SYNTHESIS, STRUCTURE AND CATALYTIC APPLICATION OF NOVEL CARBENE COMPLEXES WITH BENZOTHIAZOLIN 2 YLIDENE LIGANDS 2

Chapter two: Results and Discussion Part I Part I Section 2.2: Synthesis and Characterization of Benzothiazolium Salts – Solvent Free Synthesis Section 2.3: Synthesis of Mono-, Bis- a

aromatic N-heterocycle, pyridyl and azole ligands The coordination chemistry and catalytic activities of N,S- and N,N-NHC Pd(II) complexes are compared Part III

presents the formation of ring-opening of benzothiazolium salts (Section 2.10), the coordination of Pt(II)-NSHC-(C^N) complexes (C^N = 2-phenyl-pyridine) (Section 2.11) and the formation of 5-member and 6-member fused ring imidazolium salts via oxidative addition and reductive elimination (Section 2.12) In the following section, the synthesis and characterization of benzothiazolium salts will be discussed

Chapter two: Results and Discussion Part I

Part I

Section 2.2: Synthesis and Characterization of Benzothiazolium

Salts – Solvent Free Synthesis

Section 2.3: Synthesis of Mono-, Bis- and Dinuclear Pd(II)

Complexes with 3-benzylbenzothiazolin-2-ylidene ligand and

their Activities toward Mizoroki-Heck Coupling

Section 2.4: Synthesis and Characterization of Pd(II)

Complexes of NSHCs with Pendant and Coordinated Allyl Functionality and their Suzuki Coupling Activites

2.2 Synthesis and Characterization of Benzothiazolium Salts – Solvent Free Synthesis

The preparation of 3-benzylbenzothiazolium bromide (A) (in 65% yield) from

the reaction of benzothiazole with benzyl bromide in DMF at 95 °C has been reported

in the literature.32 However, only moderate yield was obtained with the use of dry and high-boiling point DMF as solvent In this work, when neat benzothiazole is treated with benzyl bromide in the absence of a solvent at 60 °C (Scheme 2.1), the desired

product A readily precipitates from the liquid mixture, which solidifies toward the end

of the reaction Washing with Et2O gives pure A as an off-white powder in

Chapter two: Results and Discussion Part I

Table 2.1 Selected 1H, 13C NMR and ESI data for A-D.

Salts

1

H NMR (ppm) of SCHN-proton a

13

C NMR (ppm) of SCHN-carbon a

The 1H NMR spectrum of A shows a characteristic downfield resonance at

12.27 ppm for the SCHN proton (SCHN = benzothiazolium proton), indicating the

formation of an azolium salt (Table 2.1) The formation of A is also supported by a

signal at 165.1 ppm in the 13C NMR spectrum for the SCHN carbon (SCHN = carbon

Fig 2.1 ORTEP representation of the cation of benzothiazolium salt C14H 12BrNS (A) with

50% thermal ellipsoids and labeling scheme; hydrogen atoms are omitted for clarity

carbene precursor) and a base peak at m/z 226 for the azolium cation in the positive

mode ESI mass spectrum (Table 2.1) The identity of A was further confirmed by its

molecular structure (Fig 2.1 and Table 2.2) elucidated by single-crystal X-ray

diffraction Compound A is described as a carbene precursor in Section 2.3

3-(2-Propenyl)benzothiazolium bromide B can be prepared by stirring neat

benzothiazole with a slight excess of allyl bromide at ambient temperature for a few days.13 The reaction time can be significantly shortened if the reaction is carried out at

60 ◦C, affording 81% yield in only 12 h (Scheme 2.1) However, higher temperature should be avoided in order to prevent thermal decomposition The characteristic downfield resonance at 11.87 ppm indicates the SCHN proton of the azolium salt

Fig 2.2 ORTEP representation of the cation of benzothiazolium salt C10H 10BrNS (B) with

50% thermal ellipsoids and labeling scheme; hydrogen atoms are omitted for clarity

Chapter two: Results and Discussion Part I

(Table 2.1) The 13C NMR signal for this SCHN carbon (SCHN = carbon carbene precursor) is at 165.0 ppm (Table 2.1) The positive mode of ESI mass spectrum

molecular ion is 176 m/z The crystal structure of B is shown in Fig 2.2 The use of

compound B as a carbene precursor will be described in Section 2.4

3-Propylbenzothiazolium bromide C was prepared in 78 % as a yellow

powder from 1-bromopropane and benzothiazole at 120 ◦C without solvent (Scheme 2.1) The 1H NMR spectrum of C in CDCl3 shows a pseudo-sextet at 2.17 ppm and two sets of triplets at 5.12 and 1.08 ppm characteristic of the propyl substituent The downfield resonance at 12.09 ppm for the SCHN proton (Table 2.1) indicates the formation of a benzothiazolium salt This is further supported by a 13C signal at 164.9 ppm for the SCHN carbon (Table 2.1) The positive mode ESI mass spectrum shows a

base peak at m/z = 178 corresponding to the [M − Br]+

cation (Table 2.1) The

Fig 2.3 ORTEP representation of the cation of benzothiazolium salt C10H 12BrNS (C) with

