.0 %).

7.2.1.8 (R,R)-N-[2-(Diphenylphosphino)benzyl-5-(hex-5-ynoic acid tert-butyl

ester)]-N0-(2-diphenylphosphino)benzyl-cyclohexane-1,2-diamine (12)

A solution of compound 11 (2 mmol, 0.78 g) and compound 7(2 mmol, 0.91 g) in 15 mL of

DCM was stirred at room temperature under inert and anhydrous conditions for 16 h. DCM

was then removed under reduced pressure, and the residue was dissolved in methanol (20 mL).

While the flask was being cooled in an ice bath, sodium borohydride (10 mmol, 0.38 g) was

added carefully. The mixture was stirred at room temperature, and the reaction was quenched

through the addition of water (20 mL) after 20 h. After extraction with DCM, the organic

7 Experimental

phase was dried over MgSO4. Column chromatography (eluent: CHCl3/MeOH/NH3, 95/5/1)

and solvent removal gave a slightly yellow, foamy solid in 80 % yield.

1H-NMR (400 MHz, CDCl3): δ(ppm) = 7.55–7.10 (m, 25 H, CφH), 6.82 (m, 1 H, CφH),

6.75 (m, 1 H, CφH), 4.04 (d, 1 H, J = 13.9 Hz, CH2N), 3.86 (d, 1 H, CH2N, J = 13.3 Hz),

3.94 (d, 1 H, CH2N, J = 13.6 Hz) 3.78 (dd, 1 H, CH2N, J = 1.3 Hz, J = 13.5 Hz), 2.42 (t,

2 H, CH2COOtBu, J = 7.0 Hz), 2.37 (t, 2 H, CH2Calkynyl , J = 7.5 Hz), 2.21–2.10 (br m, 2

H, CHN), 1.98 (m, 2 H, CH2), 1.84 (p, 2 H, CH2CH2CH2, J = 7.3 Hz), 1.67–1.53 (m, 2 H,

CH2), 1.44 (s, 9 H, CH3), 1.3–1.01 (m, 2 H, CH2), 0.97–0.82 (br m, 2 H, CH2).

13C-NMR (100 MHz, CDCl3): δ(ppm) = 172.5 (COOtBu), 137.0–136.5 (5 C, tert Cφ),

135.6–135.3 (3 C, tert Cφ), 134.0–133.8 (8 C, CφH), 133.3 (d, 2 C, CφH, J = 8.1 Hz), 131.9

(d, 1 C, CφH, J = 4.9 Hz), 130.0 (CφH), 129.04–128.45 (15 C, CφH), 127.0 (CφH), 124.5

(CφCalkynyl ), 90.1 (Calkynyl CH2), 81.2 (Calkynyl Cφ), 80.2 [C(CH3)3] , 61.2 (CHN), 60.9

(CHN), 49.0 (d, 2 C, CH2N, J = 22.3 Hz), 34.5 (CH2), 31.3 (2 C, CH2), 28.1 (3 C, CH3),

25.0 (CH2), 24.9 (CH2), 24.2 (CH2), 19.0 (CH2).

31P-NMR (80 MHz, CDCl3): δ(ppm) = -17.71.

HRMS (ESI): m/z calculated for C54H59N2O2P2([M +H]+) = 829.40463 (100%), 830.40798

(58.4 %), 831.41134 (16.7%); m/z found = 829.40594 (100%), 830.40881 (59.4 %),

831.41211 (16.5 %).

7.2.1.9 (R,R)-N-[2-(Diphenylphosphino)benzyl-5-(hex-5-ynoic

acid)]-N0-(2-diphenylphosphino)benzyl-cyclohexane-1,2-diamine (13)

Under inert and anhydrous conditions, compound 12 (0.96 mmol, 0.80 g) was dissolved

in DCM (6 mL), and trifluoroacetic acid (1.5 mL) was added. The solution was stirred at

room temperature over night. After concentration under reduced pressure, the residue was

dissolved in methanol (20 mL) and mixed with an aequeous solution of NaOH (0.1 M, 25 mL).

Extraction with DCM, drying over MgSO4and concentration under reduced pressure yielded

7 Experimental

the product as a white foamed solid (Y13 = 94 %).

1H-NMR (400 MHz, CDCl3): δ(ppm) = 7.70 (m, 1 H, CφH), 7.62 (m, 1 H, CφH), 7.37–7.10

(m, 23 H, CφH), 6.91 (m, 1 H, CφH), 6.75 (m, 1 H, CφH), 4.36 (d, 1 H, CH2N, J =

13.3 Hz), 4.20 (dd, 1 H, CH2N, J = 2.2 Hz, J = 13.3 Hz), 4.18 (dd, 1 H, CH2N, J =

2.0 Hz, J = 13.0 Hz), 3.84 (d, 1 H, CH2N, J = 13.0 Hz), 2.56 (m, 1 H, CHN), 2.52–2.40

(m, 5 H, CHN/CH2COOtBu/CH2Calkynyl ), 2.20–2.05 (m, 2 H, CH2), 1.94–1.85 (m, 2 H,

CH2CH2CH2), 1.76–1.63 (m, 2 H, CH2), 1.46–0.82 (m, 4 H, CH2).

13C-NMR (100 MHz, CDCl3): δ(ppm) = 176.1 (COOH), 140.7 (d, CφCH2, J = 23.8 Hz),

138.5 (d, CφCH2, J = 24.6 Hz), 136.6 (d, CφP, J = 13.8 Hz), 135.6–135.2 (5 C, tert Cφ),

134.1–133.4 (11 C, CφH), 130.8–130.7 (3 C, CφH), 130.0 (CφH), 129.2–128.7 (12 C, CφH),

125.3 (CφCalkynyl ), 91.2 (Calkynyl CH2), 81.1 (Calkynyl Cφ), 59.7 (CHN), 59.2 (CHN), 47.7 (d,

1 C, CH2N, J = 21.7 Hz), 46.7 (d, 1 C, CH2N, J = 22.5 Hz), 32.2 (CH2), 29.7 (CH2), 28.4

(CH2), 24.2 (CH2), 24.2 (CH2), 23.3 (CH2), 19.2 (CH2).

31P-NMR (121 MHz, CDCl3): δ(ppm) = -18.01/-18.62.

HRMS (ESI): m/z calculated for C50H51N2O2P2([M +H]+) = 773.34203 (100%), 774.34538

(54.1 %), 775.34874 (14.3%); m/z found = 773.34100 (100%), 774.34412 (56.2 %),

775.34741 (14.7 %).

IR (ATR): ν(cm−1) = 1717, 1665, 1588, 1434, 1386, 1196, 1132, 1026, 996, 830, 798, 743,

720, 696.

7.2.1.10 Dichloro{(R,R)-N-[2-(Diphenylphosphino)benzyl-5-(hex-5-ynoic acid)]-N0-

(2-diphenylphosphino)benzyl-cyclohexane-1,2-diamine}ruthenium(II) (14)

Compound 13 (0.5 mmol, 0.38 g) and RuCl2(DMSO)4(0.6 mmol, 0.29 g) were dissolved

in toluene (40 mL). The solution was stirred under reflux conditions for 2 h and was then

concentrated under reduced pressure. The residue, an orange-brown solid film, was purified

7 Experimental

via column chromatography (eluent: DCM/acetone, 10/2) and yielded the product, which

appeared now yellow-brown (Y14 = 65 %).

