CONVERSION OF LEVULINIC ACID TO GAMMA VALEROLACTONE USING p CYMENE RU(II) n HETEROCYCLIC CARBENE COMPLEXES

Catalytic hydrogenation of LA to GVL using molecular H2 as the hydrogen source Route 1 .... This thesis describes the synthesis and characterisation of three p-cymene RuII complexes bear

CONVERSION OF LEVULINIC ACID TO

GAMMA-VALEROLACTONE USING p-CYMENE RUTHENIUM(II)

N-HETEROCYCLIC CARBENE COMPLEXES

TAY BOON YING

DECLARATION

I hereby declare that this thesis is my original work and it has been written by me in its entirety, under the supervision of Associate Professor Huynh Han Vinh, Chemistry Department, National University of Singapore and Dr Ludger Paul Stubbs, Polymer Engineering and Catalysis, Institute of Chemical and Engineering Sciences, between 8 August 2011 and 16 September 2013

I have duly acknowledged all the sources of information which have been used in the thesis

This thesis has also not been submitted for any degree in any university previously

ACKNOWLEDGEMENT

I would like to take this opportunity to express my gratitude and heartfelt appreciation to the following people for their help, encouragement and support throughout this research journey:

Associate Professor Huynh Han Vinh (my supervisor at the National University of Singapore, NUS) for his insightful guidance, support, encouragement and friendship I would also like to thank him for his patience and understanding as well as the time spent

in the evenings after group meeting for discussions

Associate Professor Leong Weng Kee (Honours Year Supervisor) for encouraging me to continue with my graduate studies

Dr Ludger Paul Stubbs (my supervisor at the Institute of Chemical and Engineering Sciences, ICES) for giving me the autonomy in this project I would also like to thank him for his care, understanding, friendship and support

Dr Phua Pim Huat (my mentor at the ICES) for sharing with me his knowledge on nanoparticles, and for assistance with HPLC analysis

Dr Wang Cun (my mentor at the ICES) for sharing and guiding me in various aspects in synthesis and crystallographic analysis I would also like to thank him for his assistance

in solving the crystal structures obtained in this project and the discussions we had in the lab

Members of the Central Analytical Laboratory, ICES for their assistance and support:

Ms Seo Pei Nee for her guidance and assistance on the operation of TEM, Ms Chia Sze Chen for X-ray crystallographic analysis, Mr Ng Fu Song for elemental analysis, Mr Jeffrey Ng and Ms Doris Tan for the MS analysis, Ms Ong Lili for NMR analysis, Mr Lee Koon Yong and Mr Lee Ah Teck for the gas line support

My company, ICES, for the encouragement and financial support for this project

My examiners for taking their precious time in reviewing my thesis

Last but not least, I would like to express my heartfelt appreciation to my family, close friends and colleagues of the Polymer Engineering and Catalysis group, Vinh’s group members for their constant understanding, encouragement, patience, support and advice