50% thermal ellipsoids and labeling scheme; hydrogen atoms are omitted for clarity

analysis of C by X-ray single-crystal diffraction confirms the benzothiazolium

structure with N-propyl substituent (Fig 2.3) The unit cell contains two independent

molecules, which show disorder of the propyl-substituent

3-Isopropylbenzothiazolium tri-iodide D forms readily from the reaction of

benzothiazole in neat 2-iodopropane (used in excess) Unlike the related

1,3-diisopropylbenzimidazolin-2-ylidene analogues, iPr2-bimyH+I-,215 it is isolated in its tri-iodide form, presumably from iodide and iodine addition reaction The formation

of iodine, which notably appears as a purple solid on the wall of the condenser at the end of the reaction, could be traced to photo-activation (from stray light) of alkyl iodide giving alkyl radical and iodine.216 The somewhat unsatisfactory yield (32%) is attributed to base-assisted Hofmann elimination of 3-isopropylbenzothiazolium iodide

to propene and benzothiazole (Scheme 2.2) The yield of D can be raised to 44%

when iodine is added to the reaction The product salt is soluble in common organic

Fig 2.4 ORTEP representation of the molecule of benzothiazolium salt C10H 12BrNS (D)

with 50% thermal ellipsoids and labeling scheme; hydrogen atoms are omitted for clarity

Chapter two: Results and Discussion Part I

solvents (e.g halogenated solvents, methonol, THF, CH3CN, DMSO, DMF) and

water, and it is generally more soluble than A-C

N

S

H I H H H

+ I2

N

S H

I3

I2hv

Scheme 2.2 Proposed of benzothiazolium salt D and side products

The benzothiazolium proton (SCHN) resonance is characteristically downfield

(11.52 ppm) (Table 2.1) It is also more deshielded compared with

1,3-diisopropylbenzimidazolium iodide, iPr2-bimyH+I- (10.79 ppm),215 which could be due to the replacement of nitrogen by a more electropositive sulfur and that there is only one exocyclic substituent on the heterocycle The thiazolium carbon (δC = 163.1 ppm) is downfield shifted by ~20 ppm with reference to the azolium carbon in 1,3-

iPr2-bimyH+X- (X = I, 139.5;215 X = Br, 140.7 ppm).217 The positive-mode ESI mass

spectrum shows a principal peak at m/z = 178 corresponding to the thiazolium cation

Salt D is used as a carbene precursor in preparing Pd(II) mononuclear and dinuclear

NSHC- complexes The results will be discussed in Section 2.7 and 2.8 X-ray crystal

structure diffraction analysis of D has confirmed the identity of 3-isopropyl

substituted benzothiazolium cation with the linear tri-iodide anion (Fig 2.4)

Under solventless conditions, benzothiazole readily reacts with alkyl halides

give A-D, which are air-stable that can be used conveniently

Chapter two: Results and Discussion Part I

Table 2.2 Selected bond lengths [Å] and angles [deg] for A-D

2.3 Synthesis of Mono-, Bis- and Dinuclear Pd(II) Complexes with benzylbenzothiazolin-2-ylidene ligand and their Activities toward Mizoroki-Heck Coupling

3-The synthetic pathway and structure of mono-, bis- and dinuclear NSHC complexes are discussed This will be followed by their catalytic application in Mizoroki-Heck reaction

Pd(II)-2.3.1 Synthesis of Pd(II) N, S-heterocyclic Carbene Complexes

The reaction of 3-benzylbenzothiazolium bromide and Pd(OAc)2 gives the Pd(II)-NSHC complex in a one-pot synthesis This is based on the methodology

developed for Pd(II) dicarbene complexes by using in situ deprotonation of azolium

salts with basic metal precursors.33, 218 Calò et al used a similar strategy to prepare a Pd(II) bis-(benzothiazolin-2-ylidene) complex from 3-methylbenzothiazolium iodide

and Pd(OAc)2 in THF.12(a)

The orange bis(NSHC) complex cis-[PdBr2(NSHC)2] (2.1) was prepared in

good yield (91%) from Pd(OAc)2 and 2 equivalents of A in refluxing CH3CN (Scheme 2.3) The positive mode ESI mass spectrum is dominated by an isotopic

pattern centered at m/z = 637 corresponding to a monocation formed upon bromide

dissociation from 2.1 At ambient temperature, the 1H NMR spectrum (DMSO-d6) of

2.1 exhibits broad signals that are slightly shifted upfield compared to the values

observed for the benzothiazolium salt A One broad singlet is noted at 6.34 ppm for

both methylene groups, indicating a relatively high energy barrier for the rotation around the Pd-C bond arising from the bulky benzyl substituent on the carbene ligand The 13C carbon carbene resonance at 203.8 ppm is significantly more upfield than the

Chapter two: Results and Discussion Part I

N

S H Br

+ 2 Pd(OAc)2

N N

S Pd

reflux

- 2 HOAc

Br Br

N S N

N S H

OAc + 2

DMSO, 70 oC

- 2 HOAc 4

Br Br

Br

2

Scheme 2.3 Synthesis of Pd(II) carbene complexes 2.1-2.4

analogous resonance observed by Calò et al in their trans-complex 1.45 (210.5

ppm),12(a) which points to a cis-configuration of the carbene ligands in 2.1 The

identity of 2.1 has been established, although its carbon analyses in microanalysis

remains unsatisfactory due to solvent molecules contained in the crystal lattice When the reaction is carried out in DMSO at 70 °C, a bromo-bridged dinuclear Pd(II) complex [PdBr2(NSHC)]2 (2.2(a)) forms, which shows a 1:1 carbene-to-metal ratio

The absence of the downfield signal for the SCHN proton in the 1H NMR spectrum

indicates a successful complexation of 2.2(a) Although the carbene resonance could

not be detected in the 13C NMR spectrum, the FAB mass spectrometric data shows a

peak that corresponds to the dinuclear cation (m/z = 903 corresponding to [2.2(a) -

Br]+) These results also suggest that the formation of a dinuclear vs dicarbene

complex is driven by the solvent instead of substrate stoichiometry The solvent has

an additional effect on the formation of dinuclear complex In the presence of donor solvents such as CH3CN or DMF, bridge cleavage readily occurs, giving the

mononuclear solvate complexes trans-[PdBr2(NSHC)(CH3CN)] (2.3) and

trans-[PdBr2(NSHC)(DMF)] (2.4), as shown in Scheme 2.3 The 13C carbon carbene signals

of 2.3 and 2.4 occur at 191.5 and 191.9 ppm, respectively

2.3.2 Molecular Structures

The molecular structures of complexes (2.1-2.4) have been determined by

X-ray single-crystal diffraction Selected bond lengths and bond angles are summarized

in Table 2.3 The structure of complex 2.1 reveals an essentially mononuclear

square-planar Pd(II) coordinated by two NSHCs and two bromo ligands in a cis-arrangement

Chapter two: Results and Discussion Part I

C(17)-N(2)-C(16) - - - 118.1(2) N(1)-C(1)-Pd(1) 129.20(4) 130.00(6) - 129.48(19) N(1)-C(1)-S(1) 110.90(3) 111.10(5) 111.6(2) 111.31(18)

Fig 2.5 ORTEP representation of the molecule [PdBr2(NSHC) 2] (2.1) with 50% thermal

ellipsoids and labeling scheme; hydrogen atoms are omitted for clarity

Chapter two: Results and Discussion Part I

interligand interactions between the carbenes The Pd-C bond distances (1.971(5) and

1.976(4) Å) are notably shorter than those found in

trans-diiodobis-(3-methylbenzothiazolin-2-ylidene)Pd(II) 1.45 (2.06(3) Å),12(a) suggesting a stronger

Pd-C bond in complexes These bonds are also shorter than those found in a

cis-dibromo-bis(benzimidazolin-2-ylidene)Pd(II) complex (1.996 Å).219 The other complexes studied herein have even shorter, and presumably stronger, Pd-C bonds The Pd-Br bond lengths (2.454(6) and 2.469(6) Å), on the other hand, are comparable

to other cis-dibromobis-(carbene) complexes of Pd(II).219 The N(1)-C(1)-S(1) angle of

the ligand precursor A (114.14°) has contracted to 110.9 Å upon coordination to give

2.1 Other structural parameters remain unchanged, indicating that the coordination to

the Pd center mainly affects the carbene carbon as well as the neighboring nitrogen and sulfur atoms

The dimeric molecular structure of the CHCl3 solvate of 2.2(a) was also

confirmed by an X-ray single-crystal diffraction technique (Fig 2.6) Each of the two Pd(II) centers is coordinated by one carbene, one terminal bromo, and two bridging µ-

Fig 2.6 ORTEP representation of the molecule [PdBr2(NSHC) 2] (2.2(a)) with 50%

thermal ellipsoids and labeling scheme; hydrogen atoms are omitted for clarity

bromo ligands in a square-planar fashion The metal can be stabilized by the basic

bridging bromide The two carbene ligands in 2.2(a) are syn to each other across the

flat dinuclear frame The dihedral angles between both carbene ring planes and the coordination plane are 68.8° and 74.7° respectively Both 3-benzyl substituents are on the same side of the metal coordination plane The Pd-C bonds (1.936(7) and 1.960(8)

Å) are even shorter and stronger than those in 2.1, again indicating higher Lewis acidity of the Pd(II) centers in 2.2(a) as compared to that in 2.1 The six Pd-Br bonds

can be divided into three different sets with significantly different lengths The terminal Pd-Br bonds are the shortest (2.389(1)-2.415(1) Å), whereas the bridged

bonds that are trans to the carbenes are the longest (2.539(1)-2.532(1) Å)

The bridging Pd-Br bonds are readily cleaved by CH3CN and DMF This is

consistent with the fact that the Pd-Br bonds are weak The cleavage of 2.2(a) with the two coordinating solvents afford complexes 2.3 (CH3CN) and 2.4 (DMF) Fig 2.7 ORTEP representation of the molecule [PdBr2(NSHC)(CH 3CN)] (2.3) with 50%

thermal ellipsoids and labeling scheme; hydrogen atoms are omitted for clarity

Chapter two: Results and Discussion Part I

respectively, which also have been structurally characterized by single-crystal X-ray

diffraction The molecular structures of 2.3 and 2.4 are depicted in Fig 2.7 and 2.8

respectively Both are square-planar mononuclear and monocarbene Pd-(II)

complexes with a weakly coordinated solvent molecule occupying the position trans

to the carbene ligand The Pd-C bond lengths are 1.936(3) and 1.921(2) Å for

complexes 2.3 and 2.4 respectively These Pd-C bond distances are shorter than those

in 2.1, which again reflect a more Lewis acidic metal center The dihedral angles

between carbene ring planes and the PdBr2CH3CN or PdBr2DMF coordination planes are 76.5° and 73.8° respectively The adjacent 3-benzyl rings are oriented almost orthogonally to the benzothiazolyl rings with C(1)-N(1)-C(8) angles of 91.2° and

95.3° for complexes 2.3 and 2.4 respectively

Fig 2.8 ORTEP representation of the molecule [PdBr2 (NSHC)(DMF)] (2.4) with 50%

thermal ellipsoids and labeling scheme; hydrogen atoms are omitted for clarity

2.3.3 Mizoroki-Heck Coupling Reaction

With the isolation and characterization of this series of N,S-heterocyclic

carbenes complexes, it is now possible to directly compare the catalytic activities of mono- and biscarbene as well as mono- and dinuclear Pd(II) complexes The

Mizoroki-Heck coupling reactions of various aryl bromides and chlorides with

tert-butyl acrylate catalyzed by complexes 2.1-2.4 lead to the corresponding cinnamates (Table 2.4) Complexes 2.1-2.4 are air-stable, and hence there is no necessity for handling these complexes in a glovebox The dicarbene complex 2.1 is efficient

toward the coupling reaction between 4-bromobenzaldehyde or 4-bromoacetophenone and the acrylate, giving quantitative conversion (100%) (Table 2.4, Entries 1-6) Slightly lower (75-79%) conversions were observed when deactivated aryl bromides were used (Table 2.4, Entries 7-9) To further investigate the effect of the catalyst loading on the conversion rate, coupling of 4-bromobenzaldehyde with the acrylate

using complex 2.1 as precatalysts was examined As shown in entries 1-4, the loading

of catalyst from 1 mol % to 0.05 mol % gave quantitative conversion (≥ 95%)

Table 2.4 Mizoroki-Heck coupling reactionsa catalyzed by complexes 2.1- 2.4

Solvent, NaOAc O

O

R [Pd]

-HX

X = Br, Cl

R = CHO, CH3CO, CH3O, OH, CH2OH, CN

O O

Entry [Pd]

Cat

load [mol

%]

Aryl halide Solvent t

[h]

Temp [ºC]

sion b [%]

2 2.1 0.5 4-bromobenzaldehyde DMF 22 100 100

3 2.1 0.25 4-bromobenzaldehyde DMF 22 100 99

Chapter two: Results and Discussion Part I

10 2.2(a) 1 4-bromobenzaldehyde Toluene 17 100 17

11 2.2(a) 1 4-bromoacetophenone Toluene 17 100 47

12 2.2(a) 1 4-bromoanisole Toluene 17 100 25

13 2.2(a) 1 4-bromoanisole Toluenec 21 110 25

A kinetic plot of the coupling reaction between electron-withdrawing

4-bromobenzaldehyde and tert-butyl acrylate under 1 mol % catalyst 2.1 is given in Fig

2.9 The reaction is essentially complete within 3 h (monitored by 1H NMR spectroscopy), whereas up to 60% of product is formed within the first 20 min

Complex 2.1 thus promotes a remarkably fast coupling, although the reaction profile

shows an induction period for the first 10 min upon mixing It suggests a fairly rapid reduction of Pd(II) to the catalytically active Pd(0) The turnover points to an initially sluggish product formation followed by a marked increase in turnover with time A similar activation step has been reported for cis-diiodo-bis-(N,N’-

dimethylbenzimidazolin-2-ylidene)Pd(II).17 When DMF is replaced by, for example,

Fig 2.9 Concentration/time diagram (amount of substance x [%], time t [min]) for the

Mizoroki-Heck olefination of 4-bromobenzaldehyde with tert-butyl acrylate to form tert-butyl (E)-4-formylcinnamate (■) catalyzed by complex 2.1.

Chapter two: Results and Discussion Part I

toluene, the conversion decreases markedly to only 18% (Table 2.4, Entry 5)

The catalytic efficiency of the dinuclear 2.2(a) is the poorest among the

complexes studied and unsatisfactory in toluene (Table 2.4, Entries 10-13), giving the conversion of 47% (Table 2.4, Entry 11) This could be attributed to its high stability

in toluene; the poor donor character of toluene also does not cleave the Pd(II)-NSHC

dinuclear complex Use of DMF as solvent for 2.2(a) effectively gives 2.4 The

catalytic activities of 2.3 were examined in three different solvents, viz., CH3CN, DMF, and toluene (Table 2.4, Entries 14-24) The reaction was compared with those

complexes 2.1-2.2(a) and 2.4 The best activity is observed when DMF was used as

solvent, giving 95% conversion of product with 4-bromobenzaldehyde as substrate (Table 2.4, Entry 16) In CH3CN, the highest conversion recorded is 54% (Table 2.4, Entry 19) when sodium formate is added (Table 2.4, Entries 15, 19, and 21) The reaction carried out in toluene gives only 19% (Table 2.4, Entry 17) The high donor character of DMF appears to provide the best catalyst stability The catalytic activity

of 2.3 depends significantly on the temperature and solvent In most cases, catalyst

decomposition was observed, giving rise to palladium black These results suggest

that 2.3 is unstable during catalysis which could explain some low activities The

above results suggest some benefits of using DMF as a solvent or ligand solvate

When the DMF complex 2.4 is used, quantitative conversion (≥ 99%) of

4-bromobenzaldehyde, 4-bromoacetophenone, and 4-bromoanisole are obtained (Table 2.4, Entries 25-27) However, this does not apply to the deactivated aryl bromides such as 4-bromophenol and 4-bromophenylmethanol, which give low conversions

(Table 2.4, Entries 28 and 29) The catalytic activity of complexes 2.1 and 2.4 with

activated aryl chlorides and deactivated aryl chlorides such as 4-chlorobenzaldehyde, 4-chloroacetophenone, and 4-chlorobenzonitrile were studied However, the observed

catalytic activity is low (conversion below 10% and improved slightly, <20%, when tetrabutylammonium bromide was added) (Table 2.4, Entries 30-41) The catalytic activity of oxalin-2-ylidene Pd(II) complexes also showed low conversion for the activated aryl chloride.220 This is possibly due to the less electronegative sulfur as opposed to nitrogen in benzothiazolin-2-ylidene in the Pd(II)-NSHC complexes

compared to Pd(II)-NNHC Complexes 2.1 and 2.4 show a greater catalytic activity than complexes 2.2(a) and 2.3 The catalytic activities of complexes 2.1 and 2.4 are

comparable with Caló’s bis(2,3-dihydro-3-methylbenzohiazole-2-ylidene)Pd(II) diiodide complex when the reactions are carried out in molten tetrabutylammonium bromide.12(a)

2.3.4 Conclusion

A simple and direct method for the synthesis of the 3-benzylbenzothiazolium salt and the related benzothiazolin-2-ylidene complexes of Pd(II) was developed

Treatment of A with Pd(OAc)2 in CH3CN affords the bis(carbene) complex

cis-[PdBr2(NSHC)2] (2.1) (NSHC= 3-benzylbenzothiazolin-2-ylidene) In DMSO, this

reaction yields an unprecedented dinuclear N,S-heterocyclic carbene complex

[PdBr2(NSHC)]2 (2.2) Complex 2.2 undergoes bridge cleavage reactions with

CH3CN and DMF to give the mononuclear and solvated monocarbene complexes

trans-[PdBr2(NSHC)(Solv)] [Solv = CH3CN (2.3) and DMF (2.4)] The catalytic activities of 2.1-2.4 toward Mizoroki-Heck coupling reactions of aryl bromides with

tert-butyl acrylate are described and compared

Chapter two: Results and Discussion Part I

2.4 Synthesis and Characterization of Pd(II) Complexes of NSHCs with Pendant and Coordinated Allyl Functionality and their Suzuki Coupling Activities

A simple pathway to mono- and bis- as well as dinuclear Pd(II) complexes

with 3-benzylbenzothiazolin-2-ylidene ligand was discussed in the previous section

In this section, a functional side-chain [3-(2-propenyl)benzothiazolium-2-ylidene ligand] is introduced in the synthesis of Pd(II)-NSHC complexes Apart from the general tuning of electronic properties, a donor-functionalized arm in the ligand core could enhance the potential for hemilability The synthesis and structural characterizations of the novel allyl derivatized NSHC Pd(II) complexes and early studies of their activity in Suzuki–Miyaura coupling toward aryl bromides and chlorides are discussed The mixed dicarboxylato-NSHC complexes with the 3-(2-propenyl)benzothiazolium-2-ylidene ligand is also discussed

2.4.1 Synthesis and Characterization of Pd(II) Complexes (cis-,

trans-isomers and dicarboxylato-bis(NSHC) Pd(II) Complexes)

The reaction of Pd(OAc)2 with precursor B in a 1 : 2 ratio in refluxing CH3CN

gives a mixture of cis- and trans-Pd(II) bis(carbene) complexes as major products in

an overall yield of 92% (Scheme 2.4) The two geometrical isomers, cis-2.5 and

trans-2.5 have different solubility in CH3CN and can be separated by a simple

filtration

S H

Br + Pd(OAc)2

N

S Pd S N

Br Br

Br Br

N

S Pd

Br Br

Scheme 2.4 Synthesis of Pd(II) carbene complexes trans-2.5, cis-2.5, 2.6 and 2.7.

The insoluble (in CH3CN) complex cis-[PdBr2(NHSC)2] (cis-2.5, NSHC =

3-(2-propenyl)benzothiazolin-2-ylidene) can be isolated in a moderate yield of 32% as a white and air stable solid Column chromatography on the filtrate mixture afforded

trans-[PdBr2(NHSC)2] (trans-2.5) as an orange powder in 60% yield The positive

mode ESI mass spectra for both isomers is dominated by an isotopic pattern centered

at m/z = 537 corresponding to [2.5 − Br]+

Their 1H NMR spectra recorded in

DMSO-d6 are also similar with only slight differences in chemical shifts For both complexes,

the absence of the downfield signal for the SCHN proton characteristic for B indicates

a successful formation of carbene complexes The NCH2 resonance for cis-2.5 gives rise to a broad doublet at 5.92 ppm, whereas for trans-2.5 two very broad signals are

Chapter two: Results and Discussion Part I

observed from 5.92–5.85 ppm These less well resolved resonances most probably

result from the overlap of signals of trans-syn and trans-anti isomers for trans-2.5,

which are in dynamic equilibrium with each other and cannot be separated Finally, the 13C carbon carbene signal for trans-2.5 resonances at 212.2 ppm, which is significantly further downfield than the resonance at 203.0 ppm of the cis-2.5 The

chemical shifts for both isomers fall within the range reported for analogous Pd(II) complexes (Section 2.3)

Two minor products [PdBr2(NSHC)] (2.6) and trans-[PdBr2

(benzothiazole-ĸN)(NSHC)] (2.7) were isolated by column chromatography of the resultant filtrate in

very low yields (Scheme 2.4) However, these complexes could not be further separated

2.4.2 Halide substitution reaction

The isolation of cis-2.5 gives a new Pd(II)-NSHC carboxylate complex 2.8 A metathesis of cis-2.5 with AgO2CCF3 was carried out, and its trifluoroacetato

derivative 2.8 was isolated (Scheme 2.5)

S N

N S Pd Br Br

S N

N S Pd

Halide substitution of bromo with trifluoroacetate ligand only leads to insignificant changes of the 1H NMR resonances of the NSHC ligand The carbene signal, on the other hand, experiences a pronounced upfield shift from 203.0 ppm in

cis-2.5 to 192.0 ppm in 2.8 Such a behavior has also been observed for imidazole-

and benzimidazole-based NNHC systems.5(i),221 The 19F NMR spectrum shows a sharp

singlet at 2.49 ppm for the CF3-group, suggesting a structure with terminal carboxylato ligands with dissociation

2.4.3 Molecular Structures

X-Ray diffraction analysis on single crystals of cis-2.5·DMF obtained from

vapor diffusion of Et2O into a concentrated DMF solution shows a square planar Pd(II)

centre with two cis-oriented NSHC carbene ligands, each of which bears a pendant

allyl side-arm (Fig 2.10 and Table 2.5) To overcome repulsion between the two

Fig 2.10 ORTEP representation of the molecule cis-2.5.DMF with 50% thermal

ellipsoids and labeling scheme; hydrogen atoms and the DMF molecule are omitted for

clarity.

