Migratory insertion of carbene complexes of ruthenium


Material Information

Migratory insertion of carbene complexes of ruthenium preparation of ruthenabenzene
Physical Description:
vi, 120 leaves : ill. ; 29 cm.
Yang, Jing, 1959-
Publication Date:


Subjects / Keywords:
Carbenes (Methylene compounds)   ( lcsh )
Ruthenium   ( lcsh )
Organometallic compounds   ( lcsh )
Chemistry thesis, Ph. D
Dissertations, Academic -- Chemistry -- UF
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1995.
Includes bibliographical references (leaves 117-119).
Statement of Responsibility:
by Jing Yang.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 021631013
oclc - 33424643
System ID:

This item is only available as the following downloads:

Full Text







To my husband, Bill Dolbier


It is with much gratitude that I acknowledge the excellent guidance of my

research advisor, Dr. William M. Jones., for having made the fulfillment of this dream

possible. His enthusiasm for chemistry and his understanding nature have been an

inspiration to me throughout the course of my graduate studies.

I would like to give my special thanks to my committee members, especially,

Dr. Merle Battiste and Dr. James Boncella for taking the time to discuss my projects

and providing valuable insight.

It is a pleasure to thank the many friends and colleagues whom I have met and

worked with at the University of Florida, including Dr. Yin-sheng Wang, Dr. Conrad

Burkholder, Michael Bartberger, Xiao-xin Rong, Dr. Ronda Trace, Dr. Tatjana Omrcen,

Dr. Lu Zheng, and Jerzy Klosen for their helpful discussions and comraderie over the

years. The friendship and guidance of Dr. Jian-guo Yin was especially appreciated.

The assistance of Mr. Rudy Strosheim in constructing numerous glassware, including

my special low temperature apparatus, was very much appreciated, as were the essential

contributions of Dr. Roy King and Dr. Khalil Abboud to my research. Thanks also goes

to Dr. Neil Allison and Ms. J. Dixon for providing some of the starting materials.

I must also express gratitude to my parents who imbued in me those important

values in life which I will carry with me always.

Finally, I would like to thank Dr. Bill Dolbier, my best friend and my husband,
for his understanding and his loving support.


ACKNOWEDGEMENTS...................................... ...........................iii

A B STR A C T ............................................ ............ ........................v


1 INTRODUCTION....................................... ...........................1

CARBENE COMPLEX...............................................................15


AND RUTHENABENZENE........................................ .................54

5 CONCLUSIONS.................................................. ... 84

6 EXPERIMENTAL............................. ..... ........................86


TABLES OF CRYSTALLOGRAPHIC DATA....................................104

REFERENCES......................................... ...... .........................117

BIOGRAPHICAL SKETCH.................................................................120

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy


Jing Yang

May, 1995

Chairman: William M. Jones
Major Department: Chemistry

Research was carried out with the overall purpose of better understanding

migratory insertion processes of ruthenium carbene complexes.

A methoxy-stabilized metallaindene complex of ruthenium (2.04) was prepared.

Reaction of this complex with phenyllithium orp-tolylithium followed by treatment with

trimethylsilyl triflate afforded the first stable aryl substituted Fischer carbene complexes of

ruthenium (2.03a and 2.03b) which do not bear a heteroatom stabilizing substitutent on

the carbene carbon. In solution, these complexes underwent a novel thermal rearrangement

to form ruthenacene derivatives 2.11a and 2.11b. A possible mechanism for this
rearrangement which involves a ruthenanaphthalene intermediate is proposed.

Attempts to isolate and study the kinetics of the migratory insertion of acyclic
ruthenium carbene complexes, CH3(CO)Ru=C(OR)C6H5 (R = SiMe3 & C2Hs), were

unsuccessful. Neverthless, insight regarding their reactivity was obtained, including the

fact that apparently such complexes rearrange rapidly at temperatures as low as -78 OC.

Introduction of potential aromaticity into such carbene complexes by synthesis of a

ruthenaphenanthrene (4.01) was not sufficient to slow migratory insertion to the point that

the carbene complex could be observed. The rapid rearrangement of 4.01 at -50 oC was

On the other hand, introduction of a greater degree of aromaticity into a ruthenium

carbene complex through the synthesis of a diphenyl substituted ruthenabenzene (4.16)

was sufficient to permit its detection at -50 oC. 4.16 is the first second row, late transition

metallabenzene which has been reported.


A carbene migratory insertion is defined as the 1, 2-rearrangement of a hydrogen,
alkyl or aryl group from a transition metal of a carbene complex, such as 1.01, to a

carbene carbon to give an unsaturated species 1.02.

R\ / Q
M=C MC--
Lx L/ I
1.01 1.02
18-electron metal 16-electron metal
R = H, Alkyl, Aryl

Scheme 1-1

The recognition that carbene migratory insertions are fundamentally important,
discrete steps in organometallic reactions dates back twenty five years ago.1 However,
during the last few years the postulation of migratory insertions in discussions of
mechanisms of organometallic reactions has become especially common.2"3
Mechanistic studies by Bercaw14 on carbene migratory insertions of the early
transition metal tantalum, such as shown in Scheme 1-2, were instrumental in the

development of a general understanding of such processes.

Cp* a= CH w Cp*2 Ta- CHCH3]- Cp*2 Ta--"
1.03 1.04 1.05

Scheme 1-2

Complex 1.03 cleanly rearranges to xr-complex 1.05 via methyl migration from Ta
to methylidene (AG* = 30.3 kcal/mol at 140 OC ), followed by rapid P-H elimination from
the ethyl complex 1.04.
Theory predicts that migratory insertion rearrangements should be reversable, the
reverse step to form a carbene complex being known as alpha elimination (right to left in
Scheme 1-1). However, other than for rearrangements of hydrogen, such alpha
eliminations have only been rarely observed. In fact, the example of Grubbs and his
coworkers,15 shown in Scheme 1-3, was the only case in which strong evidence had been
provided for such a rearrangement until Jones' work in 1986.

C NiLx ------ C Lx-l products


Scheme 1-3

Theoreticians have rationalized the lack of observation of rearrangements of alkyl
or aryl groups from saturated carbon to metals on the basis of thermodynamics. First
Hoffmann16, but more recently Ziegler17, and Goddard13 have predicted migratory
insertions to be exothermic. Goddard's results, which dealt with hydride shifts within

ruthenium carbene complex 1.06 (in Scheme 1-4 ), predicted a strong kinetic and
thermodynamic preference for rearrangement from the metal center (i.e, migratory
insertion ), while Ziegler's calculations for manganese complex 1.07 (shown in Scheme
1-5) predicted a smaller preference for methyl migration than for hydride migration.

C1- RuHI Cl- Ru H
C H Cl- RC t H

1.06 Ea=10.9 +1.7 kcal/mol
AH=-10.5 +1.0 kcal/mol

Scheme 1-4


(CO)4Mn= CH (CO)4Mn- -
1.07 R=H AH = 31.74 kcal/mol

R=CH3 AH =-16.94 kcal/mol

Scheme 1-5

Recent research within the Jones group has been directed towards gaining insight
into those structural factors which control such equilibria. One aspect of this research has
sought to design systems which would favor alpha elimination, while the purpose of

another has been to study the migratory insertion process. Moreover, the Jones' group

has centered its attention upon the chemistry of the first and second row late transition

metals, such as manganese, iron and ruthenium etc., because there have been little data

reported on alpha elimination or migratory insertion processes of those carbene


As a matter of fact, before Jones' work, only the two stabilized iron carbene

complexes, 1.08 and 1.09 of Rosenblum'8 had been reported with alkyl groups on iron,

presumably because of the strong propensity of lesser stabilized species to undergo

migratory insertion.





In contrast, a number of stable third-row, late-transition-metal complexes have
been reported19 lately. Two examples a shown in Scheme 1-6.
been reported lately. Two examples are shown in Scheme 1-6.



L = PPh3


P = PEt3

Scheme 1-6

The observed greater stabilization of such third row, late-transiton metal carbene
complexes probably derives from their stronger carbon-metal double bond that
compensates for the strength of the carbon-carbon bond which would be formed as a
result of migratory insertion.
Recently, Jones and his coworkers4-7.1020 developed a novel strategy for isolation
of alkyl- and aryl-substituted carbene complexes of iron, ruthenium and manganese
(Scheme 1-7).

Jones' Strategy

stabilization of carbene insertion
migratory insertion
OR sOR inhibited by strain
L / slow \ |
I ) n = 2 or3 L/ -

(CHz) (CH3
1.12 1.13

Scheme 1-7

They concluded that by introducing ring strain into the migratory insertion
product 1.13, it should be possible to shift the thermodynamic equilibrium in favor of
carbene complexes such as 1.12, and thus inhibit their migratory insertion.

Moreover, they concluded that they should be able to use relief of ring strain
during the reverse, a-elimination process to provide a general synthetic method for such

carbene complexes.

General Method
stabilization of carbene
/OR hv 4- a-elm. Z

La rcH2 strain relief \(Hn CH
n=2or3 1.14 1.15

Scheme 1-8

Indeed, Jones and his coworkers have been able to design a number of systems,
generally described by Scheme 1-8, in which unsaturated small ring a-complexes of iron
and manganese, such as 1.14 formed by photoelimination of CO, undergo a-eliminative

ring expansions to produce examples of isolable carbene complexes such as 1.15.
An early and significant example of such a rearrangement of an alkyl group from
saturated carbon to metal to give a carbene complex is outlined in Scheme 1-9.5 The
significance of this example derives from the fact that there was evidence of facile a-

elimination at room temperature and because the carbene complex, 1.19, was actually
able to be isolated.

f Cp vFe-Cp
__ CO L_ CO
1.16 1.17

S1I-- [ Cp
1.21 1.20

1 RT

Cp /CO


Scheme 1-9

Compound 1.17 was prepared in situ by the photolytic decarbonylation of its acyl
complex 1.16, and apparently loses another mole of CO to generate the sixteen electron
intermediate 1.18. This electron-deficient intermediate then rearranges to give the
carbene complex 1.19.
Jones rationalized that one of the important factors in favor of formation of the
carbene complex is the 20-25 kcal/mole relief of strain energy21 which occurs upon
rearrangement of 1.18 to 1.19. Another important factor is the electronic character of the
substituent on the carbene carbon, where it is known that an electron-donating substituent
such as OCH3 provides stablilization to the Fischer-type carbene system.



Jones and Trace2 demonstrated that relief of cyclobutane strain without sufficient
carbene stabilization is apparently not enough to induce this rearrangement, since the
phenyl-substituted substrate failed to rearrange (Scheme 1-10 ).


no rearr. Fe Ph
^=- ^ y

Scheme 1-10

Resonance stabilization of the carbene complex, such as shown in Scheme 1-11
provides a reasonable explanation for the need of an ether or amine substituent at the a-

M- OMe e OMe 0 JMe
M-- M-C M--CM

Scheme 1-11

As can be seen from the conversion of 1.18 to 1.21 in Scheme 1-9, side reactions
can occur in competition with the desired alpha-elimination process. In a study of an
analogous manganese system, Jones and Crowther3 confirmed that the dominant side
reaction was a-hydride elimination (as shown in Scheme 1-12).