1H-NMR (400 MHz, CDCl3): δ(ppm) = 7.34–7.18 (m, 10 H, CφH), 7.16–7.10 (m, 4 H,

CφH), 7.08–6.94 (m, 13 H, CφH), 4.71 (m, 2 H, CH2N), 4.14–4.02 (m, 2 H, CH2N), 4.00–3.91

(br m, 2 H, CHN), 3.04–2.97 (br m, 2 H, NH), 2.82–2.75 (m, 2 H, CH2), 2.56–2.51 (t, 2 H,

CH2COOtBu, J = 7.4 Hz), 2.51–2.47 (t, 2 H, CH2Calkynyl , J = 6.9 Hz), 1.96–1.88 (m, 2 H,

CH2CH2CH2), 1.88–1.83 (br m, 2 H, CH2), 1.32–1.14 (m, 4 H, CH2).

13C-NMR (100 MHz, CDCl3): δ(ppm) = 178.4 (COOH), 140.9 (d, 2 C, tert Cφ, J = 14.5

Hz), 136.4–135.8 (6 C, tert Cφ), 134.7–127.0 (CφH), 124.9 (CφCalkynyl ), 90.4 (Calkynyl CH2),

80.9 (Calkynyl Cφ), 63.2 (2 C, CHN), 52.4 (d, 1 C, CH2N, J = 4.8 Hz), 52.2 (d, 1 C, CH2N, J =

4.8 Hz), 32.7 (CH2), 30.9 (CH2), 30.9 (CH2), 24.9 (CH2), 24.9 (CH2), 23.6 (CH2), 18.9 (CH2).

31P-NMR (121 MHz, CDCl3): δ(ppm) = 40.88.

HRMS (ESI): m/z calculated for C50H50Cl2N2O2P2Ru ([M]+) = 943.17749 (54.1 %),

944.17626 (100 %), 945.17961 (54.1 %), 946.17331 (63.9 %); m/z found = 943.17554

(62.0 %), 944.17419 (100 %), 945.17511 (66.4 %), 946.17328 (95.1 %).

IR (ATR): ν(cm−1) = 1730, 1707, 1664, 1595, 1481, 1433, 1401, 1188, 1132, 1089, 1043,

961, 831, 743, 694.

7.2.1.11 Dichlorotetrakis(dimethyl sulphoxide)ruthenium(II)

RuCl2(DMSO)4was prepared as described by Wilkinson et al. [198]. Ruthenium trichloride

trihydrate was refluxed in dimethyl sulphoxide for 10 min. After cooling down to room temper-

ature, the complex was precipitated from the dimethyl sulphoxide solution as an orange-yellow

powder with acetone, and was purified by recrystallisation from hot dimethyl sulphoxide.

7.2.1.12 Di-µ-chloro-dichlorodihydridobis(cyclco-octa-l,5-diene)di-iridium(III)

[IrHCl2(COD)]2was prepared as described by Robinson and Shaw [199]. Chloroiridic acid was

dissolved in ethanol and heated to 95 ◦C for 30 min. Cyclo-octa-1,5-diene was added, and the

mixture was stirred for further 2 h under reflux conditions. The flask was cooled down to about

0◦C, and the precipitate, a white-yellow powder, was filtered off and washed with ethanol and

diethyl ether.

7 Experimental

7.2.1.13 (R,R)-N,N0-bis[2-(Diphenylphosphino)benzyl]-cyclohexane-1,2-diamine (17)

Following a procedure used by Gao and co-workers [165], a mixture of (R,R)-1,2-

diaminocyclohexane (3 mmol, 0.34 g), 2-(diphenylphosphino)benzaldehyde (6.0 mmol,

1.74 g), and Na2SO4(18.0 mmol, 2.56 g) in DCM (15 mL) was stirred for 20 h under

inert and anhydrous conditions. DCM was then removed under reduced pressure, and the

residue was dissolved in methanol (15 mL). While the flask was cooled in an ice bath,

sodium borohydride (30 mmol, 1.13 g) was added carefully. The mixture was stirred at room

temperature, and the reaction was quenched by the addition of water (20 mL) after 20 h. After

extraction with DCM, the organic phase was dried over MgSO4. Column chromatography (elu-

ent: DCM/acetone, 10/2) and solvent removal gave a slightly yellow, foamy solid in 80 % yield.

HRMS (ESI): m/z calculated for C44H45N2P2([M +H]+) = 663.30525 (100 %), 664.30860

(47.6 %), 665.31196 (11.1%); m/z found = 663.30432 (100%), 664.30743 (47.9 %),

665.31091 (11.0 %).

7.2.1.14 Dichloro{(R,R)-N,N0-bis[2-(diphenylphosphino)benzyl]-cyclohexane-1,2-

diamine}ruthenium(II) (18)

Similar to the procedure described by Noyori et al. [86], compound 18 was prepared by the

reaction of 17 (0.48 g, 0.72 mmol) and RuCl2(DMSO)4(0.53 g, 1.08 mmol) in toluene

(20 mL) under reflux conditions for 6 h. After filtration, the solution was concentrated under

reduced pressure. Silica gel chromatography of the crude product with DCM and acetone

(2/1) afforded the product as yellow-orange crystals in 84% yield. The same product (78 %

yield) was obtained when the reaction was performed in 2-propanol instead of toluene under

otherwise equal conditions.

HRMS (ESI): m/z calculated for C44H44Cl2N2P2Ru ([M]+) = 833.14071 (54.1 %), 834.13948

(100 %), 835.14283 (47.6 %); m/z found = 833.14075 (59.6 %), 834.13971 (100 %), 835.14075

(62.9 %).

7.2.2 Synthesis of tethered Rh(III)-catalyst

A detailed description of the preparation of compound 20 is provided in the supporting infor-

mation of reference [179]. Compound 20 was received from the group of Prof. Haag and was

converted into the corresponding rhodium complex and the methylated complex, respectively.

7 Experimental

7.2.2.1 Chloro(S,S)-h2-methyl-[2,3,4,5-tetramethylcyclopentadienyl]phenyl-{1,2-

diphenyl-2-([4-amidophenyl sulfonyl]amido-hexanoic

acid)-ethyl}aminoirhodium(III) (21)

Compound 20 (0.8 mmol, 674 mg) and RhCl3·H2O (0.96 mmol, 201 mg) were dissolved in

THF (50 mL). After stirring for 20 h under reflux conditions, triethylamine (4 mmol, 554 µL)

was added. The mixture was stirred for further 3 h and then concentrated under reduced

pressure. Purification via column chromatography (eluent: dichloromethane/methanol, 9/1)

yielded a red solid as the product (Y21 = 50 %).

HRMS (ESI): m/z calculated for C42H45N3O5RhS ([M −Cl]+) = 806.21295 (100 %),

807.21630 (45.43 %), 808.21966 (10.07 %); m/z found = 806.21259 (100 %), 807.21558

(47.82 %), 808.21832 (12.27 %).

IR (ATR): ν(cm−1) = 1727, 1684, 1592, 1531, 1496, 1455, 1399, 1373, 1308, 1253, 1155,

1128, 1081, 1037, 1024, 935, 899, 835, 807, 750, 698.