TABLE OF CONTENTS

ACKNOWLEDGEMENT I

TABLE OF CONTENTS III

SUMMARY VI

COMPOUNDS NUMBERING SCHEME VIII

LISTS OF TABLES X

LIST OF FIGURES XII

LISTS OF SCHEMES XVI

LIST OF CHART XVIII

LIST OF ABBREVATIONS XIX

1 INTRODUCTION 1

1.1 Global energy demands and environmental issues 1

1.2 Biomass as an alternative energy source 2

1.3 Levulinic Acid and γ-valerolactone 3

1.4 Conversion of LA to GVL 5

1.4.1 Catalytic homogeneous hydrogenation using molecular H2 7

1.4.2 Catalytic heterogeneous hydrogenation using molecular H2 7

1.4.3 Using formic acid as the alternative hydrogen source 8

1.5 N-heterocyclic carbenes 11

1.5.1 Definition and properties 11

1.5.2 Preparation of NHC complexes and characterisation 14

1.6 Objective of the study 17

2 RESULTS AND DISCUSSIONS 19

2.1 Synthesis and characterization of p-cymene Ru(II) NHC complexes 19

2.1.1 p-cymene Ru(II) NHC complexes bearing monodentate NHC ligands 19

2.1.2 p-cymene Ru(II) NHC complexes bearing bidentate NHC ligands 22

2.2 Catalytic hydrogenation of LA to GVL using molecular H2 as the hydrogen source (Route 1) 29

2.2.1 Optimisation of reaction conditions 29

2.2.2 Catalyst screening 32

2.2.3 Kinetics 35

2.3 Catalytic hydrogenation of LA to GVL using FA as the hydrogen source (Route 2) 42

2.3.1 Optimisation of reaction conditions 42

2.3.2 Catalyst screening 45

2.4 Appearance of the reaction mixture for the two different routes 47

2.5 Methods to probe the catalytic nature 48

2.5.1 TEM Studies 48

2.5.2 Mercury poisoning test 53

2.5.3 UV Measurement 55

2.6 Effect of additive on Route 1 57

2.7 Catalyst recycling experiments 60

2.8 Effects of water and FA on nanoparticle formation 62

2.8.1 Hydrogenation of cyclohexanone to cyclohexanol 65

3 CONCLUDING REMARKS 67

4 FUTURE WORK 69

5 EXPERIMENTAL SECTION 72

5.1 Methods and materials 72

5.2 Analysis 72

5.3 Catalysis 73

5.3.1 General procedure of catalytic hydrogenation of LA to GVL using molecular H2 (Route 1) 73

5.3.2 General procedure of catalytic hydrogenation of LA to GVL using FA as the H2 source (Route 2) 74

5.4 Synthesis of Ligand Salt Precursors 74

5.5 Synthesis of p-cymene Ru NHC Complexes 76

6 REFERENCES 82

7 APPENDIX 88

SUMMARY

Fossil fuel has always been the main source of energy since the eighteenth century However due to its depletion and related environmental concerns, there has been a shift towards renewable sources of energy Biomass is a suitable alternative as it is the only renewable source of organic carbon that is essential for the production of liquid hydrocarbon fuels to chemicals One of the top twelve useful platform chemicals that can

be derived from biomass is levulinic acid (LA) that can be converted to

gamma-valerolactone (GVL) via hydrogenation

This thesis describes the synthesis and characterisation of three p-cymene Ru(II) complexes bearing monodentate N-heterocyclic carbene (NHC) ligand and three p-

cymene Ru(II) complexes bearing bidentate NHC-NHC/NHC-py ligand and their catalytic activity on the catalytic hydrogenation of LA to GVL [RuCl2(η 6 -p-cymene)( iPr2-bimy)] (2),

[RuCl2(η 6 -p-cymene)(Bn2-bimy)] (3) and [RuCl(η 6 -p-cymene)(κ 2 C,C-diNHCme)][PF6] (4)

were characterised by X-ray diffraction analyses Two hydrogen sources, molecular hydrogen (Route 1) and formic acid (FA, Route 2) were used in the catalytic hydrogenation of LA to GVL

In Route 1, we found that the catalysis proceeded in a heterogeneous fashion where the ruthenium complexes act as a precursor to catalytically active ruthenium nanoparticles (RuNPs) rather than a soluble metal catalyst Under reducing H2 atmosphere, the catalyst precursor will rapidly be converted to RuNPs At the end of the reaction, metallic plating was found on the reaction liner and magnetic stir bar RuNPs were seen during transmission electron microscopy (TEM) analysis Mercury poisoning experiments also showed positive results where there was a significant drop in the GVL yield The

disappearance of absorption maxima with time in the UV spectroscopic study indicated that the catalyst precursor was being converted to RuNPs