Chapter two: Results and Discussion Part I

carbene ligands, the two carbene planes are tilted nearly perpendicularly to the PdC2Br2 coordination plane with dihedral angles of 82° and 88° respectively The two allyl substituents are on opposite sides with respect to the PdC2Br2 plane The Pd–

Ccarbene bond distances of 1.976(2) Å are similar to those found in related NSHC complexes (cis-dibromobis(3-benzylbenzothiazolium-2-ylidene)Pd(II):

Pd(II)-1.971(5) and 1.976(4) Å ) (Section 2.3) and shorter than those in trans derivatives

(trans-diiodobis(3-methylbenzothiazolin-2-ylidene)Pd(II): 2.063(3) Å ).12(a) The two

terminal C=C bonds showed some disorder with 65 : 35 occupancy ratios

Table 2.5 Selected bond lengths [Å] and angles [deg] for complexes cis-2.5, 2.6 and 2.7

[PdBr2(benzothiazole-ĸN)(NSHC)] (2.7) in very low yields (< 10%) (Scheme 2.4)

Vapor diffusion of Et2O into CH2Cl2 solutions of both compounds yielded some single crystals, which have been crystallographically analyzed (Fig 2.11 and 2.12)

Compound 2.6 is an unusual Pd(II) monocarbene complex with coordinated allyl side

arm Such a chelating coordination mode of an allylfunctionalized NNHC ligand is rare in palladium chemistry, but has been observed for Ir(I) complexes.13, 76(a) Notably, the Pd–Ccarbene distance of 1.958(3) Å in 2.6 is significantly shorter than those of cis-

Chapter two: Results and Discussion Part I

2.5, which is in agreement with a Lewis acidic metal center in monocarbene

complexes The coordination of the allyl moiety in 2.6 has led to an elongation of the C9–C10 bond (1.374(5) Å) compared to B (1.309(3)Å ) (Section 2.2)

This phenomenon leads to a slight decrease in bond order consistent with a

weak π-basicity of the allyl arm toward the Pd(II) center This in turn suggests a labile

Fig 2.11 ORTEP representation of complex 2.6 with 50% thermal ellipsoids and labeling

scheme; hydrogen atoms are omitted for clarity.

Fig 2.12 ORTEP representation of complex 2.7 with 50% thermal ellipsoids and labeling

scheme; hydrogen atoms are omitted for clarity

coordination of the allyl arm, which in combination with the strongly coordinating carbene donor may give rise to hemilabile behavior of the functionalized NSHC ligand The Pd–C9 (2.188(3) Å) and Pd–C10 (2.181(3) Å) bond lengths are significantly longer than the Pd–Ccarbene bond Unlike in cis-2.5, there is steric

hindrance experienced by the NSHC ligand where the double bond of allyl-subtituent coordinates to Pd(II) metal center, and hence its plane is nearly coplanar (14 °) with the PdBr2 coordination plane

The molecular structure of the minor by-product 2.7 is depicted in Fig 2.12

The compound has been identified as a square planar dibromo monocarbene complex,

in which its fourth coordination site is completed by a benzothiazole-ĸN ligand

Complex 2.7 may have formed through attack of benzothiazole on 2.6 upon

elimination of the labile allyl side arm, again highlighting the hemilabile behavior of this functionalized NSHC ligand The formation of free benzothiazole from

deallylation of B has been discussed elsewhere13 and may involve a radical

[1,3]-sigmatropic rearrangement of free dimeric carbenes,31-32 which illustrates the frailty of

such a reactive substituent on the NSHC ring The Pd–Ccarbene bond of 1.945(11) Å of

2.7 is the shortest compared to cis-2.5 and 2.6, and presumably strongest which is

attributable to the absence of a strong trans-ligand or steric repulsions Both the

benzothiazole and the NSHC plane are almost perpendicular to the PdCNBr2

coordination plane with dihedral angles of 89 ° and 78 ° respectively

2.4.4 Suzuki-Miyaura Coupling Reaction

One of the incentives for this work was to find new NSHC carbene complexes

that could serve as phosphine-mimics to promote chemical reactions Complexes

cis-2.5, trans-2.5 and 2.8 have been chosen for this study due to their ease of preparation

Chapter two: Results and Discussion Part I

Complex 2.8 could represent the NSHC version of the commonly used

Pd(OAc)2/phosphine mixtures The coupling of selected aryl bromides with phenylboronic acid in the presence of Cs2CO3 as a base in DMF served as a standard test reaction for comparison and the results are summarized in Tables 2.6

Table 2.6 Comparison of cis-2.5, trans-2.5 and 2.8 as catalysts for the Suzuki-Miyaura

couplinga of aryl bromides with phenylboronic acid

Br + B(OH)2

DMF/Cs2CO3/100 oC

R = CHO, COCH3, OCH3

R

R [Pd]

Schlenk techniques Complexes cis-2.5 and trans-2.5 give quantitative conversion (≥

99%) of activated substrates (4-bromobenzaldehyde and 4-bromoacetophenone)

(Table 2.6, Entries 1-2 and 8-9) Cis-2.5 is more active than trans-2.5 in unactivated substrates such as 4-bromoanisole even after prolonging the reaction in trans-2.5 (19 hours) (Table 2.6, Entries 7 and 10) Both cis-2.5 and 2.8 give near-quantitative

conversion (≥ 99%) to electron poor substrates such as bromobenzaldehyde or bromoacetophenone, even at catalytic loads as low as 0.05 mol% (Table 2.6, Entries 4

4-and 14) Cis-2.5 affords a good conversion of 74% of 4-bromoanisole substrate,

whereas 2.8 only leads to 38% conversion of 4-bromoanisole substrate (Table 2.6,

Entries 7 and 16)

Table 2.7 Suzuki-Miyaura couplinga reaction catalyzed by cis-2.5

+ B(OH)2

Chapter two: Results and Discussion Part I

The good performance of cis-2.5 is encouraging Under the given conditions,

methyl substituted aryl bromides can be coupled with good yields of 60–80% (Table 2.7, Entries 2–4) It is noteworthy that donor groups such as alcohol or pyridine are well tolerated in this coupling reaction giving good to moderate conversions (63-100%) (Table 2.7, Entries 1, 7-9) Aryl chlorides such as 4-chlorobenzaldehyde and 4-chloroacetophenone can also be coupled with near-quantitative yield (≥ 99%) (Table 2.7, Entries 5 and 6), although prolonged reaction times are required

Table 2.8 Suzuki-Miyaura coupling catalyzed by a Pd(OAc)2/benzothiazolium

Br + B(OH)2

The Suzuki coupling catalyzed by in situ generated catalyst was studied For

activated 4-bromobenzaldehyde, the Pd : B (ratio of both 1 : 2 and 1 : 1) catalysts

which were generated in situ give quantitative conversion (100%) This indicates that

monocarbene complexes may be involved as catalytically active species (Table 2.8, Entries 2 and 3).215 This result is in agreement with the optimum Pd : salt ratio of 1 : 1

reported for a similar imidazolium based catalyst.222 When deactivated aryl bromides

are employed, the 1 : 1 ratio is less effective compared to the precatalyst cis-2.5,

which has a metal-to-ligand ratio of 1 : 2 (Table 2.8, Entries 4 and 5) Further increase

of compound B concentration leads to a decrease in conversion (Table 2.8, Entry 1)

2.4.5 Conclusion

3-(2-Propenyl)benzothiazolium bromide (B) provides a direct and simple entry

to Pd(II) complexes with N,S-heterocyclic carbene (NSHC) ligands functionalized

with an allyl pendant with hemilabile potential Addition of compound B to Pd(OAc)2

eliminates HOAc and affords the bis(carbene) complexes cis-[PdBr2(NSHC)2]

(cis-2.5, NSHC = 3-(2-propenyl)benzothiazolin-2-ylidene) and trans-[PdBr2(NSHC)2]

(trans-2.5) along with the monocarbene complexes [PdBr2(NSHC)] (2.6) and

trans-[PdBr2(benzothiazole-ĸN)(NSHC)] (2.7) as minor side products A metathesis of

cis-2.5 with AgO2CCF3 yields the mixed dicarboxylato-bis(carbene) complex

cis-[Pd(O2CCF3)2(NSHC)2] (2.8) Complexes cis-2.5, trans-2.5 and 2.8 were

characterized by multinuclear NMR spectroscopy, ESI mass spectrometry and

elemental analysis The molecular structures of complexes cis-2.5, 2.6 and 2.7 have been determined by X-ray single crystal diffraction Complexes cis-2.5 and 2.8 as

well as an in situ mixture of Pd(OAc)2 and compound B are active toward Suzuki–

Miyaura coupling of aryl bromides and activated aryl chlorides giving good conversions

Chapter two: Results and Discussion Part II

Part II

Section 2.5: Pd(II) Complexes of Mixed

Benzothiazolin-2-ylidene and Phosphine Ligands and their Catalytic Activities

toward C-C Coupling Reactions

Section 2.6: Mono- and Dinuclear Pd(II) N, S-heterocyclic

Carbene Complexes with Aromatic N-Heterocycle

Section 2.7: Formation and Structures of Pd(II) NSHC-Pyridyl

Mixed-Ligand Complexes

Section 2.8: Benzothiazolin-2-ylidene and Azole Mixed-Ligand

Complexes of Pd(II)

Section 2.9: Structures and Suzuki-Coupling of NNHC

Complexes of Pd(II) with Coordinated Solvate and PPh3

2.5 Pd(II) Complexes of Mixed Benzothiazolin-2-ylidene and Phosphine Ligands and their Catalytic Activities toward C-

C Coupling Reactions

The NNHC-palladacycles of Nolan,223 NNHC-phospha-palladacycles194 and

NNHC-phosphine complexes of Herrmann,209(e) Cloke,45 and Welton,224diaminocarbene- and Fisher-carbene complexes with PPh3 of Fürstne140(c) and

pyridylcarbene complexes of Lin225 are just a few of the many successful examples of

Pd(II) mixed-NNHC-phosphine complexes However, there is no report of Pd(II) complexes with mixed-NSHC-phosphine ligands In Part II, a series of Pd(II) complexes with mixed benzothiazolin-2-ylidene and ligand (ligand = phosphine,

aromatic N-heterocycle, pyridyl and azole ligands) and their catalytic studies will be

presented

It was mentioned in Section 2.3 that the Pd(II)-NSHC dinuclear complex,

2.2(a) is a useful precursor to monocarbene Pd(II)-solvated complexes 2.3 and 2.4

The dinuclear Pd(II)-NSHC species can undergo bridge-cleaved with various donor ligands In the following section, the results on the synthesis of Pd(II)-mixed benzothiazolin-2-ylidene and phosphine ligands as well as their catalytic activities toward C-C coupling reactions will be presented

2.5.1 Synthesis of Dinuclear Pd(II) Complexes

The synthesis of Pd(II)-NSHC dinuclear complex 2.2(a) bearing the

3-benzylbenzothiazolin-2-ylidene ligand was described in Section 2.3 Using the same

method, the dinuclear Pd(II) carbene complex 2.2(b) with ylidene ligands from the reaction of two equivalents of azolium salt C with one

3-propylbenzothiazolin-2-Chapter two: Results and Discussion Part II

equivalent of Pd(OAc)2 in 76% yield was isolated The absence of the downfield signal in the 1H-NMR spectrum of 2.2(b) for the SCHN proton characteristic for C indicates a successful complexation The propyl resonances are similar to those of C,

but the aromatic resonances are slightly shifted upfield by ∼0.3 ppm Although the signal of carbon carbene could not be detected in the 13C NMR spectrum, the FAB

mass spectrometric data agrees with a dinuclear structure (m/z = 807) corresponding

to [2.2(b) − Br]+

N S N

S

Pd Pd R

R Br

Br Br

N

S Pd Br Br

P N N

S Pd Br Br P

Scheme 2.6 Synthesis of Pd(II) mixed NSHC-phosphine complexes 2.9-2.11

2.5.2 Cleavage of 2.2(a) and 2.2(b) with triphenylphosphine (PPh3)

The bromo-bridged complexes 2.2(a) and 2.2(b) are known to be cleaved by

coordinating solvents such as CH3CN and DMF, giving the mononuclear solvate

complexes (Section 2.3) They are thus readily cleaved by various ligands to form a

range of mononuclear monocarbene complexes Reaction of complexes 2.2(a) and

2.2(b) with PPh3 gave mixed complexes 2.9(a) and 2.9(b), which have been isolated

as cis-isomers (Scheme 2.6) The 1H-NMR spectrum (CDCl3) of 2.9(a) exhibits an

AX system with two sets of doublets for the geminal methylene protons of the benzyl substituent at 6.53 and 5.00 ppm (∆δ(AX) = 1.53 ppm) with a geminal coupling constant of 2J(HH) = 15.10 Hz (Table 2.9) The geminal coupling of the benzylic protons indicates that there is hindered rotation around the Pd–Ccarbene bond due to the bulky PPh3 ligand Complex 2.9(b) behaves similarly Its propyl substituent also

exhibits an AX system with two sets of triplets for CH2CH2CH3 at 4.81 and 4.05 ppm

as well as two sets of multiplets for CH2CH2CH3 at 2.41 and 1.79 ppm (Table 2.9) The terminal methyl group of the propyl substituent remains unaffected, but shifts

slightly downfield by ∼0.4 ppm compared to C Apparently, the terminal methyl

group is further away from the Pd–Ccarbene bond and thus has greater flexibility This phenomenon once again indicates hindered rotation (in the 1H NMR time scale) of the Pd–Ccarbene bond due to the neighboring bulky phosphine ligand

The 13C NMR signals for the carbenoid carbons of 2.9(a) and 2.9(b) appear at

208.4 and 206.3 ppm respectively (Table 2.9) They are downfield compared to those

of the mononuclear solvate complexes reported earlier (191.5 ppm for

trans-[PdBr2(NSHC)(CH3CN)] and 191.9 ppm for trans-[PdBr2(NSHC)(DMF)]) (Section

2.3) This observation is in line with a lower Lewis-acidity of the Pd(II) centers in

2.9(a) and 2.9(b), which is caused by an additional strong σ-donating phosphine ligand The 31P NMR signals of the coordinated phosphine is also slightly downfield-

shifted (27.3 ppm for 2.9(a) and 27.7 ppm for 2.9(b), Table 2.9) as compared to the

1,3-diisopropylbenzimidazolin-2-ylidene analogue cis-[PdBr2(NNHC)PPh3] (26.6 ppm).217

Chapter two: Results and Discussion Part II

Table 2.9 Selected 1H, 31P and 13C NMR and crystallographic data for complexes 2.9-2.11 NMR

data 2.9(a) 2.9(b) 2.10(a) 2.10(b) 2.11(a) 2.11(b)

6.69, 6.14

5.39, 4.46

6.78, 5.26

5.04, 4.28

15.75, 15.75

5.25, 4.00

15.10, 15.80

5.03, 4.83

1.964(4) 1.958(7) 1.971(6) 1.972(3) 1.971(3) Pd1-Br1 2.459(8) 2.470(6),

2.470(6) 2.466(11) 2.466(9) 2.472(4) 2.480(4) Pd1-Br2 2.471(8) 2.480(6),

2.484(6) 2.479(10) 2.484(8) 2.467(4) 2.463(4) S1-C1 1.716(6) 1.703(4),

1.711(4) 1.709(7) 1.713(6) 1.703(3) 1.714(3) N1-C1 1.325(7) 1.328(5),

1.320(5) 1.335(8) 1.345(7) 1.326(4) 1.326(4) N1-C8 1.468(7) 1.479(5),

Br1 175.16(17)

178.11(12), 178.85(11) 175.8(2) 174.39(15) 178.61(8) 178.02(9) P1-Pd1-

Br1 87.48(5)

88.26(3), 91.40(3) 89.31(6) 91.40(5) 88.03(2) 89.85(2) C1-Pd1-

Br2 87.76(16)

86.34(12), 87.55(12) 86.2(2) 84.27(15) 87.37(9) 88.45(8)