Heterogeneous catalysis in protodecarboxylation and c c bond formation 2

!
!
1!
Chapter 1: Introduction

The aim of this thesis is to develop and study heterogeneous Pd, Cu, and Ag
catalysts for C-C bond transformation reactions. The Ullmann reaction and
decarboxylative cross-coupling are preferred methodologies for C-C bond
formation between aromatic rings. Both do not require the use of preformed
organometallic reagents and can be used to prepare biphenyl compounds. The
biphenyl moiety is a very important structural component of many industrial,
agrochemical and pharmaceutical products. Thus, the development of more
efficient catalytic systems for access to this group of molecules has very high
economical and industrial value. We also investigated the
protodecarboxylation of aromatic benzoic acids over heterogeneous supported
catalysts. The study of the expulsion of the carboxylate group is important
because it is also the rate-determining step in the decarboxylative cross-
coupling of aromatic benzoic acids with aryl halides.
The catalyst is the key to the efficient synthesis of these valuable
biphenyl compounds. Thus, in this work, various parameters that influence the
catalytic properties such as metal loading and pretreatment conditions were
studied to understand and correlate the catalyst design to its function. The
reaction mechanism was also studied to better understand the catalytic process
and to optimise the reaction conditions.

1.1 Supported nanosized transition metal catalysts in fine chemical
synthesis
Catalysis is a fundamental area of research as it is a highly important
technology used to support industrial processes. Several well defined areas of
!
!

2!
industrial catalysis include petroleum, pharmaceutical and environmental
catalysis. The petroleum industries rely heavily on catalysts to manufacture
petrochemicals from crude oil [1], while pharmaceutical companies have
started to appreciate the benefits of catalysis in the manufacture of chemical
compounds with medicinal functions [2]. Environmental catalysts are used to
remove and degrade toxic waste products from manufacturing effluents [3]
and most notably in the car catalytic converter. It has been estimated that up to
90 % of all industrial chemical processes were performed with the help of
catalysts [4]. BASF, a market leader in catalysis, reported total sales of
catalyst products worth € 6.4 million in 2011 [5]. The importance of industrial
catalysis is also reflected by the fact that the annual catalyst demand has been
projected to increase by 8 % annually from 2011-2015.
Catalysts can speed up the rate of chemical transformations without
being consumed in the process. They lower the activation energy by providing
an alternative reaction pathway which is energetically and sterically
favourable. Catalysis also lowers by-product formation and maintains high
productivity. All these factors reduce the costs and time required, and make
the processes more environmentally friendly. Hence, catalysis is an
indispensable tool in the chemical industries as it ensures optimum
efficiencies and high process yields.

1.1.1 Homogeneous catalysts versus heterogeneous catalysts
Catalysts are generally divided into two classes: homogeneous catalysts and
heterogeneous catalysts. In homogeneous catalysis, the catalysts used are
chemical compounds such as general acid, base or organometallic complexes
!
!
3!
which dissolve fully in the reaction medium to form a single phase. Thus, they

have a higher degree of dispersion. Due to their molecular nature, each
homogeneous catalyst contains only a single type of active site, and high
selectivity to a particular product can be achieved. However, homogeneous
catalysts such as organometallic complexes have poor thermal stability which
limits their working temperature to below 200
o
C; this is unlike heterogeneous
catalysts, which can tolerate higher reaction temperatures. Heterogeneous
catalysts are usually solids, whereas the reactants are liquids or gases.
Catalytic reactions using such catalysts take place between two or more
different phases. Many catalysts are based on precious metals such as
palladium, platinum, ruthenium and silver. However, in a heterogeneous
reaction at a fluid-solid interface, only atoms on the surface of the solid can
participate in the reaction. High surface area is therefore required for efficient
catalysis. Due to the high price of the precious metals, they are often dispersed
as nanoscale metal particles onto support materials to maximise and stabilise
the exposed catalytically active surface. Some molecular species such as
hydrogen and oxygen adsorb dissociatively onto these nanoparticles by
interaction with the d-electrons of the precious metals, allowing hydrogenation
and oxygenation reactions to occur under milder conditions [6]. The higher
activity afforded by the precious metal catalysts compensates for their high
price by making it possible to use a very small amount of the catalysts. Also,
precious metal catalysts have higher thermal stability and are more resistant to
oxidation, which can potentially prolong the catalyst life-time. However, there
are also many catalysts which are made up of non-precious metals such as
copper and iron, or metal oxides with Lewis or Brönsted acidic and basic sites
!
!
4!
such as zeolites, silica and alumina. These catalysts are used effectively in a

wide variety of processes, for example metallic Fe in the Haber-Bosch process
[7], Cu/ZnO/Al
2
O
3
in methanol synthesis [8], and ZSM-5 zeolites in catalytic
steam cracking of hydrocarbons [9].
The atoms at the surface of a solid catalyst have an incomplete
coordination shell. There are more highly coordinated atoms within planes,
and coordinatively unsaturated atoms in edges and at corners. The amount and
type of each active site varies according to the chemical composition,
dimension and the form of the solid catalyst particle, as seen in the example of
a gold nanoparticle in Fig. 1.1.