In the solvent selection study, we discovered that water is the best solvent for our optimised reaction conditions Water is also the best solvent in the catalytic

hydrogenation of cyclohexanone to cyclohexanol using our p-cymene Ru(II) NHC

complexes as it aids in RuNPs dispersion In the effect of additive study on Route 1, we discovered that pH has an effect on the stability of the RuNPs RuNPs tend to aggregate

at low pH p-cymene Ru(II) complexes bearing monodentate NHC ligands generally

performed better as compared to the complexes bearing bidentate ligands and catalysis

via Route 1 with p-cymene Ru(II) complexes bearing monodentate NHC ligands

proceeded via zero-order kinetics with respect to LA and first-order kinetics with respect

to H2 The best performing catalyst was 2 where almost quantitative yield was achieved

in 120 min

In Route 2, the catalysis was found to proceed in a homogeneous fashion as FA present might aid in the formation of a formate-bridged ruthenium dimers that will first catalyse the decomposition of FA to H2 and CO2 followed by the formation of GVL RuNPs were absent for most of the catalysts screened and mercury poisoning experiments showed

no decrease in GVL yield The absorption spectrum of for the Ru(II) species did not change throughout the course of the catalysis With respect to the catalytic performance,

the trend was similar to the catalyst performance in Route 1 p-cymene Ru(II) complexes

bearing monodentate NHC ligands generally performed better as compared to the

complexes bearing bidentate ligands The best performing catalyst was 2 where 90.7%

COMPOUNDS NUMBERING SCHEME

Ligand Precursor Salts

1,1'-Dimethyl-3,3'-methylene-Salt E

1,1'-Dimethyl-3,3'-ethylene-diimidazolium dibromide (diimyet·2HBr)

Salt F

3-Methyl-1-(2-picolyl)imidazolium

chloride (py-imy·HCl)

p-cymene Ruthenium(II) NHC complexes

LISTS OF TABLES

Table 1 Summary of some physical properties of LA.[11] 4

Table 2 Summary of some physical properties of GVL.[9] 5

Table 3 Summary of some thermodynamic properties of FA decomposition pathways.[18c] 9

Table 4 Selected interatomic distances (Å) and angles () 22

Table 5 Ccarbene-Ru and Ru-Cl distances [Å] for 4, 5 and 6 28

Table 6 Catalyst loading optimisation using the reported optimised condition.[a] 30

Table 7 Solvent screening using 0.1 mol% 1 30

Table 8 Pressure screening using 0.1 mol% 1 and water as solvent 31

Table 9 Pressure screening using 0.1 mol% 1, water as solvent and 12 bar H2 31

Table 10 Summary of the outcome for catalyst screening[a] 33

Table 11 Summary of the rate constant and linearity of [RuCl2(p-cymene)]2 and complexes bearing monodentate NHC ligands 1 to 3 35

Table 12 Summary of the rate constant and linearity of [RuCl2(p-cymene)]2 and complexes bearing bidentate NHC ligands 4 to 6 38

Table 13 Summary of the rate of reaction under 1, 6 and 12 bars using 0.1 mol% 1 as

the catalyst at 130 C under 1, 6 and 12 bar H2 for 160 min 40

Table 14 Temperature screening with/without water as solvent 43

Table 15 Temperature screening without water as solvent 44

Table 16 Summary of the outcome for catalyst screening.[a] 46

Table 17 Summary of mercury poisoning test for the conversion of LA to GVL via Route 1 and Route 2 54

Table 18 Summary of mercury poisoning test for the conversion of LA to GVL via Route 2.[a] 54

Table 19 Summary of the results obtained from the study of the effect of additives on Route 1 57

Table 20 Summary of the pH of reaction mixtures taken before hydrogenation via Route 1[a] 59

Table 21 Summary of the results obtained from various variations in reaction condition. 63

Table 22 Summary of the results obtained from hydrogenation of cyclohexanone[a] using water and THF 66

LIST OF FIGURES

Figure 1 World energy consumption from 1990 to 2035.[1] 1

Figure 2 Global CO2 emission from fossil fuels from 1900-2009.[3] 2

Figure 3 Organic components of lignocelluloic biomass and examples of useful chemical/intermediates that can be obtained from acid catalysed hydrolysis 3

Figure 4 Derivatives of levulinic acid.[12] 4

Figure 5 FA produced from acidic hydrolysis of lignocelluosic biomass can be directly used as the hydrogen source with a suitable catalyst to be decomposed to H2 and CO2 followed by the conversion of LA to GVL 8

