Title: Aromatic carbene transition metal complexes
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00099118/00001
 Material Information
Title: Aromatic carbene transition metal complexes
Physical Description: ix, 82 leaves : ill. ; 28 cm.
Language: English
Creator: Allison, Neil Thomas, 1953-
Copyright Date: 1978
 Subjects
Subject: Aromatic compounds -- Reactivity   ( lcsh )
Carbenes (Methylene compounds)   ( lcsh )
Chemistry thesis Ph. D
Dissertations, Academic -- Chemistry -- UF
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Statement of Responsibility: by Neil T. Allison.
Thesis: Thesis--University of Florida.
Bibliography: Bibliography: leaves 79-81.
General Note: Typescript.
General Note: Vita.
 Record Information
Bibliographic ID: UF00099118
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000079190
oclc - 04942793
notis - AAJ4494

Downloads

This item has the following downloads:

aromaticcarbenet00alli ( PDF )


Full Text














AROMATIC CARBENE TRANSITION METAL COMPLEXES


By

NEIL T. ALLISON
















A DISSERTATION PRESENTED TO iE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIRP2ENTS FOP THE
DEGREE OF DOCTOR OF PHILOSOPHY








UNIVERSITY OF FLORIDA


19 7





























to my wife,

Amelia,

for her love, understanding, and help















ACKNOWLEDGEMENTS

I would like to thank Professor William M. Jones for

his help and enthusiasm throughout my stay at the University

of Florida. I would also like to thank Dr. Edward E.

Waali who not only gave me a foundation in organic chemistry

but also encouraged me to continue my studies.

I would like to express also my gratitude for the

encouragement that I received from my parents, my wife's

parents, and my wife's grandmother, Mrs. L.E. Toole.

Thanks must also go to Dr. Yuzo Kawada, Mr. Bradley

Duell, and Mr. Thomas Baugh for their help and friendship.

Finally I would like to thank a very good friend, Mr. Frank

Hill, who provided a "special" intangible help.














TABLE OF CONTENTS


Pace
ACKNOWLEDGEMENTS iii

ABSTRACT vii

CHAPTER

I INTRODUCTION 1

II RESULTS AND DISCUSSION 11

III EXPERIMENTAL 57

General 57

Cyclopentadienylirondicarbonyl iodide 53

Preparation of tropone (20) 58

Preparation of bromotropylium bromide 58

Preparation of 1-, 2-, and 3-bromocyclchepta-
trienes (21) 58

Preparation of 1-, 2-, and 3-lithiocyclohepta-
trienes (22) 58

Preparation of 1-, 2-, and 3-(cyclopentadienyl-
irondicarbonyl)cycicheptarrienes (23) 58

Preparation of ; -cycloheptatrienylidene-;75-
cyclopentadinyienydicarbonyliron hexafluoro-
phosphate (24) 59
Reaction of 1 5
Reaction of h -cycloheptatrienylidene- -cyclo-
pentadienyldicarbonyliron hexafluorophosphate
with lithivu aluminum hydride 60

Reaction of hl-cyclcheptatrienylidene-h5-cyclo-
pentadiensyl dicarbonyliron hexafluorophosphate
with methyl lithium 61

Oxidation of -cvcloheptatrienyiidene-h'-cyvclo-
pentadienyldicarbonyliron hexaflucrohophosphate
with dime hyisulfoxide 61








Page


Preparation of 4,5-benzotropone (36) 62

Preparation of 1,2-benzo-5-bromotropylium
bromide (37) 62

Preparation of 1,2-benzo-5-bromocyclohepta-
triene (38) 62

Preparation of 1,2-benzo-5-lithiocyclohepta-
triene (39) 63

Preparation of 1,2-benzo-5-(cyclopentadienyl-
irondicarbonyl)cycloheptatriene (40) 53

Preparation of h -4,5-benzocycloheptatrienyl-
idene-h5-cyclopentadienyldicarbonyliron
hexafluorophosphate (41) 64
1
Reaction of h -4,5-benzocycloheptatrienyl-
idene-h5-cyclopeietadinyldicarbonyliron
hexafluorophosphate with lithium aluminum
deuteride 65

Reaction of h -4,5-benzocycloheptatrienyl-
idene-h5-cyclopentadienyldicarbonyliron
hexafluorophosphate with methyllithium 66

Preparation of 1-bromo-1,2,3,4-tetrahydro-
naphthalene 66

Preparation of 1,2-dihydronaphthalene 67

Preparation of 1,2-benzo-4- and 1,2-benzo-6-
bromocycloheptatrienes (48, 49) 67

Preparation of 1,2-benzo-4- and 1,2-benzo-6-
lithiocycloheptatrienes (50, 51) 68

Preparation of 1,2-benzo-4- and 1,2-benzo--6-
(cyclopentadienylirondicarbonyl)cyclchepta-
trienes (52, 53) 68

Preparation of h -3,4-benzocycloheptatrienyl-
idene-h5-cyclopentadienyldicarbonyliron
hexafluorophosphate (46) 69

Preparation of 5,6-dibromo-5H-dibenzo[a,c]-
cyclohepgene (58) 70

Preparation of 6-bromo-5H-dibenzc '[, ca cyclc-
heptene (59) 72








Page


Preparation of 6-lithio-5H-dibenzo [a,c]-
cycloheptene 73

Preparation of 6-(cyclopentadienylirondicarbonyl)-
5H-dibenzo[a,c]cycloheptene (60) 73

Attempted synthesis of n -3,4-5,6-dibenzocyclo-
heptatrienylidene-R5-cyclopentadienyldicarbon-
yliron hexafluorophosphate (56) from 6-
(cyclopentadienylirondicarbonyl)-5H-dibenzo-
[a,c]cycloheptane (60) 74

Preparation of 6-bromo-5-methoxy-3H-dibenzo[a, c]-
cycloheptene (63) 74

Preparation of 6-lithio-5-methoxy-5H-dibenzo-
[a,c]cycloheptene (64) 75

Preparation of 6-(cyclopentadienylircndicarbonyl)-
5-methoxy-5H-dibenzo[a,c]cycloheptene (62) 75

Attempted synthesis of h -3,4-5,6-dibenzocyclo-
heptatrienylidene-h5-cyclopentadienyldi-
carbonyliron hexafluorophosphate (56) from
6-(cyclopentadienylirondicarbonyl)-5-methoxy-
5H-dibenzo[a,c]cycloheptene (62) 76

Reaction of yellow solid with sodium methoxide 77

Reaction of 6-(cyclopentadienrlirondicarbonyl)-
5-methoxy-5H-dibenzo[a,a]cycloheptene with
hydrogen chloride in the presence of tetra-
cyclone 77














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



AROMATIC CARBENE TRANSITION METAL COMPLEXES

By

Neil T. Allison

December 1978

Chairman: William M. Jones
Major Department: Chemistry

Aromatic carbene transition metal complexes are

classified as carbene complexes which possess an intra-

ligand (4N+2) pi electron system. To date, no carbocyclic

carbene complex with (4n+2) pi electrons where n>l has been

reported. The synthesis of h -cycloheptatrienylidene-, -

cyclopentadienyldicarbonyliron hexafluorophosphate (Fp [cyclo-

heptatrienylidene]HFP), a six pi electron carbocyclic aroma-

tic carbene complex,is a first step in filling this void.

The same methodology was also successfully applied by

Kawada to the synthesis of cycloheptatrienylidene(penta-

carbonyl) tungsten, the first neutral transition metal

complex of cycloheptatrienylidene. This synthetic method

was a2so readily applicable to the synthesis tc two ten pi

electron carbccyclic carbene complexes, namely, i 4-4,5-

benzocyclohepta trienylidene-2 -cyclopentadienyldicarbonv i ron








hexafluorophosphate (Fp[4,5-benzocycloheptatrienylideneHFP) ,

and n h -3 ,4-benzocycloheptatrienylidene-5 -cyclopentadienyl-

dicarbonyliron hexafluorophcsphate, (Fp[3,4-benzocyclohepta-

trienylidene]HFP).

The synthesis of Fp cycloheptatrienylidene]HFP was

accomplished via hydride abstraction from 1-, 2-, and 3-

(cyclopentadienylirondicarbonyl) cyclcheptatrienes. In turn,

these cycloheptatriene complexes were prepared by reaction

of 1-, 2-, and 3-lithiocycloheptatrienes with cyclopenta-

dienylirondicarbonvl iodide. The H nmr, 1-C nmr, and ir

spectra indicate that there is a greater and more highly

delocalized positive charge on the carbene ligand of Fp[cyclo-

heptatrienylidene]HFP than h -benzocyclobutenylidene-; 5-

cyclopentadienyliron hexafluorophosphate, Fp [benzocyclo-

butenylidene]HFP, or h -phenylcarbene-i -cyclopentadienyl-

dicarbonyliron triflate, (Fp[phenylcarbene]TFA). Reaction

of Fp[cycloheptatrienylidene]HFP with lithium aluminum

hydride or .ethyllithium generated neutral cycloheptatriene

ccrr.lexes. Oxidation of Fp[cycloheptatrienylidene]HiFP with

dimethylsulfoxide produced tropone.

The spectral data of Fp[4,5-benzocyclohepoatrienylidene]-

HFP indicates that the amount of positive charge on the

carbene ligand is almost equal to that on the carbene ligand

of 'p [cvcloheptatrienylidene]HFP. However, 1C nmr data

indicate that more positive charge is located on the carbene

carbon 3f Fp[4,5-benzocycl.oheptatrienylidene]HFp than Fp[cyclo-

het.ar.rien;ylid:ene]EFP. As with ?p [cycloheptatrienylidene]HEFP


vii i









Fp[4,5-benzocycloheptatrienylidene]HFP is susceptible to

attack by nucleophiles. Reaction of Fp[4,5-benzocyclo-

heptatrienylidene]HFP with methyllithium or lithium

aluminum deuteride led to two neutral complexes.

The spectral data of Fp[3,4-benzocycloheptatrienylidene]-

HFP suggest that there is less metal to ligand back

donation than with Fp[cyclophetatrienylidene]HFP. Further-

more, 1C nmr data indicate the amount of positive charge

residing on the carbene carbon of Fp[3,4-benzocyclohepta-

trienylidene]FHP is less than that of Fp[cycloheptatrienyl-

idene]HFP and Fp[4,5-benzocycloheptatrienylidene]HFP. From

an x-ray structural determination performed by Davis, the

carbene ligand of Fp[3,4-benzocycloheptatrienylidene]HFP

appears planar with some bond alternation,

Attempts to prepare a 3,4-5,6-dibenzocycloheptatri-

enylidene complex of iron were unsuccessful. Two potential

carbene precursors, 6-(cyclopentadienylirondicarbcnyl)-5H-

dibenzo[a,.]cycloheptene and 6-(cyclopentadienylircn-

dicarbcnyl)-5-imethoxy-5H-dibenzo[a,c]cycloheptene, were

prepared. However, generation of a dibenzocyclohepta-

trienylidene complex from either was unsuccessful.














CHAPTER I
INTRODUCTION

The synthesis of aromatic carbene complexes of transi-

tion metals has for some time been restricted due to a lack

of general synthetic methods."1 2 In fact, no carbocyclic

6 pi electron or larger aromatic carbene complex has ever

been reported. The synthesis of a cycloheptatrienylidene

transition metal complex (i.e. 1) and higher (4n+2) pi elec-

tron electron homologs would not only fill a void in the

literature but would also provide a new class of transition

metal carbene complexes from which physical and chemical

properties could be obtained.





M







la lb



In general, carbene complexes are carbene species that

are coordinated to a transition metal and, on the whole, can

be formulated using the three extreme resonance forms 2

through 4. These three forms represent the vacant p

carbene orbital interacting (1) with no atom, (2) within











M M M

+ <---> I < > i

Y X Y X Y X
2 3 4



the ligand, and (3) with a metal orbital. Qualitatively,

the difference in orbital energies between the d metal

orbitals and the ligand orbitals causes the carbene p

orbital to interact more within the ligand.- This means

that with aromatic carbocyclic carbene complexes, metal back

donation (as represented by 4) should decrease to some extent

by delocalization of the (4n-2) electrons within the pi system

of the ligand. Thus aromatic carbene complexes are classified

as those complexes which possess an intraligand (4n+2) pi

electron configuration.