7.2.2.2 Chloro(S,S)-h2-methyl-[2,3,4,5-tetramethylcyclopentadienyl]phenyl-{1,2-

diphenyl-2-([4-amidophenyl sulfonyl]amido-hexanoic acid methyl

ester)-ethyl}aminoirhodium(III) (23)

Compound 20 (0.25 mmol, 175 mg) and RhCl3·H2O (0.25 mmol, 52 mg) were dissolved

7 Experimental

in methanol (20 mL). After stirring for 20 h under reflux conditions, triethylamine (70 µL,

0.5 mmol) was added. The mixture was stirred for further 3 h and then concentrated under

reduced pressure. Purification via column chromatography (eluent: dichloromethane/methanol,

9/1) yielded a red solid as the product (Y23 = 60 %).

1H-NMR (400 MHz, CDCl3): δ(ppm) = 7.51 (dd, 1H, CφH, J = 6.2Hz, J = 14.4Hz), 7.43

(d, 1 H, CφH, J = 7.1 Hz), 7.34 (dd, 2 H, CφH, J = 5.1 Hz, J = 11.3 Hz), 7.26–7.05 (m,

8 H, CφH), 6.90 (d, 1H, CφH, J = 7.4 Hz), 6.70 (m, 1 H, CφH), 6.61 (t, 2 H, CφH, J =

7.6 Hz), 6.45 (d, 2 H, CφH, J = 7.5 Hz), 5.00 (br d, 1 H, CH2NH, J = 12.5 Hz), 4.27

(d, 1 H, Ntosyl CH, J = 11.1 Hz), 4.22 (br d, 1 H, NHCH2, J = 13.8 Hz), 3.66 (m, 1 H,

NHCH2), 3.56 (s, 3 H, OCH3), 3.24 (t, 1 H, NHCH, J = 11.7 Hz), 2.40–2.25 (m, 4 H,

CH2CH2CH2CH2), 2.04 (s, 3 H, CpCH3), 1.92 (s, 3 H, CpCH3), 1.80 (s, 3 H, CpCH3), 1.67

(br s, 4 H, CH2CH2CH2CH2), 1.52 (s, 3 H, CpCH3).

13C-NMR (100 MHz, CDCl3): δ(ppm) = 174.1 (NCO), 171.2 (COOCH3), 139.9–135.3 (5

C, tert Cφ), 126.7 (Cφ), 131.5–126.6 (CφH), 118.2 (CφH), 106.4 (CCH3), 99.4 (CCH3), 97.1

(CCH3), 88.3 (CCH3), 80.7 (d, CCH3), 75.8 (CHNtosyl ), 69.8 (CHNCH2), 52.1 (CH2N), 51.6

(COOCH3), 37.0 (CH2), 33.7 (CH2), 24.8 (CH2), 24.3 (CH2), 10.6 (CH3C), 10.5 (CH3C),

10.2 (CH3C), 8.1 (CH3C).

HRMS (ESI): m/z calculated for C43H47N3O5RhS ([M −Cl]+) = 820.22860 (100 %),

821.23195 (46.51 %), 822.23531 (10.56 %); m/z found = 820.22742 (100 %), 821.23041

(48.20 %), 822.23309 (11.13 %).

IR (ATR): ν(cm−1) = 1735, 1693, 1592, 1529, 1497, 1455, 1401, 1354, 1309, 1255, 1157,

1130, 1090, 1025, 1004, 935, 912, 839, 807, 760, 731, 700.

7.3 Characterization of Supports and Coupling

7.3.1 Determination of Amine Loading

7.3.1.1 Cleavage of Protective Groups

Surface-functionalized support materials were received with Boc-protected amino groups. Cleav-

age was performed in a beaker. For example, six PCs (sinter chips, amine loading 1 µmol cm−2,

2×2 cm2each) were added to 75 mL of a 6 N solution of HCl in 2-propanol. After stirring for

45 minutes, the PCs were accurately washed with methanol and dried under reduced pressure.

7 Experimental

7.3.1.2 Qualitative Test

A solution of diisopropylethylamine (DIPEA, 10 %) in DMF and a 1 M solution of 2,4,6-

trinitrobenzenesulfonic acid (TNBS) in DMF (about 2 mL in each case) were mixed in a

beaker at room temperature. A small sample of the deprotected support material was added,

and the mixture was left for 10 min. The support material was washed, and the amine loading

was classified by the order of staining: intensely orange (high loading), yellow or slightly orange

(moderate or low loading), colorless (no loading).

7.3.1.3 Quantification of Accessible Amine-Groups

Using the example of beads, a sample of about 50 mg of deprotected PBs were accurately

weighed in an Eppendorf micro test tube and mixed with 0.25 mL of a 0.8 M solution of

DIPEA in DMF. To this suspension, 35 µL of a fresh 0.63 M solution of Fmoc-

-Alanin-OPfp

in N-methyl-2-pyrrolidone (NMP) were added, and the tube was vigorously shaken. After 20 min

further 35 µL of the same solution were added, the tube was shaken again, and left for another

20 min. Then the tube was centrifugalized, and the solution was carefully decanted. Great

importance was attached to the adherence to the following washing procedure: the particles-

containing tube was filled with 0.8 mL of an organic solvent, vigorously shaken, centrifugalized,

and the solvent was decanted. The washing was conducted five times using DMF (twice),

then ethanol (once), and again DMF (twice). As cleavage reagent piperidine (0.8 mL of a

2.3 M solution in DMF) was added, the tube was shaken and left for 45 min. The solution

was filled in a UV-transparant cuvette using a syringe filter to ensure that it was particle-free.

The concentration of the cleaved fluorenyl derivative was calculated via the Lambert-Beer law

(= 7 800 L mol−1cm−1) from the absorbtion determined at λ= 301 nm.

7.3.2 Immobilization and Determination of Catalyst Loading

7.3.2.1 General Procedure for Immobilization

The polymeric chips were added to a solution of DIPEA (145 µmol, 25 µL) in DMF (30 mL)

in an atmosphere of argon. Separately compound 21 (32 µmol, 27 mg) and o-(benzotriazol-1-

yl)-N,N,N0,N0-tetramethyluronium tetrafluoroborate (TBTU; 36 µmol, 11.6 mg) were dissolved

in DMF, and the solution was added to the PCs in DIPEA/DMF. After tentative shaking the

flask for 20 h, the PCs were removed from the solution, washed with DMF and methanol, and

dried under reduced pressure. Via ICP-based analysis of the rhodium content on the supports,

a yield of up to 60 % was determined.

IR (ATR): ν(cm−1) = 1651, 1590, 1535, 1494, 1453, 1401, 1371, 1307, 1252, 1214, 1148,

1125, 1099, 1034, 938, 895, 804, 759, 698.

7 Experimental

7.3.2.2 General Procedure for ICP-OES Analysis

Pressurized digestion of catalyst-loaded support materials was performed by microwave treat-

ment. A weighed sample of the supported catalyst (∼100 mg) was mixed with nitric acid (65 %

5 mL) and hydrochloric acid (37 % 1 mL). Within 5 min the sample was heated to 200 ◦C with

a fixed microwave power of 200 W, and the temperature was hold for another 5 min. After

digestion, the samples were filled up to 25 mL with water and subjected to ICP-OES analysis

(usually two runs for each sample).

7.4 Catalysis Experiments

7.4.1 Homogeneous Catalysts

7.4.1.1 General Procedure for ATH Experiments Using the Ru-PNNP Catalyst

Complex 18 (10 µmol, 8.34 mg) was dissolved in 2-propanol (20 mL) under inert and anhydrous

conditions. Acetophenone (5 mmol, 584 µL) was added, and the mixture was stirred at 45 ◦C.

The reaction was started upon addition of a solution of potassium propan-2-olate (0.1 M,

200 µL).