Figure 6 Examples of Fischer carbene, Schrock carbene and NHC 12

Figure 7 Grubbs second generation catalyst 12

Figure 8 Common azole rings used to tune the electronics of NHC 14

Figure 9 p-cymene Ru(II) NHC complexes with monodentate NHC ligands (1 to 3) and bidentate NHC ligands (4 to 6) used in the study 18

Figure 10 Molecular structures of 2 (a) and 3 (b) with thermal ellipsoids drawn at 30% probability level Hydrogen atoms are omitted for clarity 21

Figure 11 Molecular structure of 4 with thermal ellipsoids drawn at 30% probability

level Hydrogen atoms and hexafluorophosphate anion are omitted for clarity Selected

2.413(1), Ru(1)-Ct1* 1.739, Ru(1)-C(6) 83.3(2), N(2)-C(1)-N(3) 109.9(4), C(4)-Ru(1)-Cl(1) 86.3(1), C(6)-C(4)-Ru(1)-Cl(1) 86.3(1) *Ct1 denotes centroid formed by

C(10)-C(15) 28

Figure 12 Reaction profile of catalytic hydrogenation of LA to GVL using 1 from 0 to 160 min 32

Figure 13 Plot of LA depletion vs time for [RuCl2(p-cymene)]2 36

Figure 14 Plot of LA depletion vs time for 1 36

Figure 15 Plot LA depletion vs time for 2 37

Figure 16 Plot of LA depletion vs time for 3 37

Figure 17 Plot of LA depletion vs time for 4 38

Figure 18 Plot of LA depletion vs time for 5 39

Figure 19 Plot of LA depletion vs time for 6 39

Figure 20 Plot of reaction rate vs hydrogen pressure 41

Figure 21 Effect of base loading on GVL yield (yield reflected is an average of two runs) 44

Figure 22 Plot of GVL formation and LA depletion with time 45

Figure 23 Appearance of reaction mixture after reaction A Using molecular H (Route

bar Reaction condition: 4.31 mmol LA, 0.1 mol% 1, 10 mL water, 12 bar H2, 130 C, 160

min B Using FA as the H2 source (Route 2), yellow solution was obtained with clean liner wall and magnetic stir bar Reaction condition: 80 mmol LA, 80 mmol FA, 0.075

mol% 1, 3 mol% NaOH, 130 C, 12 h 47

Figure 24 TEM images of [RuCl2(p-cymene)]2 using molecular H2 at 30 min (left), 90 min (middle) and 160 min (right) 49

Figure 25 TEM images of 1 using molecular H2 at 30 min (left), 90 min (middle) and 160 min (right) 50

Figure 26 TEM images of 2 using molecular H2 at 30 min (left), 90 min (middle) and 160 min (right) 50

Figure 27 TEM images of 3 using molecular H2 at 30 min (left), 90 min (middle) and 160 min (right) 50

Figure 28 TEM images of 4 using molecular H2 at 30 min (left), 90 min (middle) and 160 min (right) 51

Figure 29 TEM images of 5 using molecular H2 at 30 min (left), 90 min (middle) and 160 min (right) 51

Figure 30 TEM images of 6 using molecular H2 at 30 min (left), 90 min (middle) and 160 min (right) 51

Figure 31 TEM images of catalytic hydrogenation of LA to GVL using molecular H2 as

the hydrogen source A Ru/C 4 B [RuCl2(PPh3)] 52

Figure 32 TEM images of catalytic hydrogenation of LA to GVL using FA as the

hydrogen source A Ru/C B Complex 4 C Complex 6 53

Figure 33 UV spectra of 1 at 0.0004M for catalytic hydrogenation of LA to GVL using A Route 1 from 0 – 160 min B Route 2 from 0 – 24 h 55

Figure 34 Colour changes of 0.004M of reaction mixtures throughout the course of reaction via A Route 1; B Route 2 56

Figure 35 TEM images of 1 when different additives were added 58

Figure 36 TEM image of 1 after second catalytic run 61

Figure 37 TEM images of RuNPs formed in A Scenerio 3 and B Scenerio 4 63

Figure 38 Structure of formate-bridged ruthenium dimers.[47] 64

Figure 39 TEM images obtained for 1 in the conversion of LA to GVL using different solvents 64