MeO II 0 MeO
CO, V. ____ Mn(CO)5
Mn(CO)s h n(CO)4


II li-ination L(CO)4

Mn(5 MnMn(CO)4
+ If

Mn(CO)5 Mn(CO)4

Scheme 1-12

In order to diminish the P-hydride elimination side reaction, Stenstron6 and
Crowtherto investigated the benzocyclobutene system 1.23 (in Scheme 1-13 ) in which
they predicted that P-hydride elimination should be minimized because of the
unlikelihood of forming antiaromatic product 1.25



M=FeCp(CO)2, Mn(CO)5

hv AN
- CO
M=FeCp(CO)2, Mn(CO)5

+CO hv,- CO


1.26 1.27 1.24

M=FeCp(CO) Mn(CO)4 M=FeCp(CO) Mn(CO)4

1 OMe


Scheme 1-13

As expected, no 1.25 was observed, and exclusive rearrangement to the carbene

complexes, 1.26 and 1.27, was observed.

Since these early studies, the Jones group has centered its attention on comparing
the carbene migratory insertion reactions of complexes of the first and second row, late
transition metals ruthenium and iron because both are catalytically active in the Fisher-
Tropsch reaction, a reaction in which this rearrangement is presumably a key step.13
Unfortunately, the photodecarbonylation a-elimination process which had worked
so well for iron and selected manganese alkly- and aryl-substituted carbene complexes
failed completely for the related ruthenium carbene complexes, 1.28 and 1.29 (Scheme 1-


M 'R" 85 C, 4 days




OMe Cp

Cp\ CO

+ OMe


Scheme 1-14

With the ultimate goal of comparing migratory insertions, Jones and Trace22,
therefore, developed a new method to prepare the ruthenium analogue, 1.30, of the iron
complex 1.19 (Scheme 1-15).

1) Na/Hg
[CpRu(CO)2 2) Br(CH2)3Br
CpRu(CO) --

OC u-(CH2)3Br


Cp CO Cp CO Cp

)- ) CO


Scheme 1-15

In Stenstrom and Jones' earlier study,6 deuterium-labeled iron carbene complex
1.19-d, had been used to demonstrate its reversible migratory insertion at room

temperature. (Scheme 1-16).

Cp, CO



MeO CO Cp, /0
Fe-Cp D_ OMe

26 Kcal/mol

Scheme 1-16

Unfortunately, the analogous ruthenium carbene complex, 1.30-d2, proved to be

stable to rearrangement at temperatures up to 100 oC (Scheme 1-17), and one could only

estimate a minimum rate difference of 107-108 for the rates of migratory insertion of the

two complexes.

Cp /CO

/ "^ nn, 100 0C
OMe 100 C No Scrambling Observed
56 hr


Scheme 1-17

Jones20 rationalized the difference in rate of Fe and Ru complexes on the basis of

the difference in strength of the carbon-metal double bonds in the two species. It is a

generally accepted presumption that carbon-metal bonds of second row transition metals

such as ruthenium are stronger than those of first row transition metals such as iron.

The relative reluctance of the ruthenium complex 1.30 to rearrange suggested two
interesting possible projects for further research. First, by using a combination of ring

strain and the inherent stability of alkyl substituted ruthenium carbene complexes to

retard rearrangement, it might be possible to prepare non-heteroatom-substituted (i.e. H,

alkyl and aryl substituted) carbene complexes, such as 1.31, which should be less stable

and thus more reactive. Such complexes not only would be structurally unique (that is,
the first stable alkyl or aryl substituted carbene complexes of ruthenium not stabilized by

alkoxy), but also would be more prone to migratory insertion than their alkoxy analogues.

Cp /co

R= H, Alkyl, Aryl

Secondly, it might be possible to prepare an acyclic Ru carbene complex, such as
1.32, where strain is not a factor either in the starting material or the rearrangement
product, under conditions where it would be observable, and perhaps even isolable.
Then, it might be possible to study the migratory insertion process of this complex.

Cp OEt
\ /


These ideas led to our two individual research projects which were (1) the
synthesis and study of hydrogen-, alkyl-, and phenyl-substituted ruthenium complexes
1.31 which might be stable enough for isolation and further study; and (2) the preparation
and study of the migratory insertion of an acyclic alkoxy-stabilized Ru carbene complex


In order to study the migratory insertion process of the ruthenium carbene
complexes, it was decided to prepare a non-heteroatom substituted Ru-carbene complex
(1.31) since the carbene complex 1.30 which is stablized by a OCH3 group is so stable
that it does not undergo rearrangement. The hope would be that the ground states of
such hydrogen-, alkyl-, or aryl-substituted carbene complexes would be raised
sufficiently to make the transition states more accessible, with the desired result being
that the migratory insertion process would be observable.
Since the carbene complex can be readily synthesized, it was hoped that 1.30
would undergo facile substitution of the methoxy group by nucleophiles as is outlined
in Scheme 2-1 for the reaction with phenyllithium.

Ph OMe Ph

H O Cp PhLi jX. IO cCO
> RRu
H CO OMe 1.32

kh1 .1/Cp
1.30 Ru Cp
1.30 CO --- decomposition

Scheme 2-1

However, it was realized after a failed attempt that the synthetic methodology

indicated in the Scheme 2-1 would not work because, instead of the nucleophile
attacking the carbene carbon followed by expulsion of the methoxy group, the P-

hydrogen would be preferentially deprotonated due to its acidity.

Therefore, we decided to work with a benzannelated system 2.01 (M=Ru) in
which there was no such P-hydrogen present in the molecule. This system has the added

advantage that, based on MMX calculation, a benzocyclobutene species such as 2.02

should be substantially more strained than cyclobutane (49 vs 31 kcal/mole).
Consistent with this, the barrier to rearrangement of 2.01 to 2.02 (M=Fe) has been

found to be at least 13 kcal/mol higher than that of its non-benzannelated analogue.24

The corresponding Mn complex (2.01, M=Mn) was also found to be stable.




OMe Cp


Thus, a straightforward synthetic approach to the benzannelated Ru-carbene

complex 2.04 was designed as outlined in Scheme 2-2.

r^ Rp' 1) n-BuL _C
a lS o CIBr Rp 2)Me3OBFj
H2 2.05 04

Scheme 2-2

2-Bromobenzyl ruthenium complex 2.05 was readily prepared in 66% yield as
an off-white crystalline solid by reaction of [CpRu(CO)2]- (or Rp-) with 2-bromobenzyl

bromide. Addition of 1.2 equivalents of n-BuLi to a solution of compound 2.05 in THF
at -100 oC, followed by warming to room temperature overnight and treatment with
Me30BF4 gave crystals of 2.04 in moderate yield.
Complex 2.04 was characterized by 'H and '3C NMR, IR, and elemental
analyses. To our knowledge, complex 2.04 is the first example of a benzannelated
metallacyclic ruthenium carbene complex. As with its iron and manganese
analogues,7.10 complex 2.04 is exceedingly stable. It showed no evidence of
undergoing carbene migratory insertion, which could have been readily detected by
observation of its rearrangement to isomeric metallindene 2.06 or by trapping
intermediate 2.07 with PMe3. Neither 2.06 (heating at 80 OC for 5 days) nor 2.08

(photolysis in the presence of PMe3 for 6 hours) was observed; 2.04 eventually

decomposed under such conditions.

OMe CP\ /
SRu Cp

2.04 2.07

ihv, PMe3

Cp% I
Ru- PMe3


This stable carbene complex, 2.04, appeared to be an excellent precursor for our

desired synthetic target, 2.03, and other, non-heteroatom-substituted carbene


OMe |1

A 1i /cp CF3SO3SiMe3 \ Y /C
2.04 2.09 2.03a, Ar = Ph
2.03b, Ar =p-Tolyl

Scheme 2-3

Reaction of 2.04 with phenyllithium or p-tolyllithium in THF at -78 OC,
followed by addition of 1.5 equivalents of Me3SiOSO2CF3, gave the Ru-carbene

complex 2.03a and 2.03b in moderate yield, as air-sensitive, black crystals. This
nucleophilic substitution reaction is believed to proceed by a stepwise mechanism that
involves the anionic intermediate 2.09.

The Ru complexes 2.03a and 2.03b are, to our knowledge, the first examples of
stable, aryl substituted, Fischer carbene complexes of ruthenium which do not have a

heteroatom stabilizing substituent on the carbene carbon. In the IR spectra of the

complexes 2.03a and 2.03b, terminal CO absorptions at 1968 and 1966 cm-1 were
observed, respectively, and their characteristic carbene carbon resonances in the 13C

It is interesting that we found that carbene complex 2.04 undergoes facile
substitution of the methoxy group by nucleophiles while its analogues of Mo
and W carbeen complexes 25 failed to undergo such a nucleophilic substitution

NMR were detected at 8 309.5 and 309.2 ppm respectively. Comparing these shifts

with those of the carbene resonance of 2.04 (294.2 ppm), it can been seen that the

carbene resonances of 2.03a and 2.03b are shifted to a relatively lower field, which is in
accord with a somewhat reduced electron density on the carbene carbon of the aryl
substituted molecules. This is not only because the phenyl is an inherently poorer
electron donor than methoxy but also, at least in the crystal, because the phenyl is out of
conjugation with the carbon-metal double bond.
The structure of complex 2.03a was confirmed by x-ray crystallography, the
ORTEP representation being shown in Figure 2-1. The compound has a

pseudooctahedral geometry with the Cp ring on the top. The bond distance between C6
(carbene carbon) and Ru is 1.949(5)A a typical bond length for a metal-carbon double
bond.26 Only one phenyl ring (the benzo ring) was found to be conjugated with the Ru-
carbene bond. As a result, the C6-C13 bond length of 1.444(7)A is shorter than C6-C7
[1.470(7)A], which is about the length of a normal sp2-sp2 single bond (1.48A).27 The

torsion angle of C18-C13-C6-Ru is 18.40. The dihedral angle between Ru-C6-C13 and
C8-C7-C12 is 46.20.

A specific purpose for preparation of complexes 2.03a and 2.03b was to utilize
them in a study of the process of carbene migratory insertion. Thus, an interesting
aspect of the chemistry of these molecules was the surprising fact that thermolysis of
black compound 2.03a in benzene at 50 OC failed to give rise to the expected migratory
insertion reaction. That is, there was no spectroscopic evidence for formation of the

expected carbene migratory insertion product 2.10 as illustrated in Scheme 2-5.

\ .Cp


C12 Cs

2P C2

Figure 2-1. Structure and labeling scheme for 2.03a with 50% probability of
thermal ellipsoids.