7.4.1.2 General Procedure for ATH Experiments Using the Ir-PNNP Catalyst

Compound 17 (30 µmol, 20 mg) and [IrHCl2(COD)]2(13.4 µmol, 10 mg) were dissolved in

toluene (2 mL) and stirred for 1 h at 50 ◦C. A solution of sodium formate (1.9 mmol, 0.13 g)

in water (20 mL) was added. The reaction was startet by addition of acetophenone (1.2 mmol,

0.14 mL).

7.4.2 Heterogenized Catalysts

7.4.2.1 General Procedure for ATH Experiments Using PC-supported Catalysts

The standard experiment was performed in a 25-mL flask using a magnetic stirrer with haeting.

To a stirred solution of sodium formate (5 mmol, 340 mg) and acetophenone (1 mmol, 117 µL)

in 10 mL of neat water at 40 ◦C, a catalyst-loaded PC (22 b, 2×2 cm2) was added. Samples of

50 µL were taken from the reaction mixture, diluted with ethanol, and analyzed via GC-MS for

determination of the conversion. The ee was determined from samples taken after extraction

with ethyl acetate or DCM when the reaction was finished.

7 Experimental

7.4.2.2 General Procedure for the Recycling of PC-supported Catalysts

After the reaction was finished, the catalyst chip was removed with forceps and placed in a

beaker with 10 mL of methanol for about 5 minutes. Then, the chip was rinsed with methanol

and dried under reduced pressure. When no subsequent experiment was performed, e. g., over

night, the chip was stored in a refrigerator.

7.4.2.3 General Procedure for ATH Experiments Using PB-supported Catalysts

Experiments were carried out in a jacketed 100-mL flask equipped with a mechanical stirrer

with stainless steel ringed propeller as well as an inside temperature gauge. In the standard

experiment the temperature was adjusted to 40 ◦C, and a stirring speed of 1 200 rpm was

applied. The reaction mixture consisted of 1.5 g of catalyst beads (22 a, rhodium content

≈0.105 mM) suspended in 100 mL of an aqueous solution of sodium formate (0.4 M) and

acetophenone (0.04 M). The order of adding the components was as follows: successively about

50 mL of dissolved sodium formate, the catalyst beads, and the remaining solution of sodium

formate were added to the preheated reaction vessel. A couple of minutes of stirring was

required to provide a homogeneous suspension at the designated temperature. The reaction

was started by the addition of acetophenone. The reactor was kept closed except when samples

(∼200 µL) were taken with a syringe. The samples were prepared for GC-MS analysis by

filtering off the particles with a syringe filter.

Bibliography

[1] A. M. Rouhi, Chem. Eng. News 2004,82, 47–62.

[2] R. A. Sheldon, I. Arends, U. Hanefeld, Green Chemistry and Catalysis, Wiley-VCH, Wein-

heim, 2007, p. 1.

[3] B. Cornils, W. A. Herrmann, J. Catal. 2003,216, 23–31.

[4] B. Cornils, W. A. Herrmann, I. T. Horváth, W. Leitner, S. Mecking, H. Olivier-Bourbigou,

D. Vogt (Eds.), Multiphase Homogeneous Catalysis,Vol. 1, Wiley-VCH, Weinheim, 2005,

pp. 3–23.

[5] H.-U. Blaser, B. Pugin in Handbook of Asymmetric Heterogeneous Catalysis, K. Deng,

Y. Uozumi (Eds.), Wiley-VCH, Weinheim, 2008, pp. 413–437.

[6] E. L. Eliel, S. H. Wilen, Organische Stereochemie, Wiley-VCH, Weinheim, 1998, pp.

533–583.

[7] H. Y. Aboul-Enein, I. W. Wainer, The Impact of Stereochemistry on Drug Development

and Use,Vol. 42, John Wiley and Sons, New York, 1997, pp. 10–17.

[8] FDA, Development of New Stereoisomeric Drugs,5/1/1992.

[9] J. Clayden, N. Greeves, S. Warren, P. Wothers, Organic Chemistry, Oxford University

Press, 2001, p. 386.

[10] Z. Wang, K. Ding, Y. Uozumi in Handbook of Asymmetric Heterogeneous Catalysis,

K. Deng, Y. Uozumi (Eds.), Wiley-VCH, Weinheim, 2008, pp. 1–24.

[11] E. L. Eliel, S. H. Wilen, Organische Stereochemie, Wiley-VCH, Weinheim, 1998, pp.

185–289.

[12] W. J. Lough in Chirality in Natural and Applied Scince, W. J. Lough, I. W. Wainer (Eds.),

Blackwell, 2002, pp. 179–202.

[13] R. Noyori, H. Takaya, Acc. Chem. Res. 1990,23, 345–350.

[14] H.-U. Blaser, B. Pugin, F. Spindler, J. Mol. Cat. A: Chem. 2005,231, 1–20.

Bibliography

[15] A. Behr, Angewandte homogene Katalyse, Wiley-VCH, Weinheim, 2008, pp. 34–47.

[16] J. Hagen, Industrial Catalysis, 2nd ed., Wiley-VCH, Weinheim, 2006, pp. 1–14.

[17] J. Weitkamp, R. Gläser in Winnacker/Küchler: Chemische Technik, 5th ed.,Vol. 1,

R. Dittmeyer, W. Keim, G. Kreysa, A. Oberholz (Eds.), Wiley-VCH, Weinheim, 2004,

pp. 645–718.

[18] J. Hagen, Industrial Catalysis, 2nd ed., Wiley-VCH, Weinheim, 2006, p. 9.

[19] A. M. Palmer, A. Zanotti-Gerosa in Asymmetric Catalysis on Industrial Scale: Challenges,

Approaches, and Solutions, 2nd ed., H.-U. Blaser, H.-J. Federsel (Eds.), Wiley-VCH,

Weinheim, 2010, pp. 61–78.

[20] H.-U. Blaser, G. Hoge, B. Pugin, F. Spindler in Green Catalysis - Homogeneous Catalysis,

Vol. 1 of Handbook of Green Chemistry, R. H. Crabtree (Ed.), Wiley-VCH, Weinheim,

2009, pp. 153–203.

[21] A. Behr, Angewandte homogene Katalyse, Wiley-VCH, Weinheim, 2008, p. 41.

[22] M. M. Green, H. A. Wittcoff, Organic Chemistry Principles and Industrial Practice, Wiley-

VCH, Weinheim, 2003, p. 249.

[23] W. Kersten, E.-M. Kern in Betriebswirtschaftslehre für Chemiker, G. Festel, A. Hassan,

J. Leker, P. Bamelis (Eds.), Springer-Verlag, Berlin Heidelberg, 2001, pp. 248–334.

[24] R. A. Sheldon, I. Arends, U. Hanefeld, Green Chemistry and Catalysis, Wiley-VCH, Wein-

heim, 2007, p. 37.

[25] H.-U. Blaser, H.-J. Federsel (Eds.), Asymmetric Catalysis on Industrial Scale, 2nd ed.,

Wiley-VCH, Weinheim, 2010, pp. xxix–xxxvii.

[26] P. T. Anastas, J. C. Warner, Green Chemistry: Theory and Practice, Oxford University

Press, New York, 1998, p. 30.

[27] P. T. Anastas, J. B. Zimmerman, Env. Sci. Tech. 2003,37, 94A–101A.

[28] A. Moores in Green Catalysis - Homogeneous Catalysis, Vol. 1 of Handbook of Green

Chemistry, R. H. Crabtree (Ed.), Wiley-VCH, Weinheim, 2009, pp. 1–15.