LISTS OF SCHEMES

Scheme 1 Reaction pathways of LA to GVL.[17n] 6

Scheme 2 Conversion of LA to GVL using molecular H2 or FA as the hydrogen source 6 Scheme 3 Synthesis scheme of the first stable NHC 13

Scheme 4 Electronic configuration and resonance structures of NHC.[29] 13

Scheme 5 Common synthetic routes towards NHC complexes 16

Scheme 6 Synthesis of the p-cymene Ru(II) NHC complexes 1 to 3 19

Scheme 7 Synthesis of 4 23

Scheme 8 Synthesis of 5 25

Scheme 9 Literature reported procedure for the synthesis of salt F.[34] 25

Scheme 10 Modified procedure for the synthesis of salt F 26

Scheme 11 Synthesis of 6 26

Scheme 12 Reaction scheme of catalytic hydrogenation of LA to GVL via Route 1 with optimised conditions by Yan et al [17m] 29

Scheme 13 Reaction scheme of catalytic hydrogenation of LA to GVL via Route 2 with optimised conditions by Deng et al [11q] 42

Scheme 14 Hydrogenation of cyclohexanone to cyclohexanol 65

Scheme 15 Proposed ways to convert biomass-derived LA to GVL using our optimised

routes 69

Scheme 16 Hydrogenation of acetanilide to N-cyclohexylacetamide using 1 70

LIST OF CHART

Chart 1 Methods of synthesizing MNPs and methods of stabilizing MNPs 11

LIST OF ABBREVATIONS

AcOH Acetic Acid

API Atmospheric Pressure Ionisation

et al and others (Latin et alii)

NMR Nuclear Magnetic Resonance

NREL National Renewable Energy Laboratory

1 INTRODUCTION

1.1 Global energy demands and environmental issues

Since the eighteenth century, fossil fuel has been the main energy source for human and economical developments With the increasing human population and technological advances, the energy needed to meet the economic and social demand increased (Figure 1) in order to sustain human well-being and raise the standard of living

Figure 1 World energy consumption from 1990 to 2035.[1]

Fossil fuels have been gradually declining due to increasing usage Environmental problems like high carbon dioxide (CO2) emission (Figure 2) led to increasing drastic climatic changes CO2 emission has led to global warming and increase in average global temperature of 0.8 C since 1880.[2] Hence, there is a pressing need in search of alternative energy source that is sustainable and environmentally benign

History 2013 Projection

Figure 2 Global CO2 emission from fossil fuels from 1900-2009.[3]

1.2 Biomass as an alternative energy source

Biomass is biological material that is derived from living, or recently living organisms In the context of biomass for energy this is often used to mean plant based material, but biomass can equally apply to both animal and vegetable derived material.[4] Biomass is a suitable alternative as it is the only renewable source of organic carbon, which is essential for the production of liquid hydrocarbon fuels to chemicals.[5] In the Kyoto protocol, together with the vision to reduce crude oil dependence, researchers’ attention has been directed to use biomass as a source of energy and, more specifically, for transportation fuels.[5-8]

2009

Figure 3 Organic components of lignocelluloic biomass and examples of useful

chemical/intermediates that can be obtained from acid catalysed hydrolysis

1.3 Levulinic Acid and γ-valerolactone

The National Renewable Energy Laboratory (NREL) has identified levulinic acid (LA) as

one of the top 12 building block chemicals that can be produced from sugars via

biological transformation or chemical hydrolysis of plant biomass.[10] LA, also known as 4-oxopentanoic acid, 4-oxovaleric acid and 3-acetylpropionic acid, is a C5 ketoacid and

is a product formed from the hydrolysis of cellulosic biomass with acid Formic acid (FA) and water are formed as the by-products of hydrolysis Table 1 summarises some physical properties of LA

Table 1 Summary of some physical properties of LA.[11]