Figure 1.1: A plot of the calculated fraction of Au atoms at the corner (red),
edge (blue), and crystal face (green) of a truncated octahedral gold
nanoparticle. The insert shows the top half of a truncated octahedral gold
nanoparticle and the position of the corner, edge and surface atoms [10], [11].

The various surface atoms have different catalytic activities and may catalyze
different reactions. For example, studies conducted by Medlin et. al. using
self-assembled monolayers (SAM) on supported Pd/Al
2
O
3
catalysts have
shown that furfurals react differently on various surface sites present on the
!
!

5!
Pt(111) surface [12]. It was postulated that furfural decarbonylation followed
by ring hydrogenation occurs preferentially on threefold terrace sites as it
allows the adsorption of the furan ring in a flat-lying manner. In contrast,
aldehyde hydrogenation to give furfural alcohol and hydrodeoxygenation to
give methylfuran occurs primarily at the step edge and defect sites as the
furfural binds upright at these sites. The results clearly show that several
products can be obtained using unmodified supported Pt catalysts; with furan
as the primary product. The rate of of decarbonylation was about two orders of
magnitude higher than that of the aldehyde hydrogenation and
hydrodeoxygenation.

Figure 1.2. Furfural hydrogenation pathways on Pt(111) surface [12]

This example show that due to the presence of multiple active sites present on
heterogeneous catalysts, the selectivity towards the desired products is
normally poorer than with homogeneous catalysts. Also, the degree of
dispersion, i.e., the ratio of atoms in the surface to the total number of atoms in
the particle for such solid catalysts is generally lower than one, compared to
homogeneous catalysts where all atoms are active.
The mechanism of homogeneously catalysed reactions is usually well
understood because the catalytically active atoms or molecules are well-
defined and can be studied through the use of spectroscopic methods [13]. For
!
!
6!
heterogeneous catalysts, the exact reaction mechanism is harder to elucidate.
Nevertheless, the chemical interactions between the components in the
reaction mixture and the active sites should be similar in both homogeneous
and heterogeneous catalysts. This is certainly true if the active sites of both

types of catalysts are of similar chemical nature. The active site of the catalyst
is made up of a specific type of atom or fragment of a molecule or complex.
The active site can be inherent to the chemical compound or it can be formed
upon pre-coordination or pre-chemisorption of reaction components.
In homogeneous organometallic catalysts, the transition metal ions
coordinated to ligands usually make up the active site of the catalyst. The
ligands are neutral molecules with a lone pair of electrons or charged ions. In
both cases, they can donate an electron pair into the empty d-orbital of the
transition metal ion to form a carbon-metal (C-M) bond. The presence of the
ligand changes the size, charge, coordination number and electron
configuration of the metal ion, which in turn affects its catalytic activity. The
steric bulk of the ligand can control the type and the binding of reactants to the
organometallic complex. During reaction, the reactants coordinate to the
central transition metal ion of the organometallic complex to form an activated
intermediate complex which accelerates the rate of chemical transformation.
For example, changes in the electronic configuration of the reactants such as
alkenes or carbon monoxide coordinated to positively charged metal centres
activate the carbon atoms towards nucleophilic attack. Insertion of hydride or
alkyl ligands is also favoured upon the coordination of unsaturated reactants to
the transition metal centre of the organometallic complex. β-Hydrogen
elimination, which is the reverse of insertion, can take place more easily
!
!
7!
through the formation of the intermediate complex. Oxidative addition of
reactants to the organometallic complex results in the dissociation of the bond
within the reactant and brings it into close contact with other reactants for the
reaction to take place. These examples are shown in Figure 1.3.
M
CO

CO
OC
CH
3
OC
CO
M
OC
CO
OC
OC
C
CH
3
O
M
C
CO
OC
COOC
CO
O
CH
3
+ CO
(b) Migratory insertion of alkyl group
(c) Migratory insertion of hydride/ beta-hydride elimination (reverse)
M
H
CH