[29] R. A. Sheldon, Chemistry and Industry 1992, 903–906.

[30] R. A. Sheldon, I. Arends, U. Hanefeld, Green Chemistry and Catalysis, Wiley-VCH, Wein-

heim, 2007, pp. 2–5.

Bibliography

[31] C. Hedberg in Modern Reduction Methods, P. G. Andersson, I. J. Munslow (Eds.), Wiley-

VCH, Weinheim, 2008, pp. 109–134.

[32] R. Noyori, Angew. Chem. Int. Ed. 2002,41, 2008–2022.

[33] A. Miyashita, A. Yasuda, H. Takaya, K. Toriumi, T. Ito, T. Souchi, R. Noyori, J. Am.

Chem. Soc. 1980,102, 7932–7934.

[34] R. Noyori, T. Ohkuma, Angew. Chem. Int. Ed. 2001,40, 40–73.

[35] R. Noyori, T. Ohkuma, M. Kitamura, H. Takaya, N. Sayo, H. Kumobayashi, S. Akuta-

gawa, J. Am. Chem. Soc. 1987,109, 5856–5858.

[36] T. Ohkuma, H. Ooka, S. Hashiguchi, T. Ikariya, R. Noyori, J. Am. Chem. Soc. 1995,

117, 2675–2676.

[37] T. Ohkuma, M. Koizumi, K. Muñiz, G. Hilt, C. Kabuto, R. Noyori, J. Am. Chem. Soc.

2002,124, 6508–6509.

[38] J. C. Fiaud, H. B. Kagan, Bull. Soc. Chim. Fr. 1969,8, 2742–2743.

[39] C. R. Johnson, C. J. Stark, Tetrahedron Lett. 1979,20, 4713–4716.

[40] A. Hirao, S. Itsuno, S. Nakahama, N. Yamazaki, J. Chem. Soc., Chem. Commun. 1981,

315–317.

[41] E. J. Corey, C. J. Helal, Angew. Chem. Int. Ed. 1998,37, 1986–2012.

[42] S. Itsuno in Comprehensive Asymmetric Catalysis,Vol. 1, E. N. Jacobsen, A. Pfaltz,

H. Yamamoto (Eds.), Springer-Verlag, Berlin Heidelberg, 2000, pp. 289–316.

[43] E. J. Corey, R. K. Bakshi, S. Shibata, J. Am. Chem. Soc. 1987,109, 5551–5553.

[44] N. Arai, T. Ohkuma in Modern Reduction Methods, P. G. Andersson, I. J. Munslow

(Eds.), Wiley-VCH, Weinheim, 2008, pp. 159–181.

[45] H. Nishiyama in Comprehensive Asymmetric Catalysis,Vol. 1, E. N. Jacobsen, A. Pfaltz,

H. Yamamoto (Eds.), Springer-Verlag, Berlin Heidelberg, 2000, pp. 267–288.

[46] S. Rendler, M. Oestreich in Modern Reduction Methods, P. G. Andersson, I. J. Munslow

(Eds.), Wiley-VHC, Weinheim, 2008, pp. 183–208.

[47] A. S. Bommarius, B. R. Riebel, Biocatalysis, Wiley-VCH, Weinheim, 2004, pp. 1–18.

[48] T. P. Yoon, E. N. Jacobsen, Science 2003,299, 1691–1693.

Bibliography

[49] K. Nakamura, T. Matsuada in Modern Reduction Methods, P. G. Andersson, I. J. Mun-

slow (Eds.), Wiley-VCH, Weinheim, 2008, pp. 209–234.

[50] K. Nakamura, K. Takenaka, M. Fujii, Y. Ida, Tetrahedron Lett. 2002,43, 3629–3631.

[51] J. C. Moore, D. J. Pollard, B. Kosjek, P. N. Devine, Acc. Chem. Res. 2007,40, 1412–

1419.

[52] B. Cornils, W. A. Herrmann, R. Schlögl, C.-H. Wong (Eds.), Catalysis from A to Z, 2nd

ed., Wiley-VCH, Weinheim, 2003, p. 780.

[53] M. J. Palmer, M. Wills, Tetrahedron Asymm. 1999,10, 2045–2061.

[54] S. Gladiali, E. Alberico, Chem. Soc. Rev. 2006,35, 226–236.

[55] E. Knoevenagel, B. Bergdolt, Chemische Berichte 1903,36, 2857.

[56] H. Meerwein, R. Schmidt, Justus Liebigs Ann. Chem. 1925,444, 221–238.

[57] W. Ponndorf, Angew. Chem. 1926,39, 138–143.

[58] A. Verly, Bull. Soc. Chim. Fr. 1925,37, 537–542.

[59] R. V. Oppenauer, Recl. Trav. Chim. Pays-Bas 1937,56, 137–144.

[60] W. von E. Doering, R. W. Young, J. Am. Chem. Soc. 1950,72, 631.

[61] Y. Y. Haddad, H. B. Henbest, J. Husbands, T. R. B. Mitchell, Proc. Chem. Soc. 1964,

361.

[62] M. J. Trocha-Grimshaw, H. B. Henbest, Chem. Commun. (London) 1967, 544.

[63] Y. Sasson, J. Blum, Tetrahedron Lett. 1971,12, 2167–2170.

[64] Y. Sasson, J. Blum, J. Org. Chem. 1975,40, 1887–1896.

[65] R. L. Chowdhury, J.-E. Bäkvall, J. Chem. Soc., Chem. Commun. 1991, 1063–1064.

[66] S. Gladiali, R. Taras in Modern Reduction Methods, P. G. Andersson, I. J. Munslow

(Eds.), Wiley-VCH, Weinheim, 2008, pp. 135–157.

[67] R. Noyori, M. Yamakawa, S. Hashiguchi, J. Org. Chem. 2001,66, 7931–7944.

[68] K.-J. Haack, S. Hashiguchi, A. Fujii, T. Ikariya, R. Noyori, Angew. Chem. Int. Ed. 1997,

36, 285–288.

[69] M. Yamakawa, H. Ito, R. Noyori, J. Am. Chem. Soc. 2000,122, 1466–1478.

Bibliography

[70] D. G. Blackmond, A. Armstrong, V. Coombe, A. Wells, Angew. Chem. Int. Ed. 2007,

46, 3798–3800.

[71] P. T. Anastas, M. M. Kirchhoff, Acc. Chem. Res. 2002,35, 686–694.

[72] P. Dunn in Green Solvents - Reactions in Water,Vol. 5, C.-J. Li (Ed.), Wiley-VCH,

Weinheim, 2010, pp. 363–383.

[73] R. Breslow in Green Solvents - Reactions in Water, Vol. 5 of Handbook of Green Chem-

istry, C.-J. Li (Ed.), Wiley-VCH, Weinheim, 2010, pp. 1–29.

[74] X. Wu, X. Li, W. Hems, F. King, J. Xiao, Org. Biomol. Chem. 2004,2, 1818–1821.

[75] C. Bubert, J. Blacker, S. M. Brown, J. Crosby, S. Fitzjohn, J. P. Muxworthy, T. Thorpe,

J. M. Williams, Tetrahedron Lett. 2001,42, 4037–4039.

[76] T. Thorpe, J. Blacker, S. M. Brown, C. Bubert, J. Crosby, S. Fitzjohn, J. P. Muxworthy,

J. M. Williams, Tetrahedron Lett. 2001,42, 4041–4043.

[77] H. Y. Rhyoo, H.-J. Park, Y. K. Chung, Chem. Commun. 2001, 2064–2065.