Of the wide range of derivatives that can be obtained from LA, γ-valerolactone (GVL) has been identified by Horváth as a sustainable liquid for energy and carbon-based chemicals.[4] GVL is a natural occurring compound that is present in fruits and is also used in the flavour and fragrance industry It is a colourless liquid with a low melting point of -31 C, high boiling point of 207 C and has an acceptable odour as well as low toxicity.[11-15] Table 2 shows a summary of some physical properties of GVL

Table 2 Summary of some physical properties of GVL.[9]

Scheme 1 Reaction pathways of LA to GVL.[17n]

Scheme 2 Conversion of LA to GVL using molecular H2 or FA as the hydrogen source

1.4.1 Catalytic homogeneous hydrogenation using molecular H2

Homogeneous systems for the conversion of LA to GVL like [RuCl2(PPh3)3][16a], [Ru(acac)3]/phosphine system[16b, d, g], RuCl3·3H2O/PPh3 with various bases[16c] and Ir pincer complexes[16f] have been reported Geilen et al developed a homogeneous

system based on [Ru(acac)3]/phosphine system with the use of different additives to selectively hydrogenate LA to a range of promising derivatives, including GVL.[16d]

Despite its highly selective nature, homogeneous systems face separation problems especially for GVL targeted production as GVL is a high boiling liquid (208 C), which makes separation by means of distillation uneconomical.[13]

Schuette and Thomas employed a platinum oxide catalyst in an organic solvent to hydrogenate LA at 3 bar H2 in 44 h to yield 87% GVL.[17a] Christian et al improved the

GVL yield to 94% using Raney nickel and copper-chromium catalysts.[17c] The disadvantage of using Raney nickel and copper-chromium catalysts is the need for high temperature and H2 pressure

Since then, numerous other catalyst systems have been studied Broadbent et al

reported the use of rhenium catalysts (Re black, Re(IV) oxide hydrate) for the hydrogenation of LA to GVL.[17d] Yan et al reported the hydrogenation of LA to GVL

using Ru/C and achieved a 92% conversion of LA with a 99% GVL selectivity at 130 C

and 12 bar H2 pressure in methanol (MeOH).[17m]

Among the carbon supported noble metal catalysts (Ru, Pt, Pd)[17s], 5 wt% Ru/C showed the highest catalytic activity and product selectivity The better performance of Ru/C is attributed to its high dispersion of metallic Ru on carbon in nano-sizes compared to the

Pt and Pd catalysts

1.4.3 Using formic acid as the alternative hydrogen source

Formic acid (FA), containing 4.4 wt% hydrogen, has been identified as the most promising material for hydrogen storage.[18] It is commonly adopted as an effective hydrogen source to produce GVL as it is readily available and of low cost Moreover, from the hydrolysis of biomass (Figure 5), FA is produced By using FA as the hydrogen source, the separation of the mixture of LA and FA can be eliminated; hence reducing the high separation cost and also improves the atom economy of the process

Figure 5 FA produced from acidic hydrolysis of lignocelluosic biomass can be directly used as

the hydrogen source with a suitable catalyst to be decomposed to H 2 and CO 2 followed by the conversion of LA to GVL

The decomposition of FA can occur via two different pathways, namely (A)

dehydrogenation/decarboxylation and (B) dehydration/decarbonylation Their thermodynamic properties are summarised in Table 3

Table 3 Summary of some thermodynamic properties of FA decomposition pathways.[18c]

decarboxylation HCOOH  CO2 + H2 -32.9 31.2 215 B) Dehydration/

decarbonylation HCOOH  CO + H2O -12.4 28.7 138

Pathway A is generally catalysed by transition metal catalysts[18a-g] while pathway B can

be catalysed by acid catalyst.[18i] Pathway A is favourable as H2 is produced FA decomposition generating H2 is desirable as it serves as a hydrogen source for portable application without the need to use a pressurised H2 source Pathway B is undesirable

as the CO produced may poison the catalysts

Hováth et al used [(η6-C6Me6)Ru(bpy)(H2O)][SO4] in water for the transfer hydrogenation of LA with FA as the hydrogen donor and obtained 25% GVL and 75% 1,4-pentandiol.[16b] Deng et al reported using RuCl3/PPh3/pyridine catalyst system to convert a 1:1 aqueous mixture of LA and FA selectively into GVL with high yield of about 80-90%.[16c]

1.4.4 Water as the solvating media

In a recent review on the methods of accessing environmental impact of chemical processes, Sheldon stated that “The best solvent is no solvent and if a solvent is needed then water is preferred”.[19] Water is an ideal solvent as it is non-toxic, environmentally benign, non-flammable and readily available

Recent studies have shown that water can be advantageous as solvent for the catalytic

hydrogenation of LA to GVL Al-Shaal et al hydrogenated LA to GVL in water with 5

wt% Ru/C and achieved a 99.5% LA conversion and 86.6% GVL selectivity.[17n]

Delhomme et al used [Ru(acac) 3] with various water-soluble phosphine and was able to reduce LA to GVL at 140 Cand 5 MPa H2, with a LA conversion of up to 99% and GVL selectivity of as high as 97% within 5 h.[17v] Most importantly, water can replace alcoholic solvents (such as methanol) which might form levulinate esters

1.4.5 Metal nanoparticles as catalysts

Using metal nanoparticles (MNPs) as catalysts is a growing area in homogeneous and heterogeneous catalysis as nanoparticle catalysts are efficient, selective and can be recycled, hence meeting the requirements of green catalysis.[20] Chart 1 is a simplified chart illustrating the synthesis of MNPs by chemical reduction/decomposition MNPs tend to aggregate but can be stabilized by using polymers, alcohols, ionic liquids and ligands.[21]

Chart 1 Methods of synthesizing MNPs and methods of stabilizing MNPs

Lately, the use of [Ru3CO12] as the precursor for the preparation of ruthenium nanoparticles (RuNPs) for the conversion of LA to GVL by molecular H2 or FA as the H2source has been reported.[17w] This is an example of nanoparticle formation by reductive

decomposition of an organometallic cluster Ligands like N-heterocyclic carbenes

(NHCs) have been reported to stabilise or modify MNPs.[22]

1.5 N-heterocyclic carbenes

1.5.1 Definition and properties

Carbenes are defined as neutral carbon species containing a divalent carbon atom with six valence electrons There are two different types of carbene, namely Fischer[23] and Schrock carbenes[24] Wanzlick-Arduengo carbenes,[25] which are also known as NHC, are a type of Fischer carbene (Figure 6) Fischer carbene complexes have an electrophilic carbene carbon atom while Schrock carbene complexes have a nucleophilic carbene carbon Because of the electrophilic nature of Fischer carbene carbon atom, it

Stabilisation methods

Alcohol Ionic Liquids Polymers

Ligands

ablation

metal series On another hand, the nucleophilic carbene carbon atom of the Schrock carbene will tend to prefer metal center that is electron deficient like those in the early transition metal series

Figure 6 Examples of Fischer carbene, Schrock carbene and NHC

NHC are a class of ligands with excellent steric and electronic versatility, where the carbene carbon is incorporated into a heterocyclic ring Since the first report of an NHC-Mercury complex in 1968[26], the use of NHCs as ligands in transition metal chemistry

remained relatively dormant until Arduengo et al isolated the first stable crystalline

carbene (Scheme 3) in 1991.[27] Subsequently, many NHC-metal complexes have been prepared and used in catalysis One such example is the Grubbs second generation catalyst (Figure 7), which is highly active in olefin metathesis reactions[28]

Scheme 3 Synthesis scheme of the first stable NHC

The stability of singlet vs triplet NHC can be explained by a combination of mesomeric (M) and inductive (I) effects, also collectively known as “push-pull” effect The +M effect

“pushes” the lone pair of electrons from the neighbouring nitrogen atoms into the empty

pπ-orbital, hence increasing the electron density of the carbene center On the other hand, the –I effect of the σ-electron withdrawing N atoms “pulls” electrons from the

carbene center, hence stabilizing the σ-orbital The “push-pull” effect increases the energy gap between the σ and pπ orbitals, thus stabilizing the singlet carbene (Scheme 4)

Scheme 4 Electronic configuration and resonance structures of NHC.[29]

The attractive features of using NHC as ligands for transition metal catalysis are the ability of tuning their electronic and steric properties The electronic properties of an NHC can be tuned by changing the azole ring The three common azole rings commonly used include benzimidazole, imidazole and imidazoline are shown in Figure 8 Their electron donating power increases in the order of benzimidazole<imidazole<imidazoline Benzimidazole is the least electron-donating in the series due to benzannulation Also

the sp 2 carbon of imidazole is more electron drawing as compared to the sp 3 carbon of imidazoline

Figure 8 Common azole rings used to tune the electronics of NHC

1.5.2 Preparation of NHC complexes and characterisation

Since the isolation of the first stable crystalline carbene, various NHC and their

pincer-type NHCs Three common approaches to prepare NHC complexes are highlighted (see Scheme 5) They are:

A) Reaction of transition metal complex with a free carbene

This is regarded as the most straightforward method for the preparation of NHC

complexes The free carbene can be pre-formed or generated in-situ A base is used

to deprotonate the azolium salts prior to the addition of the metal precursor Common bases used are NaH, KOtBu, NaOEt and KN(SiMe3)2

B) In-situ deprotonation of azolium salt with a metal precursor containing a basic ligand

The azolium salt is deprotonated in-situ with a basic ligand of the metal precursor

Commercially available or easily synthesizable complexes with acetate, alkoxide, hydride or acetylacetonate ligands are frequently used Wanzlick[26] and Öfele[30]

have employed this method to prepare the first imidazolylidene complexes starting

from Hg(OAc)2 and [CrH(CO)5]-

C) The silver-carbene transfer method

This method was first reported by Lin et al [31] It is a widely used method that makes use of silver(I) oxide to react with the azolium salt to form the silver carbene complex As the carbene coordinates weakly to silver, silver metal can be easily replaced by other metals such as ruthenium, gold and palladium Another advantage

of this method is that the synthesis can be carried out in the presence of moisture and oxygen The complexes discussed in this work were prepared using this method

Scheme 5 Common synthetic routes towards NHC complexes

The most convenient way to characterise NHCs is to use Nuclear Magnetic Resonance (NMR) spectroscopy As mentioned earlier, free carbenes are generated from the deprotonation of the azolium salt The best diagnostic tool would be to use 1H NMR spectroscopy to monitor the disappearance of the acidic proton resonance (typically at 8-

12 ppm) Carbene formation can be monitored by 13C{1H} NMR spectroscopy Free carbene resonates between 200-250 ppm while its corresponding salt resonates between 130-160 ppm Once the free carbene is coordinated onto a metal centre, the corresponding signal is usually shifted upfield Other characterisation methods include mass spectrometry (MS), elemental analysis (EA) as well as x-ray crystallography

1.6 Objective of the study

In this study, we are interested in comparing the differences in the activity between the

p-cymene Ru(II) NHC complexes bearing monodentate and bidentate NHC ligands on

the conversion of LA to GVL The two reaction systems (Scheme 1) under study are as follows:

Route 1  catalytic hydrogenation of LA to GVL using molecular H2 as the hydrogen source

Route 2 catalytic hydrogenation of LA to GVL using FA as the hydrogen source

Three p-cymene Ru(II) NHC complexes bearing monodentate NHC ligands (1 to 3, Figure 9) and three p-cymene Ru(II) NHC complexes bearing bidentate ligands (4 to 6,

Figure 9) were prepared using the silver carbene transfer method Optimisation studies

of Route 1 and Route 2 were carried out and under optimised conditions, we screened

their catalytic activities The kinetic studies of the six p-cymene Ru(II) NHC complexes

and [RuCl2(p-cymene)]2 for Route 1 were also carried out

Further tests (mercury poisoning experiment, Transmission Electron Microscopy (TEM) studies and UV-Vis spectroscopy) were carried out to probe the nature of catalysis and the possible catalytic pathways depending on the use of H2 source will also be discussed

In addition, the effect of additives on Route 1 was studied Finally, the effects of water and FA on the formation of nanoparticles were investigated and an investigation of the effect of water on a different substrate, cyclohexanone, was carried out