PART ONE
THE PREPARATION AND
CHARACTERIZATION OF A SERIES OF
CHLOROALKOXOBISCYCLO-
PENTADIENYLZIRCONIUMHV) AND
DIALKOXDBISCYCLO '
PENTADIENYLZIRCONIUMUV) COMPOUNDS

PART TWO
NITROGEN FIXATION UNDER MILD
CONDITIONS BY USING SOME '
ORGANOMETALLIC COMPOUNDS OF
THE SECOND-ROW TRANSITION ELEMENTS
AS THE ACTIVATING AGENTS

Thesis for the Degree of Ph.D.
MICHIGAN STATE UNIVERSITY
DONALD R. GRAY
197D

 

 

 

6,;

 

This is to certify that the

thesis entitled
PART ONE: THE PREPARATION AND CHARACTERIZATION OF A SERIES OF

CHLOROALKOXOB ISCYCLOPENTAD | ENYLZ IRCON I UM (IV) AND

D IALKOXOB ISCYCLOPENTAD I ENYLZ I RCON IUM (IV) COMPOUNDS

PART TWO: NITROGEN FIXATION UNDER MILD CONDITIONS BY USING
SOME ORGANOMETALLIC COMPOUNDS OF THE SECOND-ROW TRANSITION
ELEMENTS AS THE presented by ACTIVATING AGENTS

DONALD R. GRAY

has been accepted towards fulfillment
of the requirements for

Ph. D. Jegree in Chemistry A'r \‘1

Major professor

Date December, 8, 19 70

 

0-7639

ABSTRACT

PART ONE

THE PREPARATION AND CHARACTERIZATION OF A SERIES OF
CHLOROALKOXOBISCYCLOPENTADIENYLZIRCONIUM(IV) AND
DIALKOXOBISCYCLOPENTADIENYLZIRCONIUM(IV)-COMPOUNDS

PART TWO

NITROGEN FIXATION UNDER MILD CONDITIONS BY USING SOME
ORGANOMETALLIC COMPOUNDS OF THE SECOND-ROW TRANSITION
ELEMENTS AS THE ACTIVATING AGENTS

BY is
\,53

.

Donald REIGray
PART ONE

A new series of zirconium(IV) compounds of the type
(c5H5)22r(0R)nc12_n (n = 1 and 2; R = CH3. C2H5, ifC3H7)
were prepared and characterized. .The compounds were pre-
pared by allowing the appropriate alcohol to react with
(C5H5)2ZrC12 and triethylamine in tetrahydrofuran solvent.
All the compounds were purified by sublimation under vacuum.

The nuclear magnetic resonance spectra of the compounds
were used to show unequivocally the presence of the cyclo-
pentadienyl and alkoxide groups. The integrated areas of
the resonance peaks were used to determine 'n'. The change
in chemical shifts of (C5H5)22r(OR)nC12_n as 'n' increased
showed the opposite trend from what has been observed in

similar series of titanium(IV) compounds.

Donald R. Gray

The infrared spectra were studied to prove further the
existence of the cyclopentadienyl and alkoxide ligands and
to speculate on the electron density change about the zir-
conium as 'n' increased. The far-infrared spectra gave
clear indication of the presence of the Zr-OR and Zr-ring
bonds. The shifts in all infrared and far-infrared fre-
quencies were consistent with basicity arguments.

The hydrolysis reactions were studied. For n = 1, the
hydrolysis product was of the form (CPZCer)2O with a zir-
conoxane (Zr- 0 - Zr) linkage. For n = 2, the hydrolysis
product was of a more complicated form, probably
-[O-Zr(OH)2]n-. The infrared spectra, weight changes, and
melting point changes were consistent with these conclusions.

Polarograms of the compounds in THF were attempted but

were unsuccessful in yielding the desired reduction potential.
PART TWO

A survey of several early second-row transition metal
organometallic compounds was made to determine their ef-
ficacy as nitrogen-fixation agents under mild conditions.
Most of the compounds used were of the form (C5H5)2MClx_2,
where x is the oxidation state of M. Some of the com-
pounds prepared in Part One were also used.

AThe experiment was similar in all cases. Nitrogen was

continually passed through a solution of the complex in

tetrahydrofuran. A two-electron reduction of the compound

Donald R. Gray
'was attempted by using sodium naphthalenide. If the reduc-
tion was effected, the compound conceivably attached a
molecule of nitrogen. Hydrolysis at this point produced

ammonia.

rThe experiment gave measurable quantities of ammonia
when (C5H5)2NbC13, (c5H5)22rc12, and (C5H5)2Zr(OC3H7)Cl
were used. Several other compounds were effective in pro-

ducing very small amounts of ammonia.

PART ONE

THE PREPARATION AND CHARACTERIZATION OF A SERIES OF
CHLOROALKOXOBISCYCLOPENTADIENYLZIRCONIUM(IV) AND
DIALKOXOBISCYCLOPENTADIENYLZIRCONIUM(IV)-COMPOUNDS

PART TWO
NITROGEN FIXATION UNDER MILD CONDITIONS BY USING SOME

ORGANOMETALLIC COMPOUNDS OF THE SECOND-ROW TRANSITION
ELEMENTS AS THE ACTIVATING AGENTS

BY
0 (yoga
Donald RJ\Gray

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Chemistry

1970

DEDICATION

This work is dedicated to the following:

Joan, my wife, whose constant love and encouragement
made these years very good years;

Rick and Cindy, my children, whose existence makes
everything worthwhile;

Mr. and Mrs. W. Russell Gray, my parents, whose under-
standing and assistance have made rough times
smooth and whose general attitude has created
a close and wonderful family;

Vera Elliott, my mother-in-law, whose self-sacrifice

for her family has been inspirational.

ii

ACKNOWLEDGMENT

The author wishes to express his appreciation for the
guidance and leadership of Professor Carl H. Brubaker, Jr.
Financial support from the National Institute of Health

(Predoctoral Fellowship) was profoundly appreciated.

iii

TABLE OF CONTENTS

PART ONE

THE PREPARATION AND CHARACTERIZATION OF A SERIES OF
CHLOROALKOXOBISCYCLOPENTADIENYLZIRCONIUM(IV) AND
DIALKOXOBISCYCLOPENTADIENYLZIRCONIUM(IV)‘COMPOUNDS

Page
INTRODUCTION . . . . . . . . . . . . . . . . . . . . 2

EXPERIMENTAL . . . . . . . . . . . . . . . . . . .
A. Materials . . . . . . . . . . . . . . . . .
B. Analytical Methods . . . . . . . . . . . .

C. Experimental Apparatus and Technique . . .

q -a m a: A

D. Preparation of Compounds . . . . . . . . .
E. Hydrolysis Reactions . . . . . . . . . . . 12
F. Spectroscopic Measurements . . . . . . . . 13

G. Polarography . . . . . . . . . . . . . . . 13

RESULTS AND DISCUSSION . . . . . . . . . . . . . . . 14
A. Preparations . . . . . . . . . . . . . . . 14
B. Nuclear Magnetic Resonance Spectra . . . . 17
C. Infrared Spectra . . . . . . . . . . . . . 25
D. Hydrolysis Reactions . . . . . . . . . . . 42

E. Miscellaneous Observations . . . . . . . . 46

iv

TABLE OF CONTENTS (Cont.)

.

PART TWO

NITROGEN FIXATION UNDER MILD CONDITIONS BY USING SOME
ORGANOMETALLIC COMPOUNDS OF THE SECOND-ROW TRANSITION

ELEMENTS AS THE ACTIVATING AGENTS

INTRODUCTION . . . . . . . . . . . . . . . . .
EXPERIMENTAL . . . . . . . . . . . . . . . . .

A.
B.
C.
D.
E.

RESULTS
A.
B.
C.
D.
E.

BI BLIOGMPI-IY O O O O O O O O O O O O O O O O C

.Materials . . . . . . . . . . . . .
Analytical Methods . . . . . . . . .
.Experimental Apparatus and Technique
The Nitrogen-Fixation Experiment . .

Preparation of the Compounds . . . .

AND DISCUSSION . . . . . . . . . . . .
General Observations . . . . . . . .
Compounds of Zirconium . . . . . . .
-Compounds of Niobium . . . . . . . .
Compounds of Molybdenum and Tungsten

Ruthenocene ;. . . . . . . . . . . .

Page

49
54
54
56
57
57
6O

62
62
65
66
66
67

68

LIST OF TABLES

TABLE Page

I. Reactant Ratios and Sublimation Temperatures
for the Compounds Cp22r(OR)Cl, Cp22r(OR)2.
and zr(OR)4 O O O O O O ‘ O O O O O O ‘ O O O O 16

II. Proton NMR Chemical Shifts for szzr(OR)nc12_n
and ZI(OR)4 o o o o o o o o o o o o v o o o o 24

III. .Characteristic Frequencies (cm-1) of Alkoxide
Groups in Cp,,Zr(ORJ)nC12__n . . . . . . . . . . 36

IV. Far-Infrared Frequencies (cm—1) of Zr-O and
Zr-ring Bonds . . . . . . . . . . . . . . . 41

V. A Summary of the Results Obtained by Using

Various Transition Metal Compounds in the
Nitrogen-Fixation Experiment . . . . . . . . 63

vi

Far— infrared spectra of Cp2Zr(OC2H5)n Cl
- 2) and Zr(OC2H5) . . . .

LIST OF FIGURES

of
of
of
of
of
of
of
of

of

FIGURE

1. Proton nmr

2. Proton nmr of

3. eProton nmr of

4. Proton nmr of

5. Proton nmr of

6. Proton nmr of

7. Infrared spectrum
8. Infrared spectrum
9. Infrared spectrum
10. Infrared spectrum
11. Infrared Spectrum
12. Infrared spectrum
13. Infrared spectrum
14. Infrared spectrum
15. Infrared spectrum
16. Far-infrared spectra

(h =
17.
<n=o
18.

Far- -infrared spectra of Cp22r(OC3H7)n Cl
0-2) and Zr(oc3H7)4 . . . .

(n =

of Cp2Zr(OCH3)Cl . . .
Cp22r(OCH3)2 . . . .
CpZZrIOC2H5)Cl . . .
Cp22r(OC2H5)2 . . . .
Cp2Zr(OC3H7)C1 . . .

Cp22r(OC3H7)2 . . . .

szerlz . . .
CpZZr(OCH3)Cl
Cp2Zr(OCH3)2 .
Cp2Zr(OC2H5)Cl
szzr(0C2H5)2

Zr(oc2H5)4 . .
Cp2Zr(OC3H7)Cl
szzIIOCSH7)2

Zr(oc3H7)4 . .

of Cp22r(OCH3) c1 _
0‘2) 0 o o o o o o o o o c on 2 n

vii

2-H

2-D

Page
18
19
20
21
22
23
27
28
29
3O
31
32
33
34
35

38

39

4O

LIST OF FIGURES (Cont.)

FIGURE Page

19. Infrared spectrum of hydrolysis product of
Cp22r(OR)Cl . . . . . . . . . . . . . . . . . 43

20. Far-infrared spectra of hydrolysis products of
(A) Cp2Zr(OR)C1; (B) Cp2Zr(OR)2 . . . . . .. 45

21. Van Tamelen's proposed mechanism for nitrogen
fixation . . . . . . . . . . . . . . . . . . 51

22. Experimental appratus and procedure for the
nitrogen-fixation experiment . . . . . . . . 59

viii

PART I

THE PREPARATION AND CHARACTERIZATION OF A SERIES OF
CHLOROALKOXOBISCYCLOPENTADIENYLZIRCONIUM(IV) AND
DIALKOXOBISCYCLOPENTADIENYLZIRCONIUM(IV) COMPOUNDS

PART I
INTRODUCTION

Little research has been undertaken to synthesize
systematiCally and characterize mixed cyclopentadienyl-
alkoxide compounds [(C5H5)nM(OR)4_n] and cyclopentadienyl-
alkoxide-chloride compounds [(C5H5)mM(OR)3_mCl] of the’*
metals Ti and Zr. This void in organometallic chemistry
is noteworthy because of the increasingly widespread use of
organometallic ccmplexes in homogeneous catalysis.

More work has probably been done with titanium alk-
oxide-cyclopentadienide systems than with any other metal.
The complexes (C5H5)Ti(OC2H5)3 and (C5H5)Ti(OC3H7)3 were
synthesized by Nesmeyanov1 in 1960. The preparation of
(C5H5)2Ti(OC2H5)Cl was reported in the same article, but
attempts to prepare the biscyclopentadienyldialkoxide com-
plexes were evidently unsuccessful. Complete nuclear mag-
netic resonance2 and infrared spectra3 have been measured
for the series (C5H5)Ti(OR)3_nCln. The biscyclopentadi-
enyldialkoxide complexes have been conspicuously neglected.

The analogous zirconium complexes have also received
little attention. The complexes (C5H5)2Zr(OC2H5)Cl and
(C5H5)2ZI(OC3H7)C1 were reported but not convincingly char-
acterized by Brainina‘. This work was mainly concerned

2

3
with the study of zirconoxane (-Zr - O — Zr -) likages and
not the study of the (C5H5)2Zr(OR)Cl complexes. The only
other mixed cyclopentadienyl-alkoxide zirconium complex
reported is (C5H5)2Zr(OC3H7)2,5 but the complex has not
been obtained completely pure.

Part One of this research deals with the preparation
and characterization of the series (C5H5)ZZI(OR)C1 and
(C5H5)2Zr(OR)2, (R = CH3, c2115., i-C3H7). These complexes
Should not only be interesting from a synthetic standpoint
but may also be interesting as possible nitrogen-fixation

agents. Their role in the latter is discussed in Part Two.

EXPERIMENTAL SECTION

A. Materials

Bis(v-cyclopentadienyl)dichlorozirconium(;yl was ob-
tained from Arapahoe Chemical Company and used without
further purification.

Analysis Calcd for C10H10ZrC12: Zr, 31.22; Cl, 24.26.

Found: Zr, 31.09: Cl, 24.13.

Triethylamine was dried over barium oxide and pipetted

as needed.

Tetrabutylammonium iodide was obtained from Eastman
Organic Chemicals. The commercial product was further puri-

fied by standard procedures.6

Solvents. All solvents were refluxed with the appro-
priate drying agent under a nitrogen atmosphere unless indi-
cated otherwise.

Tetrahydrofuran (THF) was refluxed continuously with
lithium aluminum hydride and distilled immediately before
use.

Benzene was refluxed continuously with calcium hydride

and distilled immediately before use.

5

Hexane was continuously refluxed in the presence of
sodium metal and distilled immediately before use.

-Methyl alcohol was dried by distillation in the pres-
ence of magnesium. The magnesium was activated with iodine
according to the method of Lund and Bjerrum.7

Absolute ethyl alcohol was dried by distillation in
the presence of sodium ethoxide and diethylphthalate.

Isopropyl alochol was dried by distillation in the

presence of sodium isopropoxide.

Gages. Prepurified nitrogen from Liquid Carbonics was
passed through an activated copper catalyst (BTS catalyst -
BASF R-3-11 from Badische Anilin-Soda-Fabrik AG) to remove
oxygen. It was then passed through two drying towers filled
with calcium sulfate and barium oxide to remove any water

present.
B. Analytical Methods

Carbon and Hydrogen Analysis. These analyses were
performed by Spang Microanalytical Laboratories and by the
microanalytical laboratory of the Institute of Water Re-
sources at Michigan State University. The latter labora-
tory was not equipped to handle the moisture-sensitive zir-
conium complexes properly. Spang reported that the complexes
were very difficult to handle and that the analyses reported
were a result of experimenting with several handling
techniques. Identical samples gave divergent results, and

the reported analyses may not be totally reliable.

6

Chloride Analysis. About 100 mg of the sample of the
zirconium compound was destroyed by being heated in dilute
nitric acid. This solution was made basic with dilute
aqueous ammonia. The hydrous oxide of zirconium thus formed
was removed by filtration, and the filtrate was collected
in a.100 ml volumetric flask. The precipitate was washed
with several portions of dilute ammonia. IThe filtrate was
diluted with distilled water to 100.00 ml. Twenty-five ml
aliquots were made acidic with dilute sulfuric acid. Chlor-
ide was then determined by a differential potentiometric
titration with a standard silver nitrate solution.8 Dupli-

cate samples showed reproducible results.

Zirconium Analysis. Zirconium was initially analyzed
by igniting the hydrous oxide to zirconium oxide. This
method failed to give reproducible data. A modified ver-
sion of the cupferron method was finally employed, and this
method gave results of high precision. One-hundred mg
samples of the compound were decomposed by being heated in
10% sulfuric acid. This solution was cooled to 10°. An
excess of cupferron (the ammonium salt of nitrosophenyl-
hydroxylamine) solution (6 g in 100 ml water and filtered)
was added. The white, curdy precipitate was filtered and
washed with several portions of 1:10 hydrochloric acid.

The filtered precipitate was then ignited at 10000 for six
hours, and the zirconium content was determined by weighing

ZIOZ .

7

C. Experimental Apparatus and Technique

Practically all the compounds dealt with in this re-
search were extremely sensitive to oxygen and/or water vapor.
All reactions and manipulations were thus performed in an
atmosphere of dry nitrogen. Schlenk vessel techniques,
similar to those reported by Herzog,9 were used extensively.
The compounds were stored in closed vials in an inert-atmo—
qflxxe box. These manipulations included preparation of
Nujol mulls for infrared measurements, filling of nuclear
magnetic resonance tubes, and scraping sublimation products
from cold-finger condensers.

The use of a vacuum manifold in conjunction with the
nitrogen line allowed facile Schlenk manipulations which
precluded the presence of air in the reaction vessels.
Glassware was dried for several hours at 180°. It was then
evacuated and filled with nitrogen several times while it
cooled. Dry nitrogen gas was forced into the vessels
through a sidearm stopcock whenever a glassware connection

was made or broken.
D. Preparation of Compounds

Chloromethoxobis(v-cyclopentadienyl)zirconium(1y), A
mixture of methyl alcohol (32.2 mmol) and triethylamine
(32.2 mmmol) in 20 ml THF was added dropwise to a stirred
solution of zirconocene dichloride (16.1 mmol) in 80 ml THF.

This reaction mixture was allowed to reflux for three hours.

8
The dense white precipitate, (C2H5)3N'HC1, was filtered.
The filtrate, an orange-brown solution, was treated by re-
moving the solvent under vacuum. Seventy ml of hexane was
added to the remaining amorphous reddish—brown mass. This
mixture was vigorously stirred for six hours and crystals
formed. These beige crystals were filtered and dried.

1 mm) at 103° caused pure

Sublimation under vacuum (10-
white crystals to condense on the cold-finger condenser.
Yield 1.4 g. Melting Point = 111—114.5°.

Analysis Calcd for ZrClC11H130: Zr, 31.69; Cl, 12.32;
C, 45.89; H, 4.52.

Found: Zr, 31.51; Cl, 12.06; C, 45.35; H, 4.15.

Dimethoxobis(v-cyclopentadienyl)zirconium(IV). A
solution of methyl alcohol (83.6 mmol) and triethylamine
(41.8 mmol) in 10 ml THF was added dropwise to a solution
of zirconocene dichloride (20.9 mmol) in 80 ml THF. The
(C2H5)3N'HC1 was removed by filtration. A brown oil re-
mained after the solvent was removed from the filtrate,
Thirty ml of hexane was added to this oil, and beige crys-
tals formed. These crystals were removed by filtration.
The filtrate was evaporated and the remaining solid was
dried. The resulting brown solid was sublimed under vacuum

1 mm) at 85°. Pale yellow crystals of product formed

(10'
on the cold-finger condenser. Yield 2Z1.4 g. Melting
point - 63-68°.

Analysis Calcd for ZrC12H1302: Zr, 32.19: C, 50.86:

H. 5.65.

Found: Zr, 32.20; C, 48.91; H, 5.11.

Chloroethoxobis(w-cyclopentadienyl)zirconiumygyl. A
solution of ethyl alCohol (40.2 mmol) in 10 ml THF was
added to a solution of zirconocene dichloride (20.1 mmol)
in 80 ml THF. This reaction mixture was allowed to reflux
for two hours. A solution of triethylamine (40.2 mmol) in
10 ml THF was then added, and the refluxing was allowed to
continue for two more hours. The white precipitate,
(C3H5)3N-HC1, was removed by filtration. The filtrate was
clear yellow. A small amount of solid formed after the
bulk of the THF had been removed. These crystals were re-
moved by filtration. Removal of the remaining THF by evapor-
ation left a dark brown oil. About 30 ml hexane was added,
and beige crystals formed. The crystals were filtered and
dried. This product was purified by sublimation under

1 mm) at 92°. White crystals of product were

vacuum (10-
obtained. Yield :13 g. Melting Point = 72-77°.

Analysis Calcd for ZrClC12H150: Zr, 30.24; Cl, 11.75:
C, 47.77; H, 4.97.

Found: Zr, 30.09; C1, 11.87; C, 47.48; H, 4.78.

Diethoxobis(n—cyclopentadienyl)zirconium(IV). A
solution of ethyl alcohol (65.9 mmol) in 10 ml THF was added
to a solution of zirconocene dichloride (18.8 mmol) in 75 ml
THF. This mixture was allowed to reflux for one hour. A

solution of triethylamine (37.6 mmol) in 10 ml THF was added,

10

and refluxing was continued for an additional two hours.
The precipitate, (C2H5)3N-HC1, was separated by filtration.
Additional solid formed after most of the THF had been re-
moved from the filtrate. This solid was removed by filtra—
tion, and the remaining THF was removed from the filtrate
by evaporation. The resulting oil was sublimed under vacuum
(10—1 mm) at 92°. Proton magnetic resonance spectroscopy
showed the sublimation product was a mixture of
(C5H5)2Zr(OC2H5)2, (C5H5)2Zr(OC2H5)Cl, and Zr(OC2H5’)4. The
desired product (C5H5)2Zr(OC2H5)2 was isolated by resub-
liming it from this mixture at 55°. The crystalline pro-
duct was pale yellow. Yield 230.5 g. Melting point =
52-570.

Analysis Calcd for ZrC14H2002: Zr, 29.30; C, 54.00;
H, 6.42.

Found: Zr, 29.01; C, 52.06; H, 6.01.

Chloroisopropoxobis(U-cyclopentadienyl)zirconium(IV).
A solution of isopropyl alcohol (13.8 mmol) and triethyl-
amine (20.7 mmol) in 10 m1 THF was added to a solution of
zirconocene dichloride (13.8 mmol) in 60 ml THF. This re-
action mixture was allowed to reflux for two hours. The
insoluble (C2H5)3N-HC1 precipitate was removed by filtra-
tion. .The THF was then removed from the filtrate and a dark
brown liquid remained. Thirty ml of hexane was added to
this liquid, and a very small quantity of crystals formed.

These were removed by filtration and the filtrate was

11
treated by removing the hexane by evaporation. The result-

1 mm) at

ing dark brown oil was sublimed under vacuum (10-
86°. A large quantity of white crystals formed on the cold-
finger condenser. Yield :13 g. Melting Point = 81-82.5°.

Analysis Calcd for ZrClC13H17O: Zr, 28.88; Cl, 11.23;
C, 49.44; H, 5.38.

Found: Zr, 29.21; Cl, 11.21; C, 48.90; H, 5.07.

Diisopropoxobis(w-cyclopentadienyl)zirconium(IV). A
solution of isopropyl alcohol (45.1 mmol) and triethylamine
(33.9 mmol) in 10 ml THF was added to a solution of zir-
conocene dichloride (11.3 mmol) in 50 ml THF. This mixture
was allowed to reflux for two hours. The white precipitate,
(C3H5)3N-HC1, was removed by filtration. The bulk THF was
removed from the filtrate by evaporation. Twenty-five ml
of hexane was added, and the small quantity of resulting
crystals was separated by filtration. The hexane was re-
moved by evaporation and a yellow-brown solid remained.
Vacuum sublimation at 68° produced white crystals on the
cold-finger condenser. Yield 211.5 g. Melting Point - 110-
116.50.

Analysis Calcd for ZrC16H2402: Zr, 26.88; C, 56.62;

 

H, 7.07.
Found: Zr, 26.92; C, 56.29; H, 7.36.

Tetraethoxozirconium(lv). A solution of ethyl alcohol
(73.8 mmol) and triethylamine (37.0 mmol) in THF was added

to a solution of zirconocene dichloride (18.5 mmol) in THF.

12

This mixture was allowed to reflux for two hours. The pre-
cipitate, (C2H5)3N°HC1, was separated by filtratiOn. Subli-
mation under vacuum (10.1 mm) at 165° produced white crystals
of product on the cold-finger condenser. -Yield :11 g.
Melting Point = 172-180°. The physical data agreed with
those given by Bradley and Wardlaw 1° for Zr(OC2H5)4 pre-
pared by a different method.

.Analysis Calcd for ZrC8H2004: Zr, 33.62.

Found: Zr, 34.14.

Tetraisopropoxozirconium(Ty). This compound was pre-
pared in somewhat questionable purity by continued sublima-
tion of the residue from the (C5H5)ZZr(OC3H7)2 preparation
(see above). All of the (C5H5)22r(OC3H7)2 had sublimed at
90°. Further sublimation produced white crystals of pro-
duthr(0C3H-,)4. Yield z 0.5 g. .Melting Point = 180-200°
with decomposition. ‘The infrared spectrum agreed closely

with that reported by Bradley11 and Lynch.12
E. Hydrolysis Reactions

These reactions were simply performed by exposing the
compounds to air. This treatment was generally sufficient
to cause almost total hydrolysis to occur in about thirty
minutes. A more efficient method, ensuring that complete
hydrolysis had taken place, was to place the samples in air
saturated with water vapor. This was accomplished by fill-

ing the bottom section of a desiccator with water, placing

13
the samples (in open vials) on the plate, and closing the
desiccator lid. .The sample vials were periodically removed,
evacuated several times in a vacuum desiccator, and weighed

until further hydrolysis caused no further weight loss.
F. -Spectroscopic Measurements

Infrared Spectra. .The infrared spectra were deter—
mined by means of Nujol mulls on cesium iodide plates with
a Perkin-Elmer 457 grating spectrophotometer. All mulls
of the moisture—sensitive compounds were prepared in an

inert-atmosphere box.

Nuclear Magnetic Resonance Spectra. Nuclear magnetic
resonance spectra were recorded with a Varian A56/60D
Analytical NMR Spectrometer. The nmr tubes were filled
with sample in an inert-atmosphere box. All work was done

in deuterated benzene as the solvent.
G. Polaroggaphy

Attempts to obtain polarograms were made with a Sargent
Model XV Polarograph. Tetrahydrofuran was used as the sol-
vent, and tetrabutylammonium iodide (TBAI) was the support-
ing electrolyte. To obtain a 0.1§_solution of TBAI in THF
it was necessary to maintain a temperature of 55°. A drop-
ping mercury electrode was used as the indicating electrode.
The reference electrode was a saturated calomel electrode.

A three-compartment cell was used to ensure that no aqueous

solution would diffuse into the sample.

RESULTS AND DISCUSSION

Part One of this research is concerned with the prep-
aration and characterization of compounds of the forms
(U-C5H5)2Zr(OR)Cl and (U-C5H5)2Zr(OR)2 (R - CH3, C2H5,
17C3H7). The results and discussion of this work are con-
veniently presented under the following headings:

A. Preparations, B. Nuclear Magnetic Resonance Spectra,
C. Infrared Spectra, D. Hydrolysis Reactions, and E. Mis-

cellaneous Observations.
A. Preparations

The preparation of all these compounds was accomplished
by allowing zirconocene dichloride (CpZZrClz) to react with
the appropriate alcohol in the presence of triethylamine
[(C2H5)3N]. Tetrahydrofuran (THF) was used as the solvent.
If the alcohol was used as the solvent, the Zr(OR)4 com-
pound resulted. The ratio of CpZZrC12:ROH:(C2H5)3N was of
utmost importance if reasonable yields of product were to
be obtained. The presence of (C2H5)3N was necessary to
produce alkoxide ions and allow the reactions to proceed.

The reactions were as follows:

14

15

cp = 7Hgfins Cp2Zr(OR)Cl + (C2H5)3N°HC1

[A]
Cp2ZrC12 + ROH + (C2H5)3N

x y z
Cp2Zr(OR)2 + (C2H5)3N-HC1

[B]

The (C2H5)3N:HC1 is insoluble in THF whereas the products
[A] and [BL are appreciably soluble. The extent of the re-
action could thus be easily monitored by weighing the
filtered (c2H5)3N-HC1.

All of the compounds were isolated and purified by
sublimation under vacuum (10-1 mm) at a given temperature.
This temperature was critical. Too high a sublimation
temperature often caused cosublimation of undesirable pro-
duct. Table I shows the correct x : y : 2 ratios and the
proper sublimation temperature for each compound.

The absence of any indication of Cer(OR)3 should be
noted. The titanium analogs exist and have been the sub-
ject of extensive infrared3 and nuclear magnetic resonance2
studies. When the Cp2ZrC12 - C2H50H reaction (x : y : z =
1 : 4 : 2) is studied, the final product (sublimed at 165°)
shows the presence of three species. The species present
can be shown by NMR to be Cp2Zr(OC2H5)Cl, Cp2Zr(OC2H5)2,
and Zr(OC2H5)4. It appears that the two chloride ligands
are successively replaced by ethoxide ligands, but that the
cyclopentadienyl rings are displaced either simultaneously

or not at all.

16

Table I. Reactant Ratios and Sublimation Temperatures for
the Compounds Cp2Zr(OR)C1, Cp2Zr(OR)2, and Zr(OR)4
according to the reaction:

szer12 + ROH + (C2H5)3N = Product

x y 2 Compound

 

 

Subl Temp

Compound x y 2 (at 10-1 mm)

 

Class: Cp22r(OR)Cl

R = CH3 1 2 2 1030

R = C2H5 1 2 2 92

R = ifC3H7 1 1 1.5 86
Class: CpZZr(OR)2

R = CH3 1 4 2.5 105

R = c2H5 1 3.5 2 55

R = irC3H7 1 4 3.5 68
Class: Zr(OR)4

R = C2H5 1 >4 4 120

R ETC3H7 1 >4 4 120

 

17

B. Nuclear Magnetic Resonance Spectra

Nuclear magnetic resonance provided a convenient method
of characterizing the compounds and of checking for the
presence of impurities. The spectra were recorded in C6D6,
and the resonance peaks were sharp and well-defined. The
nmr spectra, along with important analytical results and
melting points, are shown in Figures 1—6. A summary of the
chemical shift data is shown ianable II.

The chemical shift results are especially noteworthy.
Nesmeyanov has studied the nmr of the series of compounds

CpTi(OCZH5)nCl (n = 0 - 3).2 His results show that the

a-n
chemical shifts of the ring, methyl, and methylene groups
shift to higher frequencies as 'n' increases. This trend
is explained by invoking electronegativity arguments. As
the ethoxide groups are replaced by the more electronega-
tive chloride ligands, the electron density is decreased
about the Cp ring and the methylene and methyl groups of
the ethoxide ligand. The result is a shift toward lower
frequencies in the nmr spectrum.

»Nesmeyanov's results are significant because the exact
opposite effect is seen with the zirconium compounds (see
Table II). As the number of alkoxide groups is decreased,
the chemical shifts of the ring and alkoxide resonances in-
crease. The electronegativity argument cannot be used to

explain this apparently anomalous behavior. The primary

differences between the Nesmeyanov work and the present

18

 

COMPLEX
C5H5\ 2 /Cl
r
/’/’ ‘\\
C5H5 0—CH3
NMR T=4,os
6.35
intensity IO 2.7
thoor IO 3.0
el.onoI. Zr Cl C H
colcd 31.69 12.32 45.89 4.52
found 31.51 12.06 45.35 4.l5

m.p. III-I146

Figure 1. Proton nmr of Cp2Zr(OCH3)Cl. Chemical shifts
in ppm .

Figure 2. Proton nmr of szzr(OCH3)2.

19

 

 

 

COMPLEX
C5H5\ Zr/O CH3
c511 / \O—CH3
NMR T=4.02
6.23
intensity 10 5,7
"100! 10 6.0
eI. anal. Zr C H

colcd 32.19 50.86 5.65
found 32.20 48.91 5.02

mp. 63-68’

Chemical shifts

in ppm.

Figure 3.

 

20

 

 

COMPLEX
r .
a’l’ .\\\
N M R If: 4.01
200
6.14
- _1Ma_ __
intensity-a 10 L9 2.8
theorn- 10 20 10
81. cool. Zr Cl C H
calcd 30.24 11.75 47.77 4.97
found 30.09 11.87 47.48 4.78
m.p. 72-77°

Proton nmr of Cp22r(OC2H5)Cl.

in ppm.

Chemical shifts

21

COMPLEX CSHS \ /0—CH2—CH3

Zr

 

 

NMR T=4-00
8.89
6.07

intensirym 10 3.8 6.4
thoor - -- 10 4.0 6.0
61. cool. Zr C H

1:0ch 29.30 54.00 6.42

found 2 9.01 5 2.03 6.01
m.p, 52-54'

Figure 4. Proton nmr of Cp2Zr(OC2H5)2.

in ppm.

Chemical shifts

22

 

COMPLEX
C5H5\z/Cl C
r H
3
9115/ \o—cH/
‘\
CH3
NMR T=4.05
'ROO
intensitym 10 5.9
rhooru' 10 60
61.61161. Zr Cl C H
colcd 28.88 11.23 49.44 5.38
found 29.21 11.21 48.90 5.07
m.p. 81-82.5'

Figure 5. Proton nmr of Cp2Zr(OC3H7)Cl. Chemical shifts
in ppm.

23

 

 

 

 

COMPLEX C5:5\zr/ 0"- C H— (C “3’2
NMR 1’=4.01
8.92
592 II
w - .
intansin 10 12
thaor no 12
e1. anal. Zr C H

calcd 26.88 56.62 7.07
found 26.92 56.29 7.36

m.p. 1105-115"

Figure 6. Proton nmr of CpZZr(OC3H7)2. Chemical shifts
in ppm.

Table II. Proton NMR

and Zr(OR)4

standard =

24

Chemical Shifts for Cp2Zr(OR)nC12_n

(solvent = C6D6; T-scale; internal
TMS).

 

 

Chemical Shifts

 

(Rpm)

Compound Ring Methyl Methylene
szer12 4.09 '-

Cp22r(OCH3)Cl 4.05 6.35

Cp2Zr(OCH3)2 4.02 6.23

Cp2Zr(OC2H5)Cl 4.01 9.00 6.14
szzr(OC2H5)2 4.00 8.89 6.07
Zr(0C2H5)4 - 8.64 5.73
Cp22r(OC3H7)Cl 4.05 9.00

Cp22r(OC3H7)2 4.01 8.92

Zr(OC3H7)4 - 8.65

 

25
zirconium work are (1) the central metal, and (2) the num-
ber of cyclopentadienyl rings bonded to the central metal.

Wales13 has published the nmr work of the only other
relevant series,.Cp2Ti(SR)nC12_n (n = 0,1,2). Because
this series contains two Cp rings, the analogy to the zir-
conium series may be more valid than the Nesmeyanov series.
However, Wales' results show the same trends in chemical
shifts as the Nesmeyanov compounds show. The presence of
thio—alkoxide groups rather than alkoxide groups makes this
comparison questionable.

Until additional work is done with the zirconium series,
it is futile to attempt an explanation of the anomalous
chemical shift observations. The reason could perhaps be
the result of using zirconium rather than titanium. The

series of compounds szTi(OR)nCl and Cer(OR)nCl

2-n 3-n’

neither of which has been characterized, should clarify this
matter. The anomaly could also be a solvent effect. The
Nesmeyanov compounds were studied in THF, the Wales compounds

in CDCl3, and the zirconium compounds in C6D6.
C. Infrared Spectra

The infrared spectra of these compounds also provide
an important characterization technique. Coordinated cyclo-
pentadienyl rings and the alkoxide groups give character-
istic infrared absorption bands in the 3500 - 700 cm"1 re-

gion. In addition, valuable information about the

26
Zr - ring and Zr - OR bonds can be obtained by a study of
the far infrared (700 - 250 cm_1) region.

The infrared spectra of the 3500 - 700 cm-1 region for
the compounds studied are presented in Figures 7 — 15. The
bands characteristic of the vibrations of the cyclopenta—
dineyl rings (3100, 1020-1010, 850-840-815) are all present

in the series of compounds Cp2Zr(OR)nCl rThere is a

2-n'
slight shift towards lower frequencies of the 1020-1010

and 850-840-815 bands as 'n' increases. This may well re-
flect the increased electron density on the ring as the
alkoxide groups replace the chlorides.

The bands characteristic of the OR groups become much
more intense (relative to the Cp bands) as "n" increases.
The band assignments, based on standard spectra for the
pure alcohol, are presented in Table III. The CH3-bending
modes which are not shown are overlapped by the Nujol
bands in the 1400 - 1500 cm"1 region.

A study of Table IIIshows that insignificant shifts
occur in the OR bands as 'n' goes from 1 to 2. The unas-
signed bands in the 730 - 770 cm"1 region, however, undergo
significant shifts to lower frequencies as 'n' goes from
.1 to 2. These bands are absent in the spectra of the pure
'alcohols and the Cp2ZrClz. It is reasonable, therefore,
to assume that they arise from the Zr - 0 bond. The pres-
ence of a chloride ligand should decrease the electron
density about the central metal, thus allowing a stronger

Zr - 0 bond to be formed. This is consistent with the

27

aauomwsuou

oon

 

.NHUHNNQU mo Eduuommm CoumnmcH .h whomflm

ATE: xucoaoo;
oom coo. oo~_ oov. oonWI ooum OOOn
a q _ _ 47 q d

 

 

 

 

 

 

28

95061411115001;

oon

 

00m

 

 

 

ooo—
_

.HoAnmUoVuNumo mo Eduuoomm CommuwcH

ATE: xucoaoo:
CON. oov. OOMW oonu
_ _ Lfl. AI

 

 

 

 

.m ousmflm

coon

 

 

 

29

aouomwsuou

 

.mAmmUOVHNmmU mo Eonuommm UmHmHMCH .m musmflm

ATE: Sconce:
oom 00m 000. cow. oov. oonw ooum ooon
_ _ _ 7 .7 a H

 

 

 

 

 

 

 

30

83001111115001;

oon

 

oom

 

 

coo—
_

.HoflmmmooVHNnmo wo Eduuoomm COHCHMCH .OH ousmflm

ATEOV Sconce;
oo~_ oov. oonw Conn
_ .I 47 a

 

 

 

 

 

 

1

ooon
fl.

31

85004111115001;

oon

 

oom

 

ooo.
_

.mflmmmuovnmmmo mo Eouuoomm UmnmumcH .HH ousmflm

A—IEUV xucoaao;
CON. oov— oonI— 00km
_ _ A_I E

ooon

 

32

33001410150011

oon

 

00m

000—
_

.vAmmmUovHN mo Esnuoomm COHCHMGH

:IEUV xucoanuo;
OON— OOV— OOMW CONN
_ _ .1 a

 

 

.Na onsmflm

oooM

 

 

 

 

33

asunIIstuou

 

.Huflnmmoovumumo mo Eouuoomm UmumumcH .mH ousmflm
ATEOV 88:06:
com com ooo— ooNp oov— oonI— OOKN coon
— _ _ . :I q 7

 

 

 

 

 

 

 

34

aauougwsuou

com 00m 000—
IIfi q

2

 

 

 

 

 

 

 

 

.mflnmmoovuwmmo mo Eduuoomm CommumcH .vH ouomflm

ATE“; xucoaooi
com. 00V. comp CORN
_ _ d7 7

 

 

 

 

 

 

ooom

 

 

35

aauomwsuou

oom

 

oom

coo—
_

 

 

 

 

 

.vflnmmoovum mo Esnuoomm CoumnmcH

:IEUV xucoaoui
OON— OOV— comm CONN
_ _ 41 a

 

 

 

 

 

.mH onsmflm

coon

 

 

 

 

36

 

 

 

 

owe cmfifiumoflfl mmmfl ooom «Asmaoovuuado
non AAHHIHoHA omms comm Hofismeoovuuado
«me «He umfifluoefls osos «Asmaoovuuado
was one «mnfiuomfifi whoa Hoflamaoovuusdo
can oNHH comm «Aemoovuusoo
«on cuss nowm Hoflemoovuuaao
AmVAOIHNV nouonum comm nouonum ocmm nuuouum
oocmsnnmco Ono 0.0 010 Tue. mno oaaomsoo
.cIAHOEAmoVuNamo

CA mmsono moflxoxad mo AH

180v moaocosooum oflumfluouomnmno .HHH manna

37
observed shift to higher frequency of the Zr - 0 bond as-
signment as chloride replaces alkoxide in each case. Since
the Zr - O stretching frequencies are found at much lower
frequencies, the frequency in the 730 - 770 cm.1 region is
possibly a bending mode.

The far-infrared spectra (700 - 250 cm-1) of the com-
pounds of interest are shown in Figures 16-18. The import-
ant features of these spectra are the assignments and
observed shifts in the Zr - O and the Zr - ring stretching
frequencies. These frequencies are summarized in Table IV.
The assignments were made by comparing the spectra of
Cp,Zr(OR)nc12_n with the spectrum of CpZZrClz.

From Table IV it is seen that the Zr - O and the
Zr - ring frequencies are shifted to lower frequencies as
the chlorides are successively replaced by alkoxides. This
is once again consistent with the electron density argument
expressed above. The Zr - O and Zr - ring bonds are
equally affected by the presence or absence of chloride

ligands.

38

Figure 16. Far-infared spectra of Cp2Zr(OCH3)nC12_n

(n = 0-2).

(C5H5)2 Zr (OCHg) 2

 

 

 

 

 

 

 

 

 

 

 

 

(C5H5)2Zr(OCH3)k W
I
I
fwt:_‘v‘~ "-VA‘VA' ‘I
(C51'15I'21r C12 I1
L 1 A 1 L 1 L 3— L
700 600 500 400 250
freq uancy (cm‘ ')

transmittance

39

Figure 17. Far—infrared Spectra of szzr(OC2H5)nCl2
(n = 0—2) and Zr(0C2H5)4.

—n

Zr(QC2H5)4

W

(C5H5)2 Z r(OC2H5)2

 

(C5 H5)2 Zr(OC2 H5) C1

 

(C5H5)2Zr 02 I

 

 

L_ LL i 1 L A J L A

transmittance

 

700 600 500 400 250

frequency (cm")

Figure 18. Far-infrared spectra of Cp2Zr(OC3H7)nC1

40

(n = 0-2) and Zr(OC3H7)4.

   
 
 

-n

 

 

(C5H5)2Zr( C
(C5H512zr(
l l I 1 I L A L J
700 600 500 400 250

frequency (cm-1)

transmittance

41

Table IV. Far-infared Frequencies (cm-1) of Zr-O and
Zr-ring Bonds. .

 

 

Compound

Zr-O Stretch

Zr-ring Stretch

 

CpZZrClz
Cp22r(OCH3)Cl
Cp2Zr(OCH3)2
Cp22r(OC2H5)Cl
szzr(0C2H5)2
Zr(OC2H5)4
Cp22r(OC3H7)Cl
Cp2Zr(OC3H7)2

Zr(OC3H7)4

507
482
532
521
(520,472,421)
572
560 (453,435)
(570,516,463)

359
342
329
338
318

344
315

 

42

D. Hydrolysis Reactions

It has been previously noted that the compounds
Cp22r(OR)Cl and Cp2Zr(OR)2 are extremely sensitive to
moisture. If water vapor is allowed to come in contact
with the compounds, the compounds are quickly destroyed.
This section will deal with the products of these hydrol-
ysis reactions.

If the compounds are exposed to air saturated with
water vapor, they lose weight rapidly. No physical change
can be visibly observed, but the melting points change
from the relatively low values reported earlier to values
in the range of 300° or greater. The infrared spectra
undergo marked changes.

-The infrared spectra and melting points indicate that
the end hydrolysis product for each of the three compounds
szzr(OR)C1 (R = CH3, C2H5, ifC3H7) is the same. Further-

more, the weight loss corresponds to the hydrolysis reaction

 

Cp\ /Cp
2Cp2Zr(OR)Cl + HOH > Cl-Zr-O-Zr:Cl + 2ROH
./

This is the same noted by Brainina4 when he performed an
elaborate hydrolysis reaction with szzr(OC2H5)Cl.

The infrared spectrum of the hydrolysis products of
Cp3Zr(OR)Cl is shown in Figure 19. The spectrum (the same
regardless of 'R') is consistent with the above reaction.

The absorption bands characteristic of the cyclopentadienyl

43

eouougwsuou

CON

COO

.Honmoqummo mo uosoonm mflmwaouohz mo Eduuoomm COHOHCCH

ATE: xucoaoo...
COO. CON. OOV. COO. CONN

.OH musmflm

coon

 

 

 

 

 

 

 

 

 

 

 

_ q E .Eci _.

 

 

 

 

 

 

H

 

 

 

44
rings (3100, 1030-1020, 840-810 cm-l) are still present.
All bands characteristic of the R-groups are absent. There
are additional strong bands at 779 and 752 cm.1 which may
be additional cyclopentadienyl bands. The far-infrared
region (see Figure 20) shows the Zr-ring stretching band
at 349 cm.1 and the lower frequency broader bands character-
istic of a Zr-Cl bond.

The hydrolysis of the compounds Cp2Zr(OR)2 [B] is
much different from and less straightforward than the
hydrolysis of Cp2Zr(OR)Cl [A]. The weight loss (relative
to weight of original product) attending the hydrolysis of
[B] is much greater than the weight loss for [A]. The
hydrolysis of [B] seems to proceed in two rather ill-de-
fined steps. On the basis of infrared spectral evidence
and weight loss calculations, the two steps can be ex-

pressed as follows:

 

 

Cp Cp
(1) n Cpgzr(OR)2 + n HOH > Zr O-Zr—O + n ROH
('29 GP
n-I
[Cl
OH OH
(2) [C] +nHOH > Zr -Zr-O + nHCp
OH OH
n-1

ID]

The above steps are undoubtedly an oversimplification. If

the hydrolysis is allowed to proceed for about 30 minutes,

45

 

 

 

 

 

A
0
u
C
2
'E
Q
C
2

B O-

L_ 1 I 1 L l 41 J I
700 600 500 400 250

frequency Icm")

Figure 20. Far-infared spectra of hydrolysis products of

(A) Cp2Zr(OR)Cl : (B) Cp22r(OR)2.

46
the weight loss corresponds roughly to the weight lost by
elimination of alkoxide groups in step (1). The infrared
spectrum shows that all bands characteristic of the R-groups
are absent. The spectrum shows, however, that the cyclo-
pentadienyl absorption bands are superimposed on the broad
band characteristic of the Zr-OH band (very broad centering
at 480 cm-l). See Figure 20. This would indicate that
step (2) has already begun. Further hydrolysis results in
a slower weight loss until a constant weight is finally
observed. This weight loss corresponds roughly to the
elimination of HCp as in step (2). The distinctive odor
of cyclopentadiene is observed during the latter stage of
hydrolysis. The infrared Spectrum of [D] shows no indica-
tion of the presence of Cp. Broad bands are observed at
3300, 1480, and 480 cm-1. The former two are indicative

of bound OH while the latter is probably due to the Zr-OH

bond.
E. Miscellaneous Observations

Reduction Potentials. It was hOped that reduction po-
tentials for each of the compounds could be determined.
This information would be particularly useful in relation
to Part Two of this work. .The polarograms were attempted
in THF solution. The resistance of THF, however, could not
be sufficiently overcome by the presence of supporting elec-
trolyte (TBAI) to obtain valid polarograms. This work is

being continued.

47
Melting Points. .Each of the compounds gave clear
melts with no apparent decomposition. Brainina4 reported
that his preparation of Cp22r(OC2H5)Cl melted with decomposi-
tion. It is possible that his compound was not sufficiently
purified. Small quantities of hydrolysis product cause the
melt to become brown. This gives the appearance of decompo-

sition.

PART TWO

NITROGEN FIXATION UNDER MILD CONDITIONS BY USING SOME

ORGANOMETALLIC COMPOUNDS OF THE SECOND-ROW TRANSITION
ELEMENTS AS THE ACTIVATING AGENTS

48

INTRODUCTION

The process by which nature converts molecular nitro-
gen to ammonia and other more useful compounds has long
been a fascinating puzzle to the chemist and the biologist.
Nitrogen can be industrially converted to ammonia only
under severe conditions of temperature and pressure.14
Micro-organisms, however, have the ability to "fix" molecu-
lar nitrogen under the mild conditions of ambient tempera-
ture and pressure. These natural systems inevitably con-
tain iron and molybdenum components, and for this reason
inorganic chemists have long tried to duplicate the nitrogen-
fixing process.

The first major breakthrough in molecular nitrogen
activation under mild conditions by a transition metal was
made by Vol'pin in 1966.15 By allowing one of several
Ti(IV) compounds to react with an alkyl magnesium halide
in the presence of molecular nitrogen, Vol'pin found that
subsequent hydrolysis produced detectable quantities of
ammonia. Since then this syStem has acted as the model for
most attempts to "fix" molecular nitrogen by inorganic or

organometallic complexes.

49

50

The mechanistic concept by which Vol'pin's system
"fixes" molecular nitrogen remains to be elucidated.
Brintzinger has suggested that the monomeric dihydride
anion [(v-C5H5)2TiH2]- may be an intermediate.1°v17 The
focus of this topic has shifted away from Vol'pin's system,
however, and little mechanistic work seems to be in process.

.Much work in the last two years has been done by van
Tamelen and co-workers.13'34 Van Tamelen has worked with
a variety of titanium(IV) systems and has achieved a sig-
nificant amount of success. He asserts that a Ti(II)
species is the activating intermediate in practically all
of his systems. A strong reducing agent, either chemical
or electrical, is employed to give the two-electron reduc-
tion. In this reduced state, the Ti(II) species perhaps
"complexes" a molecule of molecular nitrogen. Van Tamelen
suggests that the NeN bond is cleaved by further reduction
at this stage, and the result is a nitride-type structure.
Subsequent hydrolysis of this species liberates ammonia and
possibly regenerates the initial (Ti(IV) species. If van
Tamelen's suggested mechanism is correct (see Figure 21), a
true catalytic process is possible. Yields of greater than
100% (based on moles of ammonia produced per mole of initial
titanium compound) have been realized by van Tamelen's pro-
cedure, but a true catalytic process is yet to be shown.

An alternative mechanism for the van Tamelen-type pro-

cess has been proposed by Henrici-Olivé and Olivé.25 The

51

.coflumxflw COCOHDHQ How Emflcmzomfi Comomoum m.cmHOEmB Cm>

-56 + MUr21." .mIromo + -66 + N2

mot._zi._ A -3. \

1046
620%.. AK
mIzm

4

50:.

\

 

...dz +624
>

OZV

 

Oco_of;doc

OZN

.fim onsmflm

"mmouoca .oc __ocO>O

a mzmaoé

\

7

(If

 

5+9:

 

 

I.QZ +07— N

.3
< \

Dr 620%

52
Olivés' mechanism, unlike van Tamelen's, involves a hydrogen-
bridged Ti(III) dimer as the activating intermediate. Van
Tamelen has recently shown that this is highly unlikely.24
Regardless of which mechanism one chooses to believe,
it is nevertheless clear that somehow the molecular nitrogen
has been chemically activated. No actual molecular nitrogen
complex of the early transition metals has been isolated.
vThe nature of the metal-dinitrogen bond is thus not known.
It is quite probable, however, that an olefinic-type of U-
complex is formed. This type of complexation is a common
phenomenon in complex catalysis. .The usual result of such
a U-complex formation is the activation of the multiple bond.
The activation is a result of the donation of the U-elec-
trons of the multiple bond to the empty d-orbitals of the
transition metal, which causes a partial rupture of the
multiple bond and imparts a slight positive charge to the
atoms comprising the multiple bond. These effects produce
a weakening of the multiple bond and permit further reaction
to occur. It is quite probable that bonding of this type
is somehow involved in the nitrogen fixation-reduction cycle.
A large number of stable molecular nitrogen complexes
have been prepared and characterized, but these complexes
invariably involve a transition metal containing six to
eight d—electrons. .The most interesting from both a histori-
cal and theoretical perspective are [Ru(NH3)5N2]C12,26
IrCl(Ph3P)2N2f,27 HCo(Ph3P)3N338, and [RU(NH3)5]2N2-29 The

important feature of all these complexes is that all

53
attempts at further activation of the bound molecular nitro-
gen have proven futile. In each case the metal-dinitrogen
bond is nearly linear. The bonding is analogous to the
metal-carbonyl bond. Back r-bonding from the filled d-elec-
trons of the transition metal to the empty antibonding p-
orbitals of the molecular nitrogen is thought to be the
stabilizing factor. Because the electrons involved in the
multiple NEN bond are not significantly affected, it is not
difficult to understand why further activation is improb-
able. This is quite different from what would be eXpected
if the bonding were of the olefinic type as described pre-
viously. The stable molecular nitrogen complexes are fur-
ther hindered as possible intermediates in the nitrogen-
fixing cycle if one accepts van Tamelen's suggestion that
a two-electron reduction is a necessary first step. In
each case this would mean forcing electrons into the next
higher molecular orbital, but the next higher molecular
orbital in these molecules is an antibonding orbital of high
energy.

Because the titanocene dichloride system works well in
the van Tamelen experiments, perhaps the biscyclopentadienyl-
halide and the mixed cyclopentadienyl-alkoxide compounds of
the second-row transition metals also have the ability to
fix nitrogen. Part Two of this research project deals with
a survey of these organometallic compounds and their efficacy

in the fixation-reduction cycle.

EXPERIMENTAL

A. Materials

Bis(U-cyclopentadienyl)dichlorozirconium(ij was ob-
tained from Arapahoe Chemical Company and used without
further purification.

Anal. Calcd for C10H10ZrC12: Zr, 31.22; Cl, 24.26.

Found: Zr, 31.09; Cl, 24.13.

Niobium pentachloride was obtained from Research In-

organic Chemicals and used without further purification.

.Molybdenum pentachloride was obtained from Climax
Molybdenum Company. The commercial product was further
purified by vacuum sublimation. The green oxychloride im-
purities were removed by gradually increasing the tempera-
ture until no more green substance sublimed. The shiny
black crystals of pure MoC15 were then sublimed and stored
under dry nitrogen. This process was accomplished by the

use of a three-component Pyrex tube and a tube furnace.

Ruthenium trichloride was purchased from Alfa Inor-

ganics and used without further purification.

54

55
Sodium was used in the form of finely dispersed sodium
prepared according to the method of T. P. Whaley.3° The
sodium was stored under sodium-dried xylene in a dry nitro-
gen atmosphere. Transfers of sodium were effected by use
of a pipet. The xylene was removed under vacuum before the

sodium was weighed and used.

-Cyclopentadiene was freshly prepared by depolymeriza-
tion of dicyclopentadiene by heating the dicyclopentadiene
until the depolymerization occurred. .The cyclopentadiene
was fractionally distilled into a flask and stored until

used at -78°.

[W(OCH3)2C13]J2f and [W(092TH5)2C12(C2H50H)]2 were obtained

 

from D. Paul Rillema.31

Solvents. All solvents were refluxed with the appropri-
ate drying agent under a nitrogen atmosphere unless indi-
cated otherwise.

Tetrahydrofuran (THF) was refluxed continuously with
lithium aluminum hydride and distilled immediately before
use.

Benzene was refluxed continuously with calcium hydride
and distilled immediately before use.

Hexane was continuously refluxed in the presence of
sodium metal and distilled immediately before use.

Chloroform was washed with water and dried over calcium

chloride. The chloroform was then refluxed in the presence

56
of phosphorus pentoxide and distilled. The distilled

chloroform was stored in the absence of light.

gaggg. Prepurified nitrogen from Liquid Carbonics was
passed through an activated copper catalyst (BTS catalyst -
BASF R-3—11 from Badische Anilin - Soda-Fabrik AG) to re-
move oxygen. It was then passed through two drying towers
filled with calcium sulfate and barium oxide to remove any
water present.

Argon was obtained from Liquid Carbonics and used

without further purification.
B. Analytical Methods

Niobium Analysis. The solid sample was decomposed by
dissolution in very dilute nitric acid and digesting the
solution over a steam bath for about an hour. This solution
was made basic by addition of aqueous ammonia, and the di-
gestion was continued for an additional two hours. Dilute
nitric acid was added until the solution was acidic. The
precipitate was removed by filtration and washed three times
with dilute nitric acid. This precipitate was ignited for
several hours at 700°, and the niobium was weighed as Nb205.
The filtrate was saved and used for determination of chlor-

ide content.

Molybdenum Analysis. The samples were decomposed by
dissolution in BE aqueous ammonia. This solution was acidi-

fied with BN_nitric acid and heated to the boiling point.

57
The solution was then diluted to 100.00 ml. Twenty-five-ml
aliquots were analyzed for molybdenum according to the

method of Pribil and Malat.32

Ammonia Analysis. Ammonia was detected qualitatively
by the use of Nessler's Reagent. Nessler's Reagent was pre-
pared as follows: (1) Fifty grams of potassium iodide were
dissolved in 50 ml cold water; (2) A saturated solution of
mercuric chloride (about 22 g in 350 ml of water) was added
until an excess was indicated by the formation of a precipi-
tate; (3) Two-hundred ml of 5N_sodium hydroxide was added,
and the solution was diluted to one liter.

Ammonia was quantitatively determined by being passed
through standard hydrochloric acid. The remaining hydro-
chloric acid was then titrated with standard sodium hydroxide
solution, and the ammonia was determined from the amount of
acid which had been neutralized. Phenolphthalein was used

as the indicator.
C. Experimental Apparatus and Technique

Most of the materials dealt with in this research
were sensitive to oxidation and/or hydrolysis. Similar

techniques to those explained in Part One were used.
D. The Nitrogen:§ixation Experiment

The general procedure followed in the nitrogen-

fixation experiments was essentially the same for each

58
compound studied. The apparatus used and a simplified pro-
cedure are illustrated in Figure 22. .A more detailed ex-
perimental description follows:

1. The reducing solution of sodium naphthalenide was
prepared by allowing stoichiometric (1:1) quanti-
ties of finely dispersed sodium to react with re-
agent naphthalene in THF. A deep emerald-green
solution resulted.

2. The appropriate transition-metal complex was dis-
solved in THF. The quantity of complex was stoi-
chiometrically one-sixth that of the sodium naph-
thalenide. This solution of complex in THF was
added dropwise to the reducing solution. A con-
stant stream of nitrogen was bubbled through the
reaction mixture and directed through Nessler's
Reagent, which was used for qualitative detection
of ammonia in the escaping gas.

3. A slight excess of ethanol, diluted with THF, was
added dropwise to the mixture of complex and re-
ducing agent. If fixation-reduction occurred, the
mixture remained dark; ammonia was very quickly
observed in the Nessler's Reagent. If no fixa-
tion-reduction occurred, the alcohol reacted with
the sodium naphthalenide, the solution became clear,
and no ammonia was detected in the Nessler's Re-

agent.

.DCOEHHOQMO :oflumxflm
Isomouuflc map How onsooooum one msumummmm Hmucwefluomxm

 

 

 

.NN musmflm
.9.
6.3.302 K W“;
f
_/\ .w 7.5.0.5

5.0. to. .U... a. all polo;
36:3 pea .¢.Z a_> ~12 to. .8. n

IIIIIII)

 

 

 

.6: R r

9

.92 .a 5:296 to. .2... n

 

at... a. on... .xoiEoO 6.0.50.5»? tt< u

c...» c.1mzo.U+OZ .0 £3 Octave. atone... .

”mznamUOxm

60
4. If fixation-reduction occurred, the escaping nitro-
gen stream was passed through standard hydrochloric
acid solution for 24 hours. Ammonia was determined

as previously described under Analytical Methods.

 

B. Preparation of Compounds_

Trichlorobis(1-cyclopentadienyl)niobium(V). A combina-

 

tion and modification of the methods of Birmingham33 and
Brantley34 were used. A solution of sodium cyclopentadi-
enide in THF was prepared by adding 8.1 ml freshly distilled
cyclopentadiene to 2.54 g of finely dispersed sodium. This
solution was added dropwise to a suspension of 10,9 Nbcl5
in 50 ml dry benzene. Vigorous mechanical stirring and
strict exclusion of air were required. The reaction con-
tents were vigorously stirred for several hours. The sol-
vents were removed under vacuum. A brownish—black dry solid
resulted. This solid was ground to a powder. The powder
was placed in a continuous extraction apparatus and ex-
tracted with benzene. The extracted product was recrystal-
lized from benzene. The product crystals were deep purple.
Yield = 4.3 g. The infrared spectrum agreed closely with
that reported by Birmingham.33

Analysis Calcd for NbC13C103103 Nb, 28.21; C1, 32.29.

Found: Nb, 28.05; Cl, 33.16.

Dichlorobis(U-cyclopentadienyl)molybdenum IV .35135.37

Freshly sublimed molybdenum pentachloride (0.05 mol) was

61
added very slowly to an ice-cooled solution of sodium cyclo-
pentadienide (0.25 mol) and sodium borohydride (0.13 mol) in
125 ml THF. This mixture was stirred and allowed to reflux
for several hours. The solvent was removed by evaporation,

1mm)

and the residue was sublimed under vacuum (about 10-
at 120°. Bright yellow crystals of (C5H5)M0H2 were obtained.
These crystals were extremely sensitive to air, and extreme
precautions had to be taken. Exposure to air caused the
crystals to turn brown immediately. .The (C5H5)2M0H2 pro-
duct was added to 50 ml anhydrous chlorofOrm, and a dark
green solution resulted. Green crystals of (C5H5)2M0C12
crystallized from this solution. The physical properties

of the product agreed with those given by Cotton and Wilk-

inson.35

Biscyclopentadienylruthenium II .35 Ruthenium trichlor-
ide (0.01 mol) was added gradually to a solution of sodium
cyclopentadienide (0.06 mol) in THF at -80°. The contents
of the flask were warmed to room temperature. As the tem-
perature was raised the contents became purple. This reac-
tion mixture was allowed to reflux for two hours. The sol-
vent was removed by evaporation and the remaining solid
material was dried. This violet-purple solid was ground to
a fine powder and sublimed under vacuum (10-1 mm) at 120°.
The resulting product crystals of (C5H5)2Ru were light-

yellow.

RESULTS AND DISCUSSION

Part Two of this research is concerned with a general
survey of the possible nitrogen-fixing ability of some early
transition metal organometallic compounds. .It is not an
in-depth study, but is rather an overview with the purpose
of pointing out some of the possibilities for future study.
The emphasis is on the biscyclopentadienyl compounds of the
second-row transition metals. This work will be discussed
by each group metal. The discussion will thus be divided‘
into the following headings: A. General Observations, B.
Compounds of Zirconium, C. Compounds of Niobium, D. Com-

pounds of Molybdenum and Tungsten, and E. .Ruthenocene.
A. General Observations

Except as otherwise noted, the nitrogen-fixation ex-
periment was performed according to the directions prescribed
in the Experimental section (see Figure 22). The important
aspect of this experiment is to see whether a given transi-
tion metal compound is capable of activating molecular
nitrogen. The proposed mechanism for the activation has
already been reviewed in the Introduction (see Figure 21).

The results are summarized in Table V.

62

Table V. A Summary of the Results Obtained by Using Various
Transition Metal Compounds in the Nitrogen-Fixa-
tion Experiment.

'Compound' + Na+C10H8'_ + N2 big? NH3

 

 

 

Compound %NH3 Comments

CpZZrClz O NaBH4 usgd rather than
Na C10H8'

Cp2ZrC12 2.7.4.1

Cp2Zr(OC3H7)Cl 1.3

CPazIIOC3H7)2 0

CpngCl3 2.6

NbCl5 0 NbCl5 reacts with THF at 25°

CngoClz ? Solubility problems not
overcome

CngoH2 pb

W(OCH3)2C13 0.3

W(OC2H5)2C12(C2H50H) P

szRu 0

 

aBased on moles NH3:

moles of compound.

bPositive qualitative test but not measurable quantitatively.

64

It was necessary to test the validity of the experiment.
Spurious conclusions could be drawn if ammonia were already
present in the gaseous nitrogen or if some substance present
in the compound or reducing agent were to react with Nessler's
Reagent. The following tests show that the experiment is
indeed valid: (1) If nitorgen gas was bubbled through the
reducing agent and the escaping gas was directed through
Nessler's Reagent, no ammonia was detected. There is no
ammonia in the nitrogen. (2) If the experiment was con-
ducted as outlined but without the addition of transition
metal complex, no ammonia was detected. The transition
metal complex is a necessary constituent in the experiment.
(3) If a transition metal compound was found to have the
ability to fix nitrogen, the experiment was repeated but
argon was substituted for nitrogen. No ammonia was detected
when the argon was used. .Molecular nitrogen was the source
of the nitrogen in the ammonia. (4) If no reducing agent
was used, no ammonia was detected. Reduction of the transi-
tion metal complex is an important step in the mechanism.
(5) No ammonia was detected until the final hydrolysis
and thus the source of the NH3 protons is probably the

alcohol used for the hydrolysis.
B. Compounds of Zirconium

Only for zirconium was a systematic study of related
compounds possible. -The compounds which were used were

those synthesized and characterized according to Part One

65
of this research. The only compounds available in large
enough quantities to permit a valid study were Cp2ZrC12,
Cp22r(0C3H7)Cl, and Cp22r(OC3H7)2. The results of the
nitrogen-fixation experiment for these three compounds are

shown below:

 

Compound Eggs
CpZZrC12 2.7, 4.1
Cp22r(OC3H7)Cl 1.3
CpZZr(OC3H7)2 0.0

The yields of ammonia produced are disappointingly
small, and any conclusions must be drawn with considerable
uncertainty. However, it does appear that the nitrogen-
fixing ability of the series decreases as the chloride
ligands are replaced by isopropoxide groups. One can only
speculate about the effect being related to the ease of the
two-electron reduction of Zr(IV)-Zr(II). Watt and Drummond38
have shown that sodium naphthalenide is capable of reducing
szer12 to CpZZr. Unfortunately the actual reduction
potentials are not known. Attempts to obtain these reduc-
tion potentials polarographically (see Part One, Results
and Discussion, Section E) were unsuccessful.

The speculated'ease-of-reduction" series Cp2ZrC12 >
Cp22r(OR)C1 > Cp22r(OR)2 may be a result of the relative
electron density about the zirconium. As the chlorides
are replaced by the more basic alkoxide groups, the electron

density about the zirconium should increase and may have

66
the effect of raising the energy of the lowest-lying d-or-
bital slightly. Such an effect would make reduction more
difficult. This would be consistent with the trend seen in

the relative nitrogen-fixing abilities of the compounds.

C. Compounds of Niobium

 

No niobium series similar to that discussed with zir-
conium was available. The two compounds studied were
szNbCl3 and NbCls. When szNbCla was used in the experi-
ment, yields of 2.6% NH3 were realized. The NbC15 gave no
indication that NH3 had been produced. The yield of NH3
produced by szNbCl3 was too close to the yield given by
Cpaerlz to allow a significant comparison.39

During the course of the szNbCl3 experiment, several
of the possible variables were investigated for their ef-
fect on the nitrogen-fixation process. After the ammonia-
producing process had been proceeding for several hours,
the nitrogen flow was replaced by an argon flow. Ammonia
production ceased immediately. When the nitrogen flow was
once again begun, more ammonia was detected. The addition
of more sodium after the reaction had proceeded for a day
caused slightly more ammonia to be produced. Adding more

szNbCl3 had no effect.
D. Compounds of Molybdenum andgungsten

The compounds of molybdenum used were szMoClz and

szMon. The CngoClz was difficult to study because of

67
its insolubility in THF and most other organic solvents.
This insolubility necessitated a vigorously-stirred solid-
(CpaMoClz)-liquid(NaNp)-gas(N2) reaction, and the reported
results are probably not a valid indication of the ability
of Cp2MoCl2 to fix nitrogen.
The tungsten compounds used were the dimers

[W(OCH3)2C13]2 and [w(oc2H5)2c1(c2H50H)]2. Each produced

small quantities of NH3.

E. Ruthenocene (CpaRu)

 

It was thought that CpgRu would be most significant if
it could fix molecular nitrogen. Isolated molecular nitro-
gen complexes of ruthenium have been isolated,26 and it
would be significant to show that other compounds of the
same metal could work in the nitrogen-fixation experiment.
It was also hoped that szRu would be an ideal compound to
try since the ruthenium is already in a reduced state, but
szRu gave only negative results in the nitrogen-fixation

experiment.

BIBLIOGRAPHY

10.

11.

12.

BIBLIOGRAPHY

A. N. Nesmeyanov, O. V. Nogina, and A. M. Berlin, Dokl.
Akad. Nauk. SSSR, 134, 607 (1960).

A. N. Nesmeyanov, E. I. Fedin, P. V. Petrovskii, V. A.
Dubovitskii, O. V. Nogina, and N. A. Lazareva, Dokl.
Akad. Nauk. SSSR, lg; (3), 659 (1965). ""

G. G. Dvoryantseva, Y. N. Sheinker, A. N. Nesmeyanov,
O. V. Nogina and V. A. Dubovitskii, Dokl.Akad. Nauk.
SSSR, 161 (3), 603 (1965).

E. M. Brainina, R. Kh. Freidlina, and A. N. Nesmeyanov,
Dokl. Akad. Nauk.ߤSSR, 154, 1113 (1964).

P. C. Wales and H. Weigold, J. Organometal. Chem., 24,
413 (1970).

C. K. Mann in "Electroanalytical Chemistry," Vol. 3,
A. J. Bard, Ed., Marcel Dekker, Inc., New York, N. Y.,
1969, p. 132.

H. Lund and J. Bjerrum, Ber. Deut. Chem. Gesell., 64”
210 (1931).

H. F. Walton, "Principles and Methods of Chemical
Analysis," Prentice Hall, Inc., Englewood Cliffs, New
Jersey, 1964, p. 236.

S. Herzog, J. Dehnert, and K. Luhder in "Techniques of
Inorganic Chemistry," Vol. III, H. B. Jonassen and A.
Weissberger, Eds., Interscience Publishers, New York,
N.Y., 1968.

D. C. Bradley and W. Wardlaw, J. Chem. Soc., 1951, 280.

D. C. Bradley, C. G. Barraclough, J. Lewis, and I. M.
Thomas, J. Chem. Soc., 1961, 2601.

C. T. Lynch, K. S. Mazdiyasni, J. S. Smith, and W. J.
Crawford, Anal. Chem., §§'(12), 2332 (1964).

68

13.

14.
15.
16.
17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

69

R. S. P. Coutts, J. R. Surtees, J. M. Swan, and P. C.
Wales, Aust. J. Chem., 18, 1377 (1966).

B. Timm, Chem. Ind. (London), 1960, 274.

M. E. Vol'pin and V. B. Shur, Nature, 888, 1236 (1966),
H. Brintzinger, J. Am. Chem. Soc., 88, 4307 (1966).

H. Brintzinger, J. Am. Chem. Soc., 88, 6871 (1967).

E. E. van Tamelen and M. A. Schwartz, J. Am. Chem.
Soc., 81, 3277 (1965).

 

E. E. van Tamelen, G. Boche, and R. Greeley, J. Am.
Chem. Soc., 88, 1677 (1968).

E. E. van Tamelen, G. Boche, S. W. Ela, and R. B.
Fechter, J. Am. Chem.y§oc., 88, 5707 (1967).

E. E. van Tamelen and B. Akermark, J. Am. Chem. Soc.,
88, 4492 (1968).

E. E. van Tamelen and D. Seeley, J. Am. Chem. Soc.,
88, 5194 (1969),

E. E. van Tamelen, R. B. Fechter, S. W. Schneller, G.
Boche, R. H. Greeley, and B. Akermark, J. Am. Chem.
Soc., 8;, 1551 (1969).

E. E. van Tamelen, D. Seeley, S. Schneller, H. Rudler,
and W. Cretney, J. Am. Chem. Soc., 88, 5251 (1970).

G. Henrici-Olivé and S. Olivé, Angew. Chem., 81, 679
(1969).

A. D. Allen and C. V. Senoff, Chem. Commun., 621 (1965).

J. P. Collman, M. Kubota, F. D. Vastine, J. Y. Sun,
and J. w. Kang, J. Am. Chem. Soc., 22.1 5430 (1968).

A. Sacco and M. Rossi, gporganica Chimica Acta, 8,
127 (1968).

H. Scheidegger, J. N. Armor, and H. Taube, J. Am.
Chem. Soc., 88, 3263 (1968).

T. P. Whaley in "Inorganic Syntheses," Vol. V, T.
Moeller, Ed., McGraw—Hill Book Company, New York, N.Y.,
1957, pp 6r10.

D. Paul Rillema, Ph.D. Thesis, Michigan State University,
East Lansing, Michigan, 1969.

32.

33.

34.

35.

36.

37.

38.

39.

70

R. Pribil and M. Malat in "Gravimetric Analysis,"
Part II, L. Erdey, Ed., Pergamon Press, New York, N.Y.,
1965, p 540.

G. Wilkinson and J. M. Birmingham, J. Am. Chem. Soc.,
Z8, 4281 (1954).

 

J. C. Brantley, U. S. Patent 2,921,948 (1960); Chem.
Abstr., g2, 11048i. . '

 

F. A. Cotton and G. Wilkinson, Z. Naturforsch, 88, 417
(1954).

M. L. H. Green, J. A. McCleverty, L. Pratt, and G.
Wilkinson, J. Chem. Soc., 280 (1951).

M. H. Green, University of Oxford, private communica-
tion, 1969.

R. S. P. Coutts, J. R. Surtees, J. M. Swan, and P. C.
Wales, Aust. J. Chem., 18, 1377 (1966).

D. R. Gray and C. H. Brubaker, Jr., Chem. Commun.,
1239 (1969).

3’ r. .21.... Jul lu

    

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

lHllmlfllLIIWWIH

06