[78] X. Wu, J. Xiao in Green Solvents - Reactions in Water, Vol. 5 of Handbook of Green

Chemistry, C.-J. Li (Ed.), Wiley-VCH, Weinheim, 2010, pp. 105–149.

[79] X. Wu, J. Xiao, Chem. Commun. 2007, 2449–2466.

[80] J. L. Namy, J. Souppe, J. Collin, H. B. Kagan, J. Org. Chem. 1984,49, 2045–2049.

[81] D. A. Evans, S. G. Nelson, M. R. Gagne, A. R. Muci, J. Am. Chem. Soc. 1993,115,

9800–9801.

[82] J. Yang, B. List, Org. Lett. 2006,8, 5653–5655.

[83] S. Enthaler, G. Erre, M. K. Tse, K. Junge, M. Beller, Tetrahedron Lett. 2006,47,

8095–8099.

[84] S. Zhou, S. Fleischer, K. Junge, S. Das, D. Addis, M. Beller, Angew. Chem. Int. Ed.

2010,49, 8121–8125.

[85] C. Sui-Seng, F. Freutel, A. J. Lough, R. H. Morris, Angew. Chem. Int. Ed. 2008,47,

940–943.

[86] J.-X. Gao, T. Ikariya, R. Noyori, Organometallics 1996,15, 1087–1089.

[87] T. Ohkuma, R. Noyori in Comprehensive Asymmetric Catalysis,Vol. 1, E. N. Jacobsen,

A. Pfaltz, H. Yamamoto (Eds.), Springer-Verlag, Berlin Heidelberg, 2000, pp. 199–246.

Bibliography

[88] S. Laue, Dissertation, Forschungszentrum Jülich, 2002.

[89] T. Ohkuma, M. Kitamura, R. Noyori in New Frontiers in Asymmetric Catalysis,

K. Mikami, M. Lautens (Eds.), John Wiley and Sons, Hoboken, 2007, pp. 1–32.

[90] S. Hashiguchi, A. Fujii, J. Takehara, T. Ikariya, R. Noyori, J. Am. Chem. Soc. 1995,117,

7562–7563.

[91] K. Püntener, L. Schwink, P. Knochel, Tetrahedron Lett. 1996,37, 8165–8168.

[92] J. Takehara, S. Hashiguchi, A. Fujii, S.-i. Inoue, T. Ikariya, R. Noyori, Chem. Commun.

1996, 233–234.

[93] M. Palmer, T. Walsgrove, M. Wills, J. Org. Chem. 1997,62, 5226–5228.

[94] D. G. I. Petra, P. C. J. Kamer, A. L. Spek, H. E. Schoemaker, P. W. N. M. van Leeuwen,

J. Org. Chem. 2000,65, 3010–3017.

[95] D. Matharu, D. Morris, A. Kawamoto, G. Clarkson, M. Wills, Org. Lett. 2005,7, 5489–

5491.

[96] X. Li, J. Blacker, I. Houson, X. Wu, J. Xiao, Synlett 2006, 1155–1160.

[97] R. Noyori, S. Hashiguchi, Acc. Chem. Res. 1997,30, 97–102.

[98] M. Yamakawa, I. Yamada, R. Noyori, Angew. Chem. Int. Ed. 2001,40, 2818–2821.

[99] G. Zassinovic, G. Mestroni, S. Gladiali, Chem. Rev. 1992,92, 1051–1069.

[100] J.-E. Bäckvall, J. Organomet. Chem. 2002,652, 105–111.

[101] R. B. Woodward, N. L. Wendler, F. J. Brutschy, J. Am. Chem. Soc. 1945,67, 1425–1429.

[102] E. C. Ashby, Acc. Chem. Res. 1988,21, 414–421.

[103] C. F. de Graauw, J. A. Peters, H. van Bekkum, J. Huskens, Synthesis 1994, 1007–1017.

[104] R. Cohen, C. R. Graves, S. T. Nguyen, J. M. L. Martin, M. A. Ratner, J. Am. Chem.

Soc. 2004,126, 14796–14803.

[105] G. Zassinovich, R. Bettella, G. Mestroni, N. Bresciani-Pahor, S. Geremia, L. Randaccio,

J. Organomet. Chem. 1989,370, 187–202.

[106] J. S. M. Samec, J.-E. Bäckvall, P. G. Andersson, P. Brandt, Chem. Soc. Rev. 2006,35,

237–248.

Bibliography

[107] Y. R. S. Laxmi, J.-E. Bäckvall, Chem. Commun. 2000, 611–612.

[108] D. Alonso, P. Brandt, S. Nordin, P. Andersson, J. Am. Chem. Soc. 1999,121, 9580–

9588.

[109] D. G. I. Petra, J. N. H. Reek, J.-W. Handgraaf, E. J. Meijer, P. Dierkes, P. C. J. Kamer,

J. Brussee, H. E. Schoemaker, P. W. N. M. van Leeuwen, Chem. Eur. J. 2000,6, 2818–

2829.

[110] O. Pàmies, J.-E. Bäckvall, Chem. Eur. J. 2001,7, 5052–5058.

[111] J. Wettergren, E. Buitrago, P. Ryberg, H. Adolfsson, Chem. Eur. J. 2009,15, 5709–5718.

[112] X. F. Wu, J. K. Liu, D. Di Tommaso, J. A. Iggo, C. R. A. Catlow, J. Bacsa, J. L. Xiao,

Chem. Eur. J. 2008,14, 7699–7715.

[113] R. G. Wilkins, Kinetics and Mechanism of Reactions of Transition Metal Complexes, 2nd

ed., VCH, Weinheim, 1991, pp. 88–89.

[114] J. Blacker, J. Martin in Asymmetric Catalysis on Industrial Scale, H.-U. Blaser, E. Schmidt

(Eds.), Wiley-VCH, Weinheim, 2004, pp. 201–216.

[115] T. Ikariya, A. J. Blacker, Acc. Chem. Res. 2007,40, 1300–1308.

[116] A. J. Blacker, P. Thompson in Asymmetric Catalysis on Industrial Scale: Challenges,

Approaches, and Solutions, 2nd ed., H.-U. Blaser, H.-J. Federsel (Eds.), Wiley-VCH,

Weinheim, 2010, pp. 265–289.

[117] K. B. Hansen, J. R. Chilenski, R. Desmond, P. N. Devine, E. J. J. Grabowski, R. Heid,

M. Kubryk, D. J. Mathre, R. Varsolona, Tetrahedron Asymm. 2003,14, 3581–3587.

[118] H. Yamashita, T. Ohtani, S. Morita, K. Otsubo, K. Kan, J. Matsubara, K. Kitano,

Y. Kawano, M. Uchida, F. Tabusa, Heterocycles 2002,56, 123–128.

[119] Strem Chemicals, Inc.,1/20/2011. http://www.strem.com/uploads/technical_notes/

96-7650tech.pdf.

[120] Takasago,1/20/2011. http://www.takasago.com/finechemicals/ligands_catalysts/

501.pdf.

[121] Johnson Matthey,1/20/2011. http://www.jmcatalysts.com/pharma/site.asp?siteid=

540.

[122] D. E. de Vos, I. F. J. Vankelecom, P. A. Jacobs (Eds.), Chiral Catalyst Immobilization

and Recycling,Vol. 1, Wiley-VCH, Weinheim, 2000, p. 320.