Cp. uCO
Ph Ru Cp CO


2.03a 2.10

Scheme 2-5

Instead, exclusively and quantitatively, a clean yellow product which exhibited
no characteristeric carbene carbon peak in its 3C NMR spectrum was formed, and in its
1H NMR spectrum the two diastereotopic protons of 2.03a had disappeared. However,
an unusual broad peak at 2.90 ppm and a single peak at 5.19 ppm with an integration
ratio of 1:1 were observed in its TH NMR spectrum. Mixing deuterated water with this
unknown product in an NMR tube led to the disappearance of the 8 2.90 peak, a result

which implied that this peak was due to the presence of a hydroxyl group in the
molecule. Both its mass spectrum and its elemental analysis revealed that the unknown
product was an isomer of compound 2.03a.
Based upon this evidence, a ruthenacene species 2.11a was proposed for the
stucture of this thermolysis product, as shown below.


2.03a, Ar = Ph 2.11a, Ar = Ph
2.03b, Ar = p-Tolyl 2.11b, Ar = p-Tolyl

The identity of the rearranged product 2.11a was confirmed by a single crystal
X-ray diffraction study of the analogous compound 2.11b.

An ORTEP representation of the structure is given in Figure 2-2. The Cp ring

and the indenyl ring are parallel with a dihedral angle of 0.3(3). The phenyl ring is not

conjugated with the indenyl ring; the dihedral angle between these two rings is 45.6(2).

The average bond distances of Ru-Cp and Ru-indenyl (five membered ring carbons )
are 2.158(4) and 2.203(4)A, respectively.

Ar = Q


Ar = p-Tolyl

Figure 2-2. Structure and labeling scheme for 2.11b with 50% probability of
thermal ellipsoids.

C19 C18

C21 C17

A proposed reasonable mechanism for the formation of 2.11 from 2.03 is given

in Scheme 2-6.


2 \co

CO insert

Spath b




Scheme 2-6


In this mechanism the five membered ring presumably undergoes initial CO
insertion to give an intermediate acyl complex 2.12. This ring expansion step is
interesting because it is an example of preferential carbonyl migratory insertion over
carbene insertion which is contrary to what appears (from very limited examples) to be
usual. For example, Werner28 found only carbene insertion in 2.17.


oc I

Jones5.7 observed a similar result in the cyclic carbene 2.18 which gave only
carbene migratory insertion in spite of the attendant cyclobutane strain.

Fe OMe


Cp L
Fe Me

On the other hand, with enough bias as in 2.198 and presumably in 2.03 (due to
ring strain in benzocyclobutene), carbonyl insertion has also been observed to occur.



-Oe Cp L o OMe

TOMe goy.-.OM e

It is not clear why the thermal liabilities of 2.03a and 2.03b are so dramatically
increased relative to their methoxy substituted precusor 2.04. One possible explanation
is the apparent increased electrophilicity of the terminal carbonyl resulting from
replacing the methoxy group with the aryl group. Consistent with this hypothesis is the
shift of the carbonyl stretching frequency from 1968.3cm-' for 2.03a to 1961.3 cm-1 for
Interestingly, the conversion of 2.12 to 2.15 could proceed via two different
mechanistic pathways. One mechanism (path a) involves an unusual carbene insertion
into a metal acyl bond, (for which only one example has been reported.29) to give a
fourteen electron intermediate 2.14, followed by tautomerization of 2.14 to give 2.15.
A second mechanism (path b in Scheme 2-6) would involve initial enolization followed
by rearrangement of the vinyl group to give 2.15. Neverthless, by either mechanism, it
is apparent that the driving force for the overall reaction is the stability of the
metallocenes 2.11, a kind of product which appears to be characteristic of reactions
where metallaaromatic rings may be intermediates.3-31 Indeed, path b in Scheme 2-6
is particularly intriguing because of its ruthenanaphthalene intermediate 2.13.
Since intermediates 2.12, 2.13 and 2.14 are electron-deficient and could be
stabilized by complexation with a good ligand, the thermolysis of 2.03 was carried out
in the presence of PMe3.

Ph Me3P Ph
( r Me3P, RT
Ru-- Ru
y CO A, 50 OC CO

2.03a 2.20

Unfortunately, it was found that even at room temperature in C6D6, PMe3
simply added rapidly to 2.03a to give the zwitterion adduct 2.20 in quantitative yield.
When 2.20 was heated to 50 oC, it simply dissociated the PMe3 ligand and regenerated
2.03a which then rearranged to 2.11 without showing any evidence of a PMe3 adduct of
2.13 [A small concentration (ca.5%) of 2.03a could be detected during the conversion
of 2.20 to 2.11].
Complex 2.03b reacts similarly with PMe3, undergoing nucleophilic addition to
give a zwitterion adduct analogous to 2.20.
Interestingly, complexes 2.03a and 2.03b exhibited different thermal

reactivities, with 2.03a rearranging at 50 oC while 2.03b rearranged at 70 OC. Such a
difference in reactivities is consistent with expectations based upon the better electron

donor ability of the tolyl group of 2.03b, which would tend to stabilize the ground state
of the carbene complex and make the carbonyl (CO) of 2.03b less electrophilic relative
to the CO of 2.03a. Therefore, the CO insertion of 2.03b should be slowed down in
comparasion to that of 2.03a.
In conclusion, a methoxy stabilized metallaindene complex of ruthenium 2.04

was synthesized. Reaction of this complex with phenyllithium or p-tolyllithium
followed by treatment with CF3SO2OSiMe3 provided the first stable aryl substituted
Fischer carbene complexes of ruthenium (2.03a and 2.03b) which do not bear a
heteroatom stabilizing substitutent on the carbene carbon. In solution, these complexes
underwent a novel thermal rearrangement to form ruthenoence derivatives 2.11a and


2.11b. A reasonable mechanism for this rearrangement which involves a
ruthenanaphthalene intermediate or carbene insertion into an acyl bond is proposed.


Ru.- CP

6 4 L 2
I. 4 7 a.. t.. 3 .2

Si .

Figure 2-3. 'H and "C NMR spectra of 2.04.


" CO

Figure 2-4. IH and 13C NMR spectra of 2.03a.

. I I j -- r I


Ar = p-Tolyl


Figure 2-5. 'H and 13C NMR spectra of 2.03b

w uO Ru

Ii~.IjI '-
I' r..~ vi. a~I;~~ l~i~I3--3 m1~ Ir1r1;;Trr

Figure 2-6. 1H and 13C NMR spectra of 2.11a

. $ 1 I ^


Carbene migratory insertion reactions have attracted considerable attention over
the years, especially those involving carbene complexes of later transition metals such

as iron and ruthenium. This is because such complexes are catalytically active in the

Fischer-Tropsch reaction, a reaction in which the carbene migratory insertion step plays

a significant role.

However, very few direct studies of the carbene migratory insertion have been
published to date.32 Jones and his co-workers have recently found that the rate of

carbene migratory insertion is strongly dependent upon the nature of the metal, and that

the rate of rearrangement for iron carbene complex 1.19 is more than 107 times faster

than that of ruthenium carbene complex 130.

Cp OMe Cp OMe Cp OMe


1.19 1.30 132

This observation suggested that it might be possible to synthesize an acyclic
ruthenium carbene complex such as 1.32 which would be analogous to the cyclic

ruthenium carbene complex 1.30 except that no strain factors would be present either in
the starting material or in the rearrangement product. If such a complex could be made,
then its migratory insertion could be directly studied.
However, for initial studies, complex 3.01 was selected with the idea that the
thermal stability of an aryl-substituted carbene complex should be greater than its alkyl-
substituted carbene complex analogue.33
The synthetic strategy for the preparation of acyclic Ru-carbene complex, 3.01
is given in Scheme 3-1.


O Cp\ /K /CO Cp CH31
Ru3(CO)2 R Ru Na/H Ru-
3.02 3.03

CpI Cp Q-Et
I 1)PhLi, -78oC \p OEt
Ru-CO 2)Et3BF4, RT OC-Ru=C

CH3 H3C Ph H Ph
3.04 3.01 3.05

Scheme 3-1

This strategy was satisfactory up to 3.04 (formed air-sensitive Rp-Methyl 3.04
in moderate yield) but when 3.04 was treated with phenyl lithium at 780C followed by
triethyloxonium fluoroborate (ethoxy was used rather than methoxy because it could be

made in our lab)48 at room temperature and the reaction mixture worked up, no carbene
complex 3.01 was observed; the only detectable product was 3.05 (the structure was
confirmed by comparison with an authentic sample).34 Although it was somewhat
disappointing that the carbene complex 3.01 was not observed, formation of 3.05
suggested that the carbene complex 3.01 had been present as a transient intermediate
but that migratory insertion as depicted in Scheme 3-2 was too rapid to permit isolation
of the carbene complex at ambient temperature.

Cp\ OEt
OC-Ru= C,
/ Ph

Cp OEt
cp\ IO
-* Rru- C-Ph
/ 1

Cp Et
\ C-Ph
- OC- r-l
/ CH

CH= Ch


Scheme 3-2

A logical mechanism for formation of enol ether 3.05 from 3.01 involves
migratory insertion of 3.01 to form the 16-electron species 3.06 which then undergoes
P-elimination to give t-complex 3.07. Dissociation of 3.07 would yield the enol ether


From these results it became clear that in order to detect and study migratory

insertions in acyclic carbene complexes of ruthenium, the reactions would have to be

carried out and examined at low temperature. Since isolation under these conditions

would be problematic, an apparatus that would permit direct monitoring of the reaction

without work-up was required. We therefore constructed and modified a special

apparatus which had been designed by Periani and Bergman35 as shown in Figure 3-1.

a f r i t

-780C I NMR tube

Figure 3-1. Specially-Designed Low Temperature Apparatus

Furthermore, since purification of ethoxy complex 3.01 at such low temperature

was not feasible, trimethylsilyl chloride was initially substituted for triethyloxonium

fluoroborate because it gave a somewhat cleaner reaction and also because the

hyperconjugative stabilizing effect of the O-Si bond should provide additional stability

to the carbene complex by retarding the rearrangement and thus increase the chance of

observing the carbene complex.

In carrying out these low temperature reactions, the starting material 3.04 in
diethyl ether was placed in the specially designed schlenk tube and was allowed to react

with phenyllithium at -78 oC. This was followed by warming the mixture to
approximately -30 oC with evaporation of ether solvent at this temperature over a six
hour period to gave the yellow ruthenium enolate salt 3.08. The enolate salt was then
treated with one equivalent of trimethylsilylchloride in deuterated methylene chloride at
-78 OC, a procedure which we expected would lead to formation of the desired carbene
complex 3.09 as shown in Scheme 3-3.

Cp\ CO Cp O-Li+ MeSiI Cp OSiMe
Ru OC-Ru=C OC-Ru=C
C -780C / \
CH3 H3C Ph H3C Ph
3.04 3.08 3.0

Scheme 3-3

Once the above described procedure was completed, the whole apparatus was
immersed in a -78 OC dry-ice/isopropanol bath, and the reaction mixture was then
filtered directly through a frit into an nmr tube which was then sealed at liquid nitrogen
The 'H and '3C NMR spectra of a typical run are shown in Figure 3-2. At -70
OC two Cp signals were observed in the 'H NMR spectrum at 5.064 and 4.964 ppm; the
signals coalesced to give a broad peak when the temperature was gradually increased to
-40 oC as shown in Figure 3-3. The broadness of the peak clearly indicated that
fluxional behavior was being exhibited in the system. The formation of metal hydride
(-8.6 ppm) and new Cp signals at 4.86 ppm also began to be observed at -40 OC with
loss of the original Cp signal as illustrated in Figure 3-4.

Cp O-Li+

H3C Ph

+ Me3SiCI I


s?3 II

I ...

Figure 3-2. 4H and 13C NMR spectra of a typical run at -78 OC attempting to
silylate the enolate 3.08. Two Cp signals (8 5.064,4.964 ppm)
were observed. No carbene carbon was observed in its
characteristic region.

'~~''~`~i`~~'~~-~l~ ''''~~'' J'''''~~~?' ~~`; m-l I. I.,,... I..,.,.,.,,.,,~~~~~ ~

Cp O-Li*
OC-Ru =C + Me'SiCl
/ \
H3C Ph

Figure 3-3. 'H NMR spectra of mixture of 3.08 + Me3SiCl at -40 OC (top)
and -70 OC (bottom).

Ph OSiMe3
CP\ \ /

I Ih

r~"-..- '7*~'' I

-40 oC

Figure 3-4. 'H NMR spectra showing formation of the x-complex 3.11 (metal
hydride 8 -8.6 ppm and new Cp signal 8 4.86 ppm) at -400C
(bottom). Two isomers of 3.11 can be seen at 0 oC (top).


The formation of the free enol ether 3.10 was detected upon further warming of
the reaction mixtures. (In order to verify the identity of 3.10, it was synthesized
according to a known literature procedure36).

\ ,OSiMe3 >- 40C OSiMe3
OC -u= C P CHhC
3.09 3.10

Formation of the metal hydrides 3.11 and the enol ether 3.10 left little question
that the carbene species 3.09 was being formed. Unfortunately, the two resonances that
would be expected to be the most characteristic of a carbene complex (a carbene carbon
at very low field; i.e. 270 350 ppm and a methyl carbon at high field; i.e. above TMS)
could not be detected in the 13C NMR spectra.
In order to assist us in seeing these peaks, it was decided to enrich both the
carbene carbon and the methyl carbon with 13C. The methods used for preparation of
these enriched Ru-carbene complexes are shown in Scheme 3-4.

Ru3(CO)12 13o Enriched Ru3(CO)2 -

CH3 Ph

CP\ Cp OSiMe3
OC-Ru + 13CH3I ---- Rp'CH PhLi OCR
/ Me3SiC1 I /
"CH3 Ph

Scheme 3-4

Indeed, when the enriched materials were used, resonances in both the carbene
region and the methyl region in the 13C NMR were observed. Since these signals were
in the expected regions, it was assumed that the acyclic ruthenium carbene complex
3.09 had been formed at -78 OC. Surprisingly, at -90 oC, multiple resonances in the
characteristic carbene carbon region were observed, as well as in the methyl region of
the spectrum as shown in Figure 3-5. These resonances coalesced upon warming the
reaction mixtures, a result which will be discussed in more detail later.

Upon warming these reaction mixtures, it was found that the decomposition
product 3.10 was the same as from those without enriched carbon. The results, thus,
seemed consistent with carbene complex formation.

Although the low temperature synthesis of this acyclic ruthenium carbene
complex, 3.09, was technically difficult, the preparation was actually quite clean. It was
presumed that, at -40 oC, the migratory insertion process of the ruthenium carbene
complex 3.09 was being observed as depicted in Scheme 3-5. In a process similar to
that outlined in Scheme 3-2 for the ethoxy complex, trimethylsiloxy complex 3.09
could be expected to give a 16-electron intermediate which would undergo 0-
elimination to produce t-complexes 3.11, which were the first products actually to be
observed. These i-complexes which exist as isomers, then, could be observed to

dissociate to give the free enol ether product 3.10.

Cp O-Li+
\ I
"CH3 Ph

+ Me3SiCI

S30 i295 290
300 295 290

-24 -26 -28 -30

Figure 3-5. '3C NMR spectrum of enriched 3.08 with Me3SiCl in the
expected carbene carbon and methyl regions at -90 OC.

Cp OSiMe3, CP OS Cp OSiMe
OC-R=C > 40C
CH \ 0 Au- v-Ph OC-U-IH
3.09 16-electros [ 8=-8.51 ppm

I 3.11b

CH2 C CH2= Cph
8 = 8.65 ppm 3.10

Scheme 3-5

An initial goal of this research was to study the migratory insertion process.
Attempts were therefore made to study the kinetics of the reaction that was occurring at
-40 OC Unfortunately, this led to frustrating results. At first, we were quite puzzled by
the fact that we could not get reproducable kinetic data, with each attempt to treat the
process as first order leading to curved lines, (a result which in the end caused us to
suspect that the observed process might not be first order.) Slowly we realized that
instead of observing first order decomposition of 3.09, we might be observing rate-
determining reaction of lithium enolate 3.08 with trimethylsilylchloride to form the
carbene complex 3.09 which then rearranges rapidly, without detection, to the products
of migratory insertion (3.11 and 3.10) as illustrated in Scheme 3-6.

SC\ ,O-Li
OC-R=C. +

OSiMe3 slow

Me3SiCl r.d.step

\ ,OSiMe3
OC-u C, Ph




To verify this assumption, a control experiment was carried out in the same
manner as described in Scheme 3-3 except without addition of trimethylsilylchloride.
In other words, the starting material 3.04 was allowed to react with phenyllithium at -78
oC followed by removing solvent at -30 OC to give the lithium enolate 3.08. This enolate
was then dissolved in the deuterated methylene chloride at -78 OC and transferred into
an nmr tube for the comparative study.
As a result and to our amazement, it was found that the 13C NMR signals of the
lithium enolate 3.08 were virtually identical to those observed for the earlier presumed
carbene complex 3.09 at -90 oC as illustrated in Figure 3-6. It would require quite a
coincidence if the lithium enolate 3.08 and the carbene complex 3.09 were to have
identical chemical shifts in the NMR.

Cp O-Li
H3C Ph

Cp\ O-Li
H3C Ph

I', I [ i

U A.

Figure 3-6. 13C NMR spectra of Li-enolate 3.08 (top) and presumed carbene
complex 3.09 (bottom) at -90 oC.

This remarkable result induced us to compare the 13C NMR spectra of an
authentic ruthenium lithium enolate with its alkylated carbene counterpart (results from
chapter 2). For this purpose we chose the isolable ruthenium indene complex 2.04
whose substituents are analogus to those in the acyclic carbene complex 3.09 alkyll on
metal; aryl on carbene carbon) but which is unable to rearrange. This could be prepared
from its lithium enolate precusor, 3.12 (Scheme 3-7).


',Cp Et2O, -l10C Me3
I -
R \ n-BuLi \ CO
2.05 3.12 2.04

Scheme 3-7

The 2-Bromobenzyl ruthenium complex 2.05 was treated with 1.2 equivalents of
n-butylithium at -1000C, followed by warming to room temperature overnight.
Evaporation of solvent gave the reddish lithium enolate 3.12. Methylation of 3.12 was
accomplished by treatment of 3.12 with excess Me3OBF4 followed by purification and
isolation to yield 2.04. Interestingly, the 13C NMR spectra of 3.12 and 2.04 in
deuterated methylene chloride revealed that the chemical shifts of the carbene carbon
signals of the lithium enolate 3.12 and those of the methylated indene complex 2.04 are
very similar, as shown in Figure 3-7.
This result was somewhat surprising because it was expected that the chemical
shifts of a lithium enolate would be quite different from those of an alkylated carbene


ii I,

I ~ i I K i
-i -! -l .3

Figure 3-7. 13C NMR spectra of Li-enolate 3.12 (Top) and methylated indene
complex 2.04 (bottom).

complex. Normally, acyl carbons of neutral complexes appear at much higher field
than carbene carbons (-230 ppm vs -350 ppm), although examples of anionic acyl
complexes37, such as 3.13, have now been found which exhibit chemical shifts
approaching the value we have observed for 3.08 and 3.12.


"C="Fe(CO)4 13Cnmr (5 TMS) 279.7 ppm in THF

We, therefore, interpret these nmr results (the comparison of spectra of lithium
enolates 3.12 and 3.08 with those of alkylated carbene complex 2.04) as indicating that
the spectra of anionic acyl complexes derive from tightly ion-paired species which are
non-acyl-like, but rather structurally more closely resemble their alkylated carbene
complexes counterparts.
An additional aspect of the enriched lithium enolate 3.08 proved to be very
intriguing. As mentioned earlier, a number of resonances were detected in the
characteristic carbene carbon region at -90 C (Figure 3-5). These signals exhibited a
temperature dependence with coalescence being observed upon warming the lithium
enolates to -50 OC. The results from this 13C NMR temperature-dependence experiment
are illustrated in Figure 3-8.
To be certain that the phenomenon which was being observed at -50 OC was not
simply decomposition, the sample was cooled back down to -80 OC, and the original
spectrum was seen to return. This indicated that one was observing a dynamic
equilibrium process during this experiment. The enriched methyl resonances also
exhibited a similar temperature dependence with their coalescence occurring at -40 oC.

Cp Li+
[ \ ,O Li


Figure 3-8. "13C NMR Temperature Dependence Experiment of Li-enolate
Fiue -.'^^^ s* '-^aueDeedne xeimn finlt

*"** lA

The origin of these coalescing signals is unknown. In principle, such multiple peaks

could originate either from different conformations about the metal carbon bond or from

different lithium aggregates. The spectra of the lithium enolate 3.08 were found to be

markedly dependent on the nature of the cation and the nature of the solvent. It was

found that the multiple peaks in the characteristic carbene carbon region at -90 OC were
reduced to two peaks (8 292 and 282 ppm in CD2C12) upon addition of ten percent

deuterated tetrahydrofuran, a solvent which would be expected to coordinate with

lithium (as shown in Figure 3-9, top). In another experiment the multiple peaks were
reduced to a single resonance (5 284 ppm in CD2CI2) when 12-crown-4 ether was

added. Such a result could have derived from formation of either the free ion or a very
loose contact ion pair 3839 (as illustrated in Figure 3-9, bottom).

These results are most consistent with aggregates of lithium enolates being the

source of the multiple carbene carbon signals. However, conformational equilibra can

not be rigorously excluded since relative conformational stabilities might well be

affected by change of solvent or presence of additives such as 12-crown-4 ether (e.g. a

change in orientation of a phenyl substituent).

It now seems clear that what we have observed in our low-temperature attempt
to prepare and isolate acyclic ruthenium complex 3.09 was instead the formation of the

lithium enolate 3.08 at -90 oC, which when warmed to -40 C underwent rate

determining silylation to give the desired ruthenium carbene complex 3.09. Since

careful analysis of 13C NMR spectra upon warming to -40 oC did not lead to detection

of any transient species [only the gradual growth of peaks deriving from metal hydride
species 3.11 could be observed (as shown in Figure 3-4)] it was, therefore, concluded

that even at -40 C the carbene complex 3.09 rearranges too rapidly to the metal hydride

species 3.11 to be detected.

Cp ,OLi-
OC-u=C. Ph

i I i

Figure 3-9. 13C NMR spectra of Li-enolate 3.08 at -90 OC in the presence of
10% THF-dg (top) and in the presence of 12-crown-4 (bottom).

It should also be mentioned that the lithium enolate species 3.08 (in the absence

of silylating agent) was found to be stable at temperatures up to 0 OC. Therefore, 3.08

could not itself be the species which is undergoing the migratory insertion reaction at

ca. -40 C.

This result reinforced our conclusion that acyclic ruthenium carbene complex

3.09 was formed at -40 oC but rearranges and disappears too rapidly to be observed.

Since the rate determining step is silylation of the anionic oxygen, and this

apparently occurs at -40 OC in the reaction of the lithium enolate with trimethylsilyl

chloride, it is clear that the observation of an alkyl substituted carbene complex and the

consequent ability to monitor its migratory insertion would require alkylation at a

temperature below -40 OC. Therefore, better silylating or alkylating reagents were

sought. Reagents which were tried included trimethylsilyl triflate and triethyloxonium


Indeed, it was found that when trimethylsilyltriflate was used as the silylating

agent the metal hydrides 3.11 were formed even at -78 OC. The NMR spectra indicated

that no significant lithium enolate remained. Therefore, it was concluded that

trimethylsilation of the lithium enolate must have occurred at -78 OC with this powerful

reagent and that the acyclic ruthenium carbene complex 3.09 was not stable even at this


While attempting to use triethyloxonium tetrafluoroborate as the alkylating

agent, it was found that lithium enolate 3.08 remained and thus did not react with this

reagent at -78 OC. However, the lithium enolate did react with triethyloxonium

tetrafluoroborate and the metal hydrides were detected when the reaction mixture was

warmed to -40 OC.

In order to make the acyl anion a better nucleophile by reducing the bonding
between the lithium cation and acylate anion, an attempt was made to solvate the
lithium ion with 12-crown-4. The experimental results revealed that the reactivity of

the lithium enolate was enhanced in the presence of 12-crown-4. Chelation of the

lithium ion with 12-crown-4 apparently provided a more nucleophilic enolate oxygen so

as to allow reaction with triethyloxonium tetrafluorobroate even at -78 oC. However,

the expected acyclic ruthenium carbene complex 3.01 was not observed even under
such conditions. Again, the initial products that were observed (-78 OC) were the metal

hydrides 3.07. This result confirmed that ion-pairing phenomena can play a significant

role in the chemical reactivity of ruthenium carbonyl anions, but unfortunately our goal

of isolating an alkylated or silylated acyclic carbene complex was again thwarted.

Although we have not been successful in isolating and studying the kinetics of

the migratory insertion of either silylated or alkylated acyclic ruthenium carbene
complexes, we have nevertheless obtained insight into their reactivity in that apparently

such complexes rearrange rapidly even at temperatures as low as -78 OC.


In one of our earlier studies, it was found that incorporation of ring strain into
the potential rearrangement product of an alkyl-substituted carbene complex
significantly retarded migratory insertion. That is, migratory insertion of the ruthenium
carbene complex 2.04 was not observed presumably because of the 49 kcal/mol of ring-
strain which would have resulted from such ring contraction.

Ru- CO C=Ru -CO

C 6 CH3

2.04 3.09

In contrast, migratory insertion of acyclic analogue 3.09 was so facile that it
could not be spectroscopically observed when it was presumably formed at -78 C.
Thus, the 49 kcal/mol strain which would have been incorporated into the
product of migratory insertion of the cyclic carbene complex 2.04 apparently inhibited
the rearrangement too much to allow any observation of the process. (It places the


reaction out of sight thermodynamically.) But, when the strain factor was removed

entirely in an otherwise completely analogous structure, the migratory insertion process

was so rapid that the carbene complex 3.09 could not be observed at -78 "C.

Therefore, it is obvious that something in between these two extremes was

needed. In other words, something which would either destabilize the rearrangement

product or stabilize the carbene complex but not so much as to totally inhibit migratory


It occurred to us that by incorporation of a metallacarbene functionality into an

aromatic system, there might be enough resultant stabilization of the carbene system to

retard such migratory insertion and to allow the observation and kinetic study of the

metallacarbene aromatic species.

Indeed, theoretical calculations predict that a selected transition metal can take
part in (d-p) xi aromatic bonding, if it has a centrally directed d orbital of appropriate

energy and overlap to provide an electron pair to the x-system of a carbon framework

that has a low lying vacant MO.40



S~~ 2 HOMO

P/ q9
P= PMe3

Scheme 4-1

The picture as illustrated in Scheme 4-1 has generally been utilized to depict
incorporation of a selected transition metal into a six-membered aromatic ring. As
pointed out by Hoffmann,40 the LUMO of the organic fragment in a metallabenzene is
considered to be effectively a pentadienyl cation which can be a i-acceptor of d-
electrons to complete an aromatic sextet
To date, however, successful preparations of transition metal metallabenzenes
have been extremely rare. Bleeke'9h and his co-workers in 1989 reported the best
example of a metallabenzene (iridabenzene) 1.11 with all of the usual aromatic
properties. In particular, the six-member ring is planar with delocalized bonding around
the ring.

P Ir (
P = PMe3

As a matter of fact, this metallabenzene complex 1.11 is of a third row transition
metal. Related examples of first or second row transition metal metallaaromatic
molecules such as metallabenzenes, metallanaphthalenes or metallaphenanthrenes, to
our best knowledge, have not yet been reported.
The methodology which was used to prepare acyclic ruthenium carbene
complexes (3.01 and 3.09 from chapter 3) provided the original inspiration for our

synthesis of the second row late transition metal metallaaromatic species,
ruthenaphenanthrene 4.01.

\ /P
C= u-CO




Conceptually, the construction of the ruthenaphenanthrene target required
modification of the acyclic carbene complex 3.09 by first replacing the methyl with a
phenyl substituent (which would create acyclic carbene complex 4.02) and secondly
joining the two phenyl substituents at the 2,2'-positions by a single bond (which would

create the potentially aromatic carbene complex 4.01).
Thus, we decided to prepare the desired ruthenaphenanthrene 4.01 to see if there
is any aromaticity to slow down the migratory insertion enough to observe the carbene

The proposed synthesis of ruthenaphenanthrene 4.01, is presented in Scheme 4-

Cp, Co
Br Br Br Ru Br
B n-BuLi B 1) leq,n-BuL,-780C
Br THF 2) CpRu(CO)2Br
4.04 4.03

2eqt-BuLi Cpp O C OR
Et2O, -780C -78RC Ru

R =(a)Et (b)SiMe3; (c)Me

Scheme 4-2

2,2'-Dibromo-biphenyl was prepared according to the literature.47 The a-

complex 4.03 was generated in moderate yield via addition of one equivalent n-

butylithium at -78 OC to produce the mono-lithio anion followed by reaction with RpBr

(4.04). The complex 4.03 was then placed in the special low temperature apparatus

described earlier (see chapter 3 page 35), and was allowed to react with t-butyllithium at

-78 OC overnight. This reaction mixture was then evaporated to dryness at
approximately -30 OC to afford the reddish-yellow lithium enolate 4.05. Alkylation (or

silylation) of this lithium enolate in CD2Cl2 at -78 OC would be expected to give the

desired product ruthenaphenanthrene 4.01.

The tH NMR spectrum of the alkylated reaction mixture (R=Et) revealed several

broad signals in the Cp region at -90 oC and a coalesced broad Cp signal when the

temperature was raised to -60 OC (Figure 4-1). From our previous experience, the
broadness of the Cp signals

co ,e @0

+ Et30BF4


Figure 4-1. 'H NMR spectra of mixture of enolate 4.05 and Et30BF4
at -90 oC (bottom) and -60 oC (top). (signal at 3.95 ppm is
unreacted Et30+).

[Cp I ?O OEtR \
+ Et3OBF4 -------



Figure 4-2. tH NMR spectra of reacting mixture of enolate + Et 30BF4
at -50 OC. The original Cp signal at 8 5.007ppm slowly
disappeared, and a new Cp signal (8 3.75 ppm) and a doubt at
5.48 ppm (Ha) appeared.

was recognized as an indication of fluxional processes within the molecule. However,

at this point the origin of the fluxionality was unknown. Regardless, if the new
complex was the ruthenaphenanthrene, it should either be stable enough for
characterization or it should undergo rearrangement in a predictable way. Indeed, as a
sealed sample of this reaction mixture was gradually warmed to -50 C, a number of
spectral changes were observed. The original Cp signal at 8 5.04 ppm in the IH NMR

spectrum slowly disappeared, and appeared to convert to a new Cp signal (8 3.75 ppm)

as illustrated in Figure 4-2. Eventually, new and quite clean 'H and t'C nmr spectra
were obtained for the product of this reaction (as seen in Figure 4-3). An unusual
upfield proton (8 5.48 ppm) which was coupled with aromatic protons demonstratedd via

the decoupling experiment depicted in Figure 4-4) eliminated the possibility of the
product being 16e intermediate 4.06 because such a 16e intermediate should possess

only aromatic protons. Thus, the detected intermediate is probably derived from 4.06.

Cp. .OR RO ,p OR
Ru- -500C /
Ij .H8
insertion /

4.01 4.04.07
R= (a) Et; (b) SiMe3;(c) Me

The structure of this product is proposed to be the 1r3-r-benzyl complex 4.07.

The doublet belongs to Hg, part of the benzylic system, which assignment was made on
the basis of comparisons with analogous known compounds41 and 'H NMR decoupling


Me3SiO 0C

Figure -H ad N

Figure 4-3. 1H and 13C NMR spectra of 4.07b.


Me3SiO Oc

.. H,

I y :r

Figure 4-4. 'H NMR spectra of Decoupling Experiment for 4.07b.


experiments. The one upfield CH carbon and seven CH carbons in the aromatic region

instead of eight aromatic CH carbons in the 13C NMR spectra are consistent with this

proposed structure (Figure 4-3).
Brookhart41 published an analogous 13-benzyl (T13-CH2C6Hs) complex
Cp(CO)Fe(Tl3-CH(OCH3)C6Hs) 4.08, which he and his coworkers characterized by 'H

and 13C NMR spectroscopy. The ortho hydrogens of the phenyl ring Ho and H.' have

very different environments in 4.08 with chemical shifts of 2.45 ppm (Ho) and 7.85 ppm


OCH, /
OC3 Fe..C 'H NMR (Toluene-ds)

A H,,
8 2.45 ppm (Ho; d, J = 6 Hz)
Ho 68 7.85 ppm (HO; d, J =6 Hz)


The structure of the t3-benzyl complex 4.07c (OR=OMe) was not only

confirmed by NMR spectroscopy as shown in Figure 4-5, but it was actually isolated by
chromatography at -20 OC as a dark-red crystalline solid when the lithium enolate 4.05

was treated with trimethyloxonium tetrafluoroborate in a relatively large scale
experiment As with 4.07a, the 'H NMR chemical shift of the benzylic hydrogen (Hs)
of 4.07c appeared as a distinct doublet (J=6.45 Hz) at 5.57 ppm. In the IR spectrum of
the T13-benzyl complex 4.07c, a terminal CO absorption at 1967.6 cm-1 was observed.

The structure of 4.07c was further verified by elemental analysis.



Figure 4-5. 'H and 13C NMR Spectra of 4.07c

As a further confirmation of the assigned structure, reaction mixtures were
allowed to warm up in the presence of carbon monoxide and the expected stable
trapping product 4.09 was formed, (isolated at room temperature). This complex was
fully characterized by NMR spectroscopy (as illustrated in Figure 4-6), elemental
analysis and high resolution mass spectrometry.

\ o Cp oc 0 ,Cp
Eto R OEt C\ EtO Ru,

-780C- RT
4.06 4.07 4.09

Characterization of the derived T13-benzyl complex 4.07 and trapping product

4.09 led us to believe that the ruthenaphenanthrene 4.01 had been formed and had
undergone migratory insertion rapidly at -50 OC to give these products.
However, the direct observation of 4.01 is desirable. The way to do that is to
look for the characteristic low field 13C resonance of the carbene carbon. From our
previous experience, we knew that the carbene carbon would be difficult to detect
without enrichment by 13C. Therefore, it was decided to enrich the carbene carbon,
much as we had enriched earlier studied acyclic carbene complexes 3.01 and 3.09. The
procedure which was used to prepare the enriched ruthenaphenanthrene is presented in
Scheme 4-4.


c\ Cp
EtO Ru


Figure 4-6. H and 13C NMR Spectra of 4.09.


~ ;C ~ ~ ` ` ;Id ~`''~~`;b~``~'' ;:o~ ` ` ~ ~ ;dd ~ ~ 'Ib' ` ''~'~b''~~'''''~'`~ : '~';b~R;~'

3CO Enriched 1)0 1 ?
Ru3(CO)2 Ru3 (CO)2 --- OC- Ru- Br 13
2)CBr4 I Cp O1 OF
CO Ru<

BrBr Li Br

Scheme 4-4

As in the case of the acyclic carbene complex, at -90 OC, multiple resonances in

the characteristic carbene carbon region were observed in the '3C NMR spectra when
the enriched materials were used (Figure 4-7). These resonances also coalesced upon
warming, and upon further warming, the Tl3-benzyl complex rearrangement product

4.07 was detected which was consistent with the results from the non-enriched study.

As in the earlier studies of the acyclic carbene complexes 3.01 and 3.09, we
could not be certain whether we were observing the lithium enolate or the ethoxy

carbene To address this question, a control experiment in which no alkylating agent
was present was undertaken. From this experiment, it was found that the chemical
shifts in the 13C NMR spectrum without alkylating agent were nearly the same as with

the alkylating agent at -90 OC. It was additionally found that the lithium enolate 4.05

was stable at temperatures up to 0 OC. Lithium enolate 4.05 could thus not itself be the

species to undergo migratory insertion at ca.-50 OC. Therefore, it was concluded that, as
in the case of the acyclic ruthium carbene complex, what we were observing in the

below -50 OC attempt to synthesize ruthenaphenanthrene 4.01 was instead formation




300 250
300 250

Figure 4-7. 3C NMR Temperature Dependence Study of Mixture of 4.05 +

of the lithium enolate 4.05, which when warmed to -50 C underwent alkylation to
afford the desired "aromatic" complex 4.01.
When reaction mixtures were carefully followed by tH and 13C NMR
spectroscopy with a gradual increase of the temperature, no transient species could be
detected (instead only the growth of signals deriving from T13-benzyl complex 4.07 were
seen). It was therefore concluded that even at -50 OC the desired complex 4.01
undergoes migratory insertion reaction to the Tl3-benzyl complex 4.07 too fast to be
This result means that there is insufficient aromatic stabilization present in the
ruthenaphenanthrene 4.01 to prevent rapid rearrangement at -50 OC. In other words,
there was no evidence for the type of aromatic stabilization in 4.01 which we had hoped
might sufficiently retard its migratory insertion reaction to permit direct observation of
the rethenaphenanthrene.
Since the middle ring of phenanthrene 4.10 itself is not as aromatic as benzene
4.11 (20 kcal/mol vs 36 kcal/mol), it was then decided to incorporate the ruthenium into
a non annelated benzenoid system with the idea that, in such a case, there might be
enough stabilization to retard the rearrangement to the point that the ruthenabenzene
could be detected.

db o db
4.11 4.12
energy 92 kcal/mol 36 kcal/mol 72 kcallmol


Thus, our final goal was focused on the synthesis of a ruthenabenzene. Neil T.

Allison* has been interested in metallabenzenes. He reported30a-c formation of 1,3-

diphenyl-2-methoxyferrocene (4.13) via an apparent conversion from
ferracyclohexatriene or ferrabenzene (4.14) several years ago, although 4.14 could not

be directly observed.

Li Li CsHsFe(CO)2
Ph i Ph N

R.T. I

MeO Ph


Dr. Allision provided us with some of his previously synthesized precusor 4.15,

(1Z,3Z)-1,4-dibromo-l,4-diphenyl-l,3-butadiene, and we have thus pursued this project

in collaboration with him. The synthesis of ruthenacyclohexatriene or ruthenabenzene

4.16 was carried out as depicted in Scheme 4-4.

A former graduate student of Dr. William Jones who currently is a professor at
University of Arkansas.

nBuLi /C 2 eqt-BuLi
Br-78C Cp Ru Br EtO, -789C
Ph Ph ] Ph Ph-
4.15 4.17

CP Ru Li CDCI,-78
PL Ph Pl P PLPh Ph


Scheme 4-4

The stable T l-4-bromo-1,4-diphenylbutadienyl complex of ruthenium 4.17 was
prepared in a manner similar to that of complex 4.03 (Scheme 4-2). Addition of two
equivalents of t-butyllithium to a cold (dry ice/2-propanol bath) solution of 4.17 in
diethyl ether for lithium exchange, followed by cyclization, afforded the reddish residue
of ruthenaaromatic lithium enolate 4.18. Treament of this enolate 4.18 with Et30BF4 in
CD2C12 at -78 oC should lead to the desired product ruthenabenzene 4.16.
Again, at low temperature, we observed the spectrum of a species which, as it
warmed up in the presence of carbon monoxide, gave the same kind of trapping product
4.19 as we obtained from the ruthenaphenanthrene 4.01.
The isolation and characterization by high resolution mass spectroscopy of the
trapping product 4.19 (Figure 4-8) implied that the ruthenabenzene had been formed
which rapidly underwent migratory insertion.

" *t .. .. ...J ... ...^ 1 ,,,,, .I i

200 150 100 5

Figure 4-8. '3C and IH NMR spectra of trapping product 4.19.

Cp Cp Cp
c Et Migraty CO E Ru
Ru Insertion E C
Ph Ph 30 Ph Ph

16 e's

To conclusively demonstrate the presence of the ruthenabenzene, a unique
carbene carbon was sought in the 13C NMR spectrum using enriched materials. First of
all, a sample containing only the lithium enolate 4.18 was studied by NMR
spectroscopy. A broad signal at 8 293.8 ppm was observed at -78 OC in the 13C NMR
spectrum which is illustrated in the Figure 4-9. To this sample was added excess
alkylating agent (Et30BF4), and the reaction mixture was allowed to react for forty
minutes at -78 oC, followed by NMR spectroscopic analysis. At -70 oC, the signal of the
unreacted lithium enolate 4.18 was seen to be shifted from 8 293.8 ppm to 8 273.9 ppm
and a small sharp new peak at 8 290.8 ppm was observed as illustrated in Figure 4-10, a
peak which we believe is due to the desired alkylated ruthenabenzene 4.16. By
carefully monitoring the 13C NMR spectrum as the temperature of the reaction mixture
was gradually increased, a series of changes were detected in the 270-300 ppm region
as illustrated in Figure 4-11. As the temperature was increased to -50 oC, the signals
due to the presumed ruthenabenzene grew at the expense of those of the unreacted
lithium enolate. Eventually the peak at 273.9 ppm completely vanished, and there
remained only the carbene carbon peak at 8 290.8 ppm due to the ruthenabenzene.
Indeed, in the terminal carbonyl (CO) and Cp regions, corresponding new peaks at 8
201.2 ppm and 8 88.86 ppm respectively, were also observed, as illustrated in Figure 4-

CpR e

u Ph
Ph -i / Ph

Figure 4-9. 3C NMR sprctrum of Li-enolate 4.18 at -78 C.


*""T"'Z Z' = 1


50 PP

ocP o
Ph \ / Ph

Cp OEt
+EtOB~~P Ru Ph
+Et3OBF4 -omPh Ph

llllm lli i
i1 00 20ffi PP

Figure 4-10. 13C NMR spectra of Li-enolate 4.18 (bottom) and alkylated Li-
.enolate (formation of ruthenabenzene 4.16) (top) at -70 C.

Ph Ph

Figure 4-11. 13C NMR spectra of formation of ruthenabenzene 4.16 at -70
C, -70 C an hour later, -50 C, respectively, (From bottom to
top) in the Carbene carbon region.


I Ph Ph




top) in the CO and Cp regions.

In order to further verify the formation of the ruthenabenzene 4.16, changes in

the 'H NMR spectrum were also carefully examined as the reaction mixture was

gradually warmed from -70 OC to -50 oC. The results, shown in Figure 4-13, revealed a

number of changes which were entirely consistent with what was observed in the 13C

NMR spectrum. At -70 OC only a broad Cp signal of the lithium enolate 4.18 (in the
absence of alkylating agent) at 8 4.95 ppm was observed (spectrum a). After addition

of excess alkylating agent to this lithium enolate, a new Cp peak at 8 5.32 ppm (Cp of

4.16) was detected at -70 OC (spectrum b). At -50 OC the ruthenabenzene 4.16 was seen

to be completely formed from its enolate (spectrum e). New multiple peaks around

4.55 ppm 4.75 ppm in the OCH2 region were also observed.

The ruthenabenzene was stable at -50 OC for at least a half hour. However, it

underwent migratory insertion at -30 oC. The rate of decomposition of the

ruthenabenzene 4.16 was measured by monitoring the Cp absorption (5.32 ppm) in the

'H NMR spectrum at -30 OC, as shown in Figure 4-14. [The excess Et20 (3.52 ppm)

peak was used as a reference]

p OEt Cp Cp
OC Migratory O ,CO E nCO
RuInsertion Et Ru
PhP -300C Ph Ph Ph Ph


Cp &19 Cp
Ruj RU=
Ph- Ph + Et3OBF4 Ph P








Figure 4-13. (a) 'H NMR spectrum of Li-enolate 4.18 ; (b) IH NMR spectrum
of 4.18 upon addition of Et3OBF4 (formation of ruthenabenzene
4.16) at -70 oC; (c) mixture at -70 oC half hour later; (d) mixture
at -70 OC an hour later; (e) mixture at -50 oC.

Cp O Cp
O OEt Migratory C EO CO
Ru= Insertion L tO R [ Ru

16 e's




Figure 4-14. 'H NMR spectra of ruthenabenzene 4.16 undergoing migratory
insertion at -30 oC. (Loss of Cp peak at 5.322 ppm observed.)

In contrast, to our attempted kinetics in Chapter 2 where we thought we were

examining the first-order reaction only later to find that it was a second-order reaction,
decomposition of the ruthenabenzene (decrease of the Cp absorption) followed
excellent first-order kinetics with a rate constant of approximately 3.7 10 4 0.09 *

10 4 sec&' as illustrated in Figure 4-15.

Rate of decomposition of ruthenabenzene

Cp (peak high) vs Time (s)


Figure 4-15. A Plot of Ln (Ho/H) YS. Time (sec) for Determination of the Rate
of Decomposition of Ruthenabenzene.

This observation of good first-order kinetics also supports the contention that the

ruthenabenzene 4.16 was indeed formed and was the species undergoing



Peaks from the migratory insertion product 4.23 [8 200.4 ppm (CO) and 90.2

ppm (Cp) in the 13C NMR spectrum] which resemble those of the l3-benzyl complex
4.07 were detected. At 0 oC, the 113-allylic complex 4.20 survived for only a short

period of time.

In a final control experiment, it was shown that the lithium enolate was itself

stable up to 0 OC. This result further confirmed the formation of ruthenabenzene at -70


In light of all of these results, we believe that our claim of the intermediacy and

detection of the desired ruthenabenzene 4.16 is, in this case, secure.

In conclusion, aromaticity in the ruthenaphenanthrene was not sufficient to slow

migratory insertion to the point that the carbene complex could be observed. In

contrast, aromaticity in a diphenyl substituted ruthenabenzene appear to be sufficient

to permit detection of the ruthenabenzene at -50 OC.


The purpose of this research has been to study the migratory insertion

processes of ruthenium carbene complexes.

A methoxy-stabilized metallaindene complex of ruthenium (2.04) was prepared.
Reaction of this complex with phenyllithium or p-tolylithium followed by treatment

with trimethylsilyl triflate afforded the first stable aryl substituted Fischer carbene

complexes of ruthenium (2.03a and 2.03b) which do not have a heteroatom stabilizing

substitutent on the carbene carbon. In solution, these complexes underwent a novel

thermal rearrangement to form ruthenacene derivatives 2.11a and 2.11b. A possible

mechanism for this rearrangement which involves a ruthenanaphthalene intermediate or

carbene insertion into an acyl bond is proposed.

Attempts to isolate and study the kinetics of the migratory insertion of acyclic

ruthenium carbene complexes, CH3(CO)Ru=C(OR)C6H5 (R = SiMe3 & C2H5), were

not successful. Neverthless, our study provided some valuable information regarding

their reactivity, including the fact that apparently such complexes undergo migratory

insertion rapidly at temperatures as low as -78 OC.

Introduction of potential aromaticity into such carbene complexes by synthesis
of a ruthenaphenanthrene (4.01) was insufficient to slow migratory insertion to the point

that the carbene complex could be observed. The rapid rearrangement of 4.01 at -50 OC

was observed.


On the other hand, introduction of a greater degree of aromaticity into a

ruthenium carbene complex through the synthesis of a diphenyl substituted

ruthenabenzene (4.16) was sufficient to permit its detection at -50 OC. 4.16 is the first

second row, late transition metallabenzene which has been reported.


All reactions were carried out under an atmosphere of purified N2 by using

Schlenk tube techniques and a dry box. All solvents were dried by known procedures and
distilled under nitrogen prior to use. Trimethylsilylchloride was distilled over proton

sponge. PhLi, o-bromobenzylbromide, and CD2C12 were purchased from Aldrich

Chemical Company. 13CO and 13CH3I were purchased from ISOTEC INC. Ru3(CO)12
was purchased from Strem Chemical. Proton and carbon NMR spectra were recorded on
a Varian VXR-300 (300 MHz) or a Varian Gemini-300 (300 MHz) spectrometer.

Chemical shifts were referenced to the residual protons of the deuterated solvents and are

reported in ppm downfield of TMS for 'H and 13C NMR spectra. Mass spectra were run

on an AEI MS-30 spectrometer or Finigan MAT 950. Infrared spectra were recorded on

a Perkin Elmer 1600 FTIR spectrometer. Photolysis was carried out using a 450-w low
pressure Hg-Hanovia lamp in a pyrex well. The alumina used was neutral, activity I,
which was deactivated to activity Il and degassed on the vacuum line prior to use. The

chromatographic separations were accomplished by the flash chromatography method of

The following compounds were prepared as described in the literature, without
any modification: Rp2 or (CpRu(CO)2)245 CH2=C(Ph)(OSiMe3),36
CH2=C(Ph)(OCH2CH3)34, enriched Ru3(13CO).46 2,2'-dibromobiphenyl,47 and
triethyloxonium tetrafluoborate.48 The RpMe (CpRu(CO)2CH3)was prepared as in the

literature but with modification.49

Preparation of 2-Bromobenzyl Ruthenium complex 2.05

A solution of 2.2 g (4.95 mmol) of [CpRu(CO)2]2 in ca. 300 mL of THF was

treated with ca. 1.1 g of Na and 11 mL of Hg (as amalgam). The mixture was stirred for

9 h at RT, and then the dark grey solution was transferred into a Schlenk tube. To this
solution, cooled to 78 oC, was added 2.48 g (1 eq, 9.92 mmol) of 2-

bromobenzylbromide, and the mixture was stirred over night at RT. The solvent was

then removed under vacuum, and the residue dissolved in a small amount of hexane and

eluted through an alumina (HI) column with hexane, to give a colorless solution.

Evaporation of most of the solvent and crystallization at ca. -15 OC, gave 2.56 g of the 2-

bromobenzyl ruthenium complex 2.05 (66% yield) as a white crystalline solid. Anal.
Calcd for C14HIIO2BrRu: C, 42.86; H, 2.81. Found: C, 42.87; H, 2.80. Mp: 52-53
OC. MS (EI, m/e ): 391.89 (M+l). 'H NMR (C6D6): 8 4.46 (s, 5H, Cp), 3.03 (s, 2H),

6.60 (t, 1H), 6.98 (t, 1H), 7.28 (d, 1H), 7.5 ppm (d, 1H). 13C NMR: 8 210 (CO),

154-121 (phenyl region), 88.39 (Cp), 33.32 ppm. IR (hexane): CO 2025.1 cm-1, 1989


Synthesis of Benzannelated Ruthenium-Carbene Complex 2.04

To a solution of 0.30 g (0.765 mmol) of 2.05 in 20 mL of THF was added to

0.45 mL (1.2 eq, 2M) of n-BuLi at -100 oC, under N2. The solution was stirred
overnight, allowing the temperature to slowly rise to RT. The solution was then cooled

to -30 OC, and Me30BF4 was added in small portions until the pH reached 6-7 (about 0.3

g is needed). The orange color of the solution changed to reddish at ca. 0 OC within 30

min and the excess of Me3OBF4 was destroyed with 30 mL of degassed water. Then the

mixture was extracted with hexane three times, and the combined extracts evaporated
under reduced pressure to dryness. The crude product was chromatographed on alumina

(III), with the desired red band eluting very slowly with hexane. After removal of most
of the solvent, the product crystallized at -10 OC after several hours to give 110 mg of
2.04 as a red crystalline solid (44% yield). Anal. Calcd for C15H1402Ru: C, 55.05; H,
4.31. Found: C, 54.87; H, 4.21. Mp = 110-111 C. HRMS (EI, m/e): Calcd for (M+):
328.003; Found: 328.0028. 'H NMR (CDC13): 8 5.23 (s, 5H, Cp), 3.18 (d, J=13.9

Hz, 1H), 3.05 (d, 2JHH=13.9 Hz, 1H), 4.51 (s, 3H, OCH3), 7.0-7.6 ppm (phenyl
region). 13C NMR (CDCl3): 8 294.16 (Ru-C), 206.07 (CO), 86.8 (Cp), 67.0 (OCH3),

10.27 ppm (CH2).

Synthesis of Phenvl-Substituted Ruthenium Carbene Complex 2.03a

In a Schlenk tube, 150 mg (0.457 mmol) of 2.04 was dissolved in ca. 30 mL
THF and cooled to -78 oC. To this red solution was added 0.38 mL (1.5 eq)
phenyllithium, dropwise, which resulted in a gradual change in color to a dark reddish-
brown as the solution warmed to -25 OC over a 4 h period. To this solution, cooled by
liquid N2 to -196 OC, was added, via vacuum line transfer, ca. 2 equiv of trimethylsilyl

triflate (Me3SiSO3CF3) which had been distilled over a proton sponge. A dark red color
was produced when the temperature was allowed to increase from -196 oC to RT over a

period of one hour. Removing the THF under vacuum led to a black-purple residue
which was chromatographed on alumina (III), with the black-purple band eluting with
hexane. Removal of most of the solvent followed by crystallization at -30 OC led to
formation of 110 mg of black crystals of 2.03a (65% yield). Anal. Calcd for

C2oHi6ORu: C, 64.35; H, 4.32. Found: C, 64.29; H, 4.28. Mp: 114-115 oC. MS

(El, m/e): 374 (M+1). IH NMR (C6D6): 8 4.87 (s, 5H, Cp); 3.72 (d, 2JHH = 12.6 Hz,
1H); 6.7-7.4 (phenyl region); 2.58 ppm (d, 2JHH =12.6 Hz, 1H). 13C (C6D6): 8 309.53;

209.77; 166.96; 160.29; 158.12; 131.09; 129.14, 128.53; 127.29; 124.46; 122.46;
122.78; 90.32; 15.87 ppm.

Synthesis of p-Tolyl-Substituted Ruthenium Carbene Complex 2.03b

The method used was the same as that described above for the phenyl derivative,
2.03a, except p-tolyllithium was used instead of phenyllithium. Black crystals of

2.03b were obtained in 58% yield. Anal. Calcd for C21HisORu: C, 65.10; H, 4.65.
Found: C, 65.09; H, 4.68. Mp: 95-96 oC. MS (EI,m/e): 388 (M+1). 'H NMR (C6D6):
8 4.91 (s, 5H, Cp); 3.73 (d, 2JHH= 12.6 Hz, IH); 2.63 (d, 2JHH = 12.6 Hz, 1H); 2.10

(s, 3H, CH3); 6.8-7.6 ppm (m, 5H, phenyl region). 13C NMR (C6D6): 8 309.16 (C=M);

209.9 (CO); 166.84; 160.32; 156.0; 139.56; 131.05; 128.13; 124.40; 122.84 (phenyl

region); 90.2 (Cp); 21.4 (CH2); 15.6 ppm (CH3). IR (hexane) 1966.7 cm-1.

Thermolysis of 2.03a -- Preparation of ruthenacene 2.11a

A black solution of 0.30 g of carbene complex 2.03a in ca. 0.8 mL C6D6 in a
sealed NMR tube was heated at 50 OC. The reaction was monitored by 'H and 13C NMR
from time to time, and it was seen that 2.11a was formed slowly as 2.03a disappeared.
After 12 hours, 2.03a was found to have been quantitatively converted to 2.11a. The
same result can be obtained if the solution of 2.03a is allowed to sit at room temperature
for two days. The solvent was evaporated under vacuum to give 30 mg (100 %) of an

orange solid. Anal. Calcd for C2oHi6ORu: C, 64.35; H, 4.32. Found: C, 64.16; H,
4.31. Mp: 120-135 oC(decomp.). MS (EI, m/e): 374 (M+). 'H NMR (C6D6): 8 5.28

(s, 1H, CH); 4.10 (s, 5H, Cp); 2.90 (br, 1H, OH); 6.75-7.8 ppm (m, 9H, aromatic
region). 13C NMR (C6D6): 8 135.7; 131.2; 128.7; 127.1; 126.9; 126.1; 123.8; 123.0;

122.9; 90.32; 88.36; 56.7; 31.8 ppm.

Thermolysis of 2.03b-- Preparation of ruthenacene 2.11b

The method used was the same as for the preparation of 2.11a, except the
temperature used was 70 OC, and 2.11b was formed quantitatively. Anal. Calcd for
C21iHsORu: C, 65.10; H, 4.65. Found: 65.09; H, 4.68. 1H NMR (C6D6): 8 2.20 (s,

CH3); 2.8 (br, OH); 4.15 (s, 5H); 5.33 (s, 1H); 6.75-7.74 ppm (m, 8H, aromatic
region). 13C NMR (C6D6): 8 136.64-122.84 (aromatic region); 88.40; 86.68; 77.83;

71.78; 56.56; 21.20 ppm.

Thermolysis of Methoxy Ruthenium Carbene Complex 2.04

A solution of 0.35 g of Ru carbene complex 2.04 in ca. 0.8 mL of C6D6 was
heated at 80 OC under an N2 atmosphere in a sealed NMR tube. The mixture was
monitored by IH NMR and 13C NMR from time to time, and no reaction was observed
after 5 days.

Preparation of 2.20

To a solution of 20 mg (0.54mmol) of 2.03a in 1 mL of C6D6 was added 20 uL
of PMe3 at room temperature. A color change of the solution from dark red to orange
occurred immediately. The IH NMR showed the PMe3 has added to 2.03a to form

2.20 completely. The solvent was evaporated to give 23 mg (100%) of an orange solid.
'H NMR (C6D6): 6.6-7.6 (m, 9H), 4.33 (s, 5H), 3.89 (dd, 1H, 2JHH=14.8Hz,
4JpH=3.8Hz), 2.91 (dd, 1H, 2JHH=14.8Hz, 4JPH=3.8Hz), 0.9 (br, 9H). This broad
peak at 0.9 ppm (PMe3) appeared as three groups of doublets at -78 OC at 1.55, 0.58 and
0.49 ppm with coupling constants 2JpH=9.6 Hz, 11.9 Hz and 9.0Hz respectively. Two
groups of doublets were observed at -400C at 1.68 and 0.62 ppm with intensity ratio 1:2.
The two groups of doublets coalesced at room temperature to give a broad peak. 13C
NMR (C6D6): 211.95 (d, 2Jpc=5.5Hz), 162.61, 155.98, 146.99, 130.63, 129.28,
129.20, 127.83, 126.12 (d, J=3.2Hz), 125.8 (br), 123.84, 121.87, 87.26, 4.63. MS

(EI, m/e) 374 (M+). Anal. calcd. for C23H2OPRu: C, 61.48; H, 5.56. Found: C 61.04;
H, 5.58.

Crvstallomaphic Analysis

Data were collected at room temperature on a Siemens R3m/V diffractometer equipped
with a graphite monochromator utilizing MoKa radiation (X = 0.71073 A). 50 reflections with
20.00 5 20 5 22.00 were used to refine the cell parameters. 4175 and 2741 reflections, for
2.03a and 2.11b respectively, were collected using the o-scan method (1.2' scan range and 3-

6' scan speed depending on intensity). Four reflections (223, 120, 023, 202) for 2.03a, and (2
)21, 023,221, 0,i3) for 2.11b, were measured every 96 reflections to monitor instrument and

crystal stability (maximum correction on I was < 1 %). Absorption corrections were applied
based on measured crystal faces using SHELXTL plus (Sheldrick, 1990). Absorption
coefficients: 2.03a, p = 0.97 mm-l(min. and max. transmission factors are 0.789 and 0.868,
respectively); 2.11b, g = 0.96 mm-l(min. and max. transmission factors are 0.723 and 0.828,
The structures were solved by the heavy atom method in SHELXTL plus 50 from which
the positions of the Ru atoms were obtained. The rest of the non-H atoms were obtained from
subsequent Difference Fourier maps. The structures were refined in SHELXTL plus using full-
matrix least squares. All of the non-H atoms were refined with anisotropic thermal parameters.
The non-H atoms were treated anisotropically. The H atoms in 2.03a were located from a
Difference Fourier map and were refined without any constraints. In 2.11b, all of the H atoms
were located from a Difference Fourier map and refined freely; except the C16 hydrogen atoms
which were calculated in idealized positions and their thermal parameters were fixed at 0.08. 263
and 261 parameters for 2.03a and 2.11b, respectively, were refined and w ( I Fo I I Fc )2
was minimized; w=l/(a I Fo 1)2, o( Fo) = 0.5 kI -'t2 [a( I )]2 + (0.02I)2 ) /2, I(intensity)= ( I

peak Ibackground )(scan rate ), and o(I) = (I peak + I background)/2 (scan rate), k is the
correction due to decay and Lp effects, 0.02 is a factor used to down weight intense reflections
and to account for instrument instability. R = 0.0457 and wR = 0.061 for 2.03a, and R =
0.0338 and wR = 0.0393 for 2.11b in the last cycle of refinement. The linear absorption
coefficient was calculated from values from the International Tables for X-ray Crystallography
.51 Scattering factors for non-hydrogen atoms were taken from Cromer & Mann52 with
anomalous-dispersion corrections from Cromer & Liberman,53 while those of hydrogen atoms
were from Stewart, Davidson & Simpson.54

Preparation of CpRu(CO)2Me (or RpMe) 3.04

A solution of 1.6 g (3.60 mmol) of (CpRu(CO)2)2 in 300 mL of THF was treated
with ca. 0.80 g Na and 8.0 mL Hg amalgam. The mixture was stirred for 5 hr. The dark

grey solution was then transferred into a Schlenk tube. To this solution, cooled to -78 OC,

was slowly added excess Mel (1.31 mL) and stirred overnight at RT. The solvent was

then removed under vacuum. The resulting residue was extracted with hexane at least

three times, and the solution was eluted through an alumina column to give a colorless

solution. Evaporation of the solvent gave pure white RpMethyl in 74% yield. The

compound was characterized by comparison of IR and NMR spectra with those of the

known material 49.

Preparation of 13C enriched CpRu(CO*Q)CH_3.04

This compound was prepared by the same method as above, except using

Preparation of acyclic ruthenium lithium enolate 3.08

To a solution of RpCH3 3.04 (0.10 g, 0.42 mmol) in ca. 10 mL of diethyl ether

was slowly added PhLi (1.8M, 0.40 mL, 0.72 mmol) at -78 OC in the special apparatus

as shown in chapter 3 page 35. The reaction was allowed to stir for at least 5 h under N2

followed by warming the mixture to -30 OC. A bright yellow precipitate was formed.

The solvent was removed in vacuo at -30 oC over a six hour period, and then the

ruthenium enolate salt 3.08 was treated with 1 mL of CD2C12 at -78 OC for a half hour.

The solution was filtered through a frit covered with a small amount of celite into an
NMR tube at -78 OC, and then the NMR tube was sealed at liquid nitrogen temperature.
The Ru enolate, 3.08, was studied by 'H and 13C NMR spectroscopy at -78 OC. 'H
NMR (CD2C12, -78 C): 8 7.26-7.75 (m, 5H, phenyl region); 5.16 and 5.05 (s, 5H,
Cp). '3C NMR (CD2C12, -78 OC): 8 294.54; 292.23; 291.79 (C=Ru); 215.4, 212.79,

212.03 (CO); 133.2-126.9 (phenyl region); 91.40; 91.34 (Cp); -26.71; -26.77; -28.05

ppm (Ru-CH3). (Figure 3-6)

Preparation of acyclic ruthenium lithium enolate 3.08 in the presence of 12-crown-4

Lithium enolate 3.08 was prepared exactly the same as above except for the
addition of one equivalent of 12-crown-4 (0.063 mL) after the formation of 3.08. 'H
NMR (-90 OC, CD2C12) 8 7.65-7.056 ppm (m, phenyl region), 4.897 (s, Cp), 0.217 (s,

CH3); 13C NMR (-90 C, CD2C12) 8 284.13 (C=Ru), 2.10.82 (CO), 128-123.9

(phenyl region), 89.50 (Cp), -28.09 ppm (CH3-Ru).

Preparation of acyclic ruthenium lithium enolate 3.08 in the presence of ten percent

Lithium enolate 3.08 was prepared exactly the same as above except for the
addition of 10% deuterated THF. 'H NMR (-90 OC, CD2C12) 8 7.60-6.974 ppm (m,
phenyl region), 4.859 (s, Cp), 0.1098 (s, CH3); 13C NMR (-90 oC, CD2C12) 8

Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E5BL4CQJH_X5ZSX1 INGEST_TIME 2013-01-23T14:53:28Z PACKAGE AA00012916_00001