Carbene complexes have been generated using a wide range
12 2
of synthetic methods. An extensive review will not be

given even though several methods will be brought forth.

The synthetic procedure most widely used was developed by

Fisher and Maasb6il and has been utilized in the synthesis

of a large variety of hetercatom stabilized carbene complexes.

Their method involved the attack on a carbonyl ligand by an

organolithium reagent. Alkylation of this intermediate,

whicn was usually negatively charged, then led tc a carbene

complex (e.g. 5).










R
M(CO)n RLi> (CO)n_1 M=C

0-

L +
Li


(CH ) BF4
>


/R
(CO) M=C

OCH3

5


A variety of complexes containing nitrogen as the

stabilizing heteroatom have been synthesized by nucleophiiic

attack with a primary or secondary amine on the carbene

carbon in oxygen stabilized complexes. Loss of alcohol

thus led to complexes such as 6.4


/R
(OC) M=C
OCH3


R
2NR (OC) M C-
2> n>1n I
O H
CH,


R
(oC) M=C
n \


6

Also, in recent years, a few non-heteroatom stabilized

complexes have been generated from salt-like precursors.

Ofele has used metal carbonyl dianions to synthesize the


-C HOH
___--- -)








only reported carbocyclic aromatic carbene complexes.5' 6

By displacing two chloride ions from 1,2-diphenyldichloro-

cyclopropene, the synthesis of the 2 pi electron diphenyl-

cyclopropenylidene complex 7 was successful.


(OC)5Cr-2


Cl Ph


Cl h


Ph

> (OC)5Cr =(

Ph


Herrmann has developed another synthetic procedure

a diazo precursor and a labile transition metal ligand.

interaction resulted in loss of nitrogen and generation

the desired complexes. In this manner he successfully

generated the diphenyl carbene complex 8.7


TEF


++
N P

Ph Ph


OC P

Ph


More recently, Casey et al. brought some innovative

new approaches to the sYnlhesis of ncn-heteroacom stabilized

carbene complexes. By utilizing the electrophilicity of

a phenylmethcxy carbene coordinated to pentacarbonyltungsten,

9, he was able to generate three new non-heteroacom stabilized


using

Their

of








carbene complexes. Via attack with phenyliithium, the stable

anionic intermediate 10 was generated. By addition of

ethereal hydrogen chloride, protonation of the methoxy sub-

stituent resulted in loss of methanol giving the desired

diphenyl carbene complex 11.



Ph
.Ph Ph i
(OC) 5W=C PhLi (OC) 5- --OCH

3 Ph

9 10


Ph
HC1 /Ph
-> (OC) W=C
-CH OH O Ph
ii oP

11




Two less stable complexes, 12 and 13,10 were generated

in a like manner. Although not observed by spectroscopic

measurements, their decomposition products pointed to their

respective intermediacy.

Casey and Burkhardt1 found that the diphenyl carbene

complex 11 underwent methathesis type reactions. When

treated with l-methoxy-l-phenylethene, 11 generated the

methoxyphenylcarbene complex 9 via a metathesis mechanism.

Ph OCH OCH Ph
w/ __
(CC) 5 3 --> (CC)5W +
Ph Ph Ph Ph

11 9













(CC) 5 7


HC1 > (CO) 5w

CH,
12


CH Li

-780C


Ph

(OC)5 W<

OCH
11 3


h P Ph CH3
Ph
CH Ph
3


SKHB[OCHCHH) 23


Ph

(OC)5W OCH3


CF3CO2H /Ph
> (CO) 5W


It should De mentioned that Schrock2 has synthesized

some unique tantalum carbene complexes. Additionally,

Fisher et al.13 have ha great success in generating carbene

complexes from alkyne transition metal complexes.


Ph



Ph








Transition metal carbene complexes of cyclopentadienyl-

dicarbonyliron (Fp) have also recently been under investigation.

The simplest preparation of these was via protonation of an

acyl complex 14.14


OC Fe

oc /
CH3
14


H+
------->


I+
OC Fe- OH

OC
CH3


By utilizing methanol as a leaving group, the methylene

complex of Fp, 15, was generated.1' 16 Although not
isolated, its intermediacy was inferred via a reaction with

cyclohexene which gave norcarane.


Fe

OC H
H


H

OC


H
---> Fe
0C 0/ H

CC
H3


0C,


Also, two carbene complexes of Fp have been reported.

By alpha hydride abstraction, Giering et al. synthesized


(9






8

. -benzocyclobutenylidene-h5-cyclopentadienyldicarbonyliron

hexafluorophosphate (Fp [enzocyclobutenylidene]HFP) (16) .17


Fe

OC
ocN


+
Ph C PFE
--1 -- ^ -


c Fe F

OC


16


S5
.1-Phenylcarbene-h -cyclopentadienyldicarbonylircn

triflate (2p[phenylcarbene]TFA) (18)18 was prepared in a

manner analogous to that used for the methylene complex 15.

By treatment of a methoxybenzyl complex, 17, with acid or the

triphenylcarbenium ion, generation of the phenyl carbene

complex was possible.


Fe
oC/C H^ 3

CPh
Ph


H -CHF OH

or Ph3 C


Ie+ CF,SO3



OC Ph


In regard to the generation of 6 pi electron or larger

aromatic carbocyclic carbene complexes, there have been








several synthetic procedures attempted that held good poten-

tial for preparing a cycloheptatrienylidene complex. How-

ever, as discussed below, all of these attempts proved futile.

After the system outlined by Ofele video supra, DeJohn
19
et al. employed dianion transition metal carbonyls (i.e.

Na2Cr(CO)5 and Na2Fe(CO)4) in reactions with dichloro-,

dibromo-, or 7,7-dimethoxycycloheptatriene. However, these

attempts resulted in the reduction of the tropyl ring, yield-

ing a mixture of substituted bitropyls 19.


x X
H

Na2M(CO)n



19



S20
Reactions of tropylium salts with NaFe(CO) 4H,
71
modeled after another reaction scheme reported by Ofele"

also gave reduction of the tropyl ring. Furthermore, reac-

tion of heptafulvalene with diphenylcarbene(pentacarbonyl)-

tungsten(0) (11) modeled after Casey and Burkhardt's reaction

using l-methoxy-l1-phenylethene ide c C ., gave no sign of

a carbene complex or diphenylethylene.20 22




(oC) C + i -> > (OC) W = |
< \





10


Decomposition of the sodium salt of tropone tosylhydra-

zone in hopes of trapping the presumed diazo intermediate

with methylcyclopentadienyldicarbonyl (TIIF)manganese *ide

supra was also attempted. However, this also resulted in

failure.

Therefore, the primary objective of this research was

to synthesize and explore the chemistry of a transition

metal complex of cycloheptatrienylidene. If such a complex

could be synthesized, it would be the first example of a

cycloheptatrienylidene complex and an unsubstituted carbo-

cyclic aromatic carbene complex.














CHAPTER II
RESULTS AND DISCUSSION

The synthesis of the first transition metal complex

of cycloheptatrienylidene was successfully accomplished via
22
the synthetic route shown below. A detailed look at this

synthesis follows.


0 Br









2C


(2! ison~rs)


Br


(3 isomers)

21


Fe





23


The preparation of 1-, 2-, and 3-bromocyclcheptatrienas

_) from tropone (20) was previously reported.23, 24

Treatment of tropone with oxalyl bromide gave bromotropylium

broaide. Reducti n of the dibr omid with lithium aluminuin

h.drid- gavee e iserc mixture of broc cloheptatrienes,

21, as reported.












0 Er
0 Br
BrC-CBr LiA1H4




(3 isomers)

20 21


The next step, the generation of the cycloheptatrienyl-

lithiums, 22, held the key to the successful completing of

a cycloheptratriene moiety to a transition metal. Even

though their generation had not been report-ed, this type of

exchange reaction with vinyl bromides and iodides is well
25
known. It was found that by treating the bromocyclohepta-

trienes with n-butyllithium, the ea-chance reaction occurred

readily, giving a dark green colored solution.






[I V ^n-EuLi
Br /^Li

(3 isomers) (3 isomers)

21 22
The reaction of transition metal halides with alkyl
26.-31
and aryl Grignard and lithium reagents is well documented.203

Treatment of the cyciohaptatrienyllitnium.s 22 with cyclo-

pent-die.nylirrondicarbonyl iodide (FpI) resulted in the

generation of the 1-, 2-, and 3-(cyclopentadienylirondicarbonyl)-

cycicheptatrianes (Fp[cycloheptatriene]) (23; 26-29%; as a





13

yellow, air sensitive oil. The 1H nmr spectrum of the mixture

is given in Figure 1.


Li



(3 isomers)

22


FpI
--->


(3 isomers)


With the isolation of 23 in hand, hydride abstraction

from the cycloheptatriene ligands with triphenylcarbenium

hexafluorophosphate gave 24 (90%), a solid whose elemental

analysis is consistent with a Cp(CO) 2Fe(C-H )PF6 formua.

The 1H nmr spectrum of 24 is given in Figure 2.


oc /
Cc

OC


+F-
PhC PF
6
------------>---


c I N-
cc


(3 isomers)

23


~II













C)
0



Id




Cl

I C
I I.
0
c)







* 01


0
-1


0u~
CC






I LI



r-


,I
i













7TK-











7,


















A























i i












I,/


c
0
C
.0

S-1

1-3
C





ri
L "P


>i
t




I


q







C


r-i









.2 --
0
U

4' i.-







-Q
n









0 .0
S-'-
i r-



tfl)


-I








C-'
!"

P i M

o ,
-^ a
J_ s o
i ^
I 13
u 1-1
'- ID "
C. 1-
nX
-"





i s


mf
c

aj








Proof of the carbene structure of 24 lies in its spectral

data. As is seen in the 1H nmr spectrum, the doublet at

610.01 and the multiplets at 68.48-8.74 and 7.94-8.3 appear

at fairly low field. The ir spectrum exhibits two carbonyl

stretching frequencies at 2045 and 1996 cm-1. Both the H

nmr and ir absorptions are indicative of the inherent symmetry

within the complex.

However, at the time of its generation, it was thought

that the ligand could be complexed in the form of an allene

(e.g. 25) rather than a carbene. This was considered because




+
CpFe(CO) 2 CpFe(CO)2







25a 25b



(1) Cycloheptatrienylidene (26) exhibits properties consistent

with a cycloheptatetraene intermediate (27)32, 33 and (2)

there are numerous examples of Fp allene complexes.34


#0


26 27








However, from the 1H nmr, this allene structure may be

ruled out on the basis of the low field proton resonances.

Whereas Cp(CO)2Fe(allene) (28) shows resonances at 66.59,
34
6.25, and 3.3,4 the proton resonances at the C-2 and C7








3Cz
I+



OC //

28



positions in 24 appear at much lower field (610.01). From

this, it is seen that an allene structure is unlikely and

the preferred structure is the carotene complex 24.

i -Cycloheptatrienylidene-z -cyclopentadienyldicarbonyl-

iron hexafluorophosphate (Fp[cycloheptatrienylider.e]HFP) (24)

is a stable yellow-orange solid with a fairly high decomposi-

tion point (180.5-131.50C). It forms plates when recrystal-

lized from methylene chloride. Moreover, it is stable to air

as a solid and can be stored for months as a solid in a dark

bottle, in the presence of air, without decomposition.

Even after refluxing in dichlcroethane (b.p. 830C) for 6 hours,

the salt was reisolated with no apparent decomposition.

To test the generality cf this method with other transi-

tion metals, a neutral cycloheptatrienylidene complex was
prepared Kawada22 his yntheticoute iselow.
prepared by Kawada. This synthetic route is shown below.