2
CH
2
M
H CH
2
CH
2
M
C
H
2
CH
3
M
CH
2
CH
2
(a) Nucleophilic attack on coordinated ligands
2+
OH
2
M
C
H
2
H
2
C

OH
H
+
+
+
ML L
L
L
+
AX
ML L
L
L
A
X
(d) Oxidative addition

Figure 1.3: Examples of reactions of organometallic complexes

In heterogeneous catalysts, the active site is made up of a specific type
of surface atom within the larger metal particle. In analogy to the
homogeneous catalyst complex, the nearest neighbour atoms can be regarded
as permanent ligands of the central active surface atom. Solvent or reactant
molecules can also function as ligands when they adsorb next to the active site
and modify its electronic properties. The support of the catalyst can also affect
the adsorption of the reactant molecules, and the direction of electron flow at
the metal-support interface can thereby influence the catalytic properties of the
!
!
8!

metal atoms. Heterogeneously catalysed reactions begin with the adsorption of
the reactant molecules onto the active sites of the metal catalyst surface,
forming activated intermediates similar to organometallic complexes.
Adsorption induces bond dissociation, such as the dissociative adsorption of
H
2
in hydrogenation reactions, or it can result in the weakening of molecular
bonds which favours substitution reactions. In this way, the activation energy
of the bond transformation is lowered which accelerates the rate of reaction.
After the bond transformation, the product desorbs from the catalyst surface to
allow the next reactant molecule to adsorb onto the active site. The strength of
the metal-adsorbate bond however cannot be too strong, otherwise it will not
be possible for the transformed species to desorb from the catalyst surface.
The active site will be blocked, and catalytic activity will come to an end. One
example of a reaction catalysed by heterogeneous catalyst is the iron-catalysed
Haber-Bosch process for the synthesis of ammonia. The process, as shown in
Figure 1.4, begins with the dissociative adsorption of nitrogen and hydrogen
on the surface of the iron catalyst. The formation of Fe-N and Fe-H is
energetically favourable which facilitates the fission of the N≡N and H-H
bond. The adsorbed N and H atoms combine to form NH
3
which readily
desorbs from the catalyst surface.

Figure 1.4: Formation of ammonia on a heterogeneous catalyst surface

!
!
9!
Although the chemical interactions between the reactants and

homogeneous or heterogeneous catalysts are similar, heterogeneous catalysts
are preferred industrially. The reason is that homogeneous catalysts are
difficult to separate from the products. This leads to serious contamination of
the products and limits the reuse of the catalysts [14]. The loss of expensive
metals in the catalysts is also a serious economical drawback. The problem of
residual metal catalysts in pharmaceutical products is particularly crucial as
guidelines have been set by government agencies to limit the amount of metal
residues present in these products [15]. Extra purification processes using
metal scavengers have to be incorporated to reduce the metal content of these
products. This increases the cost and slows down the production process [16].
In contrast, heterogeneous catalysts can be easily retained in fixed bed reactors
or separated from the products in the reaction medium through simple
filtration, centrifugation or magnetic separation.

1.1.2 Heterogeneous catalysts: Supported metal catalysts
One important class of heterogeneous catalysts are the supported metal
catalysts. Such catalysts comprise of one or more catalytically active metal
species which are highly dispersed onto a support material to form metal
nanoparticles. The support materials have high surface area (>100 m
2
/g) and
are usually porous, and these factors help to increase the dispersion and the
surface area-to-volume ratio of the catalytically active species. The support
material also functions to stabilise the small metal nanoparticles. Various
materials such as alumina (Al
2
O
3
), hydrotalcite (Mg
6

Al
2
CO
3
(OH)
16
·4(H
2
O)),
hydroxyapatite (Ca
5
(PO
4
)
3
(OH)), silica (SiO
2
), zeolites (aluminosilicate
!
!
10!
minerals), cyclodextrins (cyclic oligosaccharides), polyamides and polyamines
are used as the solid support. The nature of the support used and the specific
interactions between the support and the active phase can have a pronounced
effect on the outcome of the catalytic reaction. This was demonstrated by
Touroude and his co-workers, who reported an increased chemo-selectivity
(70-80 %) of a Pt/ZnO catalyst used to hydrogenate crotonaldehyde to crotyl
alcohol (Figure 1.5) [17] , [18].
O
H

