DERIVATIZATION OF
NANODIAMONDS WITH OSMIUM
AND RUTHENIUM CARBONYL
CLUSTERS
OH SUAT PING
NATIONAL UNIVERSITY OF
SINGAPORE
2010
DERIVATIZATION OF NANODIAMONDS
WITH OSMIUM AND RUTHENIUM CARBONYL
CLUSTERS
OH SUAT PING
B.Sc. (Hons.), NUS
A THESIS SUBMITTED FOR THE DEGREE OF
MASTER OF SCIENCE
DEPARTMENT OF CHEMISTRY
NATIONAL UNIVERSITY OF SINGAPORE
Acknowledgements
I would like to take this opportunity to express my heartfelt gratitude to all that has
helped me in one way or another throughout this project.
I would like to thank my supervisor, A/P Fan Wai Yip, for his understanding and
invaluable advice on my project. Next, I would like to thank my co-supervisor, A/P
Leong Weng Kee, for his tolerance and invaluable guidance throughout the project.
I am also grateful to all the members of the group. In particular, I would like to
express my gratitude to Xuelin and Chang Hong for their encouragement and guidance,
Garvin for imparting me invaluable research skills, Kien Voon, Wen Ling, Kai Ning,
Boon Ying, Huifang, Xiao feng, Dr Rakesh and Felicia for their help, support and fruitful
discussions.
Special thanks to all the staff in the instrument labs of NUS and NTU for making
data acquisition possible.
Last but not the least, I would like to thank my family and friends for their
encouragement and moral support.
i
Table of Contents
Acknowledgements
i
Table of contents
ii
Abstract
v
Molecular numbering scheme
vi
Abbreviations
xi
Lists of tables
xii
Lists of figures
xiv
Lists of schemes
xv
Chapter 1
p.1
Introduction to nanodiamonds, diamondoids, osmium
and ruthenium carbonyl clusters
1.1 Nanodiamonds
p.1
1.2 Diamondoids
p.2
1.3 Chemistry of osmium and ruthenium carbonyl clusters
p.3
1.4 Aim and objectives of the work
p.5
1.5 References
p.6
Chapter 2
Osmium and ruthenium carbonyl cluster derivatives of
nanodiamonds
2.1 Osmium derivatized nanodiamonds
p.8
p.8
2.2 Ruthenium derivatized nanodiamonds
p.14
2.3 Aggregation of osmium and ruthenium derivatized nanodiamonds
p.16
2.4 Conclusion
p.17
2.5 References
p.18
ii
Chapter 3
Molecular model of the cluster-nanodiamond
interaction: The reaction of Ru3(CO)12 with 1adamantanecarboxylic acid
p.19
3.1 Purple complex [Ru2(CO)4(OOCAd)2]n (6)
p.20
3.2 Novel tetranuclear chain complex
p.28
3.3 Conclusion
p.32
3.4 References
p.33
Chapter 4
Catalytic studies on hydrogenation of quinoline
p.34
4.1 Hydrogenation of quinoline
p.34
4.2 Catalytic activities of complexes MH(CO)(!3-OCOR)(PPh3)2
p.35
4.3 Conclusion
p.37
4.4 References
p.38
Chapter 5
Conclusion
p.39
Chapter 6
Experimental
p.40
6.1 General experimental
p.40
6.2 Reaction with nanodiamonds
p.41
6.3 Thermolysis of osmium- and ruthenium-derivatized
nanodiamonds
p.41
6.4 Synthesis of Os3(CO)10(µ-H)(µ-OOCAd), 4
p.43
6.5 Thermolysis of Os3(CO)10(µ-H)(µ-OOCAd), 4
p.43
6.6 Synthesis of ruthenium polymer, 6
p.43
6.7 Synthesis of Ru2(CO)4(OOCAd)2(PPh3)2, 7
p.44
6.8 Synthesis of Ru2(CO)4(OOCAd)2(TPPTs)2, 8
p.45
6.9 Synthesis of Ru4(CO)8(OOCAd)4(PPh3)2, 9
p.45
6.10 Synthesis of higher nuclearity ruthenium chain complexes
p.46
6.11 Synthesis of Ru2(CO)6(OOCAd)2, 10
p.46
6.12 Reaction of 6 with NCCH2CH3, NEt3, Cp*Ir(CO)2 and tBuNC
p.46
iii
6.13 Synthesis of RuH(CO)(!3-OCOR)(PPh3)2 where R = Ad (12b),
ND (12c)
p.47
6.14 Synthesis of [OsH(CO)(NCCH3)2(PPh3)2][BF4], 15
p.48
6.15 Synthesis of OsH(CO)(!3-OCOR)(PPh3)2 where R = Ad (16b),
ND (16c)
p.48
6.16 Details of catalytic runs
p.49
6.17 X-ray crystallography studies of 4, 7 and 9
p.50
6.18 References
p.52
Appendices
CD-ROM
iv
Abstract
The derivatization of nanodiamonds (NDs) with osmium and ruthenium clusters has
been carried out via the reaction of various clusters with unmodified and acidfunctionalized NDs. These have been characterized by IR and mass spectrometry, as
well as elemental analyses. An attempt to model the reactions with the ND surface
was made using 1-adamantanecarboxylic acid. It was found that the reaction of
Ru3(CO)12 (5) with 1-adamantanecarboxylic acid (AdCOOH) afforded a purple
polymer [Ru2(CO)4(OOCAd)2]n (6) which dissolved reversibly in coordinating
solvents (L) to form dinuclear adducts Ru2(CO)4(OOCR)2L2. The reaction of 6 with
PPh3 in a 1:0.5 Ru:P ratio afforded a novel, electron deficient, tetraruthenium chain
complex Ru4(CO)8(OOCAd)4(PPh3)2 (9).
v
MOLECULAR NUMBERING SCHEME
The short line extending from Os and Ru (in the molecular structure diagrams)
represents a coordinative bond from carbon monoxide to osmium (Os-CO) and
ruthenium (Ru-CO).
1.
!
2.
!
3.
!
4.
!
vi
5.
6.
!
7.
!
8.
vii
9.
!
10.
!
11.
!
12.
!
R = Ph (12a), Ad (12b), ND (12c) (Ad = 1-adamantyl)
13.
!
viii
14.
!
15.
!
16.
!
R = Ph (16a), Ad (16b), ND (16c)
ix
For simplicity, we assigned the following nanodiamonds with metal clusters as
O1.
O2.
O3.
O4.
O5.
O6.
O7.
ND-COOH with cluster 1
ND-COOH with cluster 2
ND-COOH with cluster 3
ND with cluster 1
ND with cluster 2
ND with cluster 3
ND with cluster 1 at 85ºC
R1.
R2.
R3.
R4.
ND with cluster 5
ND with cluster 5 at 85ºC
ND-COOH with cluster 5
ND-COOH with cluster 5 at 85ºC
x
ABBREVIATIONS
Ad
CO
Cp*
Cy
DCM
DFT
EA
Et
ESI
FAB
GC
HOMO
HRMS
IR
KBr
LC
MeCN
MeOH
MS
ND
ND-COOH
NMR
ORTEP
Q
Rbf
Rf
RT
THF
THQ
TLC
TMNO
ToF-SIMS
TPPTs
Adamantyl, C10H15
Carbon monoxide
!-Pentamethylcyclopentadienyl
Cyclohexyl
Dichloromethane
Density Functional Theory
Elemental Analyses
Ethyl (CH2CH3)
Electrospray Ionization
Fast Atom Bombardment
Gas Chromatography
Highest Occupied Molecular Orbital
High Resolution Mass Spectrometry
Infrared
Potassium Bromide
Liquid Chromatography
Acetonitrile
Methanol
Mass Spectrometry
Nanodiamonds
Acid Functionalised Nanodiamonds
Nuclear Magnetic Resonance
Oak Ridge Thermal Ellipsoid Plot
Quinoline
Round-bottomed Flask
Retention factor
Room Temperature
Tetrahydrofuran
tetrahydroquinoline
Thin Layer Chromatography
Trimethylamine oxide
Time-of-flight Secondary Ion Mass Spectrometry
Triphenylphosphine-3,3’,3”-trisulfonic acid trisodium salt
xi
Lists of Tables
Table 2.1
Elemental analyses of ruthenium contents.
p.14
Table 4.1
Percentage conversion to THQ.
p.36
Table 4.2
% Element calculated.
p.37
Table 6.1
Solid-state IR data for clusters.
p.42
Table 6.2
Crystal data and structure refinement.
p.51
xii
Lists of Figures
Fig 1.1
Members of Diamondoids.
p.3
Fig 2.1
Molecular structure of Os3(CO)10(µ-H)(µ-OOCAd) (4).
p.10
Fig 2.2
Solid-state IR spectra of O2, 2 and 4.
p.11
Fig 2.3
Solid-state IR spectra: (a) ND-COOH samples O1-3 and (b)
ND samples O4-7.
p.12
Fig 2.4
Solid-state IR spectra in the !CO region (cm-1) : (a) Cluster 1
on MgO after heating at 150 ºC for 16 h under argon;3 (b)
ND-COOH samples O1-3.
p.13
Fig 2.5
ToF-SIMS (positive ion mode) spectrum of O4
p.14
Fig 2.6
Solid-state IR spectra: (a) derivatized ND and (b) derivatized
ND-COOH.
p.16
Fig 2.7
ToF-SIMS spectra (positive ion mode) of R2
p.16
Fig 2.8
Dispersion of derivatized NDs in DCM: (top) freshly
prepared and (bottom) after standing for 1 day.
p.17
Fig 3.1
Molecular structure of Ru2(CO)4(OOCF3)2(NCMe)2.
p.19
Fig 3.2
Structure of a fragment of [Ru2(CO)4(OOCCF3)2]n; the
fluorine atoms have been omitted for clarity.
p.20
Fig 3.3
Solid-state IR spectra of R4 (a) and 6 (b).
p.21
Fig 3.4
IR spectra of 6 in various solvents.
p.22
Fig 3.5
Molecular structure of Ru2(CO)4(OOCAd)2(PPh3)2 (7).
p.23
Fig 3.6
IR spectra of 8 in KBr (a) and 7 in DCM (b).
p.24
Fig 3.7
IR spectra of 6 in tBuNC (a), NEt3 (b) and NCEt (c).
p.25
Fig 3.8
13C{1H}
NMR (C6D6) spectra of 7 (a), 6 with NEt3 (b) and 6
p.25
NMR spectra of 6 with tBuNC in C6D6.
p.26
(c).
Fig 3.9
13C{1H}
Fig 3.10 Proposed structure of 6 with tBuNC.
p.26
Fig 3.11 IR spectra of 6 with Cp*Ir(CO)2 (a) and with NEt3 (b).
p.27
Fig 3.12 Molecular structure of [Ru2(CO)5(OOCCF3)2]2 (top) and
Ru4(CO)8(OOCC10H15)4(PPh3)2 (9) (bottom).
p.29
xiii
Fig 3.13 HOMO (top) and LUMO (bottom) calculated for
Ru4(CO)8(OOCH3)4(PPh3)2.
p.30
Fig 3.14 UV spectra of solutions of 6, 7 and 9 in DCM.
p.31
Fig 3.15 Molecular model of [Ru2(CO)4(OOCCF3)2].
p.32
Fig 4.1
p.34
Potential drugs from THQ.
xiv
Lists of Schemes
Scheme 1.1 Synthesis of 2 and 3 from Os3(CO)12 (1).
p.4
Scheme 1.2 Proposed scheme for the reaction of 2 with carboxylic acids.
p.5
Scheme 2.1 Reaction of cluster precursors with nanodiamonds and
AdCOOH.
p.9
Scheme 3.1 Reaction of 9 with PPh3.
p.28
Scheme 4.1 Synthesis of MH(CO)(!3-OCOR)(PPh3)2.
p.35
Scheme 4.2 Hydrogenation of quinoline.
p.36
xv
Chapter 1: Introduction to Nanodiamonds, Diamondoids, Osmium and
Ruthenium Carbonyl Clusters
1.1 Nanodiamonds
As nanoscience and nanotechnology advances, various carbonaceous materials such
as fullerenes, carbon nanotubes and nanodiamonds, are being used extensively due to
their unique mechanical, electrical and thermal properties. Particularly, nanodiamonds
(NDs) are gaining a lot of interests from researchers. They are mostly synthesized by
microexplosions,1 using trinitrotoluene or a mixture of explosives, followed by
isolation and purification.2 According to the core theory,3 each nanodiamond is a
supramolecule with a single-crystal diamond core coated with functional groups
which are chemically bound to the core and hence, determine the chemistry of the
surface of the nanodiamond.
The surface structure of NDs mainly depends on the fabrication conditions and can
contain oxygen-containing (hydroxyl, carbonyl, ether, anhydride, etc.), nitrogencontaining (amine, amide, cyano, nitro, etc.), sulfone, or other functional groups.3 It
was also reported that the ND surface contains methyl and methylene groups in
minute amounts.3 Furthermore, the ND surface can be chemically modified. Thus,
chemically modified NDs (designated as ND-X) such as ND-COOH, ND-OH,
ND-NH2 and ND-Br are among those commercially available.
Surface-modified NDs have a good adsorption capacity and they are considered for
use as carriers of drugs, enzymes and isotopes . They are also used as a component in
several assay systems in immunological and biochemical laboratories.4 Carboxylated/
oxidized ND is a good solid phase support for the extraction of proteins and peptides
!
1
in highly dilute solution.5,6 Polylysine-coated ND powders also serve well as an
enrichment device for DNA oligonucleotides.7 It was also reported that linear and
ring forms of DNA do interact with NDs.8 In addition, NDs are also catalytically
active in oxidation processes, for example, in the oxidation of CO to CO2 as its
surface is easily saturated with oxygen,9 as well as in organic reactions such as the
reaction of oxidative diazotization (hydrogen peroxide-1-AAP-phenol).10
One major difficulty faced in working with NDs is that they are prone to aggregation
due to the large number of functional groups present on the ND surface.3 The type of
functional groups present substantially affects the colloidal properties of the NDs;
aggregation occurs when they are suspended in different media during synthesis and
treatment.11 A few methods for the disaggregation and dispersion of NDs have been
reported, such as, ultrasonic and surface modification using inorganic electrolytes,12
graphization-oxidation method and mechano-chemical treatments for NDs in aqueous
media,13 and detergent stabilization.14 However, it is still quite difficult to completely
disintegrate NDs back to a particle size of nanometers. Thus, further studies of
disagglomeration and dispersion are required.
1.2 Diamondoids
The term diamondoid (“diamantoid”, German) was first used by Decker in 1924 in
his attempts to synthesize diamonds.15 In 1933, Kleinfeller and Frercks predicted
adamantane as the lowest diamondoid.16 Diamondoids are also known as
nanodiamonds or condensed adamantanes. It consists of one or more cages of carbon
atoms, which have the same structure as bulk diamond, with terminal hydrogen atom.
Examples are adamantane (C10H16), which consists of a single cage of carbon atoms;
!
2
diamantane (C14H20), which consists of two cages, and triamantane (C18H24); and
tetramantane (C22H28), which consist of three and four cages, respectively (Fig. 1.1).
Fig. 1.1 Members of Diamondoids
(Adapted and modified from http://en.wikipedia.org/wiki/Diamondoid)
The diamondoids are difficult to be synthesized in the laboratory, except for
adamantane. As the smallest diamondoid that has only one kind of secondary and
tertiary carbon atoms, the chemistry of adamantane has been widely studied. Its
functionalization is thus much easier compared with the higher diamondoids. Recent
mechanistic studies of adamantane assisted in rationalizing and predicting the
behaviour of the higher diamondoids in the presence of radical, electrophilic and
oxidative reagents.17
1.3 Chemistry of osmium and ruthenium carbonyl clusters
As mentioned above, a major problem with NDs is aggregation due to the functional
groups on its surface. One of the reported methods to disaggregate NDs is by surface
modification using inorganic electrolytes.12 Hence, we were interested in the
possibility of using a more bulky modifier such as metal cluster, to allow for better
disaggregation and dispersion of the NDs. In particular, triosmium carbonyl clusters
are hydrophobic and chemically fairly inert, which we hope will make them suitable
for better ND dispersion, especially inorganic media. Furthermore, these cluster
modifiers may also act as spectroscopic tags or probes.
!
3
Metal carbonyl clusters are compounds with metal-metal bonds, with carbon
monoxide as stabilizing ligands. The carbonyl (CO) ligands exhibit strong absorption
bands in the mid-IR region, from 1800 - 2200 cm-1, where most organic functional
groups do not.18 This property of the clusters has found use for them as IR imaging
tags. The parent triosmium dodecacarbonyltriosmium (0), Os3(CO)12 (1), is rather
inert, making it unfriendly as a precursor in reactions;19 high temperatures are usually
required for ligand substitution. Generally, the more labile derivatives
Os3(CO)10(NCCH3)2 (2) and Os3(CO)11(NCCH3) (3) are used (Scheme 1.1).
!
Scheme 1.1 Synthesis of 2 and 3 from Os3(CO)12 (1)
On the other hand, triosmium dodecacarbonyltriruthenium (0), Ru3(CO)12 (5), is more
reactive and similar labile derivatives, Ru3(CO)10(NCCH3)2 and Ru3(CO)11(NCCH3)
can be prepared as shown in Scheme 1.1 at lower temperatures.20
As CH3CN is a labile ligand, any stronger nucleophile like PPh3 would easily displace
it. Among relevant chemistry of these clusters is their known relativity with O!
4
containing functional groups. For example, 2 reacts with carboxylic acids, RCOOH,
to give the derivatives Os3(CO)10(µ-H)(µ-OOCR) through an intermediate species,
Os3(CO)10(µ-H)(!1-OOCR) as shown in Scheme 1.2.21 Similarly, 3 react with
RCOOH to afford derivatives Os3(CO)11(µ-H)(!1-OOCR).22 However, carboxylic
acids do not react with ruthenium derivatives in a similar way. For instance, they react
with 5 to afford dinuclear species; this is discussed in chapter 3.
Scheme 1.2 Proposed scheme for the reaction of 2 with carboxylic acids.
Besides carboxylic groups, 2 is also reactive to alcohol groups (ROH) and aldehydes
(RCHO) to afford the derivatives Os3(CO)10(µ-OR)2 and Os3(CO)10(µ-H)(COR),
respectively.23,24
1.4 Aim and objectives of the work
The aim of this project was to derivatize nanodiamonds with osmium clusters. In
particular, we sought to investigate the following:
1.
The effect of attaching bulky osmium and ruthenium clusters onto the ND
surface and their effect on the surface properties and surface chemistry of
the NDs.
2.
The employment of osmium and ruthenium carbonyl clusters as a mid IR
probe for the functional groups present on the ND surface.
3.
Use of adamantane as an inexpensive model for these reactions on the ND
surface.
!
5
1.5 References
1.
Titov VM, Tolochko BP, Ten KA, Lukyanchikov LA, Pruuel ER, Diamond
Relat. Mater., 2007, 16, 2009
2.
Pichot V, Comet M, Fousson E, Baras C, Senger A, Le Normand F, Spitzer D,
Diamond Relat. Mater., 2008, 17, 13
3.
Kulakova I.I., Physics of the Solid State, 2004, 46, 636
4.
Puzyr AP, Baron AV, Purtov KV, Bortnikov EV, Skobelev NN, Mogilnaya OA,
Bondar VA., Diamond Relat. Mater., 2007, 16, 2124
5.
Kong X.L., Huang L.-C.L., Hsu C.-M., Chen W.-H., Han C.-C., Chang H.-C.,
Anal. Chem., 2005, 77, 259
6.
Chen W.-H, Lee S.-C, Sabu S., Fang H.-C., Chung S.-C., Han C.-C., Chang H.C., Anal Chem., 2006, 78, 4228
7.
Kong X.L., Huang L.-C.L., Liau S.-C. V., Han C.-C., Chang H.-C., Anal.
Chem., 2005, 77, 4273
8.
Purtov KV, Burakova LP, Puzyr AP, Bondar VA, Nanotechnology, 2008, 19,
325101
9.
Bogatyreva G.P., Marinich M.A., Ishchenko E.V., Gvyazdovskaya V.L., Bazali"
G.A., Ole"nik N.A., Physics of the Solid State, 2004, 46,4, 738
10.
Bondar V.S., Purtov K.V., Puzyr A.P., Baron A.V., Gitel’zon I.I., Biochemistry
and Biophysics, 2008, 418, 11
11.
Jan Hou#ka, Nagender Reddy Panyala, Eladia Maria Peña-Méndez and Josef
Havel, Rapid Commun. Mass Spectrom., 2009, 23, 1125
!
12.
Chigangova GA., Colloid J., 1994, 56, 212
13.
Xu K, Xue QJ., Acta Phys. Chim. Sin., 2003, 19, 993
14.
Xu X, Yu Z, Zhu Y, Wang B. Diamond Relat. Mater., 2005, 14, 206
6
15.
Decker H., Z., Angew. Chem., 1924, 37, 795
16.
Kleinfeller H., Frercks W., J. Prakt. Chem. 1933, 138, 184
17.
Fokin A.A., Schreiner P.R., Schwertfeger H., Angew. Chem. Int. Ed., 2008, 47,
1022
18.
Salmain, M., Jaouen, G., Ed., In Bioorganometallics, Wiley-VCH Verlag GmbH
& Co., 2006, 181
19.
Johnson B.F.G., Lewis J., Pippard D., J. Organomet. Chem., 1978, 160, 263
20.
Foulds G.A., Johnson B.F.G, Lewis J., J. Organomet. Chem., 1985, 296, 147
21.
Ainscough E.W. et al., Journal of Organomet. Chem., 1998, 556, 197
22.
Frauenhoff G.R., Wilson S.R.,Shapley J.R., Inorg. Chem., 1991, 30, 78
23.
Arce A.J., Danctics Y.D., Deeming A.J., Journal of Organomet. Chem., 1986,
311, 371
24.
Johnson B.F.G., Lewis J., Odiaka T.I., Journal of Organomet. Chem., 1986,
307, 61
!
7
Chapter 2: Osmium and ruthenium carbonyl cluster derivatives of
nanodiamonds
There were two types of nanodiamonds used in this project, namely, unmodified (ND)
and acid-functionalized (ND-COOH) nanodiamonds. The reactions of these
nanodiamonds with osmium and ruthenium clusters, and the characterization of the
products, are presented in this chapter.
2.1 Osmium derivatized nanodiamonds
As mentioned in Chapter 1, the cluster Os3(CO)10(NCCH3)2 (2) reacts with carboxylic
acids to afford derivatives of the type Os3(CO)10(µ-H)(µ-OOCR). The reaction of
ND-COOH with 2 was compared to the analogous reaction with 1adamantanecarboxylic acid (AdCOOH) to serve as a model; similar reactions with the
other cluster precursors Os3(CO)12 (1) and Os3(CO)11(NCCH3) (3), and with
unmodified ND were also carried out (scheme 2.1). A higher reaction temperature
(85 oC, 6 h in toluene) allowed for more deposition of cluster, as shown by the much
higher osmium content (3.2 % for O7 compared to 0.05 % for O4).
!
8
Nanodiamond
Cluster
Derivatized nanodiamond
% Os
ND-COOH
1
O1
0.9
ND-COOH
2
O2
8.3
ND-COOH
3
O3
2.7
ND
1
O4
0.05
ND
2
O5
6.3
ND
3
O6
4.1
ND
1*
O7
3.2
* 85 oC for 6 h in toluene
Scheme 2.1 Reaction of cluster precursors with nanodiamonds and AdCOOH.
A single crystal X-ray structural analysis of 4 showed that it had the same structure as
that of other carboxylic acid analogues (Fig 2.1),1 consisting of a carboxylato-bridge
and a bridging hydride. The bond length for the Os(1)-Os(2) bond (2.8760(5) Å),
which was bridged by the carboxylato groups, was longer than that for the Os(1)Os(3) and Os(2)-Os(3) bonds (2.8673(5) Å and 2.8646(5) Å, respectively). This
lengthening of the carboxylato-bridged Os-Os bond had also been observed for the
other carboxylic acid analogues.1 The C-O bond lengths (1.245(11) Å and 1.265(11)
Å) indicated partial double bond character for both these bonds as expected bond
lengths of C-O single and C=O double were 1.36 and 1.23 Å, respectively.2
!
9
Fig 2.1 Molecular structure of Os3(CO)10(µ-H)(µ-OOCAd) (4). Selected bond lengths
(Å) and angles (o) are as follows: Os(1)-Os(2) = 2.8760(5), Os(1)-Os(3) = 2.8673(5),
Os(2)-Os(3) = 2.8646(5), Os(1)-O(1) = 2.214(6), Os(2)-O(2) = 2.117(5), C(1)-O(1) =
1.245(11), C(1)-O(2) = 1.265(11); Os(1)-Os(2)-Os(3) = 59.931(12), Os(1)-Os(3)Os(2) = 60.232(12), Os(2)-Os(1)-Os(3) = 59.837(11), O(1)-Os(1)-Os(3) = 90.23(17),
O(1)-Os(1)-Os(2) = 80.93(17), O(2)-Os(2)-Os(3) = 91.49(15), O(2)-Os(2)-Os(1) =
81.63(16), O(1)-C(1)-O(2) = 125.8(8).
The solid-state IR spectra of O2 and 4 suggested that O2 did not contain an
Os3(CO)10(µ-H)(µ-OOCR)-type moiety although the distinctive carbonyl pattern in
the mid IR region (1800 - 2200 cm-1) showed the presence of osmium carbonyl
cluster on the nanodiamond (Fig 2.2).
!
10
Fig 2.2 Solid-state IR spectra of O2, 2 and 4.
The product from analogous reactions of ND-COOH with the clusters Os3(CO)12 (1)
and Os3(CO)11(NCCH3) (3), viz., O1 and O3, respectively, showed similar solid-state
IR patterns to that of O2, except that the intensities for O1 were lower. This may be
due to the inertness of 1, resulting in a smaller amount of the cluster being anchored
onto the nanodiamond. This was supported by the elemental analyses (Scheme 2.1);
the osmium content increases in the order O1 < O3 < O2, in the order of increasing
reactivity of the cluster precursors 1 < 3 < 2.
Surprisingly, unmodified nanodiamonds (ND) reacted similarly with 1, 2 and 3, to
afford samples which showed solid-state IR spectra similar in pattern to those of
O1-3, but shifted to lower wavenumbers (Fig 2.3). The peaks at 1794 (w) and 1629
(m) cm-1 are assignable to the carbonyl- and hydroxyl-containing groups on the
surface of underivatized ND-COOH;3 similar peaks are observed at 1734 (vw) and
1628 (m) cm-1 in the solid-state IR spectrum of underivatized ND.3 This suggested
!
11
that the osmium clusters did not react with the carbonyl and hydroxyl groups on the
ND surface.
Fig 2.3 Solid-state IR spectra: (a) ND-COOH samples O1-3 and (b) ND samples
O4-7.
The IR data suggested that the surface species were all similar, but they did not
appear to be of the carboxylato-bridged Os3(CO)10(µ-H)(µ-OOCR) type. Instead, the
IR pattern matched the reported values for the species obtained from the reaction of
Os3(CO)10(µ-H)(µ-OH) with the ZnO and In2O3 surfaces,4 that has been ascribed to
mononuclear di- and tricarbonyl osmium species (Os(CO)2 or Os(CO)3 units) (Fig
2.4).5 It was reported that upon exposure to air, the chemisorbed triosmium carbonyls
fragmented to form a mixture of osmium di- and tricarbonyl species.6 It was also
found that these osmium species occurred in groupings consisting of approximately
three osmium atoms, with no Os-Os bond. It has been suggested that the osmium
clusters interacted directly with the support and subsequently underwent oxidative
fragmentation that left the osmium atoms in nearly their original position on the
surface.6 It is therefore possible that similar fragmentation may have occurred on the
ND surface.
!
12
Fig 2.4 Solid-state IR spectra in the $CO region (cm-1) : (a) Cluster 1 on MgO after
heating at 150 ºC for 16 h under argon;3 (b) ND-COOH samples O1-3.
The ToF-SIMS spectra revealed the presence of both osmium di- and tricarbonyl
species, and clusters species (Fig 2.5). Since it was mentioned earlier on that
oxidative fragmentation left the osmium atoms in their original position on the
surface, the presence of cluster species maybe due to the ensembles of the Os(CO)2
and Os(CO)3 units during ToF-SIMs. Generally, the ToF-SIMS spectra of the
compounds exhibited a broad distribution of peaks. Clusters of peaks with relatively
higher intensities than the background at m/z 769, 741, 713 and 687, were assignable
to [Os3(CO)7]+, [Os3(CO)6]+, [Os3(CO)5]+ and [Os3(CO)4]+, respectively. Another
sequential loss of CO (m/z 755, 728 and 701) suggested the presence of a bonding
mode involving a nitrogen atom or a CH2 moiety. This was also corroborated by
peaks in the lower mass range, at m/z 283, 255 and 227. Three broad peaks in the
solid-state IR spectrum (1114, 1283 and 1400 cm-1) could be assigned to NO2,
SO2OH and the OH of a carboxylic group,3 which suggested the possibility of
reaction with nitro groups present on the ND surface.
!
13
Fig 2.5 ToF-SIMS (positive ion mode) spectrum of O4
2.2 Ruthenium derivatized nanodiamonds
Derivatization of the nanodiamonds with the ruthenium cluster Ru3(CO)12 (5), using
identical conditions as for the osmium clusters above, afforded products that showed
the presence of ruthenium; a higher temperature again allowed for more deposition of
ruthenium onto the ND surface (Table 2.1).
Table 2.1 Elemental analyses of ruthenium contents.
Nanodiamond
Cluster
Derivatized nanodiamond
% Os
ND
5
R1
0.9
ND
5*
R2
10.2
ND-COOH
5
R3
1.0
ND-COOH
5*
R4
10.4
* 85 oC for 6 h in toluene
The IR spectra of ND products (R1 and R2) and ND-COOH products (R3 and R4)
were different (Fig 2.6). Those for the ND-COOH products resembled that of the
osmium-derivatized NDs; three peaks at 2131 (m), 2064 (s) and 1998 (m) cm-1 can be
assigned to Ru(CO)3 units, cf 2116 (m), 2059 (s) and 2009 (m) cm-1 for
[Ru(CO)3I2]2,7 though there were shifts in wavenumbers. On the other hand, the IR
spectra of the ND products showed two peaks at 2053 (s) and 1981 (s) cm-1, with a
!
14
shoulder at 1925 cm-1; this could be assigned to Ru(CO)2 units, cf 2050 (s), 1995 (s)
and 1975 (w, sh) cm-1 for [Ru(CO)2I2]n.7 Cluster 5 is known to be more unstable. This
was reflected in the observation that the ruthenium species deposited on the ND
surface decomposed over time, affording no peaks in the IR spectra after a few
months. Similar to the osmium analogues, the ToF-SIMS spectra (Fig 2.7) indicated
the presence of both ruthenium di- and tricarbonyl species, and ruthenium clusters on
the ND surface. The peaks at m/z = 590, 600, 637, 705 and 733, could be assigned to
[Ru3(CO)10]+, [Ru3(CO)10(µ-O)(µ-H)]+, [Ru3(CO)12]+, [Ru4(CO)11]+ and [Ru4(CO)12]+,
respectively. Fragments of higher nuclearity clusters were also observed, in contrast
to that for the osmium derivatives. Furthermore, the presence of [Ru3(CO)10(µ-O)(µH)]+ may indicate interaction of 5 with hydroxyl groups on the surface of the NDs.8
Thus it is likely that as in the case of the derivatization of nanodiamonds with 5, the
ruthenium atoms were bonded directly to the ND surface. Oxidative fragmentation, to
afford a mixture of Ru(CO)2 (in derivatized ND) and Ru(CO)3 (in derivatized NdCOOH) species, took place almost immediately.
Fig 2.6 Solid-state IR spectra: (a) derivatized ND and (b) derivatized ND-COOH.
!
15
Fig 2.7 ToF-SIMS spectra (positive ion mode) of R2
2.3 Aggregation of osmium and ruthenium derivatized nanodiamonds
As mentioned in chapter 1, one difficulty faced in working with ND is the problem of
aggregation. Generally, the aggregation is through Van der Waals interactions, which
is stronger when the particle size is smaller. In addition, functional groups such as
carboxylic acids can form interparticle hydrogen bonds.9,10 When ND aggregates,
they form bigger particles and they no longer disperse well in solutions. Therefore,
ND-COOH is naturally well dispersed in water due to its carboxylic acid group, but
ND would form aggregates in water as less hydrogen bonding are found on its
surface.
The ND samples were dispersed in DCM and allowed to stand (Fig 2.8).
Underivatized ND was more well dispersed than ND-COOH in the first few hours but
after one day, both were found to settle. The osmium derivatized NDs were found to
!
16
disperse better than underivatized ND and ND-COOH. For ruthenium derivatized
NDs, those derivatized at a high reaction temperature (R2 and R4) were found to
disperse better than those derivatized at room temperature (R1 and R3). This may be
due to more ruthenium anchored on the ND surface in the former, leading to greater
disruption of the Van der Waals interactions between ND.
Fig 2.8 Dispersion of derivatized NDs in DCM: (top) freshly prepared and (bottom)
after standing for 1 day.
2.4 Conclusion
We have successfully derivatized nanodiamonds with osmium carbonyl clusters as
shown in their IR spectra, ToF-SIMS and EA analyses. High reaction temperature
resulted in more metal species deposited on the ND surface. When nanodiamonds
reacted with osmium clusters, the triosmium core was retained and the osmium atoms
bonded directly to the ND surface. However, oxidative fragmentation took place
almost immediately, affording a mixture of osmium di- and tricarbonyl species. A
similar behaviour was observed in the derivatization of nanodiamonds with ruthenium
carbonyl clusters.
The osmium and ruthenium derivatized NDs were found to
aggregate less rapidly as compared to before derivatization.
!
17
2.5 References
1.
Johnson B.F.G., Lewis J., Pippard D., J. Organomet. Chem., 1978, 160, 263
2.
MacGillavry C.H., Rieck G.D., International Tables for X-ray Crystallography,
1962, vol 3, Kynoch, Birmingham
3.
Kulakova I.I., Physics of the Solid State, 2004, 46, 636
4.
Li C., Leong W.K., Journal of Colloid and Interface Science, 2008, 328, 29
5.
Psaro R., Dossi C., Ugo R., J. Mol. Catal., 1983, 21, 331
6.
Bhirud V.A., Iddir H., Browning N.D., Gates B.C., J. Phys. Chem. B., 2005,
109, 12738
7.
Psaro R., Ugo R., Zanderighi G.M., J. Organomet. Chem., 1981, 213, 215
8.
Zecchina A., Guglielminotti E., Bossi A., Camia M., J. Catal., 1982, 225
9.
Krüger A., Kataoka F., Ozawa M., Fujino T., Suzuki Y., Aleksen-skii A.E., Vul’
A. Yu., Õsawa E., Carbon, 2005, 43, 1722
10.
!
Xu X., Yu Z., Zhu Y., Wang B., J. Solid State Chem., 2005, 178, 688
18
Chapter 3: Molecular model of the cluster-nanodiamond interaction: The
reaction of Ru3(CO)12 with 1-adamantanecarboxylic acid
In an attempt to identify the type of interaction of the ruthenium clusters with the ND
surface for samples R1-4, we investigated the reaction of Ru3(CO)12 (5) with 1adamantanecarboxylic acid. The reaction of 5 with carboxylic acids to form
polymeric [Ru(CO)2(OOCR)]n was first reported by Lewis and coworkers.1 They
found that these polymers dissolved reversibly in coordinating solvents (L), such as
acetonitrile and tetrahydrofuran, to form the dinuclear complexes
Ru 2 (CO) 4 (OOCR) 2 L 2 . The single crystal X-ray structural analysis of
Ru2(CO)4(OOCF3)2(NCMe)2, for example, has been reported and is shown in Fig
3.1.2 It comprises a diruthenium unit bridged by two mutually cis carboxylate groups.
Upon removal of the solvent, Ru2(CO)4(OOCR)2L2 would lose the coordinated
molecules L to reform polymeric [Ru(CO)2(OOCR)]n. Phosphine derivatives can also
be obtained similarly, and they have a similar structure.3
Fig 3.1 Molecular structure of Ru2(CO)4(OOCF3)2(NCMe)2.
The molecular structure of the polymeric [Ru(CO)2(OOCR)]n complexes has also
been revealed by single crystal X-ray structured analyses, for example, for R = CF3
!
19
(Fig 3.2).4 This comprised dinuclear [Ru2(CO)4(OOCCF3)2] units held together by
strong interactions between an O atom of the carboxylato bridge and an Ru atom of
the neighbouring unit.
Fig 3.2 Structure of a fragment of [Ru2(CO)4(OOCCF3)2]n; the fluorine atoms have
been omitted for clarity.
The synthesis of polymeric [Ru(CO)2(OOCR)]n complexes included: (a) refluxing 5
in the neat carboxylic acid, (b) via the labile intermediates Ru2(CO)4(OOCR)2(THF)2
by reaction in tetrahydrofuran (THF) as solvent at 120 ºC,1 and (c) in refluxing
methanol.5
3.1 Purple complex [Ru2(CO)4(OOCAd)2]n (6)
Refluxing 5 with 1-adamantanecarboxylic acid in MeCN gave a light yellow solution.
Upon removal of solvent, a purple solid (6) was obtained, which was in sharp contrast
to the yellow colour reported for analogues such as [Ru(CO)2(OOCMe)2]n.1
Comparing the solid-state IR spectra of R4 and 6 (Fig 3.3), it was clear that the
species on the surface of ruthenium derivatized NDs did not have the structure of 6.
!
20
Fig 3.3 Solid-state IR spectra of R4 (a) and 6 (b).
Complex 6 dissolved in DCM and various solvents to give an orange yellow solution,
and reverted back to a purple solid upon removal of the solvents. The IR spectra of
these solutions were also similar (Fig 3.4). It was noticed that a solution in MeCN
required a more thorough removal of the solvent before the initially obtained yellow
residue reverted back to purple. These observations suggested the reversible
formation of dinuclear adducts such as Ru2(CO)4(OOCAd)2(MeCN)2 (Ad =
adamantyl); a stronger ligand like MeCN was harder to dissociate.
!
21
Fig 3.4 IR spectra of 6 in various solvents.
The identity of 6 was deduced from its reactivity towards coordinating solvents and
PPh3. A yellow complex Ru2(CO)4(OOCAd)2(PPh3)2 (7) was formed when 6 reacted
with excess PPh3. Unlike 6 which turned brown upon several days of exposure to air,
7 is air stable. The single crystal X-ray structural analysis of 7 showed that the PPh3
occupied the axial positions (Fig 3.5). The Ru2(CO)4 moiety resembles a sawhorse
geometry, which is similar to other analogues such as that shown in Fig 3.1. The RuRu distance is 2.71827(16) Å, which is in accordance with a metal-metal single bond,
and 7 has a total valence electron count of 36 expected for a diruthenium complex.
!
22
Fig 3.5 Molecular structure of Ru2(CO)4(OOCAd)2(PPh3)2 (7). Selected bond lengths
(Å) and angles (o) are as follows: Ru(1)-Ru(2) = 2.71827(16), Ru(1)-P(1) =
2.4295(4), Ru(2)-P(2) = 2.4502(4), Ru(1)-O(3) = 2.1099(13), Ru(1)-O(4) =
2.1427(12), Ru(2)-O(7) = 2.1370(12), Ru(2)-O(8) = 2.1218(12); Ru(2)-Ru(1)-O(3) =
83.10(3), Ru(2)-Ru(1)-O(4) = 81.76(3), Ru(1)-Ru(2)-O(7) = 82.07(3), Ru(1)-Ru(2)O(8) = 83.57(3), Ru(2)-Ru(1)-P(1) = 171.486(12), Ru(1)-Ru(2)-P(2) = 168.024(12),
O(3)-Ru(1)-O(4) = 89.22(3), O(7)-Ru(2)-O(8) = 84.46(5).
Complex 6 also reacted with water-soluble triphenylphosphine-3,3’,3”-trisulfonic
acid trisodium salt (TPPTs) to yield a water soluble yellow complex
Ru2(CO)4(OOCAd)2(TPPTs)2 (8). The molecular structure of 8 is probably similar to
that of 7 since TPPTs is a bulky ligand which would preferably substitute at the axial
position of the Ru2(CO)4 moiety, as suggested by the similarity between their IR
spectra (Fig 3.6).
!
23
Fig 3.6 IR spectra of 8 in KBr (a) and 7 in DCM (b).
The above observations suggest that the molecular formula of 6 is of the type
[Ru2(CO)4(OOCAd)2]n. The reaction of 6 with a number of weaker two-electron
donors was also investigated. Thus its reaction with neat propionitrile (NCEt),
triethylamine (NEt3) and tert-butyl isocyanide (tBuNC) afforded yellow solutions.
Their IR spectra resembled that of 7, indicating the presence of
Ru2(CO)4(OOCAd)2L2 where L = NCEt, NEt3 and tBuNC (Fig 3.7). However,
isolation of products was not feasible.
!
24
Fig 3.7 IR spectra of 6 in tBuNC (a), NEt3 (b) and NCEt (c).
The 13C{1H} NMR spectrum of 6 with NEt3 in C6D6 showed only one resonance at
around % 205 ppm, corresponding to CO (Fig 3.8), further supporting that the NEt3
was coordinated at the axial position of the Ru2(CO)4 moiety, just like in 7.
Fig 3.8 13C{1H} NMR (C6D6) spectra of 7 (a), 6 with NEt3 (b) and 6 (c).
!
25
However, the coordination of tBuNC seemed to differ; the
13C{1H}
NMR spectrum
(Fig 3.9) showed two resonances in the CO region, suggesting that tBuNC
coordinated at the equatorial positions instead; the resonances are at % 214.7 and
208.9 ppm are assigned to axial and radial carbonyls, respectively.6,7 However, there
are two possible isomers, which we have not been able to differentiate
spectroscopically (Fig 3.10).
Fig 3.9 13C{1H} NMR spectra of 6 with tBuNC in C6D6.
Fig 3.10 Proposed structure of 6 with tBuNC.
An attempt was also made at the synthesis of a bimetallic complex by the reaction of
6 with Cp*Ir(CO)2. Although a red solution was obtained, the IR spectra showed that
it was mainly a mixture of Ru2(CO)4(OOCAd)2L2 and unreacted Cp*Ir(CO)2 (Fig
3.11); an attempt at purification were unsuccessful.
!
26
Fig 3.11 IR spectra of 6 with Cp*Ir(CO)2 (a) and with NEt3 (b).
The colour changes observed with the addition of ligands suggested that 6 may be
potentially useful as a sensor. Purple films of 6, when subjected to solvent vapours
such as DCM and MeOH, turned yellow after a few minutes and reverted back to
purple upon removal of the solvents. A similar behaviour was observed upon
exposure to CO gas. It afforded a faint yellow complex Ru2(CO)6(OOCAd)2 (10), and
the reaction was reversible. The identity of 10 was proposed on the basis of the
similarity of its IR spectrum in the CO region (in DCM) with those of other acid
analogues; 2099 (s), 2078 (m), 2033 (vs), 2002 (vs), cf 2110 (s), 2080 (vs), 2038 (vs),
2000(vs) for Ru2(CO)6(OOCPh)2 and 2110 (vs), 2080 (vs), 2038 (vs), 2005 (vs) for
Ru2(CO)6(p-MeOC6H4COO)2.8 With a yellow solution of 6, bubbling CO through
turned it almost colourless. When the bubbling of CO was stopped, the colour
darkened back to yellow after a few minutes. Complex 10 was reported to decompose
!
27
spontaneously to polymers with the loss of CO,8 which would account for the colour
change.
3.2 Novel tetranuclear chain complex
On the assumption that 6 comprised [Ru2(CO)4(OOCAd)2] units, we hypothesized
that it would be possible to synthesize compounds containing two or more units of the
[Ru2(CO)4(OOCAd)2] moiety terminated with PPh3. Indeed, when 6 was reacted with
PPh3 in the ratio of 1: 0.5 (Ru:P),9 a light purple complex Ru4(CO)8(OOCAd)4(PPh3)2
(9) was obtained, together with a mixture of 6 and 7. Complex 9 dissolved in DCM
and MeCN to give yellow and light yellow solutions, respectively, and the process
was reversible. Further reaction with PPh3 afforded 7 (scheme 3.1).
Scheme 3.1 Reaction of 9 with PPh3.
Single crystal X-ray structural analysis of 9 showed that two dinuclear
[Ru2(CO)4(OOCAd)2] units are held together by a Ru-Ru bond instead of the reported
staggered Ru-O interactions (Fig 3.12).8 The total valence electron count based on the
molecular formula for 9 is 64, which requires four Ru-Ru bonds. As such, complex 9
is electron deficient. The Ru-Ru bond lengths 2.8754(7) Å and 2.6851(5) Å for
Ru(2)-Ru(2A) and Ru(1)-Ru(2), respectively, suggests a single metal-metal bond for
Ru(2)-Ru(2A) and a bond order of 1.5 for Ru(1)-Ru(2).
!
28
Fig 3.12 Molecular structure of [Ru 2 (CO) 5 (OOCCF 3 ) 2 ] 2 (top) and
Ru4(CO)8(OOCC10H15)4(PPh3)2 (9) (bottom). Selected bond lengths (Å) and angles (o)
for 9 are as follows: Ru(1)-Ru(2) = 2.6851(5), Ru(2)-Ru(2A) = 2.8754(7), Ru(1)-P(1)
= 2.3945(9), Ru(1)-O(31) = 2.138(3), Ru(1)-O(41) = 2.110(2), Ru(2)-O(32) =
2.076(3), Ru(2)-O(42) = 2.089(3); Ru(2)-Ru(1)-O(31) = 81.82(7), Ru(2)-Ru(1)-O(41)
= 82.41(7), Ru(2)-Ru(1)-P(1) = 171.44(3), O(31)-Ru(1)-O(41) = 83.64(11), O(32)Ru(2)-O(42) = 83.21(14), Ru(1)-Ru(2)-Ru(2A) = 162.510(18).
A DFT calculation on a simplified model of 9, in which the adamantyl was replaced
by a methyl, showed that the electron density of the HOMO (Highest Occupied
Molecular Orbital) was found mostly along the Ru-Ru vector within the dinuclear
[Ru2(CO)4(OOCAd)2] unit, i.e., between Ru(1)-Ru(2) (Fig 3.13). On the other hand,
the electron density of the LUMO (Lowest Unoccupied Molecular Orbital) was to be
found in between the dinuclear [Ru2(CO)4(OOCAd)2] units, i.e., between Ru(2)-
!
29
Ru(2A). This is consistent with nucleophiles attacking at the Ru(2)-Ru(2A) bond, as
found in its reactivity.
Fig 3.13 HOMO (top) and LUMO (bottom) calculated for
Ru4(CO)8(OOCH3)4(PPh3)2.
As mentioned above, 9 dissolved in a number of solvents (L) to give yellow solutions,
which are believed to have the formula Ru2(CO)4(OOCAd)2(PPh3)(L). Indeed, the
UV spectra of 7 and 9 were similar but different from that of 6 (Fig 3.14).
In
addition, the 31P{1H} NMR spectrum of 9 in MeCN showed two resonances, a major
resonance at % 8.5 ppm and a minor resonance at 13.7 ppm, which may be assigned to
9 and Ru2(CO)4(OOCAd)2(PPh3)(MeCN), respectively; the resonance for 7 is at %
14.3ppm.
!
30
Fig 3.14 UV spectra of solutions of 6, 7 and 9 in DCM.
Thus the molecular structure of 9 shows that dinuclear [Ru2(CO)4(OOCAd)2] units
held together by a Ru-Ru bond instead of the reported staggered Ru-O interactions,
imply that 6 is a polymeric compound consisting of such metal-metal bond linkages.
Using a model of [Ru2(CO)4(OOCCF3)2], which comprised staggered Ru-O
interactions,4 we replaced one of the CF3 groups with an adamantyl group (Fig 3.15)
and found that the adamantyl group would pose a steric problem. The shortest
distance between an O atom of a carbonyl on a neighbouring dinuclear
[Ru2(CO)4(OOCCF3)2] unit and the H atom of the adamantyl group was 1.2 Å. This is
to be compared with the Van der Waals radii of the O and H atoms (2.7 Å). It is
therefore plausible that the steric bulk of the Ad groups in 6 led to the formation of
Ru-Ru bonds instead of the staggered Ru-O interactions, which could account for the
purple colour of 6.
!
31
Fig 3.15 Molecular model of [Ru2(CO)4(OOCCF3)2].
3.3 Conclusion
The reaction of Ru3CO)12 (5) with 1-adamantanecarboxylic acid afforded the purple
complex 6, which is believed to be [Ru2(CO)4(OOCAd)2]n, a polymeric chain of RuRu bonds. This is different from the compounds obtained for other carboxylic acids.
Complex 6 changes colour from purple to yellow when exposed to solvents and CO
gas. Hence, it exhibits potential as a sensor. It reacted with PPh3 in different
stoichiometry to afford the novel, electron-deficient, tetranuclear chain complex 9 or
the dinuclear complex 7. The solid IR spectrum of 6 was different from that of R4,
obtained from the reaction of ND-COOH with 5, thus indicating that the species on
the ND surface was not structurally similar to 6.
!
32
3.4 References
1.
Crooks G.R., Johnson B.F.G., Lewis J., Williams I.G., Gamlen G., J. Chem Soc.
(A), 1969, 2761
2.
Bruce M.I., Skelton B.W., White A.H., Zaitseva N.N., Aust. J. Chem., 1999, 52,
621
3.
Bright T.A., Jones R.A., Nunn C.M., J. Coord. Chem., 1977, 128, 253
4.
Petrukihna M.A., Sevryugina Y., Andreini K.W., J. Cluster Sci., 2004, 15, 451
5.
Auzias M., Therrien B., Süss-Fink G., J. Organomet. Chem., 1989, 389, 139
6.
Dieter L., Robert M., Organometallics, 1991, 10, 1487
7.
Pierre G., Claude D.B., J. Organomet. Chem., 1996, 513, 155
8.
Michael R., Israel G., Uri S., Youval S., J. Organomet. Chem., 1986, 314, 185
9.
Mario B., Piero F., Ugo M., Gloria M., Franco P., Giorgio P., J. Organomet.
Chem., 1983, 259, 207
!
33
Chapter 4: Catalytic Studies on hydrogenaton of Quinoline
4.1 Hydrogenation of Quinoline
1,2,3,4-tetrahydroquinolines (THQ) are of interest due to their biological activities.
Several of these compounds are natural occurring, such as 2-methyl-1,2,3,4tetrahydroquinoline in human brain,1 or are used or tested as potential drugs, such as
osamniquine and viratmycin (Fig 4.1).1 Many THQ with various simple or
complex substituents exhibit interesting biochemical activity, for example,
2-methyl-5-hydroxy-1,2,3,4-tetrahydroquinoline has analgestic activity one eighth
as potent as morphine.1 Besides pharmaceutical applications, THQ derivatives are
also useful as pesticides, antioxidants and corrosion inhibitors.1
Fig 4.1 Potential drugs from THQ.
With the availability of appropriate quinolines (Q), direct reduction of the heterocylic
ring is still the best option to the synthesis of THQ. Hydrogenation over platinum is a
common approach,2-5 affording high yields when there are electron withdrawing
substituents on the heterocylic ring. Other metal catalysts include rhodium,6 and
recently, Borowski et al. have reported using [RuH2(!2-H2)2(PCy3)2] as the precatalyst
in the regioselective hydrogenation of the non-heterocyclic rings of Q.7
Compounds of the general formula MH(CO)(&3-OCOR)(PPh3)2 are catalytically
active in a number of industrially important chemical reactions such as
hydrodenitrogenation and hydrodesulfurization.8 They are readily obtainable from the
interaction of the Vaska and Diluzio’s complex RuHCl(CO)(PPh3)3,9 with carboxylic
!
34
acids. Examples of analogous osmium-complexes are rather scarce in the literature;
the complexes OsH(CO)(&3-OCOR)(PPh3)2, where R = CH3, CH3Cl, CH(CH3)2 or
C6H5, are reported to be only stable in argon for a few hours.10
We were interested in examining if ND and adamantane analogues of the MH(CO)
(&3-OCOR)(PPh3)2 complexes may be used as catalysts for the regioselective
hydrogenation of quinoline (Q) to 1,2,3,4-tetrahydroquinoline (THQ). The use of ND
as a catalyst support is attractive as it is expected to be fairly inert and it serves to
heterogenize the catalyst.
4.2 Catalytic activities of complexes MH(CO)(!3-OCOR)(PPh3)2
The new ruthenium and osmium complexes of ND and 1-adamantane were
successfully synthesized as shown in scheme 4.1. These osmium complexes were
stable for a few days in argon, as compared to the few hours reported for simpler
analogues.10
Scheme 4.1 Synthesis of MH(CO)(&3-OCOR)(PPh3)2.
!
35
The performance of both homogeneous (12a, 12b, 16a, 16b) and heterogeneous (12c,
16c) catalyst precursors for the regioselective hydrogenation of quinoline (Q) to
1,2,3,4-tetrahydroquinoline (THQ) under the conditions given in scheme 4.2, were
evaluated. The results are tabulated in Table 4.1.
Scheme 4.2 Hydrogenation of quinoline.
As can be seen, the ruthenium complexes showed better catalytic activities than the
corresponding osmium complexes. More significantly, although the adamantane
analogues did not show any improvement in catalytic activity compared to the phenyl
analogues, the nanodiamond analogues showed a big increase in catalytic activity.
Table 4.1 Percentage conversion to THQ.
RuH(CO)(!3-OCOR)
(PPh3)2
%[THQ]
OsH(CO)(! 3 -OCOR)
(PPh3)2
%[THQ]
12a
8.2
16a
2.4
12b
6.4
16b
0.8
12c
12.8
16c
14.9
Elemental analyses (Table 4.2) showed that the % Ru and % Os present 12c and 16c,
respectively, were much lower than that present in 12a and 16a (Table 4.2). Thus the
higher catalytic activities of 12c and 12c were not due to higher amount of Ru or Os
metal moiety present in the ND samples.
!
36
Table 4.2. % Element calculated.
RuH(CO)(! 3 -OCOR)
(PPh3)2
%Ru
OsH(CO)(! 3 -OCOR)
(PPh3)2
%Os
12a
13.0
16a
22.0
12b
12.1
16b
20.6
12c
0.198
16c
0.151
Two factors have been identified to be important in the hydrogenation of quinoline;
(a) the steric bulk of the carboxylate substituents, and (b) their electron releasing
character.10 A more sterically bulky carboxylate group affords a better catalytic
activity. Same is observed for a more electron releasing substituent on the carboxylate
group. Hence, the better catalytic activity exhibited by the nanodiamond analogues
could be accounted to the effect of their steric bulk.
4.3 Conclusion
Metal derivatized NDs exhibit better catalytic activities towards the hydrogenation of
quinoline compared to the adamantane and phenyl analogues. However, we were still
uncertain of the surface chemistry of nanodiamonds that have led to their better
catalytic performance.
!
37
4.4 References
1.
Alan R.K., Stanislaw R., Bogumila R., Tetrahedron, 1996, 52, 48, 15031
2.
Hlasta D.J., Luttinger D., Perrone M.H., Silbernagel M.J., Ward S.J., J. Med.
Chem., 1987, 30, 1555
3.
Jones R.C.F., Smallridge M.J., Chapleo C.B., J. Chem. Soc., Perkin Trans. 1,
1990, 385
4.
Shuman R.T., Ornstein P.L., Paschal J.W., Gesellchen P.D., J. Org. Chem.,
1990, 55, 738
5.
Schaus J.M., Huser D.L., Titus R.D., Synth. Commum., 1990, 20, 3553
6.
Murahashi S.-I., Imada, Y., Hirai, Y., Tetrahedron Lett. 1987, 28, 77
7.
Borowski A.F., Sabo-Etienne S., Donnadieu B., Chaudret B., Organometallics,
2003, 22, 1630
8.
Sánchez-Delgado R.A., Valencia N., Márquez-Silva R.-L., Andriollo A.,
Medina M., Inorg. Chem., 1986, 25, 1106
9.
Vaska L., Diluzio J.W., J. Am. Chem. Soc., 1962, 63, 1262
10.
Merlín R., Beatriz A., Federico A., Carlos D.L.C., Ángel G., Karely M., Otto S.,
Yslamar S., Polyhedron, 2008, 27, 530
!
38
Chapter 5: Conclusion
The derivatization of nanodiamonds with osmium and ruthenium carbonyl clusters
was achieved via the direct reaction of the cluster precursors 1-3 and 5 with
nanodiamonds. It was found that prior derivatization of the nanodiamond surface was
not necessary. The derivatized nanodiamonds were less prone to aggregation
compared to the underivatized nanodiamonds. Characterization of the surface species
by IR and ToF-SIMs suggested that clusters interacted directly with the ND surface,
but rapid oxidative fragmentation to mononuclear species took place. It is postulated
that the mononuclear di- and tricarbonyl species obtained remained close to their
original position within the cluster on the surface.
The possible interaction of the cluster precursors with surface COOH groups was
modeled through the reaction of Ru3(CO)12 (5) with 1-adamantanecarboxylic acid.
This reaction yielded a purple polymeric complex, [Ru2(CO)4(OOCAd)2]n (6) which
showed different IR spectroscopic characterization from that of ruthenium derivatized
nanodiamond, thus suggesting that the surface species was not the same. Complex 6
was found to react with PPh3 to form a novel, electron-deficient, tetranuclear
ruthenium chain complex, Ru4(CO)8(OOCAd)4(PPh3)2 (9). Both 6 and 9 were found
to change colour reversibly from light purple to yellow in the presence of solvent
vapours.
The catalytic activity of metal derivatized NDs investigated with the complexes
OsH(CO)(&3-OOCND)(PPh3)2 (16c) and RuH(CO)(&3-OOCND)(PPh3)2 (12c) for the
hydrogenation of quinolines showed enhanced activity compared with the phenyl
analogues, although the reason for the enhancement remain unknown.
!
39
Chapter 6: Experimental
6.1 General experimental
All manipulations were performed under an inert atmosphere of argon using standard
Schlenk techniques unless stated otherwise. The compounds, Os3(CO)10(NCCH3)2
(2), Os3(CO)11(NCCH3) (3), RuH2(CO)(PPh3)3 (11), OsH2(CO)(PPh3)3 (13),
RuH(CO)(&3-OCOPh)(PPh3)2 (12a), OsHCl(CO)(PPh3)3 (14) and OsH(CO)(&3OCOPh)(PPh3)2 (16a) were prepared according to reported procedures with slight
modifications.1,5,6,7 Chemicals, solvents and gases that were used for reactions were
purchased from commercial sources and used as supplied without further
purification. Nanodiamonds and carboxylated nanodiamonds were purchased from
International Technology Center (ITC), North Carolina, US. Reaction mixtures were
separated by preparative thin-layer chromatography (TLC) with 20 cm x 20 cm plates
pre-coated with silica gel K60F254, purchased from Merck.
1H
NMR,
31P{1H}
and
13C{1H}
spectra were recorded on a Bruker AC300 or
AMX500 spectrometer with chemical shifts referenced to residual solvent peaks in
the respective deuterated solvents. Infrared spectra were recorded on a Bruker Alpha
FTIR spectrometer. ESI was recorded on a Thermo Deca Max (LCMS) mass
spectrometer with an ion-trap mass detector; HRMS were recorded either in ESI
mode on a Waters UPLC-Q-Tof MS mass spectrometer, or in FAB mode in a 3nitrobenzyl alcohol matrix on a ThermoFinnigan Mat95XP mass spectrometer. TofSIMS analyses were performed on an ION-TOF SIMS 4 instrument, using bunched
69Ga+
ion pulses with impact energy of 25 keV in the Institute of Materials Research
& Engineering. Samples were prepared on copper tapes mounted on cover slips of 12
mm by 5 mm size. UV analysis was performed on a Cary 100 Bio UV-vis
!
40
spectrophotometer and the results were quantified using Varian Cary WinUV
software. Samples for metal analyses were digested with 37% aq HCl, centrifuged,
decanted, washed and made up to standard solutions with deionized water. The
analyses were performed by the microanalytical laboratory at NUS on a Dual-view
Optima 5300 DV ICP-OES system. Computational studies were carried out by A/P
Leong Weng Kee, using the Gaussian 03W computational package. All structures
were optimized using the B3LYP density functional and LANL2DZ basis set.
6.2 Reaction with nanodiamonds
Cluster 1 (8.7 mg, 0.00933 mmol) and ND (21.9 mg) were placed into a one-neck rbf
with toluene (8 ml). The suspension was left to stir overnight at room temperature.
Thereafter, it was centrifuged, decanted and washed with DCM. The same
experimental procedure was followed for the reactions with clusters 2, 3 and 5, as
well as analogous reactions with ND-COOH. The solid-state IR spectra are tabulated
in Table 6.1.
6.3 Thermolysis of osmium- and ruthenium-derivatized nanodiamonds
In a typical experiment, O5 (6.5 mg) was refluxed in octane (5 ml) for 6 h. The
resulting suspension was centrifuged, decanted and washed with octane. The IR
spectrum of the mother liquor showed no absorption in the carbonyl region while the
solid was unreacted material. For R1 and R3, no peaks were observed in the carbonyl
region of the IR spectra of both the solid and the mother liquor.
!
41
!
42
ND-COOH + cluster 1
ND-COOH + cluster 2
ND-COOH + cluster 3
ND + cluster 1
ND + cluster 2
ND + cluster 3
ND + cluster 1
ND + cluster 5
ND + cluster 5
ND-COOH + cluster 5
ND-COOH + cluster 5
O1
O2
O3
O4
O5
O6
O7
R1
R2
R3
R4
85
RT
85
RT
85
RT
RT
RT
RT
RT
RT
Temperature/ ºC
2120(vw), 2053(vs), 1979(vs), 1768(m,br), 1625(m,br)
2131(w), 2064(vs), 1998(s), 1792(s,br), 1627(s,br)
2051(s), 1978(s), 1630(m)
2053(m), 1981(m), 1718(vw), 1630(br)
2120(vw), 2025(w), 1940(vw), 1638(s)
2122(w), 2022(vs), 1937(s), 1718(vw), 1629 (s,br)
2121(w), 2022(vs), 1936(s), 1734(vw), 1628 (s,br)
2118(vw), 2085(w), 2033(vw), 1719(vw), 1629(s)
2125(s), 2053(vs), 1950(s), 1787(w,br), 1630(s,br)
2125(s), 2090(vw), 2032(vs), 1967(s), 1789(w,br), 1628(br)
2128(m),2041(s),1968(vw),1794(s,br),1629(s,br)
Solid IR (KBr)
Table 6.1 Solid-state IR data for clusters.
6.4 Synthesis of Os3(CO)10(µ-H)(µ-OOCAd), 4
1-adamantanecarboxylic acid (48.1 mg, 0.266 mmol) and 2 (110.6 mg, 0.119 mmol)
were dissolved in toluene (15 ml) and stirred overnight at room temperature. The
crude mixture was purified by TLC using hexane : DCM (4:1, v/v) as eluant. Cluster
4 was obtained as the fastest moving yellow band.
Yield: 40.0 mg (32.6%)
!CO (cm-1) (Hex): 2113 (w), 2075 (s), 2063 (s), 2027 (vs), 2015 (s), 2010 (sh), 1988
(m), 1984 (m)
!CO (cm-1) (KBr): 2113 (m), sh, 2059 (vs), 2027 (vs), sh, 1998 (vs), 1985 (sh), 1967
(s)
1H
NMR (CDCl3): % 1.26 (s, CH2CH), 1.59 (s, CH), 1.91 (s, CH2CCOO), -10.6 (s,
OsHOs) ppm
FAB-MS: m/z 1030.1 [M]+, 1002.1 [M-CO]+, 975.1 [M-2CO]+, 948.1 [M-3CO]+,
919.1 [M-4CO]+
6.5 Thermolysis of Os3(CO)10(µ-H)(µ-OOCAd), 4
Cluster 4 (4.2 mg, 0.00407 mmol) in octane (5 ml) was refluxed for 6 h. Upon
removal of solvent, some black precipitate was obtained. However, IR spectrum
showed unreacted 4.
6.6 Synthesis of ruthenium polymer, 6
Cluster 5 (112.6 mg, 0.176 mmol) and 1-adamantanecarboxylic acid (191.6 mg, 1.06
mmol) were refluxed in MeCN (15 ml) for 2 h in a one-necked rbf and then cooled
down to room temperature.2 MeCN was removed before adding DCM and the
!
43
resulting orange solution was filtered through a thin pad of silica gel. Solvent was
then removed to afford a purple solid.
Yield: 156.1 mg (87.8%)
!CO (cm-1) (DCM): 2041 (vs), 1990 (s), 1960 (vs), 1924 (w)
!CO (cm-1) (ACN) of yellow solution: 2031 (vs), 1979 (m), 1947 (vs)
!CO (cm-1) (MeOH) of yellow solution: 2031 (vs), 1976 (m), 1945 (vs)
!CO (cm-1) (EtOH) of yellow solution: 2030 (vs), 1976 (m), 1944 (vs)
!CO (cm-1) (Toluene) of yellow solution: 2038 (vs), 1988 (m), 1959 (vs), 1935 (w),
1925 (w)
!CO (cm-1) (KBr): 2032 (vs), 1990 (s), 1955 (vs)
13C{1H}
NMR (C6D6): % 29.2 (CH2CH), 37.4 (CH2CH), 40.7 (CH2CCOO), 42.8
(CH2CCOO), 191.2 (CH2CCOO), 201.6 (RuCO) ppm
6.7 Synthesis of Ru2(CO)4(OOCAd)2(PPh3)2, 7
Complex 6 (16.3 mg) and PPh3 (30.4 mg, excess) were stirred in DCM (3 ml) for 45
min. The mixture was purified by TLC using hexane : DCM (4:1, v/v) as eluant.
Complex 7 was obtained as a yellow band, following a colourless band of unreacted
PPh3. Yield: 9.7 mg (33.4%)
An analogous reaction of 8 (12.5 mg) with excess PPh3 (22.3 mg) also afforded 7.
Yield: 5.7 mg (35.6%)
!CO (cm-1) (DCM): 2021 (vs), 1977 (m), 1947 (vs), 1916 (w)
1H
NMR (C6D6): % 1.58 (s, CH2CH), 1.72 (s, CH), 1.83 (s, CH2CCOO), 7.08 (m, Ph),
7.83 (m, Ph) ppm
31P{1H}
NMR (C6D6): % 14.3 ppm
ESI-MS: m/z 1198.20 [M]+
!
44
6.8 Synthesis of Ru2(CO)4(OOCAd)2(TPPTs)2, 8
Complex 6 (7.6 mg) in DCM (2 ml) and TPPTs (5.1 mg) in deionized water (1 ml)
was stirred vigorously for 2 h. The aqueous layer was separated from DCM using
separatory funnel and water layer was washed with DCM (3x 5 ml). Removal of
water from the aqueous layer afforded 8 as a yellow solid.
Yield: 11.6 mg (56.7%)
!CO (cm-1) (KBr): 2021 (vs), 1980 (m), 1949 (vs)
31P{1H}
NMR (C6D6): % 16.9 ppm
6.9 Synthesis of Ru4(CO)8(OOCAd)4(PPh3)2, 9
Complex 6 (18 mg) and triphenylphosphine (1.7 mg) [Ratio of Ru : PPh3 = 2 : 1] was
stirred in DCM (3 ml) for 1 h.3 The reaction mixture was purified by TLC using
hexane : DCM (1:1, v/v) as eluant. Three bands were obtained; in order of eluation,
they were a yellow band of 7 (24.0 %), a light purple band of 9, and a dull yellow
band of 6 (52.0%).
Yield of 9: 5.5 mg (22.0%)
!CO (cm-1) (DCM): 2036 (vs), 1985 (m), 1963 (s), 1916 (w)
!CO (cm-1) (KBr): 2027 (vs), 1987 (s), 1966 (vs), 1929 (m)
1H
NMR (C6D6): % 1.62 (s, CH), 1.90 (s, CH2CCOO), 7.07 (m, Ph), 7.73 (m, Ph) ppm
31P{1H}
13C{1H}
NMR (C6D6): % 8.8 ppm
NMR (C6D6): % 29.2 (CH2CH), 37.4 (CH2CH), 40.2 (CH2CCOO), 43.0
(CH2CCOO), 193.8 (CH2CCOO), 206.5 (RuCO) ppm
ESI-MS: m/z 1871.23 [M]+, 1690.08 [M-C10H15COO]+
!
45
6.10 Synthesis of higher nuclearity ruthenium chain complexes
A mixture of 6 (2.0 mg) and 9 (2.5 mg, 0.00134 mmol) was stirred in DCM (4 ml) for
1 h to give a yellow solution. IR of the crude solution showed presence of 9.
Removal of solvent gave a light purple solid and ESI-MS did not showed peaks of
higher nuclearity complexes.
6.11 Synthesis of Ru2(CO)6(OOCAd)2, 10
A solution of 6 in DCM (2 ml) was placed in a carius tube, degassed by three cycles
of freeze-pump-thaw and then filled with CO gas (1 bar). Immediately, the orange
yellow solution turned pale yellow. !CO (cm-1) (DCM): 2103 (m), 2078 (vs), 2032
(vs), 1999 (vs). Within minutes in solution, it would slowly decompose to give a
darker yellow solution by losing CO.2 !CO (cm-1) (DCM): 2099 (s), 2078 (m), 2033
(vs), 2002 (vs), 1961 (w), 1935 (m). Removal of solvent afforded 6 again.
6.12 Reaction of 6 with NCCH2CH3, NEt3, Cp*Ir(CO)2 and tBuNC
Complex 6 (~5 mg) was reacted with neat propionitrile (0.1 ml) to give a yellow
solution. IR spectrum of crude showed presence of a coordinated complex,
Ru2(CO)4(OOCAd)2(NCCH2CH3)2. Removal of solvent afforded 6 again. The same
observations were made with triethylamine (NEt3) and tert-butyl isocyanide (tBuNC).
IR spectra of reaction mixture:
With NCCH2CH3, !CO (cm-1) (DCM): 2018 (vs), 1965 (m), 1931 (vs), 1898 (w)
With NEt3, !CO (cm-1) (DCM): 2031 (vs), 1980 (m), 1947 (vs), 1917 (w)
13C{1H}
NMR (C6D6): % 29.3 (CH2CH), 37.5 (CH2CH), 40.8 (CH2CCOO), 42.6
(CH2CCOO), 190.2 (CH2CCOO), 205.2 (RuCO) ppm
!
46
With tBuNC, !CO (cm-1) (DCM): 2171 (m), 2025 (vs), 1983 (m), 1954 (vs)
13C{1H}
NMR (C6D6): 156.7 (t, C'N), 184.8 (CH2CCOO), 208.9 (RuCeqO), 214.7
((RuCaxO) ppm
For !-pentamethylcyclopentadienyl iridiumdicarbonyl (Cp*Ir(CO)2), 6 (~5 mg)
reacted with Cp*Ir(CO)2 (~2 mg) in DCM (5 ml) to afford a red solution instead, but
the product could not be isolated. !CO (cm-1) (DCM): 2043 (m), 2011 (s), 1990 (m),
1960 (m), 1938 (s)
6.13 Synthesis of RuH(CO)(!3-OCOR)(PPh3)2 where R = Ad (12b), ND (12c)
Compound 11 (38.4 mg, 0.0418 mmol) and 10-fold molar excess of the appropriate
acid were refluxed in deoxygenated toluene (5 ml) for 45 min.5 The resulting yellow
solution was concentrated to a small volume and methanol (5-10 ml) was added to
precipitate out the product. The precipitate was centrifuged, decanted and washed
with MeOH. The washing was repeated till they were colorless.
For complex 12b:
Yield: 60.7%
!RuH (cm-1) (DCM): 2020 (w)
!CO (cm-1) (DCM): 1923 (vs)
1H
NMR (C6D6): % -15.4 (t, Ru-H) ppm
31P{1H}
NMR (C6D6): % 43.9 ppm
FAB-MS: m/z 833.2 [M]+
For complex 12c:
Yield: ~50% (estimated)
!CO (cm-1) (KBr): 1968 (m)
!
47
The ruthenium complexes were only stable for a few days even under argon, with
RuH(CO)(&3-OCOPh)(PPh3)2 (12a) being the most unstable.
6.14 Synthesis of [OsH(CO)(NCCH3)2(PPh3)2][BF4], 15
Compound 14 (114.4 mg, 0.110 mmol) and NaBF4 (32.5 mg, 0.293 mmol) were
refluxed in deoxygenated ACN (10 ml) in a two-necked rbf for 2.5 h.8 After cooling
to room temperature, the suspension was filtered in air and the colorless filtrate
evaporated to dryness. The residue was washed with diethyl ether (3x5 ml) to afford
15.
Yield: 80.0 mg (79.7%)
!CO (cm-1) (DCM): 1949 (s)
1H
NMR (C6D6): % -13.2 (t, Os-H) ppm
31P{1H}
NMR (C6D6): % 18.9 ppm
ESI: m/z 826.5 [M]+
6.15 Synthesis of OsH(CO)(!3-OCOR)(PPh3)2 where R = Ad (16b), ND (16c)
Compound 15 (27.0 mg, 0.0296 mmol) and the sodium salt of the appropriate
carboxylic acid (prepared by reaction of the carboxylic acid with excess sodium
carbonate in MeOH) (13.3 mg, 0.0917 mmol) were dried under vacuum in a schlenk
tube.9 A deoxygenated mixture of DCM and MeOH (6 ml, 1:1, v/v) was added and
the mixture was stirred vigorously at room temperature for 3 h. The resulting pale
yellow solution was concentrated to afford a white precipitate, which was washed
with MeOH and dried under vacuum.
!
48
For complex 16b:
Yield: 57.4%
!OsH (cm-1) (DCM): 2111 (vw)
!CO (cm-1) (DCM): 1904 (s)
1H
NMR (C6D6): % -18.2(t, Os-H) ppm
31P{1H}
NMR (C6D6): % 25.6 ppm
For 16c, compound 13 (26.9 mg, 0.0267 mmol) and ND-COOH (21.0 mg, excess)
were refluxed in deoxygenated toluene (5 ml) for 1 h and then cooled.5 The resulting
mixture was centrifuged, decanted and washed with toluene (2x 4 ml) and DCM (2x
4 ml). The solid 16c was stored under argon. !CO (cm-1) (KBr): 1945
The osmium complexes were only stable for a few days even under argon, with
OsH(CO)(&3-OCOPh)(PPh3)2 (16a) being the most unstable.
6.16 Details of catalytic runs
Gas chromatographic analyses were performed on a Shimadzu GC-2010A series
instrument fitted with a flame ionization detector (FID), AOC-20i auto sampler and
DB-5MS GC column. It was connected to a Parker Balston H2PEM-260 hydrogen
generator. N2, He and air were used as carrier gases and naphthalene was used as an
internal standard. The results were quantified using the GC solution version 2.31
software.
Samples were prepared by adding naphthalene as internal standard (1.0 ml of a
0.0180 M solution in p-xylene), quinoline (0.2 ml), 1 mol% of the appropriate
!
49
catalyst and made to the mark with p-xylene in a 10 ml volumetric flask. A 1.0 ml
aliquot was withdrawn by micro syringe and diluted to 10 ml with MeOH for GC
analysis. The rest of the solution mixture was placed in a high pressure reactor,
flushed three times with hydrogen, re-filled with 5 bars of hydrogen, and heated at
130oC for 24 h. After the reaction, the system was allowed to cool down to room
temperature and analyzed by GC as above.
The calibration curves for quinoline and 1,2,3,4-tetrahydroquinoline were obtained
from 5 standard solutions each (10 ml) with naphthalene as internal standard. The
ratios of peak areas with respect to naphthalene were used.
6.17 X-Ray Crystallography studies of 4, 7 and 9
Crystals of diffraction quality were grown by slow evaporation of DCM solution.
There was a disorder of the adamantane cage in 4. This was modeled with two
alternative positions, and appropriate distance and thermal parameter restraints were
placed on chemically equivalent carbon atoms. Disorder of one of the two
adamantane cages was also found in 9. This was modeled with two alternative
positions, and the corresponding distances between the two were restrained to be the
same. Crystal and refinement data are summarized in Table 6.2.
The crystal structure determinations were carried out by Dr Li Yongxin at the
Nanyang Technological University on a Bruker Kappa diffractometer equipped with
a CCD detector, employing Mo K( radiation () = 0.71073 Å). Data was collected at
103 K with the SMART suite of programs,10 processed and corrected for Lorentz and
polarisation effects with SAINT,11 and for absorption effects with SADABS.12
!
50
Structural solution and refinement were carried out with the SHELXTL suite of
programs.13
Table 6.2 Crystal data and structure refinement.
4
9
Empirical formula
C21 H16 O12 Os3
C63 H62 Cl2 O8 P2 Ru2
C88 H90 O16 P2 Ru4
Formula weight
1030.94
1282.11
1869.82
Temperature, K
103(2)
103(2)
103(2)
Wavelength, Å
0.71073
0.71073
0.71073
Crystal system
Orthorhombic
Triclinic
Triclinic
Space group
Pbca
P-1
P*
a = 9.8471(8) Å
a = 11.5054(2) Å
(= 92.8190(10)°
a = 11.646(2) Å
(= 97.623(5)°
b = 14.9836(12) Å
b = 14.6673(2) Å
+= 107.9450(10)°
b = 12.730(3)Å
+= 101.646(6)°
c = 33.408(3) Å
c = 17.3640(3) Å
,= 90.1070(10)°
c = 13.774(3) Å
,= 98.543(5)°
Volume, Å3
4929.2(7)
2783.86(8)
1949.6(7)
Z
8
2
1
Density (calculated), Mg/
m3
2.778
1.530
1.593
Absorption coefficient,
mm-1
15.490
0.753
0.870
F(000)
3728
1312
952
Crystal size, mm3
0.40 x 0.36 x 0.34
0.32 x 0.30 x 0.10
0.30 x 0.14 x 0.04
Theta range for data
collection, o
2.55 to 34.71
1.81 to 31.13
2.05 to 26.37
Reflections collected
69571
66624
26990
Independent reflections
10541 [R(int) = 0.0859]
17758 [R(int) = 0.0287]
7959 [R(int) = 0.0288]
Max. and min. transmission
0.0769 and 0.0628
0.9285 and 0.7947
0.9660 and 0.7802
Data / restraints / parameters
5047 / 174 / 356
17758 / 0 / 694
7959 / 24 / 578
Goodness-of-fit on F2
1.09
1.187
1.216
Final R indices
[I>2sigma(I)]
R1 = 0.0342,
wR2 = 0.0801
R1 = 0.0268, wR2 = 0.0742
R1 = 0.0309, wR2 = 0.0931
R indices (all data)
R1 = 0.0472,
wR2 = 0.0882
R1 = 0.0349, wR2 = 0.0901
R1 = 0.0444, wR2 = 0.1262
Largest diff. peak and hole,
e.Å-3
2.463 and -2.835
0.916 and -1.170
1.037 and -2.256
Unit cell dimensions
!
7
51
6.18 References
1.
Nicholls J.N.; Vargas M.D., Inorg. Synth., 1990, 28, 232
2.
Crooks G.R., Johnson B.F.G., Lewis J., Williams I.G., Gamlen G., J. Chem.
Soc. (A), 1969, 2761
3.
Mario B., Piero F., Ugo M., Gloria M., Franco P., Giorgio P., J. Organomet.
Chem., 1983, 259, 207
4.
Ahmad N., Levinson J.J., Robinson S.D., Unley M.F., Inorg. Synth., 1974, 15,
48
5.
Sánchez-Delgado R.A., Thewalt U., Valencia N., Andriollo A., Márquez-Silva
R.-L., Puga J., Schöllhorn H., Klein H.-P., Fontal B., Inorg. Chem., 1986, 25,
1097
6.
Ahmad N., Levinson J.J., Robinson S.D., Unley M.F., Inorg. Synth., 1974, 15,
54
7.
Ahmad N., Levinson J.J., Robinson S.D., Unley M.F., Inorg. Synth., 1974, 15,
53
!
8.
Rosales M. et al., Inorganica Chimica Acta, 1997, 257, 131
9.
Rosales M. et al., Polyhedron, 2008, 27, 530
10.
SMART Version 5.628, Bruker AXS Inc., Madison, Wisconsin, USA, 2001
11.
SAINT+ Version 6.22a, Bruker AXS Inc, Madison, Wisconsin, USA, 2001
12.
Sheldrick G.M., SADABS, 1996
13.
SHELXTL Version 5.1, Bruker AXS Inc, Madison, Wisconsin, USA, 1997
52
[...]... ND (16c) ix For simplicity, we assigned the following nanodiamonds with metal clusters as O1 O2 O3 O4 O5 O6 O7 ND-COOH with cluster 1 ND-COOH with cluster 2 ND-COOH with cluster 3 ND with cluster 1 ND with cluster 2 ND with cluster 3 ND with cluster 1 at 85ºC R1 R2 R3 R4 ND with cluster 5 ND with cluster 5 at 85ºC ND-COOH with cluster 5 ND-COOH with cluster 5 at 85ºC x ABBREVIATIONS Ad CO Cp* Cy DCM... acid-functionalized (ND-COOH) nanodiamonds The reactions of these nanodiamonds with osmium and ruthenium clusters, and the characterization of the products, are presented in this chapter 2.1 Osmium derivatized nanodiamonds As mentioned in Chapter 1, the cluster Os3(CO)10(NCCH3)2 (2) reacts with carboxylic acids to afford derivatives of the type Os3(CO)10(µ-H)(µ-OOCR) The reaction of ND-COOH with 2 was compared... the OH of a carboxylic group,3 which suggested the possibility of reaction with nitro groups present on the ND surface ! 13 Fig 2.5 ToF-SIMS (positive ion mode) spectrum of O4 2.2 Ruthenium derivatized nanodiamonds Derivatization of the nanodiamonds with the ruthenium cluster Ru3(CO)12 (5), using identical conditions as for the osmium clusters above, afforded products that showed the presence of ruthenium;... xiv Lists of Schemes Scheme 1.1 Synthesis of 2 and 3 from Os3(CO)12 (1) p.4 Scheme 1.2 Proposed scheme for the reaction of 2 with carboxylic acids p.5 Scheme 2.1 Reaction of cluster precursors with nanodiamonds and AdCOOH p.9 Scheme 3.1 Reaction of 9 with PPh3 p.28 Scheme 4.1 Synthesis of MH(CO)(!3-OCOR)(PPh3)2 p.35 Scheme 4.2 Hydrogenation of quinoline p.36 xv Chapter 1: Introduction to Nanodiamonds, ... 2.8 Dispersion of derivatized NDs in DCM: (top) freshly prepared and (bottom) after standing for 1 day 2.4 Conclusion We have successfully derivatized nanodiamonds with osmium carbonyl clusters as shown in their IR spectra, ToF-SIMS and EA analyses High reaction temperature resulted in more metal species deposited on the ND surface When nanodiamonds reacted with osmium clusters, the triosmium core was... retained and the osmium atoms bonded directly to the ND surface However, oxidative fragmentation took place almost immediately, affording a mixture of osmium di- and tricarbonyl species A similar behaviour was observed in the derivatization of nanodiamonds with ruthenium carbonyl clusters The osmium and ruthenium derivatized NDs were found to aggregate less rapidly as compared to before derivatization. .. observed, in contrast to that for the osmium derivatives Furthermore, the presence of [Ru3(CO)10(µ-O)(µH)]+ may indicate interaction of 5 with hydroxyl groups on the surface of the NDs.8 Thus it is likely that as in the case of the derivatization of nanodiamonds with 5, the ruthenium atoms were bonded directly to the ND surface Oxidative fragmentation, to afford a mixture of Ru(CO)2 (in derivatized ND) and... O1-3 The ToF-SIMS spectra revealed the presence of both osmium di- and tricarbonyl species, and clusters species (Fig 2.5) Since it was mentioned earlier on that oxidative fragmentation left the osmium atoms in their original position on the surface, the presence of cluster species maybe due to the ensembles of the Os(CO)2 and Os(CO)3 units during ToF-SIMs Generally, the ToF-SIMS spectra of the compounds... reaction of Os3(CO)10(µ-H)(µ-OH) with the ZnO and In2O3 surfaces,4 that has been ascribed to mononuclear di- and tricarbonyl osmium species (Os(CO)2 or Os(CO)3 units) (Fig 2.4).5 It was reported that upon exposure to air, the chemisorbed triosmium carbonyls fragmented to form a mixture of osmium di- and tricarbonyl species.6 It was also found that these osmium species occurred in groupings consisting of. .. Chem., 2005, 178, 688 18 Chapter 3: Molecular model of the cluster-nanodiamond interaction: The reaction of Ru3(CO)12 with 1-adamantanecarboxylic acid In an attempt to identify the type of interaction of the ruthenium clusters with the ND surface for samples R1-4, we investigated the reaction of Ru3(CO)12 (5) with 1adamantanecarboxylic acid The reaction of 5 with carboxylic acids to form polymeric [Ru(CO)2(OOCR)]n .. .DERIVATIZATION OF NANODIAMONDS WITH OSMIUM AND RUTHENIUM CARBONYL CLUSTERS OH SUAT PING B.Sc (Hons.), NUS A THESIS SUBMITTED FOR THE DEGREE OF MASTER OF SCIENCE DEPARTMENT OF CHEMISTRY... following nanodiamonds with metal clusters as O1 O2 O3 O4 O5 O6 O7 ND-COOH with cluster ND-COOH with cluster ND-COOH with cluster ND with cluster ND with cluster ND with cluster ND with cluster... immediately, affording a mixture of osmium di- and tricarbonyl species A similar behaviour was observed in the derivatization of nanodiamonds with ruthenium carbonyl clusters The osmium and ruthenium derivatized