+
n-Bu N
Li
nBu4 (CO) WBr PhCPF6

29 C "

22 30




(CO) 5wO



31



Reaction of the cycloheptatrienyllithiums (enriched in the

1-isomer) 22 with tetra-butylammonium bromopentacarbonyl-

tungstate (29) gave the negative cycloheptatriene complex

30. Hydride abstraction from 30 with triphenylcarbenium

hexafluorophosphate then gave cycloheptatrienylidene(penta-

carbonyl)tungsten (31).

The spectra of Fp[cycloheptatrienylidene]EFP (24) and

cycloheptatrienylidene(pentacarbonyl)tungsten (31) give

some insight into their bonding. As can be seen in Figure

2, the proton resonances of the carbene ligand in Fp[cyclo-

heptatrienylidene]HFP appear at low field. Because the com-

plex is a salt, the question may be asked, "Do these chemical

shifts primarily arise from the inherent positive charge

of the complex?" This may be answered by comparison of its

spectrum with the neutral tungsten cycloheptatrienylidene

complex 31. Whereas 24 shows resonances at 610.01, 8.48-8.74,

and 7.94-8.3, the neutral tungsten complex's resonances





19

appear at only slightly higher field (69.96, 8.1, and 7.7)

and are identical in shape to those of the iron complex.

It would appear from this data that almost equivalent

amounts of positive charge lie in their ligands.

The ir spectral data of both the iron and tungsten

complexes indicate that more positive charge is localized on

the carbene ligands than is the case with other known carbene

complexes. For example, the carbonyl stretching frequencies

of Fp[cycloheptatrienylidene]HFP at 2045 and 1996 cm-I are

substantially lower than those of Fp[benzocyclobutenylidene]-
17 -i
HFP (16) at 2065 and 2020 cm A lower carbonyl frequency

has been interpreted in terms of less electron density on

the metal35 and hence less back donation from the metal to

the carbene ligand.13 However, comparison of the two absorp-

tions of either vinylcyclopentadienylirondicarbonyl (Fp[vinyl])

(32)6 or Fp[cycloheptatriene] (23) that appear at 2020 and

1950 cm suggests qualitatively that the carbene ligand

in Fp[cycloheptatrienylidene]HFP does accept a substantial

degree of back donation.







Fe
OCi

H H








The similarity of the 1H nmr spectra of the iron and

tungsten cycloheptatrienylidene complexes suggests that the

character of the carbene ligands and hence the amount of

back donation afforded to the two ligands is roughly

equivalent. Due to the fact that tungsten has both cis

and trans carbonyl ligands with respect to the carbene

ligands, some semi-quantitative calculations can be made.
37
By using a method outlined by Cotton and Kraihanzel, the

carbcnyl stretching force constants k1 and k2 in mdyn/A
38
may be calculated. Furthermore, Darensbourg and Darensbourg3

have utilized k and k2 in a semi-quantitative method for

calculating relative amounts of back donation to carbene

ligands. Using (C0)5WH5(C6H ) as a standard, the values

shown in Table 1 can be calculated. A larger value for

v represents greater back donation into the carbene ligand.



Table 1. Carbonyl Stretching Force Constants in mdyn/A and
Parameters for Some LW(CO)5 Complexes

Compound k k2 Ak Ak2 Ref.

(CO) WNH(C6H11) 15.07 15.75 0 0 0 38

(CO) WC7H6 (31) 15.34 15.72 0.27 -0.03 0.30 22

(CO) WC(Ph)OCH, (9) 15.71 15.90 0.64 0.15 0.49 8

(CO) WC(Ph)2 (11) 15.87 16.03 0.80 0.33 0.47 8



From the values of T it is seen that although the

amount of back donation to the carbene ligand in cyclohepta-

trienylidene(pentacarbonyl)tungsten is substantial, it is








not as great as in the diphenylcarbene and phenylmethoxy-

carbene tungsten complexes, 11 and 9, respectively. It

should be noted that this agrees well with the suggested

amount of back donation for Fp[cycloheptatrienylidene]HFP

vide supra.
10
The 1C nmr spectrum of Fp[cycloheptatrienylidene]HFP

exhibits a resonance for C-l at 6242.3. This lies in the

high field range of known carbene complexes9 and is 100 ppm

higher field than that of Fp[phenylcarbene]TFA (18) at

6342.4.18 This suggests that there is less positive charge

residing on the carbene carbon of Fp[cycloheptatrienylidene]-

HFP than Fp[phenylcarbene]TFA.

In summary, it is apparent that there is more positive

charge on the ligand of Fp[cycloheptatrienylidene]HFP (24)

than Fp[benzocyclobutenylidene]HFP (16). Furthermore, the

positive charge in the carbene ligand of Fp[cycloheptatrienyl-

idene]HFP is more highly delccalized than in Fp[phenylcarbene]-

TFA (18) indicating the importance of resonance forms

represented by 33. Also the 1H nmr and ir spectra of cyclo-

heptatrienylidene(pentacarbonyl)tungsten suggests similarities

in the amount of back donation and positive charge afforded

to the carbene ligand.






OJ








It was reported that Fp[benzccyclobutenylidene]HFP

(16) reacts with nucleophiles.7 Similarly, it was found

that Fp[cycloheptatrienylidene]HFP (24) also reacts quite

rapidly with nucleophiles. For example, reaction with

lithium aluminum hydride resulted in the regeneration of

1-, 2-, and 3-(cyclopentadienylirondicarbonyl)cyclohepta-

trienes (23; 67%). Furthermore, Fp[cycloheptatrienylidene]-

HFP reacted with methyllithium to give 1-, 2-, and 3-(cyclo-

pentadienylirondicarbonyl)-7-methylcycloheptatrienes (34).

It is interesting to note that in contrast to the reactions






Fe


LiAlH4 OC

(3 isomers)

PF6 23
Fe





24 Fe
O CCH3
oc


(3 isomers)

3a

of Fp[benzocyclobutenylidene]HFP (16), these nucleophiles

gave no evidence of attack at the carbene carbon. For








instance, the reaction with lithium aluminum hydride, gave

no sign of products that would be expected from the cyclo-

heptatriene complex 35.40 However, in the future perhaps

a study should be done with more selective reducing reagents.







Fe H

oc


35




Carbene complexes are also known to undergo oxidative

cleavage when treated with dimethylsulfoxide. For instance,

diphenylcarbene(pentacarbonyl)tungsten was oxidized to

benzophenone. Fp[cycloheptatrienylidene]HFP also readily

undergoes oxidative cleavage with this reagent. After

being stirred overnight in dimethylsulfoxide, Fp[cyclo-

heptatrienylidene]HFP gave tropone (20; 17%)





1 + PF6


OC








To test the generality of this synthetic method for

preparing aromatic carbene complexes, the synthesis of a

4,5-benzocycloheptatrienylidene complex was attempted.

In fact, this test proved to be quite rewarding. The

synthesis of kl-4,5-benzocycloheptatrienylidene-h -cyclo

pentadienyldicarbonyliron hexafluorophosphate (Fp[4,5-

benzocycloheptatrienylidene]HFP) (41) is outlined below.

A detailed look at this synthesis follows.


O Br Br- Br









36 37 38

Li




Fe
OC








The preparation of 4,5-benzotropone (36) has been
41
reported by Paquette and Ewing. Even though the synthesis

of 1,2-benzo-5-bromocycloheptatriene (38) had not been

reported at the time of its preparation in this laboraotry,

F6hlisch et al. reported an almost identical procedure at
42
a later date. Treatment of 4,5-benzctropone (36) with

oxalyl bromide gave 1,2-benzo-5-bromotropylium bromide

(37) in the form of a red solid. Treatment of a solution

of this bromide with lithium aluminum hydride readily gave

38 in 65% yield.



0 Br Br
Br

I0\ 0 \
11 if + / LiAIH
BrC-CBr L 4





36 37 38







Treatment of 38 with an equivalent amount of n-butyl-

lithium gave a dark evergreen colored solution. Addition

of this newly generated lithium reagent 39 to cyclopenta-

dienylirondicarbonvi iodide (FpI) gave 1,2-benzo-5-(cyclo-

pentadienylirondicarbonyl)cycloheptatriene (40; 26%) as

a yellow, air sensitive solid. Its 1H rnm spectrum is

given in Figure 3.








K,
i' ~C)





a I.



_ii
i-if C



I i 0





K s -c~ ; o



3 C)
4- fi C

I i:n











Br



I


Li




n-BuLi


FpI >


,Fe


When a methylene chloride solution of 40 was treated with

triphenylcarbenium hexafluorophosphate, it rapidly turned

from yellow to red. Workup of the reaction solution gave

41, (93%), a fire-brick red solid, whose elemental analysis

is consistent with a Cp(CO)2Fe(Cl!H8)PF6 formula. Its IH

nmr spectrum is given in Figure 4.


Fe Ph CPF
OC /




40


I PF -
e+" 6
oc" v \
OC



41



















*H





0



CL
- )
O
r-





















0
1u
















. u

O-
aU




















0
r-,N














C
i1












4-,
o
-ld

























C)


c
Q
'-n

_. ^"











5 -








The structure of formula Cp(CO) 2Fe(C11H )PF6 could be

consistent (as in the case of Fp[cycloheptatrienylidene]HFP)

with either a carbene complex or an allene complex (i.e. 42).

However, as before, the allene structure may be ruled out




+ +
CpFe(CO)2 CpFe(CO)2











42a 42b



on the basis of the low field proton resonances of the ligand.

Whereas Cp(CO) 2F(allene) shows resonances at 56.59, 6.25,
4
and 3.3," the proton resonances of C-2, 3, 10, and 11 in

?p[4,5-benzocycloheptatrienylidene]HFP (41) appear at 69.58

and 8.26. From this it is clear that the allene structure

is unlikely and the preferred structure is the carbene

complex 41.

Complex 41 is an air stable solid with a decomposition

point of 177-1780C. Furthermore, even after being stored

as a solid for four months in a dark bottle, it showed no

sign of decomposition.

As seen in Fp[cycloheptatrienylidene]HFP, the proton

resonances of 41 are also significantly shifted downfield.








Whereas Fp[cycloheptatrienylidene]HFP shows resonances at

610.01, 8.48-8.74, and 7.94-8.3, the resonances of 41

appear at 69.58 and 8.26 (for C-2, 3, 10, and 11). Inter-

estingly, in the fused benzene ring the proton shifts at

68.58 to 8.08 are downfield from most organic aromatic

resonances. This indicates that the positive charge is

delocalized throughout the pi system.

In the ir spectrum, the carbonyl stretching frequencies

of Fp[4,5-benzocycloheptatrienylidene]HFP at 2045 and 2000
-i
cm are very close to those observed for Fp[cyclohepta-
-l
trienylidene]HFP at 2045 and 1996 cm This suggests
13, 35
almost identical back bonding to the carbene ligands. '

Thus the amount of positive charge on each carbene ligand

is roughly equivalent.

Perhaps the most intriguing data are seen in the 13C

nmr spectrum. Whereas the Fp[cycloheptatrienylidene]HFP

shows its C-l resonance at 6242.3, the resonance of

Fp[4,5-benzocycloheptatrienylidene]HFP appears at 6265.9.

This suggests that there is more positive charge on the

carbene carbon of Fp[4,5-benzocycloheptatrienylidene]HFP

than that of Fp[cycloheptatrienylidene]HFP.

As with Fp[cycloheptatrienylidene]HFP, Fp[4,5-benzo-

cyclcheptatrienylidene]HFP also undergoes reduction. Treat-

ment of the complex with lithium aluminum deuteride

generated 1,2-benzo-5-(cyclopadentadienylirondicarbonyl)-7-

deuterocycloheptatriene (43; 62%).















LiAlD4


Furthermore, nucleophilic attack with methyllithium
generated the methyl substituted complex 44 in 38% yield.


I 4+ PF,
Fe
oc


36


As indicated above, in principle, it is possible that
the complex of cycloheptatrienylidene could exist as
either a carbene or an allene species. In the case of
Fp[cycloheptatrienylidene]HFP and Fp [4,5--benzocyclohepta-
trienylidene]HFP, there can be little question but that the
carbene form is the dominant and perhaps exclusive form.


47


47







With an eye toward biasing the ligand in the direction of
an allene form, it was thought that 3,4-benzo-annelation
might favor an allene complex 45 over the carbene complex 46.


,+
Fe
0 C '__


The synthesis of a 3,4-benzo-annelated system is outlined on
page 33. A detailed discussion of this synthesis follows.
Addition of dibromocarbene to 1,2-dihydronaphthalene43
generated the dibromocyclopropane 47. This dibromo compound
was thermally unstable and upon distillation spontaneously
eliminated hydrogen bromide with a concomitant ring expan-
sion to give the mixture of isomeric bromides 43 and 49.


48
Br




49


CHBrI
maCH


(': D-1













K<-K


Br



48
> +
Br



49




I
OC




52


-> +
Li



51

-^>


Fe+ 6
o c



46








The preparation of 1,2-benzo-4- and 1,2-benzo-6-

lithiocycloheptatrienes (50 and 51, respectively) was

accomplished by treating these bromides with n-butyllithium.


48 49


Li


nBuLi




50


The solution of 50 and 51 was then added to a solution

of cyclopentadienylirondicarbonyl iodide (FpI) giving an

isomeric mixture of 1,2-benzo-4- and 1,2-benzo-6-(cyclo-

pentadienylirondicarbonyl)cycloheptatrienes (52 and 53,

respectively). The 1H nmr spectrum of the mixture is given

in Figure 5.


+
Li

NY


. Fe



i


FPI


Fe


Li



51










I A
P


2 -





I 'I














- ' I I
Si !









H-I

I- K
t I"

^ } i "1 N l
^ ' ; c *
L^ -fc- : o






36

The generation of a brick red solid whose elemental

analysis is consistent with Cp(CO)2Fe(C1,H8)PF was smoothly

accomplished via treatment of 52 and 53 with triphenyl-

carbenium hexafluorophosphate. The 1H nnr spectrum of 46

is shown in Figure 6.





Ph3CPF
Fe 3 6 PF
C/ C+ Fe > ) -Fe







52 53 46




Surprisingly, from the 1 nnr spectrum, an allene

structure may be ruled out on the basis of the low field

resonances that this material exhibits. Whereas Cp(CO)2Fe -

(allene) shows resonances at 66.59, 6.25, and 3.3,34 the

resonances of 46 appear at much lower field (510.4, 9.92,

9.93, 8.5, and 8.16). Therefore, even 3,4-benzo-annelatio-n

of a cycloheptatrienylidene complex -failed to favor an

allene complex and instead gave n- 3,4-benzocycloheptatrienyi-

idene-h -cyclopentadienyldicarbonyliron hexafluorophosphate

(Fp[3,4-benzocycloheptatrienylidene]HFP) (46).

This compound gives dark violet crystals from methylene

chloride and has a rather high decomposition point (176-1777C)

Moreover, 46 as a solid is perfectly stable to air. However,











































0


C)
U


I-,



I


a)

a












41



U (:
a?


C)
















0*H


0

4-
4JC)l

C),
C)( nc











ca





(00


C) C




rC)






iC)
C)O





38

in contrast to Fp[cycloheptatrienylidene]HFP, 46 does slightly

decompose after storing several months in a dark bottle.

The most interesting characteristic of 46 is its unusual

1H nmr spectrum. As seen in Figure 6, its H nmr spectrum

exhibits the lowest field resonance (at 610.4, C-2) of the

aromatic carbenes. In fact, this resonance is 0.4 ppm lower

field than that of Fp[cycloheptatrienylidene]HFP at 610.01.

The fused benzene ring's proton resonances appear around

68.50. These are only slightly different from the resonances

of Fp[4,5-benzocycloheptatrienylidene]HFP at 68.58-8.08.

This indicates qualitatively that the amounts of positive

charge lying on the fused benzene rings of both complexes

are roughly equivalent.

However, the carbonyl stretching frequencies in the ir

spectrum do point to a difference. Whereas Fp[3,4-benzo-

cycloheptatrienylidene]HFP shows absorptions at 2037 and

1992 cm-, the absorptions of Fp[4,5-benzocycloheptatrienyl-
-l
idene]HFP appear at 2045 and 2000 cm-1. This difference is

significant and suggests that there could be less back bonding

to the carbene ligand in Fp[3,4-benzocycloheptatrienylidene]-

HFP than in Fp[4,5-benzocycloheptatrienylidene]HFP.13

The C-l 1C nmr chemical shift of Fp[3,4-benzccyclo-

heptatrienylidene]HFP is also significantly different from 36.

As mentioned earlier, the resonance of Fp[4,5-benzocyclo-

heptatrienylidene]HFP at 5265.9 lies significantly lower

field than that of Fp[cycloheptatrienylidene] at 6242.3.

Conversely, the resonance of Fp[3,4--benzocycloheptatrienyl-

idene]HFP is higher field at 5201. It is interesting to








note that these shifts correspond to what classical resonance

theory predicts for three compounds whose resonance forms

may be represented by 54, 33, and 55.




+ + +
CpFe(CO)2 CpFe(CO)2 CpFe(CO)2
6+


+ 6 6+6

6+



54 33 55





An x-ray structural determination of FpL3,4-benzocyclo-

heptatrienylidene]HFP was performed by Davis.44 The computer

drawing shown in Figure 7 shows a fairly planar carbene

ligand. It is interesting to note that the smallest angle

in the seven membered ring is the C(2)-C(1)-C(11) angle

and the longest bond is the C(3)-C(8) bond. Furthermore,

some degree of bond alternation is present. A full tabula-

tion of bond angles and distances is given in Table 2.

From classical resonance theory, annelation of a second

benzene ring in the 5,6-position of 3,4-benzocycloheptatrienyl-

idene should further enhance a cycloheptatriene form. This
45
also is consistent with INDO calculations.45 The synthesis

of h -3,4-5,6-dibenzocycloheptatrienylidene-h -cyclopenta-

d ienyldicarbonyliron hexafluorophosphate (56) (or the




















01










4j





-4r
u,








0

C)0
'3






4A0


0
Ci.
























4j -1
0 -

U ra
>,C
0,0
NC,

00





ICI




t&( 01


Ha







u0.0







Table 2. Bond Lengths and Bond Anc les for -Benzocyclohepta-
trienylidene-hS -cyclopentadienyldicarbonyliron Hexa-
fluorophosphate (46).

a. Bond Lengths, A

Fe-C(l) 1.996 C(7)-C(8) 1.429
Fe-C(17) 1.769 C(8)-C(9) 1.400
Fe-C(18) 1.766 C(10)-C(11) 1.388
C(1)-C(2) 1.395 C(12)-C(13) 1.387
C(1)-C(11) 1.407 C(12)-C(16) 1.418
C(2)-C(3) 1.416 C(13)-C(14) 1.412
C(3)-C(4) 1.429 C(14)-C(15) 1.402
C(3)-C(8) 1.435 C(15)-C(16) 1.403
C(4)-C(5) 1.358 C(17)-0(1) 1.140
C(5)-C(6) 1.402 C(18)-0(2) 1.138
C(6)-C(7) 1.346
Fe-C(cp) 2.099, 2.101, 2.103, 2.098, 2.106


b. Bond Angles, deg.


Fe-C(1)-C(2)
Fe-C(1)-C(11)
C(1)-Fe-C(17)
C(1) -Fe-C(18)
C(1)-C(2)-C(3)
C(1)-C(11)-C(10)
C(2)-C(1)-C(11)
C(2)-C(3)-C;4)
C(2)-C(3)-C (8)
C(3)-C(4)-C(5)
C(3)-C(8)-C(7)
C(3)-C(8)-C(9)
C(4)-C(3)-C(8)


C(4)-C(5)-C(6)
C(5)-C(6)-C(7)
C(6)-C(7)-C(8)
C(7)-C(8)-C(9)
c(8)-C(9)-C(10)
C(9)-C(10)-C(11)
C(12)-C(13)-C(14)
C(12)-C(16)-C(15)
C(13)-C(12)-C0(16)
C(13)-C(14)-C(15)
C(14)-C(15)-C(16)
C(17)-Fe-C(18)








allene complex 57) was therefore attempted. If a carbene

allene isomerization could be effected in an unsaturated

seven membered cyclic system (with iron as the metal),

it was thought that this system offered the best chance

for success.


I PF





56


Fe
OC

57-
oc
57


The synthetic route for the generation of 6-(cyclo-

pentadienylirondicarbonyl)-5H-dibenzo[ac]cycloheptene (60),

a potential precursor for a carbene (or allene) complex,

is given below.
Br Br




\/P \/-S


Br Li






59


Fe

ccl

0C_






43

The preparation of 5,6-dibromo 5H dibenzo[a, acyclo-

heptene (58) was accomplished through a dibromocarbene

addition to phenanthrene followed by a ring expansion. Due

to the difficulty encountered in separating the product from

phenanthrene, a large excess of bromoform and base was used

so as to completely eliminate the phenanthrene. However,

difficulty in separating a polymeric material from the

product still resulted in fairly low yields of pure material.

Work up of the reaction mixture and fractional recrystal-

lization from benzene gave the ring expanded dibromide 58

(12.4%) suitable for further use.



Br Br Br


CHBr3 Br

NaOH-i


58






Reduction of the dibromide 58 gave the desired bromide

59. However, the reduction with lithium aluminum hydride

must be done at low temperature (-78C) with subsequent

slow warming to room temperature to maximize yields. Reduc-

tion at a higher temperature led to 1,2-3,4-dibenzocyclo-

heotatriene.









Br

Br



58


LiAlH4


Br






59


Treatment of 6-bromo-SH-dibenzo[a,c]cycloheptene (59)

with n-butyllithium gave an aqua-green solution. Addition

of this to cyclopentadienylirondicarbonyl iodide (FpI)

gave in 31% yield 6-(cyclopentadienylirondicarbonyl)-5H-

dibenzo[a,c]cycloheptene (60). Its 1H nmr spectrum is given

in Figure 8.


Br Li




59 FpI

59


(D


The attempted synthesis of the carbene complex (or

allene complex) from 60 was unsuccessful. In a typical

reaction, triphenylcarbenium hexafluorophosphate was intro-

duced into a cold (-30C) methylene chloride solution of 60.

Even after five days of stirring at low temperature, the

starting complex was recovered. Triphenylmethane or carbene

complex was not detected.



















C


Ln



V-1
-4

0
Cd









0
4
j




41









4-4
0



-4J








0-.


0~4


-4 C
C) r
C)
C:


s



















Fe
OC / Ph CPF
OC / --- NO REACTION




60




Since all attempts to generate the carbene (or allene)

complex by hydride abstraction were unsuccessful, a second

preparative method was attempted. From Brookhart's

successful alpha methoxy abstraction with triphenylcarbenium

ion in the generation of the phenyl carbene complex 18,

vide supra, and the fact that olefin complexes (e.g. 61)

are formed with relative ease via treatment of beta oxides

with acid,46 it was thought that a compound such as 6-(cyclo-

pentadienylirondicarbonyl)-5-methoxy-5H-dibenzo[a, ]cyclo-

heptene (62) might be the key to the synthesis of the desired

carbene (or allene) complex.




+ e+

OC F OC e
OC OC
6
61

















C F~Fe




62


or P C+
or Ph C+ -


The preparative route to this second potential carbene

precursor is shown below. A detailed discussion of its

synthesis follows.


Br
H

Br




58


Br



H

363

63


Li
H

OCH




64


Fe


-----7






48


Treatment of 5,6-dibromo-5H-dibenzo[a,c]cycloheptene

(58) with sodium methoxide in refluxing methanol readily

afforded 6-bromo-5-methoxy-5H-dibenzo[a, c]cycloheptene

(63; 91%). Treatment of a solution of 63 with n-butyllithium

gave the 6-lithio-5-methoxy-5H-dibenzo[a, ccycloheptene (64).


58 63 64


Addition of 64 to cyclopentadienylirondicarbonyl

iodide (Fp!) gave 6-(cyclopentadienylirondicarbonyl)-5-

methoxy-5H-dibenzo[a,c]cycioheptene (62; 19%) as a yellow,

air sensitive oil. Its IH nmr spectrum is given in Figure 9.


Li


OCH



A/ H


Fpl
>






49







I.
i






4.1
0
Lrl




>"
0








.a
a




I0
"-4





r -' "i



"L | -




-I
- .g '
.1
9 ft




- 3 -
L _. !-
"-1 I :- 1(
I 0 -S
; 0 0
1 "0,









Treatment of 62 with triphenylcarbenium hexafluoro-

phosphate resulted in the generation of a yellow solid 66

that precipitated out of solution upon addition of ethyl

ether at low temperature. Triphenylmethylmethyl ether was

present whereas the starting complex was not present in

the filtrate (by H nmr). A low temperature H nmr

spectrum (0C, CD3NO2) of the yellow solid 66 shows a

complex pattern (Figure 10). Upon warming to room tempera-

ture in the nmr tube, the spectrum dramatically changes

(Figure 11). The ir spectrum (KBr) of the yellow solid

66 shows the presence of carbonyl stretching and hexafluoro-

phosphate bands (Figure 12).

It has been shown that nucleophilic attack on Fp
17 47
carbene or allene (e.g. 65)47 complexes leads to neutral

sigma complexes. Therefore if a carbene (or allene)





I R' R R'
COC C F +Fe
R1 OC R F
OC C R'

"" ,65
R" R


complex had been generated, the reaction with a nucleophile

should lead back to a neutral species that could be stable

in solution. Reaction of the yellow solid 66 with sodium

methoxide in methanol led to two identifiable products,

triphenylmethylmethyl ether and 1,2-3,4-dibenzocyclo-

heptatriene. However, no neutral transition metal complex






































c

CL
C-)


A


CN







i

U
2

]










-0
j lj




2--
4j


















CC)
Cl~

-H
























;i I


-4

-o

C

0



-4

0

0




4i










C)

0







53






I I






-i1-





-II I





















* -


I ,cr









_i I



_---H- T -








was detected. Therefore, speculation as to the structure

of this yellow solid is almost impossible at this time.








Ph COCH,
Fe H Yellow
OC OCH3 >- solid ----






62


In the event that a labile carbene (or allene) ligand

was formed (as in the case of Cp(CO) Fe(methylene) (15)15, 16

vide supra), 62 was treated with acid in the presence of

a trap. However, the expected tetracyclone adduct 67 was

not detected.


0 Ph

Ph Ph OPh h


Fe H OC Ph Ph Ph


oC HC1


62 67
67



A collection of selected spectral data of some Fp

complexes is given in Table 3. In summary, it can be seen















OHHNN
ci -i
z:


I I ~ I
I I I


c4 cN C%4 l
I *c \
rN N ~N


. O 0
0"1 0
C1 0



un Ln
o o








0 a








I






>.
C o





ar O
O (D


ri a

r^ r*
4 H
^ U



0. N


o a
rl I
U Ls




h 0.


O
CI
O
N


Lnl


0 0




O 0
0 0

CN (N


r--4










0
I


-a

0

f.)
o














N
(U
j


.Q





56

that the amount of back donation afforded to the carbene

ligands in 24, 41, and 56 is less than that in Fp[benzocyclo-

butenylidene]HFP (16). However, comparison of the carbonyl

stretching frequencies of the aromatic carbene complexes

with those of Fp[vinyl] and Fp[cycloheptatriene] suggests

that there is still substantial back bonding present in 24,

41, and 46. The 1C nmr data indicate that the amount

of positive charge on the carbene carbons of 24, 41, and 46

is less than that of Fp[phenylcarbene] vide supra.

In conclusion, the synthesis of three aromatic carbene

complexes of iron strongly suggests the general applicability

of this synthetic method to other aromatic carbene ligands.

In addition, the synthesis of a tungsten cycloheptatri-

enylidene complex by the same procedure indicates that this

synthetic method need not be limited to only one particular

transition metal or one particular charge type.















CHAPTER III
EXPERIMENTAL

General. All melting points were obtained on a Thomas

Hoover melting point apparatus and are uncorrected. All

60 MHz nuclear magnetic resonance data were obtained on a

Varian A-60A or Jeol PMX 60 spectrometer and are reported

in units of 6 from tetramethylsilane which was used as an

internal standard. All 100 MHz nuclear magnetic resonance

data were obtained on a Varian XL 100 or Jeol FX 100

spectrometer and are reported in the same manner. Infrared

data were obtained on a Beckman IR 10 spectrophotometer,

Perkin-Elmer 137 spectrophotometer, or a single beam Perkin

Elmer E-14 spectrophotometer. Combustion analyses were

performed by Atlantic Microlab Inc., Atlanta, Georgia.

Mass spectra were obtained on an AEI MS 30 spectrometer.

Unless otherwise noted, solvents used were reagent grade

except tetrahydrofuran and methylene chloride. The tetra-

hydrofuran was purified by distillation over sodium-potassium
48
alloy as described by Ward. Reagent grade methylene

chloride was distilled from P205 under nitrogen and used

immediately.

The alumina used (except where noted otherwise) was

Fisher certified neutral alumina, Brockman Activity 1, to

which 3% w/w water was added. It was purged several times

with nitrogen and allowed to stand for two or more days.









The silica gel used was prepared by Baker and was 60-300

mesh. All reactions were performed under either a nitrogen

or an argon atmosphere. All organometallic compounds were

prepared using a Schlenk type apparatus and kept under nitro-

gen or argon at all times except where noted otherwise.

Cyclopentadienylirondicarbonyl iodide. This material

was obtained from Alfa Products.

Preparation of tropone (20). This compound was pre-

pared by a modified procedure of Radlich49 from cyclo-

heptatriene. Instead cf allowing the reaction mixture to

stand overnight, it was stirred with a mechanical stirrer

in a Morton flask. This improved the yield to 37% (reported

yield 25%).

Preparation of bromotropylium bromide. This material

was prepared according to the method of Fohlisch et al.2

Preparation of 1--, 2-, and 3-bromioccloheptatrienes (21).

This isomeric mixture was prepared according to the method

described by FShlisch and Haug.24

Prepr-ation of 1-, 2-, and 3-1ithiocycloheptatrienes (22).

To a cold (dry ice and isopropanol) addition funnel, 16.5 ml

(26.4 mmoles, 1.6 molar solution in hexane) of n-butyllithium

was added dropwise with a syringe to a stirred solution of

4.32 g (2C.4 mmoles) 1-, 2-, and 3-bromocycloheptatrienes in

30 ml of tetrahydrofuran. The dark areen solution was allowed

to stir for 30 niinutes and was used immediately.

Preparation of 1-. 2-, and 2-(cyclopentadienvliron-

dicarbo-iyl)c cclohertatrienes (23). A cold solution of

26.4 mnm es) 1-, 2-, and 3-lithiocyclcheptatrlenes in









tetrahydrofuran was added to a rapidly stirred solution of

8.2 g (26.9 mmoles) cyclopentadienylirondicarbonyl iodide

in 20 ml of tetrahydrofuran cooled by a dry ice-isopropanol

bath. The reaction mixture was stirred for 45 minutes.

It was then warmed to room temperature and stirred an addi-

tional 20 minutes. To the reaction mixture, a small amount

of alumina was added and the tetrahydrofuran was removed

in va-zco. The reaction products coated on the alumina were

chromatographed on alumina using pentane as the eluent.

Two simultaneously eluted bright yellow bands were collected.

Removal of the solvent in vacio (rotary evaporator) gave

2.02 g (29%) of a yellow, air sensitive oil.

The spectral data were as follows: 1H nmr (CDC1 ),

Figure 1, 62.15, 2.25, 2.62, (2H, t, t, d, satd. CH), 4.65,

4.79, 4.84 (5H, s, s, s, Cp), 4.98-5.50, 5.78-6.83 (SH, br m,

br n, vinyl); ir (neat) 3020, 2020, 1950, 1495, 1430, 1015,

830, 725 cm ; ms m/e 268.01841 (calcd. 263.01860).

Preparation of h l-cvcloheptatrienvlidene-?--cycloenta-

dienyldicarbocnyliron hexafluorophcsphate (24) A solution

of 2.02 g (7.55 moles) of 1-, 2-, and 3-(cyclopentadienyi-

irondicarbonr.yl;cloheptatrienes in 10 ml of methylene chloride

was cooled in a dry ice and isoprocanol bath. To the stirred

solution, 2.93 g (7.55 rnoles) of triphenvlcartenium hexa-

fluorophosphaze in 30 ml of methylene chloride was added

with a syringe. The reaction mixture was stirred for 30

minutes and then was allowed to warm to room temperature.

Immediately after 30 minutes of additional stirring, ac. 20 ml









of methylene chloride was rapidly removed in vacua and ca.

60 ml of ethyl ether (freshly opened can) was rapidly

added via syringe to precipitate the complex. The solid was

immediately collected on a medium grade glass frit and was

washed with several 15 ml portions of ethyl ether. The

remaining solvent was removed in vacuo. This yielded 2.8 g

(90%) of a yellow-orange, air stable solid. Recrystalliza-

tion from methylene chloride gave orange-yellow plates,

m.p. 180.0-180.5C (dec.).

The scectral data were as follows: 1H nmr (00C, ace-

tone d,), Figure 2, 610.01 (2H, d, J=10.0, H-2, 7), 8.48-

8.74 (2H, m, H-4, 5), 7.94-8.30 (2:;, m, H-3, 6); 5.50 (5H,

s, Cp); decouples "C nmr (0C, acetone d6) 6242.3 (C-l),

212.8 (CO), 170.0, 148.3, 138.2 (C-2, C-3, C-4), 88.3 (Cp);
-l
ir (CH2C12) 2045, 1996 cm-1 (KBr) 3130, 2040, 1980, 1460,

1005, 830, 730, 590, 555 cm ; UV-visible acetonitrilee):

ax (log E)-205 (4.41), 220 (sh), 340 (sh), 405 (3.60).

Analysis calculated for C H11 F 6FeP: C, 40.81;

H, 2.69. Found: C, 40.55; H, 2.71.

Reaction of h -cycloheptatrienylidene-h -cyclopentadienvl-

dicarbonyliron hexafluorophosphate with lithium aluminum

hydride. Powdered lithium aluminum hydride (12.9 mg (0.340

moles) was added to 0.201 g (0.489 mmoles) of h -cyclo-

heptatrienylidene-h -cyclopentadienyldicarbonyliron hexafluoro-

phosphate in 15 ml of tetrahydrofuran. After a few minutes

of stirring, alumina was added and the tetrahydrofuran was

rapidly removed in vacuo. The product was chromotographed








on alumina. Pentane eluted a yellow band which was taken

to dryness in vacuo (rotary evaporator) and yielded 88 mg

(67%) of pure 1-, 2-, and 3-(cyclopentadienylirondicarbonyl)

cycloheptatrienes (23) which were identified by 1H nmr.

No organic products were detected.

Reaction of h -cvcloheptatrienylidene-h -cyclopentadienyl-

dicarbonyliron hexafluorophosphate with methyllithium. To

a stirred solution of 0.113 g (0.274 mmoles) l -cyclohepta-

trienylidene-h -cyclopentadienyldicarbonyliron hexafluoro-

phosphate in 10 ml of tetrahydrofuran was added 0.15 ml

of methyllithium (0.31 mmoles, 2.05 molar solution in ethyl

ether) via syringe. After 20 minutes of stirring, alumina

was added and the solvent was rapidly removed in vacuo.

Column chromatography on alumina (with pentane and then a

5% v/v benzene-pentane mixture as eluents) produced a yellow

band which yielded 10.2 mg (13.2%) of a yellow oil 34 after

removal of the solvent in vacuo (rotary evaporator). The

proton spectrum matched that of the cycloheptatriene com-

plexes with the exception of the methyl signals appearing

at 60.8, 1.2 (3H, d, complex m). The saturated cyclohepta-

triene hydrogen were at too low intensity to be distin-

guished. The mass spectrum gave m/e 282 (M ), 254 (M -CO),

226 (M+-2CO), base peak at 56.

Oxidation of h -cycloheptatrienylidene-h'-cyclopenta-

dienyldicarbonvliron hexafluorophosphate with dimethyl-

sulfoxide. To 0.210 g (0.509 mmoles) of hl-cycloheptatri-
5
enylidene-h -cyclopentadienyldicarbonyliron hexafluorcphos-

phate in a schlenk tube, 10 ml of dimethylsulfoxide was









added via syringe. After stirring overnight, water was

added and the solution neutralized. Workup included extrac-

tion with two 250 ml portions of methylene chloride and

drying with anhydrous magnesium sulfate. After removal of

the solvent in vacuo (rotary evaporator), tropone was

separated by a preparatory thin layer plate (silica gel using

ethyl ether as the eluent) that had been spotted with an

original sample of tropone. Collection of the corresponding

band, extraction with methylene chloride and concentration

of the product in vacuo (rotary evaporator) gave 9.0 mg

(17%) of tropone that was identified by ir and 'H nmr

spectra.

Preparation of 4,5-benzotropone (36). This material

was prepared according to the method of Paquette and Ewing.4

Preparation of 1,2-benzo-5-bromotropylium bromide (37).

This material was prepared according to the method of F6hlisch

et al. and used immediately.42

Preparation of 1,2-benzo-5-bromocycloheptatriene (38).

This material was prepared by a procedure similar to that

described by F6hlisch et al.42 with the following modifica-

tions. Before the addition of lithium aluminum hydride to

1,2-benzo-5-bromctropylium bromide, all solvent was removed

in vacuo and the bromide salt was taken up in pure tetra-

hydrofuran. The lithium aluminum hydride was then added

over a 60 minute time period to the cold (0C) solution.

After slow addition of water, the solution was filtered,

extracted with methylene chloride (reagent grade), dried








with anhydrous magnesium sulfate and concentrated in vacuo

(rotary evaporator). Column chromatography on silica gel

(pentane) gave 4.39 g (65%) of a white solid. The spectral

data matched those of reported spectra.

Preparation of 1,2-benzo-5-lithiocycloheptatriene (39).

To an addition funnel cooled with a dry ice-isopropanol

mixture, 10.1 ml (16.1 mmoles, 1.6 molar solution in hexane)

of n-butyl lithium was added dropwise with a syringe to a

stirred solution of 3.85 g (17.4 mmoles) 1,2-benzo-5-bromo-

cycloheptatriene in 25 ml of tetrahydrofuran. The evergreen

colored solution was allowed to stir for 30 minutes and was

used immediately.

Preparation of 1,2-benzo-5-(cyclopentadienylirondi-

carbonyl)cycloheptatriene (40). A cold solution of 1,2-

benzo-5-lithiocycloheptatriene (17.4 mmoles) in 25 ml of

tetrahydrofuran was rapidly added to a stirred solution of

6.1 g (20 moles) cyclopentadienylirondicarbonyl iodide in

20 ml of tetrahydrofuran cooled by dry ice-isopropanol. The

cold reaction mixture was stirred for 30 minutes. A small

amount of alumina was added to the reaction mixture and the

tetrahydrofuran was removed in vacuo. The reaction products

coated on the alumina were chromatographed on alumina using

600 ml of pentane and then a benzene-pentane mixture as

eluents (gradually increasing the benzene content to a 20%

v/v benzene-pentane mixture). After two faint yellow bands

were discarded, the bright yellow band was collected.

Removal of the solvent in vacuo gave 1.36 g (26%) of a yellow,

air sensitive solid, m.p. 114.5-115.5C (dec.).








The spectral data were as follows: 1H nmr (CDC1 ),

Figure 3, 63.0 (2H, d, J=7Hz, satd. CH), 4.6 (5H, s, Cp),

5.63 (1H, t, J=7Hz, vinyl), 6.62 (2H, s, vinyl), 7.2 (4.5H,

br m, benzylic); ir (KBr) 3120, 3000, 2380, 2020, 1955, 1930,

1490, 1430, 945, 835, 790, 755, 740, 640, 590, 570 cm-1; ms

m/e 318 (M+), 290 (M+-CO), 262 (M+=2CO).

Analysis calculated for C 8H 4FeO2: C, 67.95; H, 4.44.

Found: C, 67.99; H, 4.47.

Preparation of h -4,5-benzocycloheptatrienylidene-h -

cyclopentadienyldicarbonyliron hexafluorophosphate (41).

A solution of 1.19 g (3.73 mmoles) of 1,2-benzo-5-(cyclopenta-

dienylirondicarbonyl)cycloheptatriene in 3 ml of methylene

chloride was cooled in a dry ice-isopropanol bath. To the

stirred solution, 1.45 g (3.73 mmoles) of triphenylcarbenium

hexafluorophosphate in 15 ml of methylene chloride was added

with a syringe over 10 minutes. The red mixture was stirred

for 30 minutes. It was then warmed to room temperature. Scme

of the solvent (ca. 8 ml) was removed in vacuo and ca. 60 ml

of ethyl ether (freshly opened can) was added to precipi-

tate the complex. The complex was collected on a medium

grade glass frit and was washed with several 30 ml portions

of ethyl ether. The remaining solvent was removed in vacuo

This yielded 1.6 g (93%) of a fire brick red solid.

Recrystailization from methylene chloride gave red-brown

plates, m.p. 177-178C (dec.).

The spectral data were as follows: 1H nmr (0C, aceto-

nitrile d3), Figure 4, 69.58 (2H, d, J=11Hz), 8.26, 8.58-8.08








(6H, d, J=llHz, m, A2B2), 5.3 (5H, s, Cp); decoupled 1C nmr

(00C, nitromethane d3) 6265.9 (C-l), 212.7 (CO), 158.5, 141.6,

140.3, 137.8, 137.1 (C-2 through C-ll), 89.5 (Cp); ir (CH2C12)
-i
2045, 2000 cm 1, (KBr) 2030, 1990, 1510, 1470, 965, 840,
-1
560 cm ; UV-visible (CH C2 ): A (log E) = 500 (3.47),
S2 max
468 (3.52) 330 (sh), 315 (sh), 282 (sh), 272 (4.21), 233

(4.20).

Analysis calculated for C 18H13FFeO2P: C, 46.79, H, 2.84.

Found: C, 46.78; H, 2.86.

Reaction of h -4,5-benzocycloheptatrienylidene-h -

cyclopentadienyldicarbonyliron hexafluorophosphate with

lithium aluminum deuteride. Powdered lithium aluminum

deuteride (0.016 g, 0.38 mmoles) was added to a cold (dry

ice-isopropanol) solution of 0.200 g (0.432 mmoles) h1-4,5-

benzocycloheptatrienylidene-h -cyclopentadienyldicarbonyl-

iron hexafluorophosphate in 5 ml of tetrahydrofuran. The

remaining reducing reagent was washed out of the solid

addition flask with ca. 5 ml of tetrahydrofuran. The

solution was stirred 10 minutes and then it was warmed to

room temperature. Upon warming, the solution turned from

red to yellow. The solvent was removed in vacuo during

which the products were coated on alumina. Rapid column

chromatography on alumina (pentane) and collection of the yellow

band followed by removal of the solvent in vacuo yielded 0.0850

g (62%) 1,2-benzo-5-(cyclopentadieny1 irondicarbonyl)-7-

deuterocycloheptatriene (43). The 1H nmr was identical to

that of 1,2-benzo-5-(cyclopentadienylirondicarbonyl)cyclo-

heptatriene (40) except for the doublet at 65.63 and









the broad double at 63.0 (1H); ms m/e 291.04630 (calcd.

291.0540).

Reaction of h -4,5-benzccycloheptatrienylidene-h -

cyclopentadienvldicarbonyliron hexafluorophosphate with

methyllithium. Methyllithium (0.25 ml; 0.51 mmoles of a

2.05 molar solution in ethyl ether) was syringed over a

a 2 minute time period into a cold (dry ice-isopropanol)

solution of 0.240 g (0.518 mmoles) of hl-4,5-benzocyclo-

heptatrienylidene-h -cyclopentadienyldicarbonyliron hexa-

fluorophosphate in 10 ml of tetrahydrofuran. The color

of the solution changed from red to yellow almost immediately.

After stirring for 15 minutes, the solution was warmed to

room temperature (whereby the color turned darker). A small

amount of alumina was added and solvent was removed in vacuo.

Column chromatography of the coated alumina with alumina

(pentane) and collection of the yellow band yielded 65 mg

(38%) of 1,2-benzo-5-(cyclopentadienylirondicarbonyl)-7-

methylcycloheptatriene (44).

The spectral data were as follows: 1H nmr (CDC13) 61.5

(3H, d, J=7Hz), 2.75 (1H, m, satd CH), 4.5 (5H, s, Cp) 5.25

(1H, d, J=7Hz, vinyl), 6.7 (2H, s, vinyl), 7.25 (4H, m, arom.).
43
Preparation of l-bromo-1,2,3,4-tetrahydronaphthalene.

A solution of 404 g (3.06 moles) tetralin and 500 ml of

carbon tetrachloride were heated to reflux. Using an

addition funnel, 443 g (2.77 moles) of bromine was added

over a 3.5 hour period. The mixture was allowed to stand

at room temperature overnight. The solvent was removed









in vacuo (rotary evaporator). The residue, 636.1 g, was

used as is in the next step.
43
Preparation of 1,2-dihydronaphthalene. An aqueous

sodium hydroxide solution (160 g NaOH in 500 ml of H20)

was brought to reflux. The above bromide was added over a

two hour period. After an additional 0.5 hour of reflux,

the mixture was allowed to stand at room temperature over-

night.

The mixture was twice extracted with 200 ml portions of

petroleum ether. The organic extracts were dried over

anhydrous magnesium sulfate and the solvent was removed

in vacuo (rotary evaporator). The residue was vacuum dis-

tilled. This distillate was a mixture of dihydronaphthalene,

tetralin, and naphthalene. The amount of the dihydronaph-

thalene (0.610 moles) was determined by nmr integration.

Preparation of 1,2-benzo-4- and 1,2-benzo-6-bromocvclo-

heptatrienes (48, 49). A mixture of naphthalene, tetra-

hydronaphthalene, dihydronaphthaiene (0.610 moles in a 218 g

mixture), benzyltriethylammonium chloride (0.8 g), and bromo-

form 459.4 g (1.8 moles) was mechanically stirred and

cooled with a water bath. To this mixture, 500 g of 50%

w/w aqueous sodium hydroxide was added dropwise. After 24

hours of stirring, 500 ml of H-O was added and the organic

layer was washed twice with 250 ml portions of water.

After drying with anhydrous magnesium sulfate, the organic

layer was filtered through celite. The bromoform was

removed by vacuum distillation (10 mm, 32C to 3.5 mm, 60C)








The residue was taken up in 500 ml of pentane and filtered

to remove the brown polymer. The organic products were

concentrated in vacuo (rotary evaporator).

Distillation of this residue (1.5 mm, 600C) removed the

tetrahydronaphthalene. Three distillations (3 mm, 140C;

2.75 mm, 92-118C; and finally with a Nester Faust spinning

band column at 0.25 mm, 90-1000C) resulted in loss in hydro-

gen bromide and gave a mixture of the two bromides (51.9 g;

38%) that was suitable for further use.

The soectral data were as follows: 1H nmr (CDC13)

63.08, 3.52 (2H, d, J=7Hz, s, satd. CH), 5.45-6.5 (2H, m,

vinyl), 6.9-7.55 (4.5H, m, vinyl and arom.); ir (neat) 3000,

1601, 1480, 1440, 1420, 1340, 1190, 1110, 1030, 980, 860,

780, 755 cm-1; ms m/e 219.98953 (calcd. 219.98850).

Preparation of 1,2-benzo-4- and 1,2-benzo-6-lithiocvclo-

heptatrienes (50, 51). To an addition funnel cooled with a

dry ice-isopropanol mixture and containing a stirred solu-

tion of 5.22 g (23.6 mmoles) 1,2-benzo-4-, and 1,2-benzo-

6-bromocycloheptatrienes in 32 ml of tetrahydrofuran, was

added 14.5 ml (2.32 roles, 1.6 molar solution in hexane)

of n-butyllithium dropwise with a syringe. The green solu-

tion was stirred for 30 mintues and used immediately.

Preparation of 1,2-benzo-4- and 1,2-benzo-6-(cyclo-

pentadienylirondicarbonyl)cycloheptatrienes (52, 53). A

cold solution of (2.32 mmoles) 1,2-benzo-4- and 1,2-benzo-

6-lithiocycloheptatrienes in 32 ml of tetrahydrofuran was

added rapidly to a stirred solution of 7.2 g (2.4 mmoles)









cyclopentadienylirondicarbonyl iodide in 20 ml of tetrahydro-

furan cooled with dry ice-isopropanol. The cold reaction

mixture was stirred for 45 minutes and then warmed to room

temperature. At this time, 40 ml of alumina was added to

the reaction mixture and the tetrahydrofuran was removed

in vacuo. The reaction products coated on the alumina

were chromatographed on alumina using pentane and then a

benzene-pentane mixture (gradually increasing the benzene

content to 15% v/v benzene-pentane). The yellow band was

collected. Removal of the solvent in vacuo (rotary evaporator)

gave 2.74 g (37%) of a yellow,air sensitive oil.

The spectral data were as follows: 1H nmr (CDC13),

Figure 5, 62.9, 3.45 (2H, d, s, satd. CH), 4.9 (5H, s, Cp),

5.35, 5.2, 6.7, 7.2 (8H, complex multiplets, vinyl and

aromatics); ir (neat) 3020, 2010, 1950, 1775, 1525, 1480,
-i
1430, 1020, 910, 830, 730 cm ; ms m/e 318.03302 (calcd.

318.03410).

Preparation of hl-3,4-benzocycloheptatrienylidene-h -

cyclopentadienyldicarbonyliron hexafluorophosphate (46).

A solution of 2.74 g (8.60 mmoles) of 1,2-benzo-4- and

1,2-benzo-6-(cyclopentadienylirondicarbonyl)cyclohepta-

trienes in 15 ml of methylene chloride was cooled in a dry

ice-isopropanol bath. To the stirred solution, 3.34 g

(8.60 mmoles) of triphenylcarbenium hexafluorophosphate in

ca. 30 ml of methylene chloride was added dropwise with a

syringe. The reaction mixture was stirred for one hour.

It was then warmed to room temperature. Partial rapid








removal of the methylene chloride (ca. 30 ml) in vacuo

followed by rapid addition of ethyl ether (ca. 60 ml) aided

the precipitation of the complex. The solid was immediately

collected on a medium grade glass frit and was washed with

several 10 ml portions of ethyl ether. The residual

solvent was removed in vacuo. This yielded 3.5 g (88%) of

a brick red solid. Recrystallization from methylene chloride

gave violet needles, m.p. 176-1770C (dec.).

The spectral data were as follows: H nmr (80C, nitro-

methane d3), Figure 6, 610.4 (1H, d, J=2Hz) 9.92 (1H, d,

J=10.6Hz), 9.23 (1H, d of m, J=10.6, 0.5Hz), 8.50, 8.16 (5H,

in, t, J=10.6Hz); decoupled 13C nmr (0C, nitromethane d3),

6215 (CO), 201 (C-l), 176, 174, 155, 145, 139, 138, 138, 137,136,
-i
133 (C-2 through C-ll), 89 (Cp); ir (CH2Cl2) 2037, 1992 cm 1
-I
(KBr) 3140, 2025, 1975, 1420, 840, 595, 560 cm ; UV-visible

(CH C1,) : max (log ) = 380 (3.52), 313 (sh), 280 (4.36),

232 (4.28).

Analysis calculated for C, H 3F FeO2P: C, 46.79;

H, 2.84. Found: C, 46.68; H, 2.86.

Preparation of 5, 6-dibromro-5H-dibenzo[a ]cycloheptene

(58). To a mechanically stirred mixture of 50 g (0.280 mole)

phenanthrene, 890 g (3.52 moles) bromoform, 1.9 g triethyl-

benzylammonium chloride and 5 ml of ethanol, 500 g w/w NaOH

(aq) solution (6.25 moles) was added dropwise during 20

minutes. After addition, a water bath was employed to

cool the slightly exothermic reaction.









After stirring overnight, 500 ml of water and 500 ml

of methylene chloride (reagent grade) was added to the

reaction mixture. The organic layer was washed twice with

500 ml portions of water and then was dried with anhydrous

magnesium sulfate. The methylene chloride was removed in

vacuo (rotary evaporator). The remaining brown residue was

transferred to a one liter flask. With rapid stirring

(mechanical stirrer), 750 ml of pentane was added and an

organic tar oiled out. This tar was saved for further

purification (next paragraph). The pentane and bromoform

mixture was decanted through a medium sintered glass funnel

and the pentane removed in vacuo (rotary evaporator). The

bromoform was distilled under high vacuum (0.025 mm, water

bath employed should not exceed 600C) and the residue was

transferred to a petri dish to evaporate traces of bromoform

vide infra.

The remaining tar in the one liter flask was taken

up in 200 ml of methylene chloride (technical grade). Once

again it was rapidly stirred with a mechanical stirrer.

Once dissolved, ca. 600 ml of pentane was added with stirring

and a brown solid precipitated out (containing a polymeric

material). This was filtered on a medium sintered glass

funnel and the filtrate was concentrated in vacuo (rotary

evaporator). This brown tar was combined with the material

isolated above in the petri dish.

After the brown oil was stirred in the petri dish for

four days under a stream of nitrogen gas, it was transferred









to a 250 ml erienmeyer flask and taken up in a minimum

amount of hot benzene. Fractional recrystallization from

benzene gave 12.2 g (12.4%) of product suitable for use

(m.p. 138.5-141.00C).

An analytical sample could be obtained by further

fractional recrystallization from benzene, m.p. 140.5-141.50C.

The spectral data were as follows: iH nmr (CDC13) 55.95

(1H, d, J = 2Hz, satd. CH), 7.45 (9H, m, aromatic and vinyl);

ir (KBr) 3000, 1620, 1470, 1430, 1200, 1150, 1010, 890, 860,

840, 810, 760, 750, 725 cm-1; ms m/e (!+-Br) 270.99283

(calcd. 270.99450).

Analysis calculated for C15H10Br2: C, 51.47; H, 2.88.

Found: C, 51.52; H, 2.92.

Preparation of 6-bromo-5H-dibenzo[a c]cycloheptene (59).

Powdered lithium aluminum hydride (0.155 g, 4.08 mmoles) was

added in small portions to a cold (dry ice-isopropanol)

solution of 5,6-dibromo-5H-dibenzo[a, c]cycloheptene in

25 ml of tetrahydrofuran. The reaction mixture was

stirred for three hours and then allowed to slowly warm

to room temperature (the dry ice-isopropanol bath should be

allowed to warm up with the flask). At this time, water

was added and the organic products were extracted with

ethyl ether and dried with anhydrous magnesium sulfate.

Chromatography with silica gel (pentane) gave 1.13 g (47%)

of a clear oil.

The spectral data were as follows: 1H nmr (CDCl3) 53.48

(2H, s, satd. CH), 6.83 (1H, s, vinyl), 7.4 (8H, m, arom.);








ir (neat) 3060, 1630, 1480, 1450, 1440, 1210, 1110, 855, 750,

730 cm-1; ms m/e 270.00354 (calcd. 270.00430).

Analysis calculated for C1H11,Br: C, 66.44; H, 4.09.

Found: C, 66.57; H, 4.13.

Preparation of 6-lithio-5H-dibenzo[a c]cycloheptene.

To an addition funnel cooled with dry ice-isopropanol,

3.3 ml (5.28 mmoles, 1.6 molar solution in hexane) of

n-butyllithium was added dropwise with a syringe to a

stirred solution of 1.42 g (5.26 mmoles) 6-bromo-5H-

dibenzo[a, c]cycloheptene in ca. 40 ml of tetrahydrofuran.

The aqua-green solution was stirred for 30 minutes and was

used immediately.

Preparation of 6-cyclopentadienylirondicarbonyl)-5H-

dibenzo[a c]cycloheptene (60). The cold solution of

6-lithio-5H-dibenzo[a,c]cycloheptene (5.27 mmoles) in te-

trahydrofuran was added to a cold (dry ice-isopropanol),

rapidly stirred solution of 1.6 g (5.26 mmoles) cyclopenta-

dienylirondicarbonyl iodide in ca. 20 ml of tetrahydrofuran.

The reaction mixture was stirred for 45 minutes whereupon

it was then warmed to room temperature. A small amount of

alumina was then added and the solvent was removed in vacuo.

The reaction products coated on the alumina were chromato-

graphed on alumina using pentane eluent first, followed

by a benzene-pentane mixture (gradually increasing the

benzene content to a 25% v/v benzene-pentane mixture). A

yellow band was collected. Removal of the solvent in vacuo

gave 0.604 g (31%) of a yellow, air sensitive glass.








The spectral data were as follows: 1H nmr, Figure 8,

63.38 (2H, s, satd. CH), 4.82 (5H, s, Cp), 6.65 (1H, s,

vinyl), 7.4 (9H, m, arom.); ir (neat) 3060, 2020, 1950,
-l
1775, 1570, 1480, 1430, 1100, 990, 825, 750 cm ; ms m/e

(M+) 340.05700 (calcd. 340.0550), (M -CO) 312.05969 (calcd.

312.06000).

Attempted synthesis of h -3,4-5,6-dibenzocyclohepta-

trienylidene-h -cyclopentadienyldicarbonyliron hexafluoro-

phosphate (56) from 6-(cyclopentadienylirondicarbonyl)-5H-

dibenzo[a,c]cycloheptene (60). Via syringe, 1.10 g (2.84

mmoles) of triphenylcarbenium hexafluorophosphate in 25 ml

of methylene chloride was added dropwise to a cooled (-350C)

solution of 1.06 g (2.87 mmoles) 6-(cyclopentadienyliron-

dicarbonyl)-5H-dibenzo[a,c]cycloheptene in 20 ml of methylene

chloride. After five days of stirring at -350C, ethyl

ether was added until a yellow salt precipitated (presumably

triphenylcarbenium hexafluorophosphate). The unreacted

starting complex was identified from the filtrate by 1H

nmr. No triphenylmethane was detected by 1H nmr.

Preparation of 6-bromo-5-mrethoxy-5H-dibenzo [a, c]cyclo-

heptene (63). 6,7-Dibromo-5H-dibenzo[a,c]cycloheptene (1.9 g,

5.43 mmoles) was dissolved in 20 ml of methanol (dried by

distillation from magnesium). To this, 0.3 g (5.55 mmoles)

of sodium methoxide was added. The solution was stirred and

refluxed for one hour. The reaction solution was extracted

with ethyl ether and washed several times with water. This

yielded 1.49 g (91%) of a clear oil which was used without

further purification.








The spectral data were as follows: IH nmr (CDC1 ) 63.33

(3H, s, methyl), 4.52 (1H, br s, satd. CH), 7.00 (1H,

s, vinyl), 7.5 (8.5H, m, arom.); ir (neat) 3060, 2940,

2840, 1620, 1480, 1445, 1350, 1205, 1135, 1120, 1090, 750,
-l
730 cm-1; ms m/e 300.01212 (calcd. 300.01490).

Preparation of 6-lithio-5-methoxv-5H-dibenzo[a,c]cyclo-

heptene (64). To a stirred cold (dry ice-isopropanol) solu-

tion of 1.89 g (6.29 mmoles) 6-bromo-5-methoxy-5H-dibenzo-

[a,c]cycloheptene in 15 ml of tetrahydrofuran, 3.9 ml

(6.24 mmoles, 1.6 molar solution in hexane) of n-butyl-

lithium was added dropwise with a syringe. This reaction

mixture was stirred for 30 minutes and was used immediately.

Preparation of 6-(cyclopentadienylirondicarbonyl)-5-

methoxy-5H-dibenzo[a,c]cycloheptene (62). Cyclopentadienyl-

irondicarbonyl iodide (2.3 g, 7.57 mmoles) in 20 ml of

tetrahydrofuran was added dropwise with a syringe to a cold

(dry ice-isopropanol) solution of 6-lithio-5-methoxy-5H-

dibenzo[a,c]cycloheptene (6.29 mmoles) in 15 ml of tetra-

hydrofuran. This solution was stirred at -780C for 30 minutes.

It was then warmed and stirred for an equal amount of time

at room temperature. The addition of a small amount of Woelm

alumina (Brockmann activity grade II, neutral) was followed

by removal of the solvent in vacuo. The products on the

coated alumina were chromatographed on Woelm alumina (activ-

ity grade II neutral alumina that had been stored under

N2 for a few days). The column was eluted first with pen-

tane and then the eluent's polarity was gradually increased









with benzene (up to a 25% v/v benzene-pentane mixture).

Collection of all but the first portion of the bright yellow

band gave 0.473 g (19%) of an air sensitive oil.

The spectral data were as follows: W nmr (CDC13),

Figure 9, 63.33 (3H, s, methyl), 4.08 (1H, d, J=2Hz, satd. CE),

4.75 (5H, s, Cp), 6.68 (2H, d, J=2Hz, vinyl), 7.45 (10H, m,

arom.); ir (neat) 3100, 3050, 2940, 2000, 1950, 1470, 1430,

1110, 905, 730 cm1; ms m/e 370.0683 (calcd. 370.06540).

Attempted Synthesis of h -3,4-5,6-dibenzocycloheota-

trienylidene-h5-cyclooentadienyldicarbonyliron hexafluoro-

phosphate (56) from 6-(cyclopentadienylirondicarbonyl)-

5-methoxy-5H-dibenzo[a,c]cycloheptene (62). Via syringe,
-3
0.400 g (1.03x10 -3moles) of triphenylcarbenium hexafluoro-

phosphate in ca. 6 ml of methylene chloride was added drop-

wise over a 10 minute period to a cold (dry ice -isopropanol)

stirred solution of 0.426 g (1.06x10-3 moles) 6-(cyclopenta-

dienylirondicarbonyl)-5-methoxy-5H-dibenzo[a,o]cycloheptene

in 7 ml of methylene chloride. After stirring for 30

minutes, the solution was warmed up to room temperature

over a 15 minute time period. It was then cooled down

again (dry ice-isopropanol) and 50 ml of cold (dry ice-

isopropanol) ethyl ether (freshly opened can) was syringed

into the reaction mixture. The yellow solid 66 that pre-

cipitated was collected on a medium grade glass frit and

washed with several 10-30 ml portions of cold ethyl ether.

The filtrate was collected and the solvent removed

in vacuo (rotary evaporator). Triphenylmethylmethyl ether

was present in the filtrate (identified by 1H nmr).








The yellow, extremely air sensitive solid 66 quickly

reacted with any nmr solvent (acetone, nitromethane, and

acetonitrile) at room temperature. However, at 0C (CD3NO2)

H nmr resonances appear as shown in Figure 10. Its ir

spectra (KBr) is given in Figure 12. Upon warming in the

nmr probe, the solid gave the spectrum shown in Figure 11.

Reaction of yellow solid 66 with sodium methoxide.

To a cold solution (dry ice-isopropanol) of 0.185 g of

yellow solid 66 in 5 ml of methylene chloride, 17.8 mg

(0.330 mmoles) of sodium methoxide was added with a solid

addition funnel. The remaining solid was washed into the

reaction mixture with ca. 5 ml of methylene chloride.

After warming to room temperature and stirring for five

minutes, no visible color change occurred and undissolved

sodium methoxide remained. At this time, 4 ml of methanol

was added via syringe. The solution's color changed from

dark to light red. After an additional five minutes of

stirring, the solvent was removed in vacuo. Methylene

chloride (10 ml) was then added, the solution was cooled

again (dry ice-iscpropanol) and 50 ml of ethyl ether was

added. The solution was filtered and the filtrate

concentrated in vacuo (rotary evaporator). Column chromatog-

raphy of the filtrate products with silica gel (pentane)

gave dibenzocycloheptatriene and triphenylmethylmethyl ether.

Reaction of 6-(cyclopentadienvlirondicarbonyl)-5-methoxy-

5H-dibenzo[a,c]cycloheptene with hydrogen chloride in the

presence of tetracyclone. Gaseous hydrogen chloride was






78


bubbled for one hour into a cold (dry ice- isopropanol)

stirred solution of 79.2 mg (0.199 imnole) 6-(cyclopenta-

dienylirondicarbonyl)-5-methoxy[a,o]cyclohept-6-ene and

0.299 g (0.596 mmole) tetracyclone in 4 ml of methylene

chloride. After 30 minutes of additional stirring, the

reaction mixture was allowed to warm to room temperature

and stirred for 30 minutes. A small amount of silica

gel was added and the solvent removed in vacuo. Column

chromatography on silica gel (pentane) gave 2.6 mg of

dibenzocycloheptatriene that was identified by 1H nmr

and mass spectrum.














REFERENCES

1. D.J. Cardin, B. Centikaya, and M.F. Lappert, Chem. Rev.,
72, 545 (1972).

2. F.A. Cotton and C.M. Lukehart, Prog. Inorg. Chem.,
16, 487 (1972).

3. E.O. Fisher and A. Massb61, Angew. Chem., Int. Ed.
Engl., 3, 580 (1964).

4. E.O. Fisher and U. Klabunde, J. Am. Chem. Soc., 89,
7141 (1967).

5. K. Ofele, Angew. Chem., Int. Ed. Engl., 8, 916 (1969).

6. K Ofele, J. Organomet. Chem., 22, C9 (1970).

7. W.A. Herrmann, Chem. Ber., 108, 486 (1975).

8. C.P. Casey, T.J. Burkhardt, C.A. Bunnell, and J.C.
Calabrese, J. Am. Chem. Sec., 99, 2127 (1977).

9. C.P. Casey, L.D. Albin, and T.J. Burkhardt, ibid., 99,
2533 (1977).

10. C.P. Casey and S.W. Polichncwski, ibid., 99, 6097 (1977).

11. C.P. Casey and T.J. Burkhardt, ibid., 96, 7808 (1974).

12. R.R. Schrock, ibid., 97 6577 (1975).

13. E.O. Fisher, R.L. Clough, G. Besl, and F.R. Kreissl,
Angew, Chem. Int. Ed. Engl., 15, 543 (1976).

14. C.H. Game, M. Green, J.R. Moss, and F.G.A. Stone, J.
Chem. Soc., Dalton, 351 (1974).

15. P.W. Jolley and R. Pettit, J. Am. Chem. Soc., 88, 5044
(1966).

16. M.L.H. Green, M. Ishaq, and R.N. Whiteley, J. Chem.
Soc., A, 1508 (1967).

17. W.P. Giering, A. Sanders, L. Cohen, D. Kenedy, and
C.V. Magatti, J. Am. Chem. Soc., 95 5430 (1973).

18. M. Brookhart and G.O. Nelson, ibid., 99, 6099 (1977).






80


19. D. DeJohn, C. Mayor, and J. Morris, University of
Florida, unpublished results.

20. W.M. Jones and Y. Kawada, University of Florida,
unpublished results.

21. K. Ofele, J. Organomet. Chem., 12, P42 (1968).

22. N.T. Allison, Y. Kawada, and W.M. Jones, J. Am. Chem.
Soc., 100, 5224 (1978).

23. B. F6hlisch, P. Burgle, and D. Krackenberger, Chem.
Ber., 101, 2717 (1968).

24. B. F8hlisch and E. Haug, ibid., 104, 2324 (1971)

25. B.J. Wakefield, "The Chemistry of Organolithium
Compounds," Pergamon Press Ltd., Elmsford, N.Y., 1974.

26. B.F. Hullman and P.L. Pauson, Chem. and Ind., 653
(1955).

27. M.L.H. Green, M. Ishaq, and T. Mole, Z. Naturforschg.,
206, 598 (1965).

28. M.L. Green and T. Mole, J. Organomet. Chem., 12, 404
(1968).

29. T.E. Piper and G. Wilkinson, J. Inorg. and Nucl. Chem.,
3, 104 (1956).

30. L. Summers, R.H. Uloth, and A. Holms, J. Am. Chem. Soc.,
77, 3604 (1955).

31. A. Eisenstadt, G. Scharf, and B. Fuchs, Tetrahedron
Lett., 679 (1971) .

32. W.M. Jones and U.H. Brinker, "Pericyclic Reactions,"
Vol. 1, A.P. Marchand and R.E. Henr. Ed., Academic
Press, New York, N.Y., 1977, Ch. 3.

33. W.M. Jones, Accounts of Chemical Research, 10, 353 (1977).

34. M. Rosenblum, B. Foxmann, D. Marten, A. Rosan, and
S. Raghu, J. Am. Chem. Soc., 99, 2160 (1977).

35. F.A. Cotton and G. Wilkinson, "Advanced Inorganic
Chemistry," 3rd Ed., Interscience, New York, N.Y.,
1972, p. 687.

36. R.B. King and M.B. Bisnette, J. Organomet. Chem., 2,
15 (1974).

37. F.A. Cotton and C.S. Kraihanzel, J. AmI. Chem. Soc.,
84, 4432 (1962).





81


38. M.Y. Darensbourg and D.J. Darensbourg, Inorg. Chem.,
9, 32 (1970).

39. M.H. Chisholm and S. Godleski, Prog. Inorg. Chem.,
20, 299 (1976).

40. C. Ciappenelli and M. Rosenblum, J. Am. Chem. Soc., 91,
6876 (1969).

41. L.A. Paquette and G.D. Ewing, J. Org. Chem., 40, 2965
(1975).

42. B. Fohlisch, C. Fisher, E. Widmann, and E. Wolf,
Tetrahedron, 34, 533 (1978).

43. E.E. Waali, University of Wisconsin, Whitewater,
personal corrunication, 1977.

44. R.E. Davis, University of Texas, Austin, personal
communication, 1978.

45. R.L. Tyner, W.M. Jones, Y. Ohrn, and J.R. Sabin, J. Am.
Chem. Soc., 96, 3765 (1974).

46. A. Cutler, D. Ehnthdt, W.P. Giering, P. Lennon,
S. Raghu, A. Rosan, M. Rosenblum, J. Tancrede, and
D. Wells, ibid., 98, 3495 (1976).

47. D. Lennon, A.M. Rosan, and M. Rosenblum, ibid., 99,
8426 (1977).

48. R.L. Ward, ibid., 83, 1296 (1961).

49. P. Radlich, J. Org. Chem., 29, 960 (1964).















BIOGRAPHICAL SKETCH

Neil Thomas Allison was born August 19, 1953, in

Athens, Georgia. At the age of one, he and his parents

moved to Macon, Georgia. He attended Lanier High School

in Macon.

In September, 1971, he began undergraduate work

at Georgia College in Milledgeville, Georgia. In 1972,

he married the former Miss Amelia Ann Hancock.

While attending Georgia College, Mr. Allison

received a Georgia College Alumni Scholarship and the

Outstanding Chemistry Major Award. In June, 1975, he

received the BS degree from Georgia College. Three

months later, he began graduate work at the University

of Florida in Gainesville, Florida.










I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.






William M. Jones, Chairman
Professor of Chemisftr





I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.



/

John F. Helling
Associate Professor of Physical
Sciences and Chemistry





I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.






William R. Dolbier, J Professor of Chemistry/











I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.






Martin T. Vala
Professor of Chemistry





I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.






Mark P. Hale, Jr.
Assistant Professor of Mathematics





This dissertation was submitted to the Graduate Faculty
of the Department of Chemistry in the College of Liberal Arts
and Sciences and to the Graduate Council, and was accepted
as partial fulfillment of the requirements for the degree
of Doctor of Philosophy.

December 1978


Dean, Graduate School




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs