OVERDUE FINES:
25¢ per day per item
RETURNING LIBRARY MAngTgQ:

Place in book return to remove
charge from circulatiur reconds

 

 

 

I.

THE DECOMPOSITION KINETICS OF NICKELACYCLOPENTANES

II. THE IN SITU DECOMPOSITION OF

NICKELACYCLOBUTANES

III. STUDIES TOWARD THE HOMOGENEOUS CATALYTIC

REDUCTION OF CARBON MONOXIDE

By

Richard Kingsley Stuart, Jr.

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirement

for the degree of

DOCTOR OF PHILOSOPHY

Department of Chemistry

1979

ABSTRACT

I. THE DECOMPOSITION KINETICS OF NICKELACYCLOPENTANES

II. THE IN SITU DECOMPOSITION OF

NICKELACYCLOBUTANES

III. STUDIES TOWARD THE HOMOGENEOUS CATALYTIC

REDUCTION OF CARBON MONOXIDE
By
Richard K. Stuart, Jr.

The thermal decomposition kinetics of a series of phos-
phine and pyridine stabilized nickelacyclopentanes were
measured. The change in activation parameters (most
notably the energy of activation) in the series was cor-
related with both the steric and electronic properties of
the ligands. It was found that the monodentate and flex-
ible chelate ligands impart lower activation energies to
the complexes while rigid chelates thermally decompose
'with higher activation energies. The trans effecting
ability of the ligands also parallel the change in activa-
tion energies. Ligands of high trans effect as in bis-
(triphenylphosphine)tetramethylene nickel(II) significantly
weaken the nickel-carbon bonds resulting in a low activaticnu

energy for thermal decomposition. Ligands of weak trans

Richard K. Stuart, Jr.

effect as in (l,lO-phenanthroline)tetramethylene nickel(II)
decompose with higher activation energies.

A difference in stability between alicyclic nickel
complexes and nickelacyclopentanes was also found. The
increased stability of nickelacyclopentanes over their
acyclic analogs is due to the steric constraints of the
metal containing ring.

A series of phosphine and pyridine stabilized nickel-
acyclobutanes were thermally decomposed in situ. The main
decomposition pathway was found to be reductive elimination
with a-B bond cleavage and B—hydride elimination pathways
of lessor importance.

Studies were undertaken to develop a homogeneous
catalytic analog to the methanation reaction of carbon
monoxide. The proposed pathway involved the intermediacy
of an oxygen stabilized "Fischer style" carbene complex.
The experimental results indicate that the formation of the
Fischer carbene under the reaction conditions was extremely
slow. Attempts to increase the rate of formation of the
metal carbene species by use of Lewis acid co—catalysts

met with limited success.

ACKNOWLEDGMENTS

I would like to express my appreciation for the
patience shown to me by my research preceptor, Professor
Robert H. Grubbs.

I also want to thank my fellow graduate students for
all the help, discussions and general good times which
have made graduate school enjoyable. Chuck Hoppin, Rose
Bartoszek and Bruce Osterby in particular have added im-
mensely to my stay at MSU.

Finally, I thank my parents and family for the confi-
dence, encouragement and support given to me throughout

my academic career.

11

TABLE OF CONTENTS

Chapter Page

LIST OF TABLES. . . . . . . . . . . . . . . . . . . iv

LIST OF FIGURES . . . . . . . . . . . . . . . . . . vi
I. DECOMPOSITION KINETICS OF NICKELA-
CYCLOPENTANES . . . . . . . . . . . . . . 1
Introduction. . . . . . . . . . . . . . . . . I
Results and Discussion. . . . . . . . . . . . 30

Experimental. . . . . . . . . . . . . . . . . 76

II. THE IN SITU DECOMPOSITION OF
NICKELACYCLOPENTANES. . . . . . . . . . . . . 87

IntrOductionl O O O 0 O O O O O O O O O O O 0 87

Results and Discussion. . . . . . . . . . . . 9A

Experimental. . . . . . . . . . . . . . . . . 101
III. STUDIES TOWARDS THE HOMOGENEOUS

CATALYTIC REDUCTION OF CARBON

MONOXIDE. . . . . . . . . . . . . . . . . . . 102

Introduction. . . . . . . . . . . . . . . . . 102

Results and Discussion. . . . . . . . . . . . llO

Experimental. . . . . . . . . . . . . . . . . 116

REFERENCES. . . . . . . . . . . . . . . . . . . . . 118

iii

Table

10

11

12

LIST OF TABLES

Rate data for (¢3P)2 N<::].

Activation parameters for
(¢3P)2N{::] . . . . . . . . . . . .
Kinetics and activation data for
(¢3P)2Ni(CH3)2. . . . . . . . . . .
Rate data and activation parameters
for (dpe)Ni . . . . . . . .

Rate data and activation parameters
for (dpp)Ni

Rate data and activation parameters
for (dipy)Ni. . . . . . . . . . .
Rate data and activation parameters

for (n-Bu3P)N{::] . . . . . .

Rate data and activation parameters

for (¢3P)3NO .

Activation parameters for nickela-

cyclopentanes . . . . . . . . . . .

Standard deviation of rate constants

and Arrhenius slopes. .
Decomposition products for nickela-
cyclobutanes.

Decomposition products for methyl

substituted nickelacyclobutanes .

iv

Page

32

33

MO

Al

A6

“9

58

6H

73

7A

95

96

Table

l3
1“

15

16

Page

Methanation mechanisms. . . . . . . . . . . 106
Summary of carbon monoxide

reduction reactions . . . . . . . . . . . . 111
Summary of CO reductions with

co-catalysts. . . . . . . . . . . . . . . . 112

Solvent degradations. . . . . . . . . . . . 115

Figure

10

11

12

LIST OF FIGURES

Cyclodimerization mechanism for
1,1-dimethylcyclopropane. . . . . . .
Decomposition of di-n-butylbis-
(triphenylphosphine) platinum(II) . .
B-Hydride elimination pathway . .
Steric requirements of the metalla-
cycle and alicyclic complex . . . . . .
Proposed decomposition pathway

for (Bu3P)2P£::] . . . . . . . . .
Characteristic reactions of
nickelacyclopentanes. . . . . .

Gas composition as a function of
added (n—Bu)3P for (Cy3P)N£::]

Gas composition versus phosphine/
nickel ratio. . . . . . . . . . . . .
Cyclodimerization of ethylene by
(¢3P)3N£::] . . . . . . . . . . . . .
Moles of decomposition gas as a
function of time. . . . . . . . . .
Log [(¢3P)2Ni J as a function of
time in minutes at “0°C ... . . . .
Log [(¢3P)2Ni ] as a function of
time in minutes at 30° (AD and 19°C (0)

vi

Page

1“

18

2O

22

2A

25

28

31

3M

35

Figure Page

13 Arrhenius plot for (¢3P)2N£::]. . . . . . . 36
1“ Log [(¢3P)2Ni(CH3)2] as a function of
time in minutes at 6° GA) and 20°C (0). . . 37
15 Log [(¢3P)2Ni(CH3)2] as a function
of time in minutes at 30°C. . . . . . . . . 38
16 Arrhenius plot for (¢3P)2Ni(CH3)2 . . . . . 39
¢2
P\\\
17 Log El: ’//Ni J as a function of
P
¢2
time in minutes at 80°C . . . . . . . . . . M2
4’2
P\\\
18 Log [ //,Ni ] as a function of
P
¢2
time in minutes at 90° (0) and 100° 0A) . . U3
°2
P\\\
19 Arrhenius plot for [ Ni . . . . . . HA
P/
4’2
20 Reorganization of the metallacycles
during thermal decomposition. . . . . . . . H5
¢2
P\
21 Log [ /,Ni ] as a function of time
P
4’2

in minutes at 90° (0), 100° (D) and

110°C(A).................147

vii

Figure

22

23

2A

25
26

27

28

29

30

31

32
33

Page

82
P\
Arrhenius plot for<: N{::]. . . . . . . . MB
Px/
°2

Log [(dipy)Ni J as a function of
time in minutes at 80° (0) and 60°C
(A)....... 51
Log [(dipy)N€::]J as a function of

time in minutes at 65° (A) and 70°C

(a) . . . . . . . . . . . . . . . . . . . . 52
Arrhenius plot for (dipy)N{::]. . . . . . . 53
Log [(1,10—phen)Ni J as a function

of time in minutes at 75° (0), 70° (0)

and 80°C Q§)r . . . . . . . . . . . . . . . 5U
Arrhenius plot for (1,10-phen)N£::] . . . . 55
Relative d orbital energy levels

for nicke1(II). . . . . . . . . . . . . . . 56
Log [(nBu3P)2Ni J as a function of.

time in minutes at 20°C . . . . . . . . . . 59
Log [(n-Bu3P)2N{::]J as a function of

time in minutes at 30°C . . . . . . . . . . 60
Log [(n-Bu3P)2Ni J as a function of

time in minutes at H0°C . . . . . . . . . . 61
Arrhenius plot for (n-Bu3P)N£::]. . . . . . 62

A possible decomposition mode for

(R3P)N{::]. . . . . . . . . . . . . . . . . 6D

viii

Figure Page

3“ Log [(¢3P)3Ni J as a function of

time in minutes at 0°C. . . . . . . . . . . 66
35 Log [(¢3P)3Ni J as a function of

time in minutes at 20°C . . . . . . . . . . 67
36 Log [(¢3P)3Ni J as a function of

time in minutes at 30°C . . . . . . . . . . 68

37 Arrhenius plot for (¢3P)3N£::]. . . . . . . 69
38 Proposed decomposition pathway of

(¢3P)3NO................ 70

39 Newman projection of the tetrahedral

complex . . . . . . . . . . . . . . . . . . 71

A0 Current mechanism for olefin
metathesis. . . . . . . . . . . . . . . . . 88
A1 Proposed mechanism for Natta-Ziegler
polymerization. . . . . . . . . . . . . . . 89

U2 Possible decomposition pathways for
L2X2PO................. 91
143 Products of (Cp)2w;> by photochemical
decomposition . . . . . . . . . . . . . . . 92
HA Proposed pathway for coupling of
a-w dihaloalkanes . . . . . . . . . . . . . 93
H5 Newman projection of nickela-
cyclobutane . . . . . . . . . . . . . . . . 98
H6 Decomposition modes for L2N(:> . . . . . . 99

ix

Figure

A7

1&8

Page

Recent proposed Fischer—Tropsch
mechanism . . . . . . . . . . . . . . . . . 107
Proposed homogeneous pathway for

CO reduction. . . . . . . . . . . . . . . . 109

CHAPTER I

THE DECOMPOSITION KINETICS OF NICKELCYCLOPENTANES

INTRODUCTION

Heterocyclic compounds are an important class of
compounds and have played a major role in the development
of organic chemistry. When the heteroatom is a transi—
tion metal the heterocycle is referred to as a metalla-
cycle. It is becoming apparent that metallacycles play
a major role in the development of organometallic chemistry.
Many transition metal catalyzed reactions of olefins and
acetylene are believed to utilize metallacycles as inter-
mediates. Metallacyclopentanes have been demonstrated
to be intermediates in many metal catalyzed cycloaddition
and cycloreversions reactions of olefins.1

In this introductory section, the reader will become
familiar with the role and extent of metallacycles in
transition metal chemistry. The preparation of metalla~
cycles as well as their characteristic modes of reaction
including their decompositions modes will be contrasted
and compared to alicyclic transition metal complexes.

A discussion of the effect of additives on the decomposi-
tion of metallacycles will also be presented.

2

In 1961 F. G. A. Stone reported the preparation of

tetrakis(difluoromethylene)Iron(II) tetracarbonyl by the

reaction of iron pentacarbonyl and tetrafluoroethylene.
This complex was characterized as a white, volatile, air
stable solid with a melting point of 76-77°. This is the
first reported preparation of a metallacyclopentane.

A decade later, Osborn3 isolated an iridacycle 2
from the 2+2 cycloaddition reaction of norbornadiene in
acetone in the presence of [Ir(l,5 cyclooctadiene)ClJ2.

The identity of the iridacycle g was confirmed by an X-ray

F F
F
r F

determination. When 2 was treated with an excess of tri—

phenyl phosphine the exo-trans-exo dimer 3 of norbornadiene

“ Excess “‘
CH 33cocn 6“; VT I I 3?
\MM

.2.

 

was formed in 35 percent yield. This was the first re-
ported case where a metallacycle was isolated from a
reaction mixture and then shown to be a true reaction

intermediate by its further reaction to give the expected

products.

Katzu and Cerefice treated exo-tricyclo[3.2.1.O2’u1-
octene Q at 90° for 2 hr with tris(triphenylphosphine)-
rhodium(I)chloride and obtained a quantitative mixture of
three isomers. Compound 6 is formally related to Q by

an electrocyclic rearrangement while compound 5 is not.

“vb—>28 A A»

4
~ 2 £5, Z,

Labeling studies showed that a specific deuterium shift
was required to convert Q to 5. To account for these
intramolecular processes a pathway involving metallacycles

8 and 9 as intermediates was proposed.

Vb —> ‘Rbéfl§x$ fix
8 v

N

éa—V‘EEP 863*?
M m: x
9

N

The highly strained cubane molecule is remarkably
stable thermally. This presumably reflects the con-
straints of the Woodward-Hoffman orbital symmetry con—
servation rules. The conversion of cubane to syn-tri—

cyclooctadienes is thermally forbidden.

 

 

 

a a n
4' “1!
—-9
’/’ //. “4".
R R
IO LI, 1?;

However, in the presence of catalytic amounts of [Rh(00)2ClJ2
the conversion of 10 to 11 and 12 proceeds quantitatively.5
When a stoichiometric amount of catalyst was added, the
rhodacycle 13 was isolated in 90 percent yield. In the
presence of excess triphenyl phosphine the rhodacycle

undergoes reductive elimination to form the corresponding

4'
[Rn(c0)2c1],9 Ph3P

 

 

 

2:

RB °
l3

ketone and Rh(CO)Cl(®3P)2. The trapped rhodacycle prompted
Halpern and Eaton to propose a non-concerted series of
oxidative addition steps to account for the observed re-
sults. That is, they proposed a mechanism involving metalla-
cycles for the conversion of 10 to 11 and 12.