Bibliography

[123] B. Pugin, H.-U. Blaser in Comprehensive Asymmetric Catalysis,Vol. 3, E. N. Jacobsen,

A. Pfaltz, H. Yamamoto (Eds.), Spinger-Verlag, Berlin Heidelberg, 2000, pp. 1367–1375.

[124] G. Oehme in Comprehensive Asymmetric Catalysis,Vol. 3, E. N. Jacobsen, A. Pfaltz,

H. Yamamoto (Eds.), Spinger-Verlag, Berlin Heidelberg, 2000, pp. 1377–1386.

[125] A. Kirschning (Ed.), Immobilized Catalysts: Solid Phases, Immobilization and Applica-

tions,Vol. 242, Springer-Verlag, Berlin Heidelberg, 2004.

[126] M. Heitbaum, F. Glorius, I. Escher, Angew. Chem. Int. Ed. 2006,45, 4732–4762.

[127] K. Ding, Y. Uozumi (Eds.), Handbook of Asymmetric Heterogeneous Catalysis, Wiley-

VCH, Weinheim, 2008.

[128] N. End, K.-U. Schöning in Immobilized Catalysts, of Topics in Current Chemistry,

A. Kirschning (Ed.), Springer-Verlag, Berlin Heidelberg, 2004, pp. 241–271.

[129] D. J. Bayston, C. B. Travers, M. E. C. Polywka, Tetrahedron Asymm. 1998,9, 2015–

2018.

[130] N. A. Cortez, G. Aguirre, M. Parra-Hake, R. Somanathan, Tetrahedron Lett. 2009,50,

2228–2231.

[131] A. J. Sandee, D. G. I. Petra, J. N. H. Reek, P. C. J. Kamer, P. W. N. M. v. Leeuwen,

Chem. Eur. J. 2001,7, 1202–1208.

[132] P. N. Liu, P. M. Gu, F. Wang, Y. Q. Tu, Org. Lett. 2004,6, 169–172.

[133] P. N. Liu, J. G. Deng, Y. Q. Tu, S. H. Wang, Chem. Commun. 2004, 2070–2071.

[134] P.-N. Liu, P.-M. Gu, J.-G. Deng, Y.-Q. Tu, Y.-P. Ma, Eur. J. Org. Chem. 2005,15,

3221–3227.

[135] C.-F. Nie, J.-S. Suo, Chin. J. Chem. 2005,23, 315–320.

[136] X. H. Huang, J. Y. Ying, Chem. Commun. 2007, 1825–1827.

[137] S. Parambadath, A. Singh, Catalysis Today 2009,141, 161–167.

[138] J. Li, Y. Zhang, D. Han, Q. Gao, C. Li, J. Mol. Catal. A: Chem. 2009,298, 31–35.

[139] F. Michalek, A. Lagunas, C. Jimeno, M. A. Pericàs, J. Mater. Chem. 2008,18, 4692–

4697.

[140] S. Laue, L. Greiner, J. Woltinger, A. Liese, Adv. Synth. Catal. 2001,343, 711–720.

Bibliography

[141] M. Aglietto, E. Chiellini, S. D’Antone, G. Ruggeri, R. Solaro, Pure Appl. Chem. 1988,

60, 415–430.

[142] P. Gamez, B. Dunjic, F. Fache, M. Lemaire, J. Chem. Soc., Chem. Commun. 1994,

1417–1418.

[143] P. Gamez, B. Dunjic, C. Pinel, M. Lemaire, Tetrahedron Lett. 1995,36, 8779.

[144] F. Locatelli, P. Gamez, M. Lemaire, J. Mol. Catal. A: Chem. 1998,135, 89–98.

[145] K. Polborn, K. Severin, Chem. Commun. 1999, 2481–2482.

[146] K. Polborn, K. Severin, Chem. Eur. J. 2000,6, 4604–4611.

[147] K. Polborn, K. Severin, Eur. J. Inorg. Chem. 2000,2000, 1687–1692.

[148] M. Tada, Y. Iwasawa, J. Mol. Catal. A: Chem. 2003,199, 115–137.

[149] R. ter Halle, E. Schulz, M. Lemaire, Synlett 1997, 1257–1258.

[150] C. Saluzzo, R. ter Halle, F. Touchard, F. Fache, E. Schulz, M. Lemaire, J. Organomet.

Chem. 2000,603, 30–39.

[151] Y. Arakawa, N. Haraguchi, S. Itsuno, Tetrahedron Lett. 2006,47, 3239–3243.

[152] Y. Arakawa, A. Chiba, N. Haraguchi, S. Itsuno, Adv. Synth. Catal. 2008,350, 2295–2304.

[153] N. Haraguchi, K. Tsuru, Y. Arakawa, S. Itsuno, Org. Biomol. Chem. 2009,7, 69–75.

[154] H.-F. Zhou, Q.-H. Fan, Y.-Y. Huang, L. Wu, Y.-M. He, W.-J. Tang, L.-Q. Gu, A. S.

Chan, J. Mol. Catal. A: Chem. 2007,275, 47–53.

[155] H. Y. Rhyoo, H.-J. Park, W. H. Suh, Y. K. Chung, Tetrahedron Lett. 2002,43, 269.

[156] F. Wang, H. Liu, L. Cun, J. Zhu, J. Deng, Y. Jiang, J. Org. Chem. 2005,70, 9424–9429.

[157] H. Yang, J. Li, J. Yang, Z. Liu, Q. Yang, C. Li, Chem. Commun. 2007, 1086–1088.

[158] S. Bai, H. Yang, P. Wang, J. Gao, B. Li, Q. Yang, C. Li, Chem. Commun. 2010,46,

8145–8147.

[159] PolyAn GmbH,1/22/2011. http://www.poly-an.de.

[160] F. Koc, F. Michalek, L. Rumi, W. Bannwarth, R. Haag, Synthesis 2005, 3362–3372.

[161] U. Schedler, unpublished results.

100

Bibliography

[162] W.-K. Wong, J.-X. Gao, Z.-Y. Zhou, T. C. W. Mak, Polyhedron 1992,11, 2965–2966.

[163] W.-K. Wong, J.-X. Gao, Z.-Y. Zhou, T. C. Mak, Polyhedron 1993,12, 1415–1417.

[164] J.-X. Gao, H.-L. Wan, W.-K. Wong, M.-C. Tse, W.-T. Wong, Polyhedron 1996,15,

1241–1251.

[165] J.-X. Gao, X.-D. Yi, P.-P. Xu, C.-L. Tang, H.-L. Wan, T. Ikariya, J. Organomet. Chem.

1999,592, 290–295.

[166] J.-X. Gao, H. Zhang, X.-D. Yi, P.-P. Xu, C.-L. Tang, H.-L. Wan, K.-R. Tsai, T. Ikariya,

Chirality 2000,12, 383–388.

[167] Z.-R. Dong, Y.-Y. Li, J.-S. Chen, B.-Z. Li, Y. Xing, J.-X. Gao, Org. Lett. 2005,7,

1043–1045.

[168] Y. Xing, J. S. Chen, Z.-R. Dong, Y.-Y. Li, J.-X. Gao, Tetrahedron Lett. 2006,47,

4501–4503.

[169] R. M. Stoop, A. Mezzetti, Green Chemistry 1999,1, 39–41.

[170] R. M. Stoop, S. Bachmann, M. Valentini, A. Mezzetti, Organometallics 2000,19, 4117–

4126.

[171] W.-K. Wong, X.-P. Chen, T.-W. Chik, W.-Y. Wong, J.-P. Guo, F.-W. Lee, Eur. J. Inorg.

Chem. 2003,2003, 3539–3546.

[172] J.-X. Gao, X. D. Yi, C.-L. Tang, P.-P. Xu, H.-L. Wan, Polym. Adv. Technol. 2001,12,

716–719.

[173] L. Greiner, S. Laue, A. Liese, C. Wandrey, Chem. Eur. J. 2006,12, 1818–1823.

[174] I. Klement, P. Knochel, K. Chau, G. Cahiez, Tetrahedron Lett. 1994,35, 1177–1180.

[175] S.-Y. Han, Y.-A. Kim, Tetrahedron 2004,60, 2447–2467.

[176] X. F. Wu, D. Vinci, T. Ikariya, J. L. Xiao, Chem. Commun. 2005, 4447–4449.

[177] D. S. Matharu, D. J. Morris, G. J. Clarkson, M. Wills, Chem. Commun. 2006, 3232 –

3234.

[178] D. Morris, A. Hayes, M. Wills, J. Org. Chem. 2006,71, 7035–7044.

[179] J. Dimroth, J. Keilitz, U. Schedler, R. Schomäcker, R. Haag, Adv. Synth. Catal. 2010,

352, 2497–2506.

101

Bibliography

[180] Y.-C. Chen, T.-F. Wu, J.-G. Deng, H. Liu, X. Cui, J. Zhu, Y.-Z. Jiang, M. C. K. Choi,

A. S. C. Chan, J. Org. Chem. 2002,67, 5301–5306.

[181] T. Koike, T. Ikariya, Adv. Synth. Catal. 2004,346, 37–41.

[182] S. Richards, M. Ropic, D. Blackmond, A. Walmsley, Anal. Chim. Acta 2004,519, 1–9.

[183] A. J. Blacker, S. B. Duckett, J. Grace, R. N. Perutz, A. C. Whitwood, Organometallics

2009,28, 1435–1446.

[184] X. Wu, X. Li, F. King, J. Xiao, Angew. Chem. Int. Ed. 2005,44, 3407–3411.

[185] J. Canivet, G. Labat, H. Stoeckli-Evans, G. Süss-Fink, Eur. J. Inorg. Chem. 2005,2005,

4493–4500.

[186] X. Wu, X. Li, A. Zanotti-Gerosa, A. Pettman, J. Liu, A. Mills, J. Xiao, Chem. Eur. J.

2008,14, 2209–2222.

[187] X. F. Wu, J. K. Liu, X. H. Li, A. Zanotti-Gerosa, F. Hancock, D. Vinci, J. W. Ruan,

J. L. Xiao, Angew. Chem. Int. Ed. 2006,45, 6718–6722.

[188] P. Wessig, unpublished results.

[189] K. A. Connors, Chemical Kinetics, Wiley-VCH, 1990, p. 220.

[190] J. Canivet, G. Süss-Fink, Green Chemistry 2007,9, 391–397.

[191] L. Arnaut, H. Burrows, S. Formosinho, Chemical Kinetics: From Molecular Structure to

Chemical Reactivity, Elsevier, Amsterdam, 2007, p. 244.

[192] L. A. Campbell, T. Kodadek, J. Mol. Catal. A: Chem. 1996,113, 293–310.

[193] M. Watanabe, K. Murata, T. Ikariya, J. Am. Chem. Soc. 2003,125, 7508–7509.

[194] M. Watanabe, A. Ikagawa, H. Wang, K. Murata, T. Ikariya, J. Am. Chem. Soc. 2004,

126, 11148–11149.

[195] G. Bartoli, M. Bosco, A. Carlone, R. Dalpozzo, E. Marcantoni, P. Melchiorre, L. Sambria,

Synthesis 2007, 3489–3496.

[196] B. Neises, W. Steglich, Angew. Chem. Int. Ed. 1978,17, 522–524.

[197] D. B. Dess, J. C. Martin, J. Org. Chem. 1983,48, 4155–4156.

[198] I. P. Evans, A. Spencer, G. Wilkinson, J. Chem. Soc., Dalton Trans. 1973, 204–209.

[199] S. D. Robinson, B. L. Shaw, J. Chem. Soc. 1965, 4997–5001.

102

Appendix

Publications out of this Work

Conference Contributions

•J. Dimroth, M. Schwarze, J. Keilitz, R. Haag, R. Schomäcker, Heterogeneous asym-

metric transfer hydrogenation, 44. Jahrestreffen Deutscher Katalytiker mit Jahrestreffen

Reaktionstechnik, 2011, Weimar; poster presentation.

•M. Schwarze, J. Keilitz, J. Dimroth, S. Nowag, U. Schedler, R. Haag, R. Schomäcker,

Recyclable Metal Catalysts for Hydrogenation Reactions, Joint International Symposium

Activation of Small Molecules – Gas Phase Clusters, Molecular Catalysts, Enzymes and

Solid Materials, 2011, Erkner; poster presentation.

•J. Dimroth, J. Keilitz, U. Schedler, R. Haag, R. Schomäcker, Oberflächenfunktionali-

sierte Polymere als Träger für chirale Katalysatoren - Anwendung in der asymmetrischen

Transferhydrierung, ProcessNet-Jahrestagung, 2010, Aachen; talk.

•J. Dimroth, J. Keilitz, R. Haag, R. Schomäcker, Heterogenisation of a Rh(III)-catalyst

and application to asymmetric transfer hydrogenation, 43. Jahrestreffen Deutscher Kata-

lytiker, 2010, Weimar; poster presentation.

•M. Schwarze, J. S. Milano-Brusco, H. Nowothnick, A. Rost, K. Seifert, S. Jost, J. Dim-

roth, R. Schomäcker, Katalyse in Tensidsystemen, 43. Jahrestreffen Deutscher Katalyti-

ker, 2010, Weimar; poster presentation.

•J. Dimroth, J. Keilitz, R. Haag, R. Schomäcker, An easily recyclable heterogenised Rh-

catalyst for the asymmetric transfer hydrogenation of prochiral ketones, ISHHC XIV,

2009, Stockholm; poster presentation.

Publications in Journals

•J. Dimroth, U. Schedler, J. Keilitz, R. Haag, R. Schomäcker, New Polymer-Supported

Catalysts for the Asymmetric Transfer Hydrogenation of Acetophenone in Water – Kinetic

and Mechanistic Investigations, Advanced Synthesis & Catalysis 2011,353, 8, 1335–

1344.

•J. Dimroth, J. Keilitz, U. Schedler, R. Schomäcker, R. Haag, Immobilization of a Mod-

ified Tethered Rhodium(III)-p-Toluenesulfonyl-1,2-diphenylethylenediamine Catalyst on

Soluble and Solid Polymeric Supports and Successful Application to Asymmetric Transfer

Hydrogenation of Ketones, Advanced Synthesis & Catalysis 2010,352, 14-15, 2497–

2506; paper featured in Synfacts 2011,1, 112.

Patent Applications

•J. Keilitz, R. Haag, U. Schedler, J. Dimroth, Immobilized Rhodium(III), Ruthenium(II),

or Iridium(III) Catalysts for Asymmetric Hydrogenation Reactions, WO/2011/026682.

•J. Dimroth, U. Schedler, Immobilisierte Rhodium(III)-Katalysatoren für asymmetrische

Hydrierreaktionen, DE 10 2007 041 965 A1.