OH
H
O
H
OH
H
x
x
Pt/ZnO

Figure 1.5: Selective hydrogenation of crotonaldehyde to crotyl alcohol

The Pt/ZnO catalysts were prepared by wet impregnation of ZnO with an
aqueous solution of tetraammineplatinum (II) nitrate, followed by calcination
in air at 400
o
C for 4 h. Upon reduction at 200
o
C for 4 h under flowing H
2
,
some of the support, ZnO, is reduced to metallic Zn and dissolves into the Pt
to form a Pt-Zn alloy. Increasing the reduction temperature to 400
o
C results in
the total conversion of the Pt into a Pt-Zn alloy. The Pt
δ-
-Zn
δ+
entities favoured

the adsorption of the C=O bond over the C=C bond of crotonaldehyde. The
phenomenon was attributed to the use of an easily reducible support and the
electronic interactions between the Pt and Zn, resulting in the loss of Pt 5d
electrons and disfavouring the adsorption of C=C bond [19].
Besides the choice of support and active metal species, other catalyst
parameters such as surface composition, morphology and size of the metal
nanoparticles will greatly influence the activity of the catalysts. Decreasing the
!
!
11!
size of the nanoparticles results in an increase in the amount of exposed metal
surface atoms which can influence the rate of reaction. Abad et al. reported a
significant particle size effect when using supported gold catalysts for the
aerobic oxidation of cinnamyl alcohol to cinnamaldehyde carried out at 90
o
C
[20]. They showed that the turnover frequency (TOF) for the cinnamaldehyde
formation increases linearly with the number of external gold atoms present on
the Au/TiO
2
catalysts. Since the relative number of external gold atoms is
inversely proportional to the size of the Au nanoparticles (Figure 1.6), the
catalytic activity of the gold catalyst could be optimised by increasing the
metal dispersion and decreasing the size of the Au nanoparticles.


Figure 1.6: The surface-to-volume ratio decreases with increasing volume of a
particle [21].

Changes in the shape or size of the nanoparticle can also influence the specific

crystallographic orientation of the exposed catalyst surface and the relative
amount of the surface metal atoms occupying the edge, corner and terrace sites
(Figure 1.7). Kaneda et al. observed that in the oxidation of alcohols over Pd
nanoclusters, the catalytic reaction takes place over the low-coordinated Pd
atoms [23]. Similar observations were reported in the Heck [24] and Suzuki
[25] coupling using PVP-stabilised Pd nanoparticles. For the hydrogenation of
alcohols [26] and nitroaromatics [27], coordinatively unsaturated Ag atoms on
the edges and corner sites were reported to be much more active than Ag
!
!
12!
surface atoms located in terrace sites.

Figure 1.7: A schematic representation of the typical features of a metal
surface [22].


1.1.3 Catalyst preparation methods
The design of a supported metal catalyst is dictated by the catalyst preparation
method. Since the number of surface metal atoms increases with decreasing
metal particle size, catalyst preparation strategies often aim to synthesise
uniform metal nanoparticles with the smallest possible size. However, for such
small clusters, there is a strong thermodynamic driving force towards
aggregation into larger entities. Thus, a wide range of techniques have been
employed to deposit and stabilise these small metal nanoparticles onto a
support material to give highly active and stable catalysts. Some of the
techniques, i.e., impregnation, co-precipitation, deposition-precipitation, and
ion-exchange are discussed below.

1.1.3.1 Impregnation

To prepare supported metal catalyst via wet impregnation, a fixed amount of
solution containing the required amount of the metal precursor is mixed with
!
Adatom
Kink
Step
Terrace
Surface vacancy (‘hole)
represents a metal atom
!
!
13!
the support material. During the impregnation process, the solutes are
transported into the pore system as well as to the outer particle surface. There
is also diffusion and uptake of the solute by the pore wall. The process
conditions and the nature of the interaction of the solute with the support
material will determine the dispersion of the active phase. There are four
different types of active phase distribution, namely uniform; egg-shell, egg-
white and egg-yolk, as shown in Figure 1.8.


Figure 1.8: Active phase distribution during impregnation, (a) uniform;
(b) egg-shell, (c) egg-white and (d) egg-yolk [28].

Using the impregnation of alumina with H
2
PtCl
6
solution as an example, an
egg-shell distribution is typically achieved as the PtCl

6
2-
interacts with the
hydroxyl groups on the alumina surface [28]. However, the addition of citric
acid into the impregnating solution will result in the formation of an egg-white
type catalyst. This is because citric acid adsorbs more strongly than H
2
PtCl
6
and will be adsorbed onto the outer layer of the support. Thus, the Pt ions will
be adsorbed in a ring at the inside of the catalyst. Increasing the amount of
citric acid added will change the active phase distribution to the egg-yolk type.
This example demonstrates the need to take into account the chemical and
physical properties of the metal precursor and support material to vary the
type of dispersion.
!
!
14!
Following the impregnation step, the solvent is slowly evaporated to
effect the deposition of the metal precursors onto the surface and the pores of
the support material. As the solvent is removed, the solution in the pores
becomes oversaturated and the metal precursors will precipitate onto the
surface of the support. The solvent in the biggest pores will evaporate first,
causing the solution to accumulate in the smaller pores which will result in a
loss of dispersion. Rapid drying and the use of well crystallising precursor
salts result in a more homogeneous distribution of the active phase. However,
boiling of the precursor solution should be prevented as it can result in
inhomogeneous distribution. Slow drying while using well crystallising
precursor salts will give an egg-shell distribution. This is because the crystals
formed at the mouth of the pore will draw the solution out of the pores to the