Binger 33 al.6 have isolated nickelacyclopentane
derivatives from the [2n+2nJ cycloaddition reaction of
strained ring olefins and nickel(o) catalysts. From the
reaction of 3,3-dimethylcyclopropene and bis(l,500D)Ni(o)
in the presence of dipyridyl, Binger was able to isolate
in addition to the cyclodimerization product a metall-

acycle 1%. Treatment of the nickelacyclopentane 1% with

A Ni(COD)Ib1b_YL; (biby)Ni law
’13 '3

activated olefins (such as maleic anhydride) results

 

in the formation of the cyclodimer 15. The metallacycle
IQ is a dark green, diamagnetic, air sensitive, crystalline
complex thermally stable to about 80°C. The proposed
mechanism for the cyclodimerization of these reactive
olefins is shown in Figure 1. The key step in the forma-

tion of metallacycles from strained olefins is the

\ ./N A N\ , N
g/NI\N) ——’ CAI/NK/ —9 ”>N, VM
4

I6
“’ 17
m

Figure l. Cyclodimerization mechanism for 1,1-dimethyl—
cyclopropane.

isomerization of the bis olefin complex 16 to the metalla-
cycle 17. Attempts to isolate the unsubstituted nickel—
acyclopentane complex from the analogous reaction with
ethylene failed. This may be in part due to the lack of
strain in ethylene and hence its lower reactivity. How-
ever, Binger was successful in preparing the parent nickel-
acyclopentane from (bipy)Ni(COD) and 1,h-dibromobutane.
Incidentally this is one of the best methods of preparing
amine stabilized nickelacyclopentanes.

Traditionally the metal-carbon sigma bond is thought

00
Ni(COD)2 + dipyridyl Br Br -+ (dipyridyl)N£:::]

to be very weak. In the last fifteen years much effort
has been devoted to preparing complexes which contain
strong field ligands in addition to sigma bonded alkyl
groups. Strong field ligands such as carbon monoxide,

tertiary phosphines, and the n-cyclopentadienyl group

are believed to stabilize metal-alkyl 0 bonds by their
ability to n-bond. A widely accepted n-bonding theory?-9
has been proposed to account for the stabilizing ability
of these n accepting ligands. Homolytic cleavage of the
metal alkyl bond occurs by excitation of an electron from
a bonding orbital to a nonbonding or antibonding orbital.
This weakens or breaks the bond to form radicals. This
process should be hindered by strong field ligands which
increase the ligand field splitting. This theory was
initially accepted because homolytic fission of metal-
alkyl bonds was believed to be the main mechanism operat-
ing for the decompositions of metal alkyl bonds.

As more evidence became available the "supporting -
ligand" theory became less attractive. Many organometal-
lic complexes are now known which contain ligands with

+2 10
5]

no n-bonding ability such as [PhCH2Cr(OH2) and

a number with no supporting ligands such as (PhCH2)uTill

12
and (Ph3C)2Ni.

Spectroscopic and bond length data
suggest that transition metal-carbon bonds are of the
same strength as main group metal-carbon bonds which are
not greatly affected by the presence of w accepting ligands.
Finally the promotional mechanism as originally proposed
is unlikely to be applicable under thermal conditions.13
It was also proposed that n accepting ligands were
necessary simply to block the coordination sites necessary

for low energy pathways of bond cleavage such as B—hydride

1” A series of metal-alkyl complexes

elimination to occur.
are known which are stable to B-hydride elimination and
do not have additional ligands to block coordination
sites.15
Many transition metal organometallic complexes are
thermally labile. This lability stems from the character-
istic transition metal properties of variable co-ordination
number and oxidation state. These two properties allow
a number of low energy decomposition pathways to be avail-
able which are denied to main group elements. The most
important of the low energy pathways will now be discussed.
The thermal decomposition chemistry of alicyclic com-
plexes is dominated by B-hydride elimination. Whitesides16
has found that the thermal decomposition of n-butyl(tri-
n-butylphosphine)copper(l) proceeds by B-hydride elimina-
tion forming l-butene and an intermediate copper hydride

which reduces another copper carbon bond to form butane,

tri-n-butylphosphine and copper metal.

LC j\/——> LCU/H Lax/fly V\

. N

In another study Whitesides17 investigated the thermal

chemistry of Di-n-butyl bis(triphenylphosphineplatinum(II).

The products on thermolysis were l—butene, butane and a
platinum(0) species. Figure 2 outlines the proposed
mechanism for the decomposition of di(n)butyl bis(tri-

phenylphosphine)platinum(II) by B-hydride elimination.

/"‘//~\\ Lg-JD ‘4/P~\//
\’,,\\’/' '_____7 '+\~/’\\~/’

l
L- P+-(\/\ <— HL: m/Q

‘\//”\\

L.
L)»

Figure 2. Decomposition of di—n-butylbis(triphenylphos-
phine)platinum(II).

The first step is the loss of a phosphine which leaves

a vacant coordination site which a hydride from a 8

carbon fills in the second step. B-Hydride elimination
yields a molecule of l-butene a bonded to platinum. The
next step is a reductive elimination of a hydride and
butyl moiety to form butane. Trapping studies in this
system failed to trap any products arising from the forma-
tion of butyl radical. Hence Whitesides concluded that

free radicals were not involved in their decomposition.

10

It is also interesting to note that no (less than 1%)
octane, the reductive elimination product, was formed.
Another possible thermal decomposition mode is reduc-
tive elimination. Semmelheck18 found that bis(l,5 cyclo—
octadiene) nickel (0) would couple aryl halides or alkenyl

halides. The following mechanism was proposed.

X A '
\\.’ V th2
N1 -€> '

” Ar Ar-Ar

L Ar
\Ni/ ArX :

. ArX
N1(COD) -————-9
2 X’/’ \\‘L.

/\

X

The key step in biaryl formation is the decomposition of
the diarylnickeldihalide by reductive elimination. This
is proposed to occur by a concerted cis elimination.

G. B. Younglg investigated the kinetics and mechanism
of the reductive elimination of diaryls from diaryl bis-
(phosphine)platinum(II) complexes. Reductive elimination
is the preferred mode of thermal decomposition in these
complexes which occurs by a concerted first order uni-
molecular reaction. Young envisioned the transition state
for this reaction to be a species where the two aromatic

rings start bond making at the same time the platinum

11

aryl bonds are breaking.

In addition to B-hydride elimination and reductive
elimination metallacycles can decompose by B-y carbon-
carbon bond cleavage. That is, a metallacyclopentane de-
composes to a bis(olefin) complex which then loses one or

both of the olefins.

L2" ——-> LM9/\——-> LM =

l,A tetramethylene bis(cyclopentadienyl)titanium(IV)

20

was thermally decomposed by Whitesides. It produced 92

percent ethylene and 8 percent l-butene.

cp\\Ti _———4> ///’ \\\//A§§~ [CPZTJJn
Cp/ A /

No reductive elimination (formation of cyclobutane) was
reported while B-hydride elimination (formation of l-
butene) occurred to only eight percent.

The change of a metallacycle to a bis olefin complex
appears to be a reversible process. That is, when bis-
(cyclopentadienyl) titanium dichloride is reduced to

titanocene in the presence of ethylene, a titanacyclopentane

12

is formed. The titanacycle was trapped with carbon monoxide

to give cyclopentanone in 20 percent yield.
\ 5’
Cp Ti’ , CP\ , CO
2 >5 ¢—- /T‘I —-—'—>
09

Another possible decomposition pathway available to
transition metal carbon o-bonded complexes is a-hydride
elimination. This involves migration of a substituent
from the a-carbon to the metal with subsequent formation
of a metal carbenoid species. The metal carbon double

bond then isomerizes via a metal hydride to the primary

LMHi:1—> LMO ——» LMMH

H l

4¢’\\/’ é—-' L_fi4’fl\\/’A\V/’IJ
F1

alkene and a reduce metal species. a-Hydride elimination

is generally less well documented for transition metal

carbon o-bonds than B-hydride elimination is. Otsuka21

found from the decomposition of NiBr(®3P)2(CH2002CH2CH3)

13

prepared in situ from ethyl a-bromoacetate and tetrakis-
(triphenylphosphine)nickel(0) that both diethyl fumarate

and diethyl succinate were formed. The main product,

 

A P
3 CH co CH cu co CH CH
Brcnzcnzcuzcn3 ‘\\ Ni,/’ 2 2 2 3 FH=CH/ 2 2 3

N1(Q P)
3 4 7
@3p"’ ‘\.Br COZCHZCH3

diethyl succinate, arises from homolytic cleavage which
produces the alkyl coupling product and a reduced nickel
species. More interestingly the diethyl fumarate arises
from a-hydride elimination with subsequent coupling as in
Figure 3.

For completeness one final thermal decomposition
pathway is presented. Homolytic cleavage of the metal
carbon o-bond produces free radicals which may then under-
go any reaction characteristic of free radicals (coupling,
rearrangement, radical abstraction, etc.). Homolysis of

transition metal carbon 0 bonds is uncommon because it

in

H
, C0 Et
N‘i _, N‘i ‘9 ‘9 l
/’ \\. \\ Br-Ni-CHCO Et CO Et
3 r H

Figure 3. a-Hydride elimination pathway.

usually is a more energetic process (higher activation
energy) than alternate decomposition pathways. Mercury(II)
complexes22 decompose by this pathway. Divalent mercury
complexes may decompose by this process in a concerted

or stepwise manner.

Slow

HgR2 -> RHg R
Fast

RHg -+ Hg° + R

or HgR2 -+ Hg° + 2R~

15

The thermal decomposition of Pt(IV) metallacyclo-
pentanes was investigated by Puddephatt.23 The decomposi-
tion of bis(dimethylphenylphosphine)diiodotetramethylene
p1atinum(IV) in the solid state gives exclusively 1-butene
generated by B-hydride abstraction. The probable pathway

involves the initial disassociation of a phosphine ligand

(CH3)20P 1 “”92”“
Pt -> At?" W
(CH3)2¢P/1 H I

followed by B-hydride elimination with subsequent reduc-
tive elimination of the hydride and hydrocarbon moiety.
The decomposition of a,a'dipyridyl(iodo)methyltetramethylene

p1atinum(IV) leads to the formation of butenes and methane.

CH N
N‘A: —-> N CIH3 C \M;
(IV/’i:::] 97>?3\ ‘-—_9 hr/

16

After initial B-hydride elimination the intermediate 11 can
reductively eliminate by cleavage of the Pt-H and Pt-CH3
bonds or by cleavage of the Pt-H and Pt-CH2 bonds. The
cis reductive elimination of the hydride and methylene
moiety predominates. l-butene is formed in forty percent
and methane is formed in fourteen percent yield. 2-butene
is formed in thirty percent yield by isomerizing l-butene.
The thermal decomposition of bis(methyldiphenylphosphine)
methyl(iodo)tetramethylene p1atinum(IV) givesaivariety of
products with l-pentene as the major product. In this case
initial reductive elimination of the methyl and a methylene
from the ring occurs to give the n-pentyl complex 18 which

B-hydride eliminates to give 1-pentene. The two pathways

CH -
Lg 3 L\
L/?3 fl L/P+:l\/\’CH3

2.8 \ L3
U7

\\v//\\‘¢9 ‘F—_— *7

operating in the preceding example also operate here.
It is interesting to note that no cyclobutane is formed
in any of the p1atinum(IV) decompositions. Puddephatt

believes the lack of cyclobutane production from a reductive

l7

elimination process is due to the higher activation energy
needed to close the strained ring cyclobutane.
Whitesideszu prepared and decomposed a series of plat-
inum(II) metallacyclopentanes and compared the rates of
thermal decomposition to the appropriate alicyclic
p1atinum(II) complexes. The decompositions were done in
methylene chloride solution and were first order to ap—
proximately sixty percent decomposition. It was found
that the rate of decomposition of bis(triphenylphosphine)

A times slower than the

tetramethylene p1atinum(II) was 10
rate of thermal decomposition of bis(triphenylphosphine)-
di-n-butyl p1atinum(II). Both complexes decompose by
B-hydride elimination. Whitesides suggests the increased
stability of the platinacyclopentane toward B-hydride
elimination stems from the unfavorable dihedral angle
between the metal and B-hydride atom. That is, the optimum
dihedral angle for B-elimination is zero degrees which an
alicyclic compound can obtain. However, the steric con-
straints of the metallacycle prevent the dihedral angle
from obtaining this optimum value as Figure A illustrates.
Again, it is interesting that cyclobutane which would come
from reductive elimination was not produced.

In a more detailed study Whitesides25 found that bis-
(triphenylphosphine) tetramethylene p1atinum(II) does not
lose a ligand previous to decomposition as both the

p1atinum(IV) and p1atinum(II) dialkyl complexes do. In

18

it
:1:
11
I

Figure A. Steric requirements of the metallacycle and
alicyclic complex.

fact it was found that added triphenyl phosphine increases
the rate of decomposition of platinacyclopentanes while
the rate of decomposition of acyclic p1atinum(II) complexes
is decreased by added phosphine.
Bis(tri-n-butylphosphine)tetramethylene p1atinum(II)
19 decomposes at 120°C in methylene chloride to yield
products very different from the triphenylphosphine p1at-
inum metallacycle. The reductive elimination product,
cyclobutane, was the major product with lessor amounts
of butenes and a trace of butadiene. The corresponding
bis(tri-n-butylphosphine)diethyl p1atinum(II) thermally

decomposed by B-hydride elimination to give ethylene and

&JP

3).». ... N/‘x j

 

 

 

 

19

ethane, hence the supporting ligand change is not respon-
sible for the change in products. It was found that the
decomposition of 19 in hydrocarbon or ether solvents gave
l-butene as the major product in an analogous manner with
the other p1atinum(II) metallacycles. This reactivity
pattern was explained26 by oxidative addition of a solvent
molecule to form a p1atinum(IV) species. Figure 5 shows
the proposed pathway to explain the product distribution.
Two generalities concerning the decomposition of platin-
acyclopentanes are obvious from this work. First, reductive
elimination of two alkyl groups occurs readily from plat—
inum(IV) but not from p1atinum(II).27 The decomposition
of p1atinum(II) metallacycles is dominated by B-hydride
elimination in the absence of other oxidative addition
processes. Second, the distribution of linear and cyclic
decomposition products from an alkyl tetramethylene platinum
(IV) complex is very sensitive to the structure of the
starting complex. Several factors may be reflected includ-
ing the strain involved in cyclobutane formation and the
increased stability of platinum-carbon 0 bonds with electro-
negatively substituted carbon.28 The production of cyclo-
alkane is favored when R is electron withdrawing. The
production of linear products is favored when R is elec-
tron donating.

Grubbs §£.§l.29 prepared and isolated bis(tertiary-

phosphine) nickelacyclopentanes. The appropriate

20

 

 

 

 

cup
PDQ-W» LAC! ~+ W
.1 i

W 0

Figure 5. Proposed decomposition pathway for (Bu3P)2P<::]

21

 

 

 

 

bis(phosphine) nickel dichloride and l,A-dilithiobutane
were mixed at -78°C in ether. As the solution was allowed
to warm to -20°C yellow crystals of the metallacycle were
formed. The nickelacyclopentanes were characterized as

shown in Figure 6.

P3P

:Ni LiflLi ——> ,NQ
P3P \Cl P3P

In an initial study on the decomposition of nickela-
cyclopentanes Miyashita and Grubbs30 found that the co-
ordination number of the nickel has a great effect on the
decomposition pathway. Three coordinated nickelacycles
decompose by B-hydride elimination to give l-butene. Four
coordinated nickelacycles decompose by reductive elimina-
tion to give cyclobutane while five coordinated nickela-

cyclopentanes decompose by B-y bond cleavage to give

22

By

 

 

 

 

Figure 6. Characteristic reactions of nickelacyclo-
pentanes.

23

LNO = LaNO -.—=-* L3NO
l l i

2=

 

 

 

\/\

 

two molecules of ethylene. Figure 9 summarizes the results
of a study on the effect of added phosphine on the decom-
position of tricyclohexylphosphine tetramethylene nickel(II).
Figure 7 is a graph of decomposition gas composition as

a function of the amount of added tri-n—butylphosphine.

With no added tri-n-butylphosphine the decomposition gas

is mostly 1-butene which arises from the three coordinated
nickelacycle decomposition with traces of cyclobutane and
ethylene. As the amount of added phosphine increases so
does the amount of cyclobutane. At a phosphine to nickel
ratio of five the cyclobutane in the decomposition gas is

at a maximum which implies that the main species in solution
is the four coordinated nickelacycle. At higher phosphine
to nickel ratios the decomposition gas is mainly ethylene
which originates from the decomposition of the five co-
ordinated nickelacyclopentane. This type of behavior is
characteristic of bis(monodentate tertiaryphosphine)tetra-
methylene nickel(II) complexes, but the behavior of chelated
phosphine tetramethylene nickel(II) complexes is very

Gas Composition (molar %)

100 -

50

 

2A

//

\ \ C7

\*“ W

I v

[(n-Bu)3PJ added

Figure 7. Gas composition as a function of added (n-Bu)3P

for (Cy3P)N{::]

25

different. The corresponding graph of decomposition gas
composition as a function of added phosphine shows a dif-
ferent trend as seen in Figure 8 for bis(diphenylphosphino)-

ethane tetramethylene nickel(II). The four coordinated

 

 

 

 

Figure 8. Gas composition versus phosphine/nickel ratio.

chelated nickelacyclopentane decomposes to cyclobutane.

The addition of free phosphine causes the amount of cyclo-
butane in the decomposition gas to decrease while the amount
of butene and butane increases. In the case of the mono-
dentate phosphine, ethylene production increased at higher
phosphine to nickel ratios rather than butene production

as in this case. This apparent incongruent phosphine

26

effect is explained by the inability of the chelated phos-
phine metallacycle to form a five coordinated complex.
The added phosphine does not complex to the nickel in this
case30 but rather acts as a base to promote B-hydride

elimination type products25 (butenes).

«>2 p [“32
[ZEfiZjAPPB—e [ENE—7C EN.—

2 2 prs (Dz

l. l_~.
‘N; :2‘ L NO : NO
If :3 3 L/

In a complimentary 31? NMR study30 on bis(triphenyl-
phosphine)tetramethylene nickel(II) it was found that free
phosphine exchanges with the five coordinated nickela-

cyclopentane but it does not involve the square planar

27

four coordinated nickelacyclopentane. Hence, Grubbs
proposed that the bis(phosphine) metallacycle in equilibrium
with the tris(triphenylphosphine)metallacycle was the tetra-
hedral species and not the isolable square planar bis(phos-

phine) nickelacyclopentane.

D

°° ”/5 —~ NC:

LNi : LNi ‘— 3 0°
3 D 203/

. ,, \ \.

D

D
D
W: <-— LsNCA’

In an elegant deuterium labeling study Miyashita and
Grubbs31 presented evidence for an equilibrium between the
tris(triphenylphosphine)tetramethylene nickel(II) and the
corresponding bis(ethylene) complex. By labeling the 2 and
5 positions of the ring they were able to obtain tetra—
deuterated butanes (after quenching) where the labels were
in the 2 and 3 positions rather than the l and A posi-
tions. Hence this is direct evidence that the tris(tri-
phenylphosphine)tetramethylene nickel(II) is in equilibrium
with the bis(ethylene)bis(triphenylphosphine)nickel(o)
complex.

Miyashita and Grubbs32 found that bis(triphenylphosphine)-

28

PBN; —> PzN‘ ——> PND

 

 

 

 

Figure 9. Cyclodimerization of ethylene by (¢3P)3N£::]

tetramethylene nickel(II) in the presence of triphenyl-
phosphine will catalyticly cyclodimerize ethylene to
cyclobutane. Figure 9 summarizes the proposed reaction
sequence. This catalyst system is very different from
traditional nickel oligiomerization catalysts in that no
alkyl aluminum co-catalyst is required.
Bis(phosphine)tetramethylene nickel(II) complexes
are a good system for the study of thermal decomposition
modes other than B-hydride elimination. The decomposition
of nickelacyclopentanes can be controlled to give pre-
dominately reductive elimination, B—hydride elimination or
B-Y bond cleavage products by changing the coordination
number of the nickel. The next section describes the
quantification of these decomposition processes. The
stability of nickelacyclopentanes makes manipulations pos-

sible without necessitating extreme temperatures for

29

decomposition. The comparisons and contrasts between the
platinum and nickelacyclopentanes will be discussed along
with experimental findings. The relative stability dif-

ferences between nickelacyclopentane and nickel alkyls

will also be presented.

RESULTS AND DISCUSSIONS

The decomposition kinetics of nickelacyclopentanes
were determined by monitoring the appearance of products
compared to an internal standard as a function of time.33
A graph of moles of gas produced versus time yields a
curve which increases to a certain value and then levels
off. As Figure 10 shows the leveling off corresponds to
complete decomposition of the metallacycle. By subtract-
ing the moles of gas produced at a particular time from
the total moles of gas produced after complete decomposi-
tion the amount of complex remaining at any particular
time can be obtained. It is this value (Mr) which was
used to obtain the rate constants. That is, the graph of
the logaritthM'the moles of complex remaining versus time
gave a linear relationship which is expected for a first
order unimolecular reaction. The relationship between
moles of complex remaining and time was linear to at least
fifty percent reaction with 80-90 percent reaction being
typical.

The activation parameters were calculated from Eyring's
transition state theory.3u The entropy of activation

AS? was calculated from

AS7 = A.576 log (K/T) + (E/T) — A9.21

30

31

 

 

 

 

Figure 10. Moles of decomposition gas as a function of
time.

where K is the rate constant at temperature T and E is
the activation energy in calories/mole. The value of T
used was a midpoint value of the temperature range studied.

a

The enthalpy of activation AH was calculated from

AH? = E7 - RT

1

and the free energy of activation AG was calculated from

A07 = ARI - TA87

32

The slopes of the rate constant plots (logarithm of
the concentration of complex remaining versus time) were
obtained from the best least square fitted line. The slope
of the Arrhenius graphs (logarithm of the rate constants
versus the inverse of the temperature) were also least
squares fitted by the use of a Texas Instrument Ti-55
calculator.

The nickelacyclopentane decomposition kinetics des-
cribed here are most conveniently divided into four
classes: three coordinated, four coordinated, five co-

ordinated, and chelated metallacycles.

Four Coordinated Nickelacyclgpentanes

Four coordinated nickelacyclopentanes decompose by
reductive elimination to give cyclobutane. Bis(triphenyl-
phosphine)tetramethy1ene nickel(II) was decomposed at a

temperature range of 19 to A0°C. Table 1 summarizes the

Table 1. Rate data for (¢3P)2 Ni::]

 

 

 

Temperature Rate Constants Half Life
19°C 3.2 x 10'5 sec'1 361 mins
30°C 1.2 x 10'" sec"1 77 mins

A -l

A0°C A.6 x 10- sec 22 mins

 

 

33

rate constants, half lives, and temperatures. Figures 11
and 12 show the first order rate plots of the logarithm

of the concentration of undecomposed nickelacyclopentane
as a function of time. Figure 13 is the Arrhenius plot.
That is, a graph of the logarithm of the rate constants
against the inverse temperature. The Arrhenius plot slope
is 1.15 x 10“. Multiplication of this by the gas constant

gives the activation energy. The activation energy was

found to be 23 Kcals/mol. Table 2 summarizes the activation

Table 2. Activation parameters for (¢3P)2N<:j

 

 

 

AE7 AS7 ART AG7
(¢3P)2Ni::J 23 Kcals/ -2.3 eu 22 Kcals/ 23 Kcals/
mol mol mol

 

 

parameters for bis(triphenylphosphine)tetramethylene
nickel(II).

It is interesting to compare these values with the cor-
responding data for bis(triphenylphosphine)dimethyl nickel(II).
This complex also thermally decomposes by reductive elimina-
tion to give ethane. Table 3 summarizes the kinetics and
activation parameters for bis(triphenylphosphine)dimethyl
nickel(II). Figure 1A and 15 are the rate constant plots
while Figure 16 is the Arrhenius plot. Two significant

3A

owoomo
1 vv‘

Guam-
1?

 

Figure 11. Log [(¢3P)2Nia J as a function of time in
minutes at A0°C.

D

 

a- ,g 1 e 1
101 (00 30:5 43'.) 50.) 603

Figure 12. Log [(¢3P)2Ni:j J as a function of time in
minutes at 30° (A) and 19°C (Cl).

Figure 13.

36

\J a) to —"
v Y '

/

 

 

5
a.
3.
2r
.no ->‘ 19°
4 fi 4
3 .3 3.4
116’ /.

Arrhenius plot for (¢3P)2 N{:] .

 

37

 

a
7
6
5
a
3
2»
1,
9
3.
7
6
5»
U
4
D
3
D
2,
TC 20 45 so 83 10a 120
Figure 1A.

Log [(¢3P)2N1(CH3)2J as a function of time in
minutes at 6° (A) and 20°C ([1).

38

h U‘C‘VWO‘

amawmo—l
,vvv

 

Figure 15- Log [(¢3P)2Ni(CH3)2J as a function of time in
minutes at 30°C.

 

39

 

Figure 16.

no' n

Arrhenius plot for (¢3P)2Ni(CH3)2.

A0

Table 3. Kinetics and activation data for (¢3P)2Ni(CH3)2

 

 

 

Temp. Rate Constant Half Life E7 A87 AH7 AGI
30° 1.0::10"3 11.5 min 19 -12.7 cu 18. 22.
20° 3.7x10’” 31 min

6° 7.0x10‘5 165 min KCal/m°1

 

 

points need to be mentioned. First, the rate of thermal
decomposition of bis(triphenylphosphine) dimethyl nickel(II)
is twelve times faster than the rate for the metallacycle.
This is similar to the trend Whitesides26 found for B-
hydride elimination for bis(triphenylphosphine(di-n-butyl)
p1atinum(II) and the corresponding metallacycle. Second,
the thermal decomposition activation energy for the dimethyl
nickel complex is A Kcals/mol lower than for the metalla—
cycle. Both of these differences can be accounted for by
considering the nature of the organic moiety formed during
the decomposition reaction. The formation of the highly
strained cyclobutane molecule is a more energetic process

than the formation of the stable ethane molecule.

Chelated Nickelacyclopentanes

Ethylenefbis(diphenylphosphine)tetramethylene nickel(II)

is a four coordinated complex which decomposes by reductive

Al

elimination to form cyclobutane. The diphos metallacycle
is considerably more stable both to heat and oxidation than
the triphenylphosphine metallacycle. This light yellow
metallacycle was decomposed at temperatures ranging from
80° to 100°C. The half lives, rate constants, and activa-
tion parameters are shown in Table A. Figure 17 and 18

are the graphs of the logarithm of the remaining complex

versus time while Figure 19 is the Arrhenius plot.

92

1?
Table A. Rate data and activation parameters for[: JV{::]

(Dz

 

 

Temp. Rate Constant Half Life E7 A87 AH7 AGI

 

80 1.5x10'5 sec-1 770 min 37 21 36 28
90 7.0xlo'r5 165 in Kcals/mol
100 2.ux10'“ us

 

 

The activation energy for thermal decomposition by
reductive elimination for ethylene bis(diphenylphosphine)-
tetramethylene nickel(II) is 37 Kcal/mol. This is sig-
nificantly higher than 23 Kcal/mol found for bis(triphenyl-
phosphine)tetramethylene nickel(II). The increase of 1A
Kcal/mol can be accounted for by considering the reorgani-

zation of the nickel phosphine complex during the reduction.

A2

mumoo

 

 

 

1'00 2'00 1700 4‘00 500 600
¢ 2

P
Figure 17. Log [[ :>N£::]J as a function of time in
P

¢2
minutes at 80°C.

..a
O

GVQQD

 

 

A3

 

Figure 18.

1m 27):; 100 4710 €00 609
¢2
P\
Log [ NiOJ as a function of time in
p/
¢2

minutes at 90° (1]) and 100° (A).

AA

2:9 -‘
Y 7

ON

 

Figure 19. Arrhenius plot for [

90° 30°

# . A
2.75 2.80 2.85
110' n

P
P>NO

A5

During the decomposition the nickel is reduced from four
coordinated nickel(II) to a two coordinated nickel(o).
V.S.E.P.R. theory35 predicts the two coordinated nickel(o)
to be linear.36 Hence as the reductive elimination occurs
the phosphine ligands swing open from their square planar
arrangement to a linear one. This opening up of the bond
angle is quite easy for the triphenyl phosphine case but

is considerably harder for the diphos metallacycle. Figure
20 illustrates the reorganization during the thermal de—

composition for both cases.

 

 

 

 

 

— J
P o
{27¢‘<::]-——5> ‘thu:1_—3 '—€9"3"DA-[3
/ p ~.___

 

 

 

 

 

 

 

. P 0
[HO —> [g:N.:;__ —> [QM
L -1

 

 

Figure 20. Reorganization of the metallacycles during
thermal decomposition.

A6

By changing the size of the hydrocarbon bridge between
the two phosphines of the chelate a variation in activa-
tion energy with ring size is expected. Unfortunately,
the bis(diphenylphosphine)methane metallacycle could not
be prepared. The bis(diphenylphosphine)methane reacts
with nickel dichloride as a monodentate ligand rather than
as a bidentate ligand,36 hence only dimeric bis(diphenyl-
phosphine) methane nickel dichloride species could be
prepared. However, the larger chelate metallacycle,
propylene bis(diphenylphosphine)tetramethylene nickel(II)
was prepared. It also has a similar reactivity as the
diphos nickelacyclopentane. It was decomposed at 90° to

110°C. Table 5 presents the kinetics and activation data.

(be

P
Table 5. Rate data and activation parameters for <<:FJN:::]

 

 

Temp. Rate Constant Half Life E7 A87 AH7 A07

 

90 2.3xio’5 502 35 15 35 29
100 8.2x10'5 1A1 in Kcal/mol
110 2.8x10-l4 Al

 

 

Figure 21 is the rate constant plot for propylene bis(di-
phenylphosphine)tetramethylene nickel(II) and Figure 22 is

A7

 

 

50 100 150 2'00 250 3.00 350 4.00 4.50 500
¢2
P\
Figure 21. Log [< /Ni:jJ as a function of time in min-
P

¢’2
utes at 90° (0), 100° (1]) and 110°C (A).

1453

mONQOO
V U 1"

GNOD"
. r

 

110° 100° 90°
F. 6 7.70 2100

:10’3/1

°2

P\ D
Figure 22. Arrhenius plot for <: [Ni ‘ .

I”

4’2

A9

the Arrhenius plot for this system.

The activation energy for propylene bis(diphenylphos-
phine)tetramethylene nickel(II) is lower than for ethylene
bis(diphenylphosphine)tetramethylene nickel(II). They
have activation energies of 35 and 37 Kcal/mol, respec-
tively. The difference in ring flexibility between the
five and six membered ring of the chelated ligand is sig-
nificant. That is, the more flexible propylene bis(di—
phenylphosphine)tetramethylene nickel(II) has a lower
activation energy than the smaller, more rigid chelate
ring compound.

The 0,0' dipyridyl tetramethylene nickel(II) also
thermally decomposes by reductive elimination to yield
cyclobutane. This complex is a dark green extremely air
sensitive crystalline solid. The decomposition tempera-

ture studied ranged from 60 to 80°C. As seen in Table 6

0
Table 6. Rate data and activation parameters for NHN:::]
o

 

 

 

Temp. Rate Constants Half Lives E7 A87 AH7 AG?
60 6.9x10’7 sec”1 279 A2 38 A2 29
65 1.6x10’6 sec'1 120 in Kcals/mol

6

“3
80 2.6x10'5 7.5 hrs

70 A.5x10-

 

 

50

the temperature range was such that the rate constants are
considerably smaller and the half lives are considerably
larger than the previous cases. The rate constant graphs
are Figure 23 and 2A and the Arrhenius plot is Figure 25.
The activation energy for decomposition is A2 Kcal/mol.

The dipyridyl ligand is more rigid than the chelated
phosphine ligands discussed earlier. The increase in
activation energy for decomposition may reflect the in-
crease rigidity of the supporting ligands.

If the stiffness of the supporting ligand is an im-
portant factor then a supporting ligand such as 1,10-
phenanthroline should give the complex an even higher
activation energy for decomposition. Initial studies
over the limited temperature range of 70 to 80°C yielded
an activation energy for reductive elimination of A5 Kcal/
mole. Figure 26 gives the rate data and Figure 27 is the
Arrhenius plot for the decomposition of 1,10—phenathroline
tetramethylene nickel(II).

The relative activation energy ordering for the five
nickelacyclopentanes discussed is supported by the trans37
influencing ability of the ligands. The ligands with a
weak trans directing effect (such as dipyridyl and 1,10—
phenathroline) do not greatly weaken the trans nickel-
carbon bonds and hence have a high thermal decomposition
activation energy. The phosphine ligands, on the other

hand, have a greater trans effect and weaken the trans

51

10

V

ONO‘D

O‘Nmmd
.vvv

A
t»
t>

c.
c.
E>
t>
J.

 

 

I 7 . I T I
50 100 150 200 250 300

Figure 23. Log [(dipy)NiOJ as a function of time in

minutes at 80° ([3) and 60°C (A).

52

U" 0‘ V 01 \D
U V T

 

 

so 700 4150 700 250 $00 350

Figure 2A. Log [(dipy)NiQ J as a function of time in
minutes at 65° (A) and 70°C (D).

30.;
Y—j’

a mwwm
'YY

 

53

 

2
:7”
at
7.
5.
5-
4b-
3b

890 100 6150 69°

2'.8 2'.9 37.0

x10'3/7
Figure 25. Arrhenius plot for (dipy)NiO .

514

—l
0‘0
t

a. N m
I

 

 

50 100 130 200 250

Figure 26. Log [(1,10-phen)N£::]J as a function of time

in minutes at 75° G3), 70° «3) and 80°C 0A).

555

10,

O‘NQQD
I

O‘NmO-J
l

 

3 L 90° 80° 70°

 

2:70 2775 2.30 2385 2190 2:95
x10’3/1

Figure 27. Arrhenius plot for (l,10-phen)N{:] .

56

nickel-carbon bonds. Because of the weakened carbon-
nickel bonds in the phosphine complexes a lower thermal
decomposition activation energy is found.

Even though the crystal field stabilization energy
interpretation of stability of organometallic complexes
is not deemed as important as it once was, it, nevertheless,
might be instructive to discuss the nickelacyclopentanes
in terms of it. In particular, a comparison of ethylene
bis(diphenylphosphine)tetramethylenenickel(II) and di-
pyridyl tetramethylene nickel(II) is interesting. In

Figure 28 the d orbital energy spacing is shown for a

Au. :5
1L 3‘
4L 1L,xz a:

Figure 28. Relative d orbital energy levels for nickel(II).

8

square planar complex. Nickel(II) is a d species so the

energy difference between the dxy and the dx2—y2 orbitals
is the energy gap of interest. That is, the dxy is the

2

highest occupied orbital and dx -y2 is the lowest un-

occupied orbital.

57

38 has measured the ultraviolet and visible

Chatt
spectra of a series of trans[L, piperidine PtCl2] complexes.
It was found that tri-n-propylphosphine is a stronger li-
gand than piperidine. The difference in energy between

2 orbitals for the two complexes was

the dxy and dx2-y
found to be approximately 3 Kcal/mol. The crystal field
effect of tri-n-propylphosphine and triphenylphosphine

are very similar. Dipyridyl39 is a stronger field ligand
than piperidine. Hence the crystal field effects of di-
pyridyl and triphenylphosphine are very similar. That is,

the energy difference between the dxy and dx2-y2

orbitals
for the triphenylphosphine and dipyridyl nickelacycles
should be very similar.

YamamotouO has measured the activation energy for the
thermal decomposition of dipyridyl di-n-propyl nickel(II).
The complex was thermally decomposed in the solid state to
give equal amounts of propane and propylene which are the
B-hydride elimination products. The activation energy
was found to be 16 Kcal/mol. A comparison with the cor-
responding dipyridyl metallacyclopentane shows the metalla-
cycle to be more stable to thermal decomposition by 26
Kcal/mol. The increased stability most likely arises
from the inhibition of B-hydride elimination in the metalla—

cycle by the steric constraints of the alkyl ring.

58

Three Coordinated Nickelacyclopentanes

Molecular weight determination by the freezing point
depression of benzene indicates that bis(tri-n-butylphos-
phine)tetramethylene nickel(II) is three coordinated in
solution.“1 The yellow complex is extremely air sensi-
tive and rather difficult to manipulate. Tri-n-butylphos—
phine tetramethylene nickel(II) decomposes in toluene
solutions by B-hydride elimination at a temperature range

of 20 to “0°C. Table 7 shows the kinetics data as well as

Table 7. Rate data and activation parameters for (n-

Bu3P)N(:]

 

 

Temp. Rate Constant Half Life E’ AS# AH# AG‘

 

20 1.7x10‘5 sec-1 680 min 35 36 eu 3M 23
30 1.3x10'14 sec"1 89 min Kcal/ Kcal/ Kcal/

mol mol mol
40 8.1x10'” sec”1 in min

 

 

the activation parameters. Figure 29 to 31 are of the
logarithm of the concentration of tri-n—butylphosphine-
tetramethylene nickel(II) as a function of time and Figure

32 is the Arrhenius plot. The activation energy for this

59

d

 

 

1 1 l
foo 3'00 §no 700 900 1100 1300

Figure 29. Log [(nBuBP)2N£::]] as a function of time in

minutes at 20°C.

60

 

 

9
7.
6.
5.
4.
3_
A
2.
A

l_
9.
8»-
7.
6,,
5..

1 Y Y I 7 I

50 100 150 200 250 300

Figure 30. Log [(n—Bu3P)2N£::]] as a function of time in

minutes at 30°C.

61

\l
I

 

 

 

as N on ‘D .a
Y - 1 ‘
D

 

 

Figure 31. Log [(n-Bu3P)2N£::]] as a function of time in

minutes at “0°C.

62

1.
3

0‘1de

30° 20°

1 J J

311 3:2 313 at: 3.5

 

X10'3/1

Figure 32. Arrhenius plot for (n-Bu3P)Ni:j .

63

process is 35 Kcal/mol. The driving force for B-hydride
elimination from a metallacycle is not clear. The nickel-
B-hydrogen dihedral angle is practically orthogonal to the
optimum dihedral angle of 0°. A crystal structure”2 of
bis(triphenylphosphine)tetramethylene p1atinum(II) shows
that the metallacycle ring is puckered with an a-carbon
.76 angstrom out of the plane and a B-carbon .h angstrom
out of the plane in the opposite direction. However, this
distortion may not be adequate to allow the migration of

a B-hydride to the nickel. Whitesides“3 also finds that
bis(tri-n-butylphosphine)tetramethylene p1atinum(II)
decomposes thermally in hydrocarbon solvents to give pre-
dominately l-butene, the B-hydride elimination product.
However, no explanation is offered for B-hydride elimina—
tion predominating over other reaction pathways in metalla-
cycles.

One rationalization for B-hydride elimination from
bis(phosphine)tetramethylene nickel(II) complexes which
lose a ligand in solution is perhaps that the disassociated
phosphine helps to promote B-hydride type products. For
instance in Figure 33 the phosphine abstracts a proton to
form 29 which then picks up the proton from the phosphine.
However, the rate determining step cannot involve the
free phosphine and still have first order kinetics.

The change in coordination number from four to three

also changes the preferred ligand-nickel-ligand bond

6A

33 g ._ ...
P3P/NI -> PSP-Na PP3—9 [93PM _’ \/\
20

.6 1°
HPP3

Figure 33. A possible decomposition mode for (R3P)N{:j .

angle from 90° to 120°. The increase in the ligand-nickel-
ligand bond angle introduces considerable strain in the
hydrocarbon ring. The increase in ring strain may facili—

tate B-hydride elimination over other decomposition paths.

Five Coordinated Nickelacyclopentanes

In the presence of excess triphenylphosphine, bis-
(triphenylphosphine)tetramethylene nickel(II) will form
tris(triphenylphosphine)tetramethylene nickel(II). This
is a light brown or gold, air sensitive crystalline
solid. The complex decomposes at a temperature range of

0° to 30°C. Table 8 summarizes the kinetics and activation

Table 8. Rate data and activation parameters for (¢3P)3N{::]

 

 

 

Temp. Rate Constant Half Life E; AS¢ AH¢ AG#
0 3.6::10‘6 53 hrs 19 -1u eu 19 23
2o u.ux10‘5 262 min in Kcals/mol
L;

30 1.2x10' 96 min

 

65

parameters while Figures guto 36 summarize the rate data
and Figure 37 is the Arrhenius plot. In order to prevent
the dissociation of a phosphine ligand from the tris(tri-
phenylphosphine)tetramethylene nickel(II) it was necessary
to carry out the decomposition at low temperatures. High
temperatures shift the equilibrium between the square
planar bisphosphine complex and trisphosphine complex
toward the bis complex. The tris(triphenylphosphine)-
tetramethylene nickel(II) decomposes thermally by B-Y
bond cleavage to give two molecules of ethylene. The
isomerization of a nickelacyclopentane to a bis olefin
complex does not change the coordination number but reduces
the oxidation state from two to zero. The trisphosphine
nickelacycle is coordinately saturated so the isomeriza-
tion to the bis olefin complex requires the prior loss of
a phosphine. The 31P NMR studies of this complex reveal
that the resultant four coordinated complex is probably
the tetrahedral complex. The tetrahedral complex rapidly
isomerizes to the bis olefin complex which in the rate
determining step loses a molecule of ethylene. The products
of the reaction are ethylene and bis(triphenylphosphine)-
ethylene nickel(O). The scheme is presented in Figure 38.
The activation energy for this process was found to be 19
Kcal/mol.

A Newman projection of the tetrahedral complex shown

in Figure 39 illustrates the geometric relation of the

66

6‘16)th
‘V‘V‘

 

1

100 200 300 400 500 600 700 800 900 10001100 1200

Figure 3“. Log [(¢3P)3N{::]] as a function of time in

minutes at 0°C,

EVY

OVmO-d

 

 

100 200 300 400

Figure 35. Log [(¢3P)3N£::]] as a function of time in
minutes at 20°C.

68

 

 

 

 

 

 

 

 

160 200 300 ‘00

Figure 36. Log [(¢3P)3N€::l] as a function of time in

minutes at 30°C.

6359

van;

A U‘ 0
1

 

\med
T Y I I

U" 0‘
I

g. unaware-0

 

 

 

l l 1
3:3 3:: 3.0 3.6 3.7
x10‘3/1

Figure 37. Arrhenius plot for (¢3P)3NO .

70

//

Figure 38. Proposed decomposition pathway of (03P)3N<:j .

ring. The 8-7 carbon—carbon bond is cis co-planar with
the a-carbon nickel bond. This is the optimum geometry
for elimination of nickel with an ethyl fragment from the

two carbon unit which becomes ethylene.

71

N}

I:
II:

Figure 39. Newman projection of the tetrahedral complex.
Hoffman and Kochim4 have experimentally investigated
the reductive elimination of alkyl groups from trialkyl
gold complexes, R3LAu, as well as theoretically calculat-
ing the energetics of the process. Triphenylphosphine—
(trimethyl)gold(III) was calculated to isomerize with an
activation energy of .H eV and to reductively eliminate
ethane with an activation energy of .8 eV. These values
were obtained by using extended Huckel methods. Normally
extended Huckel methods would not be used due to its in-
ability to account for bond stretching deformations of
molecules. However, in this case the calculated bond
lengths corresponded extremely well with the actual
lengths. The computed value of .8 eV for reductive elim-
ination in the gold system corresponds to approximately
18.u Kcal/mol. The value measured for bis(triphenylphos-
phine) dimethyl nickel(II) was 19 Kcal/mol. While perhaps

running the risk of comparing apples and oranges the

72

agreement between the two values for reductive elimina-
tion is very good.

Initially, it was hoped that a series of metallacycles
in which the substituents on the tertiary phosphines were
varied in an orderly manner could be prepared and the thermal
decompositions studied. The study of a series of phos-
phines such as ¢3P, 02CH3P, ¢(CH3)2P and (CH3)3P on nickel-
acyclopentane decomposition would give some insight into
the electronic influence of the ligands. However, the
preparation of bis(diphenylmethylphosphine)tetramethylene
nickel(II) met with insurmountable technical problems as
did the preparation of bis(dimethylphenylphosphine)tetra—
methylene nickel(II).

Table 9 summarizes the activation parameters for all
the metallacycles of this study. The entropy of activation
reflects the relative ordering of the transition state com-
pared to the initial complex. No clear trend is apparent.
The triphenylphosphine cases give negative entropies of
activation while the others are positive.

The rate constants reported are presented in Table 10
along with the standard deviations. The standard deviations
reflect the closeness of the best least squares fitted
line to the actual data points. The agreement between
the best line and the experimentally determined data
points is moderate.

Experimental results verifying a number of low energy

73

Table 9. Activation parameters for nickelacyclopentanes.

 

 

 

Complex Eact Sact act act Product
(¢3P)2Ni:j 23 —2.3 eu 22 23 D
(¢3P)2N1:g§§ 19 -13 18 22 CH3CH3
(dpe)Nifj 35 15 314 29 [:1
(dpe)NiQ 37 21 36 28 D
(bipy)NO 142 38 I41 29 D
(1,10-phen)N{:] AS -- -- —- [:3
(n-Bu3P)2Ni:j 35 36 31: 23 W
(¢3P)3N1:) 19 -111 19 23 =

 

 

7H

 

 

 

Table 10. Standard deviation of rate constants and
Arrhenius slopes.
Rate Constant
Compound Temp. (Slope) Standard Deviation
19 3.2x10'5 .33
(¢3P)2N{:j 30 1.2x10'“ .63
no n.6x10'“ .07
Arrhenius 1.15Eu 1.3
xCH '5
(¢3P)2Ni\CH§ 6 6.3xlO-u .18
20 3.7x10 .5&
30 1.0x10'3 .83
Arrhenius 9.5x103 1.3
(dipy)N{:j 60 6.9x10"7 .002
65 1.6x10”6 .012
70 14.5x10"6 .016
80 2.6x10’5 .008
Arrhenius 2.11-tx10l4 1.5
(dpe)N<:] 80 1.5x10‘5 .19
90 7.0x10’5 .91
100 2.ux10‘“ .8u
Arrhenius 1.85x10“ 1.3
(dpp)N<:j 90 2.3x10'5 .uu
100 8.2x10’l5 .31
110 2.8x10'" .1u
Arrhenius 1.78x1014 1.25
(Bu3P)N§::] 20 1.7x10’5 .22
30 1.3x10'” .63
U0 8.1xlO-l4 .58
Arrhenius 1.7x10u 1.9

 

 

75

decomposition pathways for nickelacyclopentanes have been
presented. Nickelacycles were found to be more stable than
their corresponding alicyclic analogues. Reductive elimina-
tion has been found to be the main decomposition pathway
for four coordinated nickelacycles which is a dramatic
change in reactivity from nickel alicyclic complexes. The
activation energies for this process were found to reflect
the relative rigidity of the supporting ligand. The more
rigid the ligand the higher the activation energy was

found to be. Reductive elimination in nickelacycles was
compared to reductive elimination of a non-ring analogue
which also decomposed by reductive elimination. The next
chapter describes the initial study of the decomposition

of nickelacyclobutanes which are even less stable than

nickelacyclopentanes.

EXPERIMENTAL

General

All reactions and manipulations were done under argon
which had been purified by use of a BASF catalyst and
molecular sieves. All solvents were freshly distilled
from sodium (with added benzophenone)”5 under nitrogen.
Triphenylphosphine, bis(diphenylphosphine)ethane, bis-
(diphenylphosphine)propane, tri-n-butylphosphine, and
a,a' dipyridyl, and l,lO-phenathroline were used as pur-
chased.

Analytical glc was performed on a Varian series lUOO
FID chromatograph with a 20' x 1/8" Durapak column or a
13 ft paraffin wax/5% silver nitrate on alumina column.

NMR analyses were obtained using a Varian T-6O spectrometer
and 6 values were recorded relative to TMS. Infrared
spectra were obtained on a Perkin-Elmer 237B and were cali-
brated with polystyrene. Mass Spectra were run on a

Hitachi-Perkin Elmer EMU-6 spectrometer.

Synthesis of bis(triphenylphosphine)nickel(Illdichloride

A modification of Venanzi'su6

procedure was used. A
solution of 23.8 g (.1 mol) of nickel(II) dichloride hexa-
hydrate in 250 mls of glacial acetic acid and 20 mls of

water was added with stirring to a hot solution of 52.6 g

76

77

(.2 mol) of triphenylphosphine in 500 mls of glacial acetic
acid. The drab green precipitate which formed was cooled
overnight. Vacuum filtration yielded dark blue-black
crystals which were washed with ether and vacuum dried.

The yield was 52 grams (76%).

Synthesis and Standardization of 1,A dilithiobutane

In a 500 ml three neck round bottom flask equipped
with an addition funnel, argon line and magnetic stirrer
were placed 9 g of flattened lithium wire cut into fine
pieces and 250 mls of distilled ether. A solution of 20
mls of l,A-dibromobutane in 150 ml of ether was slowly
added with stirring over the course of three or four hours.
After 2 or 3 mls were added the flask was cooled to 0°C
with an ice bath for the remainder of the addition. When
the addition was complete the mixture was stirred for an
additional half hour and stored at 0°C until the suspension
had settled. The solution was filtered under argon and
standardized.

The concentration of the l,u-dilithiobutane solution
was determined by mixing a 10 ml portion of the l,u-dilithio-
butane solution with 3 mls of trimethylchlorosilane under
argon. The bis(l,A-trimethylsilyl)butane formed was com-
pared to a known amount (2 mmol) of l,2,4,5-tetramethyl-

benzene (durene) by glc on an 8 ft D0550 column at 100°C.

78

Synthesis of bis(triphenylphosphine)tetramethylene nickel(II)

A 100 m1 side arm round bottom flask was charged with
l g (1.5 mmol) of bis(triphenylphosphine) nickel(II) di-
chloride, positive argon pressure and 3 mls of dry de-
oxygenated ether. The flask was cooled to dry ice/ethanol
bath temperature and 30 mls (2.U mmol) of l,u-dilithio-
butane/ether solution was added via syringe slowly. The
solution was stirred for 10 minutes and then the cooling
bath was allowed to warm up slowly. At temperatures between
-35 and -15°C a bright yellow precipitate was seen. The
mixture was then filtered under argon. The resulting yellow
solid was washed with ether and vacuum dried.

The yellow solid was dissolved in dry oxygen free
toluene at —20°C. The yellow solution was transferred
via syringe to another vessel at —20°C leaving behind un-
dissolved lithium salts. The clear yellow solution was
condensed to half its original volume and stored at -78°C
overnight. The yellow crystals which precipitated were
filtered under argon and vacuum dried to give 22 percent

of the product.

Analysis of bis(triphenylphosphine)tetramethylene nickel(II)

A weighed portion of the complex was transferred to a
small schlenk tube fitted with a serum cap. Concentrated

sulfuric or hydrochloric acid was added via syringe through

79

the cap. The resulting gas formed was analyzed by glc
on a 20 ft Duropak column at 65°C. The complex was con-
sidered pure if more than 95% of the calculated gas volume

as compared to an internal standard (propane) was found.

Analysis of nickel (glyxime tegt)u7

150 mls of water was added to the acid decomposed
complex. The white precipitate (¢3P or ¢3PO) which formed
was warmed slightly on a hot plate, cooled, and filtered
to give a clear solution. This solution was heated to
60°C and 20 mls of a freshly prepared one percent solution
of dimethyl glyoxime in absolute ethanol was added. Con—
centrated ammonium hydroxide was added dropwise until the
characteristic red precipitate no longer formed. The mix-
ture was kept at 60°C for one additional hour and then
cooled. The red solid was aspirator filtered into a tared
fritted glass crucible. The solid was dried to constant
weight at 130°C and the percent nickel present was cal-

culated by the following relation.
%Ni = (wt. of ppt. x 20.313)/wt. of sample

Synthesis of ethylene bis(diphenylphospine) nickel(II)-
dichloride

 

3.35 g of bis(diphenylphospine)ethane was dissolved in

250 mls of warm ethanol. An ethanol solution of nickel(II)-

80

dichloride hexahydrate (2 g in 77 mls) was added to the
warm diphos solution with stirring. A dark solution
resulted which suddenly turned orange with the production
of a metallic orange precipitate. The solution was cooled
and filtered. The product was vacuum dried to yield 3.8

g (86 percent) of product.

Synthesis of ethylene bis(diphenylphosphine)tetramethylene

nickel(II)

 

In a 100 ml side arm round bottom flask connected to
an argon line were placed 2 g (3.9 mmol) of ethylene bis-
(diphenylphosphine) nickel(II) dichloride and 3 mls of
dry oxygen free ether. The flask was cooled to dry ice-
ethanol bath temperatures and 56 mls of a .11 M solution
of l,A-dilithiobutane in ether was added via syringe.
The brownish solution was stirred for 10 minutes and slowly
warmed by letting the dry ice bath warm up. At -10 to
0°C a yellow precipitate fell out which was filtered under
argon. The solid was washed with ether, ethanol, and water.

The solid was vacuum dried to yield .7“ g (38%) of product.
Preparation of propylene bis(diphenylphosphine) nickel(II)
dichloride

A one liter Erlenmeyer flask was charged with 8.2 g

(20 mmol) of bis(diphenylphosphine) propane and 600 mls

81

of ethanol. The solution was heated to 60°C with stirring.
A solution of 0.7 g (20 mmol) of nickel dichloride hexa—
hydrate in 200 mls of ethanol was added to the phosphine
solution and removed from the heat. A red solid precip-
itated which was cooled, filtered, and dried. The product

was formed in BA percent yield.

Synthesis ofypropylene bis(diphenylphosphine)tetramethylene

nickel(II)

 

In a 100 ml round bottom side arm flask under argon
were placed 1 g (1.8 mmol) of propylene bis(diphenyl-
phosphine) nickel dichloride and 5 mls of ether. The
flask was cooled to -78°C. As the suspension was stirred
20 mls (2.9 mmol) of l,A-dilithiobutane/ether solution was
added slowly. The cold bath was allowed to warm to -10°C.
A yellow precipitate slowly formed as the solution warmed
up. When all the red starting material was gone and the
reaction was complete the solution was filtered under argon.
The yellow solid was washed with deoxygenated ether,
ethanol, and then water. The product was vacuum dried
to yield .61 g of product which is 63 percent of the theo-

retical yield.

Synthesis of bis(triphenylphosphine)dimethyl nigke1(II)u8

In a 100 ml side arm flask purged with argon was

placed 1 g (1.5 mmol) of bis(triphenylphosphine) nickel

82

dichloride, 5 mls of ether and 15 mls of tetrahydrofuran.
The flask was cooled to -78°C. An addition funnel was
charged with 20 mls of ether and 1.5 mls (3. mmol) of methyl
lithium. The methyl lithium solution was slowly added

at -78°C over the course of two hours. After an additional
half hour of stirring the solution was a light yellowish
brown. Filtration under argon at -78°C yielded a yellow
solid which was washed with ether and then with water.

After vacuum drying .59 g (58 percent) of the product was

obtained.

Synthesis of bis(tri—n-butylphosphine)ynickel(II) dichloride

9.5 g (”0 mmol) of nickel(II) dichloride hexahydrate
was dissolved in 25 mls of degassed ethanol. 22 mls (88
mmol) of tri-n-butylphosphine was added to the solution
and the mixture was stirred for 40 minutes at room tempera—
ture. The solution was cooled and filtered. The purple
crystals were recrystallized from hexane and vacuum dried

to yield 7.2 g (75 percent) of product.

Synthesis of bis(tri-n-butylphosphine)tetramethylene

nickel(II)

In a 200 ml side arm round bottom flask purged with
argon were placed 3 g (5.6 mmol) of bis(tri-n-butylphos-

phine) nickel(II) dichloride and 10 mls of ether. The

83

flask was cooled to —78°C with a dry ice/ethanol bath.

52 mls of a .18 M solution of l,A-dilithiobutane (9.9

mmol) was added slowly via syringe. The suspension was
stirred for two or three hours at temperatures below -35°C.
During this time the solution slowly turned yellow and a
precipitate formed. The solvent was removed in vacuo
below -35°C. Pentane was then added to the dried gum and
the resulting yellow solution was removed by syringe to
another vessel also at -35°C. This was repeated until
only a white residue was left in the flask. The pentane
solution was condensed to half its original volume at -35°C
and stored at -78°C for four days. The yellow crystals
were collected and dried under vacuum. The yield was 10

percent.

Synthesis of tris(triphenylphosphine)tetramethylene nickel(II)

Method A. 20 mls (2.9 mmol) of an ether solution of
l,A-dilithiobutane was slowly added to l g (1.5 mmol) of
bis(triphenylphosphine) nickel dichloride in 5 mls of
ether at -78°C. The mixture was slowly allowed to warm
to 0°C. The solution was stirred at this temperature
for “5 minutes while the solution turned from dark brown
to yellow. An excess of triphenylphosphine (1.5 g, 5.7
mmol) was slowly added to the solution. The color turned
to brown with a golden precipitate. The solution was

filtered. The golden solid was washed with ether and

8H

recrystallized at -50°C from toluene which had been
saturated with triphenylphosphine at -60°C. The golden
crystals were obtained in 98 percent yield.

Method B. To a toluene solution of 1 gram (1.5 mmol)
of bis(triphenylphosphine)tetramethylene nickel(II) at
-20°C was added an excess of triphenylphosphine (.82 g
3.1 mmol). The solution was stirred for two hours at
-10°C. The solution was then cooled to -78°C overnight
and the golden solid was collected by filtration. The
solid was washed with ether and recrystallized.

Preparation Of anhydrous bis(2,uepentanedionato) nickel(II)!49

A solution of 59.9 g (.25 mmol) of nickel(II) di-
chloride hexahydrate in 250 mls of water was added to a
solution of 50 g (.5 mmol) of 2,u—pentanedione in 100
mls of methanol. A solution of 68 g (.5 mol) of sodium
acetate in 150 mls of water was added. The solution was
heated briefly and cooled to 0°C in an ice bath. The
cooled light blue product was filtered and azeotrope dried
with toluene overnight. The dark green toluene solution
was filtered while warm to remove the sodium chloride.
The toluene was removed on a rotary evaporator until a
dark green oil remained. The oil was vacuum dried to

leave a green solid. The yield was “2 g (65%).

85

Preparation of bis(l,5-cyclooctadiene) nickel(O)50

In a 3 neck 500 ml round bottom flask fitted with a
dewar condensor, argon line, thermometer, and gas flow
adapter were placed 25.7 g (.1 mol) of anhydrous bis(2,u—
pentanedionato) nickel(II), 10 mls of toluene and 59 g
(.5 mol) of 1,5-cyclooctadiene which had previously been
passed through neutral alumina. The flask was cooled to
-20°C and a butadiene tank was connected to the gas flow
adapter. The butadiene was passed over type 3A molecular
sieves and then into the flask. About 5 g of gas was
added. The gas flow adapter was removed and an addition
funnel was attached. 100 mls of a 25% solution of tri—
ethyl aluminum in toluene was syringed into the addition
funnel. The triethyl aluminum solution was added over
the course of two hours while the flask temperature was
maintained at -10 to 0°C. The green solution slowly turned
brownish-yellow. After the addition of triethylaluminum
was complete the mixture was stirred an additional half
hour at 0° and then warmed to room temperature for three
hours. It was recooled and filtered under argon to yield
23 grams (83%) of the bright yellow product. It would

turn black upon exposure to moist air.

Synthesis Of (aha'-dipyridy1)tetramethylene nickel(II)51

.33 g (1.2 mmol) of bis(l,5-cyclooctadiene) nickel(O)

was placed in an argon filled Schlenk tube and cooled in

86

an ice bath. A solution of .56 g (3.6 mmol) of a,a' di-
pyridyl in 6 mls of tetrahydrofuran was added via syringe.
The resulting solution was a deep purple. The solution was
stirred for 30 minutes and .1“ mls (1.2 mmol) of l,A—di-
bromobutane was added via syringe. The solution slowly
turned green and a precipitate was formed. The stirring
was stopped and the solid was allowed to settle. The green
solution was transferred via syringe to another Schlenk
tube and concentrated. Pentane was added and the green
solid was filtered under argon to yield 20 percent of

product. The complex is extremely air sensitive.

Measurement of a rate constant

A 30 ml Schlenk was purged with argon and charged with
.09 to .1 g of the nickelacyclopentane. The tube was
cooled and 10 to 20 mls of xylene was added via syringe.
The tube was sealed with a serum cap and .2 mls of propane
was added as an internal standard. The tube was stirred
for 30 minutes then placed in a constant temperature bath.
Gas samples were periodically removed with a 50 micro-
liter syringe and analyzed by glc on a 20 ft Duropak column
at 60°C. Product gases were identified by matching
retention times with authentic samples and confirmed by

co-inJection techniques.

CHAPTER II
THE IN SITU DECOMPOSITION OF NICKELACYCLOBUTANES
INTRODUCTION

Metallacyclobutanes have been cited as intermediates
in a number of important organometallic reactions. Both

52 and Ziegler-Natta polymerization53

olefin metathesis
are believed to utilize metallacyclobutanes as reactive
intermediates.

The current mechanism for olefin metathesis52 proposes
the alternation of a metal alkylidene species with a metal-
lacyclobutane as shown in Figure #0. The key step, as
far as product distribution is concerned, is the decomposi-
tion of the metallacyclobutane into an olefin and alkyli-
dene moiety. HughesSu reported metathesis to be first
order in both olefin and catalyst. It was also found
that the energy of activation for this reaction is less
than 10 Kcal/mol.

A proposed mechanism53 for polymerization involves a
metallacyclobutane. Figure 41 is very similar to the
mechanism for olefin metathesis except that the alkyli-
dene complex has a hydride on the metal which assists in

the decomposition of the metallacyclobutane. In polymeriza-

tion the metallacyclobutane decomposes by reductive

87

88

 

 

Figure 90. Current mechanism for olefin metathesis.

89

H

\=/
M-CHZP .2 114:0“? = “:4=CHI?

H /‘=\

H
9 =1» =18
J1 C

.J(

Figure 91. Proposed mechanism for Natta-Ziegler polym-
erization.

90

elimination of the hydride and one side of the ring where-
as in metathesis the metallacyclobutane decomposes by
B-Y bond cleavage.

56 and molybdenum57 metalla—

Platinum,55 tungsten,
cyclobutanes have been isolated and their chemistry in-
vestigated. Brown58 has prepared dichloro(3-methyltri-
methylene)p1atinum(IV) by the reaction of methylcyclo-
propane with Zeise's dimer ([PtC12(C2Hu)]2). This is

reported to be the easiest and most convenient preparation

A (Fr—04%; CIZP+©—

of platinacyclobutanes.

Tipper59 has investigated the thermal properties of
trimethylene p1atinum(IV) complexes by differential
scanning calorimetry. He found that platinacyclobutanes

of the general formula PtX2(CH2CH2CH2) and PtX2L2(CH2CH CH2)

2
where L is a nitrogen donor ligand decompose by reductive
elimination to give cyclopropane or via a w allyl hydrido-
complex as shown in Figure 92 to give propylene. The mode
of decomposition depends on the trans influence of the
ligand. Ligands of high trans influence (such as ethyl-
enediamine, dipyridyl, tertiary phosphines) follow path

A to give cyclopropane. Ligands of low trans influence

91

X21291 é: XLZP’r i> max, A

H
I
L\l —"> __, /\ [DH-2x2

Lf/

)<

Figure A2. Possible decomposition pathways for L2X2Pt:>>.

(such as ethylenediamine, dipyridyl, tertiaryphos-
phines) follow path A to give cyclopropane. Ligands

of low trans influence (such as pyridine, methylpyridine,
halogens) follow path B to give propylene. The bond
strengths calculated for the platinum carbon bonds is
27-30 Kcal/mol which is considerably lower than other
platinum carbon bond strengths. A typical value of 39
Kcal/mol was found for the platinum methyl bond in

60 The weaker bonds in platinacyclobutanes

(05H5)Pt(CH3)3.
reflect the ring strain in the platinacyclobutane.

Green61 has studied the thermal decomposition of
bis(cyclopentadienyl)trimethylene tungsten and molybdenum.

The decomposition at 80°C yields a mixture of predominately

92

cyclopropane with some propylene. However, the correspond-
ing methyl derivatives (05H5)W(CH2CH(CH3)CH2) and (C2H5)2W-
(CH2CH2CH(CH3)) evolve isobutane and l-butene as major
products upon thermal decomposition. The methyl substi-
tuent changes the major decomposition pathway from reduc-
tive elimination to B-hydride elimination. He has also
found that photochemical decomposition yields the olefin

of one less carbon numbers as shown in Figure 93.

CH
CPS/“3 —9 CPZW; 2 ——9 CPZVII=CH2
l

~

Figure 93. Products of (Cp)2W I by photochemical decom-

position.

Hagihara62 has used nickel(O) dipyridyl complexes to
couple d,w-dihaloalkanes to cyclic hydrocarbons. He pro-
posed that the coupling proceeds through a metallacycle
as depicted in Figure UH.

It was felt that this coupling reaction warranted
closer investigation. That is, when 1,3-dibromopropane
was used did the resulting metallacyclobutane decompose
exclusively by reductive elimination or were other products

formed? The presence of ring substituents affected the

93

decomposition of platinacyclobutanes but was a similar
decomposition mode change operative in the nickel system?

The next section describes the results on further investi-

gations of this reaction.

0 who 9, 0
ON’ KNO BM; N;N;@)n HA
0” )"

 

Figure AA. Proposed pathway for coupling of a-w dihalo-
alkanes.

RESULTS AND DISCUSSION

A series of tertiary phosphine and nitrogen donor
stabilized nickelacyclobutanes were prepared by the method
of Hagihara and thermally decomposed without isolation.
The product distribution and yields are listed in Table
11 for nickelacyclobutanes with no ring substitutents.

The main product, cyclopropane, arises from a reductive
elimination process. This is the same trend Green61
found for tungsten and molybdacyclobutanes. A significant
amount of ethylene was also formed. The ethylene could
arise from B-Y bond cleavage to give a metal alkylidene
species and ethylene. The ultimate fate of the alkylidene
unit is not clear. Dimerization of the carbenoid species
would give ethylene. This has been proposed as one pos—
sible termination step in olefin metathesis.63 Attempts

to trap the metal alkylidene unit in these systems failed

in part due to its low concentration. The decomposition

of the nickelacyclobutane without any supporting ligand
(entry 5) gave a very different product distribution from
the previous examples which had supporting ligands. The
main decomposition product was ethylene with a lessor amount
of cyclopropane.

The same series of ligands were used to prepare the
2-methyl nickelacyclobutanes shown in Table 12 along with

the decomposition product distributions and yields. The

99

95

Table 11. Decomposition products for nickelacyclobutanes.

 

 

Complex
Entry Prepared in Decomposition Product Yieldsa
Number Situ C /\ CH2CH‘2

 

1 d1pyN{::> 92 —- 2
2 (03P)2N(::> 82 10 7
3 (1,10phen)N{::> 90 2 7
A (diphos)N£::> 88 3 9

5 (COD)N{::>° 15 -— 85

 

 

aglc yields. Total yields were approximately 50 percent.
bTotal yield was only 20 percent.

96

Table 12. Decomposition products for methyl substituted
nickelacyclobutanes.

 

 

Complex
Entry Prepared in Decomposition Product Yieldsa

Number Situ 1 //§\§ Zf§L. //~.~/>f===\\ ===

l (dipy)N£::> trace 80 —- A 8

2 (03P)2N1 trace 17 10 52 20

3 (diphos)Ni - an 2 -- 90

u (l,lOphen)Ni - 7O 3 -- 7
b

5 (COD)N£E> trace 9 9 5 77

 

 

aglc yields. Total yields were approximately 50 percent.
bTotal yield was 22 percent.

97

main thermal decomposition product again is the reductive
elimination product, methylcyclopropane. Only in the case
of triphenylphosphine nickela(2-methyl)cyclobutane was a
significant amount of linear butenes formed. Again the
nickelacycle without supporting ligands decomposed to
form mostly ethylene.

At least three decomposition pathways are available
to nickelacyclobutanes. Reductive elimination seems to be
the major decomposition mode. This is not surprising in
light of the decomposition pathway of nickelacyclopentanes.
The production of propylene by B-hydride elimination is
a minor pathway. The four membered metallacycle is even
less puckered than metallacyclopentanes. Thus the nickel
B-hydride dihedral angle is closer to the tetrahedral
angle than the nickelacyclopentane case, so it should be
an unfavored decomposition pathway. a-B carbon-carbon
bond cleavage to form a metal carbene species and ethylene
occurs to a small extent in these systems. This decomposi-
tion pathway is reasonable on steric grounds. The nickel
and B-Y carbon bond are cis co-planar (eclipsed along the
a-B bond), the necessary geometry for concerted elimination.
Figure 95 gives a Newman projection of a nickelacyclo-
butane looking down the B-a carbon bond of the ring.

It might be expected that the methyl substituent on
the a-carbon of the ring would increase B-hydride elimina—

tion. Only for the triphenylphosphine case is there a

98

Figure 95. Newman projection of nickelacyclobutane.

significant production of linear butenes. Whitesides25
found that an d-methyl substituent on the ring did not
greatly increase B-hydride elimination in platinacyclo-
pentanes. In the present study, triphenylphosphine is the
only monodentate ligand investigated and perhaps the in-
creased freedom over chelated ligands facilitate B-hydride
elimination.

The methyl substituent also seems to increase the
production of ethylene. The metallacyclobutane may de-
compose by a-B bond cleavage in two possible ways. As
Figure 96 shows the initial a-B bond cleavage can occur
to give either a molecule of ethylene or propylene with
the corresponding alkylidene unit. Since only a trace of
propylene is found in these reactions, path B is not a
significant pathway. The metal ethylidene species from

path A may rearrange by B-hydride migration to form a

99

J.
\

Figure 96. Decomposition modes for L2Nif>>

o-bonded ethylene complex. This complex then reductively
eliminates the hydride and o-bonded ethylene to give ethylene
as the product. This mechanism for the formation of ethyl-
ene from the alkylidene complex is similar to the proposed
mechanism for a-hydride elimination.6l4

Attempts were made to trap the alkylidene unit formed
in the decompositions. Since 1,5-cyclooctadiene was present
in the reaction mixture it was thought that it might be
reacting with the metal carbene to form bicyclo[6.l.0]-

nona-3-ene. No trace of this compound could be found by

lOO

either NMR or glc. Running the reaction in cyclohexene
also failed to trap the intermediate. In a parallel study32
on nickelacyclohexanes it was possible to trap the nickel
carbene species. However, even when the concentration of
carbene complex is fairly high the yields of the trapped
compound is low. Hence, in the present case where the
nickel carbene species is present in low concentration it
is reasonable that no carbene trapping product was de-
tected.

In this preliminary study a series of nickelacyclo-
butanes and a-methyl nickelocyclobutanes were thermally
decomposed in situ. The main decomposition pathway is

reductive elimination with a-B bond cleavage and B-hydride

elimination pathways of lessor importance.

EXPERIMENTAL

Preparation and decomposition of dipyridyl trimethylene
nickel(II)

 

In a 200 ml side arm Schlenk tube under argon were
placed a .51 g (1.86 mmol) of bis(l,5-cyclooctadiene)
nickel(O). The tube was cooled to 0°C and 10 mls of tetra-
hydrofuran was added. A solution of .87 g (5.6 mmol)
of dipyridyl in 10 mls of degassed tetrahydrofuran was
added via syringe. The mixture was stirred for 30
minutes and .11 mls (1.86 mmol) of 1,3-dibromopropane was
added via syringe. The blue solution slowly turned green
as it was allowed to warm to room temperature overnight.
Gas samples were withdrawn from the head space after 29

hrs and analyzed by glc.

101

CHAPTER III

STUDIES TOWARD THE HOMOGENEOUS CATALYTIC

REDUCTION OF CARBON MONOXIDE

INTRODUCTION

The synthesis of hydrocarbon products from carbon
monoxide and hydrogen has been studied for the last
fifty years. The synthesis of hydrocarbons, usually on
heterogeneous catalysts, by the reductive polymerization
of carbon monoxide is commonly referred to as the Fischer-

67'69 synthesis reaction. The catalytic reduction

Tropsch
of carbon monoxide to methane is called the methanation
reactions.7O During World War II a tremendous research

71 enabled Germany to produce

effort headed by Franz Fischer
large quantities of gasoline and diesel fuels from carbon
monoxide derived from coal.

The current petroleum situation has renewed interest
in Fischer-Tropsch processes. There is only one commer-
cial Fischer-Tropsch plant in operation now. It is the
SASOL plant in South Africa run by the South African
Synthetic Oil Corporation, Ltd.

In the last two years efforts to find and develop

homogeneous analogs of both the Fischer-Tropsch and

methanation reaction have only been moderately successful.

102

103

Muetterties72 has found that metal cluster complexes have
catalytic activity for reduction of carbon monoxide.
Bercaw73 has found that mononuclear zirconium complexes
will reduce carbon monoxide and Feder7u has a mononuclear
cobalt catalyst for the reduction.

Muetterties75 reports that 033(CO)12 and Iru(CO)l2
catalyze the production of methane from carbon monoxide.
The catalyst displays excellent selectivity, no additional
products were formed. The conditions employed in the re-
duction (190°C 2 atm pressure) were so mild that the reac-
tion rate was very low, only one percent conversion after
five days. However, by adding a molar equivalent of tri-
methylphosphine the rate increased three fold and the
selectivity remained high. Addition of other phosphines,
such as triphenylphosphine also increase the reaction rate
but lowers the selectivity. That is, ethane and propane
were also formed.

Muetterties76

also found that the major product of
carbon monoxide reduction could be shifted to ethane by
changing the solvent from traditional organic solvents

to molten NaC1'2AlCl3. The use of Lewis acid solvents
may result in substantial oxygen aluminum interactions
facilitating the carbon oxygen bond cleavage. The use

of Iru(CO)l2 in NaCl-2A1Cl3 at 180°C and 1.5 atm pressure
(CO/H2=l:3) for twelve to twenty-four hours resulted in

the production of substantial amounts of methane and

10“

ethane with minor amounts of propane and isobutane.
Bercaw73 has stoichiometrically reduced carbon monoxide
to methanol with (hS-CSMe5)2Zr(CO)2 at 25°C and 1.5 atm

hydrogen pressure. A formyl complex77 is proposed as an

HZ C

intermediate in the reaction. This zirconium complex is
the first reported mononuclear complex capable of reducing
carbon monoxide. However, attempts to make the reaction
catalytic have failed.

Feder and Rathke7u found that HCo(CO)u catalyticly
reduces carbon monoxide to primarily methanol and methyl
formate with lessor amounts of ethanol, ethyl formate and
l-propanol. The reaction was run in benzene at 200°C and
300 atm of an equimolar mixture of hydrogen and carbon
monoxide. It was found that in p-dioxane at 182°C carbon

monoxide hydrogenation proceeded even more rapidly and

105

selectively. Feder believes that the ethanol and l-propanol
formed are products of secondary reactions.

A number of mechanisms have been proposed for the
heterogeneous catalyzed methanation reaction. Two mechan-
isms are presented in Table 13. Fontaine's mechanism78
prOposes a metal formyl species which picks up a proton
to form an oxygen stabilized alkylidene species or a
bridging species. The other mechanism proposed by
Vlasenko79 also has an oxygen stabilized carbenoid species
as an intermediate. A more recent mechanism80 based on
known homogeneous steps is shown in Figure A7 for the
Fischer-Tropsch reaction. In step 7 of the mechanism a
metal carbene is formed.

Heteroatom stabilized metal carbene complexes have
been isolated81 and their chemistry studied. The first
preparation of moderately stable carbene complexes pro-
ceeded by the action of alkyl lithium on tungsten hexa-
carbonyl. Since the carbon of the coordinated carbonyl
is the most positive center, alkylation occurs there.

The resulting lithium salt is acidified and treated with

diazomethane or alternately directly alkylated with tri-

alkyloxonium tetrafluoroborate.81
LiR OLi OH CH N OCH
W(CO)6 + (CO) W=C/ -+ (CO) W=C/ 2 2 (CO) W=C/ 3

\& [(CH3)30]BFu

106

 

 

2 + O

m

m +
mm

a
ID A: I \z + 2
mm mmox
z + \z + z
o\ mm + mouo\
\ mm + oo
.0 + \z + 00 + z
+ 00

N
.. 5+m+m+z
mmm\

opfim coapwcfiopooo Have: u 2

0mm + emu .. N x:
mm mo mo
2 + z + z
memmox moom\ xx
szn2 so \2 + \z + \2
mo m zoom ouom m
2 + z + z + z
u \ _ \
ouom co m

am + 2m + mm
m

+ z + 00

O—EI

 

Emficmnomz m.ox:mmma>

Emflcmnomz m.mcfimucom

 

 

mEchwnooz coaumcmnpmz

.MH manme

107

H
00 H2
H-M -—3 MC-M ‘——’> M-C-M —'9 H-C M—H
1 2 H l 3 ".n
O O H 0'
in
6 5
M-H + CH CH (——- CH2-M (———— CH2-M
3 6 l H I
H H 2 OH

 

   

H2C =
8
co H2 H H2 5
CH -M -——9 CH —C—M -——9 CH -C-M -——9CH -CH-M
3 9 3 u 10 3 5 11 3 l n
O H OH H
12
-H20
CH3CH20H
CO M-H

propagation 4-——-CH3CH2-M

B-H abstraction

 

 

CH2=CH2 1%
M-H

Figure A7. Recent Proposed Fischer-Tropsch mechanism.

108

It was felt that one possible approach to developing
a homogeneous analog of the methanation reaction might
be to combine the preceding chemistries. That is, the
treatment of a metal carbonyl with a nucleophile under
hydrogen and carbon monoxide pressure might generate
"Fischer style" carbenes which may react further to give
hydrocarbons or alcohols. The proposed scheme is in
Figure A8. The scheme shows two possible pathways, one
giving hydrocarbons of increased carbon number, the other
giving alcohols of increased carbon number. There are
parallels which can be drawn between the proposed reaction
sequence and the heterogeneous reaction mechanisms. Metal
carbene species are proposed as intermediates in both
cases. Metal hydrides are present in both sequences.
Bimolecular reaction pathway are possible in the proposed
homogeneous reaction but will not be considered here.

The next section presents a discussion of the attempts
to make the proposed scheme into a viable reaction path-

way.

109

 

 

M(CO)X
Nu:
,o'
.-/
(CO) M=C\
y Nu
H2
/OH
(cc) m=c\
yH Nu
/2 ‘2
OH
(CO)yM-CH (CO)M=CH + H2O
l
Nu H Nu
co H2
NuCH20H (CO)M-CH2Nu
H
+
co
(CO)MH H2
H2
Nu (CO)M CH3Nu
1H ,70-
(CO)M'=CC\
J, Nu
,OH
(CO)M=C\
Nu

Figure A8. Proposed homogeneous pathway for CO reduction.

RESULTS AND DISCUSSION

Transition metal carbonyl complexes were treated in
solution with various nucleophilic bases. These reactions
were carried out under hydrogen and carbon monoxide pres—
sures ranging from A5 to 70 psi. The ratio of hydrogen
to carbon monoxide was varied from 1 to A. Table 1A
gives the reactants, conditions, and products for repre-
sentative reactions. The reactions listed are the results
of optimization of the reaction conditions. Entry five
gave the highest yield of methane for a run without a co-
catalyst. Longer reaction times did not increase the
yields.

In an attempt to increase the yields of products,
Lewis acid co-catalysts were added to the mixture. Aluminum
isopropoxide had virtually no effect on the reaction as
did vitride.82 However, diisobutyl aluminum hydride did
increase the yields of methane. As shown in Table 15
yields as high as 60 percent were obtained by using tung-
sten hexacarbonyl, sodium propoxide and diisobutylaluminum
hydride in hexane at 105°C under 50 psi pressure. This
modest yield of methane based on tungsten hexacarbonyl
represents the highest yield for this study. Isobutyl
aluminum hydride did not reduce carbon monoxide by itself
under the conditions of the study.

The reduction of carbon monoxide was less than

110

Handgun 0 +0
.wamme op m>fipmamh omphoamp mam mcaofiwn
.oapmh Hapme on mmmmm

 

 

111

 

smO mess» Om OOH NH om ems NO Hoem HHOOVNHONnm HH
emu sense we ONH OH em esOHOHO OmH erm OOmHmm VHOsm OH
u--- we OmH HH mm OHHOOOO me OzoeHmmOOmmO mHOOOem O
smO sense we OMH OH mm OHHseeO msH goem mAOOOem O
11:: em OO OH OH mme O xo+ OHHOszsz A
---- em mm OH mm see a HO+ OHHOOONOE O
emu HOH Hm OHH ON ON OHHOOOO OH mom OHOOVB m
emu sense mm mmH OH HO mO+ Hm xO+ OHoovz H
OmO seen» we msH OH mm see a xoem OHOOvz m
u--- we mm mH mm axe m.O HO+ OHoovz m
I--- Om mmH Om Om ems H emo+ OHOOO3 H
nmpozoopm oEHB .QEoB 00 mm pcm>Hom 2\mm ommm umzHMHOQ mwmmmwmm

 

 

.mcoHpomop COHposnog mUoncoE conpmo mo mmeESm .za oHnt

112

.NHOOOONOONOOOONOHOOZ OH OOHeOH>O

 

 

 

 

 

HNHHOOO Helmeee
I--- Om OOH OH OH -eOOOHe OO Ozo Om -eOOOHO OHOOOO O
OeO moses O: OOH Om Om OOOHOHO . uuuuu u- OOHeOH> nu: O
OeO some» Om OOH Om Om msOxOHO O eeOOz Hm OOOHeOH> OAOOvz e
seO sense NH OO OH Om manger u uuuuu In eHemAsmHO :1. O
I--- am OO ON on ems O eeoez nu eHestmHO OAOOO3 O
OeO Osm Os OHH NH Om OOOHOHO m seOOz mm eHONHsmHO OHOOO3 3
OOO OOO Os OOH OH Om messes m eeOOz Om eHemAsmHO OHOOO3 m
a--- O: OOH OH eH ems O eoee O HmHmeOOeOOOHO OAOOO3 m
nun- Os OHH OH mm OssonO m seOOz O.O eHONHsOO OHOOOee H
mposoosm 9:3. .959 00 mm pco>aom S\m mmmm 2\oo umzfiwpmooo umOHmpwo ..HwQESZ
mocmSGom
.mpmmHmpmoloo anz OCOHposcos 00 no assessm .mH magma

113

stoichiometric so a study to determine which of the pro—
posed steps was not operative or which was the difficult
step was undertaken. The oxygen stabilized tungsten
carbene complex 22 was prepared. The complex was added

to the reaction in place of tungsten hexacarbonyl and

/OL1

(CO)5W = 0\

CH3

£8

stirred for A8 hours at 130°C under 30 psi of hydrogen.
Ethanol was formed in 85 percent yield. Hence under the
conditions used, if the carbene complex was formed, it
should proceed smoothly to products. In order to see if
the carbene complex was formed under the reaction condi-
tions a 13C NMR study was performed. Tungsten hexacarbonyl
and potassium T-butoxide were stirred in tetrahydrofuran
for three hours and the spectra was obtained. A peak was
seen at 332.5 ppm relative to TMS. The chemical shift of
the peak is in the range characteristic of oxygen stabi-
lized tungsten carbene complexes.83 Hence it is possible
that the carbene complex could be formed. An aliquot of
a reaction mixture was removed and analyzed by 13C NMR.
No peak in the range of 310-3A0 ppm was seen. That is,
the carbene complex was not formed in appreciable concen-

trations under the reaction conditions. Attempts to '

11A

increase the concentration of tungsten hexacarbonyl in
the reactions were limited by the low solubility of the
complex.

It was hoped that by using stronger nucleophiles the
concentration of the carbene species might be increased.
A series of sodium and potassium alkoxides were used,
but the nucleophilicity is roughly the same for all of
them. The sodium salts of long chain thiols, such as
laurylthiol, are more nucleophilic than the alkoxides.
However they did not seem to have much affect on the product
yields.

The sodium salt of diethylene glycol monomethyl ether
was used as both a nucleophile and solvent. The results
from this system, shown in Table 16 yielded considerable
methane. However, the methane was a degradation product
of the diethylene glycol monomethyl ether rather than a
reduction product of carbon monoxide. That is, entry 6
shows that the methane was produced without the presence
of carbon monoxide.

The homogeneous reduction of carbon monoxide by metal
carbonyls and bases was unsuccessful because of low con-
centrations of "Fischer style" carbene complexes being
formed in solution. That is, in order for the scheme to
be usable a method of generating the oxygen stabilized

carbene complex in greater yield is needed.

115

.copmm25p so comma mew mpaoawn

.oo.mm mo woman CH cows< awn omm

 

 

 

 

 

OeO OOH Om OOH OHne OO 1. OzoOHmeOmeOOOmeO OHOOO3 O
OeO HON OO OOH OH OO .1 OzoOHmeOmeOOOmeO 1: O
OeOOeO .OeO OO OOH OH Om OOH OzomfimeomeOOOmeOOmO OAOOO3 O
Rm umm
OeO OOO me OOH OH Om OO OzoOHmeOmeOOOOeO OHOOO3 m
OOO OeOH Om OOH O NO OO OzoOHmeOmeOOOmeO OHOOO3 m
OeO Om OOH O OO :. OzoOHmeOmmOOOmeO OHOOvz H
nmposposm oEHB QEoB 00 mm E\m Apco>aomv ommm pmmampmo smnEsz
mocosuom
.mcoHpmomswmo pcm>Hom .OH manna

EXPERIMENTAL

Synthesis of (CO)5W=C(CH3)OL18u

A three neck one liter round bottom flask was purged
with argon and charged with 10.5 g (30 mmol) of tungsten
hexacarbonyl and 500 mls of ether. 30 mmols of methyl
lithium in 100 mls of ether was added via an addition
funnel with rapid stirring. The yellow suspension slowly
changed to a cloudy solution over the course of two hours.
The solution was then filtered under argon to remove un-
reacted tungsten hexacarbonyl and concentrated to 70 mls
of a dark oil. The oil was taken up in pentane and cooled
to -78°C. Yellow crystals slowly formed which were col-

lected and dried.

 

Hydrogenation of (CO)5W=CCH3OLi

1.8 grams (A.8 mmols) of the title compound was placed
in a glass pressure bottle along with 25 mls of tetra-
hydrofuran. The pressure bottle was charged with 30 psi
of hydrogen and was heated to 130°C for A8 hrs. The bottle
was opened and no volatile products were detected by glc.
The liquid phase contained ethanol in 85% based on start-
ing complex. The ethanol was analyzed by glc and confirmed

by a mass spectrum.

116

117

Typical Carbon Monoxige Reduction Run

25 mls of a .05 M solution of tungsten hexacarbonyl
(1.25 mmol) in tetrahydrofuran was placed in a pressure
bottle along with .A grams (A.75 mmol) of potassium
ethoxide. The pressure vessel was then pressurized with
10 psi of carbon monoxide and 35 psi of hydrogen. The
pressure bottle was then heated at 1A0° by means of an
oil bath for 2A hrs. The pressure bottle was cooled and
the gas phase was analyzed on a 10 ft molecular sieve
column and the liquid phase was analyzed on a 16 ft DC

550 column at 60°C on a purapak Q column at 85°.

REFERENCES

10.

ll.

12.

13.

1A.
150

l6.

17.

18.

19.

REFERENCES

R. Noyori, et al., J. Am. Chem. Soc. 26, 63A (197A).

 

F. G. A. Stone, T. A. Manuel, S. L. Stafford, J. Am.
Chem. Soc., 83, 2A9 (1961).

 

J. Osborn, et_al., J. Am. Chem. Soc. 25, 597 (1973).

 

T. J. Katz and S. Cerefice, J. Am. Chem. Soc. 91,
2A05 (1969).

 

 

J. Halpern and P. E. Eaton, J. Am. Chem. Soc. 22,
3515 (1970).

P. Binger, et al., J. C. S. Chem. Comm. 376 (1976).

G. E. Coates, M. L. H. Green, K. Wade, Organometallic
Compounds, Vol. 2, Methuen, London, 3rd Ed. (1968).

 

 

J. Chatt and B. Shaw, J. Chem. Soc. (1959) 705.

 

J. Chatt and B. Shaw, J. Chem. Soc. (1960) 1718.

 

F. Anet and E. Leblanc, J. Am. Chem. Soc. 79, 26A9
(1957).

U. Giannini and U. Zucchini, Chem. Comm. 9A0 (1968).

 

 

G. Wilke and H. Schott, Angew, Chem. Internat. Ed.
5, 583 (1966).

P. S. Braterman and R. J. Cross, J. Chem. Soc.
Dalton, 657 (1972).

K. Thomas, et al., J. Chem. Soc. (A), 1968, 1801.

 

B. K. Bower and H. G. Tennent, J. Am. Chem. Soc.,
25. 2512 (1972).

G. M. Whitesides, et al., J. Am. Chem. Soc. 92, 1A76
(1970).

G. M. Whitesides, et,§1., J. Am. Chem. Soc. 93,
5258 (1972)-

M. F. Semmelheck, et al., J. Am. Chem. Soc. 93, 5910
(1971).

G. B. Young, et al., J. Chem. Soc. Dalton (1977) 1892.

 

 

 

 

 

118

20.

21.

22.

23.

2A.

25.
26.
27.

28.

29.

30.

31.

32.

33.

3A.

35.

36.

119

G. M. Whitesides, et_a1,, J. Am. Chem. Soc. 99, 6529
(1976).

S. Otsuka, 33 al., J. Am. Chem. Soc. 99, 3180 (1973).

 

 

S. J. W. Price in "Comprehensive Chemical Kinetics",
Vol. A, C. H. Bamford and C} H.’H. Tipper, Ed., El-
sevier, Amsterdam, 1972, p. 197.

R. J. Puddephatt, et_a1., J. Chem. Soc., Dalton
(1976) 786.

9. M.)Whitesides, et al., J. Am. Chem. Soc. 99, AA51
1973 . ——'—_

G. M. Whitesides, J. Am. Chem. Soc. 99, 6521 (1976).

 

 

 

G. M. Whitesides, J. Am. Chem. Soc. 999, 5808 (1978).

 

Divinyl p1atinum(II) and diarylplatinum(II) complexes
thermally decompose to yield butadienes or biaryls.
These are exceptions to the generality.

P. S. Braterman, et al., J. Chem. Soc. Dalton, 1306

 

R. H. Grubbs, et al., "Organotransition-Metal Chemistry",
Y. Ishii and M. Tsatsui, Eds., Plenum Press, New York,

1975, P- 135-

A. Miyashita and R. H. Grubbs, J. Am. Chem. Soc. 99,

A. Miyashita and R. H. Grubbs, J. Am. Chem. Soc.
999, 1300 (1978).

A. Miyashita and R. H. Grubbs, J. Am. Chem. Soc. 999,
7A16 (1978).

 

 

 

A. Yamamoto, et al., Bull. Chem. Soc. Jap, 99, 1109
(1977). “"—

A. J. Gordon and R. A. Ford, The Chemist's Companion,
John Wiley and Sons, New York, 1972, p. 135.

 

(a) R. J. Gillespie, J. Chem. Educ. 99, 295 (1963).

 

(b) S. Otsuka, 33 al., J. Am. Chem. Soc. 99, 5850 (1976).

 

G. Booth and J. Chatt, J. Chem. Soc. (1965) 3238.

 

37.

38.
39.

A0.
A1.

A2.
A3.

AA.
A5.

A6.

A7.

A8.

A9.

50.
51.
52.

53.

120

F. A. Cotton and G. Wilkinson, Advanced Inorganic Chem.
3rd Ed., Interscience, New York, p. 667 and references
therein.

J. Chatt, 33.31., J. Chem. Soc. (1959) 10A7.

 

C, K. Jorgensen, Modern Aspects of Ligand Field Theory,
American Elsevier, New York, 1971, Chapter 26, p. 3A3.

 

A. Yamamoto, e£'§1., J. Am. Chem. Soc. 99, 3350 (1971).

 

M. L. Lui and R. H. Grubbs, unpublished results.
H. A. Eick, .22.El°’ Inorg. Chem. 99, 2166 (1973).

 

 

G. M. Whitesides and B. Young, J. Am. Chem. Soc. 999,
5808 (1978).

R. Hoffmann, 32931,, J. Am. Chem. Soc. 99, 7255 (1976).

 

A. J. Gordon and R. A. Ford, The Chemist's Companion,
Wiley Interscience, New York, 19, p. A39.

 

L. Venanzi, J. Chem. Soc. (1958) 719.

 

R. Belcher and A. Nutten, "Quantitative Inorganic
Analysis", AMG McDonald, Butterworth and Co., Ltd.,
on on, 1970, p. 131.

H. F. Klein and H. H. Karsch, Chem. Ber. 999, 2628
(1972)-

R. A. Charles and M. A. Pawlchowski, J. Phys. Chem.
99, AAO (1958).

R. A. Schunn, Inorg. Syn. 1g, 5 (197A).

 

 

 

 

N. Hagihara, 33 al., Chem. Letters (197A) 1363.

 

Ea) T. J. Katz, Advan. Organomet. Chem. 16, Chapter 7
1977).

(b) T. J. Katz, et al., J. Am. Chem. Soc. mm, 2519
(1976).

(c) R. H. Grubbs, et al., J. Am. Chem. Soc. 99,
3A78 (1976). _—"_'

M. L. H. Green, e£_al., J. C. S. Chem. Comm. 60A
(1978).

 

 

 

 

5A.
55.
56.

57.

58.

59.

60.

61.

63.

6A.

65.
66.

67.

68.

69.

70.

71.

72.

121

 

W. B. Hughes, J. Am. Chem. Soc. 9%, 532 (1970).
J. Chatt, £2.3l" J. Chem. Soc. (1976) 738.

 

M. L. H. Green, 33 al., J. C. S. Chem. Comm. 619
(1975).

M. L. H. Green, et al., J. Chem. Soc. Dalton, 1132
(1977).

D. B. Brown, et al., J. Organomet. Chem. 999, A31
(1978).

C. F. M. Tipper, et al., J. Organomet. Chem. 99,
A23 (197A). —_'-—

 

 

 

 

S. J. Ashcroft and C. T. Mortimer, J. Organometal.
Chem. 99, 501 (1970).

 

M. L. H. Green and M. Ephritikhime, J. C. S. Chem.
Comm. 926 (1976).

 

N. Hagihara, 32.21" Chem. Lett. 1363 (197A).

 

C. R. Hoppin, Ph.D. Dissertation, Michigan State
University, 1978.

G. W. Parshall and R. R. Schrock, Chem. Rev. 19,
258 (1976).

R. R. Schrock, J. Am. Chem. Soc. 91, 6577 (1975).

 

 

P. G. Gassman and T. M. Johnson, J. Am. Chem. Soc.

99. 622 (1977).

H. Pichler and Erdoel Kohle, Endgas. Petro. Brennst.—
Chem. 99, 625 (1973).

H. Pichler and H. Schulz, Chem. Ing. Tech. 99, 1162

H. Storch, et al., The Fischer-Tropsch and Related
Syntheses, Wiley, New York 1951.

 

 

 

 

 

Len Seglin, Ed. Methanation of Synthesis Gas, Advances
in Chemistry Series 1A6, American Chemical Society,
Washington, D.C. 1975.

 

F. Fischer: Herstellung Flfissiger Kraftstoffe aus
Kohle Borntrager, Berlin, 192A.

E. L. Muetterties, Science 196, 839 (1977).

73.

7A.

75.

76.

77.

78.

79.
80.

81.

82.

83.
8A.

85.

122

J. E. Bercaw, SE al., J. Am. Chem. Soc. 99, 6733
(1976).

H. M. Feder and J. W. Rathke, J. Am. Chem. Soc. 999,
3623 (1978).

E. L. Muetterties, et al., J. Am. Chem. Soc. 99,
1296 (1976). __ “—

 

 

 

E. L. Muetterties and G. C. Demitras, J. Am. Chem.
Soc. 99, 2796 (1977).

There is now precedent for the formation of formyl
complexes. J. P. Collman, J. Am. Chem. Soc. 99,

A089 (1973) reports the preparation of [Fe(CO)4(CHO)]
whereas C. P. Casey, J. Am. Chem. Soc. 99, 5395 (1976)
reports a general method for the preparation of metal

 

 

 

formyl complexes.

R. Fontaine, Ph.D. Dissertation, Cornell University,
1973.

V. M. Vlasenko, E£.§l° Kinet. Katal. 9, 2AA (1965).

 

G. Henrice-Olive and S. Olive, Angew. Chem. Int. Ed.

E. O. Fischer and A. Maasbol, Angew. Chem. 19, 6A5
(196A)-

Vitride is Sodium bis(2-methoxyethoxy)aluminum hydride
(NaAlH2(OCH2CH20CH3)2).

 

 

E. O. Fischer, £3 g1., J. Orgmet. Chem. 99, C37 (1973).

 

 

E. o. Fischer, 33921., Chem. Ber. 999, 257A (1977).

E. O. Fischer, Advances in Organometallic Chemistry,
Vol. 1A, p. 1, 1976.

 

           

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

WWI/I![WM/ll!!!“WHWWIIH

293 3146 1373