outer surface through capillary action. After drying of the catalyst, the metal
precursor is decomposed at an elevated temperature, by either thermal
decomposition or reduction.
Examples of supported metal catalysts prepared by wet impregnation
are the Ru/TiO
2
and Ru/γ-Al
2
O
3
reported by Ran for the catalytic
decomposition of CH
2
Cl
2
[29]. These catalysts were prepared by impregnating
TiO
2
or γ-Al
2
O
3
with an aqueous solution of RuCl
3
. The average diameter of
the Ru particles is approximately 15 nm, as estimated from Scherrer equation.
A variation of the impregnation method is the incipient wetness technique,
where the amount of metal precursor solution added is just sufficient to fill all
the pores of the support material.

!
!
15!
1.1.3.2 Co-precipitation
For co-precipitation, solutions containing the precursors for the support and
the active phase are mixed and the pH of the solution is adjusted by the
addition of acids, bases or salts. The simultaneous precipitation of the two or
more components from the same solution at a fixed pH then results in the
formation of mixed crystals. The use of a constant pH optimises the
interaction of the precipitating material with the support and also prevents the
precipitation of unwanted compounds. Upon calcination or reduction, a
catalyst forms, which consists of an active phase dispersed through the surface
as well as the bulk of the support material. This intimate mixing of the active
phase and support increases the dispersion and ensures a homogeneous
distribution of the active phase. One example of such a catalyst is a highly
efficient Cu/ZnO/Al
2
O
3
catalyst described by Zhang et al. for the low
temperature steam reforming of methanol [30]. This catalyst was prepared by
adding an alcoholic solution of oxalic acid to an alcoholic solution containing
copper nitrate, zinc nitrate and aluminium nitrate to effect the precipitation.
The gel-like precipitates were centrifuged, dried and calcined to produce a mix
of copper oxide, zinc oxide with alumina as the support; particle size ranging
from 10 to 26 nm.

1.1.3.3 Deposition-precipitation
The deposition-precipitation method involves the precipitation of the active

phase such as metals, metal hydroxides, metal oxides or metal sulfides onto
the surface and into the pores of a preformed support material. The
precipitation is brought about by the change in pH, reduction or removal of a
!
!
16!
ligand [31]. The deposition is due to the electrostatic interactions between the
metal precursors and surface groups of the support which guide the nucleation
and growth of the metal crystallites. When oxides such as alumina or silica are
used as support material, the choice of metal precursors used for each support
material will be different. This is because the surface of alumina is covered by
basic hydroxyl groups while that of silica is covered with weakly acidic
hydroxyl groups. Using Pt catalyst as

an example, a negatively charged Pt
complex, PtCl
4
2-
, is typically used for the preparation of Pt/Al
2
O
3
, while for
Pt/SiO
2
, positively charged Pt(NH
3
)
4
2+

is used.
To obtain the best results with the precipitation-deposition method,
vigorous and effective mixing should be used together with slow addition of a
basic solution in order to avoid localised oversaturation. This prevents rapid
nucleation and growth in the bulk of the solution. This could lead to
inhomogeneous distribution since large crystallites will not be able to enter the
pores and instead deposit only on the external surface. One common method
of effecting the precipitation is through the use of a base such as NaOH, KOH
or urea. The hydrolysis of urea produces hydroxide ions (equation 1.1) which
react with metal ions to produce metal hydroxide precipitates (equation 1.2).
CO(NH
2
)
2

(aq)
+ 3 H
2
O ! CO
2

(g)
+ 2 NH
4
+

(aq)
+ 2 OH
-

(aq)
(1.1)
M
2+

(aq)
+ 2 OH
-

(aq)
! M(OH)
2

(s)
(1.2)
For example, alumina-supported gold catalysts were prepared via deposition-
precipitation using HAuCl
4
as the precursor

and NaOH as the precipitation
agent at 80
o
C [32]. A highly active catalyst with long term stability and small
nanoparticle diameter of around 2 nm could be prepared. Smaller and more
active gold nanoparticles were obtained at a higher pH of the precipitation
!
!
17!
solution, while larger and less active gold nanoparticles are formed when

lower pH (pH 5-6) was used. However, under high pH conditions, less gold
could be deposited on the support due to reduced electrostatic attractions.

1.1.3.4 Ion exchange
The ion exchange technique is based on the electrostatic attractions between
the charges present on the hydroxyl groups of the oxide support and the metal
cations and anions of the metal precursor at different pH. Hydroxyl groups on
the surface of the oxide supports such as Al
2
O
3
, SiO
2
, TiO
2
, MgO become
polarised when dissolved in water.
Table 1.1: A list of PZC of some common oxides in water [33]


Common oxides
PZC
WO
3

0.5
V
2
O
5

1-2
SiO
2

2.5
TiO
2

6
ZrO
2

6.7
ZnO
8
Al
2
O
3

9
MgO
12

The type of charge and degree of polarisation depend on the pH of the solution
and the point of zero charge (PZC) of the support. PZC is the pH at which the
surface of the support is electrically neutral. Table 1.1 (above) lists PZC
values for some oxides commonly used as support material.
When the pH of the solution is below the PZC, the hydroxyl groups

!
!
18!
become protonated and can adsorb negatively charged anions (equation 1.3).
In contrast, when the pH of the solution is above the PZC, the hydroxyl groups
become deprotonated and can adsorb positively charged cations (equation 1.4).
The schematic equations below show the difference in charges under different
conditions.
M-OH + H
+
! M-OH
2
+
(1.3)
M-OH + OH
-
! M-O
-
+ H
2
O (1.4)
Thus, by changing the pH of the solution, the choice of the support and metal
precursors used, we can control and fine-tune the composition of the catalyst.
Zhu et al. reported the preparation of Cu/SiO
2
catalysts via ion exchange for
the dehydrogenation of 2-butanol to methyl-ethyl-ketone [34]. It was done by
first adding ammonia to a Cu(NO
3
)

2
solution to adjust the pH to 10.
[Cu(NH
3
)
4
]
2+
was formed and SiO
2
was added into the solution and stirred for
2 h for adsorption to take place. The solid catalyst was then filtered, washed
with deionised water, dried and calcined before use. They achieved a Cu
loading of 5 wt. %, and the average diameter of the Cu particles was estimated
to be around 4 nm. For comparison, the Cu particles on a 5 wt. % Cu/SiO
2

catalyst prepared via wet impregnation using the same precursors had a much
larger particle size of 15 nm, and lower surface area. This indicates that the
ion-exchange method can achieve a better dispersion as compared to the wet
impregnation method.

1.2 Ag, Cu and Pd nanocatalysts in cross coupling reactions
Transition metals are employed as catalysts in many chemical reactions. This
is because they have incompletely filled d-orbitals which allow them to exhibit
!
!
19!
variable oxidation states. In particular, group 10 and 11 transition metals such
as Pd, Ag and Cu have garnered tremendous interest due to their high catalytic

activity for a wide range of reactions. These elements have been known since
prehistoric times and are widely studied. Table 1.2 shows some of the physical
properties of these transition metals.

Table 1.2: Physical properties of Pd, Cu and Ag

Pd
Ag
Cu
d-block group
10
11
11
Atomic weight
106.42
107.9
63.5
Electronic configuration
[Kr] 4d
10

[Kr] 4d
10
5s
1

[Ar] 3d
10
4s
1

Possible oxidation states
0, +1, +2,
0, +1, +2, +3
0, +1, +2, +3
Crystal structure
f.c.c.
f.c.c.
f.c.c.
Metal radii/ Å
1.37
1.44
1.28
Density/ g/cm
3

12.38
10.5
8.96
Melting point/ °C
1555
962
1085
Electronegativity (Pauling scale)
2.20
1.93
1.9
f.c.c.: face-centred cubic
Pd, Ag, and Cu nanocatalysts are used to catalyse several types of
reactions such as reduction [35], [36], oxidation [37], [38], [39], [40], [41] and

cross-coupling reactions [42], [43], [44], [45], [46], [47]. Among these, the
cross-coupling reactions have attracted most attention as the selective
formation of C-C bonds is central to the synthesis of many compounds of
interest, such as compounds with biphenyl motifs. Cross-coupling reactions
have become increasingly popular over the last thirty years, but most are
carried out using organometallic reagents and an organic electrophile,
!
!
20!
catalysed by homogeneous group 8-10 transition metal catalysts.
R
1
-X + R
2
-M ! R
1
-R
2
+ M-X (1.5)
where X = halides, triflates, tosylates, R
1
and R
2
= alkyl, aryl and alkenyl,
M = Mg (Kumada coupling), Zn (Negishi coupling), SnR’
3
(Stille coupling),
B(OR’)
2
(Suzuki coupling). In general, the steps involved in the mechanism of

such coupling reactions include oxidative addition of the substrate to the metal
centre, transmetallation between the substrate and the organometallic reagent,
and reductive elimination of the product to regenerate the catalyst (Scheme
1.1).

L
n
Pd(0)
L
n
Pd
R
1
X
L
n
Pd
R
1
R
2
R
1
-X
R
2
-M
X-M
R
1

-R
2
oxidative
addition
transmetallation
reductive
elimination

Scheme 1.1: General mechanism of cross-coupling reactions catalysed by a Pd
catalyst.

The different variants of this methodology have emerged over a short
period of time in the second half of the 1970ies, and each makes use of a
different main group metal (Mg, Zn, Sn, B) for the organometallic reagent,
different types of catalysts and substrates. For example, Suzuki coupling is a
cross-coupling reaction between aryl or vinyl boronic acid with an aryl or
vinyl halide catalysed by a palladium catalyst [48]. This method is highly
versatile and has been applied widely in the synthesis of biphenyl compounds.
!
!
21!
Ullmann coupling is another type of cross-coupling reaction which does not
require the use of the organometallic reagents mentioned above. These
reactions proceed only in the presence of metal catalysts, which activate the
substrates when they attach onto the catalyst metal centre. Although these C-C
bond formation reactions are mostly catalysed by homogeneous metal
catalysts, some research has been directed to develop nanoparticle catalysts or
supported metal catalysts because of the advantages of heterogeneous catalysts
as mentioned before. Some examples of heterogeneous Pd, Cu, and Ag
catalysts developed for Suzuki coupling and Ullmann reaction are described

below.

1.2.1 Suzuki coupling
Pd is the most commonly used metal catalyst for Suzuki coupling. The first
example of a heterogeneous catalyst for Suzuki coupling was described by
Marck et al. in 1994 [49]. They used a Pd/C heterogeneous hydrogenation
catalyst, and observed that the reaction proceeded in the presence [49], [50]
and in the absence of phosphine ligands (Scheme 1.2) [49]. At a reaction
temperature of 80
o
C, good to excellent yields were obtained when aryl
bromide or iodide was used as the coupling partner (Table 1.3). In analogy to
the homogeneous catalytic system, the electron-rich triphenylphosphine ligand
interacts with the surface of the Pd nanoparticles, forming an organometallic
intermediate which facilitates the oxidative addition of the aryl halides. The
steric bulk of the ligand also promotes reductive elimination of the product
[51].
!
!
22!
X
+
B
OH
OH
Pd/C (4 mol %)
EtOH, 80
o
C
R

2
R
1
R
1
R
2

Scheme 1.2: Ligand-free Pd/C-catalysed Suzuki Coupling

[49]

Table 1.3: Scope of reaction for ligand-free Pd/C catalysed Suzuki coupling

[49]


This methodology was further improved with use of a 20:1 DMA:H
2
O solvent
system, which allowed the coupling of poorly reactive aryl chlorides [52]. The
reaction could be completed within 1.5 h and was highly selective towards the
cross-coupling product. Excellent yields of 79-95 % were obtained for a range
of aryl chlorides with electron-withdrawing substituents. Subsequently, many
other Pd-based heterogeneous catalyst such as Pd/MgLa mixed oxides [53],
Pd/NaY zeolite [54], [55], Pd/SH-SBA-15 [56], and Pd/GO (GO= graphite
oxide) [57] were also presented as efficient and reusable catalysts for Suzuki
coupling reactions.
Besides monometallic Pd catalysts, Ag and Cu were also incorporated
to form bimetallic catalysts for Suzuki coupling reactions [58]. Kim et al.

!
!
23!
prepared Pd-Ag and Pd-Cu nanoparticles supported on ZnO via γ-irradiation
[59]. They found that these bimetallic catalysts exhibited slightly higher
activity than monometallic Pd/ZnO. The Pd content of the bimetallic catalysts
(6-10 %) could be reduced considerably compared to the monometallic
Pd/ZnO (20 %) and this suggests that Ag and Cu have synergistic effects with
the Pd species. The catalytic efficiency of the catalysts for Suzuki coupling
decreases in the order Pd-Ag/ZnO > Pd-Cu/ZnO > Pd/ZnO > Pd/C. The
catalysts could be easily recovered by simple filtration, and activity of the
recycled catalyst was maintained for 5 consecutive runs. Similar Pd-Ag and
Pd-Cu supported on carbon [60] and unsupported Pd or Cu monometallic
nanoparticles, Pd-Cu bimetallic nanoparticles and Cu-Pd-Ru trimetallic
nanoparticles [61] were also reported as active catalysts for Suzuki coupling.
In 2008, a heterogeneous Cu-Ni/C catalyst was developed and found to
catalyse Suzuki coupling in good yield [62]. The reaction was conducted in
dioxane at 180-200
o
C using microwave irradiation. This discovery was a
great improvement for the catalytic coupling methodology as Cu could be
used to replace the expensive Pd in the catalyst. However, one limitation of
this catalyst is that it could not be used to couple electron-rich aryl halides.

1.2.2 Ullmann reaction
Ullmann reaction normally refers to the reductive homocoupling of aryl
halides to form biphenyl compounds. Traditionally, this reaction is mediated
by a stoichiometric amount of copper, and only proceeds under harsh
conditions such as high temperatures above 200
o

C [63]. A proposed reaction
mechanism for Ullmann reaction (Scheme 1.3) starts with the oxidative
!
!
24!
addition of aryl halide to a copper metal centre to form an aryl-copper
intermediate. The intermediate undergoes another oxidative addition with a
second equivalent of iodobenzene, followed by reductive elimination to give
the biphenyl product.

I
+
Cu
CuI
Cu
Cu CuI
oxidative
addition
Cu
+
I
Cu
I
oxidative
addition
reductive
elimination
+
CuI

Scheme 1.3: Cu-catalysed Ullmann homocoupling of iodobenzene [64].

Over the years, this methodology has been greatly improved, so that
only catalytic amounts of Cu or Pd are required and milder reaction conditions
can be employed. In particular, the use of heterogeneous Pd/C in the presence
of reducing agents such as hydrogen [65], isopropanol [66], formate salts [67],
and zinc [68] for Ullmann coupling of aryl halides has attracted much
attention. Mukhopadhyay et al. studied the reaction conditions required for the
Ullmann coupling of chloroarenes using Pd/C in the presence of zinc as
reducing reagent (Scheme 1.4) [69].
Cl
+ Zn
+ OH
-
1 mol % Pd/C
8 mol % PEG-400
60-120
o
C, 1-2 h, H
2
O
R
1
R
2
R
2
R
1
R

1
R
2
R
1
R
2
+
+ ZnO
+
H
2
O + Cl
-
where R
1
= H, CH
3
and R
2
= H, CH
3
, CF
3

Scheme 1.4: Pd/C-catalysed Ullmann coupling of aryl chlorides [69]

This methodology was presented as a green alternative as it uses water as the
reaction medium and low temperatures of 60-100
o

C. PEG 400, a polyethylene
glycol of approximate molecular weight 400, functions similarly to ligands by
modifying the physical microenvironment of the Pd catalyst surface. This
!
!
25!
facilitates the interaction between substrates and reactants with the catalyst
surface. Water acts both as the solvent and as a reagent by reacting with zinc
to generate hydrogen for regeneration of the Pd(0) catalyst from Pd(II).
However, a fine control of the amount of hydrogen is crucial, as excess
hydrogen will adsorb onto the surface of the catalyst, forming palladium
hydride which can carry out reductive dehydrohalogenation. This results in the
formation of undesired benzenes as side products. Another disadvantage of
this method is the need for a large excess of zinc; the reaction between zinc
and water results in the formation of zinc oxide on the surface of the zinc
particle, causing its deactivation. Other heterogeneous Pd-based catalysts such
as Pd/SiO
2
, Pd/MIL-101 (a zeolite-type metal-organic framework (MOF)
material), Pd/CSP (carbon sphere support) were also developed for Ullmann
coupling of chloroarenes in water at low temperatures of 30-90
o
C [51]. In all
cases, it was claimed that the catalysts were stable under reaction conditions
and could be recycled several times with no loss in activity. However,
unresolved problems such as poor selectivity towards biphenyls illustrate the
need to develop more efficient catalysts.
Copper nanoparticles prepared via reduction and in-situ capping with
citrate ions were also reported as active catalysts for Ullmann homocoupling
of iodobenzene by Samim et al. [64]. The use of capping agents is crucial to

obtain nanoparticles of small particle size as they prevent particle growth by
agglomeration. The size of the nanoparticles decreased from 12 nm to 5 nm
when the citrate concentration was increased from 0.2 to 0.5 M. The
polydispersity of the Cu nanoparticles did not change with size, indicating that
the size of the nanoparticles could be fine-tuned with the